J Mari Arch DOI 10.1007/s11457-016-9167-2 ORIGINAL PAPER
Viking-Age Sails: Form and Proportion Vibeke Bischoff1
Ó Springer Science+Business Media New York 2017
Abstract Archaeological ship-finds have shed much light on the design and construction of vessels from the Viking Age. However, the exact proportions of their sails remain unknown due to the lack of fully preserved sails, or other definite indicators of their proportions. Key Viking-Age ship-finds from Scandinavia—the Oseberg Ship, the Gokstad Ship and Skuldelev 3—have all revealed traces of rigging. In all three finds, the keelson—with the mast position—is preserved, together with fastenings for the sheets and the tack, indicating the breadth of the sail. The sail area can then be estimated based on practical experience of how large a sail the specific ship can carry, in conjunction with hull form and displacement. This article presents reconstructions of the form and dimensions of rigging and sail based on the archaeological finds, evidence from iconographic and written sources, and ethnographic parallels with traditional Nordic boats. When these sources are analysed, not only do the similarities become apparent, but so too does the relative disparity between the archaeological record and the other sources. Preferential selection in terms of which source is given the greatest merit is therefore required, as it is not possible to afford them all equal value. Keywords Sail and rigging Viking ships Sails in the Viking Age Square sail
Introduction The image of a Viking ship under full sail is an iconic one and has become a symbol for the Viking Age as a whole. Despite this, the precise form and dimensions of Viking ships’ sails are not known. No fully preserved sail, or definite indicators of their proportions have yet been found from this period. Archaeological ship-finds and elements of the rigging constitute the primary sources for understanding and reconstructing Viking-Age sails and rigging. Several ship-finds & Vibeke Bischoff
[email protected] 1
The Viking Ship Museum, 4000 Roskilde, Denmark
123
J Mari Arch
throughout Scandinavia have survived with traces of their rigging, and it is these traces that provide the framework for the form of the reconstructed sail. However, the mast and yard of the various ship-finds have not survived to an extent that would allow for the reconstruction of an entire rig based on the archaeological material alone. Therefore, the iconographic, written and ethnographic material must also be included as additional and important sources in this process. In this way, the many open questions can be clarified and a more realistic interpretation of Viking-Age sails and rigging put forward. Problems arise, however, when the use of the various sources highlights not only their common features, but also the great contrasts that can be seen between the ship-finds and the other sources. This demands that careful attention is given to which source is used as inspiration or argumentation in every single case, while also forcing us to look at both ship-finds and the other sources from a wider perspective. This article examines traces of rigging found in three central finds from the Viking Age—the Oseberg Ship, the Gokstad Ship and the Skuldelev 3 find—in order to explore the overall proportions of sails (Fig. 1). A common attribute of all three ship-finds is that the mast position, the sheet area and tacking area—which combine to define the breadth of the sail—have been preserved. They can therefore be seen as key finds in terms of our understanding of the composition of Viking-Age rigging (Fig. 2). These three finds do not represent the class of long, narrow warships. Of this type, only the Skuldelev 5 find had traces of the mast and holes in the uppermost strakes in the aft surviving. The holes were potentially used to secure clamps for the sheets, but this cannot be definitively confirmed. If this interpretation is correct, then the Skuldelev 5 find had a sail with the same dimensions as the Oseberg Ship, the Gokstad Ship and the Skuldelev 3 find (Andersen and
Fig. 1 Locations of ship-finds and other sites mentioned in the text. Image: Viking Ship Museum, Roskilde
123
J Mari Arch
Fig. 2 Reconstruction of the Skuldelev 1 find with rigging details named in old Norse and modern English. Drawing: The Viking Ship Museum in Roskilde (Vinner 1995, 295)
Andersen 1989, 212–217). Other Viking-Age finds, such as the Ladby Ship, the Hedeby 1, Skuldelev 2 and Roskilde 6 finds, have, apart from the position of the mast, no surviving traces of the rigging (Bischoff 2003, 71–80; Bischoff and Jensen 2001, 224–233; Crumlin-Pedersen 1997, 88–89; Crumlin-Pedersen 2012, 174–175; Gøthche 2006).
