This article aims to discuss a burial site in eastern Estonia in a way that has not been done before. Namely, the purpose is to look at previously known archaeological information about the burial site from the perspective of continuation in the landscape. The site in question was excavated quite thoroughly in the 1970s (Lavi 1977, 1978a, 1978b), and further research about it has been conducted and published by the author of the present article in the 2000s (Karro 2010a, 2010b, 2012).

The history of theoretical landscape research on other sites is new to Estonian archaeology. However, in other European countries it has been more widely discussed and some of this literature has been used in the theoretical framework of this article. In Estonia, mostly North-Estonian archaeological landscapes have been studied (e.g. Lang 1996; Vedru 2001, 2002, 2009, 2011, 2013a), and an overview of Estonian settlement and landscape archaeology has also been provided (Lang and Laneman 2006). In addition to North-Estonia, the settlement of Saaremaa has also been researched (e.g. Mägi 2002a, 2008). However, most of the landscape research has been done from the viewpoint of settlement archaeology, and not so much from landscape archaeology (Lang 1996, Mägi 2002a). Some examples from the latter are Gurly Vedru’s works (2009, 2011, 2013a, 2013b). Thus, this article also aims to discuss aspects in the lives of people in the past through more phenomenological notions like memory and narration.

However, these concepts have been dealt with by human geographers in Estonia (e.g. Palang 2001), and there has also been some co-operation with archaeologists (e.g. Palang et al 2005). Continuation is the main theoretical conception used in the discussion. However, there are several other notions that will be discussed in the context of continuation, for they form an essential part of continuation itself: memory, narration/stories. Some attention is also paid to the researcher’s perspective as to why the landscapes are being studied and described as they are, but this merely serves the function of setting the context and justifying the choice of topic.

In conclusion, the article aims to study one of Estonia’s landscapes in a way that has previously been practiced very little in Estonia, and seeks to understand why one place may have been in use for 1500 years.

Landscapes are continual spaces – never finished, but the result of processes and practices (Pred 1984). This type of (contemporary) thinking mostly emerged in the 1970s and 1980s, for before this geographers,historians and also archaeologists, dealt more with single objects and places, and not so much with continuity (Baker 2003). However, it began to be felt that archaeologists should actually not discuss single objects so much as landscapes as a whole. Hans Gumbrecht’s late concept of change of chronotopes in the perception of history after World War II is an appropriate place to start the present discussion. A ‘chronotope’ is the social construction of temporality. While the old chronotope considered the past as something that had to be left behind, the new chronotope suggests that the past has settled in the present, or in other words, presence is inundated by ‘pastness’ (Gumbrecht 2013). Driven by this idea, archaeological landscapes can also be seen as spaces that are inundated by pastness, and that this pastness is carried by archaeological objects/monuments and artefacts. In other words, the present is always affected by the past, because there is always something left from the past in the present landscape, and it is this that archaeologists study. The most difficult part of archaeological research is to set what remains from the past into the context of processes and development, or as Chris Gosden and Gary Lock (1998, 4) have stated: “For the archaeologists, sites are static entities, to be classified into land boundaries, burial monuments, hillforts and so on. We arrive many millennia later when the heat and urgency of daily life has cooled and cast a retrospective view over the landscape”.

So, continuation in landscapes should be investigated, because this “retrospective” view is lacking this. By understanding the concept of continuation, and some related concepts that will be discussed below, this can be to some extent achieved.

The meaning of landscape in this article should be explained, for it has many definitions. Landscapes are understood not only as natural and/or cultural, but as a system where natural, cognitive and temporal components are connected (Palang 2001). Landscape does not exist outside of the human mind (Vedru 2002), but the human mind saves what it has had contact with - this (in a very broad sense) can be called memory, and memory is a vital aspect in the continuation of anything connected with or in the human mind.

But landscape can also be explained as a network of places connected by paths, roads and stories (Tilley 1994). Network is one of the key words in this definition – different landscapes form networks, because all landscapes (geographically, temporally or perceivably distanced) form one unity to a certain extent, and it is only possible to study parts of it more closely. The landscape discussed below is also only a part of this network, for it is connected to other landscapes. Landscapes also consist of different layers, which may each form connections with different kinds of landscape. One of those layers is the ‘mental layer’ (for the layers of landscape see Karro 2010a) which enables connections to be made with distant physical landscapes, and thus make several geographically distinct landscapes continuous in relation to one another. This kind of continuation does not only appear in space, but also in time, and this idea will be argued for in this article using an Iron Age burial place in eastern Estonia as a case study.

