Holy groves in Estonian religion/Hiied eesti muinasusundis.

Jonuks, Tonno

The most important archaeological sites connected with prehistoric and historic religion--holy groves (in Estonian hiis or hued in plural)--have attracted academic interest only relatively recently (for example Remmel 1998; Kutt 2004; Valk in print).

Still, there has been interest in the subject of groves for quite a long time. So far the researchers have usually examined the groves together with other holy offering places, thus considering the groves to be merely one subtype of a wider sacrificial site. In some sense the groves could be treated in that way, but in the current article I concentrate on groves alone, by choosing the toponym with the stem of hiis on landscape and presuming that the term hiis means something more specific that cannot be compared with single offering trees, stones or other such places. However, the discussion of the definition of hiis and different historical processes resulting in the distribution of the concept of hiis has been left aside. The current article focuses on first, the dating of the grove-sites, and second, their religious interpretation.

History of the study

The first to take an interest in Estonian and Finnish hiis, were Baltic German and Finnish scholars, who were strongly influenced by the Enlightenment and antiquity. The oldest etymology, proposed for the word hiis, was presented by the Finnish priest Kristfrid Ganander in his "Mythologia fennica", published in 1789, where the author connects Finnish hiisi with Egyptian Isis (Pentikainen 1995). Such a sound-based association did not become established in academic world but is still noteworthy as the first etymology ever made.

The main sources for the 19th century Baltic German researchers in Estonia were the medieval chronicles, which became the most important subject in studying Estonian history as a whole in the 19th century. Still, chronicles offer very little data about Estonian groves and so contemporary folklore had to be used. Due to the heyday of national romanticism and idealising the cultures of ancient Mediterranean and the North, Greek and Roman texts also became important for drawing parallels. The most influential scholar to pursue such a style was the Baltic German historian and enlightener of Latvian origin Garlieb Merkel (1798) whose interpretation of Estonian groves is still popular. According to him, a holy grove is a nice oak copse, presumably on top of a hill where sacrifices to pagan gods were brought. As chronicles were the most important source material in these days, a famous text about 'beautiful forest' from the Chronicle of Henry of Livonia (see below) became one of the most fascinating materials to analyse. In 1836 a priest from Kadrina church Georg Magnus Knuppfer (1836) published a paper, supposing that the hill and forest described in the chronicle is the hill of Ebavere in the vicinity of Vaike-Maarja in Virumaa.

The nobility of German origin was not the only one to support the romantic movement. Also Estonian intelligentsia, who started to emerge in the 19th century, wanted to show the high culture of pre-Christian Estonians. As the yardstick of culture was actually Christian and/or classical culture, depending on the scholars' preferences, respective elements were "found" also in Estonian prehistoric culture and religion. As a result of these activities a vision of Estonian pagan religion was formed in the 19th century as the cult of nature gods based on polytheistic pantheon, where rituals took place in holy oak groves, sacrificies were brought to gods and dead people were burnt on pyres.

Starting from the mid-19th century, organized collecting of folklore was started in Estonia. During the process, scholars (Jakob Hurt, Matthias Johann Eisen, Jaan Jung) travelled around countryside and collected lores, but texts were sent to them also by local intelligentsia, such as teachers, priests, educated peasants. On the basis of this database, the first serious studies were made about groves. The main problem was that the vision of ancient groves was already deeply settled and it was constantly reused in contemporary textbooks and articles in calendars. The impact of these texts spread into folklore and were used as a source of grove studies. Another, and even longer-term influence was the national awakening. In this context Estonian prehistoric ancient religion was opposed to Christianity as brought with sword and fire, thus the 19th century tradition of groves became the most important ideological monument of ancient and free Estonia.

At the end of the 19th--beginning of the 20th century, folklore regarding groves was studied for the first time and initial results were presented (Jung 1879; Eisen 1920). Jaan Jung, a teacher and a classicist claimed, according to the results of his folklore collection, that groves were places on the hilltops where people came from long distances to bring sacrifices. In addition to groves every farm had their own sacrificial yards (Jung 1879, 37). The most important result concerning the study of groves was achieved by Jung by the 1890s when he collected information about prehistoric sites in Estonia and registered in this process also many detailed descriptions of grove-sites (Jung 1898; 1910).

Mostly older treatment of groves remained, according to which groves were considered to be a forest on top of a hill, although much wider understanding of groves was spreading as well. It was determined that groves could also have been islands in bogs, wet places, but also areas on completely flat ground. A summary of grove folklore was presented, where mostly prohibitions were stressed: it was prohibited to cut trees, pick berries, swear and misbehave (Eisen 1920).

Until the first part of the 20th century, groves were interpreted mostly as sacrificial places. Folklorist Oskar Loorits was the first to connect the grove with the cult of the dead and the fear of the dead, and suppose that at the beginning grove meant a grave site: "Der Hain hier auf der Erde vertritt also die alteste Totenwelt des uralischen Kulturkreises" (Loorits 1957, 12). Differently from previous researchers, Loorits paid attention to the motif in folklore, according to which the dead were buried in groves and people went to a grove to communicate with their ancestors. Behind Loorits's idea--connecting a grove with the fear of the dead and ancestors--is most likely a general trend in the European religious studies in the early 20th century which considered death and the cult of the dead the most important phenomena of religions, especially prehistoric religions. In the studies of Loorits, especially in earlier ones, grove/folk religion and church/Christianity were clearly confronted, with a grove defined as an idealised natural sanctuary which "is not harassed by stone walls" (Loorits 1932, 23).

So far the most important study of groves was completed by the Finnish researcher Mauno Koski (1967; 1970; 1990). Koski's main subject is etymology of Estonian and Finnish word hiis, but also folklore concerning the hiis-sites. According to Koski, who was strongly influenced by Loorits, groves were originally graves or they were in the same sacred sphere as graves, only later did groves develop into non-Christian cult places. Koski dates the groves to the end of the prehistoric period, when the grove was a cemetery and a cult place of one village or group of villages (1967, 85). More precisely Koski dated groves to 800-1100 AD, without dealing with the subject more deeply. During medieval times, groves were ancestors' cult places and other ritual practices were increasingly held in groves (1990, 432). Showing the decreasing role of groves during the Middle Ages, Koski points out that until the 14th century settlements and groves were closely situated, but after the 14th century settlements were no longer related with groves (1990, 415).

In choosing the site of a grove, local landscape played an important role and according to Koski groves were the most dominant landscape elements (1967). Koski also shows that a grove has two semantic levels--primary as a burial site and secondary as an exceptional natural place as a sacrificial site (1967, 101). According to Koski, the original centre of grove-tradition lay in northern and western Estonia and south-western Finland, meaning "cult place where the dead, the spirits of the ancestors, were worshipped" (1990, 432).

The most important elements of groves according to Koski are stones and thus it becomes clear why Bronze Age stone-graves (hudenkiuas in Finnish) appear in groves: as groves were situated outside the villages, they overlapped with the areas where graves were erected and so graves were integrated to the grove-tradition (1990, 429). But in conclusion Koski leaves the question unanswered and supposes that groves could have been used already during the time of older stone graves, "but we cannot determine at which point the term first came into use" (1990, 409). I will turn back to the source criticism of Koski, the most important author of this subject, later.

Finnish historian of religion, Veikko Anttonen (1992), agrees with Tette Hofstra (1988) that the word hiis is derived from Germanic sidon (side) and originally marked 'side, towards, seashore'. Also mets (forest) originally had the same meaning. For Anttonen the most important aspect is the sacredness and the border of grove, and together with these, the liminality of the grove. So grove has "originally been the name which characterizes and explains the physical essence of the place" (Anttonen 1992, 2523). Anttonen also argues with Mauno Koski and claims that interpreting groves as ancestor cult places is no longer valid (ibid., 2521). Considering the overinterpretation of death cult in Estonian and Finnish prehistoric religions, it is definitely justified. Anttonen claims that the dominating landscape of groves is not related to the choice of place for a grave but rather with the sacred-category (ibid., 2525). A sacred area was also used as a cemetery.

Estonian scholars have mostly avoided grove-subject after the major study of Oskar Loorits. On the one hand it is definitely connected with condemning religious studies during the Soviet atheism period. On the other, it is related to public opinion of groves and ideological pressure, according to which groves mark the religion of ancient and free Estonians. And this is the pressure which without doubt leaves its traces on the studies.

Heiki Valk was the first to connect groves with archaeological material (Valk 1995). He stressed that there is no connection between groves and graves (considering graves from the end of prehistoric and medieval times) and argued with Oskar Loorits and Mauno Koski, assuming that groves have been completely separated sites on landscape and their original meaning was probably not connected at all with the dead or burial site (ibid., 461). As Valk has dealt mostly with Late Iron Age and the Middle Ages, he has stressed the importance of natural holy sites also during the medieval and modern times. Medieval holy sites have been divided into three, according to Valk: 1) Christian parish churches, 2) non-Christian holy natural places/objects and prehistoric stone graves, 3) semi-Christian chapels, crosses and village cemetries (2004, 300). Earlier scholars have tackled the problem of choosing the grove site very briefly and only emphasised that groves were the most dominant landscape features. The last study of Valk (2007) deals exactly with this problem and presents the possibility that the decision to pick out a certain site for grove could have been made according to energetic fields, but we cannot study them here. Considering the variability of grove sites, especially those that are not situated on dominant landscape features, this approach is definitely justified.

Folklorist Mari-Ann Remmel, editor of the only monograph on Estonian groves (1998), focuses on publishing the source material and on analysing some folklore motifs. On the meaning of groves, Remmel agrees on the basis of Estonian material also with Anttonen and the conception of the sacred presented there. However, Remmel stresses connections between groves and graves in folklore and draws attention to the lore motif about feeding dead souls in grove (Remmel 1998, 18). But this is a tradition which can appear also without burying into the groves. For example the Udmurts have commemorated their dead souls in several places, including in lad, equivalent of hiis, without burying there (Lintrop 2003, 190).

