Harbour sites in ritual landscape--the example of Viltina in southern Saaremaa/Sadamakohad rituaalsel maastikul--Viltina naide louna-Saaremaal.

Magi, Marika

The location of landing-sites in cultural landscape seems to be determined predominantly by suitable natural conditions, in order to enable safe landing for the water vehicles. In case of harbour sites, the dimension of cultural landscape is added, meaning that the place needs to correspond to conditions that are important in the given society. However, the conditions vary depending on the function of the harbour site. In my earlier works I have demonstrated that, in the pre-state society, in addition to topographic indicators the choice of the harbour site was greatly dependant on the location of agricultural land in the radius of at least 5 kilometres. Archaeological sites in arable lands or in their vicinity enable guessing the one-time function of the harbour site. In addition to practical tasks (e.g. commercial harbour, fishing harbour, war harbour etc.), other possibilities that are more difficult to perceive by archaeological methods, for example a ritual gathering place, are suggested. Of the Estonian more closely studied Iron Age and medieval harbour sites, Viltina on the southern coast of Saaremaa Island seems to be the case in point (Fig. 1).

[FIGURE 1 OMITTED]

The settlement site of Viltina in cultural landscape The 18th century could be considered the beginning of the village of Viltina. On the cadastrial map from the end of the 17th century there are still single farmsteads (Estonian Historical Archives 308-2-64); however, a small hamlet named Filtin has already been marked on a map a century later (Mellin 1798).

Of the single farmsteads from the 17th century the biggest and perhaps the most important one was called Koffera, a slightly smaller farmstead next to it bore the name of Murke. In the 17th century the farmsteads were situated approximately 600 metres from the coast. Both Koffera and Murke as well as other farms on the territory of later Viltina at the time belonged to the manor of Audla (Haucul), the centre of which was located at 7.5 km distance (Estonian Historical Archives 308-2-64). Since the rest of the farmsteads in the Koiguste Peninsula belonged to local manors, from the perspective of the local cultural landscape the separate coastal estate of Audla manor might be considered a slightly unusual feature. It is possible that the phenomenon can be interpreted as a remnant from the period when the surroundings of Viltina were still important as a harbour, which was administered by Audla as the oldest and the most influential manor of the neighbourhood.

It is not known when the coastal farmsteads of Viltina started off. In the majority of cases identified in the Baltic Sea countries the farmsteads and villages of the second half of the Iron Age remained at one or two km distance from the coast. The permanent habitation shifted closer to the sea only in the 13th-14th century (see e.g. Crumlin-Pedersen 1996). The phenomenon is mostly explained by the activity of pirates, which was characteristic of pre-state societies and made the living too close to the coast dangerous. The situation changed with the formation of central administration that was capable of providing enough safety to the coastal villages. However, during the Middle Ages and later people never lived directly on the beach--the coastal villages of the time always remained at least 100-200 metres from the waterfront. Daily life directly on the beach would have been complicated because of storms, the rise of the water level and ridged ice. On account of the said, the archaeological settlement layer in the vicinity of the former shoreline is everywhere in the Nordic countries interpreted as a harbour site, even if its general appearance hardly differs from a common settlement site (e.g. Crumlin-Pedersen 1991; Nasman 1991; Christoffersen & Porsmose 1996; Ulriksen 1998; Magi 2004a, 147 ff.).

The arable land of Viltina village is stony and not especially fertile; its quality is slightly better in the villages of Randvere and Asva, which lie 1-2 kilometres from Viltina in opposite directions. In Randvere a settlement unit has been located since the first centuries AD at the latest. A stone grave from this period has been excavated near the village. During the Viking Age a stone grave partly with stone circles, partly without inner constructions was erected next to the earlier burial site (Magi 2002, 47 ff.). In the fields of the village of Asva stone graves marking an early settlement unit are absent; however, different graves are known from the lands of the villages of Kahtla and Laimjala.

From the 10th century the dead were buried in the territory of the present Viltina village, on a hill that back then was located directly on the shore and is now called Rutirank by the local people. This stone grave with the total area of 3236 sq m was completely excavated in 1940 in connection with the plan of establishing here a Soviet military base. The works had to be completed hastily, wherefore the excavation methods were much to be desired in places. Thus, for example, a big part of the stone layer of the Rutirank grave completely remained undocumented. The grave was characterized by conspicuously rich find material that dated it mainly to the 10th-12th centuries. The bulk of the dead were buried there during the 12th century when the deceased were also accompanied by the richest grave goods (Magi 2002, 60 ff.). Since the majority of the burials were mixed cremations, which used to be characteristic of the Estonian burial custom of the time, the more exact number of people buried in the grave remained unclear. According to the calculations of Artur Vassar, approximately fifty people had been buried in the eastern half of the Rutirank grave, the part excavated by him (Vassar 1940). Thus the number of the deceased in the whole grave might have been about a hundred. As indicated by fords, both men as well as women were among them. In case we are dealing with a burial site that functioned according to family principle, as commonly suggested about the Estonian stone graves, two-three families could have buried their dead into Rutirank. The considerable amount of weapons and horse gear among the grave goods seems to be referring to the high social status of the deceased in the contemporary society (Magi 2002, 60 ff.). Archaeologists documented a whole row of single graves in the close vicinity of the Rutirank grave, some of which were also excavated. Archaeological research at Viltina, which started off again since 1999, has shown that more graves can be found there.

About 1.5 km south-west from the Viltina Rutirank grave, on the fields of Randvere village, another and an almost contemporary stone grave was excavated in 1940. The later analysis of the find material gathered there has indicated that since the 11th or the 12th century at the latest mainly children were buried into the Randvere stone grave. It can be suggested that the dead of the most important family in Randvere who were earlier buried into the local stone grave, were buried into the Viltina Rutirank grave during the 12th century, which would explain the big number of the deceased there. The rest of the families that used Rutirank as a burial place probably came from Asva or its close surroundings. These were possibly people who had enough influence to regulate the usage of the Viltina harbour site and who demonstrated their power with a stone grave on the coast (Magi 2002, 47 ff.). The concentration of graves in Viltina thus refers to either the increase of the proportion of maritime activity in the local society or the growth of the importance of Viltina as a religiously meaningful place in the 11th-12th century.

