KING ØYSTEIN'S HARBOUR AT AGDENES; STATUS OF RESEARCH AND RE-ASSESSMENT OF PROBLEMS.

Por Marek Jasinski


Introduction


Maritime-related problems are, naturally, very important in Norwegian archaeology and have become noticeably more prominent in recent years thanks to a conscious policy on the part of the cultural heritage authorities. Nonetheless, this does not mean that the effort is adequate relative to the needs. Several important aspects of the maritime sphere are still awaiting more research (see e.g. Jasinski 1995).

One example is harbours, particularly those from prehistoric and medieval times. Efforts here have so far been concentrated in the Medieval towns, chiefly Bergen and Oslo, where waterfront studies are giving valuable results. Harbours outside these Medieval centres stand in a completely different position. These cultural monuments have not experienced any significant effort on the part of archaeologists, with one notable exception - King Øystein's harbour at Agdenes in Sør-Trøndelag. This article deals with the history of research, results and future strategies here.

The harbour at Agdenes holds quite a special status in central Norwegian cultural history. It is mentioned in several sagas, and as far as I know Agdenes is one of very few localities so far recorded in Norway where wooden constructions are preserved. A number of elements are present in three zones: on land, on the beach and on the sea floor. It consequently constitutes an excellent area for a maritime archaeological research and management project.

From a maritime viewpoint, the locality is situated in a key position where the Trondheim channel and Trondheimsfjorden meet. There is no doubt that it was precisely this location that played an essential role when this harbour was constructed in a small bay on the northern point of Agdenes.

Visible harbour structures here, on land and in the intertidal zone, have been mapped and described a number of times. In the 1960's and 1970's, some small-scale underwater mapping was also carried out here. Four articles have so far been published dealing with the harbour construction (Marstrander 1967, Sognnes 1985, Sognnes 1987, Jasinski 1993) and several unpublished reports also exist (in the Archaeological Department files at the Museum of Archaeology and Natural Science in Trondheim). The harbour is also mentioned, or briefly discussed in several other publications (Schøning 1910, Munch 1941, Christophersen 1991).

My basis for writing this article is the results of maritime archaeological investigations which I undertook on behalf of the Museum of Archaeology and Natural Science in Trondheim in 1992-93 (Jasinski 1993). These revealed some better data which place the results of previous studies in a new light and point the way towards new requirements and fresh directions for further research at this locality. Discoveries made on the sea floor at Agdenes also put a number of management questions on the agenda.


Written sources

Agdenes and the harbour in the bay at Agdenes are mentioned several times in Snorre and Håkon Håkonsson's saga (Sturluson 1984, Noregs Konge Soger 19?? ). The sources can be divided into two groups:

1. Those dating from the period before King Øystein's reign.
2. Those from King Øystein's reign and later, i.e. the 12th and 13th centuries.

Because these sources are quite important for the further interpretation of the localities, I will cite relevant fragments.

Group 1 The first group of sources concerns information from the sagas of Olav Trygvason and St. Olav (Sturluson 1984).

Olav Trygvason's Saga, Chapt. 47
"Olav now sailed east and reached the coast near Moster. It was there he first landed in Norway and he had mass sung there in a tent. A church was afterwards built on the same spot. Thore Klakka advised the king not to make it known who he was or let any news of him go ahead; he should travel as quickly as he could to the jarl and come upon him suddenly. King Olav did so. He went north by night and day as the wind permitted and let people on land know nothing of the journey he was taking. And when he came north to Agdenes, he learned that Håkon the Jarl was in the fjord and also that he was not at one with the farmers."

St. Olav's Saga, Chapt. 138
"Ragnhild sent men south to Jæren to her father Erling and asked him to send her help. From there came Erling's sons, Sigurd and Tore, and each of them had a twenty-bencher and on each were ninety men. And when they came north to Torberg he greeted them well and with great delight. Torberg made himself ready for the journey and he, too, had a twenty-bencher. They went north, and when they came to Trondheim's Minne (supposedly south of Storfosna), Finn and Arne, Torberg's brothers, lay ready there with two twenty-benchers. Torberg greeted his brothers with friendliness and said that the whetting had helped; Finn said that there was seldom need of that with him. They then went with all their army north to Trondheimen and Stein was with them on the journey. And when they came to Agdenes Kalv Arneson lay by there, and he had a well-fitted-out twenty-bencher. With this army they went into Nidarholm and lay there during the night."

St. Olav's Saga, Chapt. 139
"When Finn Arneson had stayed with King Olav for a short time, the King one day called him and several other men whom he was in the habit of having in his councils to talk with him. The King began thus: 'I have now decided on this plan. This spring I will call out a force from the whole land, both of men and ships, and then go with all whom I can muster against Canute the Mighty, for I know of the claim on the kingdom which he has made against me, and he does not intend to let it be only idle talk. Now I will tell you, Finn Arneson, that I wish you to go as my messenger north to Hålogaland, and hold there a levy and call out a force of men and ships, and lead the army to meet me in Agdenes.' The king then named the other men and some he sent into Trondheimen and some south in the land, so that his bidding went through the whole land."

In both sagas, Agdenes at the approach to Trondheimsfjorden is mentioned in connection with warfaring. The harbour is not mentioned, but we may assume that Agdenes was an important meeting place and assembly point for warships. It is impossible to say just where at Agdenes these gatherings took place - whether it was precisely in the bay that is now known as King Øystein's harbour. On the other hand, there is no doubt that it was this geographical location at the approach to the heart of Trøndelag which occasioned these military events.

