Wednesday, June 25, 2014

Review | Rewriting the (Pre) history of Ulster: A synthesis of developer led excavation, monuments and earthworks 4300 to 1900 BC | Dr Rowan McLaughlin

[** If you like this post, please make a donation to the IR&DD project using the secure button at the right. If you think it is interesting or useful, please re-share via Facebook, Google+, Twitter etc. To help keep the site in operation, please use the amazon search portal at the right - each purchase earns a small amount of advertising revenue **]

On Friday the 6th of June 2014, I wandered along to the Pat Collins Reading Room at Waterman House, Hill St., Belfast to listen to the magnificent Dr Rowan McLaughlin speak about prehistory in Ulster. Specifically, he was intent on tackling the impact that data from excavations in the last decade-and-a-half have had on our understanding of prehistory in Ulster and Ireland generally. The MRB have been running a pretty excellent lecture series over the last while and have a full schedule of speakers lined up until the end of 2014 (here). I’ve not been to any of these before, but I felt that I wanted to make a special effort for this speaker. Many years ago Rowan and I used to work together in the commercial field archaeology sector. While he’d often be relatively quiet in the site hut, he was certainly worth listening to once he started to speak. If he was giving a lecture, it would – most likely – be well worth the attention paid. I’ve not seen him in a few years and it was lovely to catch up in the few minutes before the lecture kicked off. I was, however, rather discomfited when one of his first remarks to me was: ‘are you going to write this up for your blog?’ Staying my hand from reaching for my notepad and pencil, I attempted to appear non-committal as I asked ‘would you like me to?’ He appeared to think about it for a moment and said ‘well … yes’. Everything hinged on that ‘yes’ and the post below is my attempt to convey Rowan’s research in a clear an intelligible manner. If I’ve misunderstood or erred in any way from the original delivery, I apologise to both the lecturer and you the reader.

An image of Saturn from the Hubble telescope (Source)
Rowan did his PhD on dental microwear analysis and has worked at Queen’s University Belfast as part of the INSTAR Cultivating Societies project. He began with the deceptively simple statement that he wanted to talk about prehistory and how we go about turning all the data generated through the excavation of archaeological sites and the pattern of sites in the landscape into knowledge about past societies. Without any written records from these times, archaeology is the only means of enquiry we have. With this in mind, it did seem a little odd when his first choice of image was the planet Saturn. As a slightly nervous ripple of laughter from the audience subsided, he explained that the planet had absolutely nothing to do with Irish prehistory, but that it does neatly illustrate some ideas about how information about distant things is acquired and how that data is transformed into knowledge. Although he started with a recent, detailed image of Saturn taken from the Hubble Space Telescope, he noted that while the ancient Greeks knew about the existence of various planets in our solar system, they understood them solely as star-like objects that shifted position and ‘wandered’ in relation to the other stars. In 1610, when Galileo Galilei pointed his telescope at the night sky to observe Saturn he was unsure of what he saw and was unable to rationalise and comprehend its rings. The planets he knew – Earth, the Moon, Jupiter etc. – did not have rings around them, so nothing in his experience could have prepared him for what he was observing. He incorrectly identified the rings as close-orbiting moons. To make the matter more complex, the plane of the rings changes, so when Galileo examined Saturn a year later, the ‘moons’ appeared to have disappeared … and then reappeared sometime later. Galileo, unable to comprehend what he saw, described it as a mysterious planet that appeared to grow and shrink over time. In 1655, Huygens came along with a slightly better telescope and more systematic analysis of the heavens. He was able to determine that Saturn does appear to be surrounded by ring-like material. These advances were again built on by Cassini in 1676 – again with a slightly better telescope – was able to confirm the existence of the rings and even map out various features of those rings, including what is now termed the ‘Cassini Division’ between the inner and outer bands (have a look here for a good run down on the subject). McLaughlin’s point here is that with better resolution – when we have better instruments to measure these remote things – we discover more information, and produce more date, and that this becomes ever more refined and better resolved. There comes a point along this process, where enough information has been accrued to allow you the opportunity to understand the reality of the situation. He points out that, given the relatively modest increases in telescope quality, Galileo could have understood the rings of Saturn had he the correct preconceived notion as to what he was looking at. As McLaughlin says: ‘so it is with planetary astronomy, and so it is with archaeology.’ We cannot directly view the past – in the same way as the naked eye cannot see surface details on distant worlds. But through archaeological research and study – especially of chronological data – you can make observations of this distant time. At this point McLaughlin provided a brief primer on radiocarbon dating and calibration. I don’t propose to repeat it verbatim here, but instead will direct the interested reader to some of my own writings on the topic (which, in any event, were heavily influenced by numerous conversations and correspondence with Rowan: here | here | here | here | here). To illustrate the gulf between the precision that may be gained from radiocarbon dating and the archaeological record itself, McLaughlin put up a slide with a calibrated radiocarbon date on one side and an image of a partially-excavated Bronze Age cist grave on the other. He explained that the individual pictured died and was buried. No doubt, they were mourned and seen as a loss to their community. In all likelihood, this happened within a relatively short period – more than days, but certainly within the scale of weeks, months and a small number of years. And slowly, slowly, the evidence and memory of that person began to fade away. When we come to radiocarbon date this individual, however, the best we can say is that this individual lived and died (in this particular instance) between 2100 and 1900 cal BC. His argument here is that this is pretty good – if you know nothing about Irish prehistory. On the other hand, it’s very bad if you want to put what happened in the past into precisely the right order, so that an understanding is gained of how societies changed over time.

