As noted in the introduction, this book is also a study in spatial anthropology, in the sense that it focuses on space in dealing with a territorial founding ritual that may have been known and practised in other parts of the world as well. There are countless studies that could be classified as contributions to the anthropology of space, but if this relatively new name is useful, it is because it calls for research focused on space-related problems that have not received enough attention from researchers in the past (Domenig 2014: 9). The question underlying the present book is a case in point. It seems to me that the ritual procedure for taking possession of land in early times has never been dealt with adequately because scholars were too easily satisfied with simple answers such as land-taking is boundary marking or going around the boundary.
As I would not have found it helpful to write a comparative study mentioning many different cultural traditions, selectively and superficially, in this book I focused on the ancient sources from Japan that I am most familiar with. Nevertheless, having extended my research to some other cultures as well, I wish to add a few notes about what I have read about territorial founding rituals in Western literature regarding Southeast Asia and three Indo-European cultures: ancient Greece, medieval Iceland and Vedic India.
We possess old written sources from many parts of the world that include stories telling of the opening up of land and the founding of settlements. The Vedas contain many a poetic allusion to the Aryan immigrations into India; numerous passages in ancient Greek and Latin literature tell of the founding of cities and colonies; and Iceland has the unique Landnámabók and the family sagas, which tell how, around the turn to the tenth century, the still uninhabited island in the North Atlantic was settled by immigrants in search of land. If we study and compare such sources, we can see that they differ in content, in the way they are presented and in many other respects; but the ways in which they can be studied differ too. Scholars specialised in the study of a specific culture may tend to consider only what the sources of that culture clearly tell us, but if we approach the topic with an anthropological interest in human space behaviour generally, and in human territorial behaviour in particular, we may wonder if there was perhaps a typical ritual procedure for founding territories that shows up varyingly in different traditions of the world.
More than a hundred years ago, E. S. Hartland, the author of the article on ‘foundation rites’ in the Encyclopedia of Religion and Ethics (1913: 109–115), organised his cross-cultural comparative material under the headings ‘Choice of Site’ (various forms of divination) and ‘Appropriation of the Site’ (to exorcise, mark out and consecrate the site).1 While this was an adequate way of characterising the founding ritual as a two-part ritual, Hartland did not discuss the relation between the two parts, nor did he specify to whom the rites were primarily addressed and where they were performed relative to the boundaries of a territory.
Divination and the consecration of the site were and still are often thought to have been addressed to the deity that was later worshipped in the shrine or temple that was founded, and exorcising is usually assumed to have been done by expelling evil spirits in order to get rid of them for good. This is also the impression we get if we consult the literature of Vedic India, ancient Greece or medieval Iceland. The texts may tell us that a temple or altar for a god used to be founded on the occasion of settling down. But the gods mentioned in such contexts – often Agni in India, Apollo in Greece and Thor in Iceland – are gods of mythologies, whereas these texts say almost nothing about the worship of nameless nature spirits met on the occasion of taking possession of land. Although local spirits played a part in all these traditions, it is usually not clear that they were worshipped on the occasion of taking possession of land and continued to be worshipped later.
Whereas Hartland mentioned various rites that used to be performed when taking possession of land, including divination, others simplistically assumed that it was sufficient to mark out the boundaries because these identify a territory in its size and geographic location. The custom of walking around the boundaries, which is still annually practised in parts of Europe, has therefore come to be interpreted as a repetition of the act of taking possession (Kramer and Schildt 2009). In the Middle Ages, it could indeed be a part of the legal contract in the transfer of land ownership,2 but whether and to what extent the boundaries of a territory used to be marked when one took possession of ownerless areas depended on the morphology of the landscape and other circumstances.
Scholars like Ananda Coomaraswamy (1935) and Mircea Eliade (1959) ignored the need of divination. They thought that a shrine or temple was founded when taking possession of land and that it symbolised the ‘centre of the world’ according to a ‘symbolism of the centre’ which for homo religiosus would have invoked the idea of a vertical world axis (axis mundi) connecting the three cosmic regions Heaven, Earth and Underworld. To take possession of land would therefore have meant symbolically recreating the universe.
Settlement in a new, unknown, uncultivated country is equivalent to an act of Creation. When the Scandinavian colonists took possession of Iceland, Landnāma, and began to cultivate it, they regarded this act neither as an original undertaking nor as human and profane work. Their enterprise was for them only the repetition of a primordial act: the transformation of chaos into cosmos by the divine act of Creation. By cultivating the desert soil, they in fact repeated the act of the gods, who organized chaos by giving it forms and norms. (Eliade 1959: 10)
This statement is typical of Eliade’s view, but if we wonder on what it is based we find that the author refers to an article by van Hamel (1936) that draws attention to the fact that the Völuspá of the poetic Edda is not a creation myth but a poem that describes the ordering of the world in an early state of wilderness and in the words of a poet who was inspired by the settlement of Iceland that had taken place a hundred years earlier.3 As Wolfgang Lange, a scholar of medieval skaldic poetry, has pointed out:
Mythology and lived religion are two different things. The reports make it clear that on going into the uncertainty of the new land, other forces and powers than the great gods were invoked. […] What Snorri reports and what the poems of the Edda tell about the gods is not to be put on a level with the religion of the Norwegian emigrants. (Lange 1958: 162, from the German)
To get an adequate idea of what land-taking meant for the settlers of Iceland, we must turn to the medieval sources.4
The Settlement of Iceland
The settlement of Iceland in the years between 870 and 930 CE is the classic case of a historical settlement where people took possession of land in uninhabited regions and were later said to have performed certain rites following their ‘heathen’ tradition. During the ‘land-taking time’ (landnámatiđ), more and more groups of settlers arrived from Scandinavia, mainly from Norway, some also from the British islands, to take possession of a stretch of land and live in Iceland. This was called landnám (‘land-taking’) or nema land, ‘to take (possession of) land (as a settler)’.5 When, by the year 930, most inhabitable parts of Iceland had been occupied by more and more groups of settlers, the land-taking time came to an end and the Icelandic Commonwealth was founded, with the Althing as a legislative and judiciary parliament. It then probably took more than two hundred years until the time of settlement became the subject of a literary work that came to be known as Landnámabók (‘Book of Settlements’). Together with a number of Icelandic family sagas, this work is a highly interesting source for the study of the land-taking behaviour of Nordic settlers before their conversion to Christianity.
