94 THE ARCHAEOLOGY OF DEATH AND BURIAL

proposed a triad of cults and their symbols: a warfare/cosmogony complex manifested byitems such as symbolic weaponry and motifs; a native earth/autochthony symbolismrepresented by the mounds and their periodic burial under new soil as a means ofpurification; and an ancestor cult represented by temple figurines. 113 He emphasizes thatpolitical organization and religious organization are normally so closely intertwined andcongruent that there is little sense in attempting to separate them'!"

We Can find interesting comparisons for these interpretations in sixteenth- to nineteenth­century ethnohistoric accounts of the hierarchical Natchez and Choctaw who lived in thisregion and considered their ancestors to have come from Mexico. ll5 In a culmre whichworshipped the sun, the chiefs of the Natchez were called the Sun or Great Sun and wereabsolute rulers with semi-divine status. The paramount chief, the Great Sun, is describedas living in a house, built like a temple, on top of an artificial mound. Chiefs were dressedwith copper headdresses and necklaces, and their spouses, servants and subjects weresometimes sacrificed at their funerals. There is even an account of a commoner stranglinghis infant for the dead chief. The upper torso of the Sun's corpse was painted red and thethighs were covered with alternate rows of red and white feathers, prior to burial in atemple mound. Although the Suns were men, descent was matrilineal. The workings of thesystem of class exogamy ensured that fourth to seventh generation descendants of the Sunswere classified as commoners. Yet at the same time, individuals born as commoners couldachieve noble status through prowess in battle.

We have thus recovered the true meaning of 'hierarchy' as a sacred ordering. Theplacing in mounds of certain male burials, decked out in their shiny sun-reflectingsymbols, and of others with their red and white paints and copper gorgets takes on a moresatisfying meaning with new possibilities, in contrast to the sterile enquiry into statusdifferences and complexity, in which form and content are sundered and content isrejected. The Natchez case study also highlights the futility of attempting to differentiarebetween achieved and ascriptive rank since both processes were operating at the sametime. Indeed, the quest for ranking as a measure of social organization becomes almostirrelevant within a more integrated study of ideology and power.

CONCLUSION

The investigation of social complexity and degrees of status differentiation has been theprincipal concern of funerary archaeology in the last rhitty years. The idea that burialsprovide an insight into the manifold aspects of an individual's social persona has beenespecially applied in the study of grave goods, though the complexity of their interprerationshas to be recognized. Are the grave accompaniments possessions of the deceased or giftsfrom mourners? Or are they heirlooms buried with the last of the line? Their inclusion mightalso vary according to changing rules of inheritance or broader political and ideologicalcurrents in society. Equally, the meaning as well as the form of funerary symbolism has to beinterpreted: are the 'high-status' burials at Moundville or Varna political leaders, religiousspecialists or both?"' Perhaps there is an important theme to be explored in the route toabsolute power when leaders assume political and religious power simultaneously.

Funerary studies have also seen changes in approach from role theory and evolutionarysocial complexity to theories of practice and concerns with the historical and politicalsiruatedness of funerary events, from static frames of ranking and measures of complexityat anyone moment to hisrorical trajectories of successive political acts in which the livinguse the dead as resource, vision and representation.

FIVE

GEN[IER AND KINSHIP

Gender centres on the social construction of masculinity and femininity: the socialvalues invested in the social differences between men and women. In this respectgender archaeology is part of the study of social structure, as significant as rank in thesocial stratification and the evolution of past societies.'

The study of rhe archaeology of gender has been a phenomenon largely of the 1980s and1990s, with many of its themes developed along lines found in post-processual approaches.'The archaeological study of kinship has a rather different recent history, emerging brieflywithin the early optimism of the New Archaeology in the 1960s and early 1970s and now,after years of neglect, set ro make a reappearance hand-in-hand with advances inbiomolecular studies of ancient DNA and other analyses of osteological remains.

The concept of gender and its relationship to biological sex have been widely debatedand defined. In 1975 the anthropologist Gayle Rubin stated: 'Gender is a socially imposeddivision of the sexes. It is a product of the social relations of sexuality." Its most recentarchaeological definition is simply that it is the cultural construction of sexuality.- ForAlison Wylie, gender is not a given nor a property of individuals but a construct with aclear political dimension, a dynamic historical process.5 Marie-Louise S0fensen adds thatthis social construct, of central importance in the structure of past societies, is negotiatedand maintained through material culture.'

Our preconception of predetermined categories of males and females makes problematicthe understanding of sex and gender outside our own cultural milieu. Archaeologists mustwork with an awareness of the dichotomy between narural, biological sex andconstructed, cultural gender.' As Tim Taylor and Tim Yates have pointed out, we shouldbe prepared to expect considerable complexity and fluidity of sexual identities not onlywithin the assigned primacy of the masculine-feminine pairing but also within non­heterosexual categorizations.8

THE OSTEOLOGICAL IDENTIFICATION OF SEX

The possession of male or female genitalia is usually a certain guide to an individual'sbiological sex, whereas the identification of male or female characteristics from skeletalremains is less clear cut. The average adult female skeleron has a broader pelvis, a lessprominenr chin and smoother brow ridges than the male.' There is also an average differencein height. Osreologists may be able confidently to assign sex to adult human skeletons in upto 95 per cent of cases but the sexing of child skeletons is much more problematic. Sexualdimorphism in the skeletons of children can be detected in the permanent teeth and thesciatic notch on the pelvis though these tests are not conclusive.1O Even with adult remains,

96 THE ARCHAEOLOGY OF DEATH AND BURIAL GENDER AND KINSHIP 97

assignment of biological sex may be biased towards males and may explain why so manycemetery groups often appear to contain more males than females; Weiss estimates that thissystematic bias in sexing adult skeletons is of the order of 12 per cent in favour of males.ll Afurther factor affecting sexual dimorphism in skeletal remains is caused by hormonalabnormalities: an insufficiency of oestrogen in females may lead to dwarfism while aninsufficiency of androgen in males (resulting from castration when young, for example) mayaffect the growth of joints and ends of bones, leading to thin-boned long-leggedness and anelongated mandible.!2 It has been claimed that a girl's body can maintain high levels ofandrogens by strenuous exercise, leading to narrow, male-like pelves. 13 Recovery of ancientDNA (described at the end of this chapter) provides new methods of sex determinationwhich are proving their worth in difficult areas such as the sexing of child skeletons.

There are certain categories of individual who may be biologically indeterminate, eitherfrom a congenital syndrome or as a result of culturally induced bodily transformations.The former group of inter-sexuals are rare, numbering around 1 or 2 per 1000 of thepopulation, and result from conditions such as foetal androgenation (andreno-genitalsyndrome [AGS] in genetically female foetuses), gonadal dysgenesis, Kleinfelter's syndrome(three or more sex chromosomes resulting in female hormones and male genitalia) andtesticular feminization syndrome. l4 As Beth Rega points out, such individuals often occupyimportant social positions in many cultures, yet are too rare ever to constitute a major axisof sex identification. 15 Cultural transformations of bodily sex are, of course, well knownin the modern world. Trans-sexualism or the creation of additional sexual categories maybe achieved by altering hormones or through surgery. One specific hormonal methodpossibly used in prehistoric societies is the ingestion of pregnant mares' urine. 16

FEMINIST TBEORY AND THE RISE OF A 'GENDERED' ARCHAEOLOGY

There is no consensus on the definition of feminism and feminist theory, and images of ahomogeneous, ideologically coherent frame of reference are largely resisted,17 As amovement of resistance and struggle against male oppression for womenls empowerment,feminism's theoretical goals include a critique of male supremacy and the definition of sexualdifference for women. Initial rethinkings of the new women's history, anthropology andarchaeology focused on the countering of androcentric narratives, the recognition ofpowerful individual women in the past, the search for matriarchies in past societies, and theredressing of the balance through recognition of realms of women's power hitherto ignored.S0rensen has outlined two categories of archaeological sources most useful for pursuingarchaeologies of gender. l8 Those realms commonly involved in communicating gendercategories are burial activities, individual appearance through costume (much of it fromfunerary contexts), and some types of art (some of it funerary). More problematic arearchaeological remains from domestic units and behaviour which may produce informationon food production, labour division, rubbish categories and spatial order. It is very clear thatfunerary archaeology is a crucial element of any research into past gender categorizations.

