Mapping Society: Settlement Structure in Later Bronze Age Ireland

Victoria Ruth Ginn

• Publisher: Archaeopress
• Year: 2016
• Print ISBN: 9781784912437
• eBook ISBN: 9781784912444
• Topic: Social Sciences
• Subtopic: Archaeology
• Index: Social Sciences

This study examines Middle–Late Bronze Age (c. 1750–600 BC) domestic settlement patterns in Ireland. Recent archaeological investigations have extended the knowledge of habitation, but no detailed, systematic attempts have been made to understand the domestic evidence, or to substantially revise the existing models for the development of complex Bronze Age societies. All available data relating to settlements dating to Middle–Late Bronze Age have been collated. An evidence-based chronology for settlement is established for the first time. The data are examined at multiple scales to investigate any spatial or chronological trends in settlement character or distribution. The relationships between settlements and the surrounding environmental and social landscapes are analysed through a GIS. The new data are investigated to see how domestic settlements operated, and if traditional concepts regarding the structure of Bronze Age society can still be upheld. Agent-based modelling and social network analysis provide another dimension to the discussion regarding power, regionalism, and hierarchy within the settlement network. The results reveal a distinct rise in the visibility, and a rapid adaption, of domestic architecture, which seems to have occurred earlier in Ireland than elsewhere in western and northern Europe.

Table of contents

• Cover
• Title Page
• Copyright Page
• Contents
• List of Figures
• Figure 2.1: Lough Gur, taken by author March 2011 (permission was sought to visit the house sites but was not granted by the landowner).
• Figure 2.2: Circular house sites still visible on Itford Hill (Dr Carl Griffin as scale), taken by author in May 2010.
• Figure 2.3: Corrstown, aerial photograph (after Ginn and Rathbone 2012, illus. 1.5).
• Figure 2.4: Ó Néill’s (2009) and Doody’s (2000) classification schemes.
• Figure 2.5: Radiocarbon dated Bronze Age logboats: Beta-83891: Cuilmore, Mayo; GrN-15968: Teeronea, Clare; GrN-18361: Ballyvoghan, Limerick; GrN-18751 Cloongalloon, Mayo; GrN-12618: Curraghtarsna, Tipperary; Beta-78159: Tonregee, Mayo; UB-3848: Derrybrusk
• Figure 2.6: Model of a decentralized Bronze Age chiefdom of northern Europe, at times expanding into central Europe (Earle a d Kristiansen 2010: fig 8.8).
• Figure 2.7: Models of societal development which describe the operating system (descriptive models).
• Figure 2.8: Models of societal development which aim to identify the mechanisms of an operating system (mechanism models).
• Figure 3.1: Research methodology (Kanes 2007).
• Figure 3.2: List of preliminary variables analysed; any analysis of the underlying geology has been omitted, as previous resea ch has found this to be invalid (e.g. Cooney and Grogan 1994; Eoin Grogan, pers. comm.).
• Figure 3.4: Breakdown of availability of specialist data for cereal and faunal remains.
• Figure 3.5: Summary details of maps obtained for use in the GIS (all relevant licences obtained by author, where required). ⤀
• Figure 3.6: Summary details of maps generated by the author for use in the GIS.
• Figure 3.7: Bronze Age radiocarbon dates obtained from Corrstown (funded by The Chrono Centre).
• Figure 3.8: Categories within the radiocarbon spreadsheet.
• Figure 3.9: Chart showing the chronological divisions of the radiocarbon dates.
• Figure 3.10: Summary details of the main GIS tools used within the research.
• Figure 3.11: Cost surface.
• Figure 3.12: Multi Distance Spatial Cluster Analysis (Ripley’s K) explanatory diagram (ESRI: http://webhelp.esri.com/arcgisdesktop/9.3/index.cfm?TopicName=Multi-Distance_Spatial_Cluster_Analysis_%28Ripley%27s_K-function%29_%28Spatial_Statistics%29). The R
• Figure 3.13: Land quality: to determine overall land quality the CORINE and soil datasets were re-classified, converted to ras er format, and then added together to produce a land quality map. This was then further re-classified. Areas with a value of 2 b
• Figure 3.14: Screen shot of the TravellerSim interface window with the Late Bronze Age dataset imported. The output is a low resolution by default.
