Introduction

1Our ability to investigate kinship practices in Neolithic communities has been revolutionized by combining measurements of biological relatedness through ancient DNA analyses with osteological and isotopic analyses and assessments of funerary practice, funerary architecture or cemetery development. Such results have been met with enthusiasm but also some concern, particularly over those that report detecting patrilineal descent (e.g. Cveček, 2024). The identification that Neolithic people placed the bodies of people (particularly males) who were connected by patrilineal descent close together in burial grounds by itself gives us only a narrow window into one aspect of Neolithic kinship practices. However, the studies that have reported the existence of mortuary practices tracing patrilineal descent have said far more about kinship than this, and provide yet further opportunity for thinking about kinship in the Neolithic. In this article I will summarize some of the evidence that patrilineal descent was traced during certain Neolithic mortuary practices, before explaining how the most revealing studies contextualize and relativize the tracing of patrilineal descent alongside other aspects of kinship practice. I note that such aspects are only archaeologically visible because they are silhouetted by the evidence that people from the same patriline were co-located in death. I will argue that lineal descent forms a particularly bright thread in the wider fabric of kinship – a thread that is especially easy to see using archaeogenetic methods – and that we need to contextualize this thread within the wider weave of kinship. In this way, we may be able to detect practices that, for instance, may have supported various cognatic kinship connections. At the same time, I suggest that we should avoid seeking to pigeon-hole Neolithic communities into distinct categories of kinship systems, and rather think about the range of kinship practices we can detect and consider how those changed over time. By focussing on specific practices, and the connections between them, we can best describe the fabric of kinship in each case without falling back on universalising and reductionist categories of kinship systems. Finally, I will examine the limitations and challenges involved in drawing conclusions about parenting and social and political organization from the archaeogenetic evidence that is currently available, while encouraging further studies combining aDNA, isotope analysis, and conventional archaeological (including bioarchaeological) analyses.

Terminology

2Kinship is a social phenomenon. It has been described – perhaps too broadly (cf Whittle, 2024) – as the* *‘mutual relatedness of beings’ (Sahlins, 2011): beings who may or may not be human, and may or may not be alive. In more restricted way, kinship has also been described as a way of identifying, naming and codifying relationships between people in a social system (e.g. Leach, 1982:107; Stone and King, 2019:8-9). A detailed outline of kinship systems, and a detailed review of changing anthropological approaches to kinship, lies beyond the scope of this piece (recent archaeological reflections on these topics can be found, for example, in Brück, 2021; Ensor, 2013:10-27; Fowler, 2022:68-71; 2024:192-195; Whittle, 2024:2-5), but it is important to note that there are many different ways to trace descent, and that descent is only one aspect of kinship. If kinship is cognatic, each person recognises kin from among several connected descent groups (Stone and King, 2019:152-156). In bilateral descent, each person may trace descent from either their mother’s line or father’s line; in ambilineal or bilateral descent, each person traces descent selectively from among their mother’s line and father’s line, while double descent combines both matrilineal and patrilineal connections. Different anthropologists have placed greater or lesser emphasis on the importance of lineal descent: for instance, Lévi-Strauss highlighted the significance of affines in alliances through marriage arrangements (Lévi-Strauss, 1949), and the role of ‘house societies’ in which lineage is significant in house formation but members have the potential to belong to several lineages and choose which to join (Lévi-Strauss, 1982). Whatever lineal descent relations might exist rely also on lateral ties across lineages – for instance, the social ties retained between a woman and her natal kin if she has married into a different patriline. The idea that kinship revolves around human reproductive relationships has also been heavily deconstructed (e.g. Schneider, 1984), and the place of substantial terms in the formation of kinship, such as ‘blood lines’, has been extensively examined (e.g. Carsten, 2004). While biological relatedness is certainly not the same thing as kinship, kinship incorporates an understanding of reproductive relationships and biological relatedness (Shapiro, 2016) which can play a more or less significant role in the kinship ontology of a community (Fowler, 2024).

