Wednesday, April 25, 2012

Corrstown: A Coastal Community. Excavations of a Bronze Age village in Northern Ireland: Review

Victoria Ginn & Stuart Rathbone (eds.). Oxbow books, Oxford, 2012. xv+301pp. ISBN 978-1-84217-464-7. £26.95 (via Oxbow) or £33.25 (via Amazon).

[** If you think the review is useful, please re-share via Facebook, Google+, Twitter etc. **]

Corrstown: A Coastal Community presents the results of the excavation and analysis of 76 Bronze Age structures excavated by Archaeological Consultancy Services Ltd. during 2002 and 2003. Along with the well-publicised Bronze Age village, an Early Christian rath and rock-cut souterrain were investigated, together with the recovery of a small collection of Neolithic pottery. Chapter 1: Introduction sets the scene, outlining the development-led background from geophysical prospection, surface collection of artefacts and test-trenching. This led to further test-trenches and eventually, in October 2002, an excavation directed by Malachy Conway. What was initially regarded as a small prehistoric settlement and an Early Christian rath, grew and grew with every further area topsoil stripped. The authors put it best: “After several weeks of topsoil stripping it was clear that the settlement consisted of many more buildings than could have ever been expected and the site was of far greater complexity and scale than had been previously envisioned for the Irish Bronze Age.” I would commend section 1.2: Surprising results to readers alone for the eloquent (and expletive free) description of what life on a development-led excavation is like in Northern Ireland. The houses (74 of them, plus two W-shaped structures) ranged from circular to oval in plan and most produced large quantities of pottery. Some of the houses were of the well-known construction type, defined by a slot-trench and containing internal structural postholes (Type 2). However, the majority of buildings at Corrstown were of a previously unrecognised type where wide, segmented ditches and multiple pits and postholes formed no easily recognisable pattern (Type 1). Several consisted of concentric rings in this manner and one appeared to stand within a deep, horse-shoe enclosure. Many of the roundhouses had long, sunken porches with cobbled floors. Some houses were connected to each other with similarly sunken and cobbled pathways. A roughly cobbled road (70m, by 10m wide) ran through the eastern portion of the site, while a second, unpaved, road is postulated, based on the absence of features along the western side of the site. Tantalisingly, the cobbled road runs out from the village, beyond the limits of the development. The fact that these houses were connected by paths and that only a small number of structures overlapped each other suggests that the houses are broadly contemporary. This is further confirmed by 24 radiocarbon dates from 22 of the houses, showing a defined cluster in the Middle Bronze Age (c. 1700-1200 cal BC). Taken together, this is the first indisputable evidence for an actual Bronze Age village in either Ireland or the UK. The recovered artefacts were, understandably, dominated by those things that survive best: pottery and lithics. Over 9000 sherds of pottery were recovered. For its size alone, this is an exceptional assemblage, but a previously little-known form of Middle Bronze Age coarse ware – derived from the Cordoned Urn tradition - has been identified within it, adding to its significance. The volume of recovered flint was so large that only estimates of its size may be made – over 16,500 pieces (509kg for secure contexts and a further 63kg from non-stratified contexts). Similar to the pottery, this is one of the largest Bronze Age assemblages ever recovered from an excavation on this island. Other significant finds included four stone moulds, two polished axe heads, and a polished mace head. The Early Christian portion of the excavation (easy to forget when bracketed together with such amazing discoveries) included the full excavation of the rath and rock-cut souterrain, along with an unusual rectangular structure and a kiln.