• Figure 3.15: Screen shot of the TravellerSim interface window once the model had been run with the Late Bronze Age dataset. Se tlement size is set in proportion to its importance compared against the most important settlements. Therefore, the larger the s
• Figure 3.16: Standard parameters in Gephi ForceAtlas layout.
• Figure 4.1: Assignment of broad chronological phases to single-phased sites.
• Figure 4.2: Distribution of sites with and without associated radiocarbon dates.
• Figure 4.3: High Quality Dates (dates from the same structure are given in italics). * represents dates from Early Bronze Age structures which were not included in Figure 4.5; ** represents dates from S104 on Carrigatogher (Ryan) 3; *** represents da es
• Figure 4.4: High Quality Dates unmodelled.
• Figure 4.5: High Quality Dates modelled, assuming that all dates belong to a single broad phase of acitivity, the chronological sequence of which is unknown.
• Figure 4.6: Summed probability distribution of dates shown in Figure 4.5.
• Figure 4.7: Sites with High Quality Dates according to settlement phases.
• Figure 4.8: Summed distribution of burial sites.
• Figure 4.10: Comparison of summed probabilities: a) settlement; b) burial; c) burnt mounds.
• Figure 4.9: Summed distribution of burnt mounds; dates from primary fills of trough or trough stakeholes only.
• Figure 4.11: Comparative chronological chart (1Needham 1996; 2 Plunkett et al. 2013; 3Becker 2006; Eogan 1994; 1995b; Ó Faoláun 2004; Waddell 1998; 4Kristiansen and Larsson 2005).
• Figure 5.1a: Distribution of sites according to assigned broad chronological phases: Early Bronze Age.
• Figure 5.1b: Distribution of sites according to assigned broad chronological phases: Middle Bronze Age.
• Figure 5.1c: Distribution of sites according to assigned broad chronological phases: Late Bronze Age.
• Figure 5.2: Distribution of Chalcolithic pottery, metal artefacts, and hoards.
• Figure 5.3a: Density maps combining datasets of dated sites including settlement, burial, burnt mounds, and enclosures as well as single artefact and hoards with known provenances: Early Bronze Age.
• Figure 5.3b: Density maps combining datasets of dated sites including settlement, burial, burnt mounds, and enclosures as well as single artefact and hoards with known provenances: Middle Bronze Age.
• Figure 5.3c: Density maps combining datasets of dated sites including settlement, burial, burnt mounds, and enclosures as well as single artefact and hoards with known provenances: Late Bronze Age.
• Figure 5.4a: Diachronic changes in density patterns: from the Early Bronze Age to the Middle Bronze Age.
• Figure 5.4b: Diachronic changes in density patterns: from the Middle Bronze Age to the Late Bronze Age.
• Figure 5.4c: Diachronic changes in density patterns: from the Middle Bronze Age to the Late Bronze Age with hillforts included.
• Figure 5.5: Proximity (in km) of Late to Middle Bronze Age sites.
• Figure 5.6: All domestic sites and elevation. Those sites which are located at higher elevations above 150 m OD (including D umbaun 2, Moatquarter and Carrigatogher (Abbott) 1) have a tendency to be located at the edge of higher lands.
• Figure 5.7: Sites and elevation: (a) site frequency and elevation zones; (b) difference between the observed and the expected site frequency and elevation zones, given the land mass of each of the elevation zones.
• Figure 5.8: The earlier activity on domestic sites located on elevations over 200 m OD which indicates that expansion into the uplands was not a uniquely Late Bronze Age phenomenon.
• Figure 5.10: Location of dated sites and land-use trends (difference between observed and expected plotted, given the land mass of each of the land-use types).
• Figure 5.9: Hillforts and elevation: (a) hillfort frequency and elevation zones; (b) difference between the observed and the expected hillfort frequency and elevation zones, given the land mass of each of the elevation zones.
• Figure 5.11: Relationship between dated domestic sites and soil types according to period (given as a percentage of the overall number of sites within each phase).
• Figure 5.12: Relationship between dated domestic sites and soil types according to period (difference between observed and expected plotted, given the land mass of each of the soil types).
