Introduction

1There are three main forms of Neolithic megalithic tomb in Ireland – portal tombs (largely free-standing megalithic chambers), (linear) court tombs and (circular) passage tombs. While there are clear relationships between court tombs and portal tombs in terms of their design (Evans 1940, De Valéra 1960), distribution and siting (Ó Nualláin 1983), orientation (De Valéra & Ó Nualláin 1961) and artefactual contents (Herity 1964), the only commonly shared feature between court tombs and passage tombs was their megalithic construction. In addition to these three main types, there is also a small number of Linkardstown-type cists (stone chambers under circular mounds). This variability has generally been explained as reflecting chronological developments, with the court tomb/portal tomb complex usually viewed as representing the early phase of megalithic construction, and passage tombs the later phase.

2The precise relationship between court and portal tombs has been disputed – for Evans (Evans 1940) and De Valéra (De Valéra 1960: 63–9) portal tombs were considered to have developed from the court tombs’ subsidiary chambers (external side- or end-chambers), while Corcoran (Corcoran 1972) argued that court tombs had developed from the multi-phase construction of complex portal tombs; Flanagan (Flanagan 1977) took a middle position, arguing that court and portal tombs were not typologically separate, but rather formed a continuum.

3There is a much clearer distinction, however, between the court tomb/portal tomb complex and passage tombs. From an early stage the chronological developments were expressed, from a culture-historical perspective, in terms of the court and portal tombs as having been built by the original Neolithic ‘colonisers’ of Ireland (De Valéra 1960: 40–8), and with the passage tombs being built by a second wave of immigrants spreading west from the Boyne valley (e.g. Herity 1974, Herity & Eogan 1977). From a social-structural perspective (e.g. Darvill 1979) this process was explained as the result of rapid internal changes from an egalitarian to a hierarchically stratified society.

4Until recently, the chronological sequence had been hard to pin down, and many of the early radiocarbon dates suggested that there could have been a significant degree of chronological overlap between the different monument types (Sheridan 1986, 2003, fig. 6, Woodman 1992: 304–5, Cooney 2000, Malone 2001, Scarre 2007). The fact that these monument types also overlapped spatially – court tombs with passage tombs, passage tombs with Linkardstown-type cists, and portal tombs with all of them – raised the possibility that the different tombs forms represent not a process of change, but rather ‘different responses, some regional, to various social and ritual imperatives’ (Waddell 1998: 100), and ‘particular and different ritual expressions by people who shared a broadly similar cultural identity’ (Cooney 2000: 139).

5The most recent dating evidence, however, has tended to confirm the general sequence. While portal tombs remain poorly dated, their construction appears to have started in first quarter of the fourth millennium BC (Kytmannow 2008), with the construction of court tombs beginning in the second quarter (Schulting et al. 2011), and the start of passage tomb construction tentatively dated to the third quarter, with their first use ending in the first quarter of the third millennium BC (Cooney et al. 2011). Dates from Linkardstown-type cists suggest their use mainly during the second and third quarters of the fourth millennium BC (e.g. Brindley et al. 1983, Manning 1985, Brindley & Lanting 1989-1990, Cooney et al. 2011: 637), thus overlapping with the uses of the other monuments.

6However, while the dates from the different monuments give indications as to when they were built and for how long they were actively used, both for burial and other forms of deposition, they do not reveal for how long they remained significant features in the lives of the local communities. By the end of the 4th millennium BC, therefore, portal tombs, court tombs, Linkardstown-type cists and passage tombs were all features of the Irish landscape. The questions remain, therefore: what led to passage tombs becoming the dominant form, and what do they signify?

7The author has previously argued that the regularity and variability in court tomb construction points to central themes of the social and ritual discourse that accompanied the tombs’ construction and use, with the symbolism evident in their form expressing alleged lineage relationships between the living community, its ancestors and the land (Powell 2005). In the same way, it can be asked: what does the construction of passage tombs reveal about the society which built them, and should these monuments be viewed as a novelty (within Ireland), or as some logical development of what had come before?

