The Absence of Gender Iron Age Burials in the Lake Mälaren area

The article discusses the extent to which conceptions of sex and gender roles can be studied and interpreted on the basis of I ron Age grave material. In accordance with the results of excavations of lron Age cemeteries in central Sweden, it is here argued that the articulation of sex and gender roles was a rare phenomenon in mortuary contexts. Although the articulation of sex and gender is very limited in the artefact material, it is interesting to note that women's graves appear more oAen than men's graves. Accordingly, this raises questions about women's position in the political and social life of Iron Age society.

The material studied in this article consists of excavated Iron Age graves, selected from cemeteries in the districts of Upland, Södermanland and Västmanland in the Lake Mälaren area.There is a slight bias in frequency in favour of selected graves from the Late Iron Age, which is from the Migration period to the Viking Age.It is important to note that Iron Age graves are not a uniform category.There is no homogeneous or unified burial practice that can be described as representative on the overall level and variations exist in both the external and the internal layout of the graves.However, it is not the variations within or between Iron Age cemeteries, nor the variations over time that I primarily want to discuss in this article.The main discussion will instead focus on the extent to which the buried person's sex and gender can be interpreted on the basis of the Iron Age grave material.In short, what can the results from excavated cemeteries tell us about sex and gender during the lst millennium A.D.? SEX AND GENDER -A SHORT DESCRIPTION OF THE BACKGROUND At the time of the Enlightenment, during the late 18th century, the duality model of two sexes (Sw.tvåkönsmodellen) replaced an older view, in which only one sex, the male, existed (Arwill-Nordbladh 2001).In the older "one-sex model" (Sw.enkönsmodellen), women were assumed to have the same physical character- istics as men, although not fully or properly developed and not organised in the body in the same way.During the introduction of the "duality model", the most significant characteristics of the body became polarized in terms of male and female.It is interesting to note in this context that the polarization between male and female was not restricted to features related to the physical reproduction, but also to other characteristics that were thought to be significant of either male or female.The introduction of the duality model has ever since contributed to maintain a hierarchy of values between the sexes, in which the female sex have been played a subordinate role in relation to the men.Many feminist researchers, historians and sociologists of today rightly argue that this biological essentialism has also given the inequality between the sexes a naturalized character that needs to be questioned (Arwill-Nordbladh 2001; Bourdieu 1999: Laqueur 1990).
The growing need to question the biological essentialism will, it is to be hoped, lead us further in our attempts to interpret men's and women's social relations, values and attitudes in different historical contexts.Recent research in history and women's studies has established a distinction between the biological (congenital) sex and the socially constructed sex (Ericsson 1993:6).Men and women are concepts that in a biological sense are conceived of as unchangeable over time, despite the historical context, while the socially constructed sex is a cultural phenomenon that is variable in time and space.However, it is not easy to maintain a clear division between the social and the biological concepts of sex.Several studies have stated that a great many of what have earlier been assumed to be biological characteristics of men and women should rather be conceived of as a part of the socially constructed sex role (ibid).What has been regarded as male Current Swedish Archaeology, Vol. 12, 2004 Fig. 1.Woman treading the bellows of an ore blast and simultaneously spinning (Jirlow 1931t114).and female over the years is not constant or fixed, because what are usually conceived of as "typical" men's or women's professions in one historical context may be radically altered in another (Fig. 1).
The knowledge of this has therefore led to the introduction of the term "gender" (Sw.genus), which focuses the attention not only on the socially constructed sex roles, but also on the values and norms that surround the relations between men and women (Hirdman 1993:148).
The theoretical discussion about what really defines the concepts of sex and gender is neither within easy reach nor clear-cut, and there is a risk that the discussion about terminology may lead to a dead end.Congenital, biological sex has usually been regarded as the fundamental constant that conditions the categorisation between what is male and what is female.It is also here that the heart of the matter lies.Sceptical voices argue that it is the conceptual division between "male" and "female" that is hindering us from going further in the gender discussion (for example, Strassburg 1997:93).