In addition to graves a hill fort is situated slightly more than a kilometre from the village of Viltina. A fortified settlement was located in Asva, on a former islet close to the shore as early as during the Bronze Age. Already back then it was probably not so much connected to the local agriculture as to the bronze casting and international trade. In any case a Bronze Age harbour site that could be considered a distant predecessor of the Viltina harbour had to be located somewhere near the Asva fortification. As much as can be suggested by the Bronze Age shoreline followed by contour lines, the place had to be suitable for this in every respect (Fig. 2).

However, a Bronze Age harbour site has never been found at Asva and its locating is hardly likely in the future either. The experience of the Nordic countries shows that harbour sites earlier than the 5th-6th century AD are difficult to identify archaeologically (e.g. Carlsson 1992; Ulriksen 1998, 134 ff., 194 f., 216 ff.). The reason is not exactly known. It is clear, however, that harbour sites definitely existed before the 5th-6th century. It is possible, though, that the cultural layer in these places is extremely scarce. It is suggested that there were no central harbour sites before this period, since the administration had hardly concentrated enough to assemble bigger harbour sites under its control. Thus we are only dealing with numerous small harbours that were not very intensely used and therefore left only weak traces into the ground. Another possible reason can be that, since the majority of later harbour sites have been located with the help of a metal detector, the less use of metal during the earlier periods of prehistory makes it difficult to detect these sites.

The Asva Bronze Age fortified settlement has been excavated considerably thoroughly (e.g. Indreko 1939; Lougas & Selirand 1989, 204 f.). The place was shortly used around the turn of era but only few finds have been preserved from the period. New hill fort was erected on top of the former fortified settlement probably in the 6th-7th century and it remained until the first half of the Viking Age or the 9th century. In connection with the land upheaval a former island had become a narrow cape near the shore, but the surroundings were continually suitable for a harbour site: namely here a bay was formed, which was well protected from winds and where deeper water reached close to the shore with suitable incline. Also this Pre- and Early Viking Age harbour site has not been found until now. However, it seems logical that the Pre-Viking Age hill fort was erected in Asva, in an unimportant place at the first glance, particularly to protect the harbour site.

[FIGURE 2 OMITTED]

During the 6th-9th century the harbour site could move many times when the ground rose and the old harbours became hardly accessible to ships. According to contour lines a suitable location for a harbour should have been at Viltina, on the western shore of a small bay where the farmsteads of Kao-Matsi are situated at the present, already during the 9th-10th century at the latest. However, the latter have remained on the spot already for a long time, having left a cultural layer behind. It is difficult to say whether the traces of any Viking Age activity could be found under this.

The mentioned bay started to overgrow already at the end of the Viking Age, in any case the edge of the presently swampy area a few hundreds of metres south-east from Kao-Matsi was used as a harbour during the 11th-12th centuries. This is the place where archaeological excavations took place in 2004-2006. Asva hill fort had been abandoned by the 11th century, perhaps for the reason that it would have remained already at 1400 m distance from the Viltina harbour, away from the sea border, and thus the functioning of the hill fort as a fortification protecting the harbour would have been problematic.

It is also not known where the harbour site moved in the 13th century. Gradual shifting can be followed at many historical harbour sites (Scandinavian parallels e.g. Lundstrom 1981; Carlsson 1999, 181 ff.; Magi 2004a, 147 ff.), and the 12th century was a time of an especially fast land upheaval that caused the re-location of many harbour sites considerably closer to the sea (e.g. Callmer 1991; Magi 2007c). Perhaps the medieval Viltina harbour site was situated in the vicinity of Koffera and Murke farmsteads as could indirectly be indicated by the later belonging of this area to the Audla manor. At the end of the 17th century the successor of the Viltina harbour was located already almost two kilometres south, in Ruhve, where a coastal village had formed by the time as well (Estonian Historical Archives 308-2-64).

The overview of the Viltina harbour site

Artur Vassar who excavated the grave of Rutirank was the first to notice the location of the grave on a former coast suitable for landing and suggested the existence of a harbour site in its vicinity (Vassar 1940). The harbour site was actually found 50 metres north-east from the Rutirank grave in 1999 (Magi 2000). Archaeological excavations on the site were conducted in 2004-2006. Altogether 330 sq metres divided into six separately situated excavation plots were studied at Viltina (Fig. 3; Magi 2006; 2007a). The locations of the plots had been chosen in a manner that would enable acquiring an overview of the structure of the harbour site and the functioning of its different parts.

Buildings located at the harbour site

The first and the biggest excavation comprised the southern part of the harbour site where earlier finds suggested probable building remains. After removing the thin humus layer, the lines and rows of stones appeared that marked the boundaries of the former light and considerably small buildings (Fig. 4). Similar building remains have been found from several Finnish and Scandinavian settlement sites (e.g. Westerdahl 1989, 101 ff., fig. 70). The majority of these buildings had been erected in the traditional cross-beam technique and the stone lines were apparently the stones crammed between the lower beams of the walls.

There were several buildings, the exact boundaries of which remained unclear. This was conditioned by the construction style: usually no proper foundation is established under the cross-beam buildings, especially in case of the solid ground of limestone and gravel as at Viltina. For the same reason it was not necessary to dig postholes, and the wooden parts of buildings do not preserve in the ground rich in limestone (ethnographic examples Rank 1939; Tihase 1974; archaeological examples see Lavi 2005). Now and then the stones surrounding circular holes of the diameter of approximately 20 cm could be detected. These holes apparently marked the one-time posts that had been supported by stones. It might be suggested that houses had been significantly re-built during the 150-200-year-long period. The houses were probably relatively small and erected tightly next to each other. Buildings measuring approximately 4 X 5 metres have been excavated elsewhere in Estonia as well (e.g. Lavi 2005) and these were especially distributed at hill forts where the construction space was limited. The same might be suggested for the Viltina harbour site.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

Although other archaeological excavations have confirmed that mostly stoves were built inside Estonian dwelling houses back then, no ovens were found at Viltina. In the excavation No 1 a preterit open fireplace was uncovered, the cracked stones of which had been later scattered. There was also a singe U-shaped fireplace built of stones and open on the top. Such an open fireplace of very temporary character, as well as the absence of ovens in the 11th-12th century buildings, imply that the houses were apparently needed only during the period when the weather was warmer and the sea navigable. The Viltina harbour site was not used during winters and therefore there was no need to build proper ovens in the houses.

