B. Sawyer: Women in Viking-age Scandinavia Women in Viking-age Scandinavia - or: who were the 'shieldmaidens'? Birgit Sawyer, NTNU (Norway) It has been generally supposed that in Viking-age Scandinavia women had a higher status, greater freedom and fewer restraints on their activity than later, i.e. after the conversion to Christianity. This view was already current in the nineteenth century and was closely related to the belief that the freedom and equality supposed to characterize Germanic society survived longer in Scandinavia than elsewhere. Few scholars still accept that 1 interpretation of Germanic and Scandinavian society but belief in free Nordic women has lasted better and continues to influence discussions of the period. It is therefore necessary for modern students of women's history to consider how this idea originated and on what basis. Earlier discussions of the topic has made much use of twelfth- and thirteenth-century literature, Icelandic sagas and the work of the Danish history writer Saxo Grammaticus. Icelandic authors who describe their pagan past let us meet many active, strong-willed, and often war-like women, but such dominant women are conspicuously absent from the sagas written about contemporary Iceland. In them the women are pale shadows of their predecessors; passive, submissive, and completely subordinate to their husbands and kinsmen. The same contrast between pagan and Christian women is found in the major histories that cover both periods, Saxo Grammaticus' Gesta Danorum and Snorre Sturlason's Heimskringla. The conclusion has been that this contrast reflects real changes in women's conditions, mainly as a consequence of Christianization. According to this view an early - pagan - ideal of active and martial 'shield-maidens' was replaced by the passive and submissive 'madonna' ideal favoured by the church. Few, if any, scholars nowadays believe that women warriors led war bands in Viking-age Scandinavia; they have either been relegated to the world of amazon myths or explained away in other ways. Even so, the shieldmaidens still figure prominently both in popular beliefs and scholarly discussions, and in this lecture I will address two questions: a) how are we to interpet the descriptions of these war-like women in the past, and b) 1For a survey of studies along this line see Jochens 1986, p. 36 (note 4). 1  B. Sawyer: Women in Viking-age Scandinavia does the contrast between active pagan and passive Christian women reflect real changes? 1. Who were the shieldmaidens? Hitherto three kinds of answers have been presented: first that the shieldmaiden-motif reflects pre-historic reality, secondly that it is influenced by Greek and Roman myths, and thirdly that it expresses a pre- Christian woman-ideal. In Icelandic literature stories about shieldmaidens are mainly found in the historically unreliable fornaldarsögur (sagas about the heroic past), e.g. about Hervor (in Hervor's and Heidrek's Saga), the only child of the hero Angantyr, who fell in battle before she was born. She took care of her father's sword ('Tyrfing') and used it to take revenge on his enemies. Other famous examples of war-like women are Brynhilde (in the Sigurd's Saga) and Freydís (in the Saga of the Greenlanders). Also Saxo Grammaticus has many examples of shieldmaidens and gives us a detailed description of them (my italics): There were once women in Denmark who dressed themselves to look like men and spent almost every minute cultivating soldiers' skills; they did not want the sinews of their valour to lose tautness and be infected by self-indulgence. Loathing a dainty style of living, they would harden body and mind with toil and endurance, rejecting the fickle pliancy of girls and compelling their womanish spirits to act with a virile ruthlessness. They courted military celebrity so earnestly that you would have guessed they had unsexed themselves. Those especially who had forceful personalities or were tall and elegant embarked on this way of life. As if they were forgetful of their true selves they put toughness before allure, aimed at conflicts instead of kisses, tasted blood, not lips, sought the clash of arms rather than the arm's embrace, fitted to weapons hands which should have been weaving, desired not the couch but the kill, and those they could have appeased with looks they attacked with lances. 2 Not only the Icelanders and Saxo tell us about such women; traditions about them are also found in earlier, foreign, sources; according to Adam of Bremen (who wrote his work about the archdiocese of Hamburg-Bremen in the 1070s) there was a region north of Swedish lake Mälaren, the so called 'Kvänland', populated by war-like women, and as late as in the nineteenth century historians like Alexander Bugge and Johannes 2 Gesta Danorum, translation by Peter Fisher 1979, p. 212. For the latin text see ed. by J. Olrik and H. Ræder 1932, p. 192. 