Landnámabók has been called a unique work without parallel in world literature, but several of its characteristics suggest a comparison with the eighth-century fudoki of Japan. It contains about 400 entries, while all the fudoki together contain about 650 entries, if we also count the quotations from lost fudoki in later works. Like the fudoki, Landnámabók is mainly a register that lists the settlements of administrative districts in topographical order – in this case the four quarters of Iceland established after the middle of the tenth century. In either case, this listing is combined with short stories referring to occurrences that allegedly took place when the respective settlement was founded. But while the various Japanese fudoki put the stress on the names of the territories that became administrative villages (sato) and districts (kōri) of the state initiated by the Taika Reform, the Icelandic counterpart rather focuses on the location and extension of the territories and on matters of genealogy.
A further similarity consists in the political motivation that may have led to the compilation of these books. It has been said that the introduction of a tithe system for the whole country in the year 1097 may have contributed to the first compilation of Landnámabók (Pálsson and Edwards 1972: 2, 5). This could explain why the settlements are identified in relation to the size of the estates and the course of their boundaries. The Japanese fudoki, on the other hand, identified the districts and villages mainly by their names, and the political meaning was implied in a law-symbolical ritual according to which the telling of a land name could mean a declaration of submission and loyalty.
The nature of these works as registers probably accounts for the fact that the stories are usually short and never give us a full account of the whole ritual procedure that was or should have been followed in taking possession of land. The information these sources provide about the religious beliefs of the settlers and the rites they performed consists of scattered pieces that must be compared to make full sense of them. By comparative study, the meanings of some of the story motifs can be identified as allusions to parts of a larger ritual with a common purpose and an inner logic.
In Landnámabók, initial divination is represented by numerous passages telling of settlers that threw their high-seat pillars overboard when they approached the coasts of Iceland, vowing to settle down where these pillars came ashore. This was clearly a rite addressed to the ‘land beings’ (landvættir), as Dag Strömbäck (1928) pointed out in his classical study almost a hundred years ago. As another noted scholar put it, for the ordinary people the myths might have been “very limited in their significance” and many individuals taking land in Iceland “must have paid more attention to the land spirits, for instance, than to the higher gods” (Davidson 1964: 214).
Strömbäck treated some of the most important rites mentioned in the Icelandic texts: the initial pillar floating, the bringing of fire onto the land and the placing of a staff. He stressed the belief in the existence of land beings (landvættir), noted that settlement represented an encroachment on the landvættir’s rights and argued that the belief in landvættir is more ancient than the belief in ancestral spirits. What Strömbäck apparently did not recognise is that the various rites were related to one another as parts of a meaningful ritual that should have ended with a cult dedicated to the landvættir.
If the texts do not clearly say that the landvættir were worshipped, then this is because they were written when Iceland had already converted to Christianity and it was forbidden by law to dedicate one’s property to ‘heathen’ beings. The Grágás, the legal corpus of the Icelandic Commonwealth written in 1117–1118, says:
Men are to put their trust in one God and His saints and are not to worship heathen beings. A man worships heathen beings when he assigns his property to anyone but God and His saints. If a man worships heathen beings, the penalty is lesser outlawry. (trans. Dennis and Perkins 1980: 38)
Nevertheless, some entries in Landnámabók indicate that the land spirits (landvættir) were sometimes worshipped during the period of settlement (874–930 CE). One settler sacrificed to the rocks at the upper end of his land; another one to the grove after which his home was named; and a third one is said to have sacrificed to a waterfall (Pálsson and Edwards 1972: entries 241, 237, 355). Rocks, groves and waterfalls were of course typical landscape features where land spirits used to receive offerings.
If one studies the Icelandic sources with a particular interest in the question of how territories were founded, they suggest that the usual foundation ritual corresponded quite well to that described in the introduction to this book. It began with an act of divining addressed to the land spirits (e.g. the floating of the high-seat pillars in connection with a divining oath). The claim was then marked, and after some time fire was brought onto the land. This was probably done to expel the spirit-owners of the land from the area to be taken over. The fire rite should therefore have been followed by worshipping these spirits as guardian deities, but since this was no longer allowed under Christian law, the later texts interpret the fire rite as the legal act of taking possession.6
The fire rite was called ‘at fara eldi um landit’, which is usually misunderstood in the sense of ‘to go with fire around the land’. This common misinterpretation is based on the preconceived idea that land-taking was done by marking out the boundaries. Although the Icelandic sources are sometimes quoted as evidence for this idea, they do not really support it. The preposition um in the above phrase also meant ‘about’ and ‘over’; and ‘over’ is doubtless the correct meaning in the cited phrase as well as in similar ones. Not only would it often have been impossible to go all around the boundaries because many territories in Iceland were situated along the coast and ended in steep and rocky hills (figure 82),7 it was also much easier to go with fire over the land. One entry in Landnámabók clearly describes this as a method of claiming entire catchment areas by lighting fires at each river mouth (Pálsson and Edwards 1972: entry 218). This was a reasonable method of claiming the land from the coast to the watersheds, whereas there are no stories that clearly tell of taking possession of land by riding or going all around the boundaries. The single explicit story that is often quoted in this context is from Hen Thorir Saga and regards a man who claimed land by riding around a burned-down house; but his claim was later not recognised as valid (Morris and Magnússon 1891: 144).
Some of the settlers additionally also founded a temple dedicated to a god of Norse mythology, often Thor. They apparently did so for political reasons; for such temples used to serve people of different settlements, whereas pagan settlers might also worship their own landvættir and might have done so at a place different from that of the temple, if they built one.
We get the most complete picture of a settlement in Iceland in a chapter of Eyrbyggia Saga, which describes in detail how a settler called Thorolf Mostrarskegg moved from Norway to Iceland and took possession of land on the northern coast of Snæfellsnes Peninsula in the western part of Iceland.8 The following is a list of what he allegedly did according to that story:
Back in Norway, Thorolf divines in combination with a sacrifice, asking Thor: shall I move to Iceland?
He pulls down the temple of Thor, takes some earth from underneath Thor’s seat, and prepares a ship for the journey.
He crosses the sea to Iceland.
