Traceless death. Missing burials in bronze and iron age Estonia/ Jaljetu surm. Puuduvad matused pronksi- ja rauaaegses eestis.

Lang, Valter

Departure

In an overview of the Stone Age religious beliefs written half a century ago, Lembit Jaanits (1961, 68 f.) drew attention to the fact that the majority of people who lived at that time were most likely never buried in the ground. This conclusion was made on the basis of an amazingly low number of burials known in either settlement sites or separate cemeteries outside the settlements of that time. Similar thoughts had been published by Estonian folklorist in exile Oskar Loorits (1949, 118) already 12 years earlier. Although more cemeteries and graves from the late Mesolithic and Neolithic have been discovered within the last fifty years in what is today Estonia and its neighbouring areas, this suggestion is still valid and realistic. With a reference to anthropological evidence of some Siberian peoples, Jaanits supposed that the dead could have been taken to certain places outside the settlements (located e.g. in the forest) and left there on the surface of the ground, wrapped perhaps into skins or birch bark (Jaanits 1961, 69). After some time almost no traces will remain from such exposed bodies.

As for the number of burial sites, the situation is even worse concerning the Early Bronze Age: no graves have been reported so far in Estonia, which could belong to the second millennium BC (Lang 2007, 147). This is most likely due to the stage of investigation and it is only a matter of time when the first burials will be discovered. In the southernmost neighbouring areas, for instance, several flat cemeteries with pit graves of the Early Bronze Age date have been unearthed, e.g. at Kivutkalns and Raganukalns (Graudonis 1967; Denisova et al. 1985). It was also in the late second millennium BC when the first monumental above-ground burial mounds were erected there, e.g. at Pukuli, Reznes, and Kalniesi (Graudonis 1967; Vasks 2000). In northern and western Estonia the first monumental stone graves were built slightly later, i.e. at the beginning of the Late Bronze Age (Lang 2007, 147 ff.). Since that time, at least one portion of burials has become very much visible in the archaeological record, and forms the main subject of research.

However, it was gradually understood since the early 1990s that despite large numbers of stone graves of the Late Bronze and Iron Ages, one part of prehistoric populations have never been buried there. How big that part was, is not clear. At first sight this conclusion based on palaeodemographic calculations was only made for both stone-cist graves of the Late Bronze and early Pre-Roman Iron Ages and north-west-Estonian tarand-graves of the Roman Iron Age (Lang & Ligi 1991; Lang 1995a) because the number of burials in those graves was too small even for regular nuclear families. The tarand-graves in other parts of the country yielding larger numbers of burials were regarded to correspond to burial places of single farms with either nuclear or extended families. Later research into the osteological evidence of cremated bones has clearly demonstrated that even large burial grounds of the Middle and Late Iron Ages might have belonged only to one or a few families and not to larger village communities (Magi 2002, 74; Allmae 2003; Mandel 2003). This conclusion did not suit historical documents, however, that had reported a settlement pattern consisting of relatively advanced villages in the 13th century.

The current article aims to discuss the problems connected with mortuary customs that are difficult to study or even invisible for archaeology. This is, first, a study of 'the others'--people who did not belong to the sphere of those buried finally in stone graves, sand barrows or flat cemeteries. Second, this is also a study of complexity in cultural behaviour concerning the death and mortuary customs in which the 'proper burying' has formed only one--and perhaps not the most popular--way of acting. Following the definition given by Frands Herschend (2009, 37), by the term 'burying' I mean the placing of the dead in a 'burial site', i.e. a construction (either over or beneath the ground), which was meant to last a generation or more reminding the descendants of their ancestors. Thus, 'proper' burying means some input of labour with the purpose of building a burial site, which was able to preserve the remains of the dead for longer times; and often also (but not necessarily) some investment of wealth by giving grave goods to the dead. This is what 'proper' ancient burying means in archaeological terms today, but I am well aware of the fact that prehistoric people thought and acted according to rather different concepts. However, the result of their different approaches to death and post-mortuary activities is that the burial customs of some people and communities can be and those of the others cannot be studied by archaeological means today.

Who were buried, who were not?

The Late Iron Age

As a matter of fact, there is not much to study if the subject is invisible; at least, there is not much to study empirically. The fact is that a large portion of the dead people are 'lost' not only from the earlier periods but also from the Late Iron Age; that is, they have not reached burial grounds that have been, and can be, investigated by archaeological means. Is it possible to find the grounds for differentiating between people who were buried and those who were not?

First, it seems obvious that not all settlement units had their own burial places. In order to get a general picture of known burial sites in relation to probable settlement units and population, one can use some numbers, although rather roughly. In the early 1970s, the number of registered burial places of the Late Iron Age in northern Estonia (111) was more than four times smaller than the number of settlement units at the end of prehistoric times (478) (Selirand 1974, 192). However, the former number is given forty yeas ago and several new discoveries have increased it to some extent. From another side, twenty-four of these 111 graves (21.6%) counted by Selirand (1974, 265) have not been proper cemeteries of the period, but earlier built stone graves where the excavations have mostly yielded only a few Late Iron Age burials or simply single artefacts of later origin. Such sites cannot be taken as real cemeteries of corresponding village communities; rather they reflect some aspects in the ancestors' cult. Thus the question of the fate of ca. 80% of the Late Iron Age dead people (resp. settlement units) in northern Estonia arises. The same discrepancy between the Late Iron Age burial custom and that of the medieval times (characterized by the burying of almost all dead people either in churchyards or local village cemeteries) becomes clear if one compares the numbers of burial places of those periods in southern Estonia: there are ca. 100 cemeteries known from the Late Iron Age (Selirand 1974, pl. XLII) (1) and 1200 burial places from the medieval and early modern times (Valk 2001, 18, fig. 5). The latter number is not only the result of the changes in burial customs, however, but also reflects the population increase during the 13th-18th centuries. A recent settlement-archaeological study of two south-Estonian parishes, Vonnu and Kambja, has revealed that there were only 18 villages of 40 (registered in the 1680s) that had archaeological evidence of the Late Iron Age (Roog 2008, table 6). Using these parishes as a model for southern Estonia, we can conclude that the number of settlement units in the late 17th century was ca. 2.2 times bigger than that at the end of prehistoric times. Keeping this increase in mind, we can suppose that the number of known late prehistoric burial sites forms ca. 18% (using data given by Selirand) or 24% (data by Valk) of the total number of probable settlement units of the time.