But is it really the conceptual division and the analytical categorisation that are the main problem here?Terms like "male" and "female" are definitely not to be conceived of as constant norms in a universal sense, but I am not fully convinced that it is satisfactory or sufficient to exclude them from the discussion.The whole issue is a matter of socialisation, a question about the changeable relations between the sex roles and the social groups.We also need to discuss the contextual significance of biological (congenital) sex if we wish to get closer to the socially constructed relations.We cannot disregard the possibility that biological differ- ences, as in the roles in reproduction, for example, may be utilized in our concepts of how we socially define each other in various ways.

SEX AND GENDER IN THE ARCHAEOLOGICAL RECORD
The many excavations of Iron Age cemeteries that since the 1970's have been carried out in the Lake Mälaren area, undeniably constitute a useful basis in discussing questions concerning the constitution of the buried population.Older studies of the Iron Age grave material have usually focussed on questions directed to the settlement unit and the development and change of the settlement structure (for example, Ambrosiani 1964; Ferenius 1971; Hyenstrand 1974), although the social status and the sex of the buried have also been discussed (for example,    1969).The number of theses', articles and papers discussing questions about how gender and sex affiliations may be distinguished in the archaeological record is now so extensive that it is not possible to present a comprehensive outline of what has been published on this matter here, although a few of the more important discussions can be mentioned, such as Arwill-Nordbladh (2001), Göransson  (1999), Hjerungdal (1994), Johnsen-We linder (1997), Strassburg (2001) and  Serensen (1991).
It must be declared from the start that there is no certain or absolute method for determining the attributing of sex in the archaeological or osteological record.The structural principle for the duality model of two sexes stems from an assumption that men and women, in a material sense, represent each other differently, not only in the living world but also when they are dead and buried.This principle is axiomatically implicit in most interpretations, but it also has problems and limita- tions.Bo Petré, who has analysed the grave material from Lovö in the district of Upland, writes as follows in his thesis: "Some artefacts can, according to current values, be attributed to either men's or women's graves; ornaments and jewellery belong to women and weapons to men" (Petré 1984:191).The problem in this example is, above all, connected with our current opinions and values concerning what materially expresses maleness and femaleness in these contexts.The factual biological difference that exists between men and women is not sufficient to make clear how we should understand the social significance that these differences may give rise to (Moore 200:317).The social domain is more or less always structured according to specific rules of gender, in which different tools and objects are associated with different activities and sex roles.At the same time, there are anthropologists who argue that these gender-related domains in life are simply reflections of the obvious division between the biological sexes.Of course, this is a simplifying way of describing the complexity of the matter (Moore 2000:318).Gender relations are to some extent dependent on the local context, and it is important that the perspectives of gender do not become isolated from the surrounding, socio-political and socio-economic structures that maintain and legitimate the existing relations of power within the given society.
Advocates of the gender perspective have stressed the importance of how the relations of power within the social world, define sex differently from one period to another (Hirdman 1993; Moore 2000).A person's gender is activated in certain situations and contexts in which the existing norms of society allow it.These norms may change and what are assumed to be typical female activities at one time may be assumed to be typical male activities at another.To be able to continue the interpretation of how to understand and describe ancient gender categories and how they vary over time, it is also of importance to use osteological analyses to identify the biological sex of the buried person.Although the osteological method has limitations when it comes to classifying a person's sex or age, its relevance is as important as the archaeological interpretations of the artefacts.
There are several factors in life that are involved in the creation of a person's Curren( Swedish Archaeology; Vol.l2, 2004 gender.Besides the biological sex affiliation and biological age, there are kinship relations, social competence, technical skills, etc. that can be mentioned as important factors.The surrounding society and its relations of power and reproduction of norms are just as much important, but these factors are perhaps less easy to get hold of.We have to return to the question of what is left for us to interpret in the grave material?What may we learn about the life and social position of a buried person?According to the British archaeologist Michael Parker Pearson (2000), the burial custom is often subjected to various forms of manipulation with the purpose of naturalizing or playing down the existing relations within the social structure.The treatment of the buried therefore does not always reflect the existing differences between individuals and groups in the society.