Design, symbolism and discourse

8Cooney (Cooney 2000: 91) has pointed to the increasing emphasis by archaeologists on the ‘experiential’ aspects of megalithic tomb design and use (e.g. Richards 1988, Thomas 1992, 1993). However, it is one thing to recognise that the ‘bird’s-eye’ analysis of tomb form looks at these monuments from ‘a perspective that Neolithic people could never have had’ (ibid.), but it is entirely another to suggest that ‘at many, perhaps most Neolithic monuments there is little evidence of blueprint’ (Field 2010: 8) and, consequently, that people would have started building without knowing eventual form they would take.

9There is no reason to assume that a tomb’s builders could not, and did not, visualise (or even represent) a plan view of their monument, enabling them to arrange its various structural components in the desired configuration as part of the design process. In fact, even setting aside exceptional monuments like the Newgrange passage tomb, Co. Meath, with its precise solar alignment (Patrick 1974) and geometrical shape (Powell 1994), it is apparent at many Irish tombs that care was taken to build to some preconceived design. Megalithic tomb design provided an enduring, although adaptable, material medium within which to express a relatively coherent set of symbolic relationships – relationships which may have had far more transient and intermittent expression during the performance of mortuary ritual. Moreover, the communal act of construction of a tomb may have served to integrate a community, as much as its subsequent use.

10It is argued here that it is within these design patterns that we can discern themes of the social and religious discourses which gave meaning to the tombs’ use.

Court tombs and portal tombs

11The author has previously sought to identify some of the ‘grammar’ of the ritual discourses expressed in the form and construction of court and portal tombs (Powell 2005). It was suggested that the repeated elements of court tomb design (fig. 1) – the linear mound defined by an orthostatic kerb, the linear segmented chamber on the same axis, and the concave façade of the court forming a transitional space between the outside and the inside – symbolised perceived attributes of the community of the dead inhabiting the monument.

1. The structure and component parts of a court tomb, Browndod, Co. Antrim

12It was argued that the structural regularity displayed by court tombs, when viewed in conjunction with their proximity to settlements (Caulfield 1983, Cooney 1983) and the deposition within them of artefacts found in domestic contexts (Herity 1987), and even settlement debris (Case 1973), suggests that their builders were representing the dead buried in them in terms directly relevant and recognisable to the living – portraying the present and past generations of their community as a single entity extending back from the living to the dead.

13Such ideas find close parallels in the ideologies of social groupings based on lineage and descent (Keesing 1975), in which group membership is based on a shared descent through a lineage of ancestors. The construction within the landscape of a monument to ‘house’ these ancestors symbolised the permanence of the local lineage’s association with that land, so legitimising its claims to territory. It was further argued (Powell 2005) that the variable arrangement and elaboration of the court tomb’s constituent parts (kerb, court and gallery) allowed the tomb’s design to symbolise in a ‘grammatically’ consistent way variant configurations of the idealised kinship relationships making up these local descent groups.

14It was also argued, however, that lineage relations were not the defining characteristic of the social groups that built the portal tombs (Powell 2005: 19–21). Instead, they appear to represent social units that sought to back up their claims to territory, not by reference to a continuation of long-term ancestral rights, but rather by the conspicuous display of group strength, solidarity and self-confidence as forcefully expressed in the precarious positioning of the chambers’ capstones.

Passage tombs – circularity and variability

15It is argued here, in a similar manner, that the defining attributes of passage tomb design reflect underlying symbolic principles that were expressed in the ritual discourse which accompanied those monuments’ construction and use, again relating in part to the nature and structure of the community of the dead and its relationship with the living.

16Unlike court tombs and portal tombs, there was no equivalent continuum between court tombs and passage tombs. The extreme rarity of transitional or hybrid tombs emphasises the fact that court tombs and passage tombs were structurally quite distinct. While an essential feature of court tomb symbolism was expressed in the monuments’ linearity, passage tomb design is dominated by circularity (fig. 2).