Androcentric narratives within funerary archaeology

Archaeology is a continuous struggle to excavate our own preconceptions andunacknowledged assumptions. This is perhaps clearest in our attributions of meaning tograve goods, as Meg Conkey and Janet Spector have highlighted in their critique ofWinters's double-standards in his analysis of Late Archaic burials from the American

Midwest.!' When trade goods were found in a male grave Winter considered that theyindicated the man's involvement in long-distance exchange systems, whereas in a woman'sburial such items were assumed to be gifts from male relatives. Quernstones in the gravesindicated, in the case of women, that their tasks included seed-grinding and, in the case ofmen, that they were involved in making the querns! Many of our unacknowledgedassumptions permeate our interpretations - about the universality of a sexual division oflabour, gender dimorphism (the ways that men and women's bodies look different), thecommodification of sexuality, the associations of women with the private, domestic, minor,peripheral and the natural, the definition of women by their reproductive capacity and menby their social role, the exclusion of women from hunting, and the perception of certainactivities (notably hunting) as intrinsically more important than others such as gathering.20

Ian Hodder has recently remarked on his own double-standards and reappraised hisdifferential treatment of the representation of men and of women. The elaborate femalesymbolism in the Near Eastern Early Neolithic, embodied especially in figurines, could betreated as demonstrative either of their power or their powerlessness whereas the elaboratesymbolism of men in the Late Neolithic, largely from burials, was accepted without questionas indicative of their power.2 ! We may also include feminist critiques of Levi-Strauss andothers' models of kinship; such models rest on an initial premise of the exchange of women, inwhich women are considered as powerless chattels swapped by their fathers and brothers.22

Within funerary archaeology, the principal methodological issue concerned withunexamined assumptions has probably been the ascription of biological sex on the basis ofassociated grave goods and dress. With the increasingly systematic application of rigorousosteological analyses, this ought to have been largely consigned to archaeology's own dustbinbut it is still a problem in certain quarters, reinforcing contemporary gender stereotypes andfurthering the invisibiliry of potential additional or transvestite gender categories.

An interesting case is the identity of the individual buried with a range of extraordinarygrave goods within a mound at Vix, in eastern France, around 500-480 Be during theHallstatt D period of the Early Iron Age (Figure 5.1).23 The surviving skull and otherbones were initially identified as belonging to a 30 to 35-year-old woman, henceforthknown as 'the princess of Vix', but a subsequent study suggested that the sex of thisperson was not only indeterminate but might even be male.24 This led to a reinterpretationof the Vix princess as a transvestite male priest.25 Spindler's reassessment fitted with Pauli'sprevious identifications of male transvestites in two elite burials of the same period fromStuttgart-Bad Cannstatt in Germany, in which spearheads were found with femaleornaments."' Subsequently, a third osteological study has confirmed that the Vix princesswas most probably a woman and Bettina Arnold has argued that the Stuttgart-BadCannstatt individuals are also actually women, as are remains from other importantHallstatt D/La Tene A burials that Spindler and other (male) archaeologists haveconsidered to be men, indicative of male status within a male power structure.27

Sam Lucy's reassessment of sex and gender in East Yorkshire Anglo-Saxon burials findsthat, instead of the expected binary split between women with jewellery and men withweapons, there is a large proportion of 'neutral' graves or unaccompanied graves. Althoughthe osteological analyses were hampered by poor bone preservation, three women (twocertain, one possible) were buried with weapons at West Heslerton cemetery and three of theburials with jewellery from Sewerby are possibly male28 Brush had previously come up withsimilar conclusions from the mainly cremation cemetery at Spong Hill in Norfolk,2' and Lucyargues that the absence of sex-linked artefacts in a third of the graves suggests that gendermay not have been an important structuring principle in pagan Anglo-Saxon society.3D

98 THE ARCHAEOLOGY OF DEATH AND BURIAL GENDER AND KINSHIP 99

Women's power, women's wealth

A second theme of gender studies in archaeology has been to explore rhe evidence forwomen's power in past societies. Paula Webster cites both Simone de Beauvoir andShulamith Firestone as considering that women have been universally oppressed, with theroot cause of this oppression being women's restriction to a maintaining and nurturingrole, owing to reproductive labour and associated childcare burdens.32 Can archaeologyshed light on the claims of such programmatic statements?

In her book Women in Prehistory, Margaret Ehrenberg highlights certain outstandingwomen's burials, among them the VIx princess and the woman in an Iron Age chariot burial atWetwang Slack in East Yorkshire.33 Given their unquestionably high-status grave goods, thelatter two burials must be those of members of the most powerful groups within theircommunities. Liv Helga Dommasnes points our that the identity, power and significance of theNorwegian Viking Age queen buried in the Oseberg ship is often played down in favour of themen's activities of carving, shipbuilding and sailing embodied in the boat and other associatedartefacts.34 In all these cases, other archaeological and documentary sources indicate that thesewomen held positions mostly occupied by men and can be presumed to have possessed somepower. Their existence demonstrates that women were not excluded from these positions.

The problems of interpreting women's power from rheir grave goods and funerarycostumes were addressed by Hodder in his ethnoarchaeological observations on women'sjewellery among pastoralists in the Lake Baringo region of Kenya.35 In this patrilineal,virilocal and polygamous society, the low status of women is at odds with the quantitiesand quality of the ear-rings and necklaces worn. Young women might wear up to fortynecklaces but those aged over fifty might have only twenty-five or so. Were women to beburied in the costumes they wore in life, Hodder argued, then their sratus might appear tobe greater than it actually was.

Another example of women as bearers of men's wealth can be found in MarilynStrathern's study of the Mount Hagen area of Papua New Guinea. Although women'become like men' at important ceremonies their decoration is distinctly different,signifying their role as intermediaries (bearing items of exchange) rather than negotiators(enacting the exchanges). A woman's formal dance dress for moka ceremonies symbolizesthis intermediary status, the bailer shell on her front perhaps loaned by her brothers to herhusband and the pearl shell on her back intended as a gift from her husband to herbrothers. In the past, wives used to display their husbands' wealth at moka, swathed inlong ropes of cowrie shells and wearing shell aprons. Similarly, brides also used to wearsome of the bridewealth valuables.36

Many archaeologists have failed to notice their pervasive double-standard of subjectingmale burials and female burials to differing interpretations of whose wealth and goods theycontain.]7 As these two ethnographic examples show, the ownership and use in life of gravegood wealth is one of the most challenging questions for any archaeologist attempting tounderstand funerary material culture in terms of status, power and gender roles.

9

•••••

• • 18 ••

1 M

6 16

Gj7 * D ~m(Jr17

~ ~a[I]aB

L=Jl

5.1 In the wooden burial chamber at Vix lay partially preserved human remains on a wagon box (its wheels 9 tothe right of the picture), together with an enormous bronze krater (a Greek vessel for serving wine, 1 in the

picture), imported ceramics (3, 4), bronze and silver containers (2, 5-8), a gold 'diadem' on the skull (19) andother personal grave good.~ such as beads (18), fibulae (brooches; 17) and bracelets (16).

Of course, in both the Yorkshire and Spong Hill cases, the observations we can make onthe non-perishable material culture are only a partial view of the full costume worn indeath. Conclusions on the absence of a binary malelfemale gender distinction in dressmust thus remain tentative given that the clothing does not survlve. A recent reappraisal ofcertain supposedly female Anglo-Saxon burials from different parts of England has alsoled to the identification of a possible third gender of men-women, akin to rhe NorthAmerican berdache.31

Matriarchs and Amazons

The interesr in prehistoric matriarchy stems largely from claims by J.J. Bachofen in 1861and Frederick Engels in 1884 that it formed a universal stage in human culture after aninitial stage of promiscuity and prior to what Engels termed 'the world historic defear ofthe female sex'. Engels suggested an early stage in human development characterized bygroup marriage with descent traced through women and matrilocality. Women had

100 THE AHCHAEOLOGY OF DEATH AND BURIAL GENDER AND KINSHIP 101

supremacy in the household and their high sratus derived from their central positionwithin the social relations of production.

These conclusions were based not on archaeological evidence but on ancient myths andethnographic cases. It is probable that Engels' vision of matriarchal society was heavilyinfluenced by Lewis Henry Morgan's nineteenth-century ethnographic account of thematrilineal and matrilocal Iroquois, among whom the women controlled food productionand distribution, and life in the longhouse, and had a major influence on male councilelders, ritual specialists and wat parties.3B There are many other examples of matrilinealand matrilocal societies known today from other patts of the world, such as the islandersof Dobu in Melanesia,39 but none of these, including the Iroquois, would be consideredmatriarchal, with women as a group having absolute power and authority over men.

Marija Gimbutas's interpretation of Early Neolithic (10000-5000 BC) farmingcommunities as rna trifocal and probably matrilinear, agricultural and sedentary,egalitarian and peaceful and worshipping a supreme female goddess, stems from herresearch into the symbolism of female figurines and statuary from household contexts insouth-east Europe and the Near Easr.40 Although unsupported by most archaeologists, herviews have become unassailable for certain ecofeminist groups and New Agers.41 Recentcontextual studies of figurine symbolism demonstrate very satisfactorily that, despiteGimbutas's best attempts, there are no easy equations to be made between female figurinesand matriarchal societies or mother-goddess cults.42 In fact, there are some very goodreasons, discussed in Chapter 7, why the interpretation of mother-goddess worship fromthese clay figurines is wholly bogus.