The roles that are portrayed in death ritual are expressions of status which must be seen as relating to, rather than "reflecting", social position [.. .] Thus in death ritual it is not necessarily the case that the actual relations of power are displayed.It does not follow that those social identities which embody the greatest degree of authority will always be expressed; however, it is important to understand why certain roles are expressed in death as well as in other spheres of social life (e.g.house form, dress, display of material possessions etc. ) and also to understand the extent to which they are used as social advertisements between competing social groups.(Parker Pearson 2000:249).
It cannot be taken for granted in advance to what extent the burial practise naturalizes or conceals the existing differences in society.An illuminating example of how the burial ritual sometimes plays down and masks the order of social relations in the living society is the burial practice of today.At present, we are living in an economically divided and sex-segregated society, but this is not observable either in our funeral rituals or in the graves in our cemeteries.It is, on the other hand, interesting to note that there are certain variations that appear in the burial practice during the Iron Age.These variations are related not only to the external form of the graves but also to the deposition of artefacts.Of course, there is a possibility that the death ritual in fact directly reflects the existing order in society, without any manipulation.It is also important to be aware of this when discussing the ideological background behind the burial custom.
There are several reasons why it is important to apply a gender perspective in the discussion of men's and women's relations in the social world.Many sociohistorical descriptions of men's and women's roles in the archaeological literature show a quite stereotyped pattern.An often-depicted stereotype is, for example, the "bigwig chief", the successful horseman or warrior (for example, Hedeager 1990; Simonson 1969; Sten & Vretemark 1988).In these descriptions, men are usually placed in advanced positions, in politics, exchange and public life, while the status of women, not surprisingly is, generally given a lower profile.1980:117).It is also notable that when images of women occur in the archaeological literature, they are often depicted in a passive position, often with their hands in their knees and their heads directed downwards (Fig. 2).Men never appear in such postures.
The term "chief's grave" (Sw. stoi7rtannagrav) is a well-established concept, which has served to maintain the asymmetric relation between the sex roles.One need not be too much of a critic to react suspiciously to the use of this concept.A rich and powerful man, buried in a large grave mound, together with prestigious finds, perhaps in company with his slave, is still a very often-depicted category in the interpretative discussion of the grave material (for example, Sten k Vretemark 1988:145ff).
The continuing use of the concept may at first seem simplifying and quite harmless, but it reveals both conscious and unconscious conceptions of a hierarchical and segregated order between the sexes.The cultural and historical depiction of the past, which usually tends to focus on the men's domination in social life, functions as a constant work of construction in the mind of the interpreter, male or female (Bourdieu 1999:69).It is important to be aware of this when describing sex and gender roles in the past.As I noted at the beginning, it is far from certain that we can maintain a strict division between the sexes on the basis of finds that we have traditionally interpreted as male or female.This can be illustrated by looking at the occurrences of axes and tools in 80 osteologically examined and sex-determined graves on Gotland during the Neolithic (Table I).The compilation has been made on the basis of excavated graves from seven cemeteries, and it is worth mentioning that there are local differences between the cemeteries.Anyway, the purpose of this example is to show that similar types of tools during a specific period of time may occur in both male and female graves.Accordingly, we have no reason to draw any far- reaching conclusions when it comes to the occurrences of certain artefacts in the graves because they apparently appear in the graves of both men and women.
What is interesting to note here is that the distribution of fishing tools and axes does not show any exclusive occurrence to either sex.One exception is the occurrence of harpoons, which exclusively occur in male graves.The fishing tools dominate in male graves (70 %), while stone and flint axes are quite evenly distributed between the male and the female graves (52 % and 48 %).If we were to discuss the order of the artefacts in the graves on the assumption that they directly represent the relations in the living society (Welinder 1997:72), we would be confronted with an equal division of labour and status between the sexes.If the grave material directly reflects the existing order in the contemporary society, we can conclude that Neolithic society did not stress any major differences between male and female activities.Of course, it is possible that the death ritual concealed, hid or played down the existing relations in the living society (for example, Parker Pearson 2000:249 above), but this line of argument needs to be confirmed by complementary evidence.Anyway, with reference to the Neolithic grave material from Gotland, there is nothing that indicates any major categorisation between men and women.

