Sources for the massive naval expeditions of 1066 reveal little about how Harald Hardradi, William the Conqueror, or Harald Godwinson managed to gather the great numbers of ships they required. Generally speaking, northern European fleets were assembled as needed on the basis of obligations owed one’s overlord. As far back as the first century, Tacitus referred to Germanic chiefs relying on warrior cohorts of a hundred men selected from each district under their rule, and it is possible that a similar practice applied to maritime communities in Scandinavia and the British Isles. According to the seventh-century Census [or History] of the Men of Alba, every twenty households in the kingdom of Dalriada (northern Ireland and southern Scotland) were required to supply two ships and twenty-eight crew when summoned—all told, 177 ships and 2,478 crew.
Apart from the ad hoc levies of the Census, British rulers seem to have been largely indifferent to naval affairs prior to the ninth century, and while Alfred the Great is frequently credited as the founder of the English navy, the only evidence for a fleet is a brief mention of his design for ships to be sailed against the Danes. The number of ships, where they were based, how they were administered, paid for, or manned—all these are unknown, as are details about the hundred-ship fleet deployed by Edward the Elder at Brunanburh. Later in the tenth century Edgar and Æthelred II evidently established, or continued, a system to finance the fleet by mandatory levies of ships and men, called the ship-soke, in which administrative divisions equivalent to three hundred households were required to furnish one ship and sixty men. Select crews of sailor-warriors were levied on the basis of one man per five households, each of which contributed 3½ shillings to his maintenance for two months. Æthelred also hired English and Danish mercenaries. The ship-soke had parallels in Norway, where Harald Fairhair instituted the first large-scale ship levy in the ninth century, when every three households had to provide one crewman and his maintenance for two and a half months. This evolved into a more sophisticated levy of ships, men, and provisions throughout Scandinavia called a leidang.
Under normal circumstances, it seems that almost anyone could be sent up for service with the fleet, for Olaf Tryggvason was obliged to set firm guidelines for the crew of the Ormr inn Langi (“long serpent”), his flagship at the battle of Svold: “No man younger than twenty years of age was to serve on this ship, and none older than sixty. No effeminate cowards or beggars were to come aboard, and hardly anyone was allowed aboard unless he was distinguished in some way.” Such strict guidelines probably did not apply in 1066, when Harald Hardradi gathered at least 250 ships for his invasion of England, and William’s fleet numbered between 700 and 3,000 ships (the sources do not agree) manned by 7,000 crew who ferried another 7,000 soldiers and knights together with their gear and horses.
In their brisk telegraphic style, the captions of the Bayeux Tapestry merely hint at the logistical complexities and organizational sophistication required to coordinate such a formidable undertaking: “Here William orders the ships to be built. Here they pull the ships to the sea. These men carry weapons down to the ships. And here they pull a cart loaded with weapons. Here William crosses in a large ship over the sea and comes to Pevensey. Here the horses disembark. And here the soldiers hurried to Hastings to requisition food.” Well aware of the Norman threat, Harald Godwinson had “gathered a greater ship-army and also land-army than any king in the land had ever gathered before,” but his experience highlights the drawbacks of the temporary levies. William delayed so long in putting to sea that Harald was forced to relax his watch on the coast because “the men’s provisions were gone, and no one could hold them there any longer. Then the men were allowed to go home, and the king rode inland, and the ships were sent to London.” Although disbanding the fleet freed Harald to deal with Harald Hardradi, it left his southern flank exposed.
Such a casual system for raising a fleet could work only so long as the weapons involved were what sailors would own anyway. Northern Europeans had no long-range weapons like ballista or catapults and the nature of their ship design precluded ramming, so ships sailed as transports and became platforms for hand-to-hand combat more by accident than design. Oddr Snorrason offers a lengthy depiction of Olaf Tryggvason’s last stand in the Ormr inn Langi at the battle of Svold (one of the only fleet engagements of the Viking age to have been described in any detail), where Olaf faced the scores of ships marshaled by his enemies with only four of his own. Olaf had his ships chained together, with the Ormr inn Langi in the middle because it was “much longer and higher in the gunwales than other ships. That made for a good battle stage as if it were a fort.” Against this Eirik Håkonsson had a ship called Járnbarðinn (“Ironprow”), which was “extensively reinforced with iron and sharp spikes” at bow and stern as protection against boarders, an unusual configuration for the time. Eirik’s victory is attributed to his embracing Christianity and removing the idol of Thor from the bow of his ship, and to his erecting a large tower on the Járnbarðinn, from which to drop heavy beams on Olaf’s ship. Advantageous conversions were common enough, but the latter tactic seems to have been an improvised stratagem for which Viking ship design was not well suited. Nonetheless, it appears to have worked in this case and Olaf and his eight surviving comrades ended the battle by throwing themselves into the sea. The Swedes and Danes had “posted small boats around the larger ships so that they could fish out those who dove overboard and bring them to the chieftains,” and all save Olaf were pulled from the water. Whether he drowned or escaped has been debated ever since.