3Not all social relationships are relations of kinship: a housemate or friend or colleague, for instance, may be personally close to you without being kin. The boundaries of kinship are negotiated and fuzzy; perhaps one day you may come to recognise that friend as kin. In what follows, I use the term elective kin to refer to kinship connections that are not based on biological reproduction or affinal relationships (i.e. via marriage), but may use the same terms (mother, daughter, niece, cousin, etc). A brother can be a brother without being a biological sibling, and the use of the term recognises him as kin (though, again, there is flexibility here and the same term may be used in a looser and more metaphorical way to refer to a more generalized sense of kinship or affinity). Of course, kinship is always elective to some extent in that people choose to continue to be kin or can sever kin ties – as a social phenomenon kinship is after all something that people produce, work at, build and maintain – but the term ‘elective kin’ is useful to highlight cases where kinship is produced between people who are not close biological relatives. This is particularly necessary when working with the results of archaeogenetic analyses since they provide us with biological data that need to be contextualized within the fields of social practice that produced the human beings and the patterning in the deposition of their remains from which that data is derived. Despite critique (Cveček, 2024:4), then, I retain an analytical distinction between biological progenitors (e.g. ‘biological father’, ‘biological grandmother’) and social parents (simply ‘father’, ‘grandmother’) largely in order to explore cases where Neolithic parent–child relationships may have been partly elective (e.g. as in ‘step-parents’) or wholly elective (e.g. as in ‘adoptive parents’). I will return below to further problems with inferring parenting and parenthood from such analyses, however, since the practice of parenting is clearly quite different from the acts of conceiving offspring or giving birth.

4Those buried together need not be close kin, but archaeologists and anthropologists have long been aware of a strong connection between kinship and the co-location of the dead (e.g. Fleming, 1972:65-66; Goldstein, 1981:57; Schmidt and Déderix, 2018:209). Archaeologists studying the remains of mortuary practices are working with the results of deliberate choices made by Neolithic people about whose remains should be placed where, and with whom. These are social decisions, and that is why is it possible to discuss kinship when using archaeogenetics to analyse such remains – indeed, in doing so, we are attending to the dynamics of kinship practices, which we know to be fluid, negotiated and contested (cf. Frieman, 2023:44-45). These mortuary presentations were part of how the dead were transformed, and part of how relationships between the living and the dead were transformed; they tell us about the concerns of the living as they interact with the remains of the dead. They also tell us something about how that community coped with death emotionally and psychologically; about how the living kept the dead close, continuing their bonds with them, and/or distanced themselves from the dead (Croucher, 2017). Perhaps keeping the dead close to one another may also imply relationships of care for the dead by the living in which the dead were also expected to care for one another and to join those dear to them after they died. Archaeogenetic analysis has to be open to a wide range of possibilities, including emphasis on lineal descent, emphasis on co-residence or other elective or experiential bases for kinship, and the possibility that understandings of relatedness did not involve a concept that aligns with our idea of biological relatedness. Each mortuary analysis that includes genetic data therefore needs to consider what relationships there may have been between biological relatedness and kinship practices in the community in question, and then work outwards from there to discern in more detail how kinship worked in that community. This is certainly not to posit that biological relatedness precedes kinship, but to acknowledge that kinship is practised* and* that it is the means by which biological relatedness is reflected on socially – even if in the process biological relatedness is diminished in comparison with other forms of relatedness.

• 1 Fowler et al. (2022) refers to ‘lineage’ and ‘non-lineage’ individuals where, in retrospect, I thin (...)

5While we should not assume that a genealogical model of relatedness was current in past communities (e.g. Johnston, 2020, 14), in the next section I will outline the biological and spatial relationships that support the inference that patrilineal descent was traced by those depositing the remains of the dead in three Neolithic burial grounds from northern Europe. I will then move on to explain how analyses can move from identifying pedigrees to interpreting patrilines, and then illustrate how contextualizing these results allows us to draw wider inferences about kinship. That includes putting descent in context. In what follows I use the term pedigree to refer to those connected via biological descent, and the term* patriline* to refer to the social categorization of people into a lineage1. I use the term subadult rather than child to refer to individuals of young age who have not yet reached adulthood in order to retain child as a kinship relation paired with parent. I start from the basis that lineal descent is only ever one strand in the fabric of kinship, that descent need not be unilineal, that lineal descent can be traced through both maternal and paternal lines at the same time (double descent) or selecting one or the other (bilateral descent), and that the tracing of patrilineal descent can be a dominant trend in a community where kinship as a whole is cognatic (cf. Fowler, 2022:72, 83-84). The kinship roles, authority and autonomy of women varies widely among the range of present-day communities that trace patrilineal descent, and we should be alert to such variation in the past (Stone and King, 2019:70-99; Brück, 2021; Bickle and Hofmann, 2022:110-115; Fowler, 2024). Furthermore, descent might be stressed in some contexts, such as during deposition at chambered tombs, without featuring to the same extent in all other social contexts.