Saturn as visualised by Galileo (top),
Huygens (middle),
and Cassini (bottom)
Over the last decade-and-a-half or so has seen a vast increase in the volume of archaeological data that has been produced. Much of this is the result of economic circumstances, where development-led excavations have been carried out in advance of construction. Many of these were carried out during the Celtic Tiger years, in advance of major infrastructural works such as roads, pipelines, quarries, and residential developments. In a perfect world, the excavations get written up and get stored in archives as ‘grey literature’ ... some even get published. McLaughlin estimates that over the last ten years alone some one million pages of new data for the island of Ireland have been written down. Rowan’s approach to the ‘data mountain’ has been centred on extracting an understanding of chronology from this mass of data, a task he describes as ‘the golden cord to lead out of the labyrinth’. An examination of all the available radiocarbon dates for Irish prehistory that were available in around 2001, shows that there were 1396 known from both published and grey literature sources. Plotting all of these out gives an indication of what prehistory is ‘like’ at this point. As may be expected, there are relatively few sites dating to the earliest periods and more sites from more recent times and a slow but gradual increase in between. This has led to a view of the prehistory of Ireland where there was an initial colonisation during the Mesolithic (c.10k cal BP) and that this was a relatively stable hunter-gatherer way of life until the introduction of agriculture (c.6k cal BP). After this point we see a population explosion that goes hand in hand with increasing social and religious complexities. From this point on we can witness communities evolving and adapting these beliefs and practices, until it reaches its final developed flourishing of civilization in more recent times. As he says: ‘The problem with that view is that it is entirely wrong. It’s not what the archaeological data actually indicate.’ He sees that this discovery has been the big achievement of development led/commercial archaeology in Ireland since the millennium. He then turned to another histogram of 4928 dates from prehistoric Ireland that have become available in the time since 2001.Instead of a gradual increase, there are peaks and troughs in activity. At some times it appears that there were significant episodes of large-scale archaeological deposition, and this contrasts with periods of seemingly little activity. To understand the reality of what we’re seeing here, there are a number of ideas that must be kept in mind. Firstly, all the dates must be calibrated as we cannot directly compare this archaeological data with environmental evidence from various regions. The other issue is the degree of bias in how these data points were collected. Obviously, the first bias is where excavations take place – either dictated by individuals’ research interests, or where development is planned. Further biases exist in the systematic approach that archaeologists use in the collection of this data. For example, there are certain types of features that are more likely to be dated over others – what McLaughlin describes as features that are ‘more juicy looking’ and, thus, more likely to be dated. Indeed, certain types of sites are almost completely ignored and this is an ongoing issue.

Creggandevesky court tomb, County Tyrone
© Ken Willaims/ShadowsandStone.com (
Source)
The Neolithic is an enduring focus of world prehistory. It is the period of human existence where we abandoned the hunter-gatherer approach and took on – to a greater or lesser extent – agriculture, along with various forms of religious beliefs and structures of social organisation that led to this proliferation of archaeological remains. He showed a slide of Creggandevesky court tomb in County Tyrone, arguably one of the best examples of a Neolithic monument in Ireland. During the Neolithic there is evidence for a significant number of these sites being constructed across the northern half of Ireland. At the same time, other forms of megalithic tomb were being constructed, including Portal Tombs (e.g. Legananny, Co. Down) and at least two of the earliest Passage tombs were constructed during the Early Neolithic (Carrowmore, Co. Sligo and Baltinglass, Co. Wicklow). However, the real dividend of all the development-led excavation over the last couple of decades has been the proliferation of newly discovered settlement sites from this period. Prehistoric settlement has, until recently, been difficult to find, leaving our understanding of prehistoric lives and societies incomplete. For McLaughlin, though he does admit to being biased, the single most important question in Irish archaeology is currently: ‘when did the Neolithic period start?’ The ‘orthodox’ date that’s found in most of the text books and on the informational signage at the sites is c. 4000 BC. McLaughlin suggests that one factor in its adoption is that it’s a nice round number that’s easy to remember … but, however you wish to view it, it’s just not the correct date! Instead he points to evidence such as the first domesticates arriving in Ferriter’s Cove, County Kerry, at 4300 cal BC. Confusingly, there are at least three or four Mesolithic sites with securely dated contexts that show that this lifestyle was still extant at 3900 cal BC. Thus, we appear to have a 400 year period where we have both Neolithic and Mesolithic economies running in parallel. Turning to the court tombs, such as Creggandevesky mentioned above, McLaughlin notes that until recently their chronology has been ‘somewhat elusive’. However, following a dating programme directed by Rick Schulting (and incorporating Bayesian modelling), it is now clear that their first phase of use was between 3700 and 3570 cal BC.