EXEMPLIFYING IRON AGE CASES FROM THE MÄLARDALEN AREA
Only a minor part of the graves in the Iron Age cemeteries can used for the analysis of the sex of the buried individuals.This has earlier been pointed out in various archaeological contexts, both in osteological analyses of the bone material (Iregren 1991), and in discussing the significance of the artefacts in the graves (Hjerungdal 1991; 1994).In discussing the sex of the buried individual, most archaeologists usually depart from traditional criteria.By "traditional" criteria I mean a norm or order that considers male and female objects as a universal constant (for example, Bennett 1987; Petré 1984).
Table 2 shows to what extent the grave material from the Lake Mälaren area can answer our questions about the sex of the buried individuals during the Iron Age.The table consists of 16 cemeteries of different sizes, comprising over 600 burials.The cemeteries are dated to the Early and Late Iron Age, although with an emphasis on the latter period, that is roughly from the Sth to the I lth century A.D. The result of the osteological analysis shows that approximately 9 % of the Currenl Swedish Arehaeology, Voh l2, 2004  total number of burials can be estimated according to sex.Although this is an average figure, it points to the fact that only a very small part of the burials can be determined.
If we tum to the archaeological criteria for determining the sex of the buried the result is almost equally limited.According to traditional criteria, women's graves are characterised by specific finds related to the dress, such as oval brooches, button-on-bow brooches and ornaments, such as bead-string spacers, arm rings, bracelets, beads (&10), etc. ( Hjerungdal 1991; Jansen Sellevold et al. 1984; Andersson 1998; Petré 1984).Keys and spinning whorls occur very seldom as grave finds, but when they occur, they are here related to women's graves.
"Hairpins" are not diagnostic of any specific sex and are therefore excluded from the compilation in the table.Artefacts related to men's graves are weapons and weapon accessories.The aim of putting the characteristics of men's and women's artefacts together with the osteological analysis is to find out to what extent the traditional view of how the distribution of the sex of the buried appear in the Iron Age grave material.
The average percentage of the graves that, according to traditional criteria, can be discussed in terms of either male or female is 12 % (Table 2).It is worth noting that grave finds indicating female burials occur more frequently (11 %) than burials containing weapons and weapon accessories, which occur only in  1% of the graves.The generally fragmented find situation in the Late Iron Age cemeteries in the Lake Mälaren area is exemplified in Tables 3 and 4. It is true that there are also a small number of extraordinarily rich and poor, Iron Age cemeteries, but these are far from representative in a general view.The two cemeteries in the district of Västmanland clearly show the difficulties of tracing the articulation of sex and gender in the burial material during the Late Iron Age.
When discussing the distribution of male and female graves in the Iron Age cemeteries it is also important to pay attention to the osteological analysis.The preconditions for determining the biological sex from cremated remains are also problematic and complicated because of the fragmentary condition of the bone material (Iregren 1991:104).The archaeological discussion about the distribution of male and female graves has therefore been focussed more on the interpretation of the artefact combination in the burial context than on the osteological results (Bennet 1987; Petré 1984).In those cases in which the interpretations from the osteological evidences and the artefacts do not correspond, this is often explained by "uncertainties or error of judgement" in the osteological analysis (Bennett 1987:102).
To assume that sex determination made on the basis of artefact analysis will be more credible than on the basis of osteological analysis, is, of course, hardly reliable.The archaeological and the osteological analysis should instead be combined and interact if we wish to achieve fruitful results.Anyway, my con- cluding point is that only a small number, approximately 12 %, of the Iron Age burials, are diagnostic in respect of the traditional criteria for determining male and female graves.The situation is almost the same when we look at the extensive grave material from northern Spånga in Upland (Biuw 1992).Of the 385 Iron Age graves, not more than 20 % can be discussed in terms of male and female burials.Only four of the graves contain weapons or weapon accessories, that is, I % of the burials, while artefacts associated with women occur in as many as 17 % of the graves.According to the compilation made in Table 2, it is fairly reasonable to assume that the relationship between male and female burials is significant for the general picture of the Iron Age burial tradition in the Mälardalen area.There are a few exceptions that deviate from the general description, and I will exemplify this in the next section.