Ship-Finds with Evidence for Rigging The Oseberg Ship was built in AD 820 and placed in a burial mound near Tønsberg, in present-day Norway, in 834. It was excavated in 1904 and is today on display at the Viking Ship Museum in Oslo (Bonde 1994, 138–141; 1997, 195–200). The 21.5 m-long and 5.1 m-wide ship is the most well-preserved and spectacular ship yet known from the Viking Age (Brøgger et al. 1917, 283–343; Bischoff 2012, 336) (Fig. 3). The keelson and mast partner are preserved, which mean that the mast’s exact position 1 m forward of midships is clear. Holes in the sides of the hull in the foreship indicate the position of the foremost corner of the sail (the tack), while in the aft, the sheet area is welldefined with either clamps or the holes where clamps were originally fitted. The breadth of the sail extends diagonally from the foremost hole for the tack on the port side to just forward of the surviving sheet clamp to starboard in the aft. Knowing both
123
J Mari Arch
Fig. 3 The Oseberg Ship from AD 820, on display in the Viking Ship Museum, Oslo. Photo: Kulturhsitorisk museum, Oslo
the points where the sail is secured and the mast’s position, it’s possible to pinpoint the sail’s approximate width. The tack position and the sheet clamps to port and starboard are displaced by roughly 0.5 m in relation to each other. This can be due to the side rudder’s impact on the sail balance, or a possible distortion in the hull, which has been compensated for by staggering the sail’s position. As the displacement of the clamps from both starboard and port match each other, it is still possible to estimate the sail’s breadth from the forward tack position to the sheet clamp, with perhaps a 0.5 m margin of error, to ca. 9.5 m (Bischoff 2014, 24). A possible yard was found on board the ship in the burial mound. The piece is 12.47 m long, 19 cm in diameter in the middle and 12.7 cm and 14 cm in diameter at each end. The proportions match with the proportions of the yards from the nearby and slightly later Gokstad find. The yard from Oseberg is not complete however, so its absolute length is unknown (Andersen and Andersen 1989, 246; KLNM 1980–1982). The mast’s height, and therefore the sail area, can be estimated on the basis of a mast crutch, or myke, which was found at the stern. The myke most likely served as a support for the mast when it was lowered. The 2.8 m-long myke has an opening of the same size and form as the opening in the mast partner. Here, the mast was 20 cm thick and had a flat forward edge. The mast partner is shaped with an open slot behind the mast, which indicates that the mast could be lowered towards aft—a method which is also described in the literature such as Sverri’s Saga, Haakon Haakonson’s saga recorded in the fourteenth century (Falk 1912/ 1995, 73; Andersen and Andersen 1989, 29; KLNM 1980–1982, Bind 4, 412; Bind 17,
123
J Mari Arch
552; Bind 6, 54). Mykes aft are known from other archaeological finds and from iconographic and ethnographic material (Andersen and Andersen 1989, 252–253; ˚ fjord boats McGrail 1993, 75–77). For example, on nineteenth century Nordland and A from the west coast of Norway, when fishing, the mast was laid in a myke in the top of the forestem (Eldjarn and Godal 1990, 242). The myke’s position in the stern on the Oseberg Ship provides a minimum possible height for the mast of ca. 11.5 m to the halyard hole. If it is accepted that the sail is adjusted to fit the mast, then the Oseberg Ship has a sail height of ca. 8.5 m. During excavation of the ship, several textile fragments of wool, 29 pieces in all, woven in 2/2 twill, stiff with red colour, were found. One section of textile, 11.5 9 16.5 cm of several layers of cloth, 1–3 cm thick, was tightly entangled with fragments of lime bast rope. It could have been part of a tent, but on the basis of the rope’s dimensions, it was concluded that the textile fragment was most likely part of the ship’s sail (Ingstad 2006, 199, 234–235, 244). A full-scale sailing reconstruction of the Oseberg Ship, Saga Oseberg, was built and launched in Tønsberg in Norway in 2012. It was built by Stiftelsen Nytt Osebergskip (The Foundation for a New Oseberg Ship) (www.osebergvikingskip.no). The volunteer group building the ship was assisted by 2–5 professional boat builders. The present author, in collaboration with Erik Andersen, reconstructed the rigging of Saga Oseberg with an 80 m2 right-angled wool sail, handwoven from Spelsau wool of the same quality as the fragment of textile that survived with the find (Fig. 4). The Gokstad Ship was built in AD 895 and placed in a burial mound at Sandefjord, in modern Norway, some 10 years later. It was excavated in 1880 and has since been placed on display in the Viking Ship Museum in Oslo. The ship is 23.2 m long and 5.2 m wide (Sjøvold 1985, 52–67; Bonde 1994, 140–142; 1997, 195–200) (Fig. 5). Three smaller boats were found within the ship.
Fig. 4 Saga Oseberg rigged with a right-angled square sail of wool, 80 m2 in size. Photo: Jørgen Kirsebom
123
J Mari Arch
The lower part of the Gokstad Ship’s mast stood in position in the keelson and mast partner when excavated. In the aft, the sheet area is defined with three clamps in each side, used to secure the sheets and braces. In the foreship, the corner of the sail (the tack) was fastened to a tack spar, a so called beitiass. Beitiass fittings with two indentations facing obliquely forward were found at deck-height towards the sides of the ship in both starboard and port. They were fastened to the ship side and knees two frames forward of the mast and had served as steps for the lower part of the beitiass (Nicolaysen 1882, Pl. I–IV; Andersen and Andersen 1989, 250–251) (Fig. 6). Beitiass and beita are known from written sources including the kings’ sagas and the medieval Norwegian laws recorded in the thirteenth and fourteenth centuries (Falk 1912/1995, 76; Andersen and Andersen 1989, 191, Jesch 2001, 175; KLNM 