Unfortunately the subject of grove has not attracted scholars in neighbouring countries, although the topic has come into focus in Sweden during the past couple of years (see e.g. Brink 2001). Still, few excavations have been done on land-sites (grove-sites) (see Andersson et al. 2004). Besides some studies about the connection of groves and early Christian sites (e.g. Fabech 1999) an archaeologist Nina Ingren (2005) has assumed semantic differences between the land on hills and flat grounds. She also stresses bans in land-folklore. Importantly she shows the difference of time and argues that groves could have gradually changed their meaning. Another recent study about Swedish land (Oostra 2006) dealt more with later, medieval and modern processes about land, showing the dynamic understanding of groves and how these sites turned from sacred sites into parks.

The studies of holy places are currently most efficient in Latvia and Lithuania. Juris Urtans (1988), Latvian archaeologist, has investigated several sacrificial stones and argues that the oldest of these were used already during the Early Iron Age but many of them, including stone idols, can be dated as late as 16th-18th centuries. Sacrifices in these places had mostly a personal meaning, although public rituals have been conducted, too, which Urtans connects with the cult of the dead.

In Lithuania, 70 ritual sites including sacrificial stones, hills, springs, etc. had been investigated by archaeological methods by 2003 (Vaitkevicius 2004). Studying holy sites, called alka/alkas, was started by archaeologists already in the first half of the 20th century in Lithuania (Sturms 1946). According to the most recent study (Vaitkevicius 2004) it is possible to see both a person and a group of people, a village and a group of villages behind the grove tradition. It means that interpreting "cultic" places has become more complicated. Unlike Estonian and Finnish tradition, connecting hiis with the cult of the dead, seems to have developed only during the last years in Lithuania. However, despite many studies the chronology of holy sites in Lithuania is still unclear. At least some sacred stones were taken into use around the turn of our era, but the majority of sites are dated to the extensive period of 1st-2nd millennia AD.

Sources of studying Estonian groves

Similarly with the study of Estonian prehistoric religion as a whole, quite a wide source basis has been used for studying groves. Although folkloric sources have been used most in grove studies, the researchers of different periods have set the emphasis on different kinds of sources.

Written data about Estonian groves

The earliest and probably the most famous description of a grove dates from the beginning of the 13th century, when the chronicler Henry of Livonia describes how two priests baptized Jarva and Viru counties in 1220 and how they "... baptized three villages on the border of Virumaa, where there was a hill and a pretty forest where the local people said that the big god of Osilians was born who was called Tharapita and who had flown all the way to Osel from this place. And the other priest went, breaking the figures and faces of their gods and they were surprised that they did not bleed, and believed more the sermons of the priests" (HCL 1982, XXIV, 5).

Description of priests cutting down figures and faces (imagines et similtudines) at the groves caused a big discussion. The expression is a loan from Genesis I: 26--"Then God said, "Let us make man in our image, in our likeness ..." (faciamus hominem ad imaginem et similitudinem nostram ...)". Oskar Loorits (1949, 178) assumed that these figures did not depict gods but the souls of dead people. It is the only text which describes figures at groves while most descriptions involve trees. Figures have been mentioned only in some folklore texts, (1) written down at the beginning of the 20th century. But it is likely that the god figures are secondary motifs in these texts and were already influenced by written texts from the 19th century textbooks, calendar articles and other national romantic approaches.

Throughout medieval and modern times groves have been mentioned only in connection with descriptions of local people worshipping idols. In all of these, the German term (heilige) Hain is used. As such texts concentrate only on documenting groves and bringing sacrifices to them, at the same time mentioning worshipping of stones and trees (see Sild 1937), they do not offer anything new in understanding groves. In the descriptions of groves, trees have been considered the most important component, which is referred to by German (heilige Hain) and Latin (locus sanctus) terms, but usually nothing more specific can be read from these descriptions. But a general characteristic motif in most of these texts is the ban on damaging groves. Similar depictions have been presented also in the 18th century texts of August Wilhelm Hupel. The latter in its own context is especially valuable as Hupel was the first to describe groves and what was going on there in ethnographic way (Fig. 1).

"In some places there is one, in some there are many trees--mostly spruces; these are on hills, fields, by the springs and other places. Peasants who are not afraid of punishment if it came out, bury their dead people in these places. It is strongly forbidden to go in these holy groves and to worship these. Some landlords have demanded peasants to cut the trees but with all the threats and admonitions achieved nothing and had to take an axe at last and give example to the fearfuls. Sacrificing of wool, wax, yarn, bread and other things is still a custom among them; they put these gifts to holy places or into the tree hollow. Also rivers and springs receive gifts" (Hupel 1774, 153).

Several 17th-18th century sources also mention stone crosses and chapels next to groves and stones as places of superstitious worship (see Eisen 1920, 45); it is likely that for priests who wrote down these notes, it did not make any difference if the rituals were conducted in prehistoric sacrificial places or in pre-Reformation sites. There is a possibility of course that Catholic crosses and chapels were erected in older holy places in order to bless them. Most likely not all places in the description of A. W. Hupel can be considered groves. It is very clearly mentioned in the text that people did bury in groves. But as little proof has been found about Early Modern Age burials in hiis-sites known in folklore, it is likely that in addition to groves Hupel has at the same time described stone-graves and medieval rural cemetries, where trees were also holy and all kind of damage was prohibited (see Moor 1998).

[FIGURE 1 OMITTED]

In the 17th century the word hiis appears in written sources for the first time. In 1694 a catechism was published in North-Estonian language. Describing the sin against the first commandment: "I am the Lord your God, who brought you out of Egypt, out of the land of slavery. You shall have no other gods before me", a sentence is added: "If a man gives the honour, which is meant only for God, to those, who are not worth it if figures will be worshipped, dead Saints called, gifts brought in special times, Earth-Mother or Under-Earths worshipped, some places, hills, stones, springs, steam of sauna, groves or trees in forest considered to be holy, bringing them sacrifices or in some other ways serving idols." (2)

It is a textbook for school and confirmation students in North-Estonian language, written after the original in South-Estonian language from 1684. Unfortunately the original publication has not survived. According to the context it is likely that in the 17th century all these described beliefs were alive and this passage is written precisely against the worshipping of springs, stones, etc. Surely the author of this publication, J. Hornung, must have been familiar with these beliefs and has used them as examples.

Archaeological sources

Archaeological sources have been least used for studying holy groves. It is well grounded as there is no cultural layer and also stray finds have seldom been detected in groves. So we are not able to study the groves by ordinary archaeological methods.

Finding coins from groves and sacrificial sites has been most often mentioned. Unfortunately almost all of these reports are given by amateur archaeologists and thus no proper information is available. Another problem is that these coins have not been preserved as they are immediately allowed into the circulation of the collectors without any documentation. So the groves are completely uninvestigated from this point of view. Even if it is impossible to collect all coins found from groves, it is necessary to collect at least the data about finding those which would give necessary information about using groves in medieval and modern times.

Single stray fords from the groves date from the end of prehistoric times (Fig. 2). The problem with these also lies in the documentation. So far all stray finds have been obtained while destroying the grove hills. As most of the groves are moraine hills, many of them have been dug to get gravel for road building. So the find context of the known stray finds is not clear, nor do we know if these items originate from graves or sacrificial context. As the majority of findings are Late Iron Age brooches, either explanation can be valid.

The dating of these stray finds could be noteworthy. In addition to some single finds from groves, there are also some items found from springs, known from folklore as sacrificial springs. All these items belong to the last centuries of prehistoric times, like fords from groves. The oldest finds from springs are two spearheads from Koorkula, dated to the 8th-9th centuries (Tamla 1985). Most of the items, including bracelets, pendants, etc., however, date from the following centuries. Such date, the last prehistoric centuries, fits also with Latvian sacrificial springs ([TEXT NOT REPRODUCIBLE IN ASCII] 1988, 11). The lack of medieval finds is characteristic as well, whereas finds--mostly coins--appear again during the Early Modern Age. So the date which the present state of archaeological research can provide, fits with the mainstream chronology of Estonian and Finnish grove-sites as belonging to the end of prehistoric times (Koski 1967; Valk 1995). Still, the presence or absence of stray finds is not a proof of using or not using groves in some particular period. Despite the lack of medieval stray finds it is very likely that groves were used in some way also during medieval and modern times (see Valk 2004).

[FIGURE 2 OMITTED]

Another problem in studying groves with archaeological methods is that compared to other sites, in case of groves we are not sure what to look for. It is clear that stereotypic understandings of "sacrifices" is not relevant. Considering the contemporary and recent offering practices, people have offered pieces of cloth, glass, etc. (Viidalepp 1941), but these are definitely items which could easily be regarded as waste even during archaeological fieldworks. So it is important to reconsider what kind of archaeological find we are expecting from a grove.

Archaeological investigations have been carried out so far on three sites with hiis-toponyms in Estonia: Sammaste in Viljandi county (Valk & Maesalu 2006, 140), Palukula in Rapla county and Kunda in Viru county. In addition to these there have been some investigations around sacrificial trees and chapels (Valk 2006b, 214). A brief survey has been done on some hiis-sites in Virumaa (Torma, Antu, Aburi, Mula, Kolu, Tammiku, Vaekula) but during the survey no cultural layer or stray find was revealed.

Investigations at Sammaste were concentrated on a stone-grave at a hiss-named place. The grave was dated to around the 1st century AD, but the grave was used for burials also during the Late Iron Age and the Middle Ages. In addition to the grave a wooden chapel was built during the Catholic time and the site was used for offering also in the Early Modern Age, which is referred to by many 16th-17th century coins.

