This electronic thesis or dissertation has beendownloaded from Explore Bristol Research,http://research-information.bristol.ac.uk

Author:Wilson, Jacqueline

Title:Becoming 'Irish' : The materiality of transcultural identities in the Later Irish Iron Age

General rightsAccess to the thesis is subject to the Creative Commons Attribution - NonCommercial-No Derivatives 4.0 International Public License. Acopy of this may be found at https://creativecommons.org/licenses/by-nc-nd/4.0/legalcode This license sets out your rights and therestrictions that apply to your access to the thesis so it is important you read this before proceeding.

Take down policySome pages of this thesis may have been removed for copyright restrictions prior to having it been deposited in Explore Bristol Research.However, if you have discovered material within the thesis that you consider to be unlawful e.g. breaches of copyright (either yours or that ofa third party) or any other law, including but not limited to those relating to patent, trademark, confidentiality, data protection, obscenity,defamation, libel, then please contact collections-metadata@bristol.ac.uk and include the following information in your message:

•Your contact details•Bibliographic details for the item, including a URL•An outline nature of the complaint

Your claim will be investigated and, where appropriate, the item in question will be removed from public view as soon as possible.

Becoming 'Irish':

The materiality of transcultural identities

in the Later Irish Iron Age

Jacqueline Wilson Department of Archaeology and Anthropology

University of Bristol

A dissertation submitted to the University of Bristol in March 2010, in accordance with the requirements of the research degree of

Doctor of Philosophy (PhD) in the Faculty of Arts

Word Count: 82,942

ABSTRACT

The materiality of `Celtic' Iron Age Ireland has been characterised as enigmatic and

problematic and its people as essentially invisible to the archaeologist. The creation of

the present culture-historical narrative has resulted in the use of typological and

chronological classifications that do not account for localised development and localised responses to external cultural influences during the period. A critical

assessement of the literary evidence from the Irish manuscripts and the Classical

sources is presented against a background of how and why the existing nationalist

narrative was formulated. Following recent dialogues in Roman studies that have

recognised the variability in responses to being Roman, even in the provinces, a new theoretical framework is outlined around the concept of transcultural identities. The

application of a theoretical approach which is inclusive, rather than separatory, allows

`intrusive' material to be reconciled as an essential part of the Irish narrative but

within a more holistic social history for the period. A new classification series has

been designed for the Iron Age, to separate that which we know (through absolute or

secure diagnostic parallels of similar material elsewhere), from that which remains

open to further analysis. All of the imported material for the period has been

reclassified in line with the new system for the longer Iron Age. Using strontium and

oxygen isotope analysis and new 14C dates a scientific study was undertaken on burials which have also been classified as ̀ intrusive'. There is no correlation between

new 14C dates and the relative dates proposed for these burials based on burial

morphology. Not all of the ̀ intrusive' burials were of foreigners and some of those

classified as local people originated in the Roman provinces, and were (until now) indistinguishable in death. We can no longer classify people and artefacts as `foreign' or `intrusive' and treat them as an adjunct to the ̀ Irish' or `Celtic' Iron Age,

as in reality, these are an essential part of the transcultural materiality of the period.

Acknowledgements

I could not have completed this study without the constant love and support provided

by Paul, Tippi and Peter so thank you for putting up with me. Colleagues, staff and

friends at the University of Bristol have provided the laughter, support and friendship

needed to complete the work. Professor Mark Horton and my adoptive supervisor Dr

Alistair Pike have provided all the inspiration and help a student could need, and I

thank you both. My grateful thanks also go to all the staff at the National Museum of

Ireland but especially to Ned Kelly, Maeve Sikora and Laureen Buckley who were

incredibly accommodating during my frequent trips to Dublin. Professor George

Eogan and Dr Kerri Cleary at UCD always greeted me with smiles, despite their own

heavy workloads and I am indebted to them. To Hege, Carolyn, Claudia, Phil, Sian

and my undergraduate students thank you for sharing my enthusiasm for all things

Roman. Grateful thanks also go to my sponsors the Arts and Humanities Research

Council who provided an award to complete this research and to the Royal Irish

Academy for funding four 14C dates as part of the research.