Group 2 The second group of sources consists of fragments from the saga about Magnus's sons, the Magnus Erlingsson Saga and the Håkon Håkonsson Saga (Snorre's Kongesagaer 1979).

In the first of these sagas, where King Øystein and his brother Sigurd are comparing their achievements, we find the first written information about the harbour at Agdenes which Øystein had built during his reign (1103-1123).

Magnus's sons Saga, Chapt. 14
"King Øystein had done much good in the country while King Sigurd was on his journey. He founded Munkeliv at Nordnes in Bergen and placed large estates under it. He had St. Michael's Church, a splendid stone church, built, and likewise the Apostle's Church on the royal farm, built of wood. And at the same place he had a large hall built, the finest house of wood built in Norway. He also had a church built at Agdenes, and near it a rampart and a harbour where no harbour had existed before. He also had the church of St. Nicholas built on the royal farm at Nidaros, and that building was extremely caringly built with carvings and other decorations. He also had a church built at Vågan in Hålogaland, and gave it income."

Magnus's sons Saga, Chapt. 21
"King Sigurd said: 'You've doubtless heard that I had many battles in Serkland and won all of them, and gained many kinds of treasures, the like of which has not been seen before in this land. I think I am most worthy when I am with the finest men, and I think you are still a stay-at-home.' King Øystein said: 'I've heard that you had some battles abroad, but it was more useful for our land that I, during this time, built five new churches, and made a harbour at Agdenes where there was no harbour before, and where all must pass who will travel north or south along the land. I also built the tower at Senholmsund and the hall in Bergen, while you were slaughtering Arabs in Serkland and sending them to hell, which I think was of little benefit for our kingdom.'"

The next item of information we find about Agdenes is in Magnus Erlingsson's Saga, in the account of Erling Skakke and his problems with the people from Trøndelag in 1165.

Magnus Erlingsson's Saga, Chapt. 25
"Erling was in Bergen, but when spring came his men told him of the rumour they had from some men on a cargo boat that had come from Trondheim in the north, that the people there were showing him open emnity, and were vowing at their things that if Erling came to Trondheim, he would never pass Agdenes alive. ..... On Tuesday during the procession days, Erling called his men to the ships. .... On Wednesday, in the evening, they sailed past Agdenes; there they met upon a large fleet of boats sailing together, both cargo boats, rowing boats and small sailing boats. These were people who were going to town for the Holy Day (Ascension Day, which in 1165 was on 13th May). Some were ahead of them, some behind. Therefore, the townsmen did not notice the approach of longships."

The final item of information on Agdenes is in Håkon Håkonsson's Saga (Noregs Konge soger 4, 1979). Chapter 333 mentions that a rampart was built during the reign of King Håkon (1217-1263), probably after his coronation, i.e. between 1247 and 1263.

Håkon Håkonsson's Saga, Chapt. 333
"King Håkon attached greater importance to strengthening God's Christianity in Norway than any other king before him, since the holy King Olav. He had a church built in the north in Troms and converted the entire population of the parish. Many people from Bjarmaland came to him, who had fled from the battles with the Tartars in the east, and he had a church built in Ofoten and a rampart at Agdenes."

The information from this group of sources is of great significance since it provides specific information about the establishment of the harbour at Agdenes and events from subsequent periods. However, there are interpretation problems since different publications give different translations of the saga texts. According to Snorre's Kongesagaer (1979), King Øystein built a church, a rampart and a harbour at Agdenes "where there was no harbour before, and where all must pass who will travel north or south along the land." Marstrander (1967: 583) pointed out that the term skansevoll (rampart) used in the saga translation implied that the harbour had a military purpose, whereas he thought this interpretation of the saga word virki was wrong since nothing suggests that the harbour had any military significance.

When we compare the text from Magnus's sons saga given in Snorre's Sagas (1979) with that quoted by Helland (1898: 53) we find still greater differences. According to the latter, King Øystein, in his trial of strength with his brother, says:

"at Agdenes there was a dangerous passage, and ships often stranded. I have constructed an excellent fairway for ships, whether they are travelling north or south, and likewise built a church there and set up beacons on the hills".

This version of King Øystein's statement points more towards a non-military function for the harbour. Unfortunately, Helland gives no source for his version of the text.

There are also problems with the text of Håkon Håkonsson's Saga. In Noregs Konge soger (1979) we are told that King Håkon only built a rampart at Agdenes. Sognnes (1985: 61), on the other hand, writes that he "built bulwarks and restored the harbour".

I have not studied the original texts of the sagas, nor have I the linguistic knowledge to do so. It is also necessary to pose the question of whether the sagas are of any value as historical sources. Consequently, my use of the fragments cited will be limited to the more obvious aspects. Otherwise, I will primarily rely upon the archaeological material.


The first field investigations

The investigation of King Øystein's harbour has a history lasting more than 220 years. This was in fact one of the first Medieval monuments to be investigated in Norway (Sognnes 1985: 61). As early as 1773, Gerhard Schøning tried to gather information about the harbour when he spent a night at Agdenes. He wrote (Schøning 1910: 79):

"We chose to spend the night here, and I thought I would take the opportunity of asking someone about the reported harbour and church, or find traces of them, but to no avail. There is no harbour here now, not for ships at least; the only object for their use is a large iron ring inserted in the rock on the outermost headland not long ago."