Tracing descent

6I will focus firstly on three Neolithic burial grounds that have seen recent interpretations of kinship drawing together new ancient DNA results with other lines of evidence: a cemetery at Gurgy ‘les Noisats’ (France, c. 4850-4500 cal BC), a chambered tomb at Hazleton North (England, c. 3700-3600 cal BC), and a tomb at Frälsegården (Sweden, c. 3100- 3000 cal BC). These are certainly not the first or only studies to identify groups of close kin that were deposited together in Neolithic Europe, and they are considerable distances apart in time and space. However, each of these studies involve high-coverage analyses of large numbers of individuals from the same burial ground – as such they are, I would suggest, especially valuable in providing high-quality information crossing several successive generations. Such cross-generation results are crucial for detecting trends in kinship and also changes in kinship practices, providing firm bases for comparative discussion. In the case of Hazleton North and Frälsegården the results are all the more remarkable because of the commingled nature of the human remains, which has for a long time frustrated attempts to identify the age-at-death and sex of specific individuals. The purpose of these comparisons is to highlight both similarities and differences between the three cases, but this should not be taken as suggesting they represent any kind of homogeneous large-scale shared cultural millieu.

7Analysis of 94 aDNA samples from a cemetery containing 128 burials at Gurgy ‘les Noisats’ in the Yonne valley within the Paris Basin (c. 4850-4500 cal BC) detected 76 close biological relatives spanning seven generations, and identified that closer biological relatives were buried closer together (Rivollat et al., 2023). Cemetery burial within flat graves of this kind was only one of several mortuary traditions within the region, including clusters of structures de type Passy long mounds each of which covered small numbers of burials which in some cases were grouped together according to age and sex (Chambon and Thomas, 2010), and the relationship between these varied burial modes is not well understood. No local settlement has been found. Two main pedigrees were reconstructed from the Gurgy genetic results, each of which revolved around repeated father–son connections (figure 1); individuals from pedigree A buried in a central and southern areas, individuals from the smaller pedigree B in the northern area of the cemetery, and individuals biologically unrelated to others spread throughout the cemetery (figure 2). It is striking that no half-siblings were detected in the entire burial ground, and only in two cases were a mother and adult daughter both present. Twelve individuals had no close biological relatives in the cemetery, while another six had one other biological relative present but were not part of either pedigree. Of the sampled individuals attributed to a pedigree, 49 were male and 20 were female, of which 13 of the males and 9 of the females died before reaching adulthood – in the latter case, all were among the last three generations buried at the cemetery. In some cases, subadults shared a grave with a close biological relative, such as biological father and son (grave 285) or a brother and sister who died in childhood (grave 215).

Figure 1

Biologically related and unrelated individuals from Gurgy ‘les Noisats’ (after Rivollat et al., 2023. Courtesy of Maïté Rivollat) | Individus biologiquement apparentés et non-apparentés à Gurgy ‘les Noisats’ (d’après Rivollat et al., 2023. Avec l’aimable autorisation de Maïté Rivollat)

Figure 2

Burial locations of biologically related and unrelated individuals from Gurgy ‘les Noisats’ (after Rivollat et al., 2023. Courtesy of Maïté Rivollat) | Plan des tombes des individus biologiquement apparentés et non-apparentés à Gurgy ‘les Noisats’ (d’après Rivollat et al., 2023. Avec l’aimable autorisation de Maïté Rivollat)

8A single pedigree was reconstructed connecting 27 out of 35 aDNA samples from a chambered tomb at Hazleton North, Gloucestershire (c. 3700-3600 BC), which contained at least 41 different individuals according to osteological assessments (Fowler et al., 2022). There are over a hundred chambered long cairns from the area around the Cotswolds and Black Mountains, some of which have bilateral arrangements of chambers with entrances at opposing sides of the cairn, as Hazleton North does (Darvill, 2004). The tomb was built over the remains of earlier Neolithic occupation activity including a midden and post-holes from what may have been a house (Saville, 1990). Strontium isotope analysis has suggested that the geology on which the site was built contributed to one of two sources of food consumed during the early lives of most individuals sampled and that for some of these individuals the other geological source was at least 40km away, suggesting residential mobility (Neil et al., 2016). There were no cases of biological mothers and daughters in the tomb compared with 13 biological father-to-son connections (four of which were with subadult sons). One first generation male (NC1m) had offspring with four different women, three of whom were among the sampled group. All 12 descendants of two of these women were placed in the southern chambers of the tomb, while 9 of 13 descendants of the other two women were entombed in the northern chambers. Only nine individuals of the 35 sampled were female, of which only two were born into the pedigree and both of those died in childhood. By contrast, 26 of the sampled individuals were male, of whom 7 died in childhood (figure 3).