THE APPEARANCE OF GENDER IN A CLASS-RELATED CONTEXT
It is evident that concepts of sex and gender do not seem to have been of any primary significance when we look at the artefact material in the Iron Age burial ritual.There are a limited number of graves on the Iron Age cemeteries that can be discussed in terms of male and female burials.According to Tove Hjerungdal (1991) it is plausible to think that the ancient concept of sex and gender affiliation was primarily restricted to the upper social classes in society, where strategies of constructing and maintaining gender roles may have been considered to be important.Following this line of argument, it is notable that the articulation of gender roles in the Iron Age grave material was more frequently expressed in women's graves.I find Hjerungdal's assumption most interesting because it draws attention to the gender issue in a class-related perspective.A relevant question is why only men and women in the higher social classes needed to express affiliation in terms of sex and gender?With the aim of shedding some light on the relation between gender and class, I am now going to look at the burial material from Birka in the Lake Mälaren area.The social structure at Birka was well debated during the last century.The chamber tombs (Fig. 3) and rich inhumation burials have, not surprisingly, been considered to be representative of the upper social strata in the Birka community (for example, Gräslund 1980:77ff).These rich burials have been described in terms of merchants' graves, warrior-class graves or graves of the royal admin- istration.Although chamber tombs containing both women and men occur, the significance of the women's tombs have attracted less attention in the literature describing the roles and positions of men and women on Birka (Johansson 2003).
However, it is hardly controversial to assume that the persons buried on Birka, especially those in the chamber tombs, also represented a social class with quite Current Swedish Archaeology, Vol./2, 2004 Fig. 3. Chamber graves from Birka.To the left, the male burial no.977 and to the right the female huri al no.965 (after Arbman 1940-43).
extensive power in the local community, not only with respect to economic wealth but also in terms of social and political relations.The occurrence of weapons and oval brooches in the chamber tombs can be used to illustrate the articulation of sex and gender on Birka (Fig. 4 and Table 5).The account of the chamber tombs derives from the studies of Holger Arbman  (1940-43) and Ann-Sofie Gräslund (1980), and it is interesting to note that those tombs show a polarized division between those containing weapons and those containing oval brooches.Approximately 49 % of the chamber tombs can be interpreted as male and 37 % as female burials.No osteological analysis has been made of the bone material and the determination of sex is based on traditional criteria (Gräslund, personal comment, 2003).
According to the traditional view of the way of determining male and female graves, the chamber tombs on Birka clearly articulate a polarized pattern of male Current Swedish Archaeology, Vol. 12, 2004  Fig. 4. Accessories ofornaments and weapons occurri ng in the chamber tombs on Birka (after Roesdahl  et al. l 992.239, 304).and female symbols which has no correspondence to the general burial ritual in the Mälardalen area.With reference to Tove Hjerungdal's discussion (1991), it is possible to argue that concept of sex and gender may not only be a class-related question but may also be a question of habitus or practical consciousness (for example, Moore 2000).Although a person's habitus is to a certain extent governed and maintained by his or her position in the social community, the common interests of the class to which he or she belongs are also an important factor (Bourdieu 1990).The finds of weapons and oval brooches in the chamber tombs appear in such a standardised and polarized manner that it seem reasonable to conceive of them as sex-and gender-related symbols, restricted to the conventions of a specific social milieu.Only 6 % of the chamber tombs deviate from the polarized pattern.
The question why the Iron Age graves in general articulate the individual sex in such a limited way does not perhaps have a straightforward answer.Although the Iron Age burial ritual did not generally express the identity of the buried clearly, it is evident that those of individuals of the higher classes did.Certain
Current Swedish Archaeo!ogy, Vot. 12, 2004 concepts, such as gender and sex affiliation, were closely related to the conventions of the upper classes.The limited number of Iron Age graves that can be classified as male or female instead indicates that these concepts were not supposed to be articulated outside the superior classes.In the history of archaeological research, the hierarchical division between the sexes has been central to the description of Iron Age society.The social setting has, without exception, been described as a society dominated by men, represented by petty kings, chiefs and "big men".