Ships of Northwest Europe
The Viking ships present at the battle of Svold represent the culmination of a line of development in northern European shipbuilding the origins of which were hundreds if not thousands of years old. The ships of northern Europe were distinct from those of the Mediterranean or the Monsoon Seas in two key respects: the hulls tended to be built with strakes overlapping rather than laid flush and joined to one another by clenched bolts, a style called lapstrake or clinker; and until shortly before the start of the Viking age the sail was apparently unknown on the continent north of the Rhine. When sails were developed in or introduced to Gaul and the British Isles is unknown, but Pytheas’s claim that Ultima Thule lay six days from the Shetland Islands or Great Britain assumes the use of a sail by at least the fourth century BCE. The Shetlands are four hundred miles from where the Arctic Circle intersects the coasts of Norway and Iceland, so if Pytheas and his crew sailed from the Shetlands, they would have to have averaged seventy miles per day, about twice the speed possible under oars.
The next positive evidence for sails in the British Isles comes from the Broighter “boat,” a twenty-centimeter gold model of a boat from the north of Ireland dated to the first century BCE. Fitted with eighteen oars, a steering oar, and a mast and yard for a square sail, a life-size version of the model would yield a boat of between twelve and fifteen meters long. There is little reason to believe that mariners of Pytheas’s day, or even earlier, could not rig their hide boats with a mast and sail, and it seems likely that if he had fallen in with far-ranging mariners who did not use sails, he would have mentioned it. Whether sails were used throughout the British Isles or whether there was a line of demarcation between areas of ships with sails and those without them is hard to know. Regardless, the practice of building large, unrigged vessels continued as late as the seventh century.
The Broighter boat is contemporaneous with the oldest firsthand account of northern European ships, in Caesar’s Conquest of Gaul. During a naval campaign against the Veneti on the south coast of Brittany in 57 BCE, Caesar was especially impressed with their seamanship and the differences between their ships, designed to weather the rough Atlantic littoral, and those of the Mediterranean. “The Gauls’ own ships were built and rigged in a different manner from ours,” he wrote.
They were made with much flatter bottoms, to help them to ride shallow water caused by shoals or ebb-tides. Exceptionally high bows and sterns fitted them for
use in heavy seas and violent gales, and the hulls were made entirely of oak, to enable them to stand any amount of shocks and rough usage.… They used sails made of raw hides or thin leather, either because they had no flax and were ignorant of its use, or more probably because they thought that ordinary sails would not stand the violent storms and squalls of the Atlantic and were not suitable for such heavy vessels. In meeting them the only advantage our ships possessed was that they were faster and could be propelled by oars.
Caesar had ships built on the Loire, but smaller and fitted with rams, they were no match for those of the Veneti. “Perfectly equipped and ready for immediate action,” the Veneti ships rode so high in the water that when the Romans “tried erecting turrets they found that they were still overtopped by the foreigners’ lofty sterns and were too low to make their missiles carry properly, while the enemy’s fell with great force.” The Romans finally gained the advantage by cutting the halyards of the Veneti ships with “pointed hooks fixed into the ends of long poles.” Because the Veneti ships carried no oars, the Romans could then pick off the powerless ships one by one.
Caesar gained additional experience on his two crossings to Britain, and for the second he ordered the construction of ships suited to crossing the English Channel. As he writes:
To enable them to be loaded quickly and beached easily he had them made slightly lower than those which we generally use in the Mediterranean.… To enable them, however, to carry a heavy cargo, including a large number of animals, they were made somewhat wider than the ships we use in other waters. They were all to be of a type suitable for both sailing and rowing—an arrangement which was greatly facilitated by their low freeboard.