Thus, approximately 20% of the settlements in the Late Iron Age had some kind of cemeteries, the distribution of which forms an evenly sparse pattern over the country. This is a rough average number, which in some core areas can be bigger, but in peripheral areas remarkably smaller. For instance, in north-western Estonia (east and south-east of Tallinn), on the Island of Saaremaa and in the western part of the mainland the cemeteries are located more densely (Selirand 1974, pl. XLII) but in the area between the villages of Vihasoo and Palmse in northern Estonia no certain cemeteries of the 11th-13th centuries are known (Lang 2000, 265 f.). As to core areas, such a distribution pattern in general speaks for a structure where single settlement units with cemeteries were surrounded by numerous other settlements without cemeteries. Some examples of this structure have been published for the surroundings of what is today Tallinn (Lang 1996, figs 106, 111, 116, 119, 124, 128, 131) and some for the Island of Saaremaa (Magi 2002, figs 7, 11, 13, 15, 32), and although new cemeteries can be discovered in the future, this hardly changes the general picture. According to Marika Magi's calculations, up to 240 farms could have had stone graves for their dead on the islands of Saaremaa and Muhu in the Late Iron Age (Magi 2002, 74), which however forms not more than 10% of all the farms at the time (cf. Johansen 1925; Palli 1996, 21 ff.; Lang & Valk in print). The addition of inhumation (and probable cremation) cemeteries beneath the ground naturally increases this percentage for these islands, but it is not clear how much.

Second, as already described above, it seems that the majority of investigated cemeteries belonged to relatively small communities, i.e. only to one or a few families. We unfortunately do not know in most cases what the real settlement units nearby such cemeteries were; i.e. whether they were indeed single farms or villages with many farms, one or a few of which used to bury their dead in a cemetery. The situation might have been different in different areas. The area on the lower reaches of the Pirita River can be taken as an example to describe these differences in the Late Iron Age (Lang 1996, fig. 106). There is only one 'proper' flat cemetery with a stone cover in this area located at Proosa. The cemetery is rich in cremated burials (not analysed in osteological terms) and grave goods of the 11th-13th centuries, but it most likely belonged to a rich manor-like single farm Koskil, which had 4 plots of arable land (i.e. ploughlands) in the early 13th century (see Johansen 1933, 450). There are no cemeteries reported from the neighbouring villages at Vao (13) and Nehatu (7 ploughlands). Only a small stone grave with a few bones and grave goods is known from Iru but it is hardly a cemetery of the whole village with 7 ploughlands. The same can be said about Lagedi (23 ploughlands and a relatively small number of grave goods dispersed in several earlier stone graves or found as isolated finds nearby) and Saha (12 ploughlands and a small flat cemetery with inhumations in pit graves near the chapel on the Chapel Hill2), while no cemeteries have been discovered at Ulejoe (16), Maardu (12), and some other places of smaller size. Thus it is evident for this area that (i) only one relatively small settlement unit (perhaps a large single manor-like farm) buried their dead in a 'proper' way, (ii) several villages with many farms did not bury their dead at all in proper cemeteries (or these cemeteries are still unknown, which is less believable for this well-studied area), and (iii) only one or a few farms from some other villages have buried some of their dead (or simply put some offerings without burials) either in the (earlier) stone graves or in a new way as inhumations in pit graves, while the rest of the farms of the same villages did not bury to preserve.

Many other examples can be found from northern Estonia where both the in-depth settlement-archaeological analyses (Lang 1996; 2000; Vedru 2010) and the lists of villages of the early 13th century (Johansen 1933) are available. Yet, almost all these examples seem to fit one of these three possibilities mentioned in the paragraph above. One of the few exceptions is a flat cemetery with inhumations in pit graves located at Pada, in front of the large hill fort. Only one fourth of the area of this cemetery has been excavated but 172 richly furnished skeletons (Tamla 1996) indicate a short-time burial ground of a large and wealthy community, that inhabiting the hill fort. It is likely that the majority of the dead of this hill fort community were buried in the cemetery. A similar cemetery, but probably smaller in size, was recently unearthed at Kukruse, north-eastern Estonia (Lohmus et al. 2010). One more exception comes from the sand barrows of the Votic population in north-eastern Estonia (e.g. Jouga and Kuremae) where also all people seem to have been buried (Ligi 1993). This is, however, a reflection of another, different burial custom, which seems to have been unfamiliar to the rest of the country.

It thus seems to be evident that the 'proper' burying of the dead and furnishing them with grave goods was characteristic of a few communities sharing socially outstanding positions in society. In addition, there are many earlier stone graves that have yielded limited numbers of Late Iron Age grave goods, but that could not have served as proper graveyards for whole communities living nearby. The example of Pada indicates the hill fort communities that without doubt also possessed better positions. This must have been much more widely spread in the Late Iron Age as there were numerous other hill forts with permanent settlement the cemeteries of which are still unknown. From another side, there is evidence of a number of burial sites with cremations beneath the ground. Some of them have been quite richly furnished but many are poor or empty of grave goods, which does not allow to group them among the graveyards of the elite (see more below).

Although the social difference seems to be the main feature dividing the communities to those whose dead were buried and whose were not, it does not mean that there could not be other features, e.g. differences in religious beliefs. Remarkable differences in the treatment of bodies after death certainly refer to different understanding of afterlife. Yet, it seems likely that probable differences in beliefs were socially determined.

The Late Bronze and Early Iron Ages

After this overview of burying customs at the final stage of prehistoric times, which is relatively well covered with archaeological (and partly also written) evidence, it is necessary to analyse briefly also earlier times in order to reach the roots of the phenomenon of 'not-burying'. Actually, the latter is a more primordial custom in human culture than the burying of the dead; yet, our modern stereotypical way of thinking does not accept it easily in the 'culturally developed West'.

Around the turn of the II and I millennia BC, monumental above-ground stone-cist graves with both individual and collective burials started to spread in northern and western coastal Estonia; that is, in the region, which did not know burying of the dead in the II millennium BC (3). This phenomenon coincided with other remarkable changes in society during the early Late Bronze Age, as reflected e.g. in the distribution of fields with permanent fences, fortified settlements, the knowledge of bronze casting, etc. (Lang 2007). Although the advent of stone-cist graves as such can be interpreted in the terms of religious change, it thus coincided with the sharpening of social stratification. Considering archaeological and anthropological parallels from other regions of the world, we can assume that the high status in society was not graded on a purely individual basis, but that social groups (e.g. lineages) were ranked and their status was hereditary (cf. Wason 1994, 90). However, the question of who was buried into those graves, is not easy to answer, because the evidence of contemporary settlement sites is very poor. The graves display different distribution patterns: (i) in some core areas (e.g. north-western Estonia) the groups of stone-cist graves are located rather evenly and densely reflecting, perhaps seemingly, the map of single agricultural farms; (ii) in some areas (the surroundings of lake Kahala) the graves are located so densely and in so large numbers that no such pattern of farms can be realistic; (iii) in peripheral areas they have spread so sparsely that they hardly reflect the existence of a whole settlement. We can add that (iv) large interior areas of Estonia have yielded almost no stone-cist graves (with some exceptions), although numerous stray finds, some settlement sites and pollen analytical data from bog and lake sediments prove the existence of human habitation there. As to the burial custom, we can conclude that in the areas of the 1st group all farms have probably built stone graves, but only a few members of those farm communities have been buried in those graves. The total number of burials in the graves of the 2nd group (Kahala) seems to be so large that it even exceeds the number of people who could live in that area. Only a few farms in the areas of the 3rd group have built stone-cist graves and no such graves were used in the areas of the 4th group.