Figure 3

Biologically related and unrelated individuals from Hazleton North; those with bold outline were interred in the northern chambers (after Fowler et al., 2022) | Individus biologiquement apparentés et non-apparentés à Hazleton North ; ceux avec un contour en gras étaient inhumés dans les chambres nord (d’après Fowler et al., 2022)

9The tomb at Frälsegården (c. 3100-3000 BC) in Falbygden, Sweden, consisted of two chambers separated by the end of a passage which intersected with the chambers at a right angle and connected the chambers with the entrance to the tomb. This is one of a cluster of tombs in the Falkö-ping area, seven of which were sampled in the same study. Genetic analysis of the commingled remains from individuals who were probably originally entombed as intact corpses indicated that 36 of those sampled were close biological relatives that could be arranged into two interlinked pedigrees (figures 4-5), referred to here as ‘left-hand’ and ‘right-hand’ pedigrees according to a diagrammatic representation of two ‘sub-lineages’ in the original article (Seersholm et al., 2024). The authors use the term ‘sub-lineage’ for these two pedigrees but I will refer to these as pedigrees and reserve the term lineage to describe a social grouping that can be detected by noting an intersection between pedigree and tomb location which does not exactly match the pedigrees (discussed in the next section). In four cases the same male reproduced with two different women. 28 males were identified among the aDNA samples, at least 23 of whom were born within the pedigree (of whom 7 were sub-adult), and 16 females, of whom only four were born into the pedigree (two of which died as sub-adults) and 6 are attested to have reproduced with pedigree males. The right-hand pedigree from Frälsegården also connects with individuals who were entombed c. 6km to the west: two brothers of a woman (FRA028) who reproduced with a Frälsegården male were found at the tomb of Hjelmar’s rör (HJE003 and HJE012). Overall, the composition and structure of the pedigree suggest that virilocal burial following female post-marital transfer from one lineage to another was a key kinship practice (Seersholm et al., 2024).

Figure 4

Biologically related and unrelated individuals from Frälsegården, identifying their location within the tomb (north chamber, south chamber, central chamber, or intersection between the central and south chamber) (after Seersholm et al., 2024) | * *Individus biologiquement apparentés et non-apparentés à Frälsegården, avec identification de leur emplacement dans la tombe (chambre nord, chambre sud, chambre centrale, ou au croisement de la chambre centrale et de la chambre sud) (d’après Seersholm et al., 2024)

Figure 5

Location of sampled remains for individuals from the two pedigrees at Frälsegården (after Seersholm et al., 2024) | Emplacement des restes échantillonnés des individus appartenant aux deux lignées de Frälsegården (d’après Seersholm et al., 2024)

Relativizing descent

10The previous section outlines only some of the spatial arrangements and biological connections between the dead at these three sites, and these descriptions of even the biological relationships between individuals are extremely parsimonious. However, the analyses of these sites did not stop there. This section will demonstrate how the patterns of patrilineal descent connections detected through the contextualized aDNA analyses provide a platform for drawing further inferences about kinship. These inferences take account of all of the biological relationships between individuals in the tomb *and *those cases where no biological relatedness was evident, and set the results in archaeological context. Most of what follows are interpretations made in the original analyses (i.e. Fowler et al., 2022; Rivollat et al., 2023; Seersholm et al., 2024) or subsequent publications (e.g. Cummings and Fowler, 2023; Whittle, 2024), though some are new: they illustrate why we should not reduce the results of these analyses to an identification of patrilineal descent in secondary accounts. As well as offering new observations about kinship practices, such results can be deployed in enhancing our understandings of gender, age categories, personhood, and how the dead were commemorated, and I will reflect on those issues a little too.

11Just as close biological relatives sometimes shared a grave at Gurgy, so did some biologically-unrelated individuals, and, given the co-burial of some close biological relatives mentioned above, it seems likely that the relationship between two people in one grave was often one of kinship. Perhaps the clearest example was grave 236, shared by a boy and a woman who reproduced with that boy’s adult brother – an affinal connection between these two seems likely. If we interpret the biological father and son sharing grave 285 as kin, and these two individuals as affinal kin, then should we not also consider kinship a likely connection between those sharing a grave who were not close biological relatives, such as the woman and girl in grave 232 or the adult male and adult female in grave 243? In the latter case the male had offspring with a different woman, and, if we accept the idea that it was not permitted to have multiple reproductive partners in this community, the nature of their kinship may be one other than reproductive partners: were they non-reproductive partners, elective siblings, or kin by co-residence (both have similar strontium isotope results)? Were they more distant kin or affines, buried closer together than would be usual because they had died around the same time? Was the woman in grave 232 an elective or foster parent of the girl she was buried with? We may never be able to pin down such cases entirely, but they are intriguing and we can only discuss them because they are thrown into relief by the wider patterning in the contextualized aDNA data.