1980–1982, Vol. 4, 412, 656; Vol. 5, 560; Vol. 11, 232). The use of beitiass is also known from iconographic sources as on the Gotland picture stone from Klinte Hunninge, and from ethnographic boats from northwestern Iceland (Imer 2004, 58; Andersen and Andersen 1989, 192–193). Fittings for beitiass were also found on the eleventh-century Skuldelev 1 ship-find (Crumlin-Pedersen 2002, 119). Use of the beitiass reduces strain on the side of a ship, by spreading the pressure from the sail over to the opposite side, where the forces can be spread over several frames. The beitiass itself was not found at Gokstad and the exact position of the tack can therefore not be defined. The angle of the indentations at the fitting suggests, however, that a beitiass that extends athwart the ship would cross the side of the ship roughly to the aft of the forward yard crutch. The sail’s approximate width can be determined because the mast position, when taken together with the sheet clamp position in the aft, gives half of the
Fig. 5 The Gokstad Ship from AD 894, on display in the Viking Ship Museum, Oslo. Photo: Kulturhistorisk museum, Oslo
123
J Mari Arch
Fig. 6 Fitting for a beitiass on the Gokstad Ship. Photo: Kulturhistorisk museum, Oslo
width of the sail, which is also equivalent to the point where the beitiass crosses over the ship’s side. A section of the Gokstad Ship’s yard was found during the excavation. The surviving piece is 6.2 m long, 21–22 cm thick at its broken point and 14 cm thick at the end (Nicolaysen 1882, 44). The yard has the same proportions (length and thickness) as the preserved yard from the largest of the three Gokstad Boats and the yard-piece that lay under the Oseberg Ship (Andersen and Andersen 1989, 246–251). A further indication of the Gokstad Ships’s sail width is the position of the three surviving supports to hold the yard when the sail is not in use (Fig. 7). The middle support stands beside the mast and the two outer supports stand in the middle of the seventh room, forward and after the mast, respectively, at a distance of 10.75 m from one another (Andersen and Andersen 1989, 250). The 3 m-high supports stand in a foundation on the keel and go through a reinforcement at deck-height. The position of the supports indicates that shortest possible length of the yard, and with it, the width of the sail, because the yard was most likely supported under the ends. The relatively large supports and their reinforcement must suggest that the yard and sail had a considerable combined weight. The mast has solid dimensions, and is 33 cm in diameter at the opening in the mast partner (Andersen and Andersen 1989, 251). The ship must therefore have had a relatively large rig. There are no indications of the original height of the mast from the ship-find. The rig has been reconstructed by Jon Godal on the basis of the preserved elements of the rig in combination within Norwegian wooden shipbuilding tradition’s rules of measurement. A full-scale reconstruction of the Gokstad Ship, Gaia, was built in Bjørkedalen and launched in 1990. It is rigged on the basis of Jon Godal’s calculations with a mast 14.2 m high to the halyard hole and with a 130 m2 square sail (Andersen and Andersen 1989, 250). The sail is 11.8 m in height and trapezoidal in shape, with a width of 10 m at the top and 12 m at the
123
J Mari Arch
Fig. 7 The yard crutch on the Gokstad ship set an indirect limit for the sail’s breadth. Photo: Kulturhistorisk museum, Oslo
base. Today Gaia is sailed by the Gaia boat guild in the city of Sandefjord in Vestfold, Norway. The Skuldelev 3 find was built around AD 1040 and was scuttled as part of a barrier on Roskilde Fjord, just outside Skuldelev, in present-day Denmark during the second half of the eleventh century. The find was excavated in 1962 and exhibited at the Viking Ship Museum in Roskilde. The ship is 14 m long and 3.3 m wide (Andersen et al. 1997, 7; Crumlin-Pedersen 2002, 195–243) (Fig. 8). On the Skuldelev 3 find, the position of the mast can be stated thanks to the keelson, which has survived. In the foreship, there are three holes in each side of the ship, which were used to secure the tack. As there is no sign of wear on these holes, it must be assumed that the tack was secured using belaying pins. In the aft, there is a triangular hole with clear wear marks. This is interpreted as the hole for the insertion of the sheet (Andersen and Andersen 1989, 147). By knowing these outer points for the fastening of the sail’s corners, its approximate width can be deduced (Fig. 9). As a further indicator of the sail’s width, two elongated holes through the crossbeams, or biti, respectively, three frames in front of and aft of the mast are interpreted as supports for the yard when the sail is lowered. These could have been of the same type as on the Gokstad Ship, although standing at biti height and not on the keel.
123
J Mari Arch
Fig. 8 The Skuldelev 3 find, AD 1040, on display in the Viking Ship Museum in Roskilde. Photo: Viking Ship Museum, Roskilde, photographer Werner Karrasch
When reconstructing the sail and rig on the Skuldelev 3 find, the details in the rigging and the boat handling in the Norwegian square-sail tradition was taken as inspiration. In terms of the hull, there are notable comparisons between the Norwegian traditional boats and Skuldelev 3, and the construction of the rigging, with the position of the mast, tack position and sheet position in the Norwegian boat tradition all bore similarities to the evidence seen on the hull of the Skuldelev 3 find (Andersen and Andersen 1989, 147–151) (Fig. 10). Many of the rig elements from the Viking-Age archaeological finds, such as blocks and parrels, also have parallels to the rigging of traditional Norwegian boats. These boats from Norway are still being sailed in the traditional way, making them a good reference for understanding how to approach the full-scale reconstruction of the Skuldelev 3 find.