Remains of the harbour at Agdenes were, nonetheless, found 92 years later. In 1865, the chairman of the Trondheim section of the Society for the Preservation of Norwegian Ancient Monuments, O. Krefting, reported that he had observed the remains of harbour constructions and church foundations in a small bay. His observation was confirmed by a Lieutenant Sjerstad (Nicolaysen 1868). At the request of the society, the locality was mapped and described in 1869 by an engineer, J. Meyer. His unusually precise drawings, along with his description of the locality, now form a very valuable source of information for archaeologists (Meyer 1869).
Meyer's plan comprises three principal elements: 1. the remains of a breakwater on the beach, 2. a rampart on the first marine terrace, and 3. the location of a possible church.

1. The breakwater

When Meyer undertook his work in 1869 he was able to trace a construction of stones and timbers for about 36 m across the beach at low water. Its width was about 4.7 m. Meyer was able to distinguish three layers in the construction. The deepest (layer 1) consisted of 16 huge basal timbers up to 60 cm in diameter. They lay transverse to the length of the breakwater, spaced about 2 m apart (Meyer 1869, Marstrander 1967, Sognnes 1985, Jasinski 1993).
Meyer also mentioned the presence of rectangular mortises passing right through these timbers to take vertical pales which served as stanchions holding the timbers in place and at the same time stabilising the whole construction. These probably extended right up to the upper surface of the breakwater where pairs were linked together by logs transverse to the construction (Meyer 1869, Marstrander 1967). A layer of narrower logs (layer 2) was placed on the basal timbers in the longitudinal direction of the breakwater, with more substantial ones outermost on the long sides (Meyer 1869, Marstrander 1967: 265).


According to Meyer's report, a transverse timber lay directly on the strong lateral logs of layer 2 and was inserted halfway into each pale in the pairs of stanchions, and additional transverse timbers had been placed in the same manner a little higher up. Between these cross-timbers, strong logs were placed on either side along the entire length of the breakwater (layer 3 ?? ), corresponding to the lateral logs in layer 2. The breakwater had thus been constructed as a succession of large, partly open, timber cribs or wharf piers ?? of pine, filled with stones (after Marstrander 1967: 266).


Meyer also described traces of diagonal stakes on the south side of the breakwater. Their function was probably to support the underside of the surface of the breakwater to increase its ability to resist waves (Meyer 1869, Marstrander 1967: 266).
The topography of the bay offers natural protection to winds from the south and southeast. However, the bay lies open to northerlies and northwesterlies, the prevailing winds in the area. The breakwater was oriented ENE-WSW, clearly with the objective of serving as a wave breaker when the wind and waves were directed straight into the bay (Meyer 1869, Marstrander 1967: 266).

2. Rampart
Meyer (1869) assumed that the breakwater had once continued from its present end on land at an angle landwards towards a ridge of large stone blocks, earth and smaller stones, which he interpreted as the rampart described in the sagas (Meyer 1869, Marstrander 1967: 267-268). At the time, this was approximately 29.5 m long and about 4.5 m wide (Meyer 1869, Sognnes 1985: 61).

3. Church foundations
Krefting had already in 1865 reported seeing the remains of the foundations of the church near the farm at Agdenes. Meyer did not provide documentation of these foundations, but he marked their position with a cross on his plan of the locality. According to this, the church should be located close to the rocky knoll near the present farm buildings, on the seaward side of the knoll.


Analyses of the timber

Sebelien (1929) undertook chemical analyses of timber from the breakwater to find out why the construction had been preserved and not been eaten by shipworms. Some timber showed clear evidence of attack, but this stopped after a while. Sebelien searched for signs that could suggest that the timber had been impregnated before being used in the breakwater, but the analyses failed to give a conclusive answer. He was unable to prove impregnation, but concluded that the probable reason for lack of interest on the part of the shipworms was the low content of lignin relative to cellulose.


Second phase of field investigations

This phase covers fieldwork undertaken by Marstrander in 1962 and 1964 (Marstrander 1967), Molaug probably in 1966 (S. Molaug 1967) and Møllenhus in 1967 (Møllenhus 1967).
Marstrander (1967) drew the breakwater as he saw it - the construction had almost completely collapsed and remains of the timber structure were only preserved for a length of around 20 m from the edge of the sea landwards. He found only 5 of the 16 basal timbers recorded by Meyer (1869) and 8 of the transverse timbers from layer 2. He took samples for radiometric dating from timbers which were almost completely buried in the beach sand. These gave an age of AD 870 ± 70 years, i.e. the Viking period, long before King Øystein was born.


Marstrander was in doubt whether the rampart described by Meyer was man-made or a natural feature (Marstrander 1967: 267-268). He thought the church foundations were still visible as a low grass-covered earthwork about 60 m due east of the present barn (Marstrander 1967: 270).
On his visit in 1962, Marstrander was assisted by two divers from the Trondheim Underwater Club who investigated the sea floor offshore from the breakwater towards the inner bay itself "...without observations of particular interest being made."