• Figure 5.13: Site distribution in relation to different soil categories.
• Figure 5.14: Relationship between dated domestic sites and their proximity to thin, well-drained soils (given as a percentage of the overall number of sites within each phase).
• Figure 5.15: Percentage of sites which have more than 50% of good and high quality land (4 and 5) within their 3 km radius buffer zones, within each category.
• Figure 5.16: Dated sites and the percentage of land according to quality within a 3 km-radius buffer zone.
• Figure 5.17: Difference between the expected and observed frequencies of excavated dated enclosures, RMP/SMR hilltop enclosures, and hillforts, and land quality, given the land mass of each land quality category.
• Figure 5.18: Hillforts (according to area enclosed in km2) and their associated land quality.
• Figure 5.19: Difference between the expected and observed frequencies of Middle and Late Bronze Age single finds (MBASF and LBASF) and hoards (MBAH and LBAH), and land quality, given the land mass of each land quality category.
• Figure 5.20: Difference between the expected and observed frequencies of excavated and dated Early, Middle and Late Bronze Age burials, as well as RMP/SMR cists, barrows, cairns, and land quality, given the land mass of each land quality category.
• Figure 5.21: Site distribution and the major rivers shown; the inset shows settlements with settlement sites located on higher land to the west of the River Barrow.
• Figure 5.22: Proximity of dated sites to major rivers, shown as percentage of overall number of sites within each phase.
• Figure 5.23: Lakeside sites and their details.
• Figure 5.24: Percentage of sites within each category and distance to the coast. The date ranges represent single and multi-phased sites; for the provincial data all sites have been used.
• Figure 5.25: Comparing viewsheds for hillfort and enclosure sites. Top row compares viewsheds for (a) Boley Upper hillfort, (b) Clomantagh hillfort, (c) Tintore enclosure. Second row compares viewsheds for (d) Cloneybrien hillfort, (e) Knigh hillfort, (f)
• Figure 5.26: Summary statistics for cost surface values.
• Figure 5.27: Kolmogorov-Smirnov goodness of fit test results using the known point data and one set of randomly generated points.
• Figure 5.28: Distribution of all domestic sites showing the number of houses identified within each settlement.
• Figure 5.29: Nearest Neighbour analysis for settlement sites. The null hypothesis is that no spatial pattern exists among the eatures. The Z Score represents the measure of standard deviation from the normal distribution: negative values signify clusteri
• Figure 5.30: Ripley’s K function performed on settlements within Ireland (with 50 distance bands increasing in 1 km intervals from a distance of 1 km from the each sites; 99 permutations to give a 99% confidence level; expected values are shown with the
• Figure 5.31: Ripley’s K function performed on settlements within Ulster (with 50 distance bands increasing in 1km intervals from a distance of 1 km from the each sites; 99 permutations to give a 99% confidence level; expected values are shown with the ce
• Figure 5.32: Ripley’s K function performed on settlements within Munster (with 50 distance bands increasing in 1 km intervals from a distance of 1 km from the each sites; 99 permutations to give a 99% confidence level; expected values are shown with the
• Figure 5.33: Ripley’s K function performed on settlements within Leinster (with 50 distance bands increasing in 1 km intervals from a distance of 1 km from the each sites; 99 permutations to give a 99% confidence level; expected values are shown with the
• Figure 5.34: Domestic sites with Early, Middle and Late Bronze Age activity and buffer zones in the Rivers Boyne and Liffey area.
• Figure 5.35: Proximity between burials and domestic settlement.
• Figure 5.36: Radiocarbon dates of excavated burial and settlement sites located within 100 m of each other.
• Figure 5.37: Proximity between burnt mounds and domestic settlement.
• Figure 5.38: Radiocarbon dates of excavated burnt mounds and settlement sites located within 100 m of each other.
• Figure 5.39: Late Bronze Age hoards.
• Figure 5.40: Proximity between artefacts and hoards (with provenances), stone monuments and domestic settlement.
• Figure 5.41: Cross-L analysis for settlement sites and burial sites. The X axis represents distance in km and the Y axis represents the Cross-L value.
• Figure 5.42: Cross-L analysis for settlement sites and other monuments (the X axis is distance in km).