There is somewhat fuller funerary evidence for the position of women in Early Neolithicfarming communities from central and western Europe around 5500-5000 BC, in the form ofcemeteries associated with longhouse communities of the Linearbandkeramik culture (referredto as the LBK). On the basis of cross-cultural generalizations about house floor areas,Ehrenberg suggests that these communities in their large longhouses may have been matrilinealand matrilocal.43 In societies with matrilocal residence and matrilineal descent, such as theIroquois, communities of sisters tend to stay together under one roof after marriage and thuslarge floor areas above 35 sq m for a single house are mostly recorded in such societies. As withany cross-cultural generalization there are always exceptions. For example, the LBK longhousesare generally thought to have had a single main hearth rather than a string of hearths as foundin the Iroquois houses; in this respect they may be more similar to the longhouses of patrilineal,virilocal groups such as the Barasana of Amazonia.44 Curiously, the LBK funerary evidence hasalso been interpreted as similarly resulting from a matrilineal kinship system, although theinferences made in that study are higWy questionable.45

Whereas Neolithic matriarchies are hard to pin down, Iron Age Amazons are mucheasier to identify in the funerary record. The ancient Greeks gave a special place in theircosmological beliefs to stories about societies of warlike and fierce women; Hippocratesand Herodotus wrote extensively about them, locating them in the region of Scythia andSauromatia to the east of the River Don. From the Ukraine and the Caucasus, dating fromthe fifth centuty BC and later, we now have a substantial number of warrior burials thatare considered to be females.46 Their weaponry variously includes armour, spears, arrowsand shields. In many cases, the osteological analyses of burials of this period are not tomodern standards and some of the 'male' skeletons may also conceivably be female giventhe likely alterations to the pelves of active, mounted women warriors.47 Taylor estimatesthat about forty female warrior burials are known from Scythia and reports that about20 per cent of the warrior graves from Sauromatia are thought to be female. 48 Given the

problems of methodology, in which osteological analyses are not always done in isolationfrom prior inferences about gender made on the basis of artefact assemblages, there aremany potential flaws, biases and problems in the study of gendet ftom funerary contexts.

GENDER IDENTITY AND CONTEXTUAL MEANINGS

We have to recognize that in trying to interpret a gendered past all the assumptionsand methodologies built into earlier accounts have to be re-examined, nothing can betaken for granted.49

Apart from the difficulties of osteological determinations of biological sex and the problemsof identifying examples of culturally constructed sex, the interptetation of gendets is fraughtwith difficulty. Ethnographic case studies from communities such as the Hua of the easternHighlands of Papua New Guinea illustrate the ambiguities, variability, permeability, andchangeability of gender construction and negotiation.50 The Hua classify individuals not onlyby genital characteristics but also by their amount of nu - a female and male vital essence(which has a liquid form, and resides in bodily substances such as blood, urine, fat, mucus orsemen). Nu content defines such categories as figapa ('uninitiated' or 'like women') andkakora ('young male initiates' free of all female nul. Children and women of child-bearingage are full of female nu, while young men who have abstained from heterosexual sex andwomen's food have male nu but no female nu. Post-menopausal women who have producedseveral children (and thus used up their female nu, becoming drier and harder) becomemasculinized whereas old men (who have gained female nu from years of sexual intercourse,casual contact with women and eating food prepared by women) are reclassified withadvancing years as figapa. Thus, on a sliding scale of degrees of maleness and femaleness,genitally female persons may be categorized as masculinized and vice versa.

The example of the Hua not only undermines the notion of man and woman asunchanging essences, but it also raises the possibility that gender transformations needhave no material correlates and that clothing may signify a gendet role at odds with thatadopted by the wearer. We can find similar examples from our own era: in the nineteenthcentury trousers were considered to have a clear gender-associated value so that women's'transvestite' adoption of them was seen as ambiguous and challenging.51 Today in thewestern world, trousers no longer affect the gender identity of the wearer though - pacethe Scots - skirts still do.

Within the Marxist framework of Engels' interpretation of women's changing status, it isthrough their place within the social relations of production that their gender status isdefined. Similar ideas can be found in Ehrenberg's discussion of the impact of the LaterNeolithic 'secondary products revolution', when power was supposedly wrested from thehoe-and-spade women cultivators by the ploughmen, and in John Barrett's model ofchanging gender roles in the British Late Bronze Age.52 Barrett argues that changes indomestic space, in food preparation and serving, and in inheritance systems (marked by theend of cremation burial) led to new emphases on the control of agricultural and probablyhuman fertility. Metals and the new iron technology played an increasingly subsidiary roleso that new gender roles and age sets were established outside and independent of thesphere of metal production and exchange. However, Henrietta Moore suggests that thecultural valuations given to women and men in society arise from something more than justtheir respective positions in the relations of production.53 In other words, what men andwomen respectively do has less significance than the meanings that those activities acquire.

102 THE ARCHAEOLOGY OF OEATH AND BURIAL GENDER AND KINSHIP 103

This brings us into a problematic area since we need to assess not only the culturalvaluations of those activities in the past but also to struggle to identify our own valueschemes. Linda Hurcombe points out the need for such clarity in explaining to her studentsher assumption that men were more likely hunters and women gatherers: 'To say that suchideas are sexist is to miss the point of sexual dimorphism as an evolutionary strategy and tobe biased by our own cultural experience of the status of activities. The female studentswanted women to be seen as hunters because this was the task they valued more.'54

Viking women

Something of this problem comes across in archaeologists' searches for women in high­status burials such as those at Vix and Wetwang discussed above. Anne Stalsberg's analysisof Viking tradeswomen in northern Europe provides an interesting comparison.5S Previousandrocentric interpretations of weighing scales in female graves took them to representfarewell gifts from husbands, tokens of high rank or evidence that the woman died whileher husband was away and she was temporarily in charge - effectively anything other thanthe possibility that the woman herself was a trader.56 In any case, there are many medievalScandinavian references to women active in trade at a slightly later date and, of course,women are often the principal traders in many societies in West Africa and elsewhere.Stalsberg suggests that this problem of recognition of women's trading roles is due to theshackles of tradition within archaeological scholarship.

In the earliest explicit study of gender roles and funerary remains (and probably theearliest publication on the archaeology of gender), Dommasnes examined the differentstatuses of women and men in Viking Norway.57 As well as enjoying power and authorityin the growing economic arena of foreign trade, women also held considerable power inthe household. From the tenth century onwards, 'rich' and elaborately constructedwomen's graves appeared, during a period of social unrest when men were away from thefarms and women had to take full responsibility at home.58 Yet women appear generally tohave had influence rather than direct power except within the household.

The large grave mounds of western Norway contain multiple burials but women are,with one exception, never found in primary positions wlthin them. Secondary burials (thatis, the person is buried as the second or subsequent one in the mound) are also twice ascommon for women as they are for men. If the burial mounds were established for farmfounders then these features are consonant with a gender-based hierarchical society. 59

Dommasnes also points out that, although in their graves women were provided with awide range of tools (with the exception of smithing equipment), the only items which werereserved for women were textile implements. This might suggest that Viking women notonly specialized in cloth production but maintained an independent trade-derived incomefrom its sale. Finally, she suggests that women's status was more dependent onassociations with a fertility cult than on their economic position within the relations ofproduction and that, as the position of women declined during the Iron Age, so thefertility cult lost its dominant position and was relegated from public life.

LITTLE PERlSHERS: THE ARCHAEOLOGY OF CHILDREN

In rendering the invisible visible, there has been a recent growth of interest in thearchaeological study of children and their gender development.60 As Sofaer Derevenski andothers point out, 'childhood' is a concept which is highly culture-specific, implying

sentimental western notions of learning, play and indulgence, in contrast to those manysocieties where the young work in adult projects from an early age. One of the areas ofinterest is the fact that children were almost certainly major contributors to all past societiesyet their remains are so often under-represented in cemeteries and other funerary contexts.61

This is all the more extraordinary when considered against Andrew Chamberlain's estimatethat most prehistoric populations had childhood mortality rates of at least 50 per cent andRega's estimate of 15-30 per cent mortality for children under one year of age.62 It certainlyfocuses attention on the unusualness of those children who are buried in the same way asadults. Rega's analysis of the Early Bronze Age cemetery at Mokrin (in the formerYugoslavia), for example, highlights the significantly greater mortality of girls to boys but,instead of leaping to the conclusion that young girls were neglected in favour of boys, sheargues the opposite. Given that young children and infants were normally buried under thehouses and settlements, the inclusion of these children in the cemetery distinguishes themfrom other children. The small proportion of children's graves at Mokrin (far fewer thanwould be expected from the likely mortality rate) suggests that the particular children buriedhere were fairly special and it therefore indicates that girls were more important than boys.

Ellen Pader's analysis of age-sex relations in pagan Anglo-Saxon cemeteries in easternEngland highlights the ways that age-based relations linked children and women: bothshared significantly more attributes with each other than either shared with men - malechildren were rarely found with male sex-linked artefacts whereas many children, bothmale and female, were treated as female in terms of their associated grave goods.63 It wasonly at sub-adulthood that the relationship between the sexes became apparent in theconsistent associations with sex-correlated artefacts.64 At Early Bronze Age Mokrin, incontrast, the primary axis of differentiation was by sex and even the smallest childrenwere buried according to a strongly demarcated categorization of gender.6s Malcolm Lilliecomes to rather different conclusions for cemetery populations from the UkrainianMesolithic and Neolithic where specific gender determinations had not been achieved bythe children prior to death even though the wide variety of artefacts placed with themsuggests that they were recognized as significant social actors.66

Perhaps one of the most salient points about child burials, noted by Lucy, is that theybring home to us the gap between those being buried and those doing the burying; asfunerary archaeologists we only ever see children as manipulated entities within an adultworld - they are buried by adults.'? Thus we never experience the world of children, onlythe experiences of adults coming to terms with and attempting to ascribe meaning to theirforeshortened lives and premature deaths.