However, this description does not correspond very well to the general occurrence of male and female artefacts in the grave material.
In his article "Rich or poor" (1998), Kent Andersson discusses women's and men's graves dating to the Roman Iron Age in the Mälardalen area from a social and chronological point of departure.He concludes that there are a large number of rich female burials, compared with male burials, within the studied area but that these women should not be interpreted as having been rich in real life.They belonged to a group whose legitimacy was based on the manifestation of rich female burials: In this article the graves have not been seen as material manifestations, but rather as ideological ones, using the funerals to legitimate the growth of a new elite at this time.The "rich" women's graves should therefore be seen as signals of change in the society, which are also possible to trace in e g the settlement find material.In other words, the women were neither rich nor aware [politically conscious].Instead, they belonged to a group in society which now felt the need for ideological manifestations and who came to be buried in a way, which deviated from other contemporary burials.The women can be seen as pawns in a game, which eventually led to a more organised, political structure in Upland-Västmanland.(Andersson 1998:89).
Andersson's explanation of why there is a greater number of rich female graves during the Early Iron Age is, in my opinion, not acceptable, because it postulates a hierarchical and much too static relationship between men and women.Why, for instance, are male burials interpreted by other standards than the female ones?It really seem problematic to many archaeologists to consider rich female graves as socially significant in the past.Anders Götherström, for example, who has made molecular analyses from the two cemeteries Badelunda and Alsike, states that "these two boat cemeteries are peculiar in a sense; [since] they contain women" (Götherström 2001:23, my italics).

CONCLUDING REMARKS
This article has discussed to what extent the concepts of sex and gender roles can be studied and interpreted on the basis of Iron Age grave material.According to the results revealed by excavations of Iron Age cemeteries in central Sweden, it is here argued that the articulation of sex and gender is very limited in the artefact material.Although this articulation occurs less frequently in burial contexts, only Current Swedish Archaeology, Voh l2, 2004 12 % of the burials can be discussed in terms of female or male and it is interesting to note that female graves are more often represented than male graves.When the articulation of sex and gender occurs in burial contexts, it is restricted to a small proportion of the graves, mainly in high-status burials containing weapons, ornaments and dress accessories.However, it is important to state that the traditional criteria for interpreting sex and gender relations are problematic in a general sense.Although they may be relevant for use in some contexts, as in the case of the chamber tombs on Birka, I do not find the general principle for categorising male and female artefacts in burials satisfactory.This is a sceptical standpoint but it nevertheless represents a harsh reality.Generalisation is indeed a fraudulent friend.
Anyway, in the light of the circumstances described above, it is reasonable to argue that in those cases in which sex and gender roles were articulated, they were primarily connected with class interest and especially with people who were accustomed to life in the higher social spheres.The chamber tombs on Birka may exemplify this assumption.Although the articulation of sex and gender is very limited in the grave material, it is interesting to note that women, in general, appear more often than men.It is not satisfactory to interpret the femela graves by other standards than the male ones, and this raises further questions about both men's and women's positions and their active roles in the political and social life of Iron Age society.

English revised by Neil Tomkinson.
A CKNOWLEDGEMENT My thanks go to the Helge Axion Johnsons Stiftelse, which funded my research and made it possible to carry out this study.I will also thank Elisabeth Arwill- Nordbladh for constructive comments on an earlier draft of this article.
Curreni Swedish Archaeology, Voh l2, 2004 This reconstruction of the female and ntale outfit during the Late Iran Age is an exantple of how teamen and men are depicted in the archaeological literature (after Roesdahl et ah 67, l 93).
When rich female graves appear in the Iron Age cemeteries, they have often been interpreted as those of wives or daughters of men of power and chiefs (see, for Current Sw~edish Archaeolog&:, Voh l2, 2004Fig.2.