2. The structure and component parts of a passage tomb, The ‘Druid Stone’, Ballintoy, Co. Antrim

17Externally, the circular shape of the passage tomb, emphasised by its stone kerb, was mirrored by the rounded elevation of any covering mound. Inside the monument too, it is possible to view the chamber as symbolically (if not always physically) at the centre of this circle; where subdivided into separate recesses, these were arranged not in a line but radially around that centre. The apparently linear divisions in the ‘undifferentiated’ passage tombs, such as Knowth Site 15, Co. Meath (Eogan 1986) (fig. 4a), may be more accurately characterised as a central, undivided chamber reached by an outer, divided passage.

18As with court tombs, there was a requirement for access to, or communication with (Lynch 1973), the passage tomb interior. But while the court tomb kerb was angular and indented, the passage tomb kerb was smooth, undifferentiated and unbroken, defining a symbolically unambiguous boundary. At court tombs, attention was drawn to the gallery entrance by the high portal stones which flank it and frame it, and movement was funnelled towards the entrance by the concave façade of the court. At passage tombs the kerb continued unbroken across the entrance, partly concealing it, the monument giving no other indication that it even had a front. The circle offered the perfect boundary, both physical and symbolic, to separate two distinct domains – the natural world of the living outside the tomb, and the supernatural world of the deceased inside chamber. Movement across that boundary was possible, but both physically and symbolically restricted and certainly not invited (Pl. 1).

Plate 1. Knowth Site 12

Trajectories of elaboration

19Within the constraints of this predominantly circular design, the Irish passage tombs nonetheless display considerable variation resulting from elaboration along a number of trajectories (Sheridan 1986) – in the layout of their internal components (passage, chamber and cells), in the construction of their mounds, and in their size, as well as in their decoration with carved art and in their location within the landscape.

Internal structures

20At small passage tombs, such as Ballintoy (The Druid’s Stone, Magheraboy), Co. Antrim (Mogey 1941) (fig. 2), access to the chamber was gained through a simple gap between chamber orthostats, but at larger tombs the chamber could only be reached via a structurally discrete passage (fig. 3). Although passages provided access, their architecture frequently served to emphasise the chamber’s conceptual distance from the outside world, and made the process of entering the chamber literally a ‘rite of passage’; the 35-metre long passage at Knowth East, Co. Meath (Eogan 1986) represented real physical distance. At Newgrange the visual impact was accentuated by the passage roof dropping suddenly at its inner end before entering the high corbelled chamber (O’Kelly 1982), while at both Knowth West and Carrowkeel H, Co. Sligo (fig. 3g) bends in the passages prevented any external illumination of the chamber and removed the outside world from sight. Passages could also be symbolically lengthened by the addition of sills, creating a series of thresholds to be crossed, as at many of the Carrowkeel tombs (fig. 3, a and g).

Figure 3. Variable passage tomb chamber arrangements

21Although the undivided internal space of tombs with simple polygonal chambers, such at Ballintoy, mirrored their undifferentiated external appearance, at many others the chambers were subdivided, with recesses arranged radially around the inner end of the passage, equally accessibly from it. Here too there was scope for variation and elaboration – Loughcrew S, Co. Meath, for example, had two recesses set in a ‘Y’ formation, while Carrowkeel F, had five recesses (Macalister et al. 1912) (fig. 3a), Seefin, Co. Wicklow had six (Macalister 1937) (fig. 3b), Loughcrew I had seven (fig. 3c) and Loughcrew L had eight (fig. 3d) (Frazer 1893). However, these more complex arrangements, while avoiding the linear, front-to-back divisions of most court tomb galleries, may have been considered as aesthetically less pleasing, and symbolically less powerful, than the simple and symmetrical three-recess, cruciform design. This design was the form adopted at Newgrange (fig. 3e), a monument which was laid out and constructed with unparalleled precision (Powell 1994). The cruciform chamber, therefore, appears to have been sufficient to symbolise non-linear divisions and distinctions within the community of the dead – although at some tombs a subtle distinction, in size, decoration and contents, was still made between the recesses, with the right recess more dominant than the left.