2  B. Sawyer: Women in Viking-age Scandinavia Steenstrup maintained that such women had in fact existed in prehistoric Scandinavia. Most modern scholars, however, do not believe this, and many explain the women warriors as a literary motif. According to some, the motif is a loan from antique traditions about the amazons (from Greek a mazos = without breast), women who are described as a distinct tribe in Anatolia. It was said that they cut off one of their breasts in order to be better archers, and for their survival as a tribe they were believed to couple with men from neighbouring tribes. Of the resulting children they kept only those boys that could be used for household tasks, while all girls were kept and educated in the art of warfare The problem with this 3 explanation, however, is that there are more differences than similarities In Scandinavian literature the female warriors are never described as a separate tribe; Saxo describes occasionally organized (tillfälligt hopsatta?) war-bands, and in Icelandic literature we only meet single female warriors, acting on their own. Another important difference is that Scandinavian shieldmaidens are characterized by their chastity. According to more recent research, the descriptions of these women do reflect reality but not in such a direct way as was earlier believed. One suggested explanation is that they are based on misunderstandings by visiting foreigners who have seen temporarily man-less settlements, especially in Scandinavian coastal districts. Another suggestion, made by Carol Clover, is that the descriptions of the shieldmaidens are the literary expressions of a social reality, in which brother-less women were forced to axla the role of a son. Clover interprets the shieldmaidens as women who, for practical reasons, must occasionally act like men, for taking revenge or paying/receiving wergeld (manbot). According to the Icelandic lawcode Grágás the 'ring-woman', i.e. father-, brother-, and son-less women had to function like men - as long as they were unmarried. They could be seen as 'surrogate sons', and Clover explains the popularity of the motif as due to the fascination for transvestite traditions and the transgressions of sexual borders. Thus, according to Clover, this 'collective fantasy' tells us much about the underlying tensions of the society that produced it. 4 Clover's interpretation is interesting and can explain a lot, but it does not explain why the motif remained so popular even long after the social reality had changed; most literary descriptions were written in the twelfth 3 Sobol 1972. 4 Clover 1986a, pp 35-49. 3  B. Sawyer: Women in Viking-age Scandinavia and thirteenth centuries when revenge and blood-feud had been forbidden and there were royal and ecclesiastical courts to regulate family conflicts. Neither does her interpretation explain the 'communities' of shieldmaidens in Saxo's work. 4  B. Sawyer: Women in Viking-age Scandinavia 2. Shieldmaiden and Madonna - two contrasting ideals? Whatever - reality and/or literary influences - lies behind the descriptions of the female warriors in old Norse literature, we will return to the wide- spread view that they express a pagan ideal of active and martial women, replaced by a Christian ideal of the passive and submissive 'madonna'. This view, which has greatly inluenced modern women studies, is, however, highly questionable since it lacks support in other sources. Furthermore, the very concepts 'shieldmaiden' and 'madonna' create problems since they are not univocal (entydiga?) but are used by different scholars to denote different things. Sometimes the word 'shieldmaiden' is used only about female warriors, sometimes in a figurative sense about all strong-willed and independent women. Further, the Church favoured not only one but two contrasting ideals, partly the Virgin, a Christian parallell to the strong, steadfast, and man-like woman ('virago'), partly the Wife/Mother, the submissive and self-effacing woman, who subordinated herself to her husband and devoted her life to her children and family. Of these two the Virgin ideal was most highly valued; in ecclesiastical literature it is the chaste woman - apart from the virgin, also the widow - who is honoured, but in profane literature the valuation is quite different.5 We simply do not know what pre-Christian woman-ideal the Church replaced, but it is not very likely that the shieldmaiden was ever an ideal in pagan Scandinavia; drawing their strength from chastity these female warriors were independent of men and thus uncontrollable, thereby constituting a threat to social order. I will return to the shieldmaiden-motif at the end of my lecture and now concentrate on their equally strong and independent, but less war-like sisters in Viking-age Scandinavia. 3. The saga evidence It is obvious that conversion to Christianity changed conditions for women as well as men, but it must be seriously doubted whether the effect of Christianity was so sudden and complete as the twelfth- and thirteenth- century authors make it appear. Instead of taking the contrasting depictions of dominant pagan and submissive Christian women at its face value, we ought to ask ourselves what special purposes the authors had with their descriptions of women. In the Middle Ages history writing was didactic, i.e. one should draw useful lessons from the past. Therefore, 5 Strand 1980, pp. 29-57. 