Some critiques of post-processual archaeologists' concerns with power, domination andresistance stress the omission of phenomenological aspects of individuals' lived experiences.Phenomenology, in the tradition of the philosopher Husser!, is the attempt to understandexperience 'from within the flow of life, rather than from the outside, looking at the thingswe take for granted and bringing to attention things that have been forgotten or made to

seem trivial by the theoretical tradition'." Recent approaches have stressed how 'bodies canbe addressed by social and historical processes without ceasing to be individual andmaterial'." Lynn Meskell attempts a phenomenological approach in reconstructing theexperience of burying young children within the Eastern Necropolis of the New Kingdom(1570-1070 Be) site of Deir el Medina in Egypt'?' In one of her examples, a severelydisabled boy suffered from scoliosis of the pelvis, had an abnormal left hip, one leg shorterthan the other and his legs were swollen. The body was wrapped and placed in a wickerbasket which was too short: his feet protruded through a hole in the side. Meskell points

104 THE ARCHAEOLOGY OF OEATH AND BURIAL GENDER AND KINSHIP 105

out that the burial was neither expensive nor ostentatious bur did illustrate care andconcern in the placing of the necessary food offerings within the grave pit before the vaultwas covered with large stones. She concludes that the boy's untimely death warranted careand personal responses since he was already considered an embodied person.

Her observations on the personal and immediate nature of the mourners' grief and thepathos of the troubling sighr of the sticking-out feet, however, exemplify the problems ofperception and interpretation of an alien culture. What is presented here is an interpretiveview of what we, in western society, ought to feel as a result of our attitudes towards thedisabled. The evidence could be interpreted in other ways; the disabled boy may have beenneither loved nor mourned - his death may even have been welcomed. The carefullyarranged offerings and the large roof stones may have been primarily to ensure that he didnot come back to trouble the living.

As others studying the anthropology of emotion and the archaeology of compassionhave noted,?! the evidence is often ambiguous. This is not simply a problem of lack ofevidence but of reading and understanding the complexity of contrasting emotions and theinterplay of inner feelings manipulated and orchestrated through the expectations of theritual routine. While Meskell is entirely right to pursue archaeologies of emotive andembodied experience, we tread a thin line between enquiry and empathy. Empatheticapproaches are, unfortunately, more than likely to lead us back into imposing our ownunacknowledged preconceptions.

No survey of child burial studies would be complete without considering child sacrifice,infantiCide and death through exposure or neglect. There are now a number of studies ofRoman Britain, ancient Carthage, the Inka empire, and Late Roman-Early ByzantineIsrael where very different kinds of infanticide have been claimed from archaeologicalevidence.'2 Unsurprisingly, the quality of the evidence and the strength of the claims variesbetween cases. The different contexts range from the public and highly ritualized sacrificesof the eapaeoeha ritual among the Inka to the apparently clandestine disposal of unwantedbabies at Askalon in Israel."

WOMEN, MEN AND CHILDREN IN DANISH PREHISTORY

In his provocative book The Prehistory .of Sex, Tim Taylor suggests that the emergence ofstrictly gendered clothing aided the entrenchment of social inequaliry74 By naturalizingdifference in sex through gendered dress, the existence of social hierarchy can also bemade to appear natural and unquestioned through sleight-of-hand. The Danish sequencefrom the Mesolithic to the Iron Age provides an opportunity to examine this intriguingclaim as well as to follow the changing gender relations within one part of the world overa period of five thousand years between c. 5600 BC and AD 400. It is also an opportunityto examine critically Michael Mann's sweeping statement that gender relations remainedbroadly constant, in the form of patriarchy, from earliest recorded history (c. 3000 BC inthe Near East) to the eighteenth century.75

The origins of gender's cultural construction have been estimated as dating back toarchaic Homo sapiens, abour 200,000 years ago, but are not archaeologically visible untilthe last 100,000 years with the first appearance of burials and artistic representations." Asuspiciously large proportion of the Eurasian Middle and Upper Palaeolithic burials aresexed as male, leading Whelan to question earlier conclusions such as Harrold's inferencethat Upper Palaeolithic 'males and females do not differ significantly in regard to thedistribution of grave goods or features of any category' - women may just not have been

buried by and largen Thus our search for gender categorization is limited, for themoment, to later periods and the quality of research into the Danish later Mesolithic andits cemeteries (c. 5600-4200 BC) provides an ideal starting point for study.

There is a large body of literature on gender in Danish prehistory.78 Much of thematerial on the Bronze Age has been extensively reviewed by S0rensen in a detailedcritique of the different approaches and the varying nature of their androcentric, empiricaland rheoretical difficulties so I shall present here a brief summary of trends interpretedlargely from the funerary evidence.79

The Mesolithic period

The evidence from the later Mesolithic burial groups such as Vedbaek, Nederst, StwbyEgede and Tybrind Vig indicates gender-related burial associations between males andfemales and specific types of grave goods. sO Although numbers are small and gravesoccasionally have no grave goods, some burials exhibit a marked gender distinctiondefined by the costume of the deceased or attendant artefacts placed with the body.S! Forexample, women's burials are characterized by collections of pendant tooth beads whichthe excavators interpreted as originally sewn into clothing. In only one instance atVedbaek is a single pendant tooth bead found with an adult male, behind the back of hishead. Male burials often include a flint blade placed in the pelvic area and stone and antleraxeheads around the upper part of the body. Some female burials have attendant artefactsin the form of a bone awl and flint transverse arrowhead above the head; one has a bonedagger over the pelvis. The association of axes with males and quantities of tooth beadswith femaLes is found in the other Mesolithic cemeteries of Denmark and the Skatteholmcemeteries of southern Sweden.

Yet there may have been a third gender. Within a triple grave at Vedbaek were thecorpse of a year-old child with a 25 to 30-year-old adult of indeterminate sex to its right,probably killed by the bone arrowhead lodged between its neck vertebrae (Figure 5.2). Tothe infant's left lay a 35 to 40-year-old, also of indeterminate sex, with a small flint bladebelow the lower jaw and a collection of animal and human tooth pendants, unperforatedred deer teeth, a pine marten's lower iaw and bones of roe deer. This latter individualstands out as something of an anomaly and mighr indicate rhe presence of a third gender, aritual specialist or both. To a lesser extent, some of the grave goods may have been relatedto age differences; the tooth bead assemblages found with adults are Larger and have morespecies variety (not iust red deer but also wild pig, aurochs, ox and elk).

The Neolithic and Early Bronze Age

In the earthen barrows and stone dolmens of the subsequent period, the Earlier Neolithic(c. 4200-2800 BC), skeletons are rarely preserved.S2 Although grave goods often survive,little can be said about gender relations, except that a single burial from Dragsholmcontained the skeleton of a twenty-year-old man with a pOt, amber beads and a stonebattleaxe. The same is broadly rrue of the stone passage-grave burials of the MiddleNeolithic although bones have survived in a few cases. Charlotte Damm argues thatalthough men, women and children are present in these deposirs, their often disarticulatedremains are not differentiated.83 In other words, individual identities of the dead in theMiddle Neolithic were wholly dissolved into an anonymous collectivity of ancestors(Figure 5.3). Of course, we have no knowledge of how gender was constructed within the

106 THE ARCHAEOLOGY OF DEATH AND BURIAL GENDER AND KINSHIP 107

The Middle Bronze Age

The preserved coffin burials of the Middle Bronze Age (1800-1300 Be) have attractedconsiderable interest with respect to gender since not only are certain burials exceptionallywell preserved, but there appears to have been a very strong categorization of male andfemale gendered objects and costume attributes." Men's and women's dress was differentin many ways, notably in headpieces and coiffure and the wearing of cloth. sS Randsborg'sanalysis of metal wealth in Bronze Age graves (Figure 4.3) demonstrates that both womenand men were buried with often substantial quantities of bronze and gold (the generallyheavier average weight of metal in male graves is due to the presence of swords, which arenot found in female burials).S9 The increasing quantity of female-gendered bronze dressitems deposited in votive hoards towards the later part of the Bronze Age is in inverseproportion to male-gendered artefacts and has caused much discussion as to i~s meaning.90

2M\5.3 A section and plan of <Klokkeh0j', a megalithic tomb of the Funnel Beaker period in Denmark. It initiallycontained (c. 3300 BC) the headless skeleton ofa 20-3S-year-old man (middle) with a stone 'pillow' (5), a child's

skull (6), a bone strip (1) and three pots (2-4). Around 2800 BC the disarticulated bones (bottom) of nine

children and thirteen adults were added.

sequences of funerary rites prior to final andcollective deposition, let alone in the otheraspects of everyday life, but Damm suggeststhat the ideological representationsarticulated by communal megalithic burialsare those of differences and inequalitiesbetween territorial or kin-based groups.'4An interesting feature of megalithic tombson Zeeland is that many are paired. Onechamber is always smaller and less wellconstructed than the other. Unfortunatelynone has preserved any human bones so wedo not know if this pairing served todistinguish people by age, gender or kinship.