In approaching Iceland, he throws the high seat pillars overboard in combination with an oath, saying that he will stay in Iceland where Thor lets the pillars come ashore.
Exploring the land, he finds the high seat pillars washed ashore on a peninsula (3 in figure 82).
He goes with fire from Staff River to Thor’s River (probably making fires at all the river estuaries from 1 to 2 in figure 82).
He settles his shipmates on the mainland.
He builds his estate and rebuilds the temple of Thor on the peninsula, re-installing also the earth brought from his old temple in Norway (4 in figure 82).
He declares the hill on the peninsula as a holy place where no animals should be killed and where he and his descendants would go after death (5 in figure 82).
He decides the site on the peninsula where the high seat pillars have come ashore (3 in figure 82) for a regional assembly related to the temple.
He dwells and prospers in the new place.
Thorolf’s territory at Thorsnes, according to Eyrbyggia Saga. (1) Estuary of Staff River (?); (2) Estuary of Thor’s River (?); (3) Small peninsula where the high-seat pillars were found; (4) The temple of Thor; (5) Helgafell, the sacred hill
Author’s map
The Staff River may have got its name because a staff was erected there as a claiming sign and the holy hill may have been sacred to the landvættir, which would explain why no animals were to be killed there. Wooden parts of the temple of Thor in Norway were carried to Iceland and the temple was rebuilt there. This explains why Thor is exceptionally said to have let the high-seat pillars come ashore. Basically, the landvættir were addressed in that rite, but a god of the mythology might sometimes have been invoked to accompany the floating pillars.9 The importance of addressing the landvættir in approaching the coasts of Iceland is also clear from a heathen law mentioned in the Hauksbók manuscript of Landnámabók:
This was the beginning of the heathen laws, that men must not keep a ship at sea with a figure-head on; but if they have, then they must take off the head before they come in sight of land, and not sail to land with gaping heads and yawning jaws to frighten the spirits or wights of the country.10
Since this was a pagan law, not to frighten the landvættir (the ‘spirits or wights’) meant not to forego the chance of making a deal with them and gaining their protection at the new place.
In early times of Christianisation, the importance of worshipping land spirits was still recognised by common people in many parts of Europe, and even churchmen fighting the ‘heathen’ customs sometimes had a bad conscience when they destroyed sacred groves and expelled the spirits believed to be dwelling there. The Icelandic bishop Guðmundr the Good was said to have always left ‘a place for the evil to live’, but of course he did not go so far as to worship the spirits to whom he granted a refuge. Settlers secretly still adhering to their indigenous religion, however, used to pacify and worship the spirits they had expelled.
The Japanese fudoki as sources about land-taking rites can complement the Landnámabók, because in Japan there was no reason to suppress information about the cult dedicated to the land spirits. The story of Yato no kami, quoted and discussed in chapter 2, is the best example for this because it tells how such cults were founded. On the other hand, the Landnámabók is helpful in drawing attention to the need for initial divination, which the fudoki rarely mention explicitly by using a word meaning to divine.
Founding Sacred Groves and Colonies in Ancient Greece
When it came to be known in the second half of the nineteenth century that sacred groves in northern India were regarded as remnants of the primeval forest, ethnographic reports of that kind inspired W. H. D. Rouse in his book Greek Votive Offerings (1902) to interpret the sacred precinct (temenos) of ancient Greek temples in the same way:
When the earth and its growths were regarded by the simple soul as possest [sic] or protected by unknown powers, any intrusion upon new dominions was thought to be dangerous. […] Often a plot of land is left barren, or a clump of trees unhewn, to be the abode of the spirit which has been disturbed. In Greece, when land was occupied by conquest or colonization, a portion of the land was ‘cut off’ (
τέμενος ) for the god’s habitation. (Rouse 1902: 39)
Rouse searched ancient sources to support this interpretation and, although he found a number of late ones, he concluded that “such questions can never be answered now” but “it does not follow by any means that there was nothing of the kind because we hear so little about it” (Rouse 1902: 43).
Later scholars of Greek religion ignored the sources that Rouse had cited or rejected them as being too late and unconvincing. Although there is much scholarly literature about the foundation (ktísis) of Greek cities and colonies, a scholarly encyclopaedia article on this topic noted that we are “not well informed” about the ancient Greek foundation ceremonies, although there must have been “a frame of traditional rules” regarding the actions of a founder (Cornell 1983: 1123, 1136). A few years later, Irad Malkin (1987: 2–3) nevertheless distinguished a “series of religious acts” that characterised the founding of a Greek colony:
-
An oracle, often the oracle of Apollo in Delphi, is consulted before setting out from the homeland.
-
Divination is also performed by seers (manteis) who accompany the whole foundation process (in the mother city, en route and at the new site of settlement).
-
Sacred fire from the public hearth of the mother city is transferred to the new colony.
-
Taking possession of the new site is done by founding sanctuaries for imported gods, usually at the periphery of the city.
-
The oikist or founder regulates the cults, religious calendars, etc.
-
After death the founder is worshipped as a heros in the centre of the city.
This is a convincing sequence insofar as it stresses both the importance of initial divination and the founding of sanctuaries at the periphery. However, if point 4 says that the sanctuaries were dedicated to imported gods, then this means that the author ignores or neglects the role of autochthonous local spirits. Malkin is even explicit about this point when he emphasises that the cult accorded to the deceased founder was “the first cult which was the colony’s own” (1987: 2; 1996: 231).11 He mentions that the founders had to respect local gods and heroes of earlier settlers as “Greek religious precedents” (1987: 148–60), but he does not consider that the settlers might have started by worshipping the spirit-owners of the land when they opened up land that had not been settled before. Stressing that he is concerned with historical colonies, Malkin admits that he has ‘little to say on legendary oikists and legendary foundation oracles’ (1987: 21). He therefore also ignored the fact that Jane Harrison (1962 [1927]: 277–340) had earlier come to the conclusion that there were two things that went into the making of a Greek hero: a dead man and a daimon.