22The fact, however, that Knowth, comparable to Newgrange in size and elaboration, contained both a single, undivided chamber (at the west) and a cruciform chamber (at the east), both of which were equivalent in terms of their scale of construction, and their decoration and furnishing (Eogan 1986), suggests that chamber subdivision was not an inevitable consequence of tomb elaboration, but had specific symbolic meanings relating to conceptual distinctions within the supernatural community. At some tombs, the point from which the recesses radiated became itself a structurally distinct space, and was in turn elaborated as a rounded space, perhaps to enhance the performance of internal ritual activities. At Newgrange this space had a 6-metre high corbelled ceiling making it the dominant feature of the tomb’s interior, while at Fourknocks I, Co. Meath, it was expanded to an area measuring 5.5 metres wide and 7 metres long (fig. 4b) (Hartnett 1957).

Figure 4. Variability in passage tomb construction

Mounds

23Not all passage tombs had full covering mounds, yet even in those that did not, material was often added around the chamber and within the kerb in a manner that reinforced the underlying circularity of design (Robin 2010). For example, excavations at Carrowmore, Co. Sligo (Burenhult 1980), where most of the chambers were exposed, revealed that many tombs had arcs of stones between the kerb and the chamber. Tomb 27, for instance, had a ‘mantling’ of stones built up within the kerb as well as large stones packed around, but not covering, the chamber. Similar features have been found at tombs with full covering mounds (Eogan 1984a), including the largest tombs like Newgrange (O’Kelly 1982).

24A correlation has been noted between the presence of these internal arcs and the ‘undifferentiated’ passage tombs (Eogan 1984b: 98–9), of which the most striking example is at Townleyhall, Co. Louth. Although this monument was only 13.5 metres in diameter, there were four concentric arcs of stones, with radial settings between the outer two, completely surrounding the chamber and abutting both sides (fig. 4c), with another on the mound’s earthen core, the ground beneath which was paved (Eogan 1963). It may have been that, in passage tombs whose internal structure had an apparently linear form, the laying out of these circular features around the chamber emphasised the monument’s predominant circularity, so counteracting any possible symbolic ambiguity.

25Such arcs and circles of stone are mirrored vertically in the layering of different materials, such as soil, turf and stone, within the body of the mound itself, a feature that is seen most clearly at the large well-preserved tombs, but which was also widely recorded at the Knowth satellite tombs (Eogan 1984b). The overall effect can be seen as further ‘insulating’ the chamber from the outside, so that it was not only separated from the external world by the surrounding kerb, but also by a hidden sequence of concentric internal shells (Robin 2010).

26Some mounds incorporated geologically distinct materials, such as the large quantities of quartz used at the Brú na Bóinne tombs of Newgrange, Knowth and Dowth, and at Loughcrew T (Frazer 1893). Geochemical analysis has indicated that the granite cobbles and quartz deposits at Newgrange were collected from sources in the Cooley Peninsula and the Wicklow Mountains, approximately 50 km and 75 km to the north and south, respectively (Mitchell 1992, Meighan et al. 2002 and 2003). The tombs at Seefin and Baltinglass in the Wicklow Mountains also incorporated quartz in their mounds.

27Passage tomb mounds, therefore, were not simply masses of earth and stone piled up over the chamber, but were constructed in a formalised and symbolically consistent manner. This involved the creation and deposition of a succession of distinct structures and deposits, utilising materials derived from different parts of the landscape, both locally and at some tombs more distant, expressing and reflecting the extensive nature of the social networks mobilised in the tomb’s construction.

Artefacts and art, siting and size

28As in court tombs, those whose cremated remains have been found in passage tombs are unlikely to have been viewed as simply ‘dead and gone’. The provision of continuing access (albeit restricted), as well as allowing subsequent mortuary deposits to be made in the chamber (and material to be removed), would have facilitated the continuing ritual mediation by the dead in the lives of the living. However, at passage tombs the community of the dead appears to have been represented as different in fundamental respects from the living population, and being overtly ‘other-worldly’ in essence.