123
J Mari Arch
Fig. 9 a In the foreship of the Skuldelev 3 find there are three round holes in each side of the ship, which were used to secure the tack. Photo: Erik Andersen, Viking Ship Museum, Roskilde; b The reconstruction of the sail and rigging of the Skuldelev 3 find is possible because both the keelson giving the mast’s position and the holes from the rigging are preserved. Drawing: Erik Andersen, Viking Ship Museum, Roskilde
Although the holes for the shrouds, stays and a clamp for the halyard all survived, there is no clear indicator of the mast’s height. It is therefore estimated based on an appraisal of how large a sail area the ship could carry under normal conditions and ordinary wind strengths. How large a sail a ship can carry is determined when close-hauled. Here, the pressure is greatest and the balance between the sail, hull and rudder is most acute. If the sail is too wide in relation to the hull, the ship will have a lee helm. If the ship is not in balance and
123
J Mari Arch
Fig. 10 The Nordlands boat, Rana, from Mo i Rana, Norway, 1890. It is an aid to understanding the traces of rigging on the Skuldelev 3 find. Photo: Viking Ship Museum, Roskilde, photographer Werner Karrasch
neutral on the course, the pressure on the rudder increases. The relatively slender side rudder is in itself an indication that the ship must have been in balance. In order to establish the sailing and handling capabilities of such Viking-Age ships, a full-scale reconstruction, Roar Ege, was built and launched in 1984, as part of an experimental archaeological project at the Viking Ship Museum in Roskilde, and sailed by a boat guild (Andersen et al. 1997, 8). The holes for the tack on the Skuldelev 3 find indicate that the ship was intended to sail close-hauled and therefore beat to windward. Test sailings have shown that the reconstruction is in balance with a right-angled sail, with the tack set in the foremost tack position. The reconstruction was rigged with a 45 m2 square sail of wool, which still works perfectly some 30 years later (Fig. 11).
123
J Mari Arch
Fig. 11 Roar Ege, the reconstruction of the Skuldelev 3 find, built and sailed as part of the Viking Ship Museum’s experimental maritime archaeological research. Photo: Viking Ship Museum, Roskilde, photographer Werner Karrasch
Sources for the Reconstruction of the Proportions and Shape of Sails Together with the written sources from Old Norse literature (The kings’ sagas and laws recorded in the thirteenth and fourteenth centuries), the iconographic sources from the Viking and Middle Ages (the Gotland picture stones, ships on coins the tenth century, the Bayeux Tapestry from the eleventh century, ship graffiti and frescoes) are important parameters in terms of our understanding of the design/development of the Viking-Age sail and rig in Scandinavia (Andersen and Andersen 1989, 25–50; Heide 2006, 165–172; Varenius 2006, 163; Kastholm 2009, 101; KLNM 1980–1982; KLNM 1980–1982, Bind 4, 232, 412). Some old Norse texts discuss tenth and eleventh century ships, but usually these
123
J Mari Arch
were first committed to writing several hundred years after the events they describe (Heide and Godal 2001, 129; Andersen and Andersen 1989, 157). Despite this, they contain significant details which can help to contextualise the solutions chosen when reconstructing sail and rigging (Planke 2006, 191). The so-called Gotland picture stones, which has been typographically dated to the eighth to tenth centuries, all show large ships seen from the side, with high stems, a large, right-angled and extremely broad square sail and a mast standing roughly in the middle of the ship which is approximately as high as the ship is long (Imer 2004, 47–111). There are very few representations of the slightly trapezoidal-shaped sails on the picture stones. In those examples that are found, the sail extends all the way out to the diagonal edge of the stones and is designed to fit the carving area (Andersen and Andersen 1989, 224–227; Kastholm 2009, 121). The details on the ship motifs demonstrate some parallels with the archaeological material: ships on the stones are depicted with a single, right-angled sail and the rigging is represented by ropes such as the centre sheet/sheet net (priar net), bowline, shrouds and stays. In addition, the shields can be seen along the sides, decoration can be seen on the stem tops and the side rudder has a clear elongated extension at the rear of its base (Fig. 12). The ship depictions are comparable to the archaeological finds from Oseberg from 820 and Gokstad from 895. With its large, sweeping stems, spiral-shaped stem decorations and room for shields in the shield rail along the side, the Oseberg Ship bears strong similarities to several of the ship depictions. Their details, dating and geographical location make the depictions a vital source for our understanding of the rigging from ship-finds. Ship motifs on the Hedeby coins, which date to the ninth century, depict more or less stylistic representations of ships with side rudders, high masts that are roughly as high as the ship is long and long yards or sails, which are predominantly lower than they are wide.