• Figure 5.43: Cross-L analysis for Middle Bronze Age settlement sites and hillforts and hilltop enclosures (the X axis is dis ance in km).
• Figure 5.44: Cross-L analysis for Late Bronze Age settlement sites and hillforts and hilltop enclosures (the X axis is dista ce in km).
• Figure 5.45: Summary of Middle and Late Bronze Age sites and their proximity to other monuments.
• Figure 6.1: Schematic ground plans (after Ó Néill 2009: figs 5, 7, 8). From left to right, top: GP1, GP2, GP3; bottom: GP4, GP5, GP6.
• Figure 6.2: Distribution of ground plans.
• Figure 6.3: Details for some of the structures and their associated radiocarbon dates, size and approximate floor size (where multiple radiocarbon dates were available for individual structures the date with the longest range was chosen).
• Figure 6.4: Median of the available and associated radiocarbon date against the approximate roofed floor area; trend line added.
• Figure 6.5: Frequency of hearths from sites belonging to unknown and known chronological periods.
• Figure 6.6: Probability density of summed dated hearths.
• Figure 6.7: Details of examples of different porch types. *The standard porch at Ballybrowney was undated but associated with Middle Bronze Age pottery, which was recovered from an adjacent roundhouse which was radiocarbon dated to the Middle Bronze Age
• Figure 6.8: Examples of different porch types:(a) Cuffsborough 4, S2 with no porch (Murphy 2009: fig 10); (b) Clonadacasey 2, S1 and S2 with standard porches (O’Neill 2008: fig 8); (c) Corrstown, S4 with an elongated porch (Ginn and Rathbone 2012: illus.
• Figure 6.9: Distribution of houses with very large porches.
• Figure 6.10: Main domesticates at excavated sites with significant faunal assemblages. (from left to right: Murphy 2003; McCormick 1987; Sloane 2009; McCormick 2007; Beglane 2010; Murphy and McCormick 1996; Boner 2001;McCarthy 2012; van Wijngaarden-Bakker
• Figure 6.11: Saddle querns (SQ), loom weights (LW) and spindle whorls (SW) from sites. Precise quantities have been given for sites with multiple examples.
• Figure 6.12: Percentage of charred grain species for sites with specialist reports. (a) From top left, reading across: Johnston 2004a; Johnston 2004b; Halwas 2007b; ASDU 2009a; Johnston 2004c; Halwas 2007c; Halwas 2007c; ASDU 2008; ASDU 2009b; McClatchie
• Figure 6.13: Geographic distribution of the substantial cereal assemblages, and their breakdown. (a) Middle Bronze Age.
• Figure 6.13: Geographic distribution of the substantial cereal assemblages, and their breakdown. (b) Late Bronze Age.
• Figure 6.14: Ballydrehid plan (after McQuade 2007b, plate 6).
• Figure 6.15: Killydonoghoe grain storage pits which post-dated a roundhouse (after Sherlock 2005, plate 6).
• Figure 6.16: Radiocarbon dates from four-post structures.
• Figure 6.17: Radiocarbon dates from Middle Bronze Age sites with evidence of significant grain deposition.
• Figure 6.18: Details of field systems, and possible domestic sites.
• Figure 6.19: Distribution of sites with potential exchange valuables.
• Figure 6.20: Details of the sites with notable artefacts which are shown in Figure 6.19.
• Figure 6.21: Moulds from domestic sites.
• Figure 6.22: Examples of sites with locally derived and non-locally derived material (according to specialist reports).
• Figure 6.23: Location of prominent minerals and domestic sites, in the Republic of Ireland (mineral locations for Northern I eland were not available).
• Figure 6.24: Distance of sites in the Republic of Ireland to major mineral outcrops (mineral locations for Northern Ireland were not available).
• Figure 6.25: Composite of enclosed sites (top left: Ballybrowney (after Cotter 2006, plate 2), right: Boyerstown 3 (after Clarke 2009, plate 18); bottom left: Dunbell Big (after Whitty 2009, fig. 5), middle: Rathmullan 10 (after Bolger 2000, plate 13),
• Figure 6.26: Radiocarbon dates from enclosures and enclosing elements.
• Figure 6.27: Sum of dated enclosures.