So … developer-led excavations have revealed a large number of settlement sites, and in the Early Neolithic these are in the form of rectangular structures. Many of these are comprised of large post-holes and significant wall-slot gullies deeply cut into the natural subsoil. For this reason, nothing short of a bulldozer is really capable of erasing them from the landscape. Following from this, many examples of this form of site have been discovered and, in turn, have become an enduring focus of research for well over a decade. Work by the likes of Jessica Smyth and Cormac McSparron, among others, has revealed much of the detail about domestic life in the Neolithic. This has come through myriad analyses of deposited soils, the surviving cereal grains, and the few sites that have produced the bones of domesticated animals. From this work, we can now see the Early Neolithic in terms of an economic landscape devoted to cereal farming and cattle husbandry. Other recent work, on the absorbed lipids in pottery, has shown that the secondary products from these animals were also used, including dairying as well as for meat.

Summed probability distribution of all radiocarbon dates from
Ulster contexts 
© R McLaughlin
Taken together, the corpus of radiocarbon dates from Early Neolithic sites indicate that it began after 4000 cal BC. McLaughlin showed a graph depicting the summed probability of all the available radiocarbon dates. As such, it provides a necessarily rough and imprecise idea of what  the known archaeological activity was like at this time. However, he is quick to note that the statistical justification for this approach is not proven and is still under debate. Nonetheless, it is a simple means of creating a general overview of activity and observing times when activity was particularly high. He notes that the Early Neolithic rectangular houses cause an incredible peak in activity at around 3700 cal BC. Although it appears to start slightly after 4000 cal BC, this is misleading owing to the ‘old wood effect’ where the radiocarbon dated charcoal may be from heartwood, several centuries older than the event being examined. This form of signal is particularly pronounced at the start of the Neolithic as many of these houses were built of timbers felled in the primeval forest. These large, ancient trees thus enter the archaeological record, eventually entering the samples, and influencing the radiocarbon dates. His next image showed the same graph, with all the dates from charcoal removed, and only the shorter lived samples being represented. These included cereal grains, hazelnut shells, and bones. While the latter have a greater own age, and contain material which exchanged carbon with the atmosphere several years past, it is nowhere near as extreme as the issue posed by the dating of heartwood charcoal. When examining the evidence provided by the short-lived samples, McLaughlin notes that the ‘Neolithic House Horizon’ begins at around 3750 cal BC. Using Bayesian statistics to model the available short-lived samples from Neolithic houses allows the positing of defined start and end dates to this activity. Various researchers, including McSparron (2008), Whitehouse et al. (2013), and Cooney et al. have examined this and have all produced estimations of when exactly this ‘house horizon’ began and ended. Based on this work, we can now say (with c.95% certainty) that this phase began in the period from 3720 to 3680 cal BC, and ended between 3640 and 3620 cal BC. In human terms, this equates to a period of between one and five generations – not a particularly long time! McLaughlin notes that the application of Bayesian models allows some impressive improvements in the standard calibrated chronologies. While the majority of the evidence points to the genesis of the Neolithic around 3700 cal BC, a small number of sites show evidence of slightly earlier activity at around 3800 cal BC, including Poulnabrone portal tomb, Co. Clare. When we take a more detailed look at the radiocarbon dates from Neolithic houses and plot them all together and compare it to the dates provided by the Bayesian model of the  ‘house horizon’ it is clear that the short-lived samples all converge around 3700 cal BC. However, if we examine the charcoal samples they are all earlier. McLaughlin notes that this distribution of dates is bimodal – i.e. it has two peaks. One is just after 3900 cal BC and the other dates to a little after 3800 cal BC. Such a bimodal distribution can be taken to imply that there were two separate sets of activities in operation, rather than just a single form of activity. To McLaughlin, this earlier phase of charcoal dates relates to the old wood trees used to build the Neolithic houses. He interprets the post-3800 cal BC spike as dates from younger wood, including sapwood and twigs that were being burnt in and around the houses. He suggests that, in his personal opinion, the short lived samples (primarily cereals and hazelnuts) are evidence for the last phase of use for these structures. To his mind, these short lived samples do not represent a phase of use of the sites, but in fact the evidence of their destruction. What we are seeing in the end of the Early Neolithic is the sudden disappearance of this type of rectangular house from the archaeological record. He describes it as ‘an apocalypse … it was a caprice that transcended local circumstances. It happened all through Ireland and suddenly we find that these structures were abandoned … pretty much forever’. He notes that there is one Neolithic house in County Carlow that has a burial in the middle of it that dates to approximately a century later. But this example aside, he sees no evidence that they were revisited in any meaningful way. McLaughlin argues that, better than any form of Bayesian analysis, if we accept that this abandonment was a single event rather than a more drawn-out process, we stand a reasonable chance of directly dating this event. As an illustration, he explained that the laws of probability indicate that if an individual repeatedly fires arrows at a single target, and all other variables being equal, the average of where those arrows land will be the bulls-eye … you just need enough arrows. As it happens, we do have enough arrows – in the form of many hundreds of dates from Neolithic sites. By looking at where the average of these dates lie we can date – with remarkable precision – the ‘house apocalypse.’ He has plotted the modal dates (i.e. the point of a calibrated radiocarbon date where the probability it dated to a given year was the highest) as a frequency distribution. The resulting graph shows a particularly noticeable peak at 3652 cal BC. As he says: ‘I think there was an apocalypse in the Neolithic in 3652 BC’. Admittedly, this is a relatively new statistical technique that is currently unaccepted by the archaeological establishment. McLaughlin calls the technique Computed Radiocarbon Average Probability, though he suggests that it may be better referred to by its initials.