Fig. 12 The Gotland picture stones show ships with right-angled square sails, high masts, shields and side rudders with an elongated extension. From Hejnum Ridare, Sweden. After Lindqvist 1941, Fig. 79
123
J Mari Arch
Like the ships on the picture stones, the ship motifs on the coins extend all the way out to the edge of the coin and can be interpreted as an expression of horror vacui (CrumlinPedersen 1997, 172–176; Kastholm 2009, 121) (Fig. 13). The Bayeux Tapestry from the late eleventh century depicts William the Conqueror’s invasion of England in 1066. Here, eleventh century ships are shown with details that are recognisable from Viking-Age ship-finds (Wilson 2004, 4–44). The ships that are represented, most likely built by descendants of the Vikings who settled in Normandy in the
Fig. 13 Hedeby coins depict ships and sails, which, like those on the picture stones, extend all the way to the edge of the available surface. After Malmer 1966
123
J Mari Arch
tenth century, are shown with dragon head protomes, shields along their sides and a siderudder with an elongated extension at its base. The masts, which are only half as high as the ships are long, stand slightly forward of midships, braced with shrouds and stays and with a single, right-angled sail, suspended from a short yard. The Bayeux Tapestry is almost contemporary with the Skuldelev ship-finds, and the Skuldelev 2 find was built in Dublin, an area close to the events depicted in the Tapestry (Crumlin-Pedersen 2002, 239). The details of the ships on the Tapestry are therefore used as a significant source in connection with the reconstruction of rigging on the reconstructed Skuldelev ship-finds, built at the Viking Ship Museum in Roskilde. The Tapestry depicts King Harold II’s ship with a right-angled square sail which is just as high as it is broad. The sail has many lines, or ropes, in the sail—perhaps a form for buntlines or clew lines. Descriptions of the buntlines, heflaskurd, to raise the sail up to the yard, reefing lines, sviptinger, to reduce the sail area, and handrif, to reduce the pressure of the wind on the sail, are all known from the written sources as in Konungs skuggsia´ (the Kings Mirror) from the thirteenth century and the Law of the Frostating (Falk/Varenius 1912/1995, 82–83; Andersen and Andersen 1989, 350; Heide and Godal 2001, 129–145; Heide 2006, 169; Jesch 2001, 163; KLNM 1980–1982, Bind 4, 656). On the Tapestry, the crew haul in the lines as the ships reach the coast and the sails are therefore reduced. This gives rise to the triangular sail form that is seen at several places on the Tapestry (Vinner 2009, 114–120) (Fig. 14). Ship graffiti, several examples of which are not securely dated, depict a more varied version of vessels with relatively lower masts and broad but short sails. The ship depictions from Karlby in Denmark, typologically dated to the early Viking Age (eighth century), La Grande Paoisse from France, dated to the eleventh century, and the ‘Byrding’ from Birka in Sweden, dated to the eleventh to twelveth centuruies show narrow sails, while two depictions from Gaudalen in Norway, dated to the eleventh century, depict ships with very long yards (Rieck and Crumlin-Pedersen 1988, 129; Kastholm 2011, 178; Snædal 1989, 99–104) (Fig. 15). Later frescoes are more detailed and drawn with perspective, as for example the leidang ship (military ship) in Skamstrup Church, from the western part of the island of Zealand, Denmark, which is dated to the fourteenth century. It shows a rightangled square sail, which is roughly as high as it is wide, with clearly defined panels and reefing lines (Fig. 16).
Proportions of Sail and Rigging in the Iconographic Material: A Comparison with the Archaeological Ship-Finds The majority of depictions on picture stones and coins from the Viking Age depict ships with right-angled and rather broad sails, which extend from stem to stem. The sails are half as high as they are broad (Planke 2006, 192–193). The masts, which are just as high as the ships are long, are situated roughly amidships (Kastholm 2007, 8–12; Kastholm 2011, 177–179). Here, the iconographic sources deviate from the archaeological ship-finds, where the masts clearly stand forward of amidships. The broad sails do not correlate to the ship-finds which have traces of rigging preserved. This has given rise to much debate about the degree to which ship motifs can be taken at face value, or, if it is in fact the reconstructed Viking ships that have been based upon a flawed interpretation of the material (Kastholm
123
J Mari Arch
Fig. 14 The sails on the Bayeux Tapestry’s ships have right-angled square sails, with buntlines. By special permission of the City of Bayeux
2009, 144–147; Planke 2006, 187–194; Varenius 2006, 155–163; Crumlin-Pedersen 2012, 163–172; Kastholm 2011, 180–182). Low and very wide sails have been tried on several reconstructions of Viking ships throughout the years. Practically, it gave rise to a number of challenges in handling such a long yard, while beating to windward was also difficult. The broad sail upsets the ship’s trim and balance because the centre of the aerodynamic forces on the broad sail extends
123
J Mari Arch
Fig. 15 Grafitti. a Karlby, Denmark, eighth century. Viking Ship Museum in Roskilde. b La Grande Paoisse, Seine Valley, eleventh century. After Ridel 2002. c Visby, Gotland eleventh to twelveth centuries. After Snædal 1989. d Gaudalen, Norway, eleventh centuries. After Christensen 1988
much further forward than the lateral centre of the hull. This causes the ship to have a lee helm which stresses the rudder and increases leeway (Andersen and Andersen 1989, 54–60; Planke 2006, 187–202). These experiments with the low and broad sails, were,
123
J Mari Arch
Fig. 16 In the fresco from Skamstrup Church, dated to the fourteenth century, the outermost panels are parallel with the sail’s edges from the bottom edge to the yard, underlining that the sail isn’t trapezoidal. Photo: Viking Ship Museum in Roskilde, photographer Werner Karrasch
however, solely based the reconstruction on the height/breadth proportions on the picture stones. The sail’s relatively extensive breadth in relation to the ship, has only been done once (Planke 2006, 192–193). The other proportions, such as the mast’s unrealistic height, have not been included in the experiments. The fact that the ships on the picture stones have an extremely high rig and a very large sail are significant details: these are not ships with a small, low sail, such as those that have been conducted in the sea trials. The question is, therefore, can the archaeological experiments be accepted, which have only taken some of the proportions at face value and chosen to ignore others, such as the height of the mast? The largest of the three small boats from the Gokstad find is one of the reconstructions which has been rigged with a broad sail comparable to those depicted on the picture stones. However, two partially preserved yards and a larger yard, the full length of which was preserved, were found together with the three boats. The largest yard, which presumably belonged to the largest of the three boats—known as the large Gokstad Boat—negates the