5  B. Sawyer: Women in Viking-age Scandinavia when Saxo places almost all independent and active women in the pagan past, this serves his purpose of demonstrating that such female behaviour belongs to bygone, pagan, and thus impefect times. It is clear that from this kind of history writing we cannot draw any conclusions about either contemporary or earlier reality. In the Icelandic sagas the descriptions of women are also subordinated to the special purposes and literary motifs of their authors. Here we often meet women as inciters, dangerous opponents and skilled in magic, and far from expressing a woman ideal, such women illustrate the ecclesiastical image of woman as a threat and danger to men. Both Saxo and his Icelandic colleagues were strongly influenced by Roman history writing, and as models for their descriptions of all the strong, courageous and man-like women they might as well have used examples from Antiquity as from their own past. Another consideration is the fact that a similar kind of contrast can be seen in English descriptions of women before and after the Norman Conquest of 1066. When describing women who had leading roles in public affairs, pre-conquest authors did not express any astonishment, while twelfth-century historians thought it was extraordinary. This change of attitude has been interpreted as a result of the actual shrinking of opportunities for women to play a political role as war and government left the home area to become genuinely public activities. The fact that in 6 English literature the contrasting attitudes before and after 1066 has nothing to do with the role of Christian influence should serve as a reminder when dealing with the attitudes in contemporary Scandinavian literature. Could the contrasting depictions that the Scandinavian authors give of past and contemporary women have the same cause as in England? As far as attitudes are concerned, it is highly likely that Scandinavian authors were influenced by literary modes and conventions in other parts of contemporary Europe, but it is less likely that the contrast they presented between past and present actually reflects any shrinking of opportunities for women to take active part in public life. In twelfth-century Scandinavia the social and political developments were not yet that advanced; here war and government had not yet left 'the home area'. 4. Other evidence 6 Bandel 1955, p. 114. 6  B. Sawyer: Women in Viking-age Scandinavia Since the literary sources from the twelfth and thirteenth centuries are so coloured by their authors' views and purposes, we will now turn to other sources: archaeological remains, poetry, and the only written material we have from the Viking Age itself, i.e. the runic inscriptions (see chapter 'Scandinavia in the Viking Age' in this volume). The evidence of graves, from the third or fourth centuries to the tenth or eleventh, when pagan burial customs ended in most parts of Scandinavia, show that some women, especially older ones, were treated with great respect. In Denmark the quantity and quality of the furnishing in men's graves decreased with the age of the dead man, but some of the richest burials were of women who were 50 years old or more. This contrast shows that the elaborateness of a burial was not determined by the status of the dead person's family. What ist does suggest is that respect for women increased with age and was perhaps earned by the experience gained during a long life. In other parts of Scandinavia too many of the richest burials were of women. The most lavishly furnished burial known in Sweden, at Tuna in Badelunda near Västerås, is of a woman who was buried in the fourth century with an abundance of gold and imported goods. One of the richest Norwegian graves, the ninth-century ship burial at Oseberg, was also of a woman. In Norway in this period it was women above all who were buried in large and richly furnished long-barrows. 7 There are indications that some women in pre-Christian Scandinavia were highly regarded for religious reasons. Eddic poetry shows that in Nordic mythology knowledge of the past and the unknown, especially the future, was associated with female beings, as were the arts of writing, poetry and magic. The name given to this collection of poetry, Edda, may itself originally have meant 'great-grandmother'. If so, it underlines the role of women as transmitters of tradition, as do many of the poems. In 8 Sigrdrifumal, for example, when Sigurd asks the valkyrie Sigerdriva to teach him wisdom she does so by instructing him about victory runes, healing runes and runes to protect the unborn. The collection begins with Völuspa 'Völva's prophecy'. In it Völva, a sibyl, addresses Odin and describes creation, the golden age of the gods, and their corruption. In response to Odin's request for wisdom she prophesies Ragnarök, the destruction of the world of the old gods, and a new age with one powerful ruler, in which the innocent gods are resurrected and righteous men live for ever. The poem is clearly influenced by Christian ideas and its date (if it has 7 Farbregd 1988. 