A famous story where a local daimon played an important part is the foundation story of Alexandria. According to tradition, a large serpent appeared on the site and Alexander ordered it to be killed and worshipped in a heroon as the city’s good spirit (agathos daimon). Since this story is told in the Alexander Romance (c.300 CE), it is often neglected as a source for early Greek religion. More recently, however, Daniel Ogden (2015) compared it with other stories where Alexander is associated with serpents and notes that the tale of Alexandria’s foundation “broadly aligns with a traditional Greek story type in which the founder of a city is somehow guided to the site by an animal”. Speaking of “the old Greek foundation myths”, Ogden then mentions eight cases where different animals – a crow or crows, a mouse, a fish and a boar, wolves and sheep, and a ‘metaphorical goat’ – were said to have guided the founders to the place where they settled down (Ogden 2015: 132).
Fully explicit and sufficiently old sources saying that Greek colonists worshipped local land spirits may indeed be lacking, but oracle stories mentioning animals can also support what Rouse had suggested, provided that they allow the conclusion that the animals in question were locally worshipped. A further story, which Ogden does not mention, allows this conclusion. It regards the foundation of the temple of Apollo Smintheus (the Mouse Apollo) who was mainly worshipped in Asia Minor. In that case, an oracle predicted that the settlers should stay where the ‘earth-born’ would attack them. They finally settled near Hamaxitus because there it happened that many field mice came out of the ground at night and ate the leather in their equipment. According to Strabo (13.1.48),12 this story goes back to Kallinos (about 650 BCE), and Heracleides of Pontus (fourth century BCE) said that “the mice which swarmed round the temple were regarded as sacred, and that for this reason the image [of Apollo] was designed with its foot upon the mouse”.
Since it was Apollo of Delphi, a higher god of the homeland, whose oracle was said to have predicted that the ‘earth-born’ would show the place of settlement, this oracle story suggests that the ‘earth-born’ mice were meant to be integrated as autochthonous elements into the cult of Apollo at Hamaxitus. Another ancient writer, Aelian (second/third century CE) supports this interpretation:
And those who live in Hamaxitus in the Troad worship a Mouse, and that is why according to them, they give the name Sminthian to Apollo whom they worship, for the Aeolians and the people of the Troad still call a mouse sminthus. […] And in the temple of Smintheus tame mice are kept and fed at the public expense, and beneath the altar white Mice have their nests, and by the tripod of Apollo there stands a Mouse. (Aelian, Nat.an.12.5, trans. Scholfield, Loeb Classical Library)
Here, it is clearly said that a mouse was worshipped and that white mice had their nest under the altar of Apollo. The mice were probably sacred because they were regarded as manifestations of the local spirits, and these spirits were the object of a local cult that continued to be dedicated to them as spirit-owners of the land. The foreign Apollo cult was thus established on top of a cult that was dedicated to the local spirits and continued to coexist with it.
From these and other ancient texts, it appears that Apollo sanctuaries existed along the western coast of Asia Minor, which were briefly called sminthia or ‘mouse sanctuaries’. In such places, Apollo was worshipped as a god of agriculture and in this capacity also controlled the animal pests, especially the mice, which were hated in the fields and worshipped in the smynthia. As lord of the mice, Apollo Smintheus could punish the people for some offence by sending the mice out to devastate their fields and, on the other hand, he could also listen to human requests for the removal of an already existing plague of mice.
To the north-east of the Troas was Bithynia at the southern shore of the Black Sea. An excerpt from Apuleios (second century CE) included in Geoponika (13.5.4) says that the people there knew a strange method of exorcising mice from a field. First, they would stop the holes with rhododaphne, so that the mice trying to get out would gnaw it and perish. Then they would do as follows:
Take some paper and write these words on it: “I adjure the mice taken in this place, that you do me no injury yourselves, nor suffer another to do it; for I give you this ground (and you mention which); but if I again take you on this spot, I take the mother of the Gods to witness, I will divide you in seven parts.” Having written these words, fasten the paper in the place where the mice are, before the rising of the sun, to a stone of spontaneous production, and let the letters be turned externally. (trans. Owen 1806: 141)
If the mice were ordered to remain in the field that was given to them and not to tolerate others doing harm, then this means that they were ritually turned into guardian spirits of the farmer’s fields. The method evidently corresponds to what our key story from Japan says about the foundation of a snake cult (chapter 2). Being known from ancient Bithynia, it is also relevant in connection with the sacred mice in the temple precinct of Apollo at Hamaxitus, which may likewise have been worshipped as guardian spirits of the agricultural land.
Oracle stories mentioning animals that reveal to a founder the place of settling down were apparently intended to indicate that the respective animals were to be worshipped as manifestations of the local spirits. They therefore contradict Malkin’s thesis that the cult of the deceased founder was the very first which was a colony’s own.
In a later article, Malkin (1996) proposes an interesting idea to explain the peripheral location of extra-urban sanctuaries in ancient Greece. He writes that “the question should be answered in Greek religious terms: the ‘division of the same’ between the Greeks and their gods. As with sacrifice, in which the gods got the fat and bones and man received the meat, the gods got fat land, revenue-bearing temenê, which were, however, dangerous and distant plots” (Malkin 1996: 75). The analogy is appropriate; the ‘division of the same’ is what Rouse probably also meant when he said that a portion of the land was “cut off” (
These notes indicate that the importance of initial divination is fully acknowledged in discussions of Greek foundation rites, but sources about the cults dedicated to the local spirits at the time of founding territories appear to be rare and are usually neglected because they are not considered to be sufficiently old.14
The Vedic Tradition
In the Rg Veda, the Five Aryan kindreds are spoken of as immigrants; they have come from another place across the waters, and have settled and tilled the lands on the hither shore. This process of land-taking has generally been interpreted as referring to an historical immigration of an Aryan speaking people who […] crossed the Sarasvati in the Punjab and made their home in Bharatavarsa. (Coomaraswamy 1935: vii)
Ananda Coomaraswamy wrote this in a booklet he titled The Rg Veda as a Land-náma-bók (1935). Although the title refers to the Icelandic Landnámabók, he rejected the usual interpretation that the land-taking motif in the Rig Veda concerns events in history. His interest was not in the form of the ritual but in what he called a metaphysical tradition, according to which such stories would signify the origin of time and history. Thus, the passages describing settlers in search of land and arriving at their destination ‘here’ would mean “the entering in of time from the halls of outer heaven” (1935: vii, 25). Consequently, Coomaraswamy mainly mentions the first fire, the gārhapatya, which one had to establish when taking possession of land according to the Vedic texts. He refers to a passage from Śathapatha Brahmana (Ś.B. VII, 1, 1, 1 and 4), “where it is clear that those are first settlers or ere-dwellers (viśa) who build a fire-altar on any land, the performance of this rite constituting the legal act of land-taking” (Coomaraswamy 1935: 16).