The mound burials of the Late NeolithicSingle Grave Culture (2800-2400 Be) and EarlyBronze Age Dagger period (2400-1800 Be)present us with a rigid classification of genderconformity, just as we see in eastern Europe atthat time (Figure 5.4).85 Some women's burialsare accompanied by small polished stone axes,flint blades, bone tools, ceramic pots andornaments of copper and amber. The flexedskeletons lie facing south but the majority ofmales lie on their right sides with heads to thewest and females lie on their left sides with theirheads to the east. The pattern applies to childrenas well as adults, though there is evidence of alack of fit between sex and gender in that 15 percent of males are buried in the female genderedposition.56 Of course, such pronouncements areproblematic because most of the burials surviveas body stains with bones rarely surviving andthus cannot be osteologically sexed.

5.2 The child buried between the two adults ofindeterminate sex in the Mesolithic cemeter-y at

Vedbaek. The combination of tooth beads and a flintblade with the right-hand body does not fall withinthe gender categories exhibited by other burials and

hints at the possibility that this individual was of a

third gender.

108 THE ARCHAEOLOGY OF DEATH AND BURIAL GENDER AND KINSHIP 109

/b, •

-'- --.........,............

............... ,\\\i'\

iiiI

//

.///-..... -'- --- - -'-'-'-- --- -' ---._._'-'--'-'-.__._-

1M

5.4 A burial of two young women of the Single Grave Culture from Bedinge, southern Sweden, associated withan unusually lavish assemblage of two flint carpentry axes, four pots, four flint blades, a copper ear-ornament,over a hundred amber beads, three bone bodkins, a pot-decorating tool and sheep bones.

The Iron Age

The Late Bronze Age and Early Iron Age funerary traclition is almost universally cremarion(other than the bog bodies; see Chapter 3). For the Early Iron Age the gender associations ofmetal dress fittings are not cleal; owing to a sad lack of osteological analysis on large cemeteriessuch as Arupgard. Randsborg speculates that women's labour would have been important inthis agricultural economy marked by the appearance of field systems but that their powerwould have been curtailed by the establishment of closed households, in the form oflonghouses with byres. There is no way of knowing, and the disappearance of women's bronzejewellery as offerings in bogs may signify merely the collapse of an elite or of an exchangesystem rather than a significant change in gender relations. The higWy egalitarian picturegained from Early Iron Age burials and longhouses is contraclicted only by the evidence thatsome of the bog bodies performed no manual work. There seems to have been no genderdistinction in selecting these inclividuals for the dubious honour of dying in a peat bog.

From the late first century BC, shortly after the appearance of "lite farmsteads andcremations, most cremation cemeteries were divided into male and female zones or wereentirely single sex. Although burials from the Roman Iron Age in eastern Denmark oftendo not have skeletal material preserved, enough osteological studies from other parts ofScandinavia and northern Germany indicate that gender associations of the materialculture largely conformed to sex differences but that there were substantial areas ofoverlap." For example, over 10 per cent of burials with spindle-whorls were males whilealmost 15 per cent of weapon burials were females.'"

This broad picture obscures chronological changes within the Roman Iron Age; by thesecond century AD the relatively autonomous categories of male and female equipment werebecoming intermingled within individual graves. This erosion of gender representations wasalso occurring in the increasingly mixed cemeteries. Around AD 200 there was a majorreordering of funerary practices, in line with changes in household organization, military

structure and industry." Cemeteries were re-established as gender-segregated spaces andwomen's elaborate costumes and finery marked them apart from male dress sryles. Yet by AD400 the spatial segregation in cemeteries was once again breaking down.

Taylor's claim that gender distinctions in clothing became apparent in Europe only withthe emergence of other forms of status differentiation in the Late Neolithic and EarlyBronze Age cannot be sustained.'4 The evidence shows that in the Mesolithic periodgender differentiation was probably greater than the marking of age differences. Mann'sproposition of stable gender relations is similarly without foundation. The Danishprehistoric sequence provides us with a remarkable view of the dynamic and chan~in~

relationship between men and women and the ways that genders were construc~ed~dIssolved and reformulated over time. These changes coincided with significant economic,cultural and political transformations, indicating that the politics of gender werelilseparable from other social processes and practices.

DRESS, GENDER AND KINSHIP

As discussed in Chapter 4, Sue Shennan's analysis of the Early Bronze Age cemetery atBranL (c. 2400-1700 BC) concludes with two opposing hypotheses about socialorganization. 95 She suggests that if the high status of certain women was ascribed(inherited), they were more likely to survive infancy than boys, raising the possibility thatdescent may have been calculated through the female line and that female children weretherefore vital to group continuity. Given the small size of the community represented bythe cemetery (about forty people at anyone time), men may have been brought in asmarriage partners from outside. Alternatively, she suggests that a different picture resultsfrom interpreting women's wealth as achieved (essentially as awarded at marriage). The few'rich' men, in contrast to the many 'rich' women, are explained as polygamous husbandsand the women's wealth derives from bridewealth payments or the use of wives as vehiclesfor displaying their husbands' wealth. In this scenario, descent might be patrilineal andresidence uxorilocal. Shennan was unable to decide between these two possibilities.

Subsequently, Stephen Shennan and John O'Shea have interpreted the larger quantitiesof metal goods in women's graves (as opposed to men's) from Early Bronze Age cemeteriesin the region as indicative of male wealth and prestige, thereby supporting the secondalternative." Their virocentric position is attacked by Rega who states, 'While the metalwealth in female graves may indeed represent a male contribution, the mortuary dataalone do not allow determination of "ownership", whether symbolic or actual.'97

Moktin: status and gender

Studies of gender and kinship are closely tied to studies of status and all need to be understoodtogether: O'Shea develops the notion of aSSoClattVe status, in addition to achieved and ascriptivestatus, where an individual holds or obtains a social position by virtue of a relationship (ofkinship, marriage or adoption) to another inclividual or group.98 He gives as an example theburial of certain women in head-dresses in the Early Bronze Age cemetery at Mokrin;" headornaments are regularly found with adult and mature women but are rare amon~ old womenleading O'Shea to suggest that the wearing of head-dresses was relinquished in l;er life. Fron:this he suggests that they may have had an associative character and that this might reflectconsanguineal or affinal ties to the holders of male offices.!oo Thus he perceives that the womenwearing head ornamenrs owed that right to their relationships with particular men.

110 THE ARCHAEOLOGY OF DEATH AND BURIAL GENDER AND KINSHIP 111

At the same time O'Shea argues that the general poverty of old people's dressaccoutrements and grave goods is an indicator that this was a society in which status wasaccumulated by giving away wealth and possessions; the lack of such items among theelderly may have related inversely to their status.!Ol The situation may also have beencomplicated by the different responses of mourners to untimely deaths and to those dyingin old age. At the end of a long life, personal possessions might be recycled, handed downor disposed of in other ways but, for the life cut short, the use of these items other than asgrave goods might be deemed inappropriate. In either case, we need not accept O'Shea'sinterpretation of women's head ornaments as associative status markers.

Regardless of these interpretive problems, the costumes of the deceased and the mannerof their burial reveal a very strong dichotomy in the presentation of male and femaleregardless of age or other social factors. Rega found that 94 per cent of the adults atMokrin were buried in a form accordant with their sex, and the same was probable for thechildren. While everyone faced east, females lay on their right sides with the head to thesouth or south-east and males lay on their left sides with the head to the north or north­west (Figure 5.5). The anomalous 6 per cent are considered by Rega to be due toinaccuracies in standard sexing techniques, as those results fall within the expected rangeof sex determination error.!02 Although costume and grave goods relate to some extent toage and other factors, ten of the nineteen non-ceramic artefact types relate exclusively orpredominantly to either males (three types) or females (seven types).!OJ Males and femaleswere presented as engaged in separate but equal lives with both difference andcomplementarity highlighted.

The most interesting burial at Mokrin is that of an elderly male (large and robust)buried with dress ornaments normally associated with adult females: a pair of pins, aneck-ring, four bracelets and two gold hair-rings.'o, This individual's ambiguous status isfurther highlighted in two ways. Firstly, there is a striking contrast between the poverty ofmost old people's burials and the fact that this was one of the most splendidly equippedgraves of the whole region. Its ambiguousness is also expressed by its position within thecentral overlap between the distinct clusters which sub-divide the Mokrin cemetery intofour groups. It thus embodies the fundamental significance of gender in structuring EarlyBronze Age social relationships. In such a society, ambiguity of this kind might beconsidered powerful and/or dangerous. Rather than viewing this individual as a one-offcuriosity, we might consider that his/hers was the most important burial within the wholecemetery.