29For example, they had little need for the pottery forms, flint tools and other artefact types that are found in most settlement contexts. Instead, they were accompanied by a range of objects, some highly formalised, that were appropriate only within passage tombs (and possibly a limited set of other ritual contexts) (Herity 1974). These included distinctive Carrowkeel bowls, stone balls, bone pins, mace-heads pendants and other overtly symbolic artefacts. Moreover, the actual megalithic fabric of many passage tombs was decorated with the abstract motifs of passage tomb art (Shee Twohig 1981, Robin 2008).

30This separation from everyday domestic and economic concerns may be reflected also in the locations of many tombs, sited individually or in groups on hilltops or other prominent positions in the landscape, often unrelated to the local settlement patterns. There is often a correlation between a tomb’s position, particularly if in a group, and its size relative to its neighbours. Passage tombs ranged from under 10 metres in diameter to over 80 metres, although over 80 percent of them were under 25 metres and so comparable in scale to most court tombs. Yet building larger tombs, or remodelling existing monuments, as at Baltinglass (Walshe 1941) (fig. 4d), was clearly another trajectory of tomb elaboration.

More complex than ‘from simple to complex’

31While the structures of passage tombs, therefore, could be elaborated along a number of different trajectories, there was no simple correlation between elaboration in one direction and elaboration in another. For example, although the cruciform chamber at Fourknocks I incorporated an expanded central area and contained some of the finest examples of geometric art from any passage tomb, the monument itself was only 19 metres in diameter. Similarly, while the west tomb at Knowth, approached by the second longest passage at any tomb, consisted of a simple rectangular chamber, Carrowmore 27, which was only 23 metres in diameter and built of rough glacial boulders, had a cruciform chamber. Such examples undermine the frequent assertion that passage tombs (as well as court tombs) developed from ‘simple to complex’ (Scarre et al. 2003).

32Whatever their degree or direction of elaboration, passage tombs expressed in their predominantly circular form a set of symbolic principles that appear to reflect essential characteristics of an ancestral community and represent it as distinctly separate from the living community. It had concerns significantly different from the day-to-day imperatives of social and economic life, and inhabiting an explicitly supernatural domain that could be experienced only through overtly symbolic ritual activities. The decorations of the stones (in some cases hidden) and the formal layering of materials within and under the mound suggests that the building of the monument itself was more than just a practical construction process, and was one in which symbolic and ritual activities were a necessary part of ‘sanctifying’ the very fabric of the monument.

Linkardstown-type cists

33Linkardstown-type cists, named after a burial monument in Co. Carlow (Raftery 1944) and concentrated in the south-east of Ireland (fig. 5), appear to share a number of attributes with passage tombs. They had round mounds, ranging in diameter from 18 metres at Baunogenasraid, Co. Carlow (Raftery 1974), to 35 metres at Norrismount, Co. Wexford (Lucas 1950). The mounds often had a composite, layered construction, with internal arcs or circles of stone beneath them, and Jerpoint West, Co. Kilkenny (Ryan 1973) even had radial settings similar to Townleyhall. The tombs also had stone chambers, some of which had entrance features and were similar in design to a number passage tomb chambers.

Figure 5. Distribution of passage tombs and Linkardstown-type cists

34An important difference, however, is that these chambers were sealed, the entrance at Baunogenasraid, for example, being blocked by a limestone slab (the rest of the chamber being constructed of granite boulders), and the front of the Ashleypark chamber, Co. Tipperary (Manning 1985), being filled with stones. Once the covering mound had been built, therefore, these monuments were effectively closed, preventing further burial deposits and limiting the scope and type of ritual interactions between the living and the dead. Many of the chambers had paved floors, walls of overlapping inclined slabs, and further overlapping slabs for the roofs, indicating that the sealed chamber at the core of the mound marked the primary boundary, physical and symbolic, around the realm of the dead; significantly, unlike passage tombs, these monuments had no formal kerb.

35The burials too were different, containing not the cremated remains of people now subsumed within a collective ancestral community of the dead (as suggested for both court tombs and passage tombs), but the articulated inhumations of usually single adult male – ‘remembered individuals’ (Scarre 2007: 24) who in death retained important aspects of their personal identities. Part of their status was expressed in the quality of the finely decorated necked bowls which frequently accompanied them (Herity 1982) – as well as in their selection for burial in this limited set of monuments.