12As Rivollat et al. (2023) argued, the demographic representation in the last three generations at Gurgy is notably different to the first two and to the middle two. In the final years of the cemetery, they suggest, the members of the new generation who died young were buried as usual. However, before the surviving members of that generation who reached adulthood had died the cemetery was abandoned, perhaps because the community had relocated. Perhaps a similar relocation had happened during the founding of this cemetery, since there are no subadults among those buried in the first three generations. From the fourth generation, where subadults appear, biological siblings were buried close to one another (Rivollat et al., 2023:601). While these changes do not necessarily indicate a change in a ‘kinship system’, they arguably did change the way that the mourning and remembering of kin was experienced in the community. Those in the first three generations could perhaps not easily visit the graves of their siblings or parents or even children, while those in the fourth generation could. And maybe that fourth generation was unusual, as within two generations the separation of those who died as subadults from their adult relatives would happen again. Things could have been otherwise: the dead could have been exhumed and moved when the settlement relocated, or people could have returned to a pre-existing ancestral cemetery each time there was a death, even if that cemetery was some distance from their new settlement. Instead, a spatial rupture was allowed to form between two communities of the dead. This is an important matter for archaeologists if we adopt the stance that kinship is something that people do, work at and experience – when the materiality of that work changes, then in some way so does kinship. In effect, it was important to bury people at Gurgy near to their closest biological relatives, but it was not actually important that all close relatives were* always* buried close together, and at the beginning and the end of the sequence it was not even important that they be buried in the same cemetery. Descent was traced in the co-location of the dead, but it was not so important to connect lines of descent to the extent of maintaining a single burial ground beyond the span of seven generations. Longer-term descent connections were possibly traced through oral histories, or even visits to now-distant burial grounds, but that lies beyond our reach.

13Something similar could be said about Hazleton North, which physically united those who shared descent over five different generations but no more. Patrilineal descent here was subdivided by maternal descent groups which largely aligned with the separation of the human remains into the two opposed sets of chambers either side of the cairn (Fowler et al., 2022:3-4). The intersection between age-at-death, sex and biological relationships among those buried also provided key information for interpreting age categories and gender, a point developed by Cummings and Fowler (2023). For instance, there are only two daughters with a biological progenitor in the pedigree, and they both died in childhood. This suggests that upon reaching adulthood the remains of women were treated in different ways to both boys and men from the lineage. By contrast, the analysis identified three first-generation women who reproduced with NC1m and one second-generation woman who reproduced with one of the of sons of NC1m by a fourth first-generation woman. While one of these women died before the age of 25, so would not have become a biological grandmother during her lifetime, one survived past the age of 40 and another to the age of 48-56 years. Given the absence of so many women from the tomb perhaps the inclusion of these four women indicates they were the heads of sub-lineages during life, became parents and grandparents to many individuals (see below), and/or became honoured ancestors once their sons had offspring of their own. Their presence arguably illustrates the centrality of such female ancestors in how kinship was traced during the founding generations of the tomb.

14The first- and second-generation women within the tomb also played a key role in the wider construction of the descent group beyond a simple pedigree connecting biological fathers and sons. The two first-generation women NC2f and NC3f reproduced not only with NC1m but also each with one of his third- or fourth-degree male relatives. The inclusion of these other sons of NC2f and NC3f and their offspring within the tomb may indicate elective kinship in the form of adoption into the patriline, and Fowler et al. (2022:4) suggested that social fatherhood may have been more important among lineage males than biological paternity on this basis. The connection via the mother is key here, and we might alternatively suggest that the sons of these women were a part of the descent group using the tomb by right of their mothers being placed in the tomb: this is also an elective kinship decision, but one which could be consistent with double descent or bilateral kinship (cf. Cummings and Fowler, 2023:8; Greaney, 2023:186; Whittle, 2024:8-9). Among bilateral communities each individual can elect to trace descent from either parental line at any generation, but communities that trace descent bilaterally do not normally form discrete descent groups (Stone and King, 2019:150). The collective nature of the deposit at Hazleton North suggests the formation of a distinct descent group, though it is possible that bilateral descent was practised in the first to second generation when the community using the tomb was especially inclusive and expansive. Yet, as Cummings and Fowler (2023:10) note, there is no evidence that such practices were repeated later. If descent was bilateral (or alternatively double) beyond those initial generations, then burial remained strongly patrilocal (cf. Ensor, 2021:table 2.1).

15Double descent might be considered here, wherein each person belongs to two descent groups and some features are inherited down the patriline (e.g. land) while others are inherited down the matriline (e.g. cattle) (Murdoch, 1940). Ethnographic study of double descent among patrilocal Ashanti communities suggests that a property translated as ‘blood’ flows down maternal lines, while ‘spirit’ is ‘transmitted exclusively through males’ (Murdoch 1940:557). We might consider that bodies at Hazleton North were composites of substances and attributes with differing properties, some of which might be transmitted down male lines and others inherited from mothers so that each person was formed in a complex processual intersection embodying gender, kinship and personhood (cf. Fowler, 2024). The sex identification of so many individuals at Hazleton North has highlighted that women’s remains were treated differently to men’s: women born to this lineage were not included in the tomb, so perhaps their remains were not returned to their natal community after death, and neither were all the women who reproduced with a lineage man admitted to the tomb. Perhaps the remains of women were taken to a different kind of place and perhaps they were cremated or exposed to the elements or dispersed: we do not yet know. But differing mortuary practices for women (and probably some girls) might have related to the transmission or release of properties that they embodied compared with those embodied by males.