8 See further in Steinsland 1985; Mundal & Steinsland 1989. 7  B. Sawyer: Women in Viking-age Scandinavia one) is uncertain, but it is a vivid reminder of close association of women with wisdom and prophesy. According to Snorri the gods known as Æsir learned wisdom from the goddess Freyja, and it is even more significant that Odin is said to have done so by becoming argr. That word had extremely offensive implications and was used for men who took a passive, i.e. female role in homosexual relations. To accuse a man of being argr was a grave insult not so much because of the homosexuality, but because submission to another man implied submissiveness in other ways, cowardice and loss of honour; it amounted to declaring that a man was no longer a worthy member of society. It has been suggested that such insults were considered 9 exceptionally serious in Iceland because, in the absence of a superior authority, social stability depended on the integrity of men. The emphasis on masculine qualities is also reflected in heroic poetry and tales by the common theme of women who were not only beautiful and accomplished, but also warriors. Heroes had to overcome such women in their manly role in order to win and deserve them as partners; the proper destiny of 'shield-maidens' was marriage. Another aspect of this literature is that proud and confident women were specially attractive and highly valued as wives, not only because of the prestige of winning them but also because their sons could be expected to inherit the qualities of their mothers as well as their fathers. 10 Once married, a woman was expected to assume a completely feminine role with her own well-defined responsibilities quite distinct from those of her husband. In Icelandic law her duties were entirely confined to the farm, where she had great authority, symbolised by the keys on her belt (for family, household, kinship and inheritance, see 'Scandinavia in the Viking Age' in this volume). She could not represent the farm externally and was excluded from public life, but was involved in all family business. It was in her interest to see that the honour of the family was upheld, and the common theme in sagas about early Iceland of women urging their menfolk to take revenge probably had some basis in reality. In poetry revenge appears to have been above all the concern of women. Skalds who feared that their reputation had been damaged because they had not taken revenge as they should have done, turned to women. Most examples of skalds addressing women are in the family sagas. They are less common in the king's sagas and such poems are never in drottkvætt, the prestigious 9 Sørensen 1980, pp. 24-39. 10 Mundal 1988, p. 24. 8  B. Sawyer: Women in Viking-age Scandinavia meter of court poetry. In general it was when a skald was most concerned about himself that he turned to a woman and spoke of his dreams and fears, or his wounds and impending death. The good opinion of women was valued. As Roberta Frank has but it : 'What, after all, was the point of the institutionalized male violence celebrated by the skalds, those excessive vendettas and duels, that piracy and harrying, if women were not watching you, constantly comparing you to little Alf the Stout or to Snorre Gore- Fang?'.11 Most of the literature so far mentioned was written in Iceland where conditions were in some respects very different from Scandinavia. The fact that the proportion of early medieval farms named after women was ten times greater in Iceland than in Norway may indicate that women had a higher status in the newly colonized land. It may also be significant that the most active volcanoes in Iceland all have female names. Archaeological 12 evidence and runic inscriptions, however, suggest that women had much the same roles in other parts of Scandinavia as in Iceland and that at least some were highly regarded. One way of showing concern for the honour and reputation of a family in the late tenth and eleventh centuries was to erect a rune-stone, and it is significant that almost a quarter of them were sponsored by widows. Few are explicit as the stone from Norra Härene in Västergötland 'Åsa honoured her husband in a way that henceforth no woman will ever do', but they all 13 show family pride. One woman in Uppland erected a stone in memory of a kinsman and named his killer, apparently to keep the memory of this shameful act alive and perhaps even as an incitement to revenge. There is another inscription at Bällsta in Uppland in which a widow says that she has had a lament made for her husband. These can be compared with Icelandic laments in which widows mourn and sometimes demand revenge. Carol Clover has drawn attention to other cultures with the 14 custom of blood feud in which women had a leading role in commemorating the dead and maintaining feud. The fact that lamentation 11 Frank 1990, p. 78. 12 Frank 1973, p. 483. 13 Sveriges Runinskrifter (SR) Västergötland (Vg) 59. The runic inscriptions are referred to according to their number in the standard editions. 14 Clover 1986b, p. 146. 