Time is closely connected to the concept of memory, because remembering is one aspect in making places meaningful – after all, it can be argued that this is the very way that locations are turned into places (Cresswell 2004). This meaningfulness is often mostly concerned with local people and their memories of, and roots in, a place (Hernandés et al 2007) and its monuments and natural features (Van Dyke and Alcock 2003). This is the process by which place identity forms and a place becomes a bearer of memory (Vedru and Karro 2012).

Although Gumbrecht (2013) has stated that memory is an artefact from the past, it can instead be argued that artefacts and archaeological sites such as burial places facilitate memory. Landscape can also be defined as the materialisation of memory, or the fixing of social and individual histories in time. As human memory constructs rather than retrieves, the past therefore originates from cultural memory, which is itself socially constructed (Ashmore and Knapp 2000). In this way landscape can become a collective narrative about the people living there.

Narrative is considered to be a spoken or written account of connected events (a story) or, in other words, a practice or an art of telling stories. Stories usually have a continual aspect, and can re-enact memories, as the citation at the beginning of this article illustrates, and can also be mediators of a far-away past that does not exist anymore. However, in a narrative the past can still be ‘hot’, ‘urgent’, and real, while only fractions of this past reality may be physically extant. In other words, memory is borne by artefacts, sites, objects – fragments of the past. This is basically what happens to archaeological landscapes – some components are missing, but the still existing fragments allow the story of the past to be narrated. So, while the story is the mediator or the tool that creates continuity between the past and the present, we are using information from the memory borne by objects, or in other words, the small stories narrated by single objects to retell the past.

While dealing with archaeological or historical objects, the people who made and interacted with them should not be forgotten. Even if it is possible to say that sites and objects tell stories, it is actually people who narrate them. It is thus people who give meaning to artefacts and sites, while objects and natural features help people secure their memories (Van Dyke and Alcock 2003). It is archaeologists and historians who use those memory laden objects to mediate between the people of the past and the present. In other words, they communicate with the people of the past through landscape, and this is what is meant by the continuous aspect of landscape. Furthermore, the life of past people also becomes continuous through this continuation of the landscape, and the reflection of this idea can be seen in burial sites – in places where human lives have become continuous through material manifestation.

The example discussed in this article is a cemetery in Lahepera village in eastern Estonia (see the location in Fig. 1). The cemetery is actually part of a former Estonian church parish, dating back to at least the 15^th century (for the first historical records of the church see Ederma and Jaik 1939), but it was probably formed even earlier like most Estonian church parishes.2 The physical formation of parishes in Estonia has been dated to the Latest Iron Age (1050 – 1227 AD)3, although the borders changed during the reigns of several foreign conquerors. It has been assumed that in addition to natural borders a parish was also formed on the basis of kin lines, and played the role of political, economic, and administrative unit. Ancient parishes also formed counties, but the functions of this system of counties differed considerably from the present counties of Estonia (for Kodavere parish see Karro 2010a; 2012; for ancient Estonian administrative system see Lang 2007a, 273–277).

It is not clear whether Kodavere parish was a separate county by the name of Soopoolitse, or a parish conglomerated into the larger Vaiga county in the Late Iron Age (Lang 2007a, 275). In Russian chronicles Kodavere has also been marked by the name of Subolitch (Roslavlev and Salo 2007). However, Kodavere parish area is naturally secluded from surrounding areas by the Great Emajõgi River and its bogs in the south, the Omedu River in the north, bogs and forests of central Estonia in the west, and Lake Peipsi in the east. It is also the area within the present borders of Estonia, where agricultural soils reach closest to the lake, which enabled the emergence of an agricultural settlement there from the last centuries of the Pre-Roman Iron Age (Aun 1974; Karro 2010a). The burial place in Lahepera village is about 10 km from this earliest site in the present village of Peatskivi and it was probably later connected to the hillfort which was established at the place of the early hilltop settlement (when exactly, is unclear). This hillfort remained the central place of the parish until a parish church was established in another village – the village of Kodavere. However, the centre in Peatskivi moved closer to Lake Peipsi (to Alatskivi) in the Medieval period, and a mansion centre was formed there (Karro 2012). The village of Lahepera, where the burial place is situated, is a neighbouring village to Alatskivi, but is situated on the shore of a small lake (Lake Lahepera) which used to be a bay of Lake Peipsi (Mäemets 1977), and is thus very close to Lake Peipsi.

The burial place in question is situated at the present village of Lahepera at a place on the bank of the small lake (Lake Lahepera) where the ground is high and not very boggy. The knoll where the burial place is located is the highest point in the area and is eye-catching from ground-level (Karro 2010b, 2012).