Molaug visited the locality in 1966 or early in 1967. He recorded the same elements as Marstrander described in his article. He interpreted the breakwater as a kind of bulwark of stones and timbers built in piers at right angles to the principal wave direction in such a way that the shoreward side was uneven. The vertical pales, he believed, could not have been strong enough to prevent the sea from breaking down the breakwater. To reduce the power of the sea, timbers extended perpendicularly into the sea from the north side of the breakwater. These seemed to be spaced 2-2.5 m apart. Between and above them were large stones, thus forming a slope back to the breakwater. This was the system with which Molaug was familiar from dike constructions on the south side of the North Sea (after S. Molaug 1967).
Molaug also wrote that there was no clear evidence of the church which was supposed to have stood in the field east of the farm buildings, but he noted that in certain light conditions there, nevertheless, did seem to be a rectangular area where the vegetation differed slightly from that in the rest of the field (S. Molaug 1967).
Møllenhus travelled to Agdenes on 7th September 1967 following a telephone conversation with the farmer, Kristen Selven, who reported that while levelling a terrace between the house and the shore (about 30 m from the house and somewhat further from the shore) he had uncovered a stone cist. It was made of thick flagstones standing on edge and was approximately 2.35 m long and a maximum of about 1.45 m wide. It was oriented E-W. The stones had a fairly even upper edge.
On 12th September 1967, Møllenhus excavated the cist, which was about 0.5 m below the surface. It was filled with gravel and medium-sized stones. Some stones had been cracked by heating and these mostly lay in a 10 cm thick, charcoal-rich layer between 10 and 20 cm below the upper edge of the cist. In the westernmost half of the cist, this layer contained a brownish, greasy substance, up to around 5 cm in thickness, and which was scarcely derived from bone. Møllenhus wrote that the contents of the cist seemed to indicate a fireplace. Its age was difficult to judge, but he thought it was most likely from the Christian era (after Møllenhus 1967).


The third phase

In 1982, 1984 and 1986, Sognnes (1985, 1987, 1988) undertook fieldwork at Agdenes. This consisted of documenting the remains of the breakwater on the beach, digging trial pits in search of occupation layers that could confirm permanent settlement in the area, trial excavation and documentation of the rampart near the shore, and trial excavation where Meyer had marked on his map what he interpreted as the church foundations.


Sognnes failed to find any trace of permanent settlement near the harbour at Agdenes. Trial pits on the site of the church showed that the topsoil was thicker than elsewhere at the locality. The samples from there contained small quantities of charcoal. A layer of small stones was also found in a 20 m long strip along the south side of the site. The results could, nevertheless, not be considered as confirmation of a church having once stood on the site (Sognnes 1985).
The investigations of the breakwater on the shore gave exactly the same results as Marstrander achieved. In addition, Sognnes described three attempts to make underwater observations in the area around the breakwater: in 1972 (in connection with plans to construct a ferry quay in the bay), in 1978 (Fastner & Sognnes 1983: 54, Sognnes 1985: 62), and in 1981 (Fastner & Sognnes 1983: 47). In 1972 and 1978, the divers recorded remains of the breakwater in the form of a bank of stones down to 23-24 m below the surface. The total length of the construction beneath the water was estimated to be 65 m. Poor visibility, unfortunately, made it impossible to take underwater photographs or draw the construction (Sognnes 1985: 62). An attempt in 1981 to carry out a sonar investigation unfortunately failed (Fastner & Sognnes 1983).


As regards the rampart, Sognnes confirmed that it was definitely linked to the harbour construction and was probably the object described as virki in the Magnus's sons saga. Sognnes also found it very probable that the rampart had originally been more than 75 m long (Sognnes 1985: 62). If he is correct, it means that this earthwork, and a possible palisade on the top, was built from the then shoreline all the way to the rocky outcrop, thus completely blocking the entrance to the inner part of the harbour from the northwest, as I also believe is most likely.


New investigations

One afternoon in autumn 1991 an engineer, Terje Wenaas, was diving in the bay at Agdenes. On the sea floor in the middle of the bay, at a depth of 7-8 m, he discovered the remains of a construction which looked like a timber crib filled with stones (Wenaas 1991). His report led me to undertake an inspection a few days later, on behalf of the Museum of Archaeology and Natural History in Trondheim. The submerged construction was photographed and described. It was also clear that a rescue action needed to be implemented because the construction was in danger of being totally destroyed. Because of the sea-floor topography, the parts that were still more or less intact, along with numerous stones, can slide into deep water at any time. The crib is lying at a depth of 4-7 m. The sea floor slopes steeply downwards so that part of the construction is hanging freely in the water. Several timbers and stones were observed scattered along the slope down to a depth of about 40 m.

1. 1992 field season
In 1992, the Museum began a series of underwater investigations at the site, led by the author and with Pål Aa. Nymoen and Steven P. Carpenter jointly in charge of the actual fieldwork and responsible for report writing (Carpenter 1993, Nymoen 1993).
The 1992 season was largely spent preparing the locality for more detailed study and collecting necessary additional data. Four permanent surveying points were drilled into the sea floor, delimiting an area that was to be investigated around the newly discovered crib. Within this area, the depth has a range of approximately 12 m. Nearest land in the southwest corner of the area is a pile of medium-sized stones and larger blocks. From this, two distinct, elongate mounds of stones stretch into deeper water. The crib is in the middle of the pile of stones, between these mounds.
The remains of the crib were photographed and videofilmed, and two samples of wood were taken for C-14 dating, one from the base and one from the uppermost part.
The remains on the beach, described by Meyer, Marstrander and Sognnes, were measured anew in 1992. Eleven of the sixteen basal timbers which Meyer drew in 1869 were found, i.e. six more than were reported by Marstrander (1967). On the other hand, none of the eight timbers from layer 2 which Marstrander drew were found. The eleven basal timbers were situated along an approximately 30 m long line from the west side of the bay towards the northeast. Eight of them had vertical holes, about 15 x 15 cm, passing right through them, and a few of these contained the remains of vertical pales.


In addition to the archaeological fieldwork, geological studies were carried out and data were collected to draw up a shoreline displacement curve for this part of central Norway (Rokoengen 1992a, b). Calculations were also made of the natural forces which the breakwater must have sustained whilst it functioned (Tørum 1992). Another group of scientists investigated those parts of the site that were deeper than 30 m, which is the limit set for our marine archaeological diving. They used a ROV (Remote Operated Vehicle), a minisubmarine equipped with a video camera, and observed a small iron anchor of so-far unknown age at a depth of about 34 m (Kristiansen 1995: 153).