16A final possibility, recently suggested by Whittle (2024:9), is that we ‘think in terms of Melanesian big men, bolstered by an agnatic core needing the support of many others including matrilateral kin and affines, and usually destined for a brief floruit’. Again, that might fit well for the lifespans of those in the first two generations, but less so by the time individuals from the fourth generation were entombed. Overall, we could perhaps consider some kind of cognatic kinship system within the community placing the remains of their dead at Hazleton North, but, again, only if we acknowledge that it included a strong patrilineal bias including patrilocal burial (cf. Fowler, 2022:72, 83-84).

17It is important to think, as these suggestions do, about a diverse range of ways that the results from Hazleton North could be interpreted. Yet anthropologists themselves acknowledge the difficulty with creating typologies of kinship systems, including the fine lines between different systems (Stone and King, 2019:55-56). The terminology is complex and different sources use different terms for the same relationships or the same term in differing ways. There are also significant differences among communities that have been identified as sharing a type of kinship system (Bickle and Hofmann, 2022; Cveček, 2024; Fowler, 2024:194-195). I think that rather than trying to determine on the basis of the current evidence whether kinship at Hazleton North revolved around patrilineal descent or whether that formed one key element in a double descent, bilateral, or other cognatic descent system, we are better served by focussing on the individual strands of kinship practice that we can directly detect and thinking about (a) how they were interwoven and (b) how they changed over time. Ultimately, the experience of kinship for NC1m and the women he reproduced with was probably rather different to the kin relationships experienced by third and fourth generation members of the community. While patrilocal and patrilineal burial persisted, as did, to a significant extent, the subdivision of the dead according to a female ancestor, the inclusion of mothers with their sons did not.

18Clearly, kinship was not a static framework that was ‘represented’ by the tomb at Hazleton North; rather, people engaged in kin relationships when they conceived, constructed and used the tomb, and some of those practices and experiences changed over time. We can also see a transformation in how kinship worked at Frälsegården. Some of the first- and second-generation individuals from both pedigrees were placed in the northern chamber, while among subsequent generations only individuals from the right-hand pedigree were selected for the northern chamber, and at least four left-hand pedigree individuals from the second and third generation were placed in the southern chamber (Seersholm et al., 2024:117; figures 4-5). A woman who reproduced with a second generation left-hand pedigree male was placed in the central chamber, as were most third to sixth generation individuals from the right hand pedigree. The only two individuals from the left-hand pedigree to be placed in the northern chamber were first generation father FRA011 and his son FRA012. They joined FRA021 and his son, a woman who reproduced with his son, and his grandson. I would suggest that FRA011 and FRA021, and their sons, might actually all have been part of a single ‘northern’* sub-lineage with access rights to the northern chamber. This can be contrasted with a ‘southern’ sub-lineage descending from FRA011’s unsampled brother (who reproduced with two women), all of whom for which locations are clearly reported in Seersholm et al. (2024) were placed in the southern chamber or the point where the southern and central chambers intersected (figures 4-5). By the third generation, the ‘northern lineage’, were repeatedly using both the northern chamber and the central chamber. We could therefore infer a *recursive interaction between kinship and tomb architecture, whereby the division into two chambers was initially related to a kinship distinction which became less significant in the placement of the dead by the fifth generation. Furthermore, in at least some cases the remains of the dead were also positioned close to their closest biological relatives, as with FRA01, FRA022, and FRA023, and FRA001, FRA002 and FRA003 (Seersholm et al., 2024:117). Two male and six female individuals sampled were not close biological relatives of anyone else in the tomb, and we could speculate that these were members of either the northern or southern lineage depending on their placement (whether that membership was entirely elective or due to a reproductive partnership that did not yield offspring who were entombed at Frälsegården).

19In each of these three burial grounds the majority of people’s remains were carefully deposited in relation to other individuals who were either contemporaries or predecessors, but there were changes over time in exactly how this worked. In all three cases there were also exceptions and ambiguities, exactly as we should expect with so complex a social issue as kinship. Not all individuals in each section of the site were members of the biological pedigree that included most of those they were placed alongside, but in each case we can detect that patrilineal descent was a common factor connecting the majority of those placed in the burial ground. At Hazleton North, the lineage’s tomb – and likely, therefore, the lineage – included members whose biological father was not a member of the pedigree. In each burial ground, the duration of deposition was less than or around 100 years, perhaps suggesting that certain people sought to renew the lineage – or sought to signal the centrality of their branch of the lineage – by founding a new burial ground. It is also possible that the burial grounds were abandoned as the kin group no longer had a place of habitation nearby. In that case, relationships of belonging, kinship and place must have been reworked during that relocation when at least some of the dead were left behind.