Ca. 500 tarand-graves of the Early Iron Age are known (Lougas & Selirand 1989, 71) that are distributed rather evenly across the country, except for western Estonia and the islands with very few graves. The lack of evidence of settlement sites does not allow us to draw any conclusions about the relationships between the numbers of tarand-graves and contemporary settlements. Yet, some calculations with general and rough numbers might help us to go further. The palaeodemographical considerations have shown that one tarand-grave was most likely used for burying by a community with approximately ten members (Lang 1996, 356 ff.). Although many tarand-graves have been destroyed over centuries, these 500 graves, from the other side, are certainly not contemporary and, therefore, we can use this number as the highest average number of contemporarily used tarand-graves. The total number of inhabitants in what is today Estonia in the Roman Iron Age is considered around 30,000 (Lang 1990; Ligi 1995, 222); i.e. not more than 3000 families with ten members in each. These numbers mean that only ca. 17% of the communities built and used the graves in question. But this is the average number and the situation was different in different areas; for instance, we have to subtract the population of western Estonia and the islands from this number in order to get a more real picture for the rest of Estonia with tarand-graves. But even there the situation varied and the areas with denser pattern of tarand-graves were surrounded with areas displaying much sparser patterns of graves.

One can thus conclude that the tradition of burying the dead started to spread next to the old tradition of 'not-burying' since the beginning of the Late Bronze Age (excluding here the sporadic Stone Age evidence). It has always been a custom which was only shared by a small portion of population and communities. Whatever the initial factors were according to which people were buried or not, the burying of the dead into stone graves or other cemeteries became gradually a custom of groups with a socially outstanding position. This was so perhaps already in the Late Bronze Age and certainly in the Pre-Roman and Roman Iron Ages. As to the Middle and Late Iron Ages, the numbers of reported cemeteries are even smaller than those in the Roman Iron Age (compare corresponding numbers in Vassar 1956, 183, footnote 64, fig. 42; Selirand 1974, pl. XLII; Laul 2001, 27, fig. 3 and Lang 2007, fig. 116), although the population had meanwhile increased ca. five-six times as believed by some authors (from ca. 30,000 to ca. 150,000 or even 180,000). The corresponding numbers of the Late Iron Age indicate that only ca. 20% of people were buried either in stone graves or flat cemeteries beneath the ground and, hence, 80% of people have not reached burial grounds that can be discovered by archaeological means today. In social terms, 20% of society should have involved not only the elite but probably also some outstanding portion of the 'middle class'.

A hypothesis about the social factors of burying

As burials in 'proper' graves seem to have been socially determined, we must look for some social structuring in society, which would help to understand the low numbers of 'proper' burial sites. In the Late Bronze and Early Iron Ages (until the middle of the first millennium AD), the communities in the region of the distribution of monumental stone graves were supposedly organised by groups of farms led by 'dominant' farms (Lang 1996; 2002; 2007). As the settlement sites of that period are rare, the settlement units were mostly distinguished on the basis of the location of stone graves, which means that we do not actually know the real number of surrounding communities. The analysis of grave goods has yielded, however, that only a few communities erecting and burying in stone graves could be labelled as outstanding (and thus 'dominant') while the rest were 'ordinary' in the sense of wealth invested in graves. Palaeodemographic calculations reveal a larger part of society that did not at all bury in stone graves. The territories of the systems with one dominating farm embraced ca. 100-200 [km.sup.2] until the (late) Roman Iron Age, but were then divided into smaller units (ca. 40-60 [km.sup.2]), at least in north-western Estonia (Lang 1996, 482).

During the Middle and Late Iron Ages, the socio-economical formations of farms and villages, called vakuses in Estonian, developed on the basis of former territories with one dominant farm (Lang 2002). The vakuses were prehistoric taxation and administration units that were later, after the conquest, taken over by new landlords. At the beginning of the 13th century, the average size of the vakuses in Ravala district was ca. 39 and that in Harju district ca. 30 ploughlands, being remarkably bigger in Viru district. The number of villages in such territories was most often between 3 and 6; the number of farms was presumably 25-30% smaller than that of ploughlands (Lang 1996, 366; Tarvel 1996, 246). We can assume that the chiefs or owners of the vakuses formed the main portion of the Estonian prehistoric nobility, perhaps those who were called "richer, better and elder" in the chronicle of Henry of Livonia from the early 13th century (HCL 1982).

The total number of ploughlands at the beginning of the 13th century was ca. 21,000-22,000 (Palli 1996, 21 f. and references therein). Although their number per a vakus could vary quite remarkably in different parts of the country, I think that 30-40 as an average could be acceptable in order to make the following calculations simpler. These numbers mean that there could have been ca. 525-733 vakuses in late prehistoric Estonia. (4) Jiiri Selirand has counted altogether 316 sites with Late Iron Age burials in what is the mainland of Estonia (1974, 37).

When we add the burial sites on the islands of Saaremaa and Muhu, new sites discovered during the last 40 years, and both presumably destroyed and still not discovered cemeteries, I guess the total number of cemeteries will fit rather well with that range for the number of vakuses. This means that, as an average, there could have been only one 'proper' cemetery within a territory of farms and villages of one vakus-type unit of taxation and administration.

This hypothesis can be tested by landscape archaeological studies, the first step of which is to distinguish the territories of prehistoric vakuses. Without going into details, it is possible to do this--although always with some uncertainty only in northern Estonia where we possess data on the size of settlement units in ploughlands of the early 13th century (see more Lang 2002). At least in some regions of northern Estonia this hypothesis seems to work rather well--for instance, in the areas of the lower and middle reaches of the Pirita River mentioned above. In some other regions the picture is not so clear, but this is mostly due to the lack of in-depth settlement-archaeological investigations. The discrepancy comes there from the deficiency of cemeteries of the Late Iron Age (but such can be discovered in the future) rather than from the circumstance that there are two or more outstanding cemeteries within one vakus. True, in several vakuses there can be some other features of burying in addition to a 'proper' cemetery; yet, in such cases one is mostly dealing with the limited re-use of older stone graves. Of course, there is no need to exclude the possibility of several outstanding cemeteries in one vakus either; in this case there were probably several leading (and conquering) lineages in the neighbourhood. And similarly--there is possibly no need to expect the existence of at least one 'proper' cemetery in each of the prehistoric vakus. The real life was certainly much more complicated and many-featured than one hypothesis could reflect; yet, the general distribution of late prehistoric cemeteries seems to correspond to the spread of taxation and administration units. In other words: the distribution of well-furnished cemeteries seems to reflect the geographical location of leading families.