2. 1993 field season
Field work in 1993 chiefly consisted of underwater surveying, drawing, photography and videofilming of the crib and the pile of stones. In addition, an underwater excavation began of a 2 x 1 m trench near the crib.
Investigation of the crib showed that the same technique had probably been used to build that as was used for the cribs on the beach. As mentioned above, the crib is located approximately in the middle of the pile of stones which originally filled it. An approximately 5 m long timber extends from beneath the stones uppermost on the easterly mound of stones. This marks the outer limit of the crib. Perpendicular to this are three additional timbers with diameters of between 27 and 46 cm. Between these are eight smaller logs, which probably correspond to Meyer's layer 2.


Evidence of working can be seen on several of the timbers. This is particularly distinct on timber A on which two 30 cm wide and 20 cm deep notches are spaced 2.5 m apart, and at the same position a hole has been driven right through the timber. Two 30 cm long wooden wedges were found near the hole (Nymoen 1993). The handle of a soapstone pot was found beside the crib.


The excavation of the 2 x 1 m trench revealed a thick, compact layer of material originating from human activity (hereafter called an "occupation layer", covered by 5 cm of sand. The slope of the sea floor along with the loose sand meant that material was easy to move and excavation could take place manually. However, the slope soon became a problem since unconsolidated material slid down and filled the trench during short pauses in the excavation.
Tests made southwest of the trench showed that the "occupation layer" continued for several metres towards the shore, but its northeast boundary towards sterile sand was approximately in the middle of the trench (after Carpenter 1993: 10).


Its maximum thickness was 45 cm. The material seems to have accumulated more or less at one time and to have been exposed to pressure. It was relatively homogeneous and distinct stratification could not be discerned. Pieces were cut out and taken to the surface as samples. The material can best be described as a compact, brownish-black organic mass consisting of large slivers of wood and plant remains (including moss and leaves), and in addition some fish bones, nutshells, unidentified wooden objects and an almost completely preserved baler.


Work continued on land where a detailed survey was made of the eight basal timbers on the beach and of the rampart. The aim here was to clarify the relationship between all the elements so far recorded, i.e. find out in what way the features illustrated by Meyer (1869) and Marstrander (1967) are related to those found on the sea floor. This work was subsequently extended, resulting in a precise map of the harbour (with 0.25 m contours) showing all the features so far known (Fig. 16).


3. 1994 and 1995 field work
Because of lack of funds, work in 1994 and 1995 was almost confined to the preliminary phase of a gradiometer investigation carried out on land (Binns 1995). This revealed anomalies northeast of the present farm buildings that may mark two or three houses and a group of pits containing charcoal.
In addition, six new samples for C-14 dating were taken, one from each of the five basal timbers recorded by Marstrander in the 1960's (Marstrander 1967) and one from one of the six additional timbers recorded in 1992.

Interpretations and problems

The investigations so far carried out confirm that the site at Agdenes is unique in many ways. First and foremost, this is the only harbour of its kind studied in Norway that has a construction preserved on the sea floor. The closest analogies are the harbours in the Medieval towns in Bergen and Oslo (Herteig 1990, P.B. Molaug 1993), but these differ considerably from Agdenes in building technique and function.
I believe that the archaeological data and the information from written sources regarding the harbour at Agdenes call for new evaluation and interpretation. A research and management plan for the site is also required. Even though research has taken place at the harbour for a long time, it is quite clear that the work is far from complete.

The interpretation of the site must concern four aspects:

A - the physical structure of the harbour
B - the chronology
C - the purpose and function
D - the position of the harbour in the maritime cultural landscape of central Norway.

A 1 Physical structure of the harbour


During the period from Krefting's observation in 1865 to the present day, research at the site has revolved around four aspects: the breakwater, the rampart, the site of the church, and traces of possible permanent settlement near the harbour. Of these, two (the breakwater and the rampart) have been documented and partially investigated. The search for possible church foundations has given no definite result. Traces of settlement have so far not been found, except for the fireplace and the magnetic anomalies indicating possible charcoal pits and house remains. I will now evaluate what we know about these various elements.

1. Breakwater and rampart

The bay at Agdenes is not a good natural harbour. It provides protection from only a few wind directions, and when the wind blows from the north and northwest large waves break on the beach. The locality, nonetheless, has a very important strategic position. The building of a breakwater protecting the bay from the north and northwest transformed it into an excellent harbour which could be used in various contexts.
We know how the breakwater was built. At present, the only questions are whether its remains are standing where the construction was built and whether the crib found on the sea floor in 1991 constituted the end of the breakwater, or had another function. The latter question must be posed chiefly on account of the depth at which the crib is located.


Marstrander (1967: 267) pointed out that if the breakwater was to provide adequate protection it had to be at least 1-1.5 m higher than the high-water mark. This means that in the present beach zone the breakwater must have been at least 3.5-4 m high at the low-tide level. It is difficult to disagree with this. Problems, however, arise when corresponding calculations are made concerning the newly discovered crib on the sea floor. If this marks the end of the breakwater in the bay, it must be assumed that the breakwater in this part of the bay reached a very considerable height. The crib lies at a depth of between 4 and 7 m. To reach sufficiently far above sea level, 1-1.5 m must be added. Allowance must also be made for the change in sea level from the Middle Ages to today; according to Kjemperud (1986) and Rokoengen (1992 a, b), the land in this area can be expected to have risen at least 2.5 m. All told, this gives a height of 7.5-11 m for the end of the breakwater in the sea. The breakwater at Agdenes must therefore have had quite huge dimensions.
Alternatively, the underwater crib may originally have been located closer to the shore and has subsequently slid into deeper water. This is not improbable, taking into account the bottom topography in this area. However, we lack data about the situation between the end of the construction on the beach and the crib. This question can therefore not be answered at present and must await further investigations.