What Coomaraswamy does not tell us is that the gārhapatya had a counterpart in a second fire, called āhayanīya, and that a more complex ritual is described in the text from which he quotes. Before building the gārhapatya, the ground had to be cleared by sweeping away the elements “crawling on their bellies”, lest one should “settle on those already settled”. In other words, one had to expel the local spirits that could appear as snakes. Initial divination is not mentioned explicitly; but during the sweeping episode the text says something equivalent: “Yama, the ruling power, with the consent of the Fathers, the clan, now grant to this (Sacrificer) a settlement on this earth” (Ś.B. VII. 1.1.4).
Most interestingly, the text adds that some also sweep at the āhayanīya, but that one should not do so because “by the āhayanīya he rises upwards”.15 Interpreted in the sense of what has been argued in this book, this can only mean that the ‘crawling elements’ are going to be worshipped at the āhayanīya; for by worshipping them there the settler would elevate them to the higher status of the land’s guardian deities and “rise upwards” himself, reaching the higher status of a landowner.
That the ‘crawling elements’ were supposed to be snakes and that they were indeed to be worshipped is clearly expressed in a passage of Pancavimsa Brahmana:
Through this (rite), the serpents gained a firm support in these worlds. They who undertake it get a firm support in these worlds. […] Through this (rite), the serpents vanquished death. Therefore, they (the serpents), having left aside their old hide, creep further, for they had vanquished death. The Adityas are the serpents. They who undertake this (rite), to their share falls the shining out, as it were, of the Adityas. (P.B. XXV, 15. 2 and 4)16
Here it is clear that the serpents vanquish death by being ritually transformed into deities (Adityas).17 By letting them live one could get “a firm support in these worlds” and profit of their new status as guardian deities.
What is lacking in the quoted texts is a mention of the new dwelling space of these serpents after they have vanquished death and attained a higher status as worshipped deities. The analogy with the case from Japan suggests that when such a ceremony was performed in an open landscape, a part of the landscape was kept sacred as their new dwelling space. The texts quoted do not say so, although they suggest that the snakes will live on and be worshipped. This possibly has to do with the fact that the two fire altars mentioned were built on an east–west axis in a household’s offering hall (Eggeling 1894: 298n1). The gārhapatya was associated with ‘earth’, the āhayanīya with ‘heaven’, and the whole arrangement symbolised a structured microcosm which may have represented the household’s larger territory, whereas the ritual repeated the procedure followed in creating such a territory.
For land-taking rituals performed in an open landscape, we must turn to another part of Pancavimsa Brahmana (XXV. 10; Caland 1931) where it deals with the Sarasvati sattra, a ritual that was performed along a river identified with the mythical Sarasvati. The procession starts in the west, where the Sarasvati ‘is lost in the desert’, and moves upstream and eastward over a distance of forty days on horseback towards a place called Plaksa prasravana, where the river appears and from where it flows forth. Each time, before temporarily settling down to pass the night, the participants first make the sāmnāyya sacrifice, a libation of sweet and sour milk offered to Indra:
After the sacrifice of the sāmnāyya, the Adhvaryu throws a yoke-pin (in easterly direction, from the place where the sacrifice has been performed). The place where it falls down is (the place for the) gārhapatya(-fire). From this spot, he makes thirty-six strides (in easterly direction): this is (the place for) the āhayanīya(-fire). (P.B. XXV. 10. 4, trans. Calland)
Here we have a mention of the initial act of divination; it was done by means of a yoke pin thrown to indicate the location of the gārhapatya fire. The āhayanīya fire was built thirty-six strides from there, and the temporary campsite of the participants in this ritual may have been between the two fires. The text thus mentions parts of a ritual for taking possession of land on a small scale, but this time the land was only for temporary use.
The paraphernalia needed for this ritual were movable and included sheds on wheels and a sacrificial stake that could be set up without making a hole into the ground (P.B. XXV. 10. 5). Whereas the participants ritually took possession of a small territory each time, where they stayed overnight, the procession as a whole moved upstream until they reached Plaksa prasravana after forty days:
At the distance of a journey of forty days on horseback from the spot where the Sarasvati is lost (in the sands of the desert), (is situated) Plaksa prasravana. At the same distance from here (from the earth) (is situated) the world of heaven: they go to the world of heaven by a journey commensurate with Sarasvati. (P.B. XXV. 10. 16)
The usual interpretation is that the ‘here’ in the second sentence of this passage means the location of Plaksa prasravana on earth, so that heaven would still be far from there. This has led to the interpretation that the mythical river Sarasvati was meant to flow down from the sky along ‘the plaksa tree of forth-streaming’ (the literal meaning of Plaksa prasravana), and that from there it would continue streaming westwards through the earth until getting lost in the sand of the desert (figure 83b; Hiltebeitel 2001: 143–49). However, if we assume that myths are sometimes misunderstood because they refer to conditions or concepts with which we are no longer familiar, we can make an attempt at understanding the quoted passage differently. Noticing that another ancient text, Jaiminīya Brāhmana, locates Plaksa prasravana at the extreme end of the river Sarasvati, saying that the space of heaven is situated there (Caland 1919: 201), we can convincingly interpret the quoted passage from Pancavimsa Brahmana in the same sense. We can assume that the ‘here’ in the second sentence means the place “where the Sarasvati is lost” and where the Sarasvati sattra begins (see earlier), which is what the third sentence suggests anyway. Assuming this, both texts mean the same, namely that heaven begins at Plaksa prasravana and that Sarasvati is the name of the river flowing down from there, whereas the source rivers further up had no name. Having become familiar in the present book with the concept of a ‘terrestrial heaven’, we can make sense of this and interpret the text according to figure 83a.