20We could also tentatively suggest that the division of the community into some form of duality was important in each of these three cases: in addition to the patterning noted at both tombs, individuals from pedigree B at Gurgy were consistently buried in an area of the cemetery to the north of those members of pedigree A (figure 2). While not ubiquitous, duality in tomb architecture is a common theme attested across several different regions and periods of the Neolithic in northern Europe, including bilateral chambered cairns (Fowler, 2022-76-78), opposed sets of chambers at some court tombs in Ireland, the Isle of Man and western Scotland (Fleming, 1972:63; Fowler 2022:79-81; Powell, 2005:15), divisions along the axis of the passage at some Irish passage tombs (Robin, 2010), ‘mirrored’ sets of side cells either side of an entrance passage at some Late Neolithic Orcadian passage tombs (Richards, 2005:130-131; Fowler, 2021:93), and in parallel chambers at Danish double passage tombs (Dehn and Hansen, 2000). It would be highly risky to infer too much from just three disparate examples, but the presence of a north–south duality at all three sites is also intriguing, and could be evaluated further for other sites. For instance, of the 23 out of 45 passage tombs considered by Robin (2010) which showed signs of axial opposition, the passages of at least nine were oriented in such a way that the two side chambers were located to the north and south of the main chamber. It is possible that duality was a common device in framing Neolithic kinship connections and distinctions, in some cases potentially associated with specific directions.

21Currently, we do not know how typical the results from these three cases are of other similar burial grounds, and should be wary of extrapolating too far from individual cases. Indeed, it is also vital to attend to the variation between the way that descent was traced across these cases: for example, maternal descent clusters and half-siblings were present only at Hazleton North where some men and women had multiple reproductive partners, while only men seem to have had multiple reproductive partners at Frälsegården (4/4 cases) and there were no multiple partnerships and no half-siblings at Gurgy. Perhaps some aspects of kinship were less open to negotiation at Gurgy than Hazleton North, with kinship work invested in excluding biological half-siblings from the burial ground at Gurgy or potentially even the living community – perhaps even in preventing their existence. Kinship ideals may have changed during the use of a burial ground and even varied within the community, or practice might have conformed to or diverged from those ideals to a greater or lesser extent. Perhaps monogamy was the ideal in both cases, but in practice this was frequently ignored at Hazleton North. That arguably lies beyond what we can determine, but what we can say is that depositing people connected by patrilineal descent was articulated with other concerns in each case, and that those concerns differed from case to case and may have been more open to mutation or responded to with more heterodoxy in some cases.

22To sum up, these cases allow us to explore diversities and changes in kinship practices over relatively short periods of time at the scale of a single site, and to compare across sites. It is precisely in the changing relationships between lines of kin and tomb space, in half-siblings, affines, maternal descent clusters, and potential elective kin, and in detecting that biologically unrelated individuals were folded into that fabric that we see the messy, dynamic, contingent reality of Neolithic kinship. But, even with this complexity coming into view, significant challenges remain. I would like to focus on two: parenting and social organization.

What does identifying the tracing of descent imply about parenting?

23The studies above tend to assume that where a biological progenitor and their offspring are present together it is reasonable to infer they are parent and child. Yet we know that the distribution of child-rearing, caring and teaching activities is highly varied and diverse and can involve very many different members of the child’s community. Caring for, feeding, educating, holding rituals for, and simply spending time with children are all fundamental to building ties of kinship between children and parents. Parenting is relational and processual, and might be ‘distributed among several persons’ (Alber, 2023:38-39). In some cases, kinship terms that presuppose living arrangements related to parenting have been used in archaeogenetic analyses without good evidence that they are appropriate. This is particularly the case for the term ‘nuclear family’ (e.g. Seersholm et al., 2024:118, referring to the close spatial proximity of biological parents and their offspring, such as FRA001, FRA002 and FRA003). Nuclear family structures should not be assumed simply because we see the co-presence of biological parents and one of more of their children: anthropologically, co-residential extended families, and kin groupings formed by co-residence, are more widespread. Perhaps those first-generation women whose remains were placed in the chambers at Hazleton North were key parental figures not only for the generation following them, but the generation after that too. Maternal kin often play a role in supporting mothers in patrilineal societies if they live nearby. Children can also be fostered for long periods of time, including by distant relatives (both in terms of geographical and kinship distance) (Alber, 2023). These are important considerations when interpreting the evidence set out above. For instance, perhaps NC5m was a foster child whose foster parents were part of the sub-lineage using the northern side of Hazleton North: perhaps when he died, aged 3 or 4, he was close to his foster parents and so buried among others of their lineage. An ongoing programme of analyses at Hazleton North will integrate new isotopic data with the genetic data to examine both individual and shared patterns of diet and mobility within the group, including between different maternal descent clusters and across generations (Brett Ostrum, pers comm). These integrated results may help further consider whether some of the Hazleton North individuals were fostered and which individuals may have been co-residential based on their mobility patterns.