Only 10 metres from the former coastal slope a light building with a circular ground plan had stood. A circle of boulders with the diameter of 2.8 metres and a hearth were preserved inside the building. Vertical beams probably leaned on the stones. Similar Modern Age building remains were unearthed in the settlement site of Proosa near Tallinn where it was interpreted as a granary or a depository (Deemant 1986); in ethnographic farm complexes mostly summer kitchens had circular ground plan. A semi-arch with the diameter of 2-2.75 metres, which probably derived from a similar circular building, was also found at Tornimae--another Viking Age harbour site excavated on Saaremaa (Magi 2005).

Pottery sherds and animal bones as well as single metal artefacts were collected from the territory of the building remains. The metal artefacts included numerous iron nails and rivets, the first of them presumably indicating wooden buildings. There were patches where finds were totally missing. Dozens of boat rivets, nails, rivet fragments and other pieces of iron artefacts as well as an auger were collected from the area of only approximately 6 sq metres north of the above described circular building; an axe had been found already before. At the same spot many bones were gathered, both burned and not burned. The place was interpreted as the waste pile of a carpenter's workshop. The workshop itself might have remained north of the excavation plot, on the place where, judging after the ground, similar building remains continued in the range of dozens of metres.

The remains of the piers

The excavations Nos 2 and 4 directly embraced the former coastline. From the perspective of shore buildings the first of these, No 2, was hardly efficient. The excavation was located on a relatively gentle slope that lacked finds and had apparently not been used for landing. When measuring heights on the ridge with cultural layer it was observed that the gradient towards the bog on the two 6-11 m long sections was considerably steeper, leaving the impression that in these places the former seashore had been artificially dug steeper. Excavation No 4 was established in the southern place of the two.

In the western part of the slope the unearthed stones formed an almost quadrangular construction with straight edges measuring approximately 3.5 x 3.5 metres. Inside the construction, 1.6 metres from the edge of the slope, signs of two posts surrounded by stones were found. A generally similar but a slightly smaller (measuring 2 x 2.6 m) and partly collapsed stone formation was unearthed in the north-eastern part of the excavation, 4.2 metres from the previously described one (Fig. 5). Obvious evidence of posts could not be ascertained here, although this formation was evidently manmade as well. At the foot of the slope, under the erosion layer an approximately one metre wide and 30-40 centimetres high gravel ridge appeared, which crest ran parallel to the slope in the south-west--north-east direction. Such a ridge could have formed in its current location in the course of a bigger storm when the water had already retreated to some extent and the present-day wetland was still a bay in the sea. (1)

The stone constructions in the excavation No 4 can be interpreted as the land-based parts of two piers: the piers supported by posts had been built almost parallel to each other, with the distance of 4.2 metres, transversely to the coast. In order to diminish the erosion caused by waves the posts on the slope had been supported by stones. The length of the piers into the preterit bay remains unknown; it may be suggested, however, that these buildings were repeatedly renovated. Single Viking Age and Late Iron Age pottery sherds and iron nails were gathered between the stones that lined the slope. Although the finds cannot be dated with the accuracy of a century, they indicate the using of the piers simultaneously with human activity in the surrounding area.

[FIGURE 5 OMITTED]

However, there have been more than two piers at the Viltina harbour site. While moving forward along the former coast towards the east, another, 6.5 metres long section of the slope that had been dug deeper started at the distance of 16 metres. It might be suggested that here too we are dealing with the remains of a pier similar to the previous ones. It cannot be excluded that the former shore had even more harbour constructions, although clear aboveground traces of these cannot be detected elsewhere.

Of the harbour sites of Saaremaa information of wooden coastal constructions can be found at the river harbour of Palla that is partly contemporary with the harbour at Viltina. The excavations at Palla took place in 2007, and yielded a platform of planks that rested on posts. The riverbank had been also steepened there at the quay (Magi & Nurk 2008). In the writings of the early 19th century, the possible remains of a quay or a pier were mentioned at Tornimae harbour site. Namely, while ploughing at the foot of a slope that forms the Tornimae harbour site, local peasants had found the remains of a 'wooden palisade' (Luce 1811; Magi 2005).

The surrounding fence and a meeting place

In the north-eastern end of excavation No 4, the edge of the slope or the former seashore was lined with a compound row of stones that resembled the foot of a stone fence. It might be suggested that a wooden barrier had been erected on top of the stone fence. An approximately analogous distinct zone of stones lined the slope in excavation No 2 as well, where all kinds of signs of any other coastal building were absent. An even more clear-cut foot than the previous ones, of a limestone wall or a stone fence, was unearthed in excavation No 6, from the edge of the Viking Age sea bay towards the north. Here the stones had clearly preserved in several layers. Although the edge of the slope was exposed only in single sections, it might be supposed that a similar stone fence and a possible wooden fence on top of it surrounded the whole complex of Viltina harbour site.

The area next to the piers, on the slope partly cut by excavation No 4, lacked stones almost entirely and had a relatively scarce find material. Considering the general abundance of stones in the soil in the surroundings of Viltina an impression was received that the ground here was deliberately cleared of stones. This can be explained by plentiful human activity that probably took place next to the piers and might have been impeded by the stones.

The suggestion was confirmed by excavation No 5 in the vicinity. The practically stoneless layer underneath the tilled soil consisted of hard-tramped earth. The layer contained single potsherds, burnt and unburnt bones, some charcoal and a few fragments of metal artefacts from the 10th-12th centuries. Single bigger stones in natural soil and a few remains of possible, even earlier structures, were unearthed as well. Their meaning, however, remained unclear. A careful conclusion was drawn that the ground cleaned of stones that started already in the area of excavation No 4, continued further, probably comprising roughly the area that used to be surrounded by a stone fence and perhaps served as a small field a few hundreds of years ago. The place was probably a meeting place during the Viking Age and therefore lacked permanent buildings. Since the southern part of the harbour site had been densely covered by buildings, the need was obvious for an open place meant for meetings and all sorts of activities. The area directly next to the piers was the most suitable for this purpose.

Burials

An approximately north-south directed stony elevation that resembled a stone grave is situated on the western border of the Viltina harbour site. In 1999, a set of weapons and other artefacts were found here, which were interpreted as a cenotaph or a collection of additional goods in a grave, whereas human bones could not be detected in the direct vicinity (Magi 2000). During the fieldwork of 2005 the surroundings of this find place were chosen as excavation plot No 3, hoping to ascertain the conditions for the burying of the find assemblage.