• Figure 6.28: ‘Single-generational’ and ‘multi-generational’ structures (top left: Adamstown 3 (after Russell and Ginn 2009, plate 5); bottom left: Skreen 3 (after O’Neill 2009, plate 12); right: Corrstown (after Ginn and Rathbone 2012, illus. 2.15)).
• Figure 7.1: The Mooghaun Model (after Grogan 1999, fig. 11). Different territorial extents were defined for each of the three tiers in the model (from 1.5–3 km2 and 12–25 km2 to 450 km2). Less intensively ordered zones are present between the territorie
• Figure 7.2: The case-study areas.
• Figure 7.3: Case Study 1: Tipperary.
• Figure 7.4: Case Study 2: The Boyne.
• Figure 7.5: Case Study 3: The Midlands.
• Figure 7.6: Case Study 4: The North Coast. The Lissanduff earthworks consist of a pair of concentric embanked enclosures, one with a central dry area and one with a central, spring-fed pool. Their date is unknown, but they are considered to be prehisto i
• Figure 7.7: Case Study 5: The River Suir.
• Figure 7.8: Case Study 6: The Wicklow Mountains.
• Figure 7.9: The Mooghaun Model applied to Case Study 1, Tipperary.
• Figure 7.10: The Mooghaun Model applied to Case Study 2, The Boyne.
• Figure 7.11: The Mooghaun Model applied to Case Study 3, The Midlands. The buffer zones of hillforts Ballycurragh, Cumber Lower and Letter can be seen in the bottom right of the case-study area.
• Figure 7.12: The Mooghaun Model applied to Case Study 4, The North Coast.
• Figure 7.13: The Mooghaun Model applied to Case Study 5, The River Suir.
• Figure 7.14: The Mooghaun Model applied to Case Study 6, The Wicklow Mountains.
• Figure 7.15: Natural boundary territories applied to Case Study 3, The Midlands.
• Figure 7.16: Central Place Theory. The coloured blocks represent different hillfort territories based on cost allocation and the polygons are Thiessen polygons.
• Figure 7.17: Core-periphery model applied to Case Study 1, Tipperary:
• Figure 7.18: Core-periphery model applied to Case Study 2, The River Boyne.
• Figure 7.19: Core-periphery model applied to Case Study 3, The Midlands.
• Figure 7.20: Summary of the main findings from the case-study areas.
• Figure 7.21: Least-cost paths in Case Study 1, Tipperary.
• Figure 7.22: Least-cost paths (LCP).
• Figure 7.23: Major Bronze Age trackway concentrations.
• Figure 7.24: Early medieval routeways based on Ó Lochlainn (1940).
• Figure 7.25: Routeways in Case Study 2, The River Boyne.
• Figure 7.26: Routeways in Case Study 3, The Midlands.
• Figure 7.27: Routeways in Case Study 4, The North Coast.
• Figure 7.28: U/D-shaped houses in Munster. Bottom left: Ballylegan 207.1 (after McQuade 2007a, plate 21); bottom right: Ballydrehid (after McQuade 2007b, plate 10) (see Figure 6.6d, e and g).
• Figure 7.29: Houses with segmented or double ditches. Bottom left: Cappagh Beg (Linanne 2003, fig. 10); middle: Grange 3 (Kelly 2010, fig. 8); right: Knockdomny (Hull 2007, fig. 3).
• Figure 7.30: Possible regions as defined by the settlement and case-study evidence.
• Figure 8.1: The Middle Bronze Age settlement network (via TravellerSim model (Graham 2008)). The bottom left line is a 20 km scale bar. The larger sites (and therefore the most important) include Chancellorsland A, Ballynahinch 3, Ballydrehid, Killoran 8,
• Figure 8.2: The Late Bronze Age settlement network (via TravellerSim model (Graham 2008)). The bottom left line is a 20 km scale bar. Larger sites, and therefore more important sites, include Carrigatogher (Harding) 5, Boscabell 19, and Loughfeedora. The
• Figure 8.3: Hillfort network (via TravellerSim model (Graham 2008)). The bottom left line is a 20 km scale bar. Larger sites and therefore the most socially connected include the hillforts between Leinster and Munster. Shading differentiation is not marke
• Figure 8.4: The hillfort and Late Bronze Age network (via TravellerSim model (Graham 2008)). The bottom left line is a 20 km scale bar. That many of the sites are large indicates that many are prominent within the network. No clusters of shading different
• Figure 8.5: Descriptive statistics from Freeman’s Degree Centrality function (see Hanneman and Riddle 2005; produced in UCINET (Borgatti et al. 2002)).