36However, although the distinctive bowls found in these tombs had only limited distributions in other contexts, the tombs also contained vessels, as well as bone and flint artefacts, of types found in settlements and other non-mortuary contexts. While the individual’s status may not have been derived from descent, therefore, neither does it seem to have been derived from access to a restricted body of symbolic knowledge seen as distinct from everyday life. Instead, it may have been based on some attribute of the individual’s personal achievement, such as, possibly, the control of aspects of economic production and exchange.

Tombs and mortuary ritual in a segmentary society

37In 1972, Fleming suggested that the segmentation visible in the structure of court tombs and passage tombs was ‘a fossilised record of the fission and fusion principle on which ... many primitive societies are organised’ (Fleming 1972: 62). Each tomb form, he argued, represented a different design solution to problems imposed by the structure of Irish Neolithic society, in particular to those ‘segmentary lineage systems’ (as described by Sahlins 1961) which developed with the long-term occupation of territory. The role of megalithic tombs as territorial markers within small-scale segmentary societies, and the suggestion that they displayed in their form some of the characteristics of such societies, was further developed by Renfrew (Renfrew 1976).

38Central to the Darvill’s distinction between the ‘segmentary’ and ‘chiefdom’ societies, as they related specifically to Irish court tombs and passage tombs, was his characterisation (following Renfrew) of a segmentary society as one in which ‘no hierarchical structure can be recognised’ (Darvill 1979: 318). Such a model suggested a horizontal arrangement of autonomous and equal social units that interacted communally only at periodic gatherings.

39The ‘segments’ of such a society are not only those lineage groups of equivalent scale and status forming local communities, but also those more extensive descent groupings which are clustered in a hierarchy of increasingly inclusive, but ever more remote, genealogical categories (Keesing 1975: 30). Any individual, therefore, might belong to segments at different levels in society, and so recognise membership of more than one such ideological grouping.

40It is suggested here that court tombs and the passage tombs represent social categories at different levels within such a segmentary, lineage-based society, the relative importance of which changed over time. In the context of court tombs, social status appears to have derived largely from the control of relations of production directly associated with the settlement and economic exploitation of the land. In time, however, as wider kinship networks developed through social transactions undertaken at a higher segmentary level, social status may have come to be mediated through other forms of ritual and symbolic transaction.

41The transfer of, and succession to status, however, appears to have remained firmly rooted within the mortuary context, although now expressed through the construction and use of a new and distinctive form of monument, the passage tomb. These tombs, which were not only distinctively and perhaps deliberately different to court tombs, but which also articulated a set of symbolic principle that were in opposition to those which underlay court tomb construction, served the social and ritual needs of a broader network of inter-group relations through the production and exchange of novel materials and resources, some overtly symbolic in character.

42While the wide variability of passage tomb size and elaboration should warn against any narrow social interpretation, the symbolic principles suggested above as underlying their form and structure find some parallels in the ideological structures of a high-level social category such as a ‘clan’. A clan, like a lineage, is defined by descent, but it lacks any clear genealogical connections with the apical founder from whom descent is claimed; such founders can be so remote in time that they are frequently invested with mythic, totemic and expressly supernatural attributes (Keesing 1975).

43Clan affiliation is dependent, not on the sorts of practical negotiation between local descent groups that might confer rights within a small-scale territorial corporation, as represented by court tombs, but instead is perceived as categorically exclusive, and expressed through essentially symbolic insignia. The forms of ritual mediation through which such clan relations may be legitimised, therefore, employ appropriately overt symbolic forms, such as those found in passage tomb art and artefacts. The concerns of a clan are also different from those of a local corporate descent group, with clan transactions being often restricted to specific social gatherings and religious festivals, bringing together members from local communities spread over a wide area.

44The particular nature of the passage tomb symbolism (as suggested here), discernible in the structure of even the simplest tombs as well as in the associated artefacts and in the motifs of passage tomb art, emphasised fundamental differences between the suggested ‘clan’ affiliations and the local lineage-based social relations. Indeed, passage tomb use appears to have had certain characteristics of a ‘cult’, whose rituals made symbolic references, not to the domestic and economic concerns of everyday life, but to more elevated and esoteric areas of social and religious discourse.