It should be mentioned, here, that the suggestion that a landslide has caused substantial changes to the harbour was put forward by Larsen (1871). He used this hypothesis to explain the lack of continuity between the breakwater on the beach and the rampart. Rokoengen (1992a, b) did not rule out the possibility of a landslide at Agdenes. Perhaps it could also have caused movements of the sea floor that resulted in the underwater crib sliding down.
Another feature that must be taken into consideration is the "occupation layer" found on the sea floor beside the crib. I believe there are three possible explanations for this layer:

- the material originates from the site where the crib was built and has been transported onto the sea floor by a landslide
- it derives from a building of some nature which stood in the shore zone, and has subsequently slid into the sea
- the organic material was used as packing between the timbers in the cribs, which would have enhanced their buoyancy so that they could be towed into place, filled with stones and sunk in a much more controlled manner than otherwise.

The first alternative implies that the crib discovered on the sea floor did not constitute the end of the breakwater, but was originally placed much closer to the beach and functioned, for example, as a quay or the foundation for a wharf (see, for example, P.B. Molaug's (1993) description of the harbour construction in Oslo). The second alternative cannot be ruled out either. Without knowing of the existence of such an underwater "occupation layer", Sognnes (1985: 64) mentioned the possibility that buildings may have stood close to the shoreline and been subsequently destroyed by erosion. Only further investigations can clarify this question.
According to Tørum's (1992) calculations, a breakwater constructed of stone-filled cribs would offer only limited protection against the force of the waves, the protection being greatest innermost in the harbour near the foot of the breakwater. S. Molaug (1967) pointed out that it is precisely there that constructed landing sites can be expected to be found at Agdenes. No attempts to confirm the presence of such sites have been made, but I believe the theory to be very plausible.
I think the breakwater and rampart have once been physically linked. Some large stones are still visible between them. My hypothesis is that this part of the breakwater was built entirely of stones, since it was unnecessary to use wooden cribs to contain stones in shallow water. The difference in height between the rampart and the closest basal timbers on the beach is less than 3 m. There would no difficulty in building a stone filling at such a depth. It is only at greater depths that it becomes difficult to control how the stones position themselves on the sea floor, thus necessitating the use of cribs to hold the stones. Meyer (1869) also thought that these two features were once connected, but without going into details concerning the materials used. If we accept that the breakwater was continuous with the rampart and that the underwater crib marked its end, such a slightly curved harbour wall and breakwater would have a total length of around 90 m.
Another possibility was proposed by Sognnes (1985), who did not believe there was any connection between the breakwater and the rampart. He envisaged that the latter would have hindered people who were able to land on the beach west of the breakwater from gaining access to the inner part of the harbour, and at the same time given the harbour area protection against hostile crews who wanted to land on the inner side of the breakwater. This is undoubtedly an interesting idea, but again its confirmation, or otherwise, requires more work in the area concerned.


The Agdenes bay basically offers poor protection against wind and waves. The building of the breakwater meant that it was transformed into an excellent harbour. If the sea level was approximately 2.5 m higher than now, a logical structure would be discernible with a breakwater protecting a large, calm area sufficiently deep for contemporary ships designed for warfaring or carrying cargo. The situation at the time the harbour was built, with the contemporary shoreline following the present 2.5 m contour, i.e. along the northwest margin of the rampart and curving round to the rocky bluff east of the harbour. The rampart prevented access from northwest right across to the rocky bluff in the west, i.e. the 5 m contour. On land to the southwest of the rampart, there was adequate space to build landing sites for boats and to draw the boats onto land. Higher up, the area where the present farm buildings stand provided space for houses and other settlement facilities.

A 2 The church and permanent settlement

The search for church foundations has so far not produced conclusive results. It is difficult to say whether the foundations which Krefting, Meyer, Marstrander and Molaug were able to observe really were the remains of a church, or of other objects. Test pits and trial excavations by Sognnes (1985, 1987) indicated traces of human activity in the area concerned, but could not confirm that it was really a church that had stood there.
According to chapters 14 and 21 in Magnus's sons saga, King Øystein built a harbour, a rampart and a church at Agdenes (Snorres kongesagaer 1979). The problem is that we do not know whether the harbour and the church were built at the same place.

With the information currently available, there are three possibilities:
- the church may have been located in a different place from the harbour
- a small wooden church stood on the site where Krefting believed he could discern foundations, but nothing or very little is now preserved
- the church stood where the farm buildings now stand.

The permanent settlement could also have stood in the area where the present farm buildings are located. The fireplace that was excavated by Møllenhus and the area of possible house remains and charcoal pits found by Binns suggest this possibility, even though Sognnes' hypothesis of settlement on the shore cannot be excluded, particularly in light of the data obtained through studies of the "occupation layer" on the sea floor. Settlement may, indeed, have been present in both places.

A 3 Other cultural monuments near the harbour

According to the text of the Magnus's sons saga cited by Helland (1898: 53), King Øystein constructed beacons on the surrounding hills, as well as building a harbour and a church. These possible beacons interested me greatly. Near the harbour are two hills where it would have been natural to construct beacons. One is Kamsheia (137 m a.s.l.) southwest of the harbour, and the other is Gangerheia (66 m a.s.l.) southeast of Agdenes farm.
On the summit of Kamsheia is a group of large stones which may be the remains of a beacon. Nearby there is also a cairn, recorded earlier as a grave. What is, however, quite certain is that Kamsheia offers an excellent view across the approaches to Agdenes from the west and north to Leksa, Storfosna, Ørlandet and Stjørna, as well as to Rissa and far in along Trondheimsfjorden in the east. There can be no doubt that Kamsheia acted as an outlook post for the harbour at Agdenes.