9  B. Sawyer: Women in Viking-age Scandinavia by women, often in poetic form, is a widespread feature of funeral rituals suggests that in Iceland women did so in real life as well as in literature. 15 In marriage the sexes had well-defined, distinct roles, but widows and women who were unmarried or had no near male kinsmen were able, or forced to assume some of the responsibilities that were normally men's. Some women were clearly highly regarded, whether as respresentatives of powerful families, of for their age and wisdom. There is no hint that women's abilities were doubted in the pagan period, and their association with wisdom and magic is notable. Their links with both nature and the supernatural were a source of power. Conversion to Christianity meant that many earlier beliefs and customs were condemned, and this gradually affected the attitude to women and their role in society. Christian authors regarded much magic as an evil to be eradicated with the result that they tended to depict women of the pagan past in the mould of Eve, the root of all evil. 5. Women's role in the Christianization process There are, however, many indications that women women were among the first and most eager converts; most of the early Christian graves at Birka were of women, and runic inscriptions show that the cult of Mary developed early in Svealand and was favoured by women. Further, most of the runic monuments commemorating men who were converted in their last days and 'died in white clothing' were erected by women. It is 16 not surprising, if women were especially attracted by the new faith; much Christian teaching must have been welcomed by them, a point obscured by the misogyny that colours so much medieval literature. They must have found the prospect of the Christian Paradise far more attractive than the gloomy realm of Hel to which they had previously been consigned. Many of them must also have been glad to believe that in the sight of God they were men's equals and that their worth did not depend on their fertility, family or social status; the community of Christians had room for all, including women who were barren or unmarried, as well as orphans and the poor. Christian teaching that all had an obligation to help those in need was especially welcome to women without near kinsfolk, for they had far more limited opportunities to support themselves than men in a similar situation. It may also be supposed that many mothers were 15 Clover 1986b, pp. 180-3. 16 B. Sawyer 2000, p. 140. 10  B. Sawyer: Women in Viking-age Scandinavia gladdened by the attempts of the Chruch to prohibit, or at least severely restrict, the custom of infanticide, despite the increased burden that this must often have imposed. It is no accident that one of the chapters in Rimbert's Life of St Anskar was devoted to the piety and steadfast devotion of Frideborg, a rich widow in Birka, and the care she took to ensure that her wealth would be distributed as alms in a suitable manner for the sake of her soul. She is said to have lived to a great age and always been a generous almsgiver. As death approached she enjoined her daughter Catla to distribute all that she possessed to the poor. According to Rimbert she said 'because there are here but few poor, at the first opportunity after my death, sell all that has not been given away and go with the money to Dorestad. There are there many churches, priests and clergy and a multitude of poor people. On your arival seek out faithful persons who may teach you how to distribute this, and give away everything as alms for the benefit of my soul'. Catla did so. 17 Since there must have been many deserving poor in Birka and its neighbourhood, Frideborg insisted that Catla should go to Dorestad because she wanted the recipients to be Christians and the distribution to be under the guidance of clergy. She seems to have included among the poor the pauperes Christi, the servants of Christ who were vowed to poverty, and at the time of Frideborg's death there was only one priest in Birka. Another consideration was almost certainly that Frideborg feared that, after her death, Catla would be vulnerable to pressure from relatives who did not share her enthusiasm for the new religion and were, more likely, hostile to it. By going abroad Catla could fulfil her mother's wish without interference. It can safely be assumed that in Scandinavia, as in other parts of Europe, women were not only among the earliest converts but were also generous donors to the infant church and were also active in the work of evangelism, encouraging their husbands to convert and teaching the new faith to their children. A rune-stone at Enberga in Uppland, erected by two brothers in memory of their parents, implies that only their mother was Christian. It was erected in a pagan cemetry and the inscription reads ' Gisl and Ingemund, good "drengs", had this monument made after Halvdan, their father and after Ödis their mother. May God now help her soul well'. 18 All the prayers in the inscriptions are concerned with the soul of the dead and are based on the funeral liturgy of the western Church. The 17 Rimbert: VA 20. 