Of Ireland's 1,500 megalithic tombs, passage tombs are our most extravagant monuments. They seem to generate superlatives: biggest, most spectacular and, inevitably perhaps, oldest. And now bold claims are emanating again from Carrowmore in Co Sligo, where archaeologist Goran Burenhult believes he has found some of the earliest megalithic tombs in Europe.
About 1,500 megalithic tombs are known in Ireland. They break into four basic classes: court tombs, portal tombs, passage tombs and wedge tombs. The distribution of each group around the countryside represents the spread of a particular community in the Irish neolithic period. In approximately the same way, moates, for example, stand as reminders of the Norman invasion.
Passage tombs often occur in clusters known loosely as cemeteries. The four largest cemeteries are the Boyne Valley and Loughcrew in Co Meath, and Carrowkeel and Carrowmore in Co Sligo.
In the late 1970s Goran Burenhult from Sweden began excavating at Carrowmore. It was not long before claims of dramatically early dates began to appear. Two significant themes ran through the claims: (1) the radiocarbon determinations from Carrowmore, when calibrated towards real dates, represented a building phase about 1,000 years older than the Boyne Valley; (2) the earliest megaliths at Carrowmore may have been built by communities whose subsistence was based largely on fishing and hunting, for which evidence exists in the region.
The Carrowmore claims were startling; they seemed to stretch the archaeological evidence, and they needed to be tested. Analysis of the dates has concentrated on the crucial link between the charcoal samples and the construction of the tombs. Not everyone is happy that the link is watertight, and there is a general tendency to treat the Carrowmore dates with caution, which does not mean that the dates are necessarily wrong. As for the second part of Prof Burenhult's theory, sceptics have pointed out that no pre-neolithic artefacts are known from the Carrowmore tombs.
The case rested ultimately on the dates. A radiocarbon date is produced by taking a sample, usually charcoal, from an archaeological site and processing it in a specialised laboratory. At its best the date produced in the laboratory provides a fixed point in the chronological profile of the particular site. Critical to this exercise is the precise context of the sample. It is virtually impossible in many cases to guarantee absolutely that a particular sample dates the layer in which it occurs. Significant amounts of material can be dragged from one layer to another by the activity of rodents, worms and tree roots. For this reason excavation teams keep a record charting the purity of various layers, and the risk of intrusion from other layers.
As a result archaeologists are very cautious in the importance they give to any one radiocarbon date. They prefer to operate in a situation where a consistent pattern emerges from a series of dates, and they also like to match the radiocarbon evidence with the other archaeological data.
In the Boyne Valley, especially at Knowth, the sequence of neolithic activity is now clearer than it was 20 years ago. The Carrowmore investigations represent the best chance of a parallel sequence against which to compare the Boyne Valley chronology. It would be most informative to know whether the Carrowmore and Boyne Valley cemeteries arose in the same way.
Both the Court Tomb and the Portal Tomb seem to have appeared early in the fourth millennium B.C. There is no earth-shattering reason why Passage Tombs would not have begun to appear in the landscape at the same time or even beforehand. In many ways Irish Passage Tombs are so obviously different from their mainland European cousins that a reasonable span of time has to be allowed for their separate evolution. The evidence from the Boyne Valley, however, suggests that they appeared at a relatively late date in the neolithic. There is no conflicting evidence apart from the contested Carrowmore dates.
In recent years Goran Burenhult has been back at Carrowmore. All those with an interest in the topic are hoping that his team can find ways of extracting charcoal samples from unambiguous layers, so that a comprehensive sequence of activity in the Carrowmore area can be established.
That said, however, it is important to emphasise that archaeological excavation is not primarily a search for dates. The Carrowmore excavations are part of a general investigation into the early prehistory of the surrounding region. They also fit into a wider study of Irish and European prehistory, spanning the later stone age and the transition from the stone age to the early bronze age. In this more important sense the benefits of Prof Burenhult's work at Carrowmore are immeasurable.
Archaeological reconstruction is a fluid and inexact process. It is dominated by prevailing images of society and current models of explanation. At one time the favoured explanation for the appearance of passage tombs and court tomes in Ireland assumed that the people responsible for them originated somewhere else in Europe and arrived at a specific point on the Irish coast, the tradition then flourishing in its purest original form near the point of entry, and going slightly native as it diffused into the hinterland.
This colonising model seemed particularly logical if one accepted that the pre-neolithic landscape was more or less uninhabited.
During the past quarter of a century, however, accumulating evidence has shown that settlement in the Irish neolithic was much more complex than the waves-of-immigrants model could explain. Nowadays the emergence of the megalithic tradition in Ireland has to be set against a network of settlement that took several thousand years to evolve.
The image of a virgin landscape with a few hunter/gatherers eking out a living at the margins seems like a distortion today. Deconstructing the older model, however, has proved much easier than producing an alternative. Any new model has to allow for a more complex set of relationships among and between communities in the era 5000-3000 BC. There is now a greater interest in exploring the mesolithic/neolithic transition and the emergence of contrasts between regions and social entities during the neolithic. The first efforts to farm the Irish landscape are now thought to have occurred in the fifth millennium BC, several centuries ahead of previous indications. The complex transition from a preneolithic to a neolithic situation is better understood. There are hints that the earliest farming communities lived side by side with pre-farming settlements. Neolithic Ireland was not homogeneous. The world represented by Ceide Fields is different from that in the Boyne Valley.
Dr Muiris O'Sullivan lectures in the Department of Archaeology, UCD