Simplified plan of the Corrstown Village site
Chapter 2: Excavation results summarises the findings and gives detailed definitions of the Type 1 and Type 2 structures. This is followed by a detailed gazetteer of the structures. The authors are to be complimented for providing clear, readable results in a consistent format. The description of each structure is accompanied by a full page (A5) plan and occasional section drawings or profiles. While I would have liked to see more section drawings, it must be remarked that the plans are of the highest quality and, like the written descriptions, are clear and consistent. Weighing in at 143 pages, this is easily the largest part of the book. I cannot stress how valuable this is – it represents the original data that all subsequent interpretations of the site depend on. I am sure it would have been attractive (even ‘easier’) to present a synthesis of the site that omitted this primary data. However, as times progress and scholarship inevitably moves on, the primary data will still be available for future re-interrogation and reinterpretation. This approach has been a personal obsession throughout my archaeological life. Admittedly, I am normally chastised for writing in this way, so when I see other writers obviously engaged in the same project, I do wish to applaud their efforts. In Chapter 3: Material Culture and Environmental Analysis the authors note that all the recovered artefacts were manufactured from locally available sources and, while they demonstrate that the community here was a vibrant, thriving place, there is no evidence for contact with the wider world. The first specialist analysis is Maria O’Hare’s examination of the lithic assemblage. She concludes that the assemblage, despite its size, is typical of domestic contexts from the Irish Bronze Age. The most common tool types produced include utilised pieces and scrapers (both formal and ad hoc types). Flaked and more formal tools were also present, but in much smaller quantities. Helen Roche and Eoin Grogan explain that this is the single largest collection of Middle Bronze Age pottery yet found on this island. This represents a minimum number of 492 vessels: nine Early Neolithic bowls and 483 Middle Bronze Age vessels. The Neolithic vessels are pretty standard and can be paralleled with numerous sites, including both domestic (e.g. Ballygalley, Ballyharry & Ballynagilly) and burial sites. The dispersed nature of the evidence across the site is taken to suggest that there had been some small-scale domestic activity in the Neolithic, but that this had been largely disturbed during the main period of Bronze Age occupation. Almost all of the houses produced Bronze Age pottery, though the quantities ranged dramatically, from as few as six sherds up to 711 sherds. The majority of the vessels (386) were plain, but some (97) possessed some form of decoration, most notably cordons. This suggests that the vessels derive from the funerary-based Cordoned Urn tradition, though their recovery from domestic structures is significant. Based on the work of Anna Brindley (2007), Roche and Eogan argue that the funerary version of the Cordoned Urn dates to the period 1750-1450 cal BC, but that the domestic version continues in use beyond this, possibly as late as the beginning of the Late Bronze Age (c. 1250 cal BC). The importance of the Corrstown assemblage is that it allows the re-dating of this form of relatively well-made plain vessels from the Late Bronze Age back into the latter portion of the Middle Bronze Age. While the authors note that this has serious implications for the dating and interpretation of many sites, they note that, in purely ceramic terms, a number of sites dated to the Late Bronze Age solely on the presence of coarse pottery may need to be re-examined. Eoin Grogan also contributes an analysis of the stone artefacts. One miniature axe head, five large axe fragments, and one small flake were recovered during the excavations. Unfortunately, the large fragments were all recovered from the topsoil. The miniature example was recovered from a secure context within structure S37. Grogan sees all the axes recovered as being paralleled within the Irish series and argues that all relate to the Neolithic phase of site use. However, given the relative size of the Bronze Age settlement, it may well be justified to reinterpret the evidence in light of Jolliffe’s (2010) research, where he argues for the evidence of polished stone axeheads being made during the Bronze Age. The perforated macehead (of Largs-type) was made of gabbro and was recovered from a posthole in structure S1. Four stone moulds, all of dolerite, for the casting of bronze artefacts were recovered. These included part of a mould for a palstave, two for the casting chisels, and one for a ribbed, bladed implement, currently unparalleled in the excavated record. Stephen Mandal also contributes a brief note on the stone types and their probable sources near the site. Örni Akeret contributes a note on the recovered seed types. He notes the absence of chaff in the samples, but cautions that this may have been due to a recovery bias at the flotation stage of sample processing. The range of cereals identified is seen as typical of this period: with a dominance of barely and, in particular, naked barley (Hordeum vulgare). While approximately 150 barely grains were recovered from one stakehole in structure S58, only a single rye grain was recovered from the entire site. Akeret suggests that the rarity of rye on this site and others suggests that it occurred as a weed, rather than as a cultivated crop. A small number of wheat grains (Triticum) were also recovered. A number of these could be further identified as emmer wheat (Triticum dicoccum). A further table gives all the primary information regarding cereal types and the features they were recovered from. In a final note on this chapter, I would like to draw attention to the excellent artefact photography by John Sunderland – it manages to fulfil my personal dictum for all archaeological photography: it should be archaeologically informative and aesthetically pleasing.