24Where we find a biological progenitor and their offspring buried together I do actually think it is reasonable to interpret their relationship as parent and child – albeit without seeing such a parenting relationship as exclusive. Many others could have been involved in parenting that child in different ways. At Gurgy, males were buried closest to their biological fathers than to anyone else, suggesting this connection was valorized (Rivollat et al., 2023:601); the existence of this pattern enables us to consider specific spatial relationships that stand out from the pattern, such as female 262 who was buried closer to a distant male relative from the previous generation, 263, rather than to her biological parents and siblings (and see above for graves shared by those who were not close biological relatives). At Hazleton North, biological father–son connections are repeatedly present but in three cases Fowler et al. (2022) suggested inferring father–son connections even where the son is not the biological offspring of the father. While it is possible that men in general played parental or avuncular roles to all of the males of the next generation, it usually seems to have mattered here who their biological mother (or grandmother) and father were when it came to deciding where to lay their remains to rest. Descent seems to have been important at an ideological level, even if we cannot at present connect that with parenting practices.

What does identifying the tracing of descent imply about social organization and political relations?

25There are, then, problems with drawing inferences about parenting and care-giving from such evidence – which is certainly not to say that it is impossible to do (e.g., for a later period, see Rebay-Salisbury et al., 2023). There are also significant problems with drawing inferences about wider social organization based on the results from specific sites. Identifying that lineal descent was traced in mortuary practices is not the same as identifying a specific form of social and political organization. In his classic deconstruction of descent theory, Adam Kuper (1982) highlighted not only that societies which had been identified as tracing patrilineal descent (e.g. the Nuer) also valued matrilateral kinship, but also that any connections between patrilineal descent groups and either political authority or territory were weak to non-existent. This is important for any further inferences we might draw based on the results of studies like Gurgy or Hazleton North or Frälsegården. Indeed, I think that Neolithic political organization is probably the most difficult sphere to interpret – harder than cosmology or kinship – in part because the concept requires the articulation of so many different social relationships and in part because those relationships might have varied contextually (cf. Thomas, 1996:178-179; 2002; Smyth et al., 2025).

26One reason I think that genetic results and kinship inferences bring us no closer to determining whether Neolithic communities were hierarchical is that we do not yet know what proportion of the community living in a certain area built or used a tomb, and do not know whether having access to a tomb was held in regard in a way that translated into social ranking that carried across all social contexts. In some regions of Britain people did not build tombs at all; in others, tombs were abandoned by the Middle to Late Neolithic. It is unclear, at the moment, whether tomb construction was fundamental in the emergence of distinct lineages, or whether patrilineal descent was practised widely in areas where tombs were not built or in the period prior to their construction. If the latter, then we need to consider why certain lineages engaged in tomb construction at specific points in time. If lineal descent groups came together to build tombs and to entomb their dead then perhaps in the process they made claims to belong to certain places, but we do not know if such places were claimed as *belonging to them *as part of a territory (and what extent such a territory might have), or as places for which they were custodians, for instance. We do not know whether or how burial at Gurgy related to burial in Passy-type mounds in the wider region, which may be contemporary or slightly later (c. 4700-4300 cal BC – Chambon and Thomas, 2010): the higher number of burials and broad demographic spread at Gurgy seems more inclusive than Passy burials, but it is not clear whether Passy burial grounds were used by different social groups, or stem from the selection of certain individuals from a single group, and if the latter we do not know whether rank was a feature in that selection. Strontium isotope results from Gurgy suggest that members of at least pedigree A probably resided in the same locale (Rivollat et al., 2023:605), but we do not yet know which of those entombed at Hazleton North lived together or for how long, though forthcoming isotopic results may help here. Nor do we know what kinds of authority, rights or property were inherited by whom. Given that burial practices are often strongly ideological, as Bloch (1971) found in his classic analysis of Merina tombs, it may be that patrilineal descent loomed large in people’s minds during formal occasions such as mortuary ceremonies, but was not something that accurately described all aspects of people’s lived experiences of kinship.

27Given these limitations, I think we need to be careful about making wide-reaching statements about Neolithic socio-political organization based on identifying patrilineal descent practices. Dulias et al (2022), reporting on the persistence of Y-chromosome haplogroup I2a1b-M423, first seen in the Neolithic in this region, in 8 of 22 individuals