123
J Mari Arch
possibility of a reconstruction with a broad sail. Rather, the proportions of the sail bear much closer relation to the three key finds named previously (Andersen and Andersen 1989, 231). As stated previously, in terms of both type and chronology, the Oseberg Ship corresponds neatly with the Gotland picture stones. If the Oseberg Ship were reconstructed with a sail and mast of the proportions suggested by the picture stones, the preserved traces of rigging that were found on the hull would be made redundant. The sail would be 21 m long and 10.5 m high with a sail area of 230 m2. The stems—which are already relatively large and heavy—would be three times as high and the mast would be 21 m high. All of this is physically impossible and demonstrates that the ship motifs cannot be taken at face value. If a ship is seen from the side, the ship will appear large while the sail will appear relatively small. Seen head-on, the opposite is the case and the sail will appear large and the ship itself small. So if the objective of the iconography has been to depict a large ship with a large sail, there is a chance that the proportions of the ship and sail on the Gotland picture stones have been exaggerated, so that both ship and sail look as big as possible (Fig. 17). Part of the debate around Viking ship rigging has focused on the fact that it is only the sails’ proportions as depicted on the picture stones that can be taken as fact (Kastholm 2009, 121). If such a sail, with a height/breadth ratio of 1:2 should be fitted to the preserved tack position and sheet position, the sail on the Oseberg Ship would measure ca. 9.5 9 4.5 m = 40 m2. With such a small sail area, the ship would have been under-rigged and lack propulsion under normal wind conditions. This raises some concerns in relation to how the sail on the picture stones can be combined with the evidence from, in this instance, the Oseberg Ship, without denigrating either the find or the carving.
Trapezoidal or Right-Angled Sails? The collective iconographic material from the Viking and Middle Ages depict ships with right-angled sails. In the fresco from Skamstrup Church, the outermost panels are parallel with the sail’s edges from the bottom edge to the yard, underlining that the sail isn’t trapezoidal.
Fig. 17 If the ship is seen from the side, the sail appears small. If it is seen head-on, the sail appears large in relation to the ship. Photo: Viking Ship Museum, Roskilde, photographer Werner Karrasch
123
J Mari Arch
All the way up to the seventeenth century, the iconographic material only depicts rightangled square sails in images of ships with one or two square sails. The trapezoidal sail appears in depictions from the sixteenth century, but only as the uppermost small sails at the top of the masts of ships with several masts. The lower sails are still right-angled (Bohlmann 2014, 113–119) (Fig. 18). It is uncertain exactly when the trapezoidal sail first came into use on square-sailed boats in Scandinavia, but judging from the iconographic material, it was probably sometime around the seventh century (Fig. 19). Within the more recent square-sail tradition, trapezoidal sails have been used on smaller boats in areas with squalls, such as Western Norway, the Faroe Islands and Iceland. The larger and more voluminous square-sailed boats, continued to carry square sails, more likely than not, in order to avoid missing sail area (Christensen 1979, 46, 78). They also have a weight and a size that enabled them to carry these sails, making them less vulnerable to gusts and squalls. ˚ land Islands in the Baltic Sea, where squalls are not a problem, right-angled On the A sails were used almost exclusively up until the nineteenth century (Andersen et al. 1997, 164–165). The advantage with a trapezoidal sail is that the yard is shorter, so the wind pressure on the upper part of the sail is reduced. This means less heeling with increasing winds or gusts. The boat is more stable, with a lower sail-centre and it can carry its sail for a longer period without having to reef (Andersen and Andersen 1989, 224).
Use of Ethnographic Material There are striking parallels between nineteenth century Norwegian square-rigged boats and the ship-finds from the Viking Age in terms of their hulls, the structure and handling of their rig, but not in relation to the shape of the sails. They are open, square-sailed boats, with the mast position slightly forward of midships, as seen on the majority of the archaeological finds (Eldjarn and Godal 1988, 47–67; Eldjarn, G. and J. Godal. 1990, 256–299). These boats have therefore been important sources in the interpretation of recovered ship-finds or the traces of sail and rig. The sails’ forward corner, the tack, is
Fig. 18 The trapezoidal sail first appeared around the sixteenth century, as topsail on ships with several masts. Coloured illustration: Rudolf van Deventer: Bericht von Pulver und Feuerwerken, 1585, handwriting dedicated to Frederik 2. Det Kongelige Bibliotek, NKS 101 folio
123
J Mari Arch
Fig. 19 a Right-angled and b trapezoidal sail types. Drawings: Vibeke Bischoff, Viking Ship Museum, Roskilde
secured using holes in the foreship when close-hauled and beating to windward or on a broad reach as seen on Skuldelev 3 and on the Oseberg Ship. This suggests that the archaeological finds, just like the ethnographic parallels, have sought balance between the rudder, hull and rig (Andersen and Andersen 1989, 199–210; Andersen et al. 1997, 147–151). Many of the rig elements from the Viking-Age archaeological finds also have ethnographic parallels. The same is true of the shape of the various blocks, shroud pins and parrels, found in the Oseberg and Gokstad Ships and also in the harbour areas in Hedeby and Bergen (Crumlin-Pedersen 1997, 132–139; Christensen 1985, 132–148). In the finds catalogue and the publication concerning the Oseberg Ship, the function of the parrel was not specified, as it was neither set fast around the mast nor tied to the yard
123
J Mari Arch
(Brøgger et al. 1917, 321). The same applies to the shroud pin from the Gokstad Ship, which was not assessed as being a part of the rigging (Nicolaysen 1882, IV). Examination of ethnographic parallels would have revealed similar parrels and shroud pins still in use on the open, square-sailed Norwegian fishing boats. The material is so identical that it is hard to comprehend that there is 800–1000 years of difference. It therefore makes sense to use the Nordic square-sail tradition to aid in the interpretation and understanding of the make-up of the rigging on Viking-Age ships.