The largest collection of Roman-era vessels on the Rhine was discovered during the construction of a hotel in the 1980s, when workers uncovered the remains of five fourth-century vessels adjacent to what had been the Classis Germanica base at Mainz. Four were slender, open lusoriae, general-purpose transports and patrol boats measuring about 21 by 2.5 meters, with a single mast rigged for a square sail, and thirty oars. (The fifth and smallest vessel was an inspection boat with a small cabin for officials.) In building these, shipwrights fastened planks around a temporary frame, which was then removed before permanent frames were inserted into the complete hull, a process that would have facilitated the mass production of these river craft, which made up a growing proportion of the imperial fleet in the post-Augustan era.
This hybrid method of hull construction is similar to the “Romano-Celtic” design associated with vessels found in Roman Gaul and Britain. Described as a “frame-based” shipbuilding technique, this is known from the remains of the second-century Blackfriars barge excavated at London, and from the third-century sailing merchantmen found at St. Peter Port, on the island of Guernsey in the Channel Islands. Neither hull was built using the shell-first method found in the Mediterranean and elsewhere in northern Europe; nor were complete frames erected before planks were fastened to them. Instead, the frame was erected in stages: a section of framing was assembled and after planks were attached to this, the frames were extended beyond the last completed strake. In this way, the framing rather than the planking was the dominant element in determining the shape of the hull. Neither this style, nor that of the Mainz vessels, developed into a fully frame-first sequence, which was only introduced from the Mediterranean, where it evolved, in the later Middle Ages.
The Blackfriars, St. Peter Port, and Mainz vessels all carried sails, but although Roman patrol craft especially would have been well known to people living along the Rhine-Danube corridor, and in fact many of the crews were recruited from the indigenous population, there is little evidence of sails being used north of the Rhine until well after the end of Roman rule in Gaul. One of the best preserved vessels of the period is a fourth-century vessel found in the Nydam bog near Schleswig, about eighty kilometers north of Kiel. The twenty-two-meter-long hull has thole pins for thirty oars and was steered with a quarter rudder, but there is no evidence of either mast or rigging. The layout of the shell-first, lapstrake Nydam boat conforms to a first-century description of Germanic vessels by Tacitus: “The shape of their ships differs from the normal [that is, Roman] in having a prow at each end, so that they are always facing the right way to put into shore. They do not propel them with sails, nor do they fasten a row of oars to the sides. The rowlocks are movable, as one finds them on some river-craft, and can be reversed, as circumstances require, for rowing in either direction.”
The Nydam boat may represent the kind of vessel in which the Angles, Saxons, and Jutes crossed the North Sea to Britain in the fourth and fifth centuries. Although such a large-scale, long-distance migration under oars alone would have been challenging, large vessels powered solely by oars were still being used in England in the seventh century as we know from the remains of the Sutton Hoo ship. When excavated in 1939, the twenty-seven-meter-long clinker hull no longer existed, because in reacting with the soil its timbers and iron fastenings had dissolved, leaving clear casts, traces, and impressions of their form and placement, and even evidence of repairs. The ship was manned by twenty-eight rowers but although there is no evidence of a mast or sail, the hull shape and construction details like a keel suggest that it could have been rigged, and a half-scale model built in 1993 demonstrated excellent sailing ability. Regardless of whether the Sutton Hoo ship was rigged, it clearly belongs to the transitional phase in northwest European shipbuilding that led to the ships of the Viking age.
Written and visual depictions of Viking-era ships are no more detailed than Oddr Snorrason’s descriptions of the Ormr inn Langi and Járnbarðinn at the battle of Svold but the rich archaeological record makes up for this: more than twenty ships from the critical ninth and tenth centuries have been excavated across an area running from Oslofjord to the coast of Jutland and eastward to the Vistula. The Vikings had a variety of vessels for both warfare and general use, although merchantmen and warships had several features in common. Like the Nydam and Sutton Hoo vessels, the double-ended hulls were of shell-first, clinker construction and steered with a single quarter rudder, and relatively flat bottoms allowed them to be run up on shore. The incorporation of a keel made it possible to rig a mast and square sail; yet while sailing was useful on long passages, when maneuvering inshore or against fickle winds, or going into battle, rowing was imperative. Unlike warships, which were generally open-hulled throughout, knarrs and other merchantmen were decked forward and aft, where the rowers sat, with a break amidships in which were crowded the passengers and cargo, everything from food and tools to trade goods and livestock, including sheep, cattle, and horses, which were routinely carried by ship. The Norse introduced horses to Iceland, and for his invasion of England, William embarked two or three thousand knights who would have traveled with the same number of warhorses in ships of similar design.