Dick Proenneke’s cabin under construction in Twin Lakes, Alaska (Source)
From all this, McLaughlin reckons that we’ve got a Neolithic that kicks off around 3800 cal BC, but doesn’t get into its full flourishing, in terms of deposition of materials, until about a century or so later. Drawing inspiration from Ray Mears’ latest TV show, How the Wild West was Won, he observes how the North American forests are known to have been cleared as part of the westward migration of people. In the 17th and 18th centuries was a specialist class of men who went forth into the wilderness, chopped down the trees, built farmsteads … but didn’t actually do anything with them. It was only when a second wave of settlers came along, they sold the freshly cleared land and cabins to the new arrivals as readymade farms. Clearing the Neolithic primeval forest using only stone tools would have been a hugely difficult undertaking, but McLaughlin notes that there may be evidence for this early forest clearance in the bimodal distribution of the radiocarbon dates. At this point he mentions one of his heroes, Dick Proenneke, the amateur naturalist who built a log cabin in the remote Twin Lakes region of Alaska. Across the world, where anthropological and historical records exist, the evidence shows that people living in heavily forested environments build rectangular houses from logs – the readymade availability of substantial timbers just appears to be readymade for converting into rectangular houses! Thus, when people stop building in this way it suggests to McLaughlin that they had simply run out of trees – the primeval forest had been cut down and a farmed, managed landscape replaced it.