Conclusion This article is not intended as an argument in favour of the use of one source over another. All sources, the archaeological, written, iconographic and ethnographic, have their merits and disadvantages, regardless of the specifics of their potential dating, time of commitment to record or other open questions. The sources are a testimony from a time about which we have many blatant gaps in our knowledge concerning maritime technology. The number of archaeological ship-finds with preserved evidence for rigging is few in relation to the large amount of surviving ship depictions. With any luck, new ship-finds will be found, which can contribute to a more nuanced understanding. In terms of the reconstruction of archaeological ship-finds, it should however be stressed that it is the finds themselves and the related artefacts that constitute the primary source material for research into the composition of the rigging. The archaeological artefact and traces from rigging is a direct greeting from the past, that is sometimes however best understood or explained in interplay with different evidence sources. Combined with the potential validity shown through the experimental work it is possible to arrive at a measured conclusion, because as noted a sail must actually work. Ship depictions from the Viking Age show long yards that extend from stem to stem, and/or sails which are predominantly half as high as they are broad, with a very high mast and rig and a voluminous sail area. This simply does not marry with the surviving evidence from the archaeological ship-finds. The iconographic material shows that Viking-Age sails were right-angled and not trapezoidal, as those found on boats from the North Atlantic area in more recent times. The written sources give account of ships with rope and reefs— sviptingr, handriff—and beitiass upon which the tack was fastened. These elements all bear witness to the size of the sail and the ability of the ships to beat to windward. Reconstructions of Viking ships which are rigged with a right-angled square sail which is roughly as high as it is broad, and whose breadth is defined in relation to the hull, such as the Oseberg Ship, the Gokstad Ship and the Skuldelev 3 find perform well under sail, both in relation to handling, speed and ability to beat to windward. This does not prove that the present interpretations are correct, but it is a fairly strong indicator. The fact that the Vikings sailed to England, Greenland, North America and the lands around the Mediterranean testifies that the ships must have performed well under sail and were able to both tack and beat to windward. No matter how large the crew is, it’s a long way over the North Sea or the North Atlantic by oar.
References Andersen B, Andersen E (1989) Ra˚sejlet—Dragens vinge. Roskilde Andersen E, Crumlin-Pedersen O, Vadstrup S, Vinner M (1997) Roar Ege—Skuldelev 3 skibet som arkæologisk eksperiment. Vikingeskibshallen i Roskilde
123
J Mari Arch Bischoff V (2003) The reconstruction of the ladby ship—using a working spline model and computer. In: Beltrame C (ed) Boats, ships and shipyards. Proceedings of the 9th international symposium on boat and ship archaeology Venice 2000. Oxford, pp 71–80 Bischoff V (2012) Reconstruction of the Oseberg ship: evaluation of the hull form. In: Gu¨nsenin N (ed) Between continents. Proceedings of the 12th symposium on boat and ship archaeology. Istanbul 2009, pp 337–342 Bischoff V (2014) Rekonstruktion og Reflektion. Kysten 2014, Nr. 2, pp 21–26 Bischoff V, Jensen K (2001) Ladby II. The ship. In: Sørensen AC (ed) Ladby. A Danish Ship-grave from the Viking Age. Ships and boats of the North 3. The Viking Ship Museum, Roskilde, pp 181–245 Bohlmann J (2014) Segel und ihre Herstellung im 17. Jahrhundert. Rekonstruiert am beispiel eines skandinavischen Lastbootes. Dissertationen fu¨r den grad philosophiae doctor. NTNU, Norwegische Technish-Naturwissenschaftliche Universita¨t Bonde N (1994) De Norske Vikingegraves alder. Et vellykket norsk-dansk forskningsprojekt. Nationalmuseets Arbejdsmark, København, pp 128–148 Bonde N (1997) Dendrochronological dating of the Viking age ship burials at Oseberg, Gokstad and Tune, Norway. In: Sinclair A, Slater E, Gowlett J (eds) Archaeological sciences 1995. Proceedings of a conference on the application of scientific techniques to the study of archaeology, Oxbow Monograph 