Schematic view of an older and a newer concept of ‘heaven’ in the myth of the river Sarasvati in Vedic texts. 83a (left): Plaksa prasravana as the boundary tree between a terrestrial heaven in the mountains and the ‘earth’ extending down along the river Sarasvati. 83b (right): The tree was interpreted as symbolising a vertical axis connecting the earth with the heaven vertically up and the river Sarasvati was thought to have flowed down from the sky
Diagrams by the author
We therefore have a case where a text referring to the horizontal worldview (the river) has been partly interpreted according to the vertical one, which accounts for the strange idea of a river originating in the sky and flowing vertically down from there along the stem of the tree at Plaksa prasravana. This strange interpretation can only be upheld because it is still common to think that anything can make sense in myths and because one does not usually dare to assume that a word like the Sanskrit svarga or svargaloka, meaning ‘heaven’ or ‘heavenly space’, might sometimes also have been used to designate a sacred area down on earth. Even if these words were only known in the common sense, the traditions regarding the Sarasvati sattra should be a reason to consider that an interpretation as ‘terrestrial heaven’ (figure 83a) makes better sense in this particular case.
The lack of understanding for the evolution of the two worldviews – from horizontal to vertical – is responsible for numerous misinterpretations of ancient texts that used to be interpreted according to the vertical view but really were meant to refer to the horizontal one. Approaching this problem with an open mind, we can imagine that with the continued extension of territories, the terrestrial ‘god land’ (chapter 2) often came to be moved higher up the side of a mountain. It could even be separated from the territory and connected only with a pathway, so that it eventually also became possible to think of it as being situated in the sky above the horizon. Once the land of the gods was imagined above the horizon, it also became possible to think of it as being located in the highest part of the sky vertically above the earth.
An ancient Indian myth that refers to the first stage of this kind of change is the story of King Videgha Māthara’s eastward migration along the river Sarasvati. The king first held Agni Vaisvānara in his mouth, and when they arrived at the river Sarasvati, Agni fell out onto the ground. After this allusion to initial divination (the falling motif), the story continues by saying that, from there, Agni “went burning along this earth towards the east”, and that the king followed him. Agni burnt over all these rivers, but the Sadānīrā, which flows from the northern Himalaya, “he did not burn over. That one the Brāhmans did not cross in former times, thinking: ‘it has not been burnt over by Agni Vaisvānara’”. Later, however, the Brahmans “caused (Agni) to taste it through sacrifices” (Śathapatha Brahmana I:4:1:13–16; trans. Eggeling 1894).
If the text calls the last river Sadānīrā, this is probably its later name; at the time when Agni ‘burnt over’ all other rivers it must still have been a nameless source river of the Sarasvati. Understanding the phrase ‘burning over rivers’ in the sense of a technical term for making land-taking fires at intervals – as in the Sarasvati sattra and in medieval Iceland (see p. 272) – we can understand that the last river was originally left unclaimed on purpose and that the Brahmans respected this. Therefore, this appears to be another allusion to the terrestrial ‘heaven’ in the source region of the Sarasvati.18 But in this case the text adds that the region was later cultivated too, presumably because by that later time the old concept of the terrestrial heaven in the source region of a river had come to give way to that of a mythical heaven up in the sky. Regarding the possibility of this kind of change, we must again consider that it could have involved either the use of different words to denote the different concepts or the semantic change of a single word that was used to denote two different concepts.19
As regards, finally, the dwelling space of the snakes that were worshipped as guardian deities of the land, there is no lack of later sources from India that associate snakes with sacred groves, and which also do so in stories about the founding of territories. A legend from Kerala in southern India says that the origin of the custom of worshipping snakes in sacred groves goes back to the time when Sri Parasurama reclaimed the land from the sea, 640 miles from north to south, and fixed images of Varuna in 108 places because he had received the land from Varuna, the god of the waters. But when he brought Brahman settlers from the north of India to colonise the new land they took to flight after a few months because the land was full of snakes and they were afraid of them. After the land had been in the possession of the serpents for some time, Sri Parasurama divided it into 64 colonies, brought another group of Brahmans from the north and advised the settlers to reserve special areas for the serpents and to worship them as household deities. Then offerings came to be made regularly and people were freed from the fear of snakes.20
Since at least the nineteenth century, we also have ethnographic accounts describing sacred groves in India. Some of them say that such groves were believed to be remnants of forests that were left standing for the local deities when the land was opened up to cultivation:
Every [Munda] village has in its vicinity a grove reputed to be a remnant of the primeval forest left intact for the local gods when the clearing was originally made. […]. The grove deities are held responsible for the crops, and are especially honoured at all the great agricultural festivals. They are also appealed to in sickness. (Dalton 1872: 188)
The Sarna or sacred village grove […] is a remnant of the primeval forest still retained by the first founders of permanent villages, who carved them out of the forest. The village dwellings are placed close to it, and round this centre ran the ring of cultivated land called by the Gonds the ring of the guardian snake. This separated the home of civilized life from the world of death, the land of the uncleared forest. (Hewitt 1899: 338)
This was the kind of source that inspired Rouse to interpret the sacred precincts of ancient Greece in the same way. Today, many similar ethnographic field reports are known from other parts of Asia and the world, but scholars of ancient Greek religion still seem to hesitate to let themselves be inspired by non-Greek sources when they deal with the ancient Greek concept of alsos.
Opening Up Land in Shifting Cultivation
The ritual transformation of spirits of the wilderness into local cult deities has also been discussed in reports about ethnic groups of Southeast Asia. As Anthony Walker has pointed out regarding the Lahu of northern Thailand, the spirits of the wilderness are propitiated through ritual to become specific guardian spirits and “lords of the place” of the settlements and swiddens (Walker 2015). Frederic Kris Lehman, another anthropologist familiar with northern Thailand, has described the ideology thus:21
The basic idea is simple enough – the original and ultimate owners having dominion over the face of the land are spirit lords commonly associated with more or less prominent features of the landscape. […] [W]hen the first human settlers began to clear any given tract of land, they had to make a contract with these spirit lords. This contract ensured that there would continue to be communication between the two parties as long as both sides lived up to the requirements of that contract. The exclusive right to serve as mediators between the lords and human settlers is supposed to pass to the hereditary heirs and successors of the original founders of the settlement, in perpetuity. This is the very essence of the founders’ cult. (Lehman 2003: 16)22
Apart from the founders’ cults in the sense of cults where the founder worships the spirit-owner/s of the land with whom he has made a cult contract, there were also many cases where the founder later came to be worshipped too and both identities then came to be conflated. As Lorraine Aragon pointed out in an article dealing with Central Sulawesi (Indonesia):
Founders’ cults concern two different, but often conflated, types of spirits. The first are the unseen guardian deities, often referred to as “owners” or “lords”, of a land area (pue’, Uma). These spirits permit the land’s initial clearing and its transfer from the “wild” to the human domain. The second type are ancestral spirits of those individuals who made successful agreements with the guardian deities. The conflation of these spirit types occurs both among the followers of founders’ cult and foreign observers […]. In areas such as Central Sulawesi, the spirit of founders become so closely associated with the land guardian deities that their discrimination often serves no symbolic or practical purpose. The agency of these two types of spiritual beings is simply merged. The metaphoric collapsing of land guardian and founding ancestors holds symbolic and practical value by allowing descendants of founders to claim descent from the guardian spirits and deified ancestors as well as mere mortals. (Aragon 2000: 87–88)
Robert Wessing has shown that this other type of a founder’s cult, where the spirit of the founder is merged or identified with the tutelary spirit, is “often found throughout Southeast Asia” (2017: 529).