If we plot all the known Early Neolithic houses, it is clear that there is a predominantly easterly distribution in Ireland. This is, in part, itself an artefact of the areas in which recent development has occurred and where development-led excavations have been undertaken. However, the distribution of contemporary megalithic monuments is generally towards the west. This too is probably quite biased as there has been less modern settlement and development to degrade and destroy these upstanding monuments outside of the major urban centres. Even accounting for these various forms of bias and distortion of the record, McLaughlin argues that distinct regional variations may be noted. For example, there is a distinct preference for megalithic tombs in the northern half of Ireland, and along the valley of the River Barrow, stretching from Waterford harbour to Dublin harbour. When this is plotted against where development-led excavations have discovered Early Neolithic settlements, there is a noticeable overlap. Examples include the vicinity of what McLaughlin terms ‘Derry-squiggle-Londonderry’ (though I believe it may be more appropriate in this context to go for Derry±London), the Antrim coast, Dundalk harbour, and along the ‘Barrow corridor’. Looking at what type of sites are being constructed, in the north there are court tombs and rectangular houses, while in the south-east there are portal tombs and rectangular houses. In the south-west – which until relatively recently wasn’t thought to have a Neolithic at all [see: here] – just has rectangular houses. McLaughlin suggests that we’re only now beginning to get a sense of the different superposition of depositions going on in prehistory, even if we’re not yet at a point where we can satisfactorily explain what it all means. He says that a major change happened at around 3600 cal BC – or at around 3652 cal BC if you follow his C.R.A.P. model. At this time the Neolithic houses suddenly fell into disuse and, instead, deposition happened at pit complexes and similar sites. These are spreads of material that don’t make particularly interesting looking (‘juicy’) sites, which has led to relatively few of them being radiocarbon dated. He showed a map of sites that haven’t been dated but might be of this period and argues that they appear to be relatively widespread across the Island. Added to this are a small number that have been dated. From these sites we get a picture, though one that is necessarily more ephemeral owing to the lack of data, of a Neolithic economy that is largely unchanged from the preceding, Early Neolithic. He pointed to one particular dot in the vicinity of Larne, looked at me and asked ‘is that Ballyboley?’ This was an absolute pig of a site that I directed in 2004 to 2005 … just a giant, unending series of pits and random postholes that took forever to excavate and I still can’t explain what’s going on. One of the reasons I can’t explain what happened there was that the post-excavation never progressed beyond a basic stratigraphic report and the thousands of decorated sherds of pottery and flint tools have never been examined and professionally analysed. Just to give one clue as to the extraordinary wealth of this site: there was one feature that looked utterly un-special … an amorphous blob about a metre wide and perhaps two metres long. Two guys were sent off to clean it up and one took one side and one took the other. I came over to find that although they were just giving it a light trowel down, the feature was so incredibly rich that they were able to play a game of ‘who’s found the most hollow scrapers?’ … if memory serves, the final score was something like 8 to 7, and that wasn’t even the richest feature. A further point, if slightly tangential, is that although this was a ‘development-led’ excavation the samples were processed for free at QUB as a training exercise for students of various grades, and the radiocarbon dates were also done as part of the INSTAR Cultivating Societies project. That aside, McLaughlin was able to tell me (and the assembled audience) that the radiocarbon dates from the site show that it’s the latest Early Neolithic site to produce cereals. Anyway … returning to the pits … he showed just what one of these typical pits looks like in excavation … unimpressive, uninspiring, frequently lacking in artefacts, and unlikely to have the expense of a radiocarbon date lavished on it.

Density of Early Neolithic sites in Ireland, arrows represent hypothetical points of entry for the Neolithic and/or points of contact with Britain. Regions of special interest: 1 Derry±London, 2 Antrim Glens, 3 Dundalk Bay, 4 Barrow Valley, 5 Munster, 6. North Mayo © R McLaughlin
Turing to the types and numbers of features that have been dated, it can easily be seen that there are a very large number of these pit and spread complexes in comparison to, say, rectangular houses. Although the data in the graph is slightly out of date, it still serves to illustrate the general point that although there were three times as many pit complexes as rectangular houses, a very large percentage of the houses have been dated (74%), in comparison to the pits (34%). This is thrown into even more stark relief when we look at it in terms of those sites whose chronology we have a good understanding of. A significant proportion of the rectangular houses have well understood chronologies (48%), while only 10% of pit and spread complexes are comprehended with the same degree of precision and clarity. Indeed, other site types have, as Rowan states, ‘similarly diabolical levels of understanding associated with them.’

Known
Dated
Good Chronology
Early Enclosure
4
3
2
Rectangular house
54
40
26
Other enclosure & timber circles
55
17
7
Neolithic pits & spreads
159
54
16
Burnt mounds
7
7
3
Court tombs
398
12
4