Chapter 4: Analysis of the Corrstown Site provides the overarching synthesis of the site, drawing on the primary data presented in Chapters 1 and 2. Detailed discussions are presented on the Type 1 structures, examining individual components of the structure’s morphology, including walls, entrances, drains, etc. The Type 1 roundhouses are broken down into several types, including single segmented ditch roundhouses (1A), concentric segmented ditch roundhouses (1B), conjoined concentric segmented ditch roundhouses (1C), freestanding structures (1D), small abutting structures (1E), and undetermined segmented ditch roundhouses (1F). Only six structures fall into the category of the Type 2, narrow slot roundhouses. While in the minority here, they are the most common type found elsewhere and a comprehensively paralleled with other excavated evidence, including Kilmurray, Co. Wicklow, Lisheen Site A, Co. Tipperary, and Ballyrennan, Co. Antrim. This is followed by a detailed note on the reconstruction drawings of Type 1 and Type 2 houses by Ella Hassett. The authors explain how the drawings were produced after detailed discussions between artist and archaeologists. They go on to discuss such variables as roof coverings and slope, along with the whole host of necessary compromises, guesses and speculations that are required to produce such drawings, but are not recoverable through excavation. I have written before of the importance of this form of discussion, where the process of reconstruction is dissected and becomes a further means of interrogating the excavated data and the assumptions we place upon it. An analysis of the rebuilding cycles at Corrstown, where individual buildings appear to have been continually rebuilt on the same spot, are interpreted in terms of the sequential replacement of a single homestead. Somewhat fewer than 50% of the Corrstown houses appear to have been rebuilt, with the extent of such renovations running from major overhauls to complete rebuilds. The authors suggest, based on a number of strands of evidence, that the Corrstown houses may have had life spans ranging from 30 to 80 years. Although fraught with difficulty, the authors bravely attempt the caveat-laden procedure of producing an estimate of the Corrstown population. While various means of calculating an estimate produced results from 200 to 1400 people living in the village during its peak, the authors prefer the lower population range of 200 to 300. In any event, they point out that, even at the lowest estimate, the Corrstown population would have been far above that of any previously known Bronze Age settlement in Britain or Ireland. As noted above, on page 7 of the introduction there is a reference to 24 radiocarbon dates from 22 roundhouses. However, on page 231 it is clear that there are some 33 radiocarbon dates listed – the reason for the difference appears to be that the additional dates have been carried out as part of Victoria Ginn’s PhD research. This may seem like a minor point to bring up, but it does demonstrate that although the report is written and the book is published, the research goes on. Four further Early Christian dates, taking the total to 37, are listed separately in a table on page 292. All of these are new to the IR&DD Catalogue [come and 'like' the IR&DD Facebook Page]. Unfortunately, the charcoal samples for original batch of 28 dates were not identified as to wood species. The authors raise the possibility that some of these included old-wood samples, making the dating less reliable that would be desired. It is a minor point, but one worth noting that the authors perpetuate the belief that oak wood should not be dated as it is likely to be significantly older than the feature being dated – one sample was dismissed for dating precisely for this reason. Instead, selection of samples should be focused on identifying short-lived, single entity items, including charred twigs and sapwood etc. (Chapple 2008; 2010a). In this way, even taking dates on oak would be ‘safe’ and eliminate the possibility of the ‘old wood’ effect. Although not the fault of the authors, it is worth noting that the current requirements for post-excavation work in Northern Ireland do not demand the identification of wood to species prior to radiocarbon dating. I believe that a change in this policy, while not eliminating all the caveats of radiocarbon dating, would bring much more clarity and robustness to the process. I am delighted to see that the curve (IntCal09) used to calibrate the dares is included, though it would be nice to have had the name and version of the computer software used included too. However, one personal bug-bear that I would like to see eliminated from archaeological writing is the inclusion of the lengthy reference to the publication of the IntCal calibration curve directly associated with the radiocarbon dates. In this volume it is included beneath both tables of radiocarbon determination on pages 231, 292, and 301. All that is required is a note to indicate the curve used and the reference (Reimer et al. 2009), while the full citation should be confined to the bibliography. Such minor annoyances aside, McSparron’s analysis of the dates (fully elaborated upon in Appendix I) identified three main phases of activity. Taken together, the returned determinations demonstrate that the Corrstown settlement was founded around 1500 cal BC, most likely after 1460 cal BC. This Growth Phase lasted up to 260 years. The Village Phase commenced around 1360-1270 cal BP and lasted for 35-155 years. A final Decline Phase began after 1150 cal BC and lasted from 80-300 years. This analysis, based purely on the radiocarbon data, is supported by the spatial organisation of the village, indicating that the most central structures were among the earliest etc. McSparron’s analysis also indicates that there was no abrupt beginning or end to the village – it appears to have started with one or two structures and grew organically. Similarly, the evidence points to a long, slow decline as opposed to any cataclysmic destruction of the village. A spatial analysis of the village indicates that the northern and southern portions of the site appear to have been organised in different manners, suggesting that they developed at different times. The northern portion is characterised by pairs or small groups of structures, while the southern side is generally formed on three well-laid-out lines of buildings. This is further supported by the artefact analyses and the radiocarbon determinations. Further sections examine the domestic unit and the significance of the roads and pathways.