A densely packed stone layer with burnt bone fragments, metal finds and pottery that had been exposed to fire was typical of a stone grave with cremations in Saaremaa during the end of the prehistoric period; Rutirank belonged to the same grave type as well. In addition to the previously discussed assemblage of weapons a collection of female jewellery characteristic of Saaremaa in the same period, i.e. the 11th century, was found only 1.5 metres away (Fig. 6). Human bones were missing also in the direct neighbourhood of the jewellery, neither had the artefacts been in fire.

In addition to the stone grave four cremations were detected in the excavated area of the Viltina harbour site. The cremations were located in the built area or directly beside the buildings. It cannot be said for sure whether these people were buried there during the use of the buildings or directly after. Single burials directly inside harbour sites, as well as possibly sacrificed artefacts and sometimes even hoards have been found at the harbour sites of Finland, and Scandinavia as well (e.g. Lundstrom 1981, 117 ff.; Carlsson 1999).

Three individual cremations were found from excavation No 1. All were unearthed between big boulders and were covered by limestone slabs that had been collapsed lengthwise and crosswise. It cannot be excluded that the burials had initially been laid in a certain wooden box covered with limestone slabs that collapsed after the decay of the box. In case of one burial it can be suggested that the remains brought from the pyre had been wrapped in a shawl or an apron as evident from numerous scattered bronze rings--these were mostly used to decorate the mentioned accessories. No other finds were obtained at this burial. The remaining two burials were found from excavation No 1, 1-2 metres south and south-west from the building with a circular ground plan, probably belonged to women, as much as can be determined after the melt metal artefacts consisting mostly of female jewellery (Fig. 7). The grave goods dated the burials to the 10th-11th century.

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

A single cremation alien for Saaremaa was unearthed in excavation No 2, next to the foot of the fence lining the coastal slope. Burnt human bones had been placed into a clay vessel and covered by a flat stone. Cremations in urns have been distributed in many areas; however, in the graves of late prehistoric Saaremaa the bones were always scattered between stones (Magi 2002, 129 ff.).

Discussion

Hill forts, 'towns' and harbour sites

On the map of Viltina from 1874-1875, in relation to the former coastline the gently descending Linnamge pold (Hill Fort Field) lies next to the Viking Age bay. A small farmstead beside the field that has disappeared by now was called Linna (Hill Fort or Town) farmstead (Estonian Historical Archives 3724-5-2946). According to the local folklore gathered from elderly people a real town had been situated once upon a time in the field next to the Viltina Rutirank grave. (2)

On the basis of the excavation results of the Viltina harbour site, it can be stated that the obscure memory of a 'town' could be more feasible than one might guess at first sight. The place was at least partly, but very likely entirely, surrounded by a fence, which encircled small buildings that were erected tightly next to each other. In periods when people gathered there, the Viltina complex might have looked like a small town or a hill fort--that apparently used to mean one and the same thing.

The research of the coastal areas of Saaremaa has indicated a phenomenon that the former more or less definitely localised harbour sites have often been marked with toponyms that refer to a town or a hill fort (see also Magi 2004a, 145 f.). For example the hillock under the Tornimae church, which according to several archaeological excavations used to be a Viking Age harbour site, was in the early 19th century still known as Linnamagi (Hill Fort). During the 19th century the name Linnamagi was replaced by the name Sillamagi (Bridge or Pier Hill), later only the label Tornimagi (Tower Hill) remained (Estonian Historical Archives 2072-3-419; Luce 1811; Holzmayer 1868; Magi 2005). Local people have also named some higher bog islands with steep slopes hill forts; however, traces of human activity have never been found there. A legend of a former town was written down in the 1920s on the southern coast of Saaremaa in the surroundings of the Kogula village (SMM 1924, 88), which, according to the topographic indicators, would be suitable for a harbour site at the end of prehistory. Test excavations there in 2006 proved that a relatively intense human activity had really taken place on the former coast in late prehistory; moreover, according to folk tales boat ribs and an anchor were found from the bog there (Magi 2003; 2007b). In some cases the supposed prehistoric harbour sites have been marked by legends of manors although according to the data of historical sources, a manor had never existed in the given place (e.g. Magi & Magi 2002; Magi 2004b).

The characteristic feature of the prehistoric harbour sites of Saaremaa, namely that they are now on dry land, does not contribute to their being remembered as harbour sites in local tradition. If there was a vague memory in folklore of the one-time significance of the place, it could easily have transformed into stories of a former manor or a hill fort. Especially if a more dominant landscape element existed in the neighbourhood that, in addition, bore a general resemblance to the known hill forts, the meaning of a former power centre was ascribed to it. Place names associated with power structures can be handled as symbols that demonstrate the power on cultural landscape (e.g. Westerdahl 2002). There is no doubt that town, hill fort or manor can be considered the kind of symbols--their actual existence would assume their control over the surrounding landscape and the people using it. Connection to former harbour sites can definitely be observed here, especially with those that had a wider meaning and the range of users than just a farmstead or a hamlet. At certain times a large number of people could have gathered in harbour sites, which would provide a basis for the hazy memory of a former 'town'. In addition, harbour sites were connected with the local elite who controlled these, hence the connection with power centres. Actually, hill forts were situated near bigger harbour sites, ensuring order and security and apparently functioning as political and religious centres (Magi 2007c).

Find material

The comparison of the find material of Viltina with the finds from other similar sites sheds some light on the function of the Viltina site complex. The find material dates Viltina mainly to the 11th-12th centuries, i.e. the end of the Viking Age and the Late Iron Age. On Saaremaa a Viking Age harbour site has been excavated at Tornimae, and a late Iron Age one has been investigated at Palla (Magi 2005; Magi & Nurk 2008). Both Tornimae and Palla were characterized by 30-40 centimetres thick intense cultural layer. At Viltina, dark soil indicating the cultural layer occurred only in places, and most of its area was covered with only a very thin humus layer. The Late Iron Age horizon was uncovered directly under it. Compared to Tornimae and Palla, the number of pottery sherds and animal bones was considerably more modest. The scarce intensity of the cultural layer left the impression that Viltina was inhabited only short-term in certain episodes, at the same time Tornimae and Palla harbour sites were apparently used during longer periods, probably at the time when sea was navigable.

However, the find material of Viltina contained rather more metal finds than Tornimae and Palla as well as common (agricultural) settlement sites. The fords gathered with a metal detector during surface survey trips included weights as well as a small deposit of a fragment of a silver pin and Arab silver coins, which had not been exposed to fire and thus apparently derived from the settlement layer and not the grave. Several small artefacts found at the built area of Viltina, included the metal parts of belts and bridles that might have been accidentally lost. Certain items, especially the collections of artefacts, may also be considered offerings.