One of the good things about working with calibrated radiocarbon years is that we may compare the patterns we see with archaeological signals from other regions, with tree-ring data, with ice-cores etc. These allow us to gain understandings of past climate and environment. Phil Barratt of QUB has been working to re-examine every available prehistoric palaeoenvironmental study from Ireland. One of his objectives is the seeking out of potential environmental events that correlate with the information in the archaeological record. One of his areas of study relates to the ‘Plantago Gap’. Plantago lanceolata (ribwort) is a weed that grows in association with arable farming. Thus, we can identify points in time where arable farming was especially important by the amount of ribwort pollen recovered from environmental samples. However, there is a period in the Middle Neolithic, after 3500 cal BC and before 2500 cal BC where pollen from the ribwort simply disappears from the record. The implication is, of course, that the importance of cereals diminished during this period. This is backed up by the archaeological record itself, where cereal grains also cease to be found between these dates. One of the challenges facing this form of analysis is the difference in scales of resolution between the fine grained chronology that can be achieved in the archaeological record and the much less precise one of the environmental data. The Middle Neolithic period simply doesn’t have all that many identified settlements and the pollen record does seem to indicate that arable agriculture just stopped. Despite the monitoring and excavation of literally thousands of sites across Ireland, there is remarkably little well-dated, secure evidence of settlement sites in the period from c.3400 cal BC to c.3300 cal BC. However, there are plenty of passage tombs – arguably, the most famous of all Neolithic monuments. These give an insight into the period where sites like Newgrange, Knowth, Dowth, etc. are large-scale, well-constructed monuments with art and astronomical alignments etc. and carefully placed in the landscape, but almost all evidence of domestic activity is absent [see also a review of a lecture by Robert Hensey: here]. As McLaughlin says: ‘There are no settlement sites. In fact the only thing we can really get at of day-to-day life is by looking at the human remains themselves, which are actually quite few and far between.’ Where these remains have been examined, it appears that there is little difference between them and other people at different times in the Neolithic, certainly in terms of what they were eating. Nonetheless, we still have no real idea as to how the highly ritualised and organised forms of behaviour that produced the Passage tombs related to day-to-day life. Acknowledging the old chestnut that ‘absence of evidence does not equal evidence of absence’, McLaughlin proposes that one solution to this issue may lie in a form of nomadic pastoralism, perhaps similar to that experienced by the modern Qashqai in Iran. Part of their lifestyle includes the herding and movement of large herds of domesticated animals by relatively small, but highly mobile, communities. Coming from this approach, McLaughlin sees the possibility of Passage tombs being located prominently in the landscape – on hills and ridges – where they were points in an annual cycle. In this way, the annual journey of the kin group intersected in significant ways with the annual journey of the spirits/ancestors. However, as McLaughlin wryly observes: ‘Archaeological theory really does reach is full flourishing when you don’t actually have any raw data.’

In the Late Neolithic we have plenty of data and we see a return to site deposition. One of the styles of pottery that becomes popular in Ireland at this time is ‘grooved ware’, a flat-based or ‘bucket shaped’ form, introduced from Orkney. In terms of sites, we see the introduction of large earthen henges. While we can’t currently see it with chronological precision, it does appear that the introduction of henges was a relatively sudden occurrence. Unfortunately, this point is just at one of the ‘wiggles’ in the calibration curve for radiocarbon dates, making it difficult to gain precise dates and insights. As McLaughlin says ‘It does seems like a large number of people sprang out of the ground in the Late Neolithic.’ Similarly, the environmental evidence from these sites indicate that there is a return to cereal cultivation and animal husbandry.

In the Beaker period we see the development of wedge tombs (a uniquely Irish development) and burnt mounds (known from all over northern Europe, but especially found in Ireland). If we look at the numbers of these burnt mounds known versus those dated, we see that some 75 have been discovered and excavated, but only four have been particularly well dated. This is in strong contrast with cist burials, 40 of which have been excavated and dated, of which 35 may be described as having a well understood chronology. Beaker pottery arrives in Ireland around 3260 cal BC and has died out by 2200 cal BC. This is of interest as the same pottery forms continue on in Britain until 1800 cal BC, so whatever was happening in Early Bronze Age Ireland caused Beakers to fall into disuse – perhaps Ireland got more insular more quickly at this time. Many of the cist burials have been discovered as part of agricultural works and, because they were so obviously a stone box containing bones, they were more frequently recorded and at least the artefacts saved. Through the work of certain researchers, the chronology is particularly well understood and they are now known to have been in use from 2200 cal BC to 1800 cal BC. In terms of the Early Bronze Age, we have burnt mounds (that are created and used across the whole period), wedge tombs that date to the start of the period, and cists that essentially date to its end … or at least ‘the end of the middle part of the Early Bronze Age’. If we analyse the distribution of radiocarbon dates from these cist burials, we find that in the south of the island they appear at around 2170 cal BC and move into the northern portion of the island around 2110 cal BC. Because so many of these burials have been so well dated, we can see genuine demographic patterns in the data. McLaughlin says ‘we can see these Early Bronze Age people that brought with them the insular pottery style, the cist burial tradition – they appeared in the south of the island and they moved north.’ If all classes of prehistoric monuments were as well dated as the cist burials, it would simply revolutionise our understanding of those times.