18 SR Uppland 808. 11  B. Sawyer: Women in Viking-age Scandinavia most common are 'May God help his/her soul/spirit' and 'May God and God's mother help his soul'. It is significant that Mary is never invoked as a virgin but always as a mother; this may have been due to the high esteem in which fertility had been held in pagan Scandinavia and facilitated the acceptance of the new faith. Britt-Marie Näsström has even suggested that Mary may have replaced the fertility goddess Fröja; many of Fröja's functions were taken over by Mary, in rites connected with childbirth, weddings and fertility of the soil. In pagan Scandinavia women 19 had played an important role in Fröja's cult, and it is plausible that Mary's popularity was especially great among female converts. Anne-Sofie Gräslund has pointed out a clear correlation between female sponsors and the use of the prayer to 'God and God's mother'; while less than one third of the shorter 'May God help' inscriptions involve women as sponsors or deceased, women are represented in almost half of the 'God's-mother' inscriptions.20 Conversion to Christianity meant many fundamental changes (see ch. 'Scandinavia in the Viking-age' in this volume), but the revolution was not complete. Some features of the old religion survived if in modified forms. This was partly because Scandinavian beliefs had been long influenced by Christianity. The Church condemned many rituals and practices but it had to accomodate some. The survival of magic in Christian form is well illustrated by the inscription on a Norwegian rune-stick:'Mary gave birth to Christ, Elizabeth gave birth to John the Baptist, Be delivered in their names. Come ourt child. The Lord is calling you into light!' The female helpers of 21 pagan times were replaced by Christian saints, above all by the Virgin Mary. The 145 inscriptions that refer to bridge building confirm the leading role of women in the Christianization period. Missionaries taught that it was a meritorious act to build a bridge or causeway 'over deep waters and foul ways for the love of God'. A surprisingly large proportion of these 'bridge-inscriptions' were commissioned by, or erected in memory of, women. There is also one eleventh-century inscription commemorating a 22 woman who planned to make a pilgrimage to Jerusalem. One Swedish 23 inscription, sponsored by two daughters in memory of their father, reports 19 Näsström 1996, pp. 335-348. 20 Gräslund 1996, p. 327. The figures given are 31% and 47% respectively. 21 Jacobsen 1984, pp. 104-5. 22 B. Sawyer 2000, pp. 134-6 23 B. Sawyer 2000, pp. 139-40. 12  B. Sawyer: Women in Viking-age Scandinavia that their father, Tore, had had a seluhus (soul-house) built after his wife Ingetora. 24 The inscriptions were predominantly Christian monuments, manifesting the acceptance of the new religion by the sponsors or the dead, and implying a willingness to give the missionaries active support. It is therefore significant that almost a quarter of all inscriptions involved women as property owners. Most of them were widows who could expect support from churchmen who urged them not to remarry, presumably in the hope that at least part of their property would be given to them as an endowment for their churches. The runic inscriptions offer many examples of sole heiresses, and obviously their number increased during the eleventh century, partly because of women's higher longevity (see ch. 'Scandinavia in the Viking- age' in this volume), partly because, after the conversion, the exposure of children gradually ceased. Since female babies had run the greatest risk of being exposed (abandoned or just actively negleced), conversion resulted in an increased number of women in the population. Secondly, male mortality was very high in the early Middle Ages; apart from extremely high infant mortality, dangerous expeditions, feuds and warfare took a heavy toll. The life expectancy of women who survived their fertile period was especially favourable. The contrast - a construct /The constructed contrast (???) Could it be concluded that in Viking-age Scandinavia women had a higher status, greater freedom and fewer restraints on their activity than later, i.e. after the conversion to Christianity? In our source material we mainly meet women from the upper strata of society, so the answer is valid only for them, not the anonymous mass of women, who occupied less exalted positions. It is true that many Viking-age women had both freedom and many opportunities to act, to take responsibility, administer farms, inherit and dispose of property, but so had their medieval sisters. The main difference seems to be the fact that after Conversion women's reputation of wisdom was undermined and their magical skills were condemned. On the other hand, new opportunities for women's influence opened, e.g. as supporters of missionaries, builders of bridges, donors to the church and religious houses, and abbesses of nunneries. 24 SR Uppland 996 (Karberga). 