In Palukula a stone fence was investigated, which was assumed to be connected with the grove-tradition. Charcoal was collected under the wall and according to [sup.14]C analyses it belonged to the end of the 18th century. Questions still remained unanswered as it is possible that charcoal came from some later fire and the fence itself is earlier. The assumption is supported by the fact that the fence is heavily buried under later erosions. The function of the fence remained unclear as well: it could be connected to a probable field next to it or to the grove as some kind of symbolic construction. Also many folklore texts refer to fences or walls surrounding the groves. Stone walls and constructions have been found also in Latvian and Lithuanian holy sites (Storms 1946, 19-20).

At Kunda grove hill together with the excavation of a stone grave, an extensive survey was carried out, but nowhere on the hill was it possible to detect a cultural layer or anything else that could be investigated by archaeological methods. Still, a line consisting of four stone-graves was detected on the slope of north-eastern side of the hill.

In analysing groves in connection with relevant archaeological sites they have often been considered to be related to settlements from the late prehistoric or medieval times (Koski 1967; Valk 1995). There can be several reasons, among which the most important one is that mostly only graves are known from earlier periods. Comparing groves with settlements fits also well with anthropological parallels.

It still seems possible that besides villages also older stone-graves, from the Late Bronze Age up to the Roman Iron Age (1100 BC-AD 500) could be connected with groves. Although there are several reports about such graves beside the grove, no systematic investigations have been yet carried out. The connection between stone-grave and grove can be followed on several sites, for example in Torma, Purtse and Iila.

The connection between the grave and the grove has been sought in a direct way so far, i.e. the graves have been looked for within the grove. However, most studies of landscape archaeology show that ritual sites are hardly isolated on the landscape, but rather they constitute a whole, a system of other archaeological sites and also natural elements (see more Ingold 1993; Barrett 2000; Bradley 2000; Garwood 2003). Vykintas Vaitkevicius has argued that the ritual places in Baltic prehistoric religion could be related to each other by myths (Vaitkevicius 2004, 48). Following the location of the grove-sites familiar from folklore texts on the landscape, the connection between grave and grove does exist, which we will see more closely below. However, this system is not absolute and there exist many examples which show that some groves or graves have never been connected with each other.

Linguistic sources

Linguistic sources were employed more widely starting from the studies of Oskar Loorits. Although the first etymologies about the origin of the word hiis had been done before, only by the mid 20th century theories of Finno-Ugric languages were developed enough to form the theoretical base to study the origin and age of the discussed stems. Most of these studies were made according to migration theories. The most important study so far is also based on linguistic sources (Koski 1967; 1990).

The word appears in two forms: hiis and iis. The 'h' at the beginning of the word comes from North-Estonian language, from where it spread to other dialects and was attached to original iis. So it does not have a semantic meaning. (3) Still we can find a dating nuance. If we consider the origin of the word in Scandinavian hiidi, hiipi (Koski 1970, 246, see more about etymology below), it could be suggested that the loan had to occur before our era or at least during the first centuries AD, when there was no 'h' at the beginning of the word in Finno-Ugric languages (Sutrop 2004, 51). I return to the problems of dating of the grove later in the article.

Although grove-like places could be found in large areas, it is important to follow the spread of the word hiis. We can argue that behind the same stem is a notion with a similar meaning, rituals and symbols. Speaking more generally, hiis-stemmed word spread in northern and western Estonia, south-western and south-eastern Finland, Karelian Isthmus and around Lake Ladoga and on the western shore of Lake Onega. (Fig. 3) At the same time hiis-stemmed word did not spread in Livonian language, neither is it known in South-Vepsian, actually the word is not important in Vepsian at all.

As said before, in Estonia hiis-stemmed toponyms spread mostly in northern and western Estonia. There are sacred places in central and southern Estonia as well, but single hiis-toponyms there are considered as secondary loans (Valk, personal communication). Only parishes of Hargla and Rapina are emphasised as areas where hiis-stem has originally spread (see Sarv & Vladokin 1988, 154). It is difficult to make more precise generalizations. There are very different landscapes with hiis-word, including hills, but at the same time places on completely flat ground. Although there are several exceptions, groves are generally unusual places. For example karsts, bog islands and other places, exceptional and anomalous in this particular landscape could be chosen for grove (cf Anttonen 1992). It is possible also that some other criteria were important in choosing grove sites that we do not know, such as energetic fields (Valk 2007).

[FIGURE 3 OMITTED]

Etymology

Another important subject besides the spread of the word is its etymology. No final etymology has been proposed although (or may be exactly because of this) the etymology of hiis has been a subject of many scholars (e.g. Koski 1967; 1970; Hofstra 1988).

In principle, etymological directions could be divided into two. The first direction suggests North-Germanic as the origin language and two possible original stems are stressed. First of them is hiidii, hiipi, and its main translation in all North-Germanic languages is a nest, a place to hide, rest or lay which is thus also related with a grave--the corpse in a grave resembles an animal in a nest. The word also has an additional meaning, i.e. a bouldered place, thicket, hill or ridge with stones. In some Swedish dialects it also marks a place with a negative meaning, like a hiding place for outlaws and criminals. Often the word appears in toponyms marking a wild area, a place out of the ordinary world (Koski 1967).

Another Germanic word considered as the original stem is sid, sidon, with the meaning of side, brink, also seashore (Hofstra 1988). This etymological direction has been stressed by Veikko Anttonen, who assumes anomality and liminality as the main characteristic features of groves (Anttonen 1992).

Another direction of etymologies has used Saami languages as the source language, and two stems have been suggested. According to the first and the most widespread tradition, the original form of hiis has been reconstructed to the form sieida/sejte (Kulonen et al. 2005, 391) with the meaning 'stone or rock which is the object of worship; sacrificial stone; sacred stone considered to be a god figure; sometimes even god or spirit who lives there'. Although the meaning of sieida is very close to hiis, Mauno Koski draws attention to the semantic difference: sieida is itself the object of worship and not the place where to do it. And another important phenomenon--sieida has never been connected to the cult of the dead, which was an important part in Koski's interpretation of groves (Koski 1990).

The second stem from the Saami language, considered as the original is sii'da: village, camp. Sii'da is phonologically closer to the word hiis, but the meaning is very different. There is a possibility to reduce the word sii'da to its original meaning sej, with the meaning of 'dance step' and in this case the original meaning of the word would be the place for (ritual) dance (Kulonen et al. 2005, 392).

So the etymology of the word hiis is ambiguous. Semantically it would be tempting to connect hiis with either of the Saami words. But besides semantic differences, there are also phonological differences, which do not allow to make such a connection. Direction from Scandinavia fits better both the semantics and linguistic geography.

Folkloric sources

Folkloric sources have been most used in studying groves. Already the first results at the end of the 18th century in Mythologia Fennica by Cristfrid Ganander were based on folklore. So folklore could be considered the best source material (see also Koski 1967; Remmel 1998; Kutt 2004), but I will not examine them in more detail. Still, it does not mean that modern studies of the folklore motifs are not needed. Definitely several motifs of the grove tradition need new and more thorough treatment, especially from the point of view of source criticism.

There are several problems in using grove-lore in analysing prehistoric religion. Some difficulties, especially the national romantic motifs, have been already stressed (Kaasik 2004). The grove-tradition mostly includes prohibition-motifs, which forbid to cut trees, break branches, etc. Similar grove-lore is known also about contemporary holy places, like kusoto among the Mari people (Toidybek 1998) and lud among the Udmurts (Lintrop 2003).

Although grove-lore has been much studied, it is usually interpreted directly and without any source-critical approach. Lore has been considered authentic, something which shows us adequately the behaviour of people and their attitude regarding the grove. In recent years some new movements have appeared in this discussion. Ulo Valk has shown that grove-lore, which has usually been directed to the past, does not reflect "direct attitude to nature, but rather the idea of how it should have been" (Valk, U. 2005, 40). These kinds of bans--not to cut trees and bushes, pick berries and in any other way disturb the grove--were not for everyday compliance but showed how people should behave in a grove. The fact that customary law does not derive from what people actually do, but reflects other issues, has been demonstrated also by historians of law (Watson 1995). In some ways it is similar to the whole process and aim of folklore--to present the ideal world (see Honko 1998). Ideal culture was defined by Lauri Honko as it "does not present the collective only from inside out. It also works in the community as a control mechanism. While studying folklore process, it has an important role, among other things, in stressing central values as well as key-symbols, mythologies, rituals and explaining social value-system. Ideal culture is one way to define group identity" (Honko 1998, 78). So with these grove laws, people defined their relationship to grove every time, and thus they do not present the rules of how people actually behaved in a grove. I am far from saying that it was normal to cut the trees in a grove, rather that in further studies the source criticism should be taken more into account while observing grove-lore.

Another problem that should be considered is the changing of the grove-lore. The tradition and toponym could get lost in time in some places (cf Fabech 1992) or be secondarily attached in others. A good example for secondary grove-lore can be found in Virumaa, a centre of grove-tradition in Estonia.

"Neeruti Sadulamagi (Saddle Hill) was initially a holy grove, later there was a hillfort" (RKM II 196, 490/1 (3)).

The hillfort of Sadulamagi has been dated to the late first millennium AD and it is very likely that grove-lore was attached to the hill only in the 19th-20th century, due to the general national romantic mentality according to which there had to be a grove in the vicinity of every village. Unfortunately, no general rules can be suggested to recognise the secondary grove-tradition.