Chapter 5: From Inception to Abandonment first attempts to place the Corrstown village within its landscape setting. This is followed by an examination of the evidence for the diet of the villagers. In the absence of surviving animal bones, this short section relies on the recovered charred seeds. Like many excavated sites on this island, the agrarian economy appears to have been centred on barley production. An analysis of the orientation of building entrances shows a pronounced, but not overwhelming, orientation towards the south-east, similar to many other excavated Bronze Age and Iron Age houses. Although functional concerns, such as avoiding the prevailing winds are one possibility for this decision, other factors may have included cosmological concerns and deliberate alignments on such phenomena as the midwinter sunrise. The evidence for structured deposition in the form of foundation deposits or closing offerings is assayed. While the placement of a pot base in a pit in the entrance of structure S58 and the macehead from S1 may be easily interpreted in this way, other deposits may be more difficult to reconstruct. One possibility is that, at least some, of the charred barley recovered from the site was intended as ritualised deposits. This bears some similarity with my own excavations at Site 19, Gransha, Co. Londonderry (Chapple 2010b). Here a series of atypical cists – broadly contemporary with Corrstown - held only a few unidentifiable scraps of bone, but produced many thousands of charred barley grains. The authors also suggest that structural deposition of pottery may have taken place in the ditches surrounding some of the houses. Again, this is paralleled at Gransha, where the surrounding ditch appears to have been used for selective deposition of both pottery and flint. The discussion on this point is well worth reading in depth for its balanced approached to this contentious area. Other sections examine the place of the roads and movement of goods and people into and out of the village, along with the evidence for burial at this period. The question of the decline and abandonment of the village is the last section in this chapter. Although the possibility of abandonment due to the environmental effects of the Hekla 3 eruption is assessed, it is discounted in favour of other aspects of climate change. These include the several episodes of colder or wetter weather that occurred in the period from the 16th to 6th centuries cal BC. This is backed up with evidence from pollen analysis of Gary Bog, Co. Antrim, and Gortcorbies, Co. Londonderry, and is also set in the wider context of population decline throughout Ireland, Britain and the Mediterranean world.

Chapter 6: Corrstown in context first attempts to fit this unparalleled site within the economic and social landscape of the north coast. The authors assess the archaeological situation of the 1990s where there was much speculation as the density of prehistoric settlement in the area, but relatively little evidence to support it. More recent excavations, such as Crossreagh West, Crossreagh East, and Cappagh Beg, etc., have radically altered this perception. The authors also note that the excavation in a field adjacent to the Corrstown site, in Magheramenagh townland, produced similar structures and dates. For these reasons, they regard it as part of the Corrstown village settlement. The Magheramenagh excavation was undertaken in a number of phases from 1999 to 2000, and I was present for one of these during 1999. While I was there we excavated a rock-cut souterrain, a suspected Neolithic house (that later turned out to be Early Christian) and a spread of Neolithic pottery. Unfortunately, the site remains unpublished; though the five radiocarbon dates from the site are available (Chapple 2008). A number of my photographs (scanned from slides) from that phase are also accessible to the public. Further discussion places the village within its island-wide context and attempts to evaluate the status of the settlement and how it should be categorised.

In Appendix I, Cormac McSparron lays out the detailed methodology for his interrogation of the radiocarbon dates and how the three Phases of Growth, Village and Decline were arrived at. Perhaps it is not for the faint-hearted, but for those who persevere it is an excellent example of the power of Bayesian analysis in refining chronologies (See review of Gathering Time for further references and resources for Bayesian applications in archaeology). In Appendix II: Medieval Corrstown the excavation of the rath, souterrain, and kiln are detailed. Although a little too brief for my taste, it is an efficient summary of the Early Christian evidence from the site. I would pick the authors up on one point, though. My reading of the text suggests that the data provided on the Magheramenagh excavation included two dates for the souterrains there. The dates cited are 1160±40 BP (Beta-186549) and 1280±40 BP (Beta-186551). Unfortunately, the authors have been provided with the Measured Radiocarbon Age as opposed to the Conventional Radiocarbon Age. I know that I bang on about this all the time, but this is important (see Chapple 2010a; and here). Beta Analytic Inc. provides two types of data. The Measured Radiocarbon Age records the 14C surviving in the sample, while the Conventional Radiocarbon Age takes account of other factors such as isotopic fractionation. Although the Measured Age is useful in certain circumstances, it is the Conventional date that should be published and used in all analyses. While there are not always differences in the two dates, in this particular instance there is a divergence of 30 and 20 Radiocarbon years, respectively. Also, neither date is associated with a souterrain. To be fair to the authors, they do state that they were given the dates directly, but were not able to consult the report text, so the fault does not lie with them. To set the record straight, the details of these two dates are as follows (in both cases the dates were calibrated with Calib 6.1.0, using the IntCal09 curve (Reimer et al. 2009)):