I myself have dealt with clearing rituals in a book about the relation between architecture and sacrificial rites performed in traditional houses of ethnic groups of Indonesia. I came to the same conclusion as in the present book: the clearing of forests had to begin, in principle, with an act of divining (or an equivalent act) and the spirits expelled were later worshipped in an adjacent grove as guardian deities of the land opened up to agriculture (Domenig 2014: 21–96). Regarding the dwelling places of earth spirits that were invited to an offering, I noticed contradictory ideas caused by not keeping the two worldviews apart. Some sources say that these spirits dwell in the earth beneath the ground, whereas others say that they dwell in a horizontally distant place and therefore approach the offerings horizontally and from above. This problem has also been noticed in other parts of Southeast Asia. Andrew Turton, for example, observed an interesting case where it led to disagreement between two ritual experts. When the spirits had been expelled from a site for building a house in Thailand, the ‘nagas under the ground’ (imagined as snake spirits) were to receive offerings of red and white cloth in a rite called ‘exorcism of the ground’. The ambiguity in this name led to different interpretations. One expert said the offerings should be buried under the two main posts of the house, whereas the other one argued that it should be taken outside the building site ‘as an exorcism’ (Turton 1978: 116). In this particular case, the spirits were finally assumed to dwell outside the site but were invited to taste the offerings inside it.
I have called this kind of contradiction the ‘earth spirit paradox’, because the problem disappears if we understand that the first statement locates these spirits in a vertical worldview and the second one sees their place in a horizontally structured world (Domenig 2014: 120–23). The idea that the earth or land spirits dwell outside is of course conditioned by the land-taking ritual, which required that they were expelled horizontally and worshipped as spirits or deities dwelling then outside the site.
Another significant point that I noticed is that the Japanese concept of yorishiro has Indonesian correspondences in temporary offering stands whose leafy decoration is intended to attract spirits and deities to their offerings.23
From Terrestrial Heavens to the Heaven in the Sky
Figure 83 may stand as a visual representation of an important point that has occasionally been mentioned in the pages of this book. The land-taking ritual makes us doubt that the vertical worldview has been the primary one of homo religiosus, as Mircea Eliade always used to proclaim; it suggests, to the contrary, that the horizontal view of the world – the view of the world as a horizontally structured whole – is typologically older. Although in building houses and the like, vertical structure was always important, wider spaces were probably first organised in the horizontal dimensions of landscapes before it became possible to imagine the macrocosm as vertically structured in the same sense.
The times of land-taking in unoccupied regions are long past in almost all parts of the globe, and the sources for learning about it have been so neglected that religions are usually associated with the idea of gods that created the great World and exist somewhere in a far-off mythical heaven above us. By contrast, the present book is an argument in favour of not continuing to overlook or misinterpret the numerous sources that relativise this Eliadean view.
To give an example from Africa, the Barotse of western Zambia say that their god Nyambe originally lived on earth but moved away because he no longer wanted to be reached by Kamonu, a man who had repeatedly killed his animals. Nyambe withdrew to an island in the river, then to the top of a mountain. When human beings had multiplied and were living everywhere he searched for a place still further away. With the help of an animal diviner and a spider he eventually found a new place in the sky, and the spider stretched a string so that he could climb up. Kamonu and his people tried to build a tower to climb up too, but as they failed, they finally had to be content with seeing Nyambe from afar. At sunrise they would greet him “with loud shouts and clappings of hands”.24
Historians of religion believing in primitive monotheism would argue that Nyambe was a supreme being who dwelt in the sky from the beginning but temporarily manifested himself as a local deity to Kamonu and later retreated to the sky. In this particular case there is, however, the problem that Nyambe only moves to the sky when human beings are already living everywhere and that he needs something material to climb up. The story therefore clearly conveys the idea of a process by which the local deity of a terrestrial microcosm eventually became the supreme being of a religious macrocosm.25 If later people should have believed that Nyambe was a high god from the beginning, this would be understandable, but scholars of the history of religions should be able to also see things the other way round.
Having briefly dealt with Japanese creation myths at the end of chapter 8, I am inclined to conclude that many territories in pre-Taika Japan were still organised as bipolar microcosms with their own ‘terrestrial heaven’ in the form of an area that was kept sacred as the dwelling space of the local kami. This kami land can be seen as the prototype of the sacred grove, which later survived as an institution even when the deity to which it was dedicated was newly thought to dwell in a more distant heaven. Speculations about greater worlds, however, were probably made by thinking in terms of myths and legends that may originally have told of the making of territories down on earth.
The basic problem with the interpretation of such ancient traditions is the correspondence of microcosm and macrocosm. At the end of the nineteenth century, many scholars noticed that there have been influences from the human microcosm to the view of the macrocosm and vice versa. Ferdinand von Andrian, for example, distinguished two stages in human spiritual development in his cross-cultural study of mountain worship. In one, the terrestrial conditions were transferred to the sky, in the other, later one, the phenomena of the sky in turn served as a measure for the terrestrial world (Andrian 1891: xii). Similarly, Otto Gruppe, in a discussion of mythical geography, suggested that in ancient Greece certain features of a topography that played a part in ritual – gardens, caves, mountains – received secondarily their fictitious counterparts in far-removed regions which were then claimed as their mythical prototypes: the divine orchard at the end of the world, the subterranean Hades, Olympus as the dwelling place of the gods (Gruppe 1906: 384–88). With regard to the relation between architecture and world images, similar ideas were also expressed, for example, by Robert Eisler (1910: 615).