45The rise to social predominance of such a clan cult could have developed over a long period. The more extensive, higher-level social groupings may have become an increasingly important force for social cohesion within the wider society, by maintaining ties of kinship between people otherwise dispersed in local communities across the landscape, by facilitating economic and social transactions between lineage segments in different settlements, and by providing contexts and mechanisms for those interactions that may have been inappropriate or unviable locally. In a segmentary society it is often the clan that defines the outer limits of exogamy, with marriage having to take place outside it; such a requirement would make local descent groups interdependent, and tie them together into the wider society by bonds of mutual kinship obligation. Over time, it is suggested, these wider clan networks may have come to dominate social and economic exchange.

Cemeteries, scale and status

46One of the clearest contrasts between court tombs and passage tombs is their very different distributions within the landscape, and within Ireland itself. While there are significant variations in the density of court tombs, with a small number being very closely positioned, there are no court tomb ‘cemeteries’ (for want of a better term) in the manner of the distinct clustering found with passage tombs. Many passage tombs were clearly sited with reference to existing passage tombs (Cooney 1990), the establishment and development of cemeteries being the inevitable consequence of this process. At Loughcrew, for example, four of the thirteen extant tombs with surviving passages are orientated on other tombs, all of which are at higher elevations within the cemetery (Prendegast 2011).

47Whereas a court tomb expressed in its form the limits (and component parts) of the corporate descent group it represented, a passage tomb in a cemetery expressed not only an individual identity but also the symbolic affiliation with a larger grouping, associated with a special, often physically elevated location, possibly associated with its perceived mythical origins. The development of the largest cemeteries suggests that particular social, economic and demographic circumstances led to certain socially dominant groupings, with widespread memberships, competing to outdo each other in their claims of access to authentic ancestral powers through which, perhaps, other forms of status and wealth could be legitimised.

48The four large passage tomb cemeteries (Brú na Bóinne, Loughcrew, Carrowkeel and Carrowmore), as well as some of the smaller ones, appear to be strung out in a line between the Boyne Valley and Sligo Bay (fig. 5). As well as marking the northern edge of the poorly drained central lowlands, this line correlates closely to the southern limit of the main distribution of court and portal tombs (with the exception of those in North Connaught, west of Sligo) (Powell 2005, fig. 3), suggesting that the cemeteries developed at a topographic, demographic and social-structural interface. The region to the north of this line, where the density of court tombs suggests a population which, although competing for land and resources, was bound together by close ties of kinship and affiliation, contrasts with the region to the south, where there were far fewer contemporary megalithic monuments, and where there was the potential to develop other forms of social structure, perhaps represented in part, for instance, by the Linkardstown-type cists.

49However, the scale of the elaboration found at a small number of tombs suggests that some social groups sought and achieved socio-political primacy, perhaps legitimising their dominance by claiming to have re-established, through the performance of increasingly elaborate ritual, and the expression of an ever more esoteric religious discourse, a genealogical link back to their clan’s mythical founder, and so to being the true inheritors the clan’s authority. The authenticity of such a claim could be demonstrated not only by the scale of the social and economic resources mobilised for the monument’s construction, but also by the symbolic resources that were expressed in the mysterious art, the cosmological references in some of the tomb’s alignments (Patrick 1974), and in ever more precise and restricted bodies of knowledge as at Newgrange (Powell 1994).

Conclusion

50The aim of this paper has been to consider one possible explanation for patterns of regularity and variability that are evident in the design of Irish megalithic tombs, and to consider the possible social mechanisms which saw court tombs supplanted by passage tombs in the later part of the 4th millennium BC. Rather than resorting to an invasion hypothesis to explain the new form of monument and its associated artefacts and symbolic art, it may be more profitable to seek the roots of passage tomb construction and use in the structure of the society that preceded it, and that society’s symbolically loaded tradition of megalithic tomb construction. The model outlined above does no more than suggests one possibility.