In conclusion we cannot avoid using the grove-lore, it is still an essential source. Differently from several previous studies, here lore has been used only as a toponym. I believe that grove-lore concentrated in places which were important for a long time. Even so long that the content of the present grove-lore might have nothing to do with the original. But still the lore has remained on landscape in these important places and this gives us the basis to study them from their sacral meaning.

Other sources

Other sources have been used comparatively less, and not systematically. One of the more efficient could be working through the material of historical maps. On the basis of this method, Gustav Vilbaste supposed there was a grove-site in the village of Saunja where only the toponym Huealuse has survived. However, analysing the 17th century maps, Vilbaste thought the initial grove was next to Huealuse and demonstrated how on the village map of 1693 the grove-site was marked as a thicket near the village, encircled by a stone fence and next to the farm of Hihe Hans (HUe Ants) (see Vilbaste 1947).

Dating

After Mauno Koski (1967) no other scholars have tried to date the concept of hiis. As there are very few archaeological finds and they do not offer any adequate dating possibility, no archaeologists have studied the subject and according to the research of Mauno Koski, groves have generally been regarded as the phenomenon of late prehistoric times.

There are several problems with the dating Koski offered. The original version comes from his major treatment from 1967, when migration theories were prevailing in Finnish archaeology. So there was a theory according to which a migration emanated from Karelian Isthmus in the 8th century and together with that several new influences were brought to the southern part of Finland (Kivikoski 1961). As the archaeological material of this time, or rather its interpretations, did not allow any other possibilities, it was likely that grove tradition was reputedly formed on Karelian Isthmus and spread from there to Finland and Estonia. And besides, such a late date fitted with other sources. The medieval chronicles confirmed the worshipping of groves, thus the custom had to exist at least at the end of prehistoric times. What is even more important--there are many archaeological sites known from that period. Most indigenous villages in Estonia and Finland were established during the Viking Age or later (Lang 1996). At the same time rather few sites are known from earlier periods and these are predominantly stone graves. In addition there are also ethnographical parallels from Finno-Ugric tribes in Russia with a continuous grove-tradition where it is a general pattern that groves are in the vicinity of villages, not in the village and not far from it. So Udmurtian groves--lud--are situated at some distance from the village but are still connected with them (see Lintrop 2003). This gave more reason to look for the connection between groves and late prehistoric-medieval settlement sites. And this connection does exist. So it was confirmed that grove-tradition was alive during the late prehistoric times.

Archaeological studies in the past few decades have shown that there was no migration from Karelian Isthmus to Finland. In addition, nothing like this has ever been claimed about Estonia. So the first, and the strongest part of the statement is not valid any more. The connection between groves and settlement sites can also be the reflection of the concurrence of different processes what will be discussed more closely in the next chapter.

Discussion

According to anthropological parallels, almost all indigenous societies have used different religious places to conduct their rituals (see for example Lintrop 2003; Jordan 2003 as parallels). Using different places has mostly been caused by different deities and spirits that were worshipped. Thus these ritual places can be divided according to their function and social role. In addition to personal/ communal rituals these religious places can be differentiated on the grounds of areal distances, where the grove might be important for the people of one village, for a bigger territory or even a set of different territories (see Vaitkevicius 2004, 51).

Different functions and dates of holy places can be the key to explain the diversity of our groves and other offering sites regarding their looks, as well as sites connected to them. In addition we should consider offering springs, trees and stones that are not discussed in the current study. This means that the ancient religious places should be interpreted starting from the wider religious background. This is indicated by the folk tradition of the 19th and the 20th century which includes several texts describing how the village community carried out their common rituals in the groves but personal practices were conducted in offering gardens adjacent to every farmhouse (Jung 1879, 37).

In order to briefly summarize some of the points relevant for the discussion, the following should be emphasized--while dating the grove-sites, their connection with the settlement sites of the end of the prehistoric times and the Middle Ages has been emphasized and the rare connection with contemporary cemeteries has been pointed out. However, at the same time the connection between some graves and groves is referred to, as well as the role of the groves associated with the dead, which can be seen from the folk tradition. Owing to the present dating, the researchers have not paid attention to earlier stone-graves. Only Mauno Koski has mentioned the Bronze Age stone graves in the Finnish context (hudenkiuas), explaining it with the tradition of locating grove-sites in places that were earlier used to erect stone graves since stones and stone heaps played an important role in the groves (Koski 1990).

In my paper I determine three main groups of groves: 1) groves situated on higher locations and associated with stone graves from Late Bronze and Pre-Roman Iron Age; 2) groves connected with indigenous settlements where association with graves and outstanding landscape is not important; 3) groves situated outside the everyday world on bog islands, deep forest, etc. where separation and liminality seem to be important factors. Of course this classification of grove-sites does not aim to be adequate and reflect the dating and function in the most precise way. Nevertheless, this is a possibility that can be used as the basis of the following assumptions. The purpose of this paper is not to create a typology of holy groves. The connections I am trying to show indicate at rather general trends, but not distinctive types. Of course, creating links between different sites from different periods and different societies is speculative and difficult to prove. However, the following associations seem to appear when we look at the sites on landscape.

Hiis-sites connected with stone graves

Following the concurrencies of graves and grove-sites, the most apparent connection occurs between the groves and the Late Bronze Age stone-cist graves and the pre-Roman Iron Age tarand-graves. However, not all stone-cist and tarand-graves can be related with grove-sites. Their connection is apparent in several cases but is not conclusive.

Regarding the location of the stone graves of Estonian Bronze Age and the earlier part of the Iron Age (until the 5th century AD) on landscape, what catches the eye is their concentration on the areas that are visually outstanding, rising from the surroundings by prominent land formations (klint slopes, hills) or having an unusual background (karst). It is clear that they include many exceptions but the location of the sites is remarkable. Until now it has been explained with the need to dominate, to emphasize that the land belongs to the family and kin that uses the grave (Ligi 1995, 216), or with the fact that people chose emotionally powerful landscape which created a sense of holiness, thus when erecting the grave the connection between the people and the holiness of the landscape was emphasized (Lang 1999).

Picking out Virumaa as one of the most important centres of the grove-tradition, in several cases it is possible to observe the concentration of graves in the neighbourhood of the grove-sites known from oral tradition. While the earlier researchers have looked for graves in the groves, it could be stated that this connection is not there. At the same time it is not very likely that landscape features exist in the tradition and religion isolated from the surroundings. Several studies on the British Isles, Scandinavia and elsewhere indicate that many objects have been considered significant at the same time, and the visual contact between the sites has been important. For example an area covered with stone graves and other sites has been integrated into the Stonehenge tradition in the radius of several kilometres (Parker-Pearson et al. 2006). Thus it is likely that in the case of Estonian groves, the location of the grove, as well as its surrounding have been considered vital as well. It is likely that the majority of the data initially emphasized with the groves, cannot be observed any more. But it is probable that one of the important aspects were the graves. Especially when we bear in mind the interpretation that has become popular only lately regarding the stone graves with constructions as the place for conducting rituals and not the place to bury in the first place (see Lang 1999; 2000; Magi 2005).

Not all traditional groves are connected with graves. At this point it is possible to see the relationship between those graves and groves that are outstanding formations on the landscape. But this connection is not absolute and it is possible to find several exceptions, but the connection between older stone graves and groves on the hills seems to be a general trend. Different interpretation is needed when analysing the groves that are located on completely flat ground and which do not have any prominent natural site in the surrounding.

Following the concurrencies of graves and grove-sites, the most apparent connection is between the groves and the Bronze Age stone-cist and the Pre-Roman and Roman Iron Age tarand-graves. However, there is no statistics and thus no measurable connection can be presented. I believe that creating of this adequate and wide connection is not even possible since we have to consider that the graves and the tradition can become lost in time. The earlier graves, from the Stone Age, do not seem to have any connection with known groves. We can have a look at two examples. The first comes from the Late Neolithic cemetery in Sope, Virumaa. This is a grave-field where altogether about 6 graves and in addition three stone-axes have been found (see more Johanson 2006). The grave-field is situated on flat ground without any landmarks. In the vicinity, about 2 km northwest is one of the most imposing Purtse grove-hills, together with stone graves which obviously have a connection to the hill. But no connection between Sope cemetery and Purtse grove-hill is obvious. Another example can be found on Saaremaa, where a grove named "Puhha metz" (Sacred Forest) is known from the 18th century maps (see Ligi 1984). Again, about 2 km away, to south-southwest, there is a Stone Age cemetery in Koljala, where at least 3 burials from the Middle and Late Neolithic have been found. And again--we cannot find a connection between Stone Age cemetery and a grove.

Later stone graves without superconstruction that were first built in the 5th century AD, can be associated with the groves only in single cases which has been pointed out earlier (Valk 1995). Also not all stone-cist and tarand-graves can be related with grove-sites. Their connection is apparent in several cases but not conclusive. This would mean that there is no single interpretation that can be used about stone-graves.

Analysing the material of Virumaa, examples can be found. A wonderful example of the traditional grove and a stone-grave is offered by the settlement complex of Torma in Virumaa (Fig. 4). Approximately one kilometre from the Iron Age settlement site there is a hill that clearly emerges from the surrounding wavy landscape. There are traditonal grove-stories known of the hill and there is a stone grave, although not investigated but apparently a stone-cist grave, on the southern slope of the hill. It is true of course that it is very risky to assume the type and the dating of a stone grave only by its visual shape (Lang 2000, 97, 161) but the round layout of the grave points at a stone-cist grave. In the neighbourhood of the hill, at about 500 m, there are other stone graves, one of which has been studied and the erecting of the grave has been dated to the 1st-2nd century AD. It is quite obvious that the grave on the southern slope of the grove-hill and the graves in the surroundings are not there by accident but the hill has been important when they were built.