Among the earliest and most aesthetically dramatic Viking-era ship finds are the Norwegian Oseberg ship (21.6 meters by 5.1 meters broad), built around 815–20 and excavated in 1904, and the Gokstad ship (23.3 meters by 5.2 meters), unearthed from a burial mound in 1880 and dated to about 890–95. Once considered archetypes of the longship (langskip), these are now believed to be karvi, a type smaller than either the longship or the merchant’s knarr. Both ships were rigged, and the Oseberg ship is the oldest northern Scandinavian ship for which there is indisputable evidence of a sail. As was typical of Viking ships, shields could be fastened to racks that ran above the single row of oarports on either side of the Oseberg and Gokstad hulls. That they were found in burial mounds suggests they probably belonged to a chieftain or other important personage, as did the Sutton Hoo ship.
Of slightly later date are five ships scuttled off Skuldelev to block the approaches to Roskilde, Denmark, during the wars between Harald Hardradi and Svein III. The fragmentary remains have been identified as belonging to two knarrs (Skuldelev 1 and 3), two warships (Skuldelev 2 and 5), and a fishing boat (Skuldelev 6), whi
ch date from between 930 and 1030.c The newest and best-preserved vessel, Skuldelev 3, was fourteen meters long and could carry about five tons with a crew of five to nine people. With a capacity of fifteen to twenty tons, the sixteen-meter Skuldelev 1 was probably built in Norway and is the sort of ship that would have been used for overseas trading. Skuldelev 5, “the small warship” (seventeen meters), and Skuldelev 2 are notable for their roughly 7-to-1 length-to-beam ratio, much narrower than the 4-to-1 ratio of the other ships. Vessels of comparable design seem to have been copied from the Baltic to Normandy and Ireland, where the thirty-meter-long Skuldelev 2 was built. The bottom planks of both warships have been worn thin from being run up on beaches. Skuldelev 2 is the longest Viking ship yet found and although the total number of oars is unknown, its complement has been estimated at between fifty and a hundred men. The Icelandic sagas indicate that ships with between thirteen and twenty-three pairs of oars were considered longships, and the Skuldelev 2 would have been at the bigger end of the spectrum.
The Gokstad ship of 895 is a large clinker-built karvi intended for ocean sailing. Built primarily of oak, the hull measures twenty-three meters by five meters and could carry about sixty-five people. The ship was found in a burial mound near Sandefjord, Norway, with three smaller boats, a bed, cooking implements, and twelve horses, six dogs, and other animals. Courtesy of the Vikingskipshuset, Oslo.
Those that carried more than twentyfive pairs were called “great ships,” the most celebrated of which is Olaf Tryggvason’s Ormr inn Langi. As described in Oddr’s twelfth-century Saga of Olaf Tryggvason, the massive ship was built near Trondheim and the slip where it was laid down was still visible in Oddr’s day: “seventy-four ells [thirty-six meters] long, not counting the raised portions at stem and stern.” Ships built for the king were sumptuously decorated, and Olaf “had the ship painted all sorts of colors, then had it gilded and adorned with silver. On the prow of the ship there was a dragon head.” The ships in Svein Forkbeard’s expedition to England were more lavish still. “On one side lions moulded in gold were seen on the ships, on the other side … dragons of various kinds poured fire from their nostrils. Here there were glittering men of solid gold or silver nearly comparable to live ones, there bulls with necks raised high and legs outstretched were fashioned leaping and roaring like live ones.” Such embellishment was intended to exalt the king and terrify the enemy. Of greater utility, though still ornamental, were wind vanes and on Svein’s ships carvings of “birds on the tops of the masts indicated by their movements the winds as they blew.”
The Sea and Civilization: A Maritime History of the World Page 36
