Navan Fort - Emhain Macha of the Ulaidh, The Men of Ulster

Articles by C.J. (Chris) Lynn
Articles by J.P. Mallory
Articles by Ann Hamlin
Articles by Helen Steele
Articles by The Northern Ireland Environment Agency (NIEA)
Video of the Figure of Eight Structure
'Time Team' (English television) video about the Navan Fort area (YouTube)

Knockaulin - Dún Ailinne of the Laighin

Articles by Gerald A. John Kelly
Articles by Bernard Wailes

Rathcroghan - Cruachain of the Connachta, the Féini in Connacht and the O'Neills in Ulser

Articles by John Waddell

Tara - Teamhair na Rí, sacred to all three kinships

Articles by Edel Bhreathnach

The Irish Royal Sites
From “Revisiting the Irish Royal Sites”
Susan A. Johnston

Research emerging at a steady pace in recent years has demonstrated that the historical texts which grouped the “royal sites” of prehistoric Ireland – Knockaulin (Dún Ailinne), Tara, Navan Fort (Emain Macha), and Rathcroghan (Cruachain) – did so for good reason. While it is interesting to speculate how long it would have taken archaeologists to define this group without the historical record, it is becoming clear that these sites share features that would have placed them in a single class even in the absence of that record.

The Archaeology of Royal Sites

Despite recent research, the data from the royal sites is somewhat variable, a situation which is rapidly improving. Navan, Co. Armagh, would have to be the site with the most published research. Dún Ailinne, Co. Kildare, has also been excavated, and the excavation report is currently in preparation. Tara, Co. Meath, has had a more uneven record; while the early excavations carried out by Ó Ríordáin remain unpublished, more recent work has been described (Roche 1999, 2002), providing interesting new data on activity there. In addition, extensive archaeological and geophysical surveys have been carried out at Tara. Rathcroghan is the only royal site to elude large-scale excavation, though extensive geophysical  and archaeological surveys have been ongoing.

This impressive amount of data allows a discussion of what the Irish royal sites have in common, as well as how they differ. The first aspect to consider is chronology. What evidence do we have that the sites are indeed contemporary and thus provide a valid comparison? To begin with Dún Ailinne, its chronology is composed of both artifact evidence and a C14 sequence, as well as a single TL date. Taken together, the list of dates, comprised of some 21 age-determinations, contributes perhaps six which can be considered as reliable for the main phases of Iron Age activity. Calibrated, the C14 dates range from the final century and a half BC through the first few centuries AD. In addition to providing evidence for Neolithic and Bronze Age activity, the datable artifacts correspond reasonably well with the Iron Age C14 dates.

Navan Fort is broadly similar in its periods of activity. Apart from the scattering of Neolithic material, the majority of the artifacts date from the later Bronze Age and Iron Age, with the larger number in the latter period. Altogether, they indicate the bulk of activity at the site from the 3rd-1st centuries BC. Navan has a significant C14 sequence from Sites A, B, and C. Summarizing the data for Site B, Phase 3 overall was considered no earlier than the 4th century BC and continued into the 1st century BC, while Phase 4 (the 40m-structure) could be assigned to the latter half of the 2nd century BC. Site A had a rather broader range; combining both slot trench and ditch data, the various phases of site construction covered the 4th century BC possibly through the 5th century AD.

Four C14 dates from Site C, attached to Site A in a figure-of-eight pattern, range from 400 cal BC to AD 26. In addition, Navan has its gratifyingly specific dendrochronological dates. The latter came from the central post of the 40m structure (94/95 BC) and from a timber lying at the base of the enclosure ditch (94 + 9 BC).

There is comparatively less dating evidence available for Tara and Rathcroghan. The more recent excavations at Tara produced both artifacts and C14 dates; the former indicated a broadly Iron Age date for both the metalworking activity pre-dating the bank of the Ráith na Ríg and the lower levels of the enclosure ditch, while the latter produced a series ranging from 370 BC – AD 406 for the metalworking furnace. Being largely unexcavated, dates for Rathcroghan are not yet numerous. These were somewhat inconsistent, but ranged from the last two centuries BC to the 2nd-3rd centuries AD.