• Figure 8.6: Sites shown to be influential or prominent by Freeman’s Degree Centrality: (a) Middle Bronze Age; (b) Late Bronze Age; (c) Late Bronze Age and hillforts.
• Figure 8.7: Descriptive statistics from the Reach Centrality function (see Hanneman and Riddle 2005; produced in UCINET (Borgatti et al. 2002)).
• Figure 8.8: Sites shown to be well-placed according to the Reach Centrality function: (a) Middle Bronze Age; (b) Late Bronze Age; (c) Late Bronze Age and hillforts.
• Figure 8.9: Results of the Reach Centrality function (see Hanneman and Riddle 2005; produced in UCINET (Borgatti et al. 2002)).
• Figure 8.10: Sites shown to be well-placed according to the Eigenvector function: (a) Middle Bronze Age; (b) Late Bronze Age; (c) Late Bronze Age and hillforts.
• Figure 8.11: Descriptive statistics from the Eigenvector function (see Hanneman and Riddle 2005; produced in UCINET (Borgatti et al. 2002)).
• Figure 8.12: Descriptive statistics from the Freeman Betweenness Centrality (Node Betweenness) function (see Hanneman and Riddle 2005; produced in UCINET (Borgatti et al. 2002)).
• Figure 8.13: Sites shown to be well-placed according to the Freeman Betweenness Centrality: (a) Middle Bronze Age; (b) Late Bronze Age; (c) Late Bronze Age and hillforts.
• Figure 8.14: Sites shown to be well-placed according to the Flow Centrality function: (a) Middle Bronze Age; (b) Late Bronze Age; (c) Late Bronze Age and hillforts.
• Figure 8.15: Descriptive statistics from the Flow Centrality function (see Hanneman and Riddle 2005; produced in UCINET (Borgatti et al. 2002)).
• Figure 8.16: Sites which emerged as important throughout the entire SNA: (a) Middle Bronze Age; (b) Late Bronze Age; (c) Late Bronze Age and hillforts.
• Figure 8.17: Middle Bronze Age settlement network as defined by connectivity and centrality (produced in Gephi (see Bastian e al. 2009)). (Differential colouring in the model output represents differences in degrees of connectivity; unfortunately, some o
• Figure 8.18: Middle Bronze Age settlement network as defined by community and influence (produced in Gephi (see Bastian et al. 2009)). (Differential colouring in the model output represents differences in communities; unfortunately, some of these nuances
• Figure 8.19: Late Bronze Age settlement network as defined by connectivity and closeness (produced in Gephi (see Bastian et al. 2009)). (Differential colouring in the model output represents differences in degrees of connectivity; unfortunately, some of
• Figure 8.20: Late Bronze Age settlement network as defined by community and influence (produced in Gephi (see Bastian et al. 009)). (Differential colouring in the model output represents differences in communities; unfortunately, some of these nuances ar
• Figure 8.21: Close-up of the components of the LBA network shown in Figure 8.8 (produced in Gephi (see Bastian et al. 2009)). Only the most important sites, as defined by the model, have been labelled.
• Figure 8.22: Late Bronze Age settlement and hillfort network as defined by connectivity and closeness (produced in Gephi (see Bastian et al. 2009)). (Differential colouring in the model output represents differences in degrees of connectivity; unfortu ate
• Figure 8.23: Late Bronze Age settlement and hillfort network as defined by community and influence (produced in Gephi (see Bastian et al. 2009)). (Differential colouring in the model output represents differences in communities; unfortunately, some of the
• Figure 9.1: Percentage of single artefacts and hoards found within each elevation zone.
• Figure 9.2: Distribution of artefact types: (a) Middle Bronze Age.
• Figure 9.2: Distribution of artefact types: (b) Late Bronze Age.
• Figure 11.1: Propositions (Chapter 3): summary and analysis.