In the 2^nd or 3^rd century AD a stone grave (probably a tarand-grave, see e.g. Lang 2007a; 2007b) was established at the peak of the knoll. It is possible that the burial place is even older, for one of the finds, that was recently re-discovered from archaeological collections by the author, is an iron shepherd’s stick shaped pin4, of a kind which were mostly worn and deposited in Estonian graves in the Pre-Roman Iron Age (Lang 2007b). However, other finds suggest the erection of the grave in the first centuries AD.

About 85% of the grave was excavated in 1977–1978, when only evidence of cremations was found (Lavi 1977; 1978a). The site had been formerly excavated by 19^th century hobby-archaeologists, and there are several finds archived in the University of Tartu archaeological collections, but those finds come without a report which would connect them to certain areas in the grave. The above mentioned pin is also one of those finds, therefore it is not clear whether there might have been an earlier cemetery with pit graves under the later stone grave. Such cremations in pits are very characteristic to southern and southeastern parts of Estonia in the Bronze and Early Iron Ages (Lillak 2009).

The stone grave lacks Middle Iron Age deposits. There are only some artefacts that may date back to the Pre-Viking Age. However, the grave was intensively used again in the Viking Age, when cremations with intentionally broken artefacts (mostly jewellery, but also some burnt weapons) were undertaken (Karro 2008). This is very common to Estonian Viking Age burials (see e.g. Mägi 2002b).

It seems that in the second half of the11^th century, or perhaps even earlier, a shift in burial customs took place - the people started to make inhumations in ground pits to the east of the stone grave (Lavi 1978b). This kind of shift is considered to be a result of Christian influence (see e.g. Mägi 2002b; Valk in press), but further discussion of that issue is not the topic of this article. When village cemeteries started to emerge in southern Estonia, the same site was taken into use for this purpose (for village/rural cemeteries on southern Estonia see Valk 2001). The latest burial in the cemetery can be dated back to the 16^th century (Yurina 2011), but many of the excavated burials are without grave goods and the bones have not been carbon dated, so it is unclear when they were buried.

At present there are fields on the northern bank of Lake Lahepera, but those lands were probably drained in the 1930s and 1960s-1970s because, according to a 17^th century map (Anonymous 1684), there were only very small fields in that area. Bigger fields appear some kilometers to the west – further from lakes Lahepera and Peipsi. The houses on the 17^th century map have formed a fishing village. Estonian stone graves are very often connected to fields, but also with roads or harbour sites (Mägi 2004), and the one in question seems to be of the latter type.

Usually, landscapes cannot be discussed using one site or object, but in the case of the Lahepera-Peatskivi area, this site seemed to have been a very important one. The importance of the landscape of Lahepera is definitely a social construction, however, the fact that it is a burial site provides some possibilities for discussing personal aspects of this landscape as well.

Social construction of a landscape can also be expressed through the definition that human-made and human-perceived objects in the landscape express collective social structure. This kind of social structure is passed on by collective memory, which means that social structure does not only have a collective aspect but is also continual. Artefacts and objects that reflect this structure often stay untouched in the landscape, even after the society that created them has gone. But those artefacts and objects still carry the collective memory of this society, and thus tell a story of that society. So, according to that archaeological landscapes can narrate the story, or at least a part of the story of the past.

The long-term usage period of the burial place in Lahepera reflects that the site was remembered for a long time, perhaps for as long as one-and-a-half thousand years, suggesting the long-term continuation of the collective memory. Evidence suggests that Estonian society changed many times in prehistory (see e.g. Mägi 2002b; Lang 2007b), and this collective memory may have outlived several of these changes.

Where the Early Iron Age is concerned, the human settlement in Kodavere parish was probably spread out because of the different settlement logic of the fisher-hunter and agricultural societies of the time. It is also quite probable that in the Pre-Roman Iron Age, fishing and hunting were still quite important, for the soils are quite heavy and not very easy to cultivate in eastern Estonia. This is said to have caused the spread of crop cultivation to inland Estonia later than in North-Estonian limestone-based soils (Kihno and Valk 1999; Lang 1999), sometime around the year 0. Stone graves were probably a part of this agriculture-based culture. The 2^nd century AD, or maybe some centuries earlier (if the hypothesis of cremations in ground pits under the stone grave is correct) seems the most likely date for construction of the stone grave at Lahepera (see above).

Land cultivation is a field of activity that needs former experience. Of course, hunting and fishing also requires some previous knowledge from the older people of the society, but land cultivation is an activity that makes people settle, so the dead also stay with them when they have been deposited in a permanent place. Slash-and-burn agriculture causes the people to move around to some extent, but it is still connected to arable soils, and it seems that the overall area of arable land was not that large in Kodavere parish during this early period. So, it is probable that the centre for people who were mainly engaged in that kind of activity was in the area of Peatskivi which has the best soils of the parish. The continuation of settlement in that part of the parish, based on soils and agricultural activity, can be dated from at least the Early Iron Age on this basis. The natural advantages of a landscape thus affect the social construction of the landscape, and collective memory makes such landscapes continual.