The situation in some neighbouring areas

The grave deficit characterizing prehistoric periods is a phenomenon of which the archaeologists in many countries have increasingly become aware within the last 20-30 years (e.g. Whimster 1981; Barrett 1996; Kaliff 1997; 1998; Parker Pearson 1999, 5, 132; Armit & Ginn 2007). Frands Herschend has recently (2009, 33 ff.) summarized and analysed the more common indications of this deficiency in an earlier research in Scandinavia. According to him, the researchers have pointed out, first, the lack of children's graves, characteristic of almost all prehistoric periods. Second, some certain geographical areas have lacked the graves datable to some specific periods. And third, certain taphonomic factors such as unfavourable preservation conditions, unlucky find circumstances, lack of money or professional interest, the devastating effects of some human activities, etc. could also be responsible for the too small numbers of graves. Herschend himself is convinced that the deficiency of graves is not occasional because

... only a fraction of the Early Iron Age dead was buried in such a way that we can recognise the context of their funeral as indeed a burial. In most cases of death, we may suspect that after a short time, there were no context let alone any remains to be found (Herschend 2009, 44).

Early graves have often reflected settlements of outstanding, leading farms, thus being class-biased and upper-class bound. Cemeteries can also be 'thematic' in a way that, for instance, only children were buried in certain places or armed men in some other places (op. cit., 117 ff.).

Recent studies have also stressed the deficiency of graves in the reflection of prehistoric settlement in Finland. The number of Late Iron Age cemeteries in south-western Finland, for instance, is considered to be many times smaller than could be expected on the basis of medieval settlement units (e.g. Pihlman 2004; Asplund 2008, 311 ff.). Thus the situation in south-western Finland resembles very much what is known from Estonia. According to Sirkku Pihlman (2004), only one group of farms buried their dead in cemeteries--these were well-established farms in core areas, which most likely belonged to outstanding families of upper classes. Other farms were dependent on these elite farms, were located in more peripheral areas and did not bury in archaeologically detectable contexts. This interpretation, in principle, is quite acceptable; yet Pihlman's estimation for the number of 'non-burying' farms (equal to that of burying ones) seems clearly too small. It is similarly questionable that only peripheral farms did not bury their dead in cemeteries; the Estonian example proves that a number of villages located in old-established settlement cores and had a long agricultural history did not use cemeteries either in the Late Iron Age. The situation might naturally have varied in different regions and it was different also in the areas of Finland; one exception has been the Eura area with its rich burial evidence where at least the cemetery at Luistari has been used for burying by a whole village (LehtosaloHilander 1984). In the Crusade Period Karelia (both on north-western and southern/ south-eastern shores of the Lake Ladoga), on the other hand, richly furnished inhumation cemeteries and barrows have been interpreted as belonging to a small upper class, while the cemeteries of 'common people' are still not known (Moora 1956, 102, note 133; Uino 1997, 117).

Due to the lack of relevant analyses it is difficult to estimate the situation in Latvia and Lithuania. In the Early Metal Ages (until the Roman Iron Age), the numbers of known cemeteries are so small that they hardly reflect the distribution of a whole settlement. For instance, only nine sites with barrows of the late Pre-Roman and early Roman Iron Ages have been registered in what is now western Lithuania, giving a reason to suggest a very sparse settlement pattern (Michelbertas 1986, fig. 83; Couronians 2008, 38). Although the number of burial grounds there increased remarkably in the following centuries of the Roman Iron Age--interpreted as an evidence of population increase (Couronians 2008, 39)--they still hardly reflect the location of all settlement sites. The same can also be said for the rest of Lithuania (cf. Michelbertas 1986, figs 83-87) and Latvia (e.g. Snore 1993) during the Roman Iron Age. Since the Middle Iron Age, however, large cemeteries with hundreds of burials occurred in both Lithuania and Latvia (e.g. Tautavicius 1996, 44-100; Radins 1999) that refer to different directions in the development of burial customs in comparison to what we know in Estonia. How representative are the numbers of cemeteries in the relation to those of settlement sites, is not however clear.

Alternative ways in mortuary customs

There are several possible ways how the dead could have been treated alternatively to the burying in above-ground stone graves or flat cemeteries with inhumations in pits. Some of them are documented by archaeological means; others are purely speculative but still highly possible due to ethnographic parallels elsewhere. They all refer to highly developed complexity in post-mortuary customs.

Underground cremation cemeteries

First, some observations are necessary about the burial customs that have led to the making of real cemeteries which, however, are very difficult to discover. These have been, first of all, cremations in pit graves and cremations on prepared plots.

There is some limited data about cremations in pit graves that belong to the Late Bronze and Early Iron Ages (Lang 2007, 217 f.). As the majority of this data has come from the excavations of stone graves, it cannot be excluded that such a custom was somehow linked only to those graves or--which is perhaps more relevant--places of the graves that through continuous reuse represented significant foci in cultural landscape (cf. Barrett 1996; Bradley 2002). More data about the cremations in pit graves (sometimes connected with structures of stones) comes from the Middle and Late Iron Ages. A thorough overview of such later cemeteries, both certain and presumable (34 altogether), is given by Mati Mandel (2003, 141 ff.). In most cases these cemeteries have been provided with grave goods. Several burial sites of this kind have been discovered only recently, e.g. at Harmi (Tamla & Kivistik 2000; Tamla 2002), Raasiku (Ulle Tamla, pers. comm.), and Kodasoo (Vedru 2004); the burials at these north-Estonian sites were richly furnished as well.

As for the later Iron Age, it is also highly relevant in this context that a new type of burial custom was recently distinguished in south-eastern Estonia: cremations on prepared plots that were covered with a layer of earth. At Kirikumagi and Kalmetemagi in Siksala, cremated human bones were dispersed over the top of the hillocks, rather densely at some places and quite sparsely in some others (Laul & Valk 2007, 18; Valk & Allmae 2010). Topsoil was removed before the burying and larger plots were prepared for burying, while up to half a metre thick layer of sand was heaped on the bones and these plots after the burying. Grave goods were either missing (Kirikumagi) or few in number (Kalmetemagi), but both burials sites were dated to the Late Iron Age (Valk & Allmae 2010). The burial custom just described has close parallels with that reported by the excavators of sand barrows (e.g. Aun 1980). The problem is, of course, that traces of such cemeteries are very hard to discover and thus we simply do not know how widely the custom in question had spread. The same can be said about the burial field with cremations underground recently discovered at Uugla, western Estonia. Scattered pieces of burnt bone together with a few grave goods can be found there over a large area around a richly furnished stone grave with cremations (Mandel 2008; 2010). As both the stone grave and scattered cremations around were contemporary (11th-13th centuries), we may speculate about social relationships between these two groups burying in different ways (Mandel 2008, 87). It has to be noted, however, that the data of such sites is so limited today that it cannot solve the questions about the burial custom of the majority of population.