13  B. Sawyer: Women in Viking-age Scandinavia The conclusion is thus that 'the strong Nordic woman" did not disappear after the Viking-age, even if they disappear in the literary sources dealing with Christian times. Why did the twelfth-century authors have them disappear? A good starting-point is to see what Saxo actually thinks about strong women. As far as the shieldmaidens are concerned he stresses how un- womanly they were, condemning their wish to defend their chastity with weapons. In living independently, not subordinating themselves to any man they were presumptuously defying the order of Nature, and it is with ill-disguised satisfaction Saxo has them all defeated in the end, either in battle, or forced by male heroes to marry and become obedient wives and mothers. In his history things always turn out badly for women who want to maintain their independene. Life-long chastity is not considered a virtue, a woman's natural career is to be a wife and mother. Saxo also objects to the female inheritance rights, which guaranteed also unmarried women a share of the family property. Apparently married women's property was not thought to be endangered as long as their husbands had control over it, and this is no doubt the implication of Saxos' propaganda for the subordination of women, a view that was supported by patristic arguments proving that women were inferior to men in every respect. It should also be stressed that in Saxos work contemporary chaste women, i.e. nuns - are conspicuous by their absence (as are indeed the monks). The Icelandic family sagas portray remarkably few unmarried women, and this has been explained as due to the weight the authors put on marriage and procreation. We actually only know about six Icelandic nuns 25 altogether between ca 1000 and ca 1300; the first, Gudrun (in Laxdœla saga) had had four husbands, a few lovers and children, and she did not become a nun till fairly late in life. Two other nuns are also said to have had husbands and children, and of the remaining three, two nuns actually lived in celibacy. The author follows up this information with anecdotes illustrating the fatal effects of such life-long chastity: one had problems with her eye-sight, and the other developed mental trouble. It is obvious that in Iceland chastity as such was not highly valued. To quote Roberta Frank, who has studied marriage in twelfth- and thirteenth century Iceland: In the family sagas [...] one way to find out who is the villain is to locate the nearest bachelor, if you can find one. He is usually 25 Frank 1973, pp. 473-84. 14  B. Sawyer: Women in Viking-age Scandinavia an outlaw, a thug, a poet, or worse [...] When there are two unmarried protagonists in a saga, on the one hand the man who openly declares that he wants nothing to do with women will be the greater scoundrel. On the other hand, the bachelor who demonstrates his heterosexual interests - however crudely - is judged redeemable. To make a bachelor like Grettir more respectable, the saga-author places him in a variety of amorous encounters - represented in most English translations of the saga by a series of blank spaces and dots [...] 26 The contempt with which the Icelanders regarded celibacy is reflected in their language; the Icelandic for bachelor (einhleypingr) means vagabond or scoundrel, and the term used for a spinster also means 'one who has bad luck' (úgiptr).27 Saxo and his Icelandic colleagues thus agree in their opposition to life-long chastity, and as far as women are concerned their message is simple and unambiguous: a normal woman marries and is praiseworthy when she devotes herself to her husband. Most ecclesiastical writings present a striking contrast; in them the ideal woman is the Virgin - or widow - who, devoted only to God, withdraws from an active life and gives her property to the church. This contrast well illustrates the resistance that church still had to fight in thirteenth century Scandinavia. The misogynistic propaganda of the day, so obvious in Saxo's Gesta Danorum, opaque but present in the Icelandic sagas, is, I suggest, an expression of the fear secular society must have felt for the consequences that the economic politics and moral ideals of the church might lead to. Such propaganda would not have been needed if women really were as weak, inferior and powerless as they are often said to be.28 Since misogynistic propaganda was obviously needed, we can conclude that women, by alienating family property to the church, were felt to constitute a real or potential threat to society. As Christian writers, neither Saxo nor his Icelandic colleagues could criticize the ecclesiastical ideal of virginity openly. They had to express their disapproval in other ways. From holy virgins to pagan 'shield- maidens' may seem to be a large step, but in fact both groups shared the same basic intention, to abandon their traditional role and defend their independence, using chastity as their weapon. This is probably the reason 26 Frank 1973, p. 481. 27 Frank 1973, p. 482. 28 Nordberg 1984, p. 118. 