KINSHIP AND THE NEW ARCHAEOLOGY

One of the earliest applications of a New Archaeological approach to archaeological datawas Binford's analysis of funerary deposits in Galley Pond Mound, a Late Woodland­Mississippian mortuary site in Illinois.'05 His approach required acceptance of a nestedseries of assumptions or inferences. As a piece of creative thinking, it illustrates clearlyBinford's early optimism that all aspects of past societies were accessible through thearchaeological record and could be addressed by framing the appropriate hypotheses.However, Binford's logic is, in this case, tortuous and hard to follow and his inferences arespeculative in the extreme.

On the floor of the mortuary structure at Galley Pond (Figure 5.6), there were bonebundles (groups of disarticulated bones, especially skulls and long bones). Binford dividedthe bone groups into different categories:

I5.5 At the Early Bronze Age cemetery of Makrin in the former Yugoslavia, the arrangement of male and female

burials shows that gender was a central structuring element of funerary representation. This body of an adult

female faces east with her head to the south. She wears a gold-foil crown, gold hair-rings, and bracelets and aneck-ring of copper.

A. Bundles consisting of a skull and long bones and rearticulated extended burials.These were complete individuals who shared the same undifferentiated treatmentof skull and body proper.

B. Burials in which the skull faces one way and the rearticulated legs another.'o6 Thesewere complete individuals whose skull and leg bones were differentiated at burial.

C. Incomplete individuals - either just skulls or just long bone bundles.

He found that eight out of the ten Category A bone groups were oriented north-south, aswere all the individnal skulls (part of Category C). In contrast, all the long bone bundles(part of Category C) were aligned east-west. His first inference was that the east-west andnorth-south orientations of the bundles, skulls and bones related to a dual social division.Binford's second inference was that the north-south orientation was symbolic of the localcommunity and therefore the east-west orientation must be symbolism used by anothercommunity or by a minor segment of this one. He supported this inference by pointing outthat the mortuary structure, domestic structures and all 'high-status' burials (apparentlyburials from nearby sites which were given elaborate treatment) were aligned north-south.

From the fact that the skulls had not been nibbled but the long bones had, Binfordsurmised that the two categories of bones (skulls and long bones) had different pre­interment histories and thus may have had two separate custodians or groups ofcustodians. Inference 3 was that orientation was determined by the moiety affiliation ofeach custodian, the person responsible for bringing already disarticulated bones for burialin the mound. Inference 4 was that interment of bodies in the mortuary structure was

112 THE ARCHAEOLOGY OF DEATH AND OURIAL GENDER AND KINSHIP 113

a) the only right-sided burialb) the only burial facing north and extended on its backc) greater variation than males in hand positioningd) a wider angle of leg flex between 45 and 180 degrees (average 80 degrees)

whereas male flexion was less than 45 degrees.

1. A complete skeleton of skull and bones (both custodians were local residents).2. Just a skull or just bones (the other portion retained by a non-resident custodian).3. Skull and bones from different bodies but on the same orientation (the resident

custodians were both members of the same moiety).4. The bones one way and the skull the other (the resident custodians were members

of opposing moieties).

done only by the local community. If a non-local custodian retained either the skull or thelong bones, that element would not be placed in the mortuary structure. Fourcombinations of bone groups could be recognized:

Since all the remains identified as female were represented by complete skeletons (bone groupcombination 1), Binford's Inference 5 was that residence and descent were matrilocal andmatrilinear. Through this ingenious but tortuous logic Binford tried to reconstruct the majoraspects of this Late Woodland community's kinship; from a series of inferences he claimed toidentify this community as a dual organization of matrilineal and matrilocal moieties.

Binford's breathtaking leaps of faith at each stage in his study are no greater than thosein certain post-processual case studies, and perhaps such soaring optimism is a feature ofmany of the earliest case studies within the grip of a new archaeological paradigm.Binford's study may also have relied on unacknowledged similarities with ethnographicmaterial. As Allen and Richardson point out, it is difficult to conceive of such an elaboratereconstruction without recourse to ethnographic data, in this case probably historic south­east North American burial practices and kinship patterns,!07 If we were to begin again onthe Galley Pond Mound material, we might pursue the appropriateness and applicabilityof those ethnographic cases more explicitly. It is also most probable that the north-southsymbolism is a feature not just of the Galley Pond community but of many Late Woodlandcommunities in that region, and we would need to reconsider the meanings of the differentorientations. It is doubtful that we could support Binford's conclusions today.

In 1971 Saxe published his analysis of Mesolithic burials at Wadi HaIfa in Egypt. Thelack of major differences in grave good provision and grave construction led Saxe to inferthat this had been an egalitarian society. His best evidence for potential social differenceswas in the degree of flex of the corpses' legs, in other words the variations in the anglebetween the vertebral column and the femur. The bodies were laid mostly on their leftsides, with their hands in the area of the chest or face and with their knees bent into atight flex of sometimes less than 45 degrees to the vertebral column. Saxe noted thatfemales were treated more variably than males and included:

•

.,!I

I

I;' I'......-..1

······,4~

"

....-

r'-'-'-'-'-',I

.....•..

(~.~, .....

,;

'.",'

"

•

8M

L._._._._._·

5.6 Galley Pond Mound contained a rectangular mortuary structure with an outer wall trench, three rows of postholes and a sunken central area. The bone bundles were scattered around its interior.

Saxe considered that these variations were due to differences in pre-interment and intermentrituals. He hypothesized that post-marital residence was patrilocal, that the women werewives who were incomers. Male corpses were treated uniformly, presumably by the localpopulation adhering to the same traditional methods, whereas women were interred by theirown non-local kin with their separate and various traditions of laying out and burial.!08

THE ARCHAEOLOGY OF DEATH AND BURIAL GENDER AND KINSHIP 115

Again, this case makes a number of dubious assumptions but not to the extent ofBinford's study. We may speculate juSt as plausibly [hat the differences resulted from menburying the males and women burying [he females, each with their own cultural traditionsof interment. Less speculatively, these differences may be most simply explained as themeans by which the corpses' genders were categorized and communicated. Other NewArchaeological approaches to reconstructing prehistoric kinship patterns can be found instudies of Natufian society,109 and of Californian Native American societies. l1O

Problems with archaeological reconstructions of kinship

Social anthropologists Allen and Richardson have criticized archaeologists' outmodedconceptions of kinship classifications and realities, pointing out that residence and descentwithin particular societies are often not clearly patterned exclusively or evenpredominantly into single forms such as virilocal or matrilineal. 111 There is a distinction to

be made between kinship rules and reality and there need not be any correlation betweentypes of residence patterning and types of descent (i.e. matrilocal does not presupposematrilineal). Their criticisms are, however, largely aimed at non-funerary studies whichwere attempting to extrapolate kinship from ceramic design variation.

Pader reiterates the criticisms of Allen and Richardson and specifically criticizes the'purity argument' that proposes that insiders can be distinguished from outsiders byfunerary treatment.1l2 Her other main criticism, relevant for Shennan's notions aboutcommunity size and the need for exogamy (out-marriage), is that the populations within asingle cemetery may well derive from a variety of different settlements: the community ofthe dead may well be very different from the communities of the living.

As we have seen, attempts to use mortuary remains to infer kinship and other relatedaspects of social organization studied by social and cultural anthropologists have been largelyunsuccessful; descent, residence and notions such as matriarchy not only elude our search buthave been, in any case, demonstrated to be far more complex and problematic entities thanarchaeologists have been prepared to accept. However, thete are two areas where at least acrude understanding of such elements of social organization may be possible. The first is inthose instances where stratigraphic sequences of burials may allow us to draw certainconclusions about gender and kinship. The second is through the use of human biologicaldata, in conjunction with funerary analyses of archaeological context and association.

STRATIGRAPHIC SEQUENCES AND KINSHIP

Stratigraphy is the study of sequences of deposits, notmally vertically, where one layer lieson top of the other. In undisturbed sequences the lower layer always pre-dates the higher.In such Situations archaeologists can establish not only relative dating but also sequencesof deposition. By studying such sequences of burials it is possible to make inferences aboutsocial precedence and succession from the vertical ordering of the dead.