Scattered cremations

In addition to these more or less 'proper' burials of cremated bones, there is also some evidence on cremated human bones scattered over a larger area on the ground; in some cases such bones might have been covered with a thinner layer of soil. It is understandable that scattered burnt bone can survive in soil and is later discovered only when protected with later deposits. Therefore it is not surprising that our knowledge about such a custom comes only from the excavations of later monuments. Scattered cremated bones were unearthed, for instance, when excavating a stone grave with inhumations at Ilmandu, north-western Estonia (Lang 1995b). These bones together with pieces of charcoal were discovered there under the bottom stones of the grave in the depth of 5-6 cm and were radiocarbon-dated to the beginning of the Late Bronze Age. Traces of a very similar custom were later also discovered at Tougu, northern Estonia, where cremated human and animal bones (without pieces from skull)5 were found scattered beneath the Pre-Roman tarand-grave IIC with inhumations (Lang 2000, 113 f.). At Kukruse, north-eastern Estonia, scattered cremations together with some pieces of burnt grave goods of bronze were discovered between the pit graves with richly furnished inhumations of the 12th-13th centuries (Lohmus et al. 2010, 75). It seems that the place was first used for scattering the cremated burials and later turned into a 'proper' inhumation cemetery.

Scattered burnt bones beneath the stone graves, sand barrows or in the area of flat cemeteries have been discovered at many other places as well. (6) It is clear that without later buildings protecting such earlier burials they disappeared. If these burial sites contain cremated burials, it is not always clear, however, whether such scattered bones originate from those later graves (post-depositional replacement) or initial burials that were put there before the monuments in questions were erected. The datable contexts refer to a long timespan, reaching from the Late Bronze Age to the Middle Ages. The use of the same places over centuries and even millennia for certain ritual purposes--involving the burying of selected human bones, erection of monumental stone graves, barrows and cemeteries, building of chapels, etc.--emphasizes the focal sacral importance of these places for the surrounding communities.

In addition to place-related dispersal of cremated bones, scattering ashes in the wind and water has also been suggested by some authors (e.g. Magi 2007, 9 f.). Empirical data of such a custom in Estonia is not available, but the underwater cremations of the 12th-15th centuries are reported from the territories of the Couronians in western Latvia (Couronians 2008, 64 f.) and there are some still earlier water burials explored in Finnish Ostrobothnia (Wessman 2010, 27). Throwing ashes into water is also mentioned in old Icelandic sagas (Kaliff 1997, 92). There are several other ways how (cremated) bones could 'disappear' from ordinary burial places, e.g. the use of bones for carbonizing iron in smitheries (Gansum 2004), sacrificing 'raw, cooked and burnt humans' to gods (Oestigaard 2000), placing different parts of cremated bodies in different places, both in graves and ritual buildings (Kaliff 1997; 1998), etc.

Exposure

One of alternatives to burying in 'proper' cemeteries is considered to be the exposure of the dead or the so-called open air burying--i.e. leaving the dead bodies on the ground, on top of trees or elsewhere. As the experiments with corpses of dead animals have demonstrated, there will be almost no visible traces (except some bigger bones) of such 'burials' on the surface of ground already after a few years (Jonuks & Konsa 2007; Tonno Jonuks, pers. comm., January 2011). Of course, until we do not possess direct archaeological evidence of the mortuary custom like this, it is nothing but pure speculation to presume the existence of that custom in prehistoric times.

However, I would like to refer here to isolated human bones found outside the burial contexts. The occurrence of such bones in settlement layers does not prove the existence of the custom of exposure by itself, because they can also originate from ordinary proper graves and speak about some kind of manipulation with the bones of dead ancestors.7 For instance, numerous human bones found in Scottish Iron Age settlement sites have been partly interpreted in this way (Armit & Ginn 2007). According to Ian Armit and Victoria Ginn (2007, 129), there is a strong evidence that such human bones often reached the final archaeological contexts after a long period of time (up to several centuries, perhaps), participating meanwhile in ritual activities of the living communities. This is also what can be said about Estonian prehistoric cemeteries, as it is often reported that the bodies are not complete and many bones are missing (e.g. Vedru 1998; Lang 2000; Jonuks 2009, 173 ff.). But human bones without other evidence of particular burial or settlement sites have been often discovered all over the country and our topographic archives are full of such reports. Such findings and reports may naturally come from the destroyed 'proper' cemeteries--particularly from medieval village cemeteries where grave goods were not always obligatory. Until we have not excavated any of these sites, it is impossible to provide more detailed analysis of the subject.

Ethnographic parallels can also be drawn. As already mentioned above, some Siberian peoples take their dead to certain places outside the settlements (located e.g. in the forest) and leave them there on the ground, wrapped perhaps into skins or birch bark (Jaanits 1961, 69 and references therein). Still in the early 20th century some indigenous Siberian tribes used to bury their dead far away from the villages in order to avoid the ghosts coming back to home (Donner 1979, 147). Oskar Loorits (1949, 118) suggested that ancient Estonians left dead bodies on top of trees. There are early written sources telling that in Mongolia, four different mortuary customs were known before the times of Genghis Khan: bodies could be buried, cremated, left exposed to wild beasts or in trees, whereas exposure could be followed by the collection of bones and their subsequent burial (Crubezy et al. 2006).8 Even today the dead have been left there in open air, on mountain slopes,

where birds feed on the earthly remains of the deceased and thus take their soul up to heavenly spheres (Jonuks & Konsa 2007, 97, fig. 1).

We can thus conclude that prehistoric mortuary customs have been rather diverse and only some of them have led to the 'proper burying' of dead. The traces of the burying have completely disappeared in many cases of 'alternative' burying, while in many others they have been found only by chance. It must be stressed here once again that the main difference between the 'proper' and 'improper' burying comes from the circumstance whether there is something preserved that can be studied by archaeological method today or not. In this way, the 'proper' and 'improper' burials are our constructs, not theirs in the past. But it should also be clear that there had to be some essential difference in social and/or religious sphere as well, which determined such a difference in mortuary customs.

Consequences

Although the situation has slightly varied through different periods of the Metal Ages and in different regions of Estonia, we can still conclude that the major part of population has never been buried in a way, which could preserve the remains of dead bodies until today. This fact has been briefly acknowledged by many authors during the last decades (Lang & Ligi 1991; Magi 2002; Valk 2009); yet it seems that this matter has not been considered in further discussions about prehistoric burial customs or related topics.