64, pp 195–200 Brøgger AW, Falk HJ, Shetelig H (1917) Osebergfundet vol I. Kristiania Christensen AE (ed) (1979) Inshore craft of Norway. From a manuscript by Bernhard and Øystein Færøyvik. Oslo Christensen AE (1985) Boat finds from Bergen. The Bryggen papers, main series. Bergen, vol I, pp 47–278 Christensen AE (1988) Ship Grafitti and models. In: Wallace PF (ed) Miscellanea 1 Crumlin-Pedersen O (1997) Viking-age ships and ship-building in Hedeby/Haitabu and Schleswig. Ships and boats from the North, vol 2. Schleswig & Roskilde Crumlin-Pedersen O (2002) Description and analysis of the ship as found. In: Crumlin-Pedersen O, Olsen O (eds) The Skuldelev ships I. Ships and boats of the North 4.1, pp 97–301. The Viking ship museum, Roskilde Crumlin-Pedersen O (2012) Ikon eller realitet? Om ‘sandhedsværdien’ af vikingetidens skibsbilleder. Aarbøger for nordisk Oldkyndighed og Historie 2010. København, pp 163–172 ˚ fjordsba˚ten, vol 1. Lesja Eldjarn G, Godal J (1988) Ba˚ten i bruk. Nordlandsba˚ten og A ˚ fjordsba˚ten, vol 2. Rissa Eldjarn G, Godal J (1990) Nordlandsba˚ten og A ¨ versa¨tning av Bo Varenius. Ba˚tdokgruppen. 1995 Ska¨rhamn Falk H (1995/1912) Fornnordisk Sjo¨fart. O Gøthche M (2006) The Roskilde ships. In: Blue L, Hocker F, Englert A (eds) Connected to the sea. Proceedings of the 10th international symposium on boat and ship archaeology Roskilde ISBSA Roskilde 2003. Oxford, pp 252–258 Heide E (2006) La˚ge og breie segl, likevel. Vurdert særleg ut fra˚ segl-krympningsma˚tar, etymilogi og norrøne skriftlige kjelder. In: Arisholm T, Paascheog K, Wahl TL (eds) Klink og Seil—festskrift til Arne Emil Christensen. Oslo, pp 165–174 Heide E, Godal J (2001) Gammalnorsk handrif pa˚ skip-«Hand-rev» eller «rekkverk» Norsk Sjøfartsmuseum ˚ rbok 2000. Oslo, pp 128–145 A Imer LM (2004) Gotlandske billedsten—dateringen af Lindqvists gruppe C og D. Aarbøger for nordisk Oldkyndighed og Historie 2001, København Ingstad AS (2006) Brukstekstilene. In: Arne Emil Christensen og Margereta Nockert (eds) Osebergfunnet Bind IV Tekstilene. Oslo, s. 185–277 Jesch J (2001) Ships and men in the late viking age. The vocabulary of runic inscriptions and skaldic verse. Woodbridge Kastholm OT (2007) Viking age iconography and the square sail. In: Maritime archaeology newsletter from Denmark, no. 22, pp 8–12 Kastholm OT (2009) De gotlandske billedsten og rekonstruktionen af vikingeskibenes sejl. Aarbøger for Nordisk Oldkyndighed og Historie 2005. København, pp 99–159 Kastholm OT (2011) The rigging of the viking age warship. The Skuldelev find and the ship motifs. In: Boye L (ed) The iron age on Zealand. Status and perspectives. Nordiske Fortidsminder, Serie C, Copenhagen, vol 8, pp 175–183 KLNM (1980–1982) Kulturhistorisk Leksion for nordisk middelalder: fra vikingetid til reformationstid. In: Karker A (ed) 2. Oplag. København Malmer B (1966) Nordiska mynt fo¨re a˚r 1000. Acta Archaeologica Lundesia, Series in 8°, no 4 Lund, Bonn McGrail S (1993) Medieval boat and ship timbers from Dublin. Medieval Dublin excavations 1962–1981. Ser. B, vol 3. Dublin Nicolaysen N (1882) Langskibet fra Gokstad ved Sandefjord. Kristiania
123
J Mari Arch Planke T (2006) Lave brede Seil allikevel?—En diskusjon av paradigmer og tolkninger av kildegrunnlaget. In: Arisholm T, Paascheog K, Wahl TL (eds) Klink og Seil—festskrift til Arne Emil Christensen. Oslo, pp 187–204 Ridel E (2002) Bateaux de types scandinaves en Normandie (Xe-XIIIe sie`cle). In: Ridel E (ed) L’Heritage maritime des Vikings en Europe de l’ouest. Presses Universitaires de Caen, Caen, pp 299–333 Rieck F, Ole Crumlin-Pedersen (1988) Ba˚de fra Danmarks Oldtid. Vikingeskibshallen 1988 Sjøvold T (1985) Vikingeskipene i Oslo. Universitetets Oldsaksamling, Oslo Snædal T (1989) Rodiaud gjorde mig.—en kvinnlig runristare pa˚ Gotland. Gotla¨ndskt Arkiv. Visby, pp 99–104 Varenius B (2006) Gotla¨ndske bildstenar och kyrkoristningar som ka¨lle till fo¨rsta˚else av rig och ba˚ttyper 500–1500 e Kr. In: Arisholm T, Paascheog K, Wahl TL (eds) Klink og Seil—festskrift til Arne Emil Christensen. Oslo, pp 155–164 Vinner M (1995) A viking-ship off cape farewell 1984. In: Olsen O, Skamby Madsen J, Rieck F (eds) Shipshape. Essays for Ole Crumlin-Pedersen. Roskilde, pp 289–304 Vinner M (2009) Unter vollem Segeln. In: Wamers E (ed) Die Letzten Wikinger. Frankfurt am Main, pp 106–132 Wilson DM (2004) The Bayeux tapestry. Singapore
123