While the historical record consistently groups these sites together, it has also been noted for some time that the royal sites share a series of general archaeological characteristics. A number of authors have described these aspects. All are located on hilltops, with impressive views to the surrounding countryside. Within that countryside, all four lie in larger archaeological landscapes, which may be characterized by a high proportion of ritual monuments. The remarkable archaeological and geophysical survey of Tara carried out by Newman lists almost 100 monuments in its catalogue, while an archaeological survey in the area of Rathcroghan lists 53. There are a number of major monuments noted in the vicinity of Navan, including Haughey’s Fort, the King’s Stables, Loughnashade, and the Dorsey, as well as many other sites and finds in the larger area. While the dates and chronological sequence of these various monuments await further study, it does seem that in all four cases, the royal site was placed in a landscape already populated by markers of ritual significance.

 As for the site components themselves, first and foremost, all have enclosures surrounding the main area of activity. The enclosures (at three of the sites; the nature of the enclosure at Rathcroghan is still unknown) are composed of an outer bank and an internal ditch, something which the royal sites share with other ceremonial monuments such as henges, and which differentiates them from more obviously defensive structures, like hillforts. The significance of the inner ditch has been the subject of speculation. It has generally been argued that an inner ditch is not as effective in defense as an outer one, since attackers would be able to surmount the bank more quickly and then have the advantage of height. Whether ceremonial or defensive, the fact remains that this is a somewhat unusual feature in enclosed sites overall, and the three excavated royal sites share it.

 Further excavation and geophysical survey has demonstrated that the similarities at a more detailed level are even greater than suspected. At this level, however, the notion of “similarity” becomes more complex. While the overall impression of the components of the sites is very similar, they actually vary quite a bit in detail. In some cases, for example, features might be shared by only some of the sites while in others, a shared general feature might vary in detail. This is a significant point, because it is becoming possible to define, on the one hand, what the necessary components of a “royal site” might have been, and on the other, where it was possible to tailor the site to local needs and/or expectations. This is of course further compounded by the fact that Rathcroghan has not been excavated, and so for the most part we don’t know exactly what is there. Remote sensing has helped this situation considerably, but Rathcroghan remains largely unknown. As such, any conclusions about “royal sites” as a class must remain provisional, with Tara, Navan, and Dún Ailinne providing the bulk of the available data.

 Moving within the enclosure, then, the first characteristic that all the sites appear to share is the presence of a large, circular, timber structure. Dún Ailinne had three successive structures, each composed of palisades with timber posts bedded in slot trenches, and each having been dismantled before the next began. The first (White) was the simplest, and was succeeded by more complex palisade arrangements (Rose and Mauve phases). At Navan, the actual number of structures represented is somewhat complicated, depending on how the various slot-trench buildings are understood and counted. In any case, there was a series of timber buildings comprising both Site A, phase A, and Phase 3 of Site B, as well as the 40m-structure, short-lived though it may have been. At Tara, an enclosure comprised of two concentric circles of pits with a ditch in between was revealed in a recent geophysical. In the absence of excavation, it cannot be stated for certain that these pits held posts, but it seems likely. The situation at Rathcroghan needs further evaluation, but again, recent geophysical work has provided evidence of a number of circular features which may be the remains of timber structures. These include the large enclosure around the main mound and also several circular features within it.

 A second shared feature is rather more nebulous, but there also seems to be a similar pattern of what have now been dubbed in the literature “figure-of-eight” structures. Lynn has already drawn a detailed comparison between the figure-of-eights at Dún Ailinne and Navan, and in his figure the resemblance is really quite remarkable. Both have the larger enclosure to the north and the smaller to the south; both enclosures at both sites have three slot trenches marking their outer circumference (Warner 1994c); at each site, both enclosures have separate entrances as well as an entrance between the larger and smaller sections; and the entrance into the larger enclosure in both cases is marked by a funnel-shaped feature. A second figure-ofeight structure at Navan has been uncovered which also appears to share the majority of these features.