In the final section of the lecture, McLaughlin ventured into the quagmire that is suggesting population numbers at various times in Irish prehistory. While it is not an easy topic to tackle, it is one that is frequently asked by the public, and is one that deserves a thoughtful answer. He began with a graph of radiocarbon dates from North America, explaining that in such a large landmass the relatively localised demographic and economic changes that occurred before the arrival of Europeans can be averaged-out. For this reason it is a useful tool with which to compare the radiocarbon dates from archaeological activity in other places as it, essentially, provides a ‘null hypothesis’. To a large extent, it allows us to say, all other factors being equal, what would the distribution of radiocarbon dates look like? – the only factor that really influences this distribution is the exponential decay of datable material. This expresses itself in a graph that shows lots of recent radiocarbon dates, with somewhat fewer further back in time, and fewer and fewer as one travels further and further into the past. As McLaughlin says: ‘this decay channel is provided by erosion and diagenesis allows you to see this exponential decay.’ Such a null hypothesis allows us to apply statistical tests to some of the apparent patterns in the Irish archaeological record. McLaughlin explained that he fitted an exponential decay function on the available northern European data set and compared it to the Irish evidence. Although not many pits, relatively speaking, have been dated they are more robust against the research interests and biases of individual archaeologists and institutions. His graph indicates that there are vastly more radiocarbon dates from the Early Neolithic than what one would expect under the null hypothesis. McLaughlin describes the increase in radiocarbon dates in the Late Neolithic as ‘a wiggly line’ that increases gradually and at approximately the same rate as the null hypothesis line. It markedly increases at around 3300 cal BC. He argues that this may be explained in terms of there being another immigration into Ireland around 3300 cal BC. Presumably there was an immigration of some form at the start of the Irish Neolithic and there was another one in the time after 3000 cal BC. He sees these as being characterised by the movement of a significant number of people. This would have been at a time when the great passage tombs like Knowth were still in use, leading him to the vastly understated remark that ‘this must have been an interesting time to live.’ After this point, the radiocarbon signal increases gradually. He sees that there may have been successive waves of people, each bringing with them new technologies, languages, and religious ideologies, but in terms of the absolute rate of archaeological deposition – the number of sites, and ultimately, the number of people – there do not appear to be too many changes after 3000 cal BC. In a prodigious feat of ‘blue sky thinking’ he attempts to turn all this data into hard figures of population. He notes the development-led excavations have discovered Early Neolithic rectangular houses at almost the same rate as we know the surviving pattern of Early Christian ringforts. There are approximately 60,000 ringforts known from the island of Ireland and using this as a very crude analogue, he suggests that there were once the same number of Early Neolithic houses. From the radiocarbon evidence we can say, with a great degree of certainty, that each of those houses lasted for between one to five generations. If we put a number out of mid-air (or swag) and say that there were 10 people at every site, we get a population in Ireland in the period from 3750 cal BC and 3600 cal BC of 120,000 to 600,000 people. To put this in context, he suggests that the landscape was almost as heavily populated in the Early Neolithic as the modern Irish countryside of, say, Tyrone. McLaughlin argues that, in many respects, this makes an awful lot of sense. He notes that he once calculated that, given the density of the primeval Neolithic forest, it would have required every man, woman, and child to cut down trees at a rate of only one every week to clear the entire island within that 150 year period. In the Early part of the Middle Neolithic or the transition from the Early to Middle Neolithic (3600 cal BC to 3400 cal BC) there are three times the number of pit deposition sites as there are Early Neolithic houses. However, he suggests that fewer people are responsible for each site as these are simple holes in the ground in comparison to the more complex rectangular structures. Also, Bayesian modelling indicates that these sites were in uses from around four to eight human generations. Extrapolating from all these limits and caveats, suggests a population of between 100,000 and 200,000 people. In the Middle Neolithic proper, during the period of the Passage tombs, we just don’t know what the population would have been like as we simply do not have the identified settlement sites. If we accept – even for a few moments – the huge assumption that they became nomadic pastoralists, anthropological studies indicate that the average group size would have been in the region of 40 to 150 people. If we then make the simple-minded (McLaughlin’s term) assumption that there was a one-to-one relationship between Passage tombs and kin groups … and there were approximately 150 tombs … we get a population of only 6,000 to 23,000 people in Ireland at this time. Although there are more caveats than you can shake a dirty trowel at, McLaughlin is firm in his assertion that the population was markedly smaller than in the Early Neolithic. He sees them as completely different societies that have nothing in common other than the fact that we today call them both Neolithic. In the Late Neolithic the rate of pit deposition is approximately that of the period from 3600 cal BC to 3400 cal BC, if we factor in the exponential decay of radiocarbon dates. Thus, he estimates that the Late Neolithic population was in the range from 50,000 to 100,000 people. While there are more sites of Beaker and Early Bronze Age date, what McLaughlin terms the ‘taphonomic correction’ again suggests a population of between 50,000 to 100,000 people.