On the summit of Gangerheia is a previously recorded cairn of huge dimensions (28-30 m in diameter). It was mapped and described by Ertsås (1968), who wrote that it seemed to have been built in two phases. A relatively small, round, cone-shaped mound of stones rests upon a much broader, circular and fairly flat base. According to Ertsås, this rather unusual form may have arisen through extensive damage to the cairn by treasure seekers or the like. The report also states that a square-shaped pile of stones has been built on the cairn in recent times - as a beacon or similar feature (Ertsås 1968).
After visiting the site myself, I consider it most probable that the unusual form is a result of a signal beacon having been placed on top of the cairn. Such practice is not unusual. Signal beacons (piles of wood periodically lit as a signal) were often placed on old cairns so that the lowermost pieces of wood would not rot on damp ground, etc. In connection with this, the cairns were often flattened out to form a platform on which the beacon could stand (see, for example, Scheen 1951: 264). The square of stones which Ertsås mentioned could have been the stand for the signal beacon on Gangerheia.

B. Chronology of the harbour

The Magnus's sons saga reports that the harbour at Agdenes was built during the reign of King Øystein (1103-1123). In the comparison of the achievements of King Øystein and his brother, King Sigurd, Øystein claims that he built the harbour whilst his brother was on a journey. This took place in 1108-1111. Hence, if we are to believe this Snorre saga, this was when the harbour was built.


As mentioned previously, Marstrander obtained a C-14 dating of one of the timbers on the beach. The result was 1080 ± 70 BP, i.e. the tree was felled sometime between about AD 800 and 940 - long before King Øystein's time (Marstrander 1967: 271).
This result led to a debate as to whether there really was "no harbour before" before King Øystein's time, as he himself claimed (Marstrander 1967: 271, Sognnes 1985, Christophersen 1993, Jasinski 1993). Marstrander (1967: 271) pointed out that a possible explanation was that old timber was used for the breakwater. Sognnes (1985: 62) wrote that Marstrander's dating was not calibrated. I have applied the OxCal v. 2.15 calibration program to this dating, and obtained a much broader age of AD 770-1120 (with 2 sigma - 95.4% probability). Hence, this does not exclude the possibility that the tree was felled during the time when the harbour is supposed to have been built, i.e. 1108-1111.


As mentioned earlier, in 1992 I took two samples for C-14 dating from wood used to build the underwater crib. These gave two different ages, the first AD 603-874, and the second AD 1002-1219 (both with 2 sigma, 95% probability). The first is in poor agreement with the age of the harbour according to the sagas, but King Øystein's reign falls midway in the span of the other dating.
Six additional samples were taken in 1993, five from the basal timbers which Marstrander (1967) recorded and one from one of the six additional timbers found in 1992. These gave the following results (all with 2 sigma, 95% probability):

AD 875-1010
AD 885-1035
AD 895-1170
AD 1030-1280
AD 1035-1270
AD 1075-1155

Two stem from the period prior to King Øystein's reign and, again, old wood may explain these. The other four agree well with the period reported in the saga. Two even stretch across the reign of King Håkon.


I believe that this series of datings probably confirms that King Øystein was responsible for the initial constructional work on the harbour at Agdenes. I consider that the possibility that the harbour was originally built during the Viking period and that King Øystein merely made improvements to it is now very unlikely.
As regards subsequent developments, information that can be interpreted is extremely sparse. The harbour was most probably functioning in 1165 when Erling Skakke sailed to Nidaros (Magnus Erlingsson's saga, chapt. 25). According to chapter 333 in Håkon Håkonsson's saga, King Håkon (who reigned from 1217 to 1263) also built a rampart at Agdenes. We know nothing of where this was built or what form it had, if we ignore the possibility that the rampart near the shore was built by Håkon, not Øystein.


C Purpose and function of the harbour

It was Marstrander (1967) who first took up the question of the reason for building the harbour, and its function. His conclusion was clear and unambiguous: "There is nothing to suggest that this simple construction of stone and timber had any military significance" (Marstrander 1967: 268). He added that the harbour at Agdenes is one of numerous places along the coast where seafarers could find a haven, whether to shelter from a storm or bad weather, or to wait for a favourable wind. He believed that the Agdenes harbour must have been one of the more important harbours of this type along the channel since it had specially built facilities. He also pointed out (Marstrander 1967: 269-270) that the harbours along the coast were not only visited in cases of emergency, but also served as resting places during long sea journeys and as meeting venues.


When Marstrander wrote, he knew nothing about the underwater portion of the harbour construction. The divers who searched the sea floor for him made no observations of interest. Marstrander was also uncertain whether the rampart was really the work of man. All he saw was five timbers surrounded by stones, lying on the beach. His conclusion that "this simple construction of timber and stone" had no military function is therefore understandable in view of the data available to him.
Sognnes (1985, 1988) reached another conclusion. He showed that the rampart near the shore really was a rampart and he learnt from divers that the breakwater also continued beneath the water. His conclusion was that, in addition to being an emergency haven from storms and while waiting for a favourable wind, the facility at Agdenes could have acted as a control post in times of strife. He pointed out that there is no evidence that can confirm other functions, such as a trading centre, etc.