One of the consequences is that the available numbers of prehistoric graves or burials in the graves cannot be used in palaeodemographic calculations for the whole population once living in what is now Estonia. As the known grave structures of the Late Bronze and Iron Ages seem to be socially determined and belong to the groups of elite or 'higher middle class' (Ligi 1995), then all fluctuations in the number of corresponding graves can be explained by changes in social strategies of these groups forming only 20% of the entire population. This means that research into further 'social stratification' within the populations buried in archaeologically known graves has to consider that those communities shared more or less outstanding positions anyway and there is no sense to look for lower strata in those graves. The earlier research has distinguished both dominant and ordinary farms among the communities buried in the Roman Iron Age tarand-graves while the rest of society was buried elsewhere (Lang 2007, 225 ff.). The decrease in the number of archaeologically known graveyards during the Middle Iron Age and a too slight increase in the Late Iron Age on the background of remarkable growth of population during the same time (from ca. 30,000 to ca. 150,000-180,000) speaks most likely about the general narrowing of the circle of people buried in 'proper' cemeteries. This, in turn, means the sharpening of social stratification. Differences within the burying units (richer and poorer burials within a grave or cemetery) are mostly occasional and natural variations between the members of leading lineages.

The question about possible differences in religious beliefs between the elite groups and common people is not easy to answer because there is not enough data on the mortuary customs of the latter. What we know almost for certain is that the religious practices of common people were not connected with the erection and use of monumental graves; i.e. the cult of the ancestors was probably not very well established--or it was practised otherwise. Concerning the Bronze and Early Iron Ages I have suggested earlier (Lang 2007, 249) that fertility cult rituals, which were performed by making cup-marked stones, might be connected with common people, whereas monumental stone graves were used for religious practices by leading groups. The use of sacred groves also fits better the world of beliefs of common people--the more so if their use started together with the advent of stone graves, as believed by Tonno Jonuks (2007, 23; 2009, 209). It does not mean that elite groups had to be somehow excluded from the rituals performed either near the cup-marked stones or in groves and most likely the beliefs of both social strata suited one another well (as proved, for instance, by the sharing of the custom to treat differently the bodies and the head; see footnote 5). Estonian folk religion knows natural sacred places of different levels, reaching from a very individual sphere (located in the corner of a farm garden) to village and larger communities' levels (see Eisen 1996, 110 ff.); the rituals in the latter were most probably carried out also with the participation of leading social groups.

Abandoning late prehistoric cemeteries in the early 13th century and the subsequent foundation of numerous village cemeteries and cemeteries in churchyards are usually interpreted in the light of the advent of Christianity. Considering what was said above, I would add and stress one more reason: the termination of the Estonian elite in the course of the German-Danish conquest. After the conquest, the remains of the elite families were gradually blended in common rural population, and lost their traditions to bury in their own graveyards--except, perhaps, the occasional later burials and offerings in some of these older cemeteries, which are reported by both written sources and archaeological evidence. The continuation of burials at some late prehistoric places well into the medieval era (e.g. at Kaberla and Siksala; see Selirand 1962; Laul & Valk 2007) can refer to the continuity and survival of the leading families. The new religion brought about the fading of the earlier mortuary customs involving the 'improper burying' of the dead, and led to the establishment of local village cemeteries and cemeteries in the churchyards. The village cemeteries can be interpreted as syncretistic occurrences that included something from the Christianity, something from the 'proper' prehistoric burying, and something from sacred groves and earlier mortuary customs of common people. The main archaeological consequence of that all was that principally the majority of society was now buried in the ground; that is, their death was no longer traceless.

I would like to finish this essay with the conclusion that death leaves traces only if it is interpreted through culture. The burying or cremation of a dead body is a culturally determined act. Birds and animals do not bury their dead, and therefore, with the exception for those whose skeletons are occasionally preserved due to favourable natural conditions, they disappear completely. Humans, too, have not buried their dead during the major part of history. And although the first human burials occurred very early, already in the Palaeolithic, the burying as such has been selective for a very long time. Therefore the questions, who were buried and what became of the rest of the dead, are essential. The answers, I guess, depend on times and places we are talking about; i.e. the answers are culture-related.

doi: 10.3176/arch.2011.2.03

Acknowledgements

I am grateful to Dr Heiki Valk and Dr Tonno Jonuks for their valuable comments to the initial versions of this article. This study was funded by the Estonian Ministry of Education and Research (SF0180150s08) and the European Union through the European Regional Development Fund (Centre of Excellence in Cultural Theory).

References

Allmae, R. 2003. Laanemaa 5.-13. sajandi kalmete antropoloogiline aines.--Mandel, M. 2003. Laanemaa 5.-13. sajandi kalmed. (Eesti Ajaloomuuseum. Toid ajaloo alalt, 5.) Tallinn, 260-262. Armit, I. & Ginn, V. 2007. Beyond the grave: human remains from domestic contexts in Iron Age Scotland.--Proceedings of the Prehistoric Society, 73, 113-134.

Asplund, H. 2008. Kymittae. Sites, Centrality and Long-Term Settlement Change in the Kemionsaari Region in SW Finland. (Turun Yliopiston julkaisuja. Sarja--ser. B osa--tom. 312. Humaniora.) Turun Yliopisto, Turku.

Aun, M. 1980. [TEXT NOT REPRODUCIBLE IN ASCII]. Valgus, Tallinn.

Barrett, J. C. 1996. The living, the dead and the ancestors: Neolithic and Early Bronze Age mortuary practices.--Contemporary Archaeology in Theory. Eds R. Preucel & I. Hodder. Blackwell Publishers, Oxford, 394-412.

Bradley, R. 2002. The Past in Prehistoric Societies. Routledge, London.

Couronians = Kursi senatne. Couronians in Antiquity. Latvijas Nacionalais Vestures Muzejs, Riga, 2008.

Crubezy, E., Ricaut, F. X., Martin, H., Erdenebaatar, S., Coqueugnor, H., Maureille, B. & Giscard, P. H. 2006. Inhumation and cremation in medieval Mongolia: analysis and analogy. Antiquity, 80: 310, 894-905.

Denisova, R., Graudonis, Ya. & Gravere, R. 1985. [TEXT NOT REPRODUCIBLE IN ASCII].

Donner, K 1979. Siperian samojedien keskuudessa vuosina 1911-1913 ja 1914. Otava, Helsinki.

Eisen, M. J. 1996. Esivanemate ohverdamised. Teine trukk. Tallinn.

Gansum, T. 2004. Role of bones--from iron to steel.--Norwegian Archaeological Review, 37: 1, 41-57.

Graudonis, Ya. 1967. [TEXT NOT REPRODUCIBLE IN ASCII].

HCL = Heinrici Chronicon Livoniae. Henriku Liivimaa kroonika. Ed. E. Tarvel. Tallinn, 1982. Herschend, F. 2009. The Early Iron Age in South Scandinavia. Social Order in Settlement and Landscape. (Occasional Papers in Archaeology, 46.) Uppsala Universitet.