The large Viking Age (tenth to twelfth centuries AD) grave mounds of western Norwaycontain large numbers of bodies. Dommasnes has shown that although mare women's gravesthan men's are found in the larger mounds, they are all, with one exception, secondary burials.For all sizes of mound, secondary burials are twice as common among women as among men.She concludes that the relative lack of women's primary burials may be due to their socialposition within families. The mounds, often built close to the farmhouses, are interpreted asfamily tombs constructed over the grave of the 'founding father' of each farm. ll3

A similar set of stratigraphic patterns is found among the Early Bronze Age roundbatrows of Wessex in southern England during the Beaker period. Inferences about thesocial structure can be made from grave goods and from the relative placing of humanremains in burial mounds. Some of the grave goods can be divided on gender lines,suggesting a certain division of labour between men and women, symbolized in death.While male graves contain arrowheads, daggers, wristguards, belt rings, amber buttons,stone axes and fire-making tools, female graves are associated with shale and jet beadsand the majority of awls and antler picks; cettain items (flint blades, ear-rings and pebblehammers) are sbared equally.!!4 Whereas daggers, ornaments and small tools wereregularly placed in graves, other items were generally deposited elsewhere. Metal axes,spearheads and halberds are found invariably in boggy contexts as hoards or single finds.These are often in rocky and impressive locations which were evidentially special places.It seems clear that such deposits were votive offerings, occasionally broken and, forcertain unknown reasons, inappropriate accompaniments to buried individuals. us

Further insights into Bronze Age gender and age distinctions can be gained from therelationship between primary and secondary burials within barrows (Figure 5.7). Whereadult males are buried first, later burials may be of other adults or children, yet whete adultfemales are buried first they are rarely followed by adult males. 116 This pattern can be foundin the excavated barrows within linear barrow cemeteries such as Shrewton in Wessex ormote dispersed groups such as Irthlingborough-West Cotton in eastern England. ll7

These patterns, in which adult women as primary burials are followed only by children,other women or adolescents while adult male primaries are followed by any combination,closely match the interment sequences in Tandroy tombs in southern Madagascar. llS

Within this strongly patrilineal society, women's corpses are often returned to their father'scommunity for burial. When an adult married woman dies before her husband she will beburied in a single grave which is then covered by a walled cairn. If she dies after he does,she will be buried eithet as a secondary within his tomb or as a secondary within herfather's tomb. A man will normally have his own tomb if he has married and established afamily, otherwise his body will be interred as a secondary burial within his fathet's romb.Other secondary burials inserted into the tomb coveting the primary grave of eithet a manor a woman are their dependants - young women and men, children and adolescents(although children below about six are buried separately in children's cemeteties).

The striking similarity of these two sequential orderings of burial is very suggestiveof a shated emphasis on the male line, which we might interptet as symbolizingpatrilineal descent. In the Tandroy case, the tombs may be many miles from the villagesof their occupants and those buried in the same tomb may be either blood relatives orkin by marriage, and may have spent relatively little time in the same community. Nonethe less the tombs fix patrilineal ancestry into the landscape. In both cases therelationship between primary and secondary burials sets forth an idealizedrepresentation of how kinship and family should be organized, thereby fixing in ritualthe ways that evetyday life should be led. Thus we can infer that the funerary rites ofthe Wessex Bronze Age embodied and represented a patrilineal notion of succession.

Mass graves and stratigraphy

An intriguing postsctipt on Eatly Bronze Age burial and kinship in north-west Europe canbe constructed from a coastal site at Wassenaar in the Netherlands. 119 Here a mass gravedating to c. 1700 BC contained twelve individuals (Figure 5.8). Unlike the jumbled chaos

116 THE ARCHAEOLOGY OF DEATH AND BURIAL GENOEH AND KINSHIP 117

5.7 A plan of multiple interments within an Early Bronze Age round barrow at Shrewton, Wiltshire, showing

cremation burials (dots), two off-centre inhumations, and a central burial pit. A secondary burial (Interment 2)

has been dug into the top of the primary central burial (Interment 1).

O.5m

5,f30-40

I

3,f

7

1$-16

III,IIIII

iMG/

1 "-

.'. -. ,( ,

\),

"',

/\9i~ 'j( '-/ \.' ," -' ;\ ',~--~ ......" ,

0 ,• • !,

'~ __ 'II,•.

...~,_ ..... ,.-_""'- 'v"

12M

of a mass grave of the Early Neolithic discovered at Talheim in Germany,120 the Wassenaarbodies were carefully laid out in two opposing rows, one to the east and one to the west,with their legs overlapping. The excavators were able to ascertain from the arrangement oflimbs that the corpses were laid out in sequence from the north end to the south (Figure5.9). The arrangement of the bodies exhibited a symmetrical structure in which the adultmales were placed centrally while the two women (one certain and one probable) and thevery young children were laid at the edges. Additionally, the two women were laid on theirfronts and children under twelve were laid on their sides; the rest were laid on their backs.This grotesque tableau presents a remarkable insight into Bronze Age classifications ofgender and possibly kinship, presenting a horizontal framing of relationships notdissimilar to the vertical differentiation found in the British round barrows.

INTEGRATING FUNERARY AND BIOLOGICAL APPROACHES

There are three major biological techniques available to archaeologists studying kinship.The first is the analysis of the form and shape of bones and teeth, either as measuredelements (metric traits) or as recorded minor anomalies (non-metric traits). There are morethan 400 non-metric variants (as opposed to metric or measured variance) described forthe human skeleton. They are not caused by disease but their development may be affectedby nutrition and factors operating during the growth period. The causes of most non­metric traits are still unknown but some have been found to be genetically inherited tovarying degrees, notably cranial features.'21 Non-metric traits have been used to studypotential genetic relationships within cemetery populations to test whether pre-identifiedsub-groups or spatial clusters might have corresponded to kin groupings.122 Certainepigenetic traits are sometimes associated with physical attributes through the mechanism

5.8 The Early Bronze Age mass grave at Wassenaar, the Netherlands. Skeleton 10, a young man, had a flint

arrowhead between his ribs, whilst the skeletons of three other men showed evidence of blows. The skull of

Child 4 had apparently been separated from the body at burial.

of pleiotropism (the phenomenon whereby the same genes code for different aspects of thephenotype). There is little certain evidence that many non-metric traits are necessarily theresult of genetic inheritance. Also problematic for the archaeological study of kinship isthe consensus that certain traits are widely dispersed within large populations of hundredsand thousands of people and are rarely limited to a particular descent line.

The identification of blood groups from ancient human remains - palaeoserology - ispossible from blood cells surviving in bone as well as in preserved soft tissue butdetermining an individual's blood group from surviving bone alone is difficult, given

118 THE ARCHAEOLOGY OF DEATH AND BURIAL GENDER AND KINSHIP 119

5.9 The skeletons at Wassenaar under excavation, seen from the north.

problems of non-specific absorbtion, contamination and possible technical error. 123 Onestudy of blood types has confirmed the likely kinship between Tutankhamen and themummy identified as Smenkhkare.l24

DNA (deoxyribonucleic acid) is a molecule which contains inherited genetic informationand is mostly located in the chromosomes. Humans have 23 pairs of chromosomes bearing50,000 genes. DNA is also located in the mitochondria, outside the nucleus - mitochondrialDNA is inherited solely from the mother whereas chromosomal DNA is inherited equallyfrom both parents. Mitochondrial DNA is more useful for studying ancient remains since itis a small and well-characterized molecule, it is more readily amplified in the PCR technique(polymerase chain reaction) than single-copy chromosomal DNA, and it is highly variableallowing easy identification of differences between individuals unless closely related.!25 Itnow appears that the survival of ancient DNA is not correlated with the age of the remainsbut is more dependent on factors of histological preservation. l26 Although ancient DNA isinvariably partial and fragmented, frozen and mummified bodies may produce the bestresults. The major problem still remains that of contamination by modern DNA andconvincing results must involve adequate controls, should be repeatable (ideally inindependent laboratories), and must be statistically verifiable phylogenetically or bycomparison with DNA-independent controls. 127

Matrilineal kinship in prehistoric Thailand

Evidence for matrilineal succession has been found in Btonze Age Thailand at KhokPhanom DiP' Burials were superimposed in tight clusters (Figure 5.10) within a largemound of ash and food debris which probably derived from funerary feasts and gradually

accumulated during the cemetery's use, providing a well-defined vertical sequence.Analysis of non-metric traits in skulls, teeth and limb bones showed that people buriedwithin a cluster were more likely to share certain abnormalities. Stress marks on the pelvesof women buried near infants were interpreted as showing that the women could havecarried at least that number of children (though there is now considerable doubt whetherparturirion can be detected from so-called 'parturition scars' on the pelvis). Pots inadjacent graves sometimes had similar decorative styles. Finally, what was probably aritual of tooth removal had been practised on a different set of teeth for people in oneclnster as opposed to another. If these clnsters were genetically related kin groups thentheir vertical sequences could provide genealogical histories over five hnndred years.

Until abont the twelfth generation (out of twenty) there was little difference in gravegoods disttibuted between men and women but this changed with the appearance ofelaborate female graves containing thonsands of beads, beautiful pots, red ochre, a purplesubstance possibly used as a nipple ornament, clay anvils and cylinders of potting clay.There is an evident association here between women and pottery: perhaps these womenwere celebrated in death as potters whose skills had brought them esteem and wealth.Charles Higham and Rachanie Bannanurag consider that the value of these individuals tothe community was so great that they would have remained in the village of their birthand thus would have participated in a matrilocal pattern of residence (though of coursethey could have moved on marriage and been returned after death for burial).

In addition, these richly endowed women were often accompanied by elaborate infantgraves though they were also succeeded by poorer descendants, presumably daughters andgranddaughters. The 'rich' infant burials might hint at inherited status but the plain gravesof immediate descendants indicate that high status was attained through achievement andwas not inherited; we might describe this as the female version of a 'Big Man' society. The'rich' infant burials were thus not reflections of these children's inherited personal standingin life but were perhaps expressions of their mothers' standing and esteem and of maternalgtief at their deaths.