[FIGURE 4 OMITTED]

Another example of the same kind can be seen in case of Purtse grove-hill where a long ridge of the hill ends with a klint outcrop. There is a group of graves at the foot of the hill on its western side with both stone-cist and tarand-graves (Tamla 1996). Moreover, there is an offering spring known from the oral tradition as "Uku allikas" (the spring of god Uku) in the same complex. Also it is apparent that the graves were erected bearing the grove-hill in mind.

Also a third example is offered by Virumaa material. In 2003 stone-graves were found from Kunda grove-hill but not on foot of the hill as the previous examples, but on top of the hill, on its edge (Fig. 5). The graves are situated in a row along the northern part of the southwest-northeast directed hill and have been orientated to the area where we know a settlement site contemporary with the graves.

For now a part of the first grave dated from the 7th-5th centuries BC has been excavated. The place of the Kunda prehistoric lake is situated between the grove-hill with the graves and the settlement site. (4) Even now the ancient lake site is filled with water at the time of high water in springs and has been a wet place until the middle of the 20th century. Also this complex shows that the graves were erected considering the place, which later become tradition. Observing the landscape around the settlement, grove-hill and graves, it can be assumed that the shore of the Kunda prehistoric lake was the connecting link. As mentioned earlier, the connections between the cemeteries of the late prehistoric times and the grove-sites are scanty. The grove-hill of Tammiku builds an example of this tradition with a 12th-13th century inhumation cemetery at its foot.

[FIGURE 5 OMITTED]

Considering the connections between the sites discussed above it can be concluded that the first and oldest groves were first used during the Late Bronze Age--Pre-Roman Iron Age (11th century BC-1st century AD) and in case of these separate hills or klint slopes have been considered important. It has been supposed that the location of graves and settlements along the North-Estonian klint is connected with the extraordinary landscape and the blessing gained through the revelation (Lang 1999) that can be expanded to the groves situated further from the klint.

The etymology of the word indicates that the grove-sites were first used during that period. If we consider the original stem of the word hiid and its meaning 'outlying place, stony hill' likely, then the word suits semantically, and there is no need to look for a religious charge in the initial meaning of the word. In this case the date of the word and the term suit too: the word has been taken over at the time when the letter 'h' at the beginning of a word was not used in the Finnic languages (before the first centuries AD) and the initial form of the word, that has spread in Estonia has actually been iid/iis. This date has been confirmed also by Veikko Anttonen's suggestion that the word puha, 'the sacred, holy', was spreading during the Bronze Age (Anttonen 1992) and the words puha and hiis were probably part of the same cultural complex. Anttonen's hypothesis is therefore also plausible, according to which groves were considered to be important as areas which are sacred, not as burial sites, and due to the sacredness rituals were carried out there, during which stone-graves were built inside and in the vicinity of holy groves.

It is characteristic that the grove-tradition on klint edges concentrates on these places where we see groups of stone-graves. This all indicates the possibility that the groves which have an outstanding landscape, were first used at the time when stone-cist graves were being built, that is the Late Bronze Age. The connection of the graves to groves seems to show that the function of the groves was related to the cult of the dead or ancestors, the most important role of which was probably emphasizing kin genealogies. Arguments against this connection have been pointed out too (see Anttonen 1992; Valk 1995) but these studies have regarded the places with the name hiis in general and sought a connection with the late prehistoric and medieval cemeteries.

Following the dates of the graves, it is obvious that mostly the stone-cist graves, as well as the Pre-Roman Iron Age early tarand-graves are connected with the groves. The Roman Iron Age classical tarand-graves are less related to groves, and since the Migration Period no graves have been erected in the vicinity of the groves. Thus the material refers again to the possibility that the groves connected with the outstanding natural phenomena have initially been holy places and related to the ancestor cult. Similar processes have been suggested also for Norwegian material, where the Bronze Age and Early Iron Age graves were built mostly on topographically exceptional places, often on the shores or on the coast of fjords. Since the Roman Iron Age the situation has changed and graves were built in the vicinity of settlements (Sognnes 2000).

This connection between the groves and the ancestor cult probably did not disappear after the Migration Period. Although new graves have not been erected, burying has continued into several of them and this means that these graves have been valued continuously (although not constantly used for burying) for a long time. Mostly it is possible to observe the burials from the Late Iron Age in these graves as well (Iila, Aseri). Distinguishing the burials from the Middle Iron Age is problematic since many of these have been buried without any grave-goods, which is why their distinguishing and dating is complicated. Still we might be dealing with a characteristic tradition of namely the Middle Iron Age as the absence of burials from the Migration Period in stone graves is characteristic of a wider area (Larson 2005, 111).

Burying into the graves close to the groves did not end in the 13th century either, but continued through the Middle Ages. A good example is provided by the Aseri stone-grave that was founded around the first centuries AD but where burials from the Late Iron Age and the Middle Ages have been distinguished. The most interesting artefact from the grave was a tin pendant of the shape of Antonius cross from the 15th century that depicted the Crucified in relatively clumsy style (5) (Fig. 6). Finding such a Christian material indicates that burying into the graves that were related to groves did not express the religious identity, but rather the connection with old families and kinships. This would explain the long usage of other stone graves--these graves were the markers of families and burying periodically in them was used to fix the relations with previous generations, history and via that also with the land, thus presenting the identity of the kin-group. In addition to burying into the graves, which, considering the archaeological material gathered in the course of research, happened quite rarely, the rituals of the same kind must have taken place much more frequently in the groves, but no traces have unfortunately survived from these, or perhaps we are not able to recognize them.

[FIGURE 6 OMITTED]

At this point we again come to the question that has been discussed widely --what exactly is a stone grave? It has been usually claimed that a grave is a ritual communication place between the dead and the living society (Lang 1999; Bolin 1998). At the same time it has been suggested that every new generation built a new grave, or a part of it (Lang & Ligi 1991; Lang 1996; 1999), which would mean that the connection between the previous generations would be cut off. It seems more likely to me that such graves were used for a longer period and not all dead members of society were buried there, but only a few chosen ones. Even not all chiefs or other important persons in social or religious sense could have been buried there. The choice who to bury into the grave might have been made according to rather different bases and thus the time gap between two burials could have been long. It would explain the breaks in burying into stone grave that can be observed according to archaeological material. And thus we actually have no breaks but periodical confirmations of legends and genealogy myths.

Hiis-sites connected with indigenous villages

Nevertheless the connection between the groves and graves is not general. There are many (holy) sites named hiis without graves and where graves have probably never been. These sites appear to have a stronger connection with the settlement sites from the last centuries of prehistoric times. Most of the Estonian indigenous villages have started, or at least we are able to follow them, during the Viking Age (Lang 1996), and according to the find material in Virumaa these villages existed until the Middle Ages, often until the present day. Many traditional grove-sites are connected with these villages. Aburi grove situated 1-1.5 km from the initial village centre and located on a completely flat ground with no graves known from its neighbourhood is one of several examples of the groves of this kind where the connection between the groves and the indigenous village is clearly visible (Fig. 7).

[FIGURE 7 OMITTED]

As mentioned, graves are rare in the surroundings of this kind of groves and it seems that we cannot use the same interpretation here as previously. Considering the changes in the construction of the stone graves that started in the Late Roman Iron Age (around the 5th century) when the structures of the graves disappeared and the graves erected afterwards were mostly burial areas covered with disarranged stone layer, it is possible that rituals that took place on the graves partly lost their significance as well. Thus it is possible that the rituals have started to change during the Roman Iron Age and Migration Period, regarding their contents and the place of their occurrence, and have moved to settlement sites and groves connected to these (see also Bagenholm 1999; Sundgvist 2002; Widholm 2006). Probably the rituals now do not include so many features of the ancestor cult (although these have not disappeared) but gods and aspects connected to them became more important. It is also distinctive that grove-places seem to become more "secular" now, and the subject of natural holiness is no longer important. Instead, the connection with living society is stressed more.

This kind of grove-places and grove-tradition was most likely integrated into 13th century chronicles and their reflections have reached the folk tradition of near past and today. Nevertheless, also this grove tradition preserved the rituals related to ancestors and grove-graves were continuously used for single burials. It is likely that some local differences developed and rituals connected with ancestors were still preserved in areas where the older grove tradition originally formed, i.e. in North-Estonian coastal areas. While spreading to western and central Estonia, where the older connection was not important, ideologies and rituals concerning ancestors did not become important either.

Hiis-sites away from everyday life

The latter subtype of the groves described above comprises a group of localities with the toponym of hiis that are connected neither with settlements nor with graves. These groves are situated outside the common surroundings in isolated places and are separated from the everyday world by a landscape difficult to access. They can be situated in primeval river valleys (for example Kongla), bog-islands (Varudi Big--and Small Hiis) etc. In the case of these places it is possible to follow their location close to exceptional natural phenomena, whereas isolation seems to be an important motif. While offering tradition has been used here as well, and even more than the grove-tradition, it is possible that different traditions and different functions entwine.

While the tradition for the majority of those also includes an offering motif, it is likely that these groves are the places for conducting personal rituals, the significance of which is increased by the prominence of the landscape. It is possible that these places have become important only during the official deploration of the grove tradition since the 13th century. However, the medieval grove-tradition cannot be taken in a simplified way--i.e. since the country was officially baptized then all worshipping of groves took place secretly. It is more likely that several groves were continuously considered important during the Middle Ages and rituals were still conducted there like in previous periods. The situation varied, depending on the area and the official government of the land. The main material example about the importance of groves during Medieval and Modern Ages are provided by the medieval burials in the grove-graves and coins thrown into groves since the end of the Middle and during the Modern Ages, possible offerings. Unfortunately, as we saw, our knowledge about coins from groves is very limited because the majority of them have been found by amateur archaeologists and no such coins have reached museum collections, thus making it difficult to draw any conclusions. Nevertheless it is clear that the situation could not be as it was before the official Christianization, and probably this caused the usage of these hidden grove-sites.