Beta-186549      1130±40 BP       780-992 cal AD  Charcoal (unidentified) from fill of post-hole (Cut 52, Context 53) in arc of post-holes/circular house (Area 1).

Beta-186551      1260±40 BP       668-869 cal AD  Charcoal (unidentified) from fill of southern wall-slot (Cut 63, Context 64) of rectangular structure (Area 1).

Appendix III: Cappagh Beg, by Steve Linnane and Victoria Ginn, details a further set of excavations carried out by ACS in Co. Londonderry. The features excavated here may be interpreted in terms of two or three houses, two boundary ditches and a cremation pit. Structures A and C are interpreted as houses Bronze Age type. The absence of internal features at Structure B is taken as evidence that it may never have been roofed, and possibly for ritual use. This is a charming little theory, but I see no particular merit in it. Two radiocarbon determinations place the site use in the period from the 11th to 15th centuries cal BC. The authors note that although large quantities of flint and pottery were recovered here, none of it has been analysed. While this report is a valuable interim statement, I would hope that arrangements could be made to fully research and publish this site.

In the forgoing review I have detailed some problems I see with the volume. I realise that these are pretty-much in the order of nit-picking and only obsessives like me worry about these things. I have only one criticism that rises above this level of minor irritation, and even still it is not significant. The colour plates of the artefacts are, as I have mentioned above, excellent, but I would dearly have loved to see some colour photographs of the site. Obviously, such considerations come down to the cost of production, but I cannot help but think that a couple of the photographs of the stone axes could easily have been sacrificed for some general shots of the excavation. I was lucky enough to visit the Corrstown excavation during 2003 and took a number of candid shots of the excavation in progress. For the benefit of anyone who wishes, these are now available for public viewing, as are some shots from the Magheramenagh excavation. Even taken together, my short list of minor criticisms should in no way be seen as detracting from the achievement of this excellent book. It is a remarkable piece of scholarship – each aspect of the site is well described and discussed and parallels are sought to put each in its place within the corpus of excavated sites. This is a remarkable feat on its own, given the complexity of the evidence, not to mention the plain fact that the Corrstown site simply rewrites most of what we thought we knew about Bronze Age settlement in these islands. As I have noted above, there is a strong commitment to making available the primary data so that future researchers may re-examine and reinterpret this site. One line in the blurb on the back of the book probably illustrates the importance of this site better than anything else: “It is intended that this volume represents a beginning of the study of the Corrstown village”. As I said above: although the report is written and the book is published, the research goes on. It is not simply that, if you are interested in the Irish Bronze Age, you should have this book in your library – you must have it.

Brindley, A. L. 2007 The dating of food vessels and urns in Ireland. Galway.

Chapple, R. M. 2010a 'Just an expensive number' Archaeology Ireland 24.2, 29-31.

Reimer, P. J., Baillie, M. G. L., Bard, E., Bayliss, A., Beck, J. W., Blackwell, P. G., Bronk Ramsey, C., Buck, C. E., Burr, G. S., Edwards, R. L., Friedrich, M., Grootes, P. M., Guilderson, T. P., Hajas, I., Heaton, T. J., Hogg, A. G., Hughen, K. A., Kaiser, K. F., Kromer, B., McCormac, F. G., Manning, S. W., Reimer, R. W., Richards, D. A., Southon, J. R., Talamo, S., Turney, C. S. M., van der Pilcht, J. & Weyhenmeyer, C. E. 2009 ‘IntCal09 and Marine09 radiocarbon age calibration curves, 0-50,000 years Cal BPRadiocarbon 51.4, 1111-1150.

If you’re going to do any shopping through Amazon, please go via the portal below. It costs you noting, but it will generate a little bit of advertising revenue for this site! Thank you.

[** If you like this post, please consider making a small donation. Each donation helps keep the Irish Radiocarbon & Dendrochronological Dates project going! **]