Jaanits, L. 1961. Jooni kiviaja uskumusest.--Religiooni ja ateismi ajaloost Eestis. Artiklite kogumik, II. Ed. E. Jansen. Eesti Riiklik Kirjastus, Tallinn, 5-70.

Johansen, P. 1925. Siedlung und Agrarwesen der Esten im Mittelalter. Dorpat.

Johansen, P. 1933. Die Estlandliste des Liber Census Daniae. Kopenhagen.

Jonuks, T. 2007. Holy groves in Estonian religion.--EJA, 11: 1, 3-35.

Jonuks, T. 2009. Eesti muinasusund. (Dissertationes Archaeologiae Universitatis Tartuensis, 2.) Tartu Ulikooli Kirjastus, Tartu.

Jonuks, T. & Konsa, M. 2007. The revival of prehistoric burial practices: three archaeological experiments.--Folklore, 37, 91-110.

Kaliff, A. 1997. Grav och kultplats. Eskatologiska forestallningar under yngre bronsalder och alder jarnalder i Ostergotland. (Aun, 24.) Uppsala.

Kaliff, A. 1998. Grave structures and altars: archaeological traces of Bronze Age eschatological conceptions.--European Journal of Archaeology, 1: 2, 177-198.

Lang, V. 1990. Eesti rahvastik enne ja parast aastat 1200. (Eesti Arheoloogiaseltsi teataja, 3.) Tallinn, 8-12.

Lang, V. 1995a. Die fruheisenzeitliche Gesellschaft im Gebiet der estnischen Steingraber. Archaeologica Baltica, 1, 107-120.

Lang, V. 1995b. A Pre-Roman tarand-grave and late medieval fossil fields of Ilmandu, NW Estonia.--TATU, 44: 4, 429-436.

Lang, V. 1996. Muistne Ravala. Muistised, kronoloogia ja maaviljelusliku asustuse kujunemine Loode-Eestis, eriti Pirita joe alamjooksu piirkonnas, I-II. (MT, 4.) Tallinn.

Lang, V. 2000. Keskusest aaremaaks. Viljelusmajandusliku asustuse kujunemine ja areng Vihasoo--Palmse piirkonnas Virumaal. (MT, 7.) Tallinn.

Lang, V. 2002. Vakus ja linnusepiirkond Eestis. Lisandeid muistse haldusstruktuuri uurimisele peamiselt Harjumaa naitel.--Keskus--tagamaa--aareala. Uurimusi asustushierarhia ja voimukeskuste kujunemisest Eestis. Ed. V. Lang. (MT, 11.) Tallinn, 125-168.

Lang, V. 2007. The Bronze and Early Iron Ages in Estonia. (Estonian Archaeology, 3.) Tartu University Press, Tartu.

Lang, V. & Ligi, P. 1991. Muistsed kalmed ajaloolise demograafia allikana.--Arheoloogiline kogumik. Eds L. Jaanits & V. Lang. (MT, 1.) Tallinn, 216-238.

Lang, V. & Valk, H. In print. Archaeological reading of Henry's Chronicle: events, traces, contexts and interpretations.

Laul, S. 2001. Rauaaja kultuuri kujunemine Eesti kaguosas (500 e.Kr.--500 p.Kr.). (MT, 9; Opetatud Eesti Seltsi kirjad, 7.) Tallinn.

Laul, S. & Valk, H. 2007. Siksala. A Community at the Frontiers. Iron Age and Medieval. (CCC Papers, 10.) Tallinn.

Lehtosalo-Hilander, P.-L. 1984. Luistari, I-III. (SMYA, 82: 1-3.) Helsinki.

Ligi, P. 1993. Vadjaparased kalmed Kirde-Eestis (9.-16. saj.).--Vadjaparased kalmed Eestis 9.-16. sajandil. Ed. V. Lang. (MT, 2.) Tallinn, 7-175.

Ligi, P. 1995. Uhiskondlikest oludest Eesti alal hilispronksi- ja rauaajal.--Eesti arheoloogia historiograafilisi, teoreetilisi ja kultuuriajaloolisi aspekte. Ed. V. Lang. (MT, 3.) Tallinn, 182-270.

Lillak, A. 2006. Maa-alused poletusmatused Louna-Eestis. Bakalaureusetoo. Tartu. Manuscript in the Chair of Archaeology at the University of Tartu.

Lillak, A. 2009. Maa-alused poletusmatused Setomaal.--Setomaa, 2. Vanem ajalugu muinasajast kuni 1920. aastani. Eds A. Lillak, A. Selart & H. Valk. Eesti Rahva Muuseum, Tartu, 82.

Lohmus, M., Jonuks, T. & Malve, M. 2010. Arheoloogiatalv Kukrusel. 12.-13. sajandi maahaudkalmistu ja 18. sajandi maantee.--Muinsuskaitse aastaraamat, 2009. Eds K. Matteus & L. Janes. Tallinn, 73-76.

Loorits, O. 1949. Grundzuge des estnischen Volksglaubens, 1. (Skrifter utgivna av Kungl. Gustav Adolfs Akademien for Folklivsforskning, 18: 1.) Lund.

Lougas, V. & Selirand, J. 1989. Arheoloogiga Eestimaa teedel. Teine, parandatud ja taiendatud trukk. Valgus, Tallinn.

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.) Tallinn.

Magi, M. 2007. Kivikalmetest kirikaedadeni. Matmiskombestik muinasaja lopu ja varakeskaegsel Saaremaal.--Artikleid usundi- ja kombeloost. (Sator, 6.) Eesti Kirjandusmuuseumi Teaduskirjastus, Tartu, 9-28.

Maldre, L. In print. Archaeozoological material from the settlement sites at Linnaaluste and the hill fort at Keava.

Mandel, M. 2003. Laanemaa 5.-13. sajandi kalmed. (Eesti Ajaloomuuseum. Toid ajaloo alalt, 5.) Tallinn.

Mandel, M. 2008. Archaologische Ausgrabungen in Uugla.--AVE, 2007, 85-90.

Mandel, M. 2010. Probegrabungen im unterirdischen Grabfeld mit Feuerbestattungen zu Uugla.--AVE, 2009, 147-149.

Michelbertas, M. 1986. Senasis gelezies amzius Lietuvoje. Mokslas, Vilnius.

Moora, H. 1956. Eesti rahva ja naaberrahvaste kujunemisest arheoloogia andmeil.--Eesti rahva etnilisest ajaloost. Artiklite kogumik. Ed. H. Moora. Tallinn, 41-119.

Oestigaard, T. 2000. Sacrifices of raw, cooked and burnt humans.--Norwegian Archaeological Review, 33: 1, 41-58.

Palli, H. 1996. Eesti rahvastiku ajalugu aastani 1712. Teaduste Akadeemia Kirjastus, Tallinn.

Parker Pearson, M. 1999. The Archaeology of Death and Burial. Texas A&M University Press.