This centre of settlement is connected to the burial site in question, but the burial site itself also provides possibilities for discussion where continuation of landscape is concerned. A burial site is a place where a society buries their dead, and the traditions of this activity are said to reflect the social structure of the society (e.g. Mägi 2002b). Thus, burials can be considered a significant aspect of social activity. Cemeteries are, of course, connected to permanent settlements associated with arable land, and this is also the case where Lahepera-Peatskivi is concerned. But land cultivation was probably not the only activity these people conducted, because of surrounding natural advantages that provided possibilities for other kinds of activities, such as fishing (lakes Peipsi and Lahepera). Activities connected to a water body, however, require places of embarkation and disembarkation, or, in other words, a harbour site. It has been discovered in Scandinavia, but also in the Estonian island of Saaremaa, that harbour sites in the Iron Age were often marked by close proximity to a burial site (e.g. Karro, 2012; Mägi 2004, 2008, 2010, in press).

It seems very likely that the Lahepera burial site also marked a harbour. Lake Lahepera is a narrow and low lake with a boggy southern and western shore and a high bank on the northern shore, and there is still a connection between this lake and Lake Peipsi. The burial site is on the highest knoll on the northern bank of the lake, and it is quite possible that in an open environment it was possible to notice it when the shore was approached. Of course, the knoll is not very high, but it still catches the eye against quite flat surroundings. The shores of Lake Lahepera experienced constant human activity before 1684, when the above mentioned map was compiled, and it is quite difficult and maybe even impossible to discover through archaeology an ancient site under this activity layer. The soil layer is thin and has been washed constantly by the waters of the lake. The search for a preserved cultural layer from the Iron Age was conducted in autumn 2012 and spring 2013, but it is not yet possible to locate a landing place archaeologically. However, the geomorphological situation and the human geographical reasons described above make it highly probable that such a place existed during the usage period of the burial site.

Of course, it remains debatable how much this postulated landing place was used for boats going fishing to Lake Lahepera or Lake Peipsi, for local transportation, or as a stopping place for trade vessels after Lake Peipsi probably became a waterway from the Finnish gulf to Pskov and other inland areas approachable by rivers that start from Lake Peipsi (for the trade route see Сорокин 1999; Karro 2010a, 2010b, 2012; Mägi in press). The existence of the trade route and its dating is not the topic of this article, but the existence of a landing site near the burial could well indicate the use of Lake Lahepera and Lake Peipsi by the people living on the shore for any of these activities over a long continuum. Although it is unlikely that the site was in use as a harbour in the village cemetery period, some artefacts of foreign5 origin dating back to the Viking Period and Latest Iron Age might suggest that foreigners who had stopped at Lahepera and died there were also buried there, strengthening the likelihood that the area was used as a stopping point on a long distance communication route.

It is also possible to discuss the burial site from the social point of view. As stated above, this burial site probably outlived social changes, but remained in use. Up to the 11^th century, cremations (very typical to Estonian society during all periods of the Iron Age) were conducted, but from the second part of the 11^th century another type of burial began to be practised used – inhumation. Inhumation cemeteries from this period have been uncovered from several places in Kodavere parish (they are always locationally connected to Lake Peipsi), but also from the North-Estonian shore, and from Western Saaremaa. The oldest are in Virumaa and by Lake Peipsi (Mägi-Lõugas 1995a, 1995b), and the artefacts of foreign origin mentioned above are found in inhumation graves, not cremations. The inhumations at Lahepera can be interpreted as a shift in the local social structure (as a result of foreign influence) or burials of people from other areas, but the site itself was still used.

The cemetery was taken into use again (or continued to be used - it is not possible to say if there was a gap in utilisation or not after the 14^th century, because most burials of that time are unfurnished) in the Medieval and Early Modern periods as a village cemetery. By this point in time the social and political background had changed so dramatically that the power of the former kin lines, who had used this place to bury their dead as a manifestation of their power over the landscape, most likely was not valid any more.6 But the place in the ground was still remembered, probably considered sacred, and still used. In this sense, the collective memory was preserved, despite profound historical change.

The narration above illustrates how the ancient burial site in the present village of Lahepera can be interpreted as a place where life continued despite changes in different spheres of people’s lives. The continuality of landscape can be caused by surrounding environment (soils, climate), land forms (lakes, knolls), and the social structure of the society living there.

Pieces of this distant time that have been preserved until the contemporary period narrate their stories, and based on them a version of the story of past people can be compiled. The story narrated by features of the landscape can be seen as a reflection of collective memory brought to present times by those features. And this is how landscape is continual, as is the life of past societies.