Conspicuously numerous were iron rivets and nails or their fragments altogether approximately two hundred, including over a hundred possible boat rivets. The more frequent occurrence of rivets at harbour sites than at common settlement sites has been noted in Scandinavia and, for example, in the excavations at Tornimae in the 1960s. Boat rivets were absent at Palla, which might be explained by the relative smallness of the excavation that contained the prehistoric layer there. Nails at Viltina were more numerous in areas with light wooden buildings: by the building remains in excavation No 1 and near the piers on the slope. Alleged boat rivets were primarily gathered from the built areas, to some extent also from elsewhere.

Meeting place in the ritual landscape

Viltina in the beginning of the Koiguste Peninsula remained at least 15 kilometres away from the more important centres of Saaremaa during the prehistory, as well as later periods (Magi 2007d). This used to be and still is a periphery in the sense of cultural landscape, a considerably limited cluster of agricultural lands that in most parts were lined with bogs. Frozen bogs are easily passable in winter but throughout the majority of the year the communication with other areas of Saaremaa was definitely more complicated by land than by waterways along the coast. The surroundings of Viltina were the typical settlement cluster of maritime culture, which relations with the outer world ran via the naturally favourable harbour sites.

At first glance the barely fertile ground of Viltina agrees neither with the big number of graves nor the abundant luxurious grave goods. The possible fishing harbour is excluded by the total absence of fish bones in a situation where the bone material was otherwise quite well preserved. The peripheral location of Viltina hardly suggests that we might be dealing with a trading place--harbours located in the vicinity of centres, e.g. Tornimae, were significantly more suitable for that. The collection of silver, coins and weights found at Viltina during a survey trip might derive from a lost purse (Fig. 8). These artefacts do not certainly prove trade at the site, but indicate activities that required paying with silver.

Although graves have been found in the vicinity of many Estonian harbour sites, the concentration of burial sites in the surroundings of Viltina is conspicuous. Graves and possible offerings refer to the fact that the place was sacral in nature--apparently graves were not merely used for burying certain people but also for performing various rituals. It is almost impossible to say, which kind of rituals these were--we hardly know anything of the pre-Christian religion of Osilians. The abundance of graves in what seems like a marginal area is characteristic of some other Estonian regions, too--for example Kurevere on Saaremaa Island or Muuksi in northern Estonia. The common trait of these areas seems to be a suitable harbour site that used to be or still is located in the vicinity. When choosing ritual meeting places, apparently the spots accessible by sea were preferred in coastal regions. One might suggest that rituality was also accompanied by secular functions--for example judicial proceedings, joining in agreements and marriages, undoubtedly certain buying and selling transactions. Which of these is primary in this case, the ritual or the practical side, is impossible to establish. Probably both sides were represented more or less equally. In any case the sacral function of the Viltina harbour site offers a possible explanation to the abandoning of the place in the 12th century or perhaps at the beginning of the 13th century--the society changed more than the coastline and the rituals of the baptized Osilians generally took place in churches and churchyards. Even if the place was still used during some time, it lost its earlier importance and was soon entirely deserted.

[FIGURE 8 OMITTED]

The most likely interpretation of the Viltina harbour site is that we are dealing with a specific place for the assemblages (karajad) in the end of the prehistoric period, the Estonian equivalent for Scandinavian thing-places. It is natural to presume that in the area so tightly connected with the sea as the Island of Saaremaa, people gathered to the assemblages by sea and not by land. As elsewhere in Nordic countries the gathering places were densely intertwined with conducting rituals there as well. A certain type of people were buried into the graves in the surroundings of Viltina or directly at the harbour site. Who were these people or how they deserved the honour, must however remain a secret. On the basis of the burials excavated so far it can only be stated that both men and women were buried at Viltina, which suggests that the representatives of both sexes attended the rituals and perhaps the assemblages as well.

Conclusion

One can visualize the Viltina site as an area surrounded by a stone fence and a wooden barrier by the cove where small light buildings stood close to one another and at least three piers supported by poles extended into the sea. There was an open area by the piers where the ground had been heavily trampled. People used to gather at Viltina only occasionally, perhaps a few times a year, and then the place resembled a small town enclosed by a fence. These meetings were probably participated not only by the inhabitants of the nearby villages but the residents of a wider area.

Viltina has probably been one of the places that played an important role in the social life of pre-Christian Osilians. It would be difficult to believe that we were dealing with a single one of its kind in the Viking and Late Iron Age Saaremaa, let alone the rest of Estonian coast. A roughly similar concentration of graves a little away from the agricultural lands, but at the same time in the vicinity of a former naturally suitable landing place, can be found in Kurevere on Saaremaa and Mala on the Island of Muhu. It is possible that in these places karajad or ritual gatherings took place as well. In any case the set of sites excavated at Viltina refers to the symbiosis of the sacral cultural landscape with the profane one. On the one hand, the ritual coastal landscape can be seen as one of the main causes for establishing the harbour site, on the other, the place apparently became ritual due to the favourable access from the sea.

Resumee

Artiklis on lahema vaatluse alla voetud Viltina sadamakoht Louna-Saaremaal, kus aastatel 2004-2006 toimusid komplekssed arheoloogilised kaevamised. Leitud muinasjaanused sobituvad muinasaja lopu kontekstis suhteliselt tavaparasesse sadamakohta, mille olulisus vois olla eelkoige piirkondlik voi ka naaberpiirkondi haarav, kuid mitte regionaalne voi rahvusvaheline. Esmapilgul voiks eeldada, et samasuguseid kohti leidus 11.-12. sajandi Eesti rannikul rohkesti. Muististe kompleks Viltinas osutus aga seni teadaolevatest erinevaks just oma ulesehituse poolest, voimaldades vaita, et tegu oli eelkoige rituaalse kogunemiskoha, mitte kauba-, kala- voi muu praktilisema tahendusega sadamakohaga. Sellisel kujul on Viltina esimene omataoline arheoloogiliselt uuritud objekt Eestis, mis laiendab meie teadmisi kunagise merendusliku kultuurmaastiku ulesehitusest ja uhiskonna erinevatest sotsiaalsetest tahkudest.