Conclusion

Although a big part of the main sources used to study groves, predominantly folkloristic, offer a diverse picture of the attitude towards grove-like places, using them for making conclusions about prehistory is problematic. However, it can be supposed that although the grove-lore written down in the 19th and 20th centuries itself does not date from prehistoric times, the places connected with the stories have been important for much longer than the lore-motifs themselves. Nevertheless, the dating of grove-lore according to folkloristic methods has so far not been possible. Written sources, predominantly descriptions in medieval chronicles, can be dated more precisely, but the oldest of these still belong only to the 13th century. It is acknowledged, however, that the motifs in the chronicle texts are dated to somewhat earlier period than the time of their writing down. Some help for dating and understanding the initial meaning of groves can be provided by linguistic sources but their results are too vague for deeper analysis. Archaeological sources that would be best for dating the groves are unfortunately most scarce. Until now only a few findings are known from grove-sites and even these are the jewellery from the end of prehistory, mainly penannular brooches. A totally new perspective is offered by landscape archaeology and associating of folkloric grove-sites with surrounding sites.

In conclusion, according to the current state of study, three groups of hiis-named places can be distinguished.

The earliest grove-sites were most likely first used in northern and western Estonia and the tradition probably started during the Late Bronze Age. Holy groves of this period probably situated on prominent landscapes, hills, klint slopes, etc. Communal rituals were most likely connected with groves with purpose to confirm and stress the connection between the living society and the dead and thus to connect past with the present and to create and confirm the identity of the community. Places for such rituals have supposedly been stone graves which were erected in the groves or in their vicinity.

[FIGURE 8 OMITTED]

Starting from the 5th century AD big changes in religion took place everywhere around the Baltic Sea. In Estonia, erecting of monumental stone-graves ended during the period and important developments probably took place in the concepts of soul and the Otherworld. Also the role of groves started to change then. Although a selection of previous grove-sites remained in use, as is referred by folkloristic grove-tradition, rituals conducted there were different now and these were less or not at all connected with ancestors. Although commemorating the dead members of society still remained, belief in gods and rituals related to chiefs became more important. The connection of groves and settlements was increasingly stressed. During this period it is possible to observe the tendency of groves moving further from the sphere connected with death, and the increasing importance of the living society and probably beliefs in gods as well.

The problem of groves outside everyday life is much more complicated. As these are hard to connect with other archaeological sites, it is also difficult to speculate about their date and function. It is likely that unlike previous sites, isolation, and marginality were considered important and thus these groves may represent some other tradition, not connected with either ancestors or living society.

It is clear that distinctions presented above cannot be taken as a clear-cut classification and it is impossible to create a "typology of groves". What has been presented rather indicates blended concepts which appear at different locations but can also occur at the same places on landscape.

Acknowledgements

The author is thankful to Heiki Valk for specifying remarks and complements and to Kristuna Johanson for translation.

References

Anderson, G., Jorpeland, L. B., Duner, J., Fritsch, S. & Skyllberg, E. 2004. Att fora gudarnas talan--figurinerna fran Lunda. (Riksantikvarieambetet arkeologiska undersokningar skrifter, 55.) Itiksantikvarieambetet.

Annist, A. 2005. Friedrich Reinhold Kreutzwaldi "Kalevipoeg". Ed. U. Tedre. Eesti Keele Sihtasutus, Tallinn.

Anttonen, V. 1992. "Puha" moiste rahvausundi uurimises.--Akadeemia, 12 (45), 2514-2535.

Bagenholm, G. 1999. Arkeologi ock sprak i norra Ostersjoomradet: En kritisk genomgang av de senaste arens forsok att finna synteser mellan historisk lingvistik och arkeologi. (Gotarc. Series B. Arkeologiska skrifter, 12.) Institutionen for arkeologi, Goteborg.

Barrett, J. C. 2000. The mythical landscapes of the British Iron Age.--Archaeologies of Landscapes. Contemporary Perspectives. Eds W. Ashmore & A. B. Knapp. Oxford, 253-265.

Bolin, H. 1998. Activating the monuments. The ritual use of cairns in Bronze Age Norrland.--Current Swedish Archaeology, 6, 7-16.

Bradley, R 2000. An Archaeology of Natural Places. Routledge.

Brink, S. 2001. Mythologizing landscape. Place and space of cult and myth.--Kontinuitaten and Bruche in der Religionsgeschichte. Festschrift fur Anders Hultgard zu seinem 65. Geburtsdag am 23.12.2001 in Verbindung mit Olof Sundgvist and Astrid van Nahl. (Erganzungsbande zum Reallexicon der Germanischen Altertumskunde. Hrsg. von H. Beck, D. Geuenich & H. Steuer. Band 31.) Hrsg. von M. Stausberg. Walter de Gruyter, Berlin; New York, 76-112.

Catechism 1694 = Katekismus. Onsa Lutemsse Laste Oppetus Luhhidalt Puhha kirja jarele arrasellitud, ning kussemisse ja kostmisse kombel kokku saetud. Jummala auuks, ning temma koggudusse Kasvuks. Druckts Johann Georg Wilcken, Riga.

Eisen, M. J. 1920. Esivanemate ohverdamised. Eesti Kirjanduse Selts, Tartu.

Fabech, C. 1992. Sakrale pladser i sydskandinavisk jernalder. En evaluering.--Sakrale navne. Rapport fra NORNAs sekstende symposium i Gilleleje 30.11.-2.12.1990. Eds G. Fellows-Jensen & B. Hohnberg. NORNA-forlaget, Uppsala, 141-167.

Fabech, C. 1999. Centrality in sites and landscapes.--Settlement and Landscape. Proceedings of a conference in Arhus, Denmark, May 4-7 1998. Eds C. Fabech & J. Ringtved. Jutland Archaeological Society, 45573.

Garwood, P. 2003. Round barrows and funerary traditions in the Late Neolithic and Bronze Age Sussex.--The Archaeology of Sussex. Ed. D. Rudling. King's Lynn: Heritage for the Centre for Continuing Education, University of Sussex, 478.

HCL 1982 = Henriku Liivimaa kroonika. Translated from Latin by R. Kleis. Edited and commented by E. Tarvel. Eesti Raamat, Tallinn.

Hofstra, T. 1988. Hiisi: uusi ehdotus sen germaanisesta alkuperasta. Omaa vai lainattua. (Itamerensuomen germaanisiin lainasanoihin liittyvia kirjoitelmia, 1.) Eds S.-L. Hahmo, T. Hofstra, A. D. Kylstra & O. Nikkila. Suomalaisen Kirjallisuuden Seura, Helsinki, 17 23.

Honko, L. 1998. Folklooriprotsess.--Maetagused, 6, 56-84.

Hupel, A. W. 1774. Topographische Nachrichten von Lief- and Ehstland. Gesammelt and herausgegeben durch A. W. Hupel. Erster Band. Riga.

Ingold, T. 1993. The temporality of the landscape.--World Archaeology, 25: 2, 152-175.

Ingren, N. 2005. Hendalundum. Lunden som beror. En studie i heliga lundar samt betydelsen av ett namn. Med utgangspunkt fran Hedelunda i Vadsbro socken, Sodermanland. Sodertorns Hogskola. Magisteruppsats i Arkeologi.

Jaanits, L., Laul, S., Lougas, V. & Tonisson, E. 1982. Eesti esiajalugu. Eesti Raamat, Tallinn.

Johanson, K. 2006. The contribution of stray finds for studying everyday practices--the example of stone axes.--EJA, 10: 2, 99-131.

Jordan, P. 2003. Material Culture and Sacred Landscape. The Anthropology of the Siberian Khanty. Altamira Press.

Jung, J. 1879. Eesti rahwa wanast usust, kombedest ja juttudest. Walja annud J. Jung. (Kodumaalt, 6.) Tartu.

Jung, J. 1898. Muinasaja teadus eestlaste maalt, II. Kohalised muinasaja kirjeldused Liiwimaalt, Pernu ja Wiljandi maakonnast. Kogunud ja valjaannud J. Jung. Jurjew.

Jung, J. 1910. Muinasaja teadus eestlaste maalt, III. Kohalised muinasaja kirjeldused Tallinnamaalt. Kogunud ja valja annud J. Jung. Tallinn.

Kaasik, A. 2004. Hiis kui parandmaastik.--Eesti Loodus, 7, 12-16.

Kivikoski, E. 1961. Suomen esihistoria. Porvoo.

Knupffer, G. M. 1836. Der Berg des Thorapilla: Ein historicher Besuch.--Das Inland, 22, 361-376.

Koski, M. 1967. Itamerensuomalaisten kielten hiisi-sanue. Semanttinen tutkimus, I. Turun yliopiston julkaisuja. (Sarja C. Scripta lingua fennica edita.) Turun yliopisto, Turku.

Koski, M. 1970. Itasnerensuomalaisten kielten hiisi-sanue. Semanttinen tutkimus, II. Turun yliopiston julkaisuja. (Sarja C. Scripta lingua fennica edita.) Turun yliopisto, Turku.

Koski, M. 1990. Finnic holy word and its subsequent history.--Old-Norse and Finnish religions and cultic place-names: Based on papers read at the symposium on encounters between religions in old Nordic times and on cultic place-names held at Abo, Finland, on the 19th-21st of August 1987. (Scripta Instituti Donneriani Aboensis, 13.) Abo, 404-440.

Kulonen, U., Seurujarvi-Kari, I. & Pulkkinen, R. 2005. The Saami. A Cultural Encyclopedia. Suomalaisen Kirjallisuuden Seura.