 While no comparable figure-of-eight structure has yet been found at Tara, it is worth noting that the two enclosures within the Ráith na Ríg, while structurally quite different from those just noted, still form a similar pattern overall. No similar structure has yet been found at Rathcroghan, but the presence of the figure-of-eight pattern, worked out in different ways at the other three sites, suggests that this held some significance in the larger category of royal sites.


 Considering all the component features of the royal sites, these are really the only ones they share – aspects of siting, the configuration of the enclosure, the presence at some point in the sequence of at least one timber structure, and the arrangement (in at least some periods of site use) of structures in a figure-of-eight pattern. Beyond this, the architectural arrangements of such features as timber and figure-of-eight structures vary considerably. The differences between the various figure-of-eight structures have already been noted. In addition, the placement of the timber structures is quite different relative to the enclosure. The post-pit ring at Tara is not within the enclosure of the Ráith na Ríg, though it does overlap it to a significant degree. While there are a variety of features within the enclosures at Tara, Rathcroghan, and Navan, at Dún Ailinne there appears to have been only one at any given time. Tara and Navan both have a central mound within their enclosures, though Navan’s covers the burned remains of the 40m timber structure. Dún Ailinne lacks such a mound, but does have a central structure within the Mauve phase complex which may have been functionally similar but stylistically different.

The figure-of-eight structures at Navan produced occupation material, prompting the excavator to suggest originally that it had been inhabited. No such material was recovered from Dún Ailinne.

There is also a great disparity in size among the various sites, as shown in Table 1 below.

The most obvious difference is the truly massive timber structure at Tara, five times larger than the next largest at Dún Ailinne. The timber structures at Navan and Dún Ailinne are closely comparable. For the other structures, Dún Ailinne is the largest, though not always by much. The Rose phase figure-of-eight structure and the diameter and area of Dún Ailinne’s enclosure are each about 20% larger than the next largest examples (respectively Navan Site C, Tara, and Rathcroghan, assuming the estimate for the latter is reasonably correct). What the size differences mean is unclear. Obviously larger structures are more imposing, more visible, attest to the greater power of their builders, etc. It might also have something to do with the number of people available to work on construction. In addition, it might be suggested that the larger size was deemed necessary to accommodate the activities held within, whether it be the enclosure itself or the structures within it. Any or all of these could have contributed to this aspect, with the resulting significance in terms of interpretation.

The meaning of difference?

 While there is no direct source of information as to what did actually take place at these sites, some clues are available which suggest that this also was varied. As far as is known, while the periodic dismantling of structures and the construction of new ones happened at all four sites, the dramatic destruction of the 40m structure at Navan is unique. Similarly, the recently discovered evidence of industrial activity at Tara, in the form of evidence for iron- and bronze-smithing and possible manufacture of glass objects (Roche 2002) is not directly paralleled, though there is evidence of low-level manufacture in a variety of materials at Dún Ailinne (Wailes and Johnston, Forthcoming). This is turn suggests the likelihood that the rituals associated with the royal sites were also variable, though they may well have contained some core rite(s) shared by all. This is certainly suggested by the fact that the sites are considered equivalent in historical texts. Their grouping was presumably not based entirely on the fact that they looked the same but also that they inhabited the same cultural or social context based on their use.

 The exact meaning of this variety within similarity is probably beyond recovery, but it does raise the obvious question – what would an Iron Age person have to have seen on one of these hills to recognize it as a “royal site”, or at least as similar to the site with which that person was familiar? This comparison suggests several observations by way of answer. Assuming that the sites were indeed contemporary, a royal site had, at a minimum, a large hilltop enclosure with an exterior bank and an interior ditch. It is interesting in this regard that the presence of an enclosure (regardless of shape) has been identified as one of the main defining characteristics of sacred sites in the larger Celtic world. This is based on both archaeological (Brunaux 1987; Wait 1985) and classical documentary (Webster 1995) evidence. Thus, if there is in any sense a “Celtic world”, the royal sites clearly form the basis of its northwestern extension.