Eeldatavalt vahemalt piirkondliku tahtsusega sadamakoha jaoks on ebatavaline juba Viltina asukoht: nii muinasaja lopu kui ka tanapaevastest keskustest eemal. Vaikekula tekkis Viltinasse alles 18. sajandil, enne seda kuulusid sealsed uksiktalud Audla moisale. Kuna koik ulejaanud Koiguste poolsaare talud kuulusid kohalikele moisatele, voib Audla moisa eraldi seisvat rannamaavaldust pidada kohaliku kultuurmaastiku seisukohalt veidi ebatavaliseks jooneks. Voimalik, et siin oli tegemist jaanukiga ajast, mil Viltina umbruskond oli veel sadamana oluline ja seda haldas Audla kui umbruskonna vanim ning mojukaim mois.

Viltina kula pollumaa on kivine ja vaheviljakas, veidi parem on selle kvaliteet Randvere ning Asva kulas, mis jaavad Viltinast eri suundades 1-2 km kaugusele (joon 1). Randveres on asustusuksus paiknenud hiljemalt alates meie ajaarvamise esimestest sajanditest, mis ajast on kula lahistelt teada kivikalme. Asva kula poldudel puuduvad varast asustusuksust markeerivad kivikalmed, kull aga on kalmeid teada Kahtla ja Laimjala kula maadelt. Pronksiajal paiknes Asvas kindlustatud asula, mille kohale ehitati eelviikingiajal linnus. See oli kasutuses veel viikingiaja algul.

Viltina umbruskond on silmatorkav sealsete rohkete kivikalmete poolest, millest tuntuim on Viltina Rutirank. 1940. aastal kaevati see kalme koos mitme laheduses paikneva uksikkalmega peaaegu taielikult labi. Tegu oli Saaremaale tavaparaste poletusmatustega kivikalmetega, milles leidus aga silmatorkavalt palju luksuslikke esemeid, sealhulgas relvi ja ratsanikuvarustust ning importkeraamikat. Umbruskonna vaheviljakad pollumaad ja perifeerne asukoht olid ilmselges vastuolus kalmete ning neisse asetatud panuste rikkusega. Kalmete paiknemine 11. sajandi mererannal kohas, kus tuulte eest kaitstud sugava veega lahesopp voimaldas ohutut ja mugavat randumist, viis motted voimalikule sadamakohale. Viimane leitigi 1999. aastal koos veel uhe kivikalmega.

Puudes rekonstrueerida kultuurmaastiku arengut, voib oletada, et varaseim Viltina sadamakoha eelkaija jai kaevatud sadamakohast umbes 1,2 km kaugusele Asva pronksiaegse kindlustatud asula juurde. Leitud seda pole, kuid uurimistoo mujal Pohjamaades on naidanud, et meie ajaarvamise 5.-6. sajandist varasemaid sadamakohti ongi aarmiselt raske lokaliseerida, nahtavasti arheoloogiliste jaanuste vahese leidumisega nendes. Seoses maatousuga nihkus sadamakoht sajandite jooksul ilmselt hilisemale Viltina kulale lahemale ja viikingiaja algul vois see paikneda praeguste Kao-Matsi talude kohal 11.-12. sajandi sadama korvale jaava kunagise lahesopi aares.

Kuhu nihkus sadamakoht 13. sajandil, pole jallegi teada. Jarkjarguline nihkumine on jalgitav vaga paljudes ajaloolistes sadamakohtades. Keskaegne kohalik sadam vois ehk paikneda Viltina uksiktalude lahikonnas alal, mis veel 17. sajandil kuulus Audla moisale. 17. sajandi lopul paiknes Viltina sadama jarglane igal juhul juba peaaegu kaks kilomeetrit louna pool, Ruhves, kuhu selleks ajaks oli tekkinud ka rannakula.

1999. aastal inspektsiooniga leitud 11.-12. sajandi sadamakoht Viltinas jaab Rutiranga kalmett 50 m kirde poole. Aastatel 2004-2006 kaevati seal kuues eraldi kaevandis hibi 330 [m.sup.2] (joon 2).

Lounapoolseimas, koige suuremas kaevandis leiti vaikeste kergete hoonete jaanused, mis olid ehitatud tihedalt uksteise korvale (joon 3). Need hooned olid enamasti olnud ehitatud ilmselt rohtpalktehnikas ja neist olid sailinud vaid alumiste palkide alla hikatud kividest voondid ning moned arvatavad postikohad, kui postid olid umbritsetud kividega. Uks hoone oli ummarguse pohiplaaniga ja 2,8 m sisemise labimooduga; hoone seest leiti tulease. Uhtki korralikku ahju ei leitud, mis viitab arvatavasti sellele, et vaikesed puithooned olid moeldud kasutamiseks vaid soojemal aastaajal. Uhest kohast tuli valja ajutise iseloomuga U-kujuline kolle ja teist kunagist tuleaset markeerisid laiali aetud tugevasti polenud raudkivid.

Hoonejaanuste alalt saadi savinoukilde ja loomaluid, sekka uksikuid metallesemeid. Viimaste hulgas oli rohkesti raudnaelu ja neete, millest eriti esimesed viitavad samuti puithoonestusele. Kohati oli laike, kus leiud taielikult puudusid.

Otseselt sadamakohast annavad tunnistust kahe paadisilla alguse jaanused omaaegsel rannikul. Need paiknesid kohas, kus kallas oli ilmselgelt jarsemaks kaevatud. Kaks postidele tuginevat paadisilda olid ehitatud paralleelselt, teineteisest veidi enam kui 4 m kaugusele ja tugevdatud nolvakul kividega (joon 4). Kui pikalt paadisillad omaaegsesse merre ulatusid, polnud kaevamistega voimalik kindlaks teha. Teine jarsemaks kaevatud loik kunagisel kaldal osutab igal juhul sellele, et sadamakoha kasutusajal oli seal paadisildu ilmselt rohkemgi.

Otse paadisildade korvale jai toenaoliselt kividest puhastatud ala, kust saadi uksikuid muinasaja lopu savinoukilde ja moned raudnaelad. Sinna rajati ka proovikaevand, kust saadi moned 11.-12. sajandi esemete fragmendid kovaks trambitud ja kividest puhastatud pinnasest. Ihnselt oli siin tegu kogunemiskohaga, kuhu hooneid ei pustitatud.

Mitmelt poolt kaevanditest tuli valja madal kiviaia tald piki omaaegset randa. Voib oletada, et kasutusajal oli kogu kompleks umbritsetud kiviaiale tugineva voi kiviaiaga aaristatud puittaraga.