Since the 1930s at the latest, however, a different kind of interpretation has been advocated. Under the influence of the hypothesis of an original monotheism (Lang 1898; Schmidt 1912) and of C. G. Jung’s archetype theory, it became fashionable to think that only the reflections from the big World to the small world should be considered as relevant for homo religiosus. Mircea Eliade in particular insisted that the myth of the creation of the great world or universe was primary and came to be symbolically repeated and imitated in human works. He argued that it is wrong to rationalistically “reduce” the religious beliefs in the cosmic prototypes to a level of understanding where they might appear to be conditioned by still earlier prototypes from the small worlds of human life (Eliade 1989b: 263n18). To look for the origins of certain cosmological ideas in the tangible elements of man’s dwelling, for example, would be wrong because such elements, he assumed, must necessarily be of a kind that has to be explained by reference to a rational need, such as adaptation to the environment. The roof of a dwelling may afford protection against rain, cold and heat, but if it is also regarded as sacred, then according to Eliade this is only because it symbolises the sky and therefore participates in the sky’s sacredness.
It would lead too far away to enter into a further discussion of Eliade’s view here. Suffice it to say that the ritually established kami land of an ancient territory did not have to symbolise the far-off sky to be experienced as sacred: the land-taking ritual discussed in this book fully explains why and how it came to be a sacred dwelling space of local deities.
This regards the founding of territories; a third part of Hartland’s article deals with laying the foundation of buildings and distinguishes actions such as scaring away evil spirits, destroying spells, conciliating the local spirits by animal or human sacrifices and providing a new tutelary power.
Fiedlerová and Razim 2018.
Van Hamel 1936: 21. Eliade refers to this Dutch article as cited by van der Leeuw, which suggests that he had not read it himself.
The following notes on medieval Iceland are based on my intensive two-year study of the subject in the late 1980s. For this study, I learned to read the texts in their Old Norse original, which explains why I do not always follow common interpretations.
Modern German ‘Landnahme’ has been testified since the eighteenth century and translates the Old Norse word landnám. For a discussion of the term Landnahme and its use in German historical studies, see Eglinger and Heitmann 2010: 11–17.
The Viga-Glums Saga, chapter 26, tells of a law case according to which a certain Glum was sentenced to selling half of his land and to depart from the land within a fixed period of time. A certain Einar then bought that land and started cultivating it, but Glum stayed to the very end of the time that had been granted him. When that time had come, he still stayed until he was chased away by bringing fire on the land. https://sagadb.org/viga-glums_saga.en.
See also the maps in Pálsson and Edwards 1972.
Morris and Magnússon 1892: 7–9.
Compare this with the story from Japan quoted at the end of chapter 1. It tells that Kannon came to Kashima with three logs which the founder of the temple floated there, but this was only possible because the founder had first asked for and received permission from the deity of Kashima. The fact that, in the above story from Iceland, a god of the mythology accompanied the floating high-seat pillars does not preclude that the rite was basically addressed to the landvættir, although some scholars have argued to the contrary.
Translation from Origines Islandicae: 318, https://archive.org/stream/originesislandic01gudb/originesislandic01gudb_djvu.txt.
Similarly, Cornell (1983: 1140) writes that there was no better-suited object for a local cult in a colony than the founder.
Strabon 13.1.48, trans. H. C. Hamilton and W. Falconer, 1903. Copied from Perseus Digital Library, http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.01.0239%3Abook%3D13%3Achapter%3D1%3Asection%3D48. For living animals being kept in shrine precincts of Japan as ‘messengers’ of the main deity, see the section ‘Nature Spirits Can Become Manifest in Wild Animals’ in chapter 8.
A rather strange interpretation has been given by Walter Burkert. He derived the word alsos from a root meaning ‘to feed’ (or ‘to grow’) and explained the meaning as “a grazing area for the pack animals and mounts of the participants in the festival”, adding that “this in no way precludes a certain feeling for nature, especially as the grove is reserved for sacral use” (Burkert 1985: 86).
This was quite different in the case of the Roman tradition, which is not dealt with here. Among the Romans, the importance of worshipping land spirits is well known. In the context of agriculture, such spirits or minor deities were called lares and worshipped everywhere. Their cult was related to the domestic hearth, and its function inspired Fustel de Coulanges to state that “it was not the laws that first guaranteed the right of property, it was religion” (1877: 86).
Ś.B. VII. 1.1.1. and 3.1.7. trans. Eggeling 1894. Quoted from: https://sacred-texts.com/hin/sbr/sbe41/sbe4161.htm.
Pancavimsa Brahmana, trans. W. Calland 1931: 640, 642, https://archive.org/details/pancavimsabrahma032052mbp/page/n683/mode/2up.
Coomaraswamy mentions this passage (1935: 17) and adds a footnote drawing attention to his separate article titled ‘Angel and Titan, an Essay in Vedic Ontology’, where he shows that “the Devas are Assuras and Serpents sacrificially transformed or ‘turned about’”. This other article appeared the same year in the Journal of the American Oriental Society, 55 (1935): 373–419.
The apparently usual interpretation is that Agni’s ‘burning over’ the rivers means drying the rivers up, but this would not make any sense for a land-taking god like Agni.
An example of the latter kind of change is Old Japanese ama, which could mean ‘heaven’ (above the earth) or ‘attic’ (of a house) or ‘sea’, depending on the context (chapter 5).
My paraphrase of a quotation in Maity (1965: 50–55) from an article by C. K. Menon that appeared in the Calcutta Review, 113 (1901): 19–20. Another version of this story says that Parasurama made the serpents accept an arrangement, according to which an eighth of each garden around the houses would be left to the serpents (Tarabout 2019: 230).
Tannenbaum and Kammerer 2003. See also Mus 1975 [1933]; Aragon 2000; Wessing 2006; Domenig 2014, Walker 2015.
For other studies about founders’ cults among various ethnic groups of Southeast Asia, see Tannenbaum and Kammerer 2003.
Domenig 2014: 131–238, with numerous illustrations.
My summary of Smith 1950: 159–61.
For another discussion of such a case, see Domenig 2014: 124–28.