 Within this enclosure there would have been a timber structure, which at times might have stood alone and at times might have assumed a figure-of-eight form. At Navan and Dún Ailinne, the figure-of-eight structure would have been the earlier, replaced by a later, single circular structure (though the one at Navan was shortlived), while at Tara, the reverse was true – the circular timber structure might have been replaced by others in a figure-of-eight form. In fact, the variety of timber structure arrangements raises the possibility that what is perceived today as a single (though variable) class (i.e., timber structure) might have been seen as sharing nothing of significance by their builders and users. Alternatively, it may have been something even more nebulous that was considered significant about the timber structures, e.g., their overall circularity or the orientation of their entrances. Indeed, given this variability, it may be argued that, for the royal sites, it was the enclosure that formed the primary defining characteristic for the contemporary population. Since the enclosure would have been the first aspect of the site that a person would have encountered (first visually then physically), this makes some sense. The characteristics of the enclosure invoked whatever thoughts, behaviours, feelings, a person was supposed to have on their arrival at the site, setting the stage in a sense for what happened within it.

 Further elaborations among the royal sites must be assumed to be on the level of local preference. Whether the structure was roofed or not, walled or not, large or small, may have been more in the way of a regional pattern, perhaps having to do with local religious beliefs, local histories, stylistic fashion, political whims, or some combination of all of these. In this sense, then, while the concept of “royal site” may reflect some kind of ideology shared through the island, the elaborations within that class spoke to what made an inhabitant of, say, Leinster different from one of Ulster. This might have had further resonance if, as the historical documents imply, the royal sites had something to do with the construction and maintenance of political power. The relevance of those documents to the prehistoric past is of course unclear (though I suspect that the real situation is somewhere between totally irrelevant and a direct window on the past – perhaps a “kaleidoscope on the Iron Age”, to coin a phrase?). Nevertheless, even without those documents, a number of factors – e.g., the archaeological evidence for the presence of elites, the location of the royal sites on hilltops, the size of the monuments, the large number of people who must have been involved in their construction, the possibility that they were designed for assembly – all might have been used to argue this function in the absence of supporting texts. If they were the sites of political rituals, e.g., inauguration, then it would be expected that both levels of meaning – the ones shared throughout Ireland and the ones that worked on the local level – would be symbolized in the places where those rituals were carried out. Thus a royal site was one where a king was made, while, say, Dún Ailinne was the only place where a king of Leinster could be made.

 Speculating further, this could be interpreted in light of what Lynn has argued recently (2003) about the human interaction with Navan Fort (e.g., the directions participants would have walked when entering the 40mstructure, or the appearance of the mound as someone approached the enclosure uphill). If the enclosure was the broader, basic, generalized marker of a royal site, which anyone could encounter from the outside, then the more specialized configuration of the structures within was physically encountered only by the actual participants. Thus as one moved from outside to inside, one also literally encountered the physical manifestations of those levels of meaning in the correct order. One could speculate further about the kinds of meaning experienced by those who moved even further within the existing structures, until, perhaps, only one person entered the centre, encountering the most deeply meaningful set of symbols. But then perhaps that is one speculation too many.

Aerial view of the Emain Macha area.


 All speculation aside, as research continues, the definition of a royal site can be gradually refined. Questions which remain to be answered include the characteristics of Rathcroghan, the nature of the apparent timber structure at Tara, and how the idea of a royal site was manifested in a local idiom at these places. In addition, sites like Raffin Fort (Newman 1993a, 1993b, 1995) in Co. Meath, raise the possibility of levels below the royal sites in the hierarchy – smaller sites representing the subordinate polities of, say, client kings (Wailes 1982). As new data emerge, we have greater cause to argue that, even without the historical documents, we would have been able to understand these sites as representing something important: an ideology shared within Iron Age Ireland that could also reflect the symbols meaningful to the communities who built, used, and ultimately abandoned the Irish royal sites.