McLaughlin stresses that the act of gathering together all the data from the available archaeological excavation literature allows incredible new insights into past lives and activities, particularly when those things happened. While it sounds simplistic, he is keen to stress that this form of research is actually a radical departure in modern archaeology. While there has been so much good research and theoretical discourse about, for example, what grooved ware meant to the people who created it, or how the landscape was encountered by the agriculturalists, or how and why power and communities were transformed at the start of the Bronze Age, this new wealth of data allows us to ask questions about the realities of domestic life in prehistory that could not even have been framed a decade ago. He sees this as a direct result of the boom in development-led archaeology that has created a ‘data set that is almost the best resolved in the wide world.’

Returning to the opening analogy of the planet Saturn, McLaughlin notes that we still do not know where on this ‘scale of knowledge’ we currently are. He says: ‘We’ve built the telescope – we have radiocarbon dating – we’ve amassed knowledge about the distant past at a rate that was unimaginable a few years ago. What we don’t know though is have we resolved the Cassini Division? Can we see surface detail on the surface of Saturn? Or are we still being foiled by what we don’t understand about the structures? I would like to think we are somewhere in between Huygens and Cassini. And soon, if I were to come back in maybe five years’ time, and present again, we will have a much better resolved idea of what is going on.’ McLaughlin has amassed a total of 1206 radiocarbon dates from Ulster but, he argues, there should be ten to one-hundred times that number. He argues that there is little point anymore in excavating archaeological features if they are not going to be radiocarbon dated. Citing Bayliss et al. (2007), he notes that ‘a date is just a number and a radiocarbon date is just an expensive number’ … but McLaughlin goes beyond this to argue that ‘an archaeological feature without a radiocarbon date is just an expensive hole in the ground’. He believes that excavation resources should be weighted in favour of understanding chronology, because it is only an understanding of chronology that leads to a true understanding of the past. He concludes that what has been discovered so far is remarkable, but improvements are definitely needed. Commercial archaeological companies must produce more radiocarbon dates, they must take better samples [you’ll find no argument from me on this one! here | here | here], in particular we need to date more ‘ordinary’ pits and not just the ‘juicy’ features!

With that, the lecture ended and the floor was thrown open to a variety of questions and a hearty round of applause and profuse thanks to the speaker. Regular readers of this blog will know that I’ve made my views on the value (or otherwise) offered by the commercial archaeology sector (at least in Northern Ireland) exceedingly clear. Rather than bore you all with a rehash of my arguments, I’ll direct the reader in search of a counterpoint to here. Even though Rowan is coming at the utility of the commercial sector to enhance our shared knowledge from a totally opposite direction (he sees the doughnut, I’m looking at the hole), he still reaches the conclusion that what we’ve got to work with is much less than we should have. As he notes above, we should have vastly more radiocarbon dates from Irish sites – in the range of two orders of magnitude. Leaving those kinds of questions aside, we must turn to the archaeological insights he provides and ask if he’s right. I have no doubt that McLaughlin is – in the long run – going to be proven wrong about most of his points. As he says, our understanding may well be somewhere in between Huygens and Cassini, but that fails to countenance the modern images of Saturn taken from Hubble. Please, don’t believe that this is in any way a criticism of Rowan’s work – it’s exactly the opposite! If future researchers see further and understand deeper, they will only have done it through standing on the shoulders of Rowan – and those like him – who are constantly pushing the boundaries of our knowledge of chronology. One way or another, when McLaughlin speaks, he’s certainly worth listening to!

Reference
Bayliss, A., Bronk Ramsey, C., van der Plicht, J. & Whittle, A. 2007 ‘Bradshaw and Bayes: towards a timetable for the NeolithicCambridge Archaeological Journal 17.1 (suppl.), 1-28.

McSparron, C., 2008 ‘Have you no homes to go to: calling time on the early Irish NeolithicArchaeology Ireland, 22.3.

Schulting, R.J., Murphy, E., Jones, C., & Warren, G., 2012 ‘New dates from the north, and a proposed chronology for Irish court tombsProceedings of the Royal Irish Academy , 112C.

Whitehouse, N. J., Schulting, R. J., McClatchie, M., Barratt, P., McLaughlin, T. R., Bogaard, A., Colledge, S., Marchant. R., Gaffrey, J. & Bunting, M. J.  2013 ‘Neolithic agriculture on the European western frontier: the boom and bust of early farming in Ireland’ Archaeological Science


Note
I am grateful to Rowan for having a read over this post and helping correct a number of errors and for providing clarification on a number of points. However, any that remain are mine alone. Rowan has also been kind enough to provide me with two of the images from his presentation, for which he has my sincere thanks.