Britons and Germans in pagan Anglo-Saxon England

One of the main questions for archaeologists working on the Anglo-Saxon period inEngland focuses on the relationship between the indigenous Romano-Britons and theGermanic incomers who arrived from the Continent during this period (c. AD 400-600).Did the two populations live side by side and if so, how did they integrate? Did the Britonslargely move west? Were many of them wiped out by disease? How many immigrantscame from Germany and Denmark? Recently Heintich Harke has produced evidence,gathered from forty-seven pagan Anglo-Saxon cemeteries, that men buried with weaponswere, on average, taller and stronger than those men buried withour. 129 The armed menwere taller and stronger, he supposed, because they were better fed and looked after. Bycorrelating the dates of these weapon burials with armed conflicts recorded in the Anglo­Saxon Chronicle, Harke concludes that the periods when these spears, swords and shieldswere placed in men's graves were largely times of peace; what was being signalled was notmen's status as warriors but the construction of their masculinity.13o

Looking at cemeteries such as Berinsfield in Oxfordshire, Harke further wonderedwhether these armed men might have been of Germanic descent while those withoutweapons were of British descent.!31 An analysis of the distribution of non-metric traits onthe skeletons from Berinsfield produced clusters of certain traits among the armed men

120 THE ARCHAEOLOGY OF OEATR AND BURIAL GENDER AND KINSHIP 121

5.10 A 'rich' woman's grave in the Bronze Age cemetery of Khok Phanom Di, Thailand. The graves werearranged spatially into dusters (middle) and kinship links within the clusters could be reconstructed as family

trees (bottom).

Non-mettic traits and kin groups in the Neolithic: La Chaussee-Tirancourt

Neolithic long barrows and chambered tombs occasionally survive with the human bonesundisturbed, though in most cases these survivals probably represent only the final phaseof deposition after one or more periods of clearing Out. At the remarkably well-preservedmegalithic tomb of La Chaussee-Tirancourt in the Somme region of northern France,excavations recovered over 360 burials, accumulated between its construction around2800 BC and its closure and alluvial inundation around 2100 BC.134 Burials have survivedmostly from the second and third of the three main phases of the tomb's use andalteration. During the second phase, earlier burials were cleared away and some sixtyburials were added, deposited in three spatial groupings which could be furtherdistinguished by associated non-metric traits. The western group within this east-westoriented tomb had high frequencies of hypotrochanteric fossae (a feature on the proximalend of the femur) and low frequencies of a third trochanter on the femur; this latter traitwas found in the other two bone clusters. The bone group on the south side of thechamber was characterized by a high incidence of incrustation on the humeral trochlea.

In the third phase of the tomb's use, these deposits were not cleared out but becamecovered by a soil layer on top of which another three hundred bodies were deposited inwhat had been a series of eight wooden cases arranged around the chamber. The presenceof non-metric traits was spatially restricred to four of the cases. There was a highincidence of incrustation on the humeral trochlea in rhe case on the south side ­immediately above the earlier bone deposits with similar traits. This and two other casesin the south-west of the tomb contained bones where the female arm bones lackedolecranon fossa perforations and, as in the deposits below, the femurs had a high incidenceof hypotrochanteric fossae. In contrast, many of the tibiae in the case in the north-west ofthe tomb had an additional articular facet.

Owing to these fortuitously high incidences of non-metric traits, Chris Scarre hassuggested thar we may identify distinct kin groups sharing a communal tomb, an idea thathas long been considered for such chambered tombs.'35 More extraordinarily, thelongevity of deposition is suggestive of lineages each interring in their particular part ofthe tomb over some seven hundred years. The spatial organization within the tomb wasthus a mapping of lineage relationships. The expansion of bone groups from three to eight

and others among the unarmed ones, with very little overlap.'32 Here was evidence tosuggest that the two groups were of different descent.

There are, however, some problems with the interpretations based on the osteologicalobservations. The average difference in height between armed and unarmed males is morethan 2cm, but this difference is small when compared to the variation in stature within each ofthe groups with and without weapons. Even where the weapon burials show less variation inheight than those without, as at Berinsfield, tall stature may just have been an importantcriterion for bearing arms. Furthermore, the distributions of non-metric traits were consideredonly for males, yet the female burials show different associations of traits to the males. Twofemales possess pairs of traits which are associated with males both with and withoutweapons, posing difficulties for Harke's model of co-residential non-interbreeding families.

A programme of DNA extraction was attempted on these groups so that comparisonsmight be made with modern German and British populations. Unfortunately, extractingancient DNA is extremely difficult and the failure to find remains of ancient DNA in theskeletal material led to the project's abandonment. 1JJ

1

6

5

4

2

3

Female

Male

Child

2

6

3

4

Mortuary phase

122 THE ARCHAEOLOGY OF DEATH AND BORIAL GENDER AND KINSHIP 123

also suggests that this may represent the creation of sub-lineages at the time of the tomb'slast remodelling. As the lineages and their sub-lineages grew, so larger numbers of peoplehad rights of burial within the tomb. While children were normally excluded from burial,the number of adults interred suggests that the living population from which the deadwere drawn was initially very small in the tomb's earliest phases, possibly just three smallhouseholds. However, the same reservations can be made as in the case of Harke's study;postcranial traits especially are influenced by environmental factors such as physicalactivity and nutritional status, so the patrerning within the tomb migbt signify differencesother than kinship.

Blood gronps and collagen groups at Mozs

Some remarkable results were obtained on a small fifth century AD Pannonian cemeterynear Mozs in Hungary.l36 Blood-typing and collagen-typing linked twenty-five of thetwenty-eight skeletons, leading to the identification of three generations of four families.Blood groups were valued over collagen types since the latter could not be clearly relatedto genetic inheritance and then these results were used to construct probabilities ofrelationships according to the constraints of their inheritance. The permutation into whichmost of the burials fitted and which fulfilled the condition of maximum biologicalprobability was determined as a genealogical table (Figure 5.11). Three of the four familieswere considered to have intermarried in the second generation. Children (whether fullygrown or not) were buried close to their mothers rather than their fathers, suggesting astrongly matrifocal family structure.!3? Another interesting feature of the cemeterypopulation was the presence of artificially deformed skulls from eleven of the skeletons.This is a bodily alteration performed on infants and is considered to have been commonamong the Huns and related ethnic groups of the Migration Period. The results indicatedthat this practice was initiated on one individual within the second generation and wasadopted for all ten of the third generation, rather than representing the presence of anentirely separate ethnic or kin group.

DNA: mummies, bog bodies and skeletons

Initial atrempts to extract DNA from two Egyptian mummies of the first millennium BCdemonstrated that degraded and chemically modified DNA sequences could be obtainedfrom such ancient remains. 138 Later, small quantities of DNA were recovered from threeout of an experimental sample of fourteen Egyptian and Peruvian mummies.139 These firsttests used cloning techniques in which a modern DNA sequence was inserted into anexisting genome. Subsequently the PCR (polymerase chain reaction) method was adopted,allowing the rapid amplification of any DNA sequence140 DNA was also recovered frompreserved human brains from North American bog bodies of the Archaic period dating toc. 6000 BC at Windover and c. 5000 BC at Little Salt Spring.!4! However, tests on Britishbog bodies indicated that these and their European counterparts would not provideDNA. 142

As mentioned above, the usual forensic method for establishing familial relationships ­chromosomal DNA 'fingerprinting' - is not usually possible with archaeological remainssince the relevant DNA sequences are often poorly preserved.!43 Although the retrieval ofancient mitochondrial DNA from archaeological samples of bones and teeth is feasibleand well established, the recovery of reliable results is difficult. Chromosonal analysis can

I

• •25 26 28•

3RD GENERATION

10 M

5.11 The plan of the Migration period cemetery of Mozs and the kinship tree derived from blood grouping ofskeletal remains (circle = female, triangle = male). The dramatic results must, however, be accepted with caution

since the family tree is only a 'best fit' model.

be used to determine sex by identifying products from the X and Y chromosomes, asdemonstrated on nineteen out of twenty skeletons from the Norris Farm Native Americancemetery dating to c. 1300 AD.!44 An earlier study had successfully recoveredmitochondrial DNA sequences from fifty skeletons in this cemetety.!45

The technique appears, so far, to be most useful in documenting prehistoric migrationssuch as the possibility of a post-Jomon migration into Japan in the first millennium BC, thenature of Polynesian colonization of the Pacific between c. 1600 BC and AD 700 as far asEaster Island, and the post-glacial populating of the Americas.14' There are few studies asyet in which kinship and burial practices have been investigated. One of these is Shinodaand Kunisada's combined analysis of mitochondrial DNA sequences and funerary ritesfrom fifty-five burials of the Yayoi period (first century BC-AD) cemetery of Kuma­Nishoda on the Japanese island of Kyushu,!47 Related work on Yayoi period remains fromKyushu has demonstrated a relationship between burial style and haplotype (geneticsequence) which may be interpreted either as burial practices varying according to kinshipor as a change in burial styles accompanying a change in the genetic structure of thepopulation.!48 Undoubtedly, funerary archaeology is poised on the edge of a new era ofbiological reconstruction of kinship and sex. These opportunities also bring with themissues of interpretation which will be complex and often problematic in political andethical terms,!4' We have opened something of a Pandora's box.