Kutt, A. 2004. Puhade puude ja puistutega seotud kaitumisnormid parimustekstides. Bakalaureusetoo. TU semiootika osakond. http://www.kodu.ee/~haldjas/BA/

Lang, V. 1996. Muistne Ravala: Muistised, kronoloogia ja maaviljelusliku asustuse kujunemine Loode-Eestis, eriti Pirita joe alamjooksu purkonnas, 1-2. (MT, 4.) Eesti Teaduste Akadeemia, Tallinn.

Lang, V. 1999. Kultuurmaastikku luues: Essee maastiku religioossest ja sumboliseeritud korraldusest. -EAA, 3: 1, 63-85.

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

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

Larson, L. 2005. Hills of the ancestors. Death, forging and sacrifice on two Swedish burial sites.--Dealing with the Dead. Archaeological Perspectives on Prehistoric Scandinavian Burial Ritual. (Riksantikvarieambetet arkeologiska undersokningar skrifter, 65.) Eds T. Artelius & F. Svanberg. Stockholm, 99-124.

Ligi, H. 1984. Kaali katastroof ja Puha kihelkonna kohanimed.--Keel ja Kirjandus, 5, 286-293.

Ligi, P. 1995. Uhiskondlikest oludest Eesti alal hilispronksi ja rauaajal.--Eesti arheoloogia historio-graafilisi, teoreetilisi ja kultuuriajaloolisi aspekte. (MT, 3. Toid arheoloogia alalt, 3.) Ed. V. Lang. Teaduste Akadeemia Kirjastus, Tallinn, 182 270.

Lintrop, A. 2003. Udmurdi usund. (Eesti Rahva Muuseumi sari, 5.) Tartu.

Loorits, O. 1932. Eesti rahvausundi maaihnavaade. (Slav teadus, 12.) Eesti Kirjanduse Selts, Tartu.

Loorits, O. 1949. Grundzuge des Estnischen Volksglaubens, I. (Skrifter utgivna av Kungl. Gustav Adolfs akademien for folklivsforskning, 18: 1.) Uppsala.

Loorits, O. 1957. Grundzuge des Estnischen Volksglaubens, III. (Skrifter utgivna av Kungl. Gustav Adolfs akademien for folklivsforskning, 18: 3.) Uppsala.

Magi, M. 2005. Mortuary houses in Iron Age Estonia.--EJA, 9: 2, 93-131.

Merkel, G. 1798. Die Vorzeit Lieflands. Ein Denkmahl des Pfaffen--and Rittergeistes. Erstes Band. Mit Kupfern and eine Karte. Berlin.

Moor, A. 1998. Hingepuu. Vanade muutide jalgedes. Elmatar, Tartu.

Oostra, S. 2006. Om lunden. Bidrag till kannedomen om begreppet lund och om lunden som foreteelse. Fakulteten for landskapsplanering tradgards--och jordbruksvetenskap. Institutionen forlandskapsplanering Alnarp. Doctoral thesis. Swedish University of Agricultural Sciences. (Acta Universitatis Agriculturae Sueciae, 2006, 25.) Alnarp.

Parker Pearson, M. 2000. The Archaeology of Death and Burial. Texas A & M University Press.

Parker Pearson, M., Pollard, J., Richards, C., Thomas, J., Tilley, C., Welham, K & Albarella, U. 2006. Materializing Stonehenge. The Stonehenge Riverside Project and new discoveries.--Journal of Material Culture, 11: 1/2, 227261.

Pentikainen, J. 1995. Mythologia Fennica by Kristfrid Ganander. Klaukkala.

Remmel, M. 1998. Hue ase: Hiis Eesti rahvaparimuses. Eesti Kirjandusmuuseum, Tartu.

Sarv, H. & Vladokin, V. 1988. Puudesalu tuupi ohvrikohtadest eestlaste ja volgasoomlaste juures. (Etnograafiamuuseumi aastaraamat, XXXVL) Valgas, Tallinn, 138-155.

Sild, O. 1937. Kirikuvisitatsioonid eestlaste maal vanemast ajast kuni olevikuni. Tartu.

Sognnes, K 2000. Det hellige landskapet: religiose og rituelle landskapselementer i et langtids-perspektiv.

--Viking, bind LXIII. Norsk arkeologisk arbok. Utgitt av Norsk arkeologisk selskap, 87-121.

Sturms, E. 1946. Die Alkstatten in Litauen. (Contributions of Baltic University, 3.) Hamburg.

Sundqvist, O. 2002. Freyr's offspring: Rulers and religion in ancient Svea society. (Historia Religionum, 21. Acta Universitatis Upsaliensis.) Uppsala.

Sutrop, U. 2004. Taarapita--the great god of the Oesilians. (Folklore, 26.) Tartu, 7-38.

Tamla, T. 1985. Kultuslikud allikad Eestis.--Rahvasuust kirjapanekuni: uurimusi rahvaluule ja proosaloomingu kogumisloost. (Emakeele Seltsi toimetised, 17.) Ed. U. Tedre. Eesti NSV Teaduste Akadeemia, Tallinn, 122-146.

Tamla, T. 1996. Virumaa muinasaeg. (Virumaa: koguteos.) Ed. K. Saaber. Tallinn, 206-244.

Toidybek, L. 1998. Mari usund. Translated by T. Vassiljeva & R. Gaidatsuk. Virgela, Tallinn.

Vaitkevicius, V. 2004. Studies into the Balts' Sacred Places. (BAR International Series, 1228.) Oxford.

Valk, H. 1995. Louna-Eesti XIII XVII/XVIII sajandi kulakalmistud rahvatraditsioonis ja uskumustes.--Rahvausund tanapaeval. Eds M. Hiiemae & M. Koiva. Eesti TA Eesti Keele Instituut, Eesti TA Eesti Kirjandusmuuseum, Tartu, 454-471.

Valk, H. 2004. Christian and non-Christian holy sites in medieval Estonia: a reflection of ecclesiastical attitudes towards popular religion.--The European Frontier. Clashes and Compromises in the Middle Ages. International symposium of the Culture Clash or Compromise (CCC) project and the Department of Archaeology, Lund University, held in Lund October 13-15 2000. Lund University Department of Archaeology and Ancient History. Gotland University College Centre for Baltic Studies. Visby. Almgvist & Wiksell International, Lund, 299-310.

Valk, H. 2006b. Medieval and Post-Medieval archaeology of the native rural population.--Archaeological Research in Estonia 1865-2005. (Estonian Archaeology, 1.) Eds V. Lang & M. Laneman. Tartu University Press, 205-221.

Valk, H. 2007. Choosing holy places.--Journal of Romano-British Archaeology. Continuity and Innovation in Religion in the Roman West, I. Archaeological and Regional Studies. In print.

Valk, H. In print. Looduslikud puhapaigad kui muistised. Arheoloogi vaatenurk.

Valk, H. & Maesalu, A. 2006. Research into the Late Iron Age.--Archaeological research in Estonia 1865-2005. (Estonian Archaeology, 1.) Eds V. Lang & M. Laneman. Tartu University Press, 127-158.

Valk, U. 2005. Eesti rahvaluule loodusenagemised.--Eesti looduskultuur. Eds T. Mayan & K. Tuur. Eesti Kultuuriloo ja Folkloristika Keskus, Eesti Kirjandusmuuseum, Tartu, 29-53.

Viidalepp, R 1941. Eesti ohvrikivid. MA thesis. Manuscript in the Library of Tartu University.

Vilbaste, G. 1947. Vanade kaartide abi muistsete kultuspaikade kindlakstegemisel. (Eesti Rahva Muuseumi aastaraamat, I (XV). Eesti NSV Teaduste Akadeemia, RK Teaduslik Kirjandus, Tartu, 157-161.

Watson, A. 1995. An approach to customary law.--Folk Law. Essays in the Theory and Practice of Lex Non Scripta, I. Eds A. Dundes Renteln & A. Dundes. University of Wisconsin Press, New York; London,141-157.

Widholm, D. 2006. Sacred Sites. Burial Customs in South Scandinavian Bronze and Iron Age. Sweden.

[TEXT NOT REPRODUCIBLE IN ASCII].

Eesti muistse usundiga seonduvad tahtsaimad kinnismuistised--hiied--on teaduslikku tahelepanu palvinud alles usna viimasel ajal. Senistes uurimustes on hiisi vaadeldud tavaliselt koos teiste puha- ja ohvrikohtadega, pidades hiit justkui laiema ohvrikoha uheks alaliigiks. Mones mottes tulekski hiisi just nii kasitleda--osana laiemast usundipraktikast--, kuid artiklis on keskendutud uksnes hiitele, valides oluliseks kriteeriumiks hiie-tuvelise toponuumi ja eeldades, et termin hiis tahistab midagi erilisemat, mida ei saa vorrelda uksikute ohvripuude ja -kividega. Kindlasti ei tohiks hiisi vaadelda kui midagi eraldiseisvat ning soltumatut ja toenaoliselt on mitmes paigas poimunud ka erinevad traditsioonid hiitest, ohvrikohtadest voi muidu parimuslikest paikadest. Autor peab siiski oluliseks anal uusida esialgu uksnes hiisi ja integreerida tulemused uldisesse ohvripaikade traditsiooni alles jargmistes uurimustes. Kull aga on allpool kasitletud hiie-tuvelisi kohti kooskolas teiste samaaegsete usundiliste motiividega.