15  B. Sawyer: Women in Viking-age Scandinavia Saxo and his Icelandic colleagues (perhaps inspired by classical traditions about Amazons) introduced female - chaste! - warriors into Scandinavian pagan history, and by criticizing their wilful strength and independence they were able to express opposition to their Christian counterparts. In doing so they created the myth of the Nordic 'shield-maiden'. 29 Conclusion The answers to the two questions presented in the introduction will thus be that whatever the real or literary background of the shieldmaiden motif is, these un-womanly female warriors served important purposes in twelfth- and thirteenth- century literature, illustrating the dangers with uncontrollable and independent women. The contrast between active pagan and passive Christian women was constructed by the medieval authors in order to teach their contemporaries that female activity belonged to bygone times and ought to be suppressed now that divine order had been estabalished. It is obvious that the high valuation of life-long chastity met with strong resistance in Scandinavia; the woman ideal that was favoured - the wife and mother, subordinated to a man - was certainly not a novelty, introduced with the church, even if patristic teaching on women's inferiority could support and inspire authors like Saxo. A new economic reality had made a stricter control of women necessary; from the eleventh century onwards many more women reached adulthood, some of whom remained unmarried, and of those who married many survived their husbands and could as widows dispose of great properties. By means of legal rules and progaganda landowners tried to möta (?) the threat posed by the inclination of single women to donate property to churches and religious houses. If we are to use the concepts 'shieldmaiden' and 'madonna' at all, we can finally conclude that the woman-ideal that was honoured in pagan as well as in Christian Scandinavia cannot have been very different from the one still honoured, i.e. the wife and mother with features of the madonna in subordination and self-sacrifice, and with features of the shieldmaiden in strength and perseverance! 29 B. Sawyer 1998. 16  B. Sawyer: Women in Viking-age Scandinavia References Bandel, Betty (1955)'The English Chroniclers' Attitude toward Women', Journal of the History of Ideas 16, pp. 113-18. Clover, Carol (1986a), 'Maiden Warriors and Other Sons', Journal of English and Germanic Philology, pp. 35-49. Clover, Carol (1986b), 'Hildigunnr's Lament', in Structure and Meaning in Old Norse Literature: New Approaches to Textual Analysis and Literary Criticism, Odense, pp. 141-83. Farbregd, Oddmunn (1988), 'Kvinneliv i vikingtid; kven var kvinnene som ligg i langhaugar?', Kvinner i arkeologi i Norge (K.A.N.), pp. 3-23. Frank, Roberta (1973) 'Marriage in Twelfth- and Thirteenth-Century Iceland', Viator 4, pp. 473-84. Frank, Roberta (1990) 'Ornithologoy and the Interpretation of Skaldic Verse', Saga-Book of the Viking Society 23, pp. 81-3. Gräslund, Anne-Sofie (1996), 'Kristnandet ur ett kvinnoperspektiv', in Kristnandet i Sverige; gamla källor och nya perspektiv, ed. Bertil Nilsson, Uppsala, pp. 313-34. Jacobsen, Grethe (1984) 'Pregnancy and Childbirth in the Medieval North: a Topology of Sources and a Preliminary Study, Scandinavian Journal of History 9, pp. 91-111. Jochens, Jenny (1986), 'The Medieval Icelandic Heroine: Fact or Fiction?', Viator 17, pp. 35-50. Mundal, Else (1988) 'Forholdet mellom born og foreldre i det norrøne kjeldematerialet', Collegium Medievale 1988, pp. 9-26. Mundal, Else, and Steinsland, Gro (1989) 'Kvinner og medisinsk magi', in Kvinnors Rosengård: föredrag från nordiska tvärvetenskapliga symposier i Århus aug. 1985 och Visby sept. 1987, ed. Hedda Gunneng et al., Stockholm. Nordberg, Mikael (1984) Den dynamiska medeltiden, Stockholm. Näsström, Britt-Marie (1996) 'Från Fröja till Maria: det förkristna arvet speglat i en folklig föreställningsvärld', in Kristnandet i Sverige; gamla källor och nya perspektiv, ed. Bertil Nilsson, Uppsala, pp. 335- 48. Rimbert (1961) Vita Anskarii (VA). In Trillmich and Buchner, 1-133. English translation by C. H. Robinson, Anskar, the Apostle of the North, London: Society for the Propagation of the Gospel in Foreign Parts, 1921. 17  B. Sawyer: Women in Viking-age Scandinavia Sawyer, Birgit (1998) 'Sköldmön och madonnan; om kyskhet som ett hot mot samhällsordningen', in Myter og humaniora, ed. K.E. Haug and B. Mæhlum, Oslo, pp. 97-122. Sawyer, Birgit (2000) The Viking-age Rune-stones: Custom and Commemoration in Early Medieval Scandinavia, Oxford. Saxo, Saxonis Gesta Danorum (1932), vol. 1. Ed. J. Olrik and H. Ræder, Copenhagen. English translation of books 1-9 by Peter Fisher, Saxo Grammaticus History of the Danes, vol. 1, Cambridge. Sobol, Donald, The Amazons of Greek Mythology, 1972. Steinsland, Gro (1985), 'Kvinner og kult i vikingetid', in Kvinnearbeid i Norden fra vikingtiden til reformasjonen; foredrag fra et nordisk kvinnehistorisk seminar i Bergen, augusti 1983, ed. Randi Andersen et al., Bergen. Strand (now Sawyer), Birgit (1980), Kvinnor och män i Gesta Danorum, Göteborg. Sveriges Runinskrifter (SR) (1911 -.) Kungliga Vitterhets Historie och Antikvitetsakademien, Stockholm. Sørensen, Preben Meulengracht (1980), Norrønt nid, Odense. 18