Uks Viltina kaevanditest rajati koos sadamakohaga leitud kivikalmesse, kivisesse korgendikku otse rannaaarse sadamakoha taga. Siin tuli valja Saaremaale muinasaja lopul tuupiline poletusmatustega kalme, kust muuhulgas leiti kalmesse toenaoliselt peidetud saaremaaparaste naisteehete kogum 11. sajandist, mis polnud tules olnud ja mille vahetus laheduses puudusid inimluud (joon 5). Sellest vaid 1,5 m eemal leiti 1999. aastal samaaegne poletamata relvakogum.

Esmapilgul ootamatud olid hoonestatud alalt leitud uksikud poletusmatused. Suski on uksikute matuste esinemine otse viikingiaegse sadamakoha sees iseloomulik ka mitmetele Soome ja Skandinaavia vastavatele muististele. Viltina hoonete seest tulid valja kaks esemete jargi naistele kuulunud selgepiirilist individuaalset poletusmatust 10. voi 11. sajandist (joon 7). Veel uks poletusmatus jai panuste puudumise tottu ebaselgeks, kuigi rohkearvulised pronksrongakesed selles viitavad kaudselt samuti sellele, et surnu vois olla naine. Uks Saaremaa oludes vooraparane poletusmatus tuli valja kaevandist nr 2, rannanolvakut aaristava tara aluse korvalt. Polenud inimluud olid pandud savinousse ja kaetud lameda kiviga.

Vorreldes Viltina leiumaterjali teiste Saaremaal kaevatud umbes samaaegsete sadamakohtadega (Tornimae, Palla), torkab silma, et sealne kultuurkiht oli ulejaanutega vorreldes ohuke ja vaheintensiivne, sisaldades aga keskmisest rohkem metallesemeid, hobeasju, kaaluvihte ning eriti paadineete. Taielikult puudusid Viltinas kalaluud, kuigi luumaterjal oli uldiselt hasti sailinud, viidates seega sellele, et kalasadamaga ilmselt tegu ei olnud. Kaubasadama funktsiooni valistab asukoht perifeersel alal, koigist keskustest eemal.

Viltina leiumaterjal ja uldine iseloom viitavad seega kohale, mida kasutati hooajaliselt ning vaid luhikeste perioodide valtel, kuid siin viidi labi toiminguid, mis eeldavad pidulikku riietust (voo- ja ratsmete osad) ning hobedaga maksmist (joon 8). Osa inimestest ka maeti vahetult sadamakoha korvale voi lausa sadamakoha sisse. Kalmete suur hulk ja rikkalikud panused lubavad oletada, et need inimesed ei parinenud mitte uksnes lahedal asuvatest voimalikest taludest, vaid toenaolisemalt nendest lahikonna eliidiperedest, kel oli siinse sadamakoha ule teatud kontroll. Voimalik, et siia maeti ka inimesi kaugematelt aladelt.

Arvatakse, et kalmetega olid muinasajal seotud ka rituaalid. Kalmete kontsentratsioon just Viltina sadamakoha umbruses viitab seega toenaolistele rituaalidele, mis siin toimusid ja kuhu koguneti meritsi. Osa siinsest leiuainesest vois kujutada enesest ohverdusi; eriti kehtib see kivikalmest leitud kahe leiukogumi koha.

Koht oli ilmselt traditsiooniliseks kogunemiskohaks, nagu osutavad kergete hoonete jaanused ja umbritsev tara, st koht oli ehituslikult organiseeritud. Ilmselt voeti hooned kasutusele vaid luhikestel perioodidel, mil siia koguneti riitusi labi viima, aga toenaoliselt ka labiraakimisi ja nou pidama, lepinguid solmima ning voib-olla ka kohut moistma. Tegu on seega kohaga, mida voiks nimetada karajakohaks, Skandinaavia tingi-kohtade Eesti vasteks. On ootusparane, et merendusliku tegevusega seotud aladel, nagu Saaremaa rannik seda kahtlemata on, tekkisid sellised kohad meritsi kergesti ligipaasetavatele aladele. Koha vahemalt osalt sakraalne iseloom pakub voimaliku seletuse ka selle mahajatmisele 12. sajandil voi veidi hiljem: lisaks jatkuvale maatousule olid saarlased votnud vastu ristiusu ja uued uldtunnustatud kogunemis- ning riitusekohad jaid ilmselt kirikute ja kabelite lahikonda.

doi: 10.3176/arch.2009.2.03

Acknowledgements

Archaeological excavations as well as the writing of this article were conducted in the framework of grant No. 6998 of the Estonian Science Foundation. The author is grateful to Kristiina Johanson for the translating of the paper.

References

Callmer, J. 1991. Platser med anknytning till handel och hantverk i yngre jarnalder. Exempel fran sodra Sverige.--Hovdingesamfund og Kongemagt. Fra Stamme tit Stat i Danmark, 2. Eds P. Mortensen & B. M. Rasmussen. (Jysk Arkaeologisk Selskabs Skrifter, XXII: 2.) Aarhus universitetsforlag, Arhus, 29-47.

Carlsson, D. 1992. Harbours and trading places on Gotland AD 600-1000.--Aspects of Maritime Scandinavia AD 200-1200. Proceedings of the Nordic Seminar on Maritime Aspects of Archaeology, Roskilde, 13th-15th March, 1989. Ed. O. Crumlin-Pedersen. The Viking Ship Museum, Roskilde, 145-158.

Carlsson, D. 1999. "Ridanas". Vikingahamnen i Frojel. (ArkeoDok Skrifter nr 2; CCC Papers, 2.) ArkeoDok, Visby.

Christoffersen, J. & Porsmose, E. 1996. Den fynske kystzones bebyggelsesmonstre i jernalder, vikingetid og middelalder.--Atlas over Fyns kyst i jernalder, vikingetid og middelalder. Eds O. Crumlin-Pedersen, E. Porsmore & H. Thrane. Odense universitetsvorlag, Odense, 154-160.

Crumlin-Pedersen, O. 1991. Maritime aspects of the archaeology of Roman and Migration-Period Denmark.--Aspects of Maritime Scandinavia AD 200-1200. Proceedings of the Nordic Seminar on Maritime Aspects of Archaeology, Roskilde, 13th-15th March, 1989. Ed. O. Crumlin-Pedersen. The Viking Ship Museum, Roskilde, 41-54.