The possibility that Agdenes was a trading centre was, nevertheless, taken up by Christophersen (1993). His hypothesis is solely based on the uncalibrated C-14 dating from the Viking period obtained by Marstrander, and his own theory that prior to the establishment of Nidaros a trading centre of the Kaupang type existed beside Trondheimsfjorden. An important requirement for such a centre would be a good harbour, and this is why Agdenes was proposed (Christophersen 1993: 165-166).


As regards the function of the harbour, my opinion is that it was built by King Øystein in 1108-1111 principally for political and military reasons. The form of the construction required a great deal of manpower and technological know-how. The choice of the location also confirms strategic thinking. Agdenes bay, as pointed out earlier, is not a good natural harbour, but its geographical position can scarcely be better if the desire is to establish a control and defence post. With lookouts on Kamsheia and Gangerheia, and men-of-war and their crews in the harbour, an effective control of the approach to Trøndelag could be practised. Such control posts are not a new thing (see, for example, Crumlin-Pedersen 1993). Nidaros at this time functioned as both a trading and a political centre in central Norway. It must also be remembered that the harbour was built during the period when civil wars had still not completely ceased and Trøndelag was still only somewhat loosely attached to the rest of the kingdom. In this context, the construction at Agdenes had, in addition to purely practical purposes, a clear symbolic value signalling royal power.


Sognnes (1985: 64) was undoubtedly correct when he wrote that because of the currents around Agdenes, boats that intended to enter Trondheimsfjorden often had to wait for a favourable wind. They would then seek shelter in the vicinity and would naturally favour the bay at Agdenes. I believe there is no better way of maintaining a check on who is sailing into the fjord, and why, than to establish a permanently manned harbour close to a difficult approach. The presence of a garrison is also, I believe, implied in Chapter 333 of Håkon Håkonsson's saga where there is mention of the townsmen who were unable to distinguish a longboat among many other vessels when Erling Skakke in 1165 attacked Nidaros, even though the people of Trøndelag claimed that he would never succeed in passing Agdenes alive. The expression the townsmen cannot mean anything other than men sent from Nidaros or men in the service of Nidaros. What were men from Nidaros doing at Agdenes? It is very likely that the expression is referring to a military garrison at Agdenes.


The military and political importance of the harbour is also confirmed by the Håkon Håkonsson saga which states that the king built a rampart in the area. In this context, it is important to point out that on several old maps from the 16th to 19th centuries the northern tip of Agdenes is equipped with a dot, an anchor or a symbol for a fortress (see, for example, the map of northern regions published in 1570 by Abraham Ortelius, the map of Russia published by Mercator in 1597, and the map of the southerly parts of the Trondheim diocese published by Johan Blaeu in 1662, etc.).


This means that a tradition concerning Agdenes as a harbour and as a fortified locality was maintained for quite a long time after the period dealt with in the sagas. The place has such strategic importance that a small battery with six cannons was placed on one side of Agdenes bay in 1814, and in 1895-97 the so-called Agdenes fortifications were built, partly at Agdenes itself and partly across the sound at Ørlandet. Their purpose was undoubtedly identical to that of King Øystein's harbour, to guard the approach to Trondheimsfjorden.

D. Position of the harbour in the maritime cultural landscape

An important problem which requires further research is the question of whether the harbour at Agdenes existed and functioned quite independently of other maritime localities in this part of central Norway, or whether it was just a cog in a larger system. I have only just begun such a study and it will take time to reach any conclusions. Nonetheless, it is very likely that several control posts existed around Agdenes, for instance at Storfosna, in Kråkvåg and at Uthaug on Ørlandet, all on the opposite coast of the approach to Trondheimsfjorden.


It us also probable that Agdenes bay functioned as part of a warning system. The beacon on Gangerheia could have functioned contemporaneously with the harbour. In the area around Agdenes and inwards along both sides of Trondheimsfjorden many names of hills on nodern maps include the word vete (signal beacon). The closest one is Vetaliheia on the south side of Selvbukta, This observation may be of great importance since Selvbukta is the closest natural harbour en route from Agdenes into the fjord, and it is probable that the nearest king's farm was located precisely there in the early Middle Ages. In the other direction, towards the north, we find another vete name on Storfosna, clearly visible from Gangerheia beside Agdenes harbour, as Marstrander (1967) showed without drawing any conclusions (his figure legend reads "The bay at Agdenes seen from the southeast. The remains of the breakwater are on the small headland. On the horizon, Vetaheia on Storfosna").
I am convinced that a study of place-names combined with topographical investigations will, in the future, uncover a complete system of signal beacons from Agdenes all the way to Nidaros. By lighting the Gangerheia beacon in a dangerous situation observed from the outlook post on Kamsheia it would be possible to warn Nidaros relatively quickly that enemy men-of-war were passing the approach to Trondheimsfjorden.


Conclusions

The marine archaeological investigations carried out at Agdenes have contributed some valuable new data. Nonetheless, studies here are certainly far from complete, and many unsolved problems and questions remain. Investigations both on land and under the water should continue.
The investigations have also put some management problems onto the agenda. These are first and foremost related to the question of how to stabilise and preserve the underwater constructions at Agdenes for future generations. Because of the power of Nature and the sea-floor topography, the constructions will disappear into the depths relatively soon if nothing is done. I have already mentioned the problem of obtaining funds for further research at the site. The same applies to the necessary protective measures. Each time I dive in the harbour I see changes. More and more elements have become detached and disappeared. Agdenes needs a new "King Øystein" willing to put effort and money into preserving the last surviving traces of the construction.

 

 


Buscar en esta seccion :