Pihlman, S. 2004. Vaestorajahdys historiallisen ajan taitteessa? Voisiko ainestoja tulkita toisinkin? ABOA. Turku Provincial Museum. Yearbook, 2002-2003, nos 66-67, 47-98.

Piho, M. In print. Setode usk ja Setomaa tsassonad.

Radins, A. 1999. 10.-13. gadsmita senkapi Latgalu apdzivotaja teritrija un Austrumlatvijas etniskas, socialas un politiskas vestures jautajumi. (Latvijas Vestures Muzeja raksti, 5. Arheologija.) Riga. Roog, R. 2008. Arheoloogiliste ja kirjalike allikate kasutusvoimalustest vanema asustusloo uurimisel: Vonnu ja Kambja kihelkond. Bakalaureusetoo. Manuscript in the Chair of Archaeology in the University of Tartu.

Selirand, J. 1962. Kaberla maa-alune kalmistu (13.-17. sajand).--Muistsed kalmed ja aarded. Arheoloogiline kogumik, II. Ed. H. Moora. Eesti NSV Teaduste Akadeemia Ajaloo Instituut, Tallinn, 131-168.

Selirand, J. 1974. Eestlaste matmiskombed varafeodaalsete suhete tarkamise perioodil (11.-13. sajand). Eesti Raamat, Tallinn.

Snore, E. 1993. Agra dzelzs laikmeta uzkalnini Latvijas austrumu dala. Zinatne, Riga.

Tamla, T. 1996. Virumaa muinasaeg.--Koguteos Virumaa. Ed. K. Saaber. Laane-Viru maavalitsus, Ida-Viru maavalitsus, Tallinn, 206-244.

Tamla, U. 2002. Excavations of the destroyed grave at Harmi.--AVE, 2001, 135-143.

Tamla, U. & Kivistik, A. 2000. New archaeological sites in south Harjumaa: the settlement site, cup-marked stones and the base of a grave at Harmi.--AVE, 1999, 100-107.

Tarvel, E. 1996. Virumaa Pohjasoja lopuni.--Koguteos Virumaa. Ed. K. Saaber. Laane-Viru maavalitsus, Ida-Viru maavalitsus, Tallinn, 244-269.

Tautavicius, A. 1996. Vidurinis gelezies amzius Lietuvoje (V-IX a.). Vilnius.

Uino, P. 1997. Ancient Karelia. Archaeological Studies. (SMYA, 104.) Helsinki.

Valk, H. 1992. Louna-Eesti talurahva matmiskombestik ja selle uskumuslikud tagamaad 13.-17/18. sajandil. Magistritoo. Manuscript in the Chair of Archaeology, University of Tartu.

Valk, H. 2001. Rural Cemeteries of Southern Estonia 1225-1800 AD. (CCC Papers, 3.) Visby.

Valk, H. 2009. Hilisrauaaeg (1000/1050-1225).--Setomaa, 2. Vanem ajalugu muinasajast kuni 1920. aastani. Eds A. Lillak, A. Selart & H. Valk. Eesti Rahva Muuseum, Tartu, 126-199.

Valk, H. & Allmae, R. 2010. Kirikumagi at Siksala: evidence of a new grave form in south-eastern Estonia?--EJA, 14: 1, 40-55.

Vasks, A. 2000. Bronzas laikmeta kapulauks Pukulos.--Petijumi par Kurzemi un Zemgali. Eds A. Caune, E. Mugurevics & I. Ose. (Arheologija un etnografija, XX.) Riga, 88-106.

Vassar, A. 1956. Lisandeid Eesti hoimude uurimisele Laane- ja Edela-Eestis I-IV sajandil.--Eesti rahva etnilisest ajaloost. Artiklite kogumik. Ed. H. Moora. Eesti Riiklik Kirjastus, Tallinn, 160-190.

Vedru, G. 1998. New archaeological data of the prehistory of Lake Kahala area.--AVE, 1997, 62-66.

Vedru, G. 2004. Archaeological excavations on the cemetery of Kodasoo and the settlement site in Lagedi.--AVE, 2003, 90-95.

Vedru, G. 2010. Pohja-Eesti arheoloogilised maastikud. Archaeological landscapes of North-Estonia. (Dissertationes Archaeologiae Universitatis Tartuensis, 3.) Tartu Ulikooli Kirjastus, Tartu.

Wason, P. K 1994. The Archaeology of Rank. (New Studies in Archaeology.) Cambridge University Press.

Wessman, A. 2010. Death, Destruction and Commemoration. Tracing Ritual Activities in Finnish Late Iron Age Cemeteries (550-1150). (Iskos, 18.) The Finnish Antiquarian Society, Helsinki.

Whimster, R. 1981. Burial Practices in Iron Age Britain: A Discussion and Gazetteer of the Evidence, c. 700 BC AD 43. (British Archaeological Reports, 90.) Oxford.

(1) True, in 1992 Heiki Valk has counted in the same area altogether 131 burial sites. However, only 61 of them were certain while the rest of them were uncertain, only presumable burial sites distinguished on the basis of a few isolated finds or non-professional reports (Valk 1992, fig. 2, 34 f.).

(2) This cemetery was founded not before the late 12th century.

(3) Only a few small cemeteries with Corded Ware cultural background are reported from this region prior to the Late Bronze Age.

(4) As a matter of fact, not all farms and villages were subordinated to the vakus-institution (Lang 2002) and, therefore, the real number of vakuses was to some extent smaller.

(5) A cremation without cranial bones was also discovered in the stone cist of grave IIA nearby, radiocarbon-dated to the final Late Bronze Age (Lang 2000, 99). Different post-mortem treatment of the head and the body was characteristic of the Early Metal Ages in Estonia (and elsewhere) and it seems that it was shared by both the burying in monumental stone graves and outside.

(6) A thorough overview of prehistoric cremations in southern Estonia is produced by Anti Lillak (2006; 2009). In addition to the cremations on the ground there is also some ethnographic data about the cremations in clay urns put on the top of posts and covered with small roofs; such a custom is reported from the eastern neighbouring areas of south-eastern Estonia (Piho in print). How widely and when this custom was spread is still unknown.

(7) The osteological assemblages of Estonian settlement sites have not been comprehensively studied, and therefore we cannot give here any thorough survey of the findings of human bones. Some teeth and a burned piece of human skull found at Viking Age Linnaaluste I settlement site and two fragments of human humerus from the Late Iron Age Keava hill fort can be some examples, for instance Maldre (in print). The phenomenon of exhumation (i.e. digging out the deceased's remains) in the culture of death is considered by Arukask (this volume).

(8) 'Ordinary' graves (tombs) are found there only in small clusters (ca. 2-20 graves) and show considerable variation in burial rites (Crubezy et al. 2006); thus, the majority of the dead is 'lost' for archaeology also in this region.