Varaseimad uurijad alates 18. sajandi lopust, mil Eesti rahvausundit hakati uurima senise ebausukommete kirjeldamise korval, kasitlesid hiisi jaanukitena muistsest vabade eestlaste usundist. Paljuski olid need kasitlused mojutatud ule-euroopalikust rahvusromantika vaimust ja nii monegi puhul on margatav klassikalise antiikmaailma hiiekirjelduste tugev moju. Siiski on selle perioodi uurijad olulised, pannes aluse mitte ainult hiite uurimise traditsioonile, vaid ka mitmele stereotuupsele kasitlusele, mis kohati on pusinud tanapaevani. 19. sajandi lopul pandi alus sustemaatilisele rahvaparimuse kogumisele, millega alustati ka seni olulisima allikakorpuse moodustamist Eesti hiieuurimustes. Ehkki hiite uurimisel olid sellest ajast domineerivamad eesti soost uurijad, sailis osa varasematest kasitlustest. Oma osa lisas ka noore Eesti Vabariigi ideoloogia valjendamine, kus muinasaegseid hiisi nahti osana vabast Eestist.

Tosisema teadusliku huvi palvisid hiied alles parast II maailmasoda, mil valiseesti folklorist Oskar Loorits pakkus valja idee hiitest kui surnu- ja esivanemakultuse paikadest. Labi 20. sajandi ahvatles see teooria osa uurijaist, kuni alates 1990. aastatest sai idee hiite seotusest surnukultusega tugeva kriitika osaliseks. Sealtpeale on peetud oluliseks pigem hiite liminaalsust, eraldatust, hiie kui erilise koha staatust.

Valdavalt Virumaa materjali pohjal on artiklis eristatud kolm ruhma hiietopon uumilisi paiku: 1) hiied, mis asuvad efektsetel looduslikel kungastel, klindiservadel jm ning millega seonduvad noorema pronksiaja ja eelrooma rauaaja kalmed; 2) hiied, mille seos efektse loodusliku asendiga ei ole oluline, samuti pole oluline seotus kalmetega. Kull aga on nende puhul hasti jalgitav seos muinasaja lopu ja kesk-/uusaja asulatega; 3) hiie-toponuumilised kohad, mis asuvad argimaastikust eemal, naiteks soosaartel jm raskesti ligipaasetavates paikades. Sellise eristamise eesmargiks pole hiie-tupoloogia koostamine ja vaevalt see ka kunagi voimalik on. Kull aga on puutud juhtida tahelepanu monedele seostele.

Jalgides Eesti pronksi- ja rauaaja varasema poole (kuni 5. sajandini pKr) kivikalmete paiknemist maastikul, on iseloomulik nende koondumine aladele, mis on visuaalselt efektsed--need tousevad umbritsevast piirkonnast esile kas voimsate pinnavormide (klindinolvad, kunkad) voi ebahariliku umbruse (karstialad) tottu. On ka hulk erandeid, kuid muististe selline paiknemine on silmatorkav. Samal ajal on mitmel juhul voimalik jalgida kalmete koondumist ka parimuslike hiiekohtade lahedusse. Kui varasemad uurijad on hiisi ja kalmeid seostades otsinud kalmeid hiiekohtadest, siis sellist seost ei teki. Samas on rituaalsed paigad toiminud alati seoses umbritsevate kohtadega ja olulisteks on peetud paljusid objekte uheaegselt. Oluline on olnud ka muististe omavaheline silmside. Seega on ilmne, et ka Eesti hiite puhul on tahtsustatud nii hiiekohta ennast kui selle umbrust. On toenaoline, et suuremat osa andmestikust, mida algselt on hiite juures tahtsustatud, me enam jalgida ei suuda. Kull aga voib oelda, et uheks oluliseks objektiks olid kalmed, kui arvestada viimasel ajal ka Eestis levivat kalmete interpretatsiooni konstruktsiooniga kivikalmetest kui rituaalide labiviimise paikadest ja mitte eelkoige matmiskohtadest.

Selliseid muististevahelisi seoseid arvestades voib arvata, et varaseimad hiied voeti kasutusele enne meie ajaarvamist ja tahtsustati eraldiseisvatel magedel voi klindiservadel olevaid hiisi. Hiiemuististe kasutuselevotule sel perioodil naib viitavat ka sona etumoloogia. Kui pidada toenaoliseks sona algset tuve hiid ja selle tahendust--eraldiolev koht, kivine magi--, siis sobib sona Eesti ja Soome hiie-magedega ka semantiliselt. Samuti sobib sel juhul ka sona ja moiste dateering: sona on ule voetud ajal, kui sonaalgulist h-d veel ei kasutatud (paar sajandit eKr) ja sona algne kuju Eestis oli iis.

Seos hiite ja esivanemakultuse vahel ei kadunud ilmselt ka parast konstruktsioonidega kivikalmete rajamise loppemist. Ehkki uusi kalmeid enam ei ehitatud, maeti paljudesse endiselt nii rauaaja hilisematel perioodidel kui keskajal. See tahendab, et neid kalmeid vaartustati pusivalt (ehkki neid ei kasutatud matmiseks pidevalt) pikka aega. Autor usub, et neid kasutati suguvosade markeritena, millesse perioodiliselt mattes kinnistati oma sidemeid eelnevate polvkondade, ajaloo ning sealtkaudu ka maaga ja selle abil esitati oma suguvosa identiteeti. Lisaks kalmetesse matmisele, mida uuringute kaigus kogutud arheoloogilise materjaliga arvestades tehti harva, toimusid toenaoliselt hoopis tihedamini samateemalised rituaalid hiites, millest kahjuks pole jalgi sailinud voi ei suuda me neid ara tunda.

Teise suurema hiieruhmana on eristatud hiied, mis seonduvad muinasaja lopu, enamasti viikingiajal alguse saanud asulatega. Nende hiite laheduses on kalmeid harva ja tundub, et varasemat interpretatsiooni pole enam voimalik kasutada. Arvestades rooma rauaaja lopul kivikalmete ehitamisel toimunud muudatusi, kui kadusid kalmete konstruktsioonid ja edaspidi rajatud kalmed olid enamasti korratu kivilademega kaetud matmisalad, on voimalik, et kalmetel toimuvad rituaalid kaotasid osaliselt oma tahtsuse. Nii on toenaoline, et rahvasterannuajal hakkasid muutuma rituaalide sisu ning toimumiskohad, mis paiknesid umber pigem asulatesse ja nendega seonduvatesse hiiekohtadesse. Toenaoliselt olid siis esivanemakultuslikud jooned rituaalides marksa vahem esindatud (ehkki loplikult need ara ei kadunud) ja neis tahtsustusid enam jumalad ja nendega seonduv. Need hiiekohad ja selline hiietraditsioon oli suure toenaosusega ka see, mis joudis 13. sajandil kroonikatesse ja mille kajastused on joudnud lahimineviku ning nuudisaja rahvap arimusse. Siiski sailisid ka sellises hiietraditsioonis endiselt esivanematega seonduvad rituaalid ja hiiekalmeid kasutati jatkuvalt uksikuteks matusteks.

Viimane hiite alaliik, mida eespool on eristada puutud, on hiied eraldatud ja raskesti ligipaasetavates paikades, soosaartel, joeorgudes ning -luhtadel. Kuna paljude nendega seondub ka ohverdusmotiiv, on toenaoline, et selliste hiite puhul on tegemist pigem isiklike rituaalide labiviimise paikadega, millele lisas oma osa ka maastiku erakordsus. Samuti on voimalik, et sellised paigad tahtsustusidki rohkem alles hiietraditsiooni ametliku taunimise perioodil, st alates 13. sajandist. Siiski ei tohiks keskaegset hiietraditsiooni liigselt lihtsustatuna votta: kuna maa oli ametlikult kristianiseeritud, toimus igasugune hiite austamine salaja. Hoopis toenaolisem on, et keskajal peeti mitmeid hiisi endiselt oluliseks ja seal viidi selliseid rituaale labi. Tosi, olukord varieerus soltuvalt piirkonnast ja maa ametlikust valitsusest. Hea naite keskaegsest hiietraditsioonist pakuvad ka keskaegsed matused hiiekalmetes ja keskaja lopul ning uusajal hiide visatud mundid--oletatavad ohvrid. Siiski on selge, et olukord ei saanud olla enam selline nagu enne ametlikku kristianiseerimist. Ehk pohjustaski see peidetud ohvritoomispaikade kasutuselevotu?

(1) For example Manniku village at Viru-Nigula parish: Beside Manniku village there is an alder brush on the shore of the sea. This is called Hiie lepik (Alder Grove). There was a grove in old times. There were big figures of idols and people were worshipping these and to the biggest figure sacrifices were brought (ERA II 216, 179 (9)).

(2) "Kui Innimenne se Auu, mis Jummalalle ukspaine peab annetama, sellele annab, kennele se ei sunni; kui: Kujud palluma, arasurnud Puhhad appi huudma, saetul Ajal umbusklikud Tootused vima, Ma-Emma ehk Ma-allusid tenima, monda Paika, Maggesid, Kivvi, Hallikuid, Sauna-Leili, Hied ehk Puid Metsas puhhaks piddama, neile vahest Ohvrid vima, kahja teggema, ehk teisitau vorad Jummalad piddama" (Catechism 1694, 9).

(3) Besides these two versions a grove can also appear in forms like hiid and iid (means also giant in Estonian). So the grove could be connected with giants as well (Annist 2005). And even more--hiis-word can appear in other meanings, like gust, bird, etc., but these associations are not important in this context.

(4) The lake was formed during the 8th millennium and dried in the 5th millennium BC (see Jaanits et al. 1982, 36).

(5) True, due to partly destroyed grave it was not possible to connect the item with any particular

Tonno Jonuks, Department of Folkloristic, Estonian Literary Museum, 42 Vanemuise St., 51003 Tartu, Estonia; tonno@folklore.ee skeleton but the pendant was found from the same area with the medieval burials.