Crumlin-Pedersen, O. 1996. Kystforsvaret.--Atlas over Fyns kyst i jernalder, vikingetid og middelalder. Eds O. Crumlin-Pedersen, E. Porsmore & H. Thrane. Odense universitetsvorlag, Odense, 182-193.

Deemant, K, 1986. Der Siedlungsplatz von Proosa.--Eesti NSV Teaduste Akadeemia Toimetised. Uhiskonnateadused, 35: 4, 398-400.

Holzmayer, J. B. 1868. Osiliana, I. Das Kriegswesen der alten Oeseler. Arensburg.

Indreko, R. 1939. Asva linnus-asula.--Muistse Eesti linnused. 1936.-1938. a. uurimiste tulemused. Opetatud Eesti Selts, Tartu, 17-52.

Lavi, A. 2005. An addendum to the study of smoke cottages.--EJA, 9: 2, 132-155.

Lougas, V. & Selirand, J. 1989. Arheoloogiga Eestimaa teedel. Valgus, Tallinn.

Luce, J. W. L. 1811. Das Schloss Mone auf Osel, oder nahere Bestimmung der Lage derjenigen Festung, auf deren Einnahme durch die Deutschen 1225 die Unterwerfung der ganzen Inset folgte. Einladungsschrift der Arensburgischen Kreisschule, Riga.

Lundstrom, P. 1981. De kommo vida ... Vikingars hamn vid Paviken pa Gotland. Uddevalla.

Magi, M. 2000. Weapons find and an ancient harbour site at Viltina Kao-Matsi.--AVE, 1999, 92-99.

Magi, M. 2002. At the Crossroads of Space and Time. Graves, Changing Society and Ideology on Saaremaa (Osel), 9th-13th Centuries AD. (CCC papers, 6.) Gotland University College Centre for Baltic Studies, Institute of History, Tallinn.

Magi, M. 2003. Inspektsioonid Kogula kulla Karla kihelkonnas Saaremaal. Tallinn. Manuscript in the Topographic Archives of the Institute of History, Tallinn University.

Magi, M. 2004a. "... Ships are their main strength." Harbour sites, arable lands and chieftains on Saaremaa.--EJA, 8:2, 128-162.

Magi, M. 2004b. Inspektsioonid Kurevere ja Tammese kulla. Tallinn. Manuscript in the Topographic Archives of the Institute of History, Tallinn University.

Magi, M. 2005. Viking Age harbour site at Tomimae, eastern Saaremaa.--AVE, 2004, 65-75.

Magi, M. 2006. Excavations on the coasts of prehistoric and medieval Saaremaa.--AVE, 2005, 65-82.

Magi, M. 2007a. Archaeological excavations at Viltina--a Viking Age harbour site and meeting place.--AVE, 2006, 121-134.

Magi, M. 2007b. Aruanne proovikaevamistest Kogula asulakohal Karla khk Saaremaal, 11.-14. juulil 2006. Tallinn. Manuscript in the Topographic Archives of the Institute of History, Tallinn University.

Magi, M. 2007c. Im linnusest Puha Birgitta kloostrini. Merenduslik kultuurmaastik Pirita joe alamjooksul.--Studies on Art and Architecture, 4, 17-40.

Magi, M. 2007d. Two paths towards Europe: Iruste and Viltina, Saaremaa.--The Significant Detail: Europanization at the Base of Society: the Case of the Baltic Rim 1100-1400 AD. (CCC papers, 9.) Gotland University, Visby, 216-225.

Magi, M. & Magi, T. 2002. Archaeological fieldwork around Tonija and Roosa on southern Saaremaa.--AVE, 2001, 56-64.

Magi, M. & Nurk, R. 2008. An archaeological complex of river harbour and 16th-17th century stone buildings at Pallamois on the southern coast of Saaremaa.--AVE, 2007, 91-106.

Mellin, L. A. 1798. Atlas von Liefland and Ehstland, oder von den beyden Gouvernements and Herzogthumern Lief- and Ehstland and der Provinz Oesel. Riga.

Nasman, U. 1991. Sea trade during the Scandinavian Iron Age: its character, commodities, and routes.--Aspects of Maritime Scandinavia AD 200-1200. Proceedings of the Nordic Seminar on Maritime Aspects of Archaeology. Roskilde, 13th-15th March 1989. Ed. O. Crumlin-Pedersen. The Viking Ship Museum, Roskilde, 23-40.

Mink, G. 1939. Saaremaa taluehitised. Etnograafiline uurimus, I. Ehitiste uksikosad, elamu ja korvalhooned, mis osalt taidavad elamu ulesandeid. Opetatud Eesti Selts, Tartu.

SMM 1924--Saaremaa ja Muhu muinasjaanused. (Tartu Ulikooli Arkeoloogia Kabineti Toimetused, II.) Odamees, Tartu, 1924.

Tihase, K. 1974. Eesti talurahvaarhitektuur. Kunst, Tallinn.

Ulriksen, J. 1998. Anlobspladser. Besejling og bebyggelse i Danmark mellem 200 og 1100 e. Kr. En studie af sofartens pladser pa baggrund af undersogelser i Roskilde Fjord. Vikingeskibshallen, Roskilde.

Vassar, A. 1940. Kaevamisaruanne Poide khk. Laimjala vl. Viltina kl. Peerna tl. "Rutirangal" 13. VI--27. VII 1940. a. Tallinn. Manuscript in the Topographic Archives of the Institute of History, Tallinn University.

Westerdahl, C. 1989. Norrlandsleden: The Norrland sailing route, 1. Kallor till det maritima kulturlandskapet: en handbok i marinarkeologisk inventering. Harnosand Lansmuseet Murberget, Harnosand.

Westerdahl, C. 2002. The cognitive landscape of naval warfare and defence. Toponymic and archaeological aspects.--Maritime Warfare in Northern Europe. Technology, Organisation, Logistics and Administration 500 BC--1500 AD. Papers from an International Research Seminar at the Danish National Museum, Copenhagen, 3-5 May 2000. Eds A. Norgaard Jorgensen et at. National Museum, Copenhagen, 169-190.

(1) At this point we would like to thank Hannes Tonisson, a researcher at the Institute of Ecology of the Tallinn University, who helped us with valuable advice in the questions regarding beach formation.

(2) Told by the locals in the summer of 2006.

Marika Magi, Institute of History, Tallinn University, 6 Ruutli St., 10130 Tallinn, Estonia; Marika.Magi@mail.ee