A very special thank you is reserved for two formidable archaeologists: Dr Tamar

Hodos, my supervisor and Dr Elizabeth O'Brien, my unofficial mentor. Dr Hodos'

approach to supervision is founded on a support base of warmth, constancy and intellectual rigour and I hope that the finished work is worthy of her role in its

creation. Without the ground breaking work completed by Dr O'Brien over many

years my research would not have been possible. Her willingness to contribute her

valued opinion to my own work and her enthusiasm for collaboration will always be

remembered. Their constant encouragement serves to remind me that truly great

scholars have big hearts as well as open minds. Go raibh mile maith agaibh a chairde.

"I declare that the work in this dissertation was carried out in accordance with the Regulations

of the University of Bristol. The work is original, except where indicated by special reference in the text, and no part of the dissertation has been submitted for any other academic award. Any views expressed in the dissertation are those of the author.

...... Icýy......... DA'I'S......... ...... SIGNED* . 11

Contents

Chapter 1 Introduction

1.1 Introduction

Chapter 2 Literary Evidence and the History of Scholarship

1

2.1 Irish manuscript evidence 15 2.2 The Classical sources 23 2.3 The study of Roman material in Ireland 28 2.4 The historical narrative 34 2.5 The Harvard Expeditions 38 2.6 The ̀ Celtic' Debate 40

Chapter 3 Fixing the Past

3.1 Identity, origin and materiality 46 3.2 Identity in life and death 53 3.3 Romans and ̀ Romanisation': a critical dialogue 59 3.4 Static boundaries or fluid frontiers 62 3.5 Tranculturality as an analytical tool 67

Chapter 4 Problems with the Iron Age Transition

4.1 The Longer Irish Iron Age: problems with dating temporal transition 70

4.2 Relative chronology: artefact classification and typology 71 4.3 Hencken's chronology and typology 77 4.4 Refining the chronology with new absolute dates 85 4.5 Circularity in interpretation 97 4.6 A classificatory system for the Iron Age 98

Chapter 5 Imported Material in Ireland

5.1 Reclassifying Roman material from Irish sites 102 5.2 The later imported ceramics 126 5.3 A reconsideration of the ̀ Hibernia Romana' debate 132

Chapter 6 Death and Burial in the Later Irish Iron Age

6.1 Background to the Research 137 6.2 Burial in the Later Irish Iron Age 138 6.3 Transition in burial practice 143 6.4 `Intrusive' burials 146 6.5 Sites chosen for scientific analysis 150 6.6 Site B, Carbury Hill, County Kildare 152 6.7 Knowth, County Meath 154 6.8 Rossnaree, County Meath 158 6.9 Betaghstown, County Meath 159 6.10 Räith na Senad, Hill of Tara, County Meath 162

Chapter 7 Strontium and Oxygen Isotope Analysis on ̀ Intrusive' Burials

7.1 The archaeological background to the sampling strategy 166 7.2 Overview of strontium and oxygen isotope research 167 7.3 Carbury Hill, County Kildare 178 7.4 Knowth, County Meath 183 7.5 Rossnaree, County Meath 190 7.6 Betaghstown, County Meath 192 7.7 Rd ith na Senad, County Meath 197 7.8 Review of the analysis and discussion 201

Chapter 8 Towards a Materiality of Transcultural Identities in the Later Irish Iron Age

8.1 Overview 206 8.2 Challenging the underlying constructs and assumptions 206 8.3 The environment and the economy 212 8.4 Trade and imports 214 8.5 Settlement and society 221 8.6 Ritual and religion 226 8.7 Language and literacy 231 8.8 At the edge of the known world 232 8.9 Becoming 'Irish': a transcultural process 234

List of Illustrations

Chapter 1. Introduction

Chapter 2 The Literary Evidence

2.1. Extract from 12th century Irish manuscript showing translation of Plato's Timaeus, with Latin annotations (after O'Cröinfn 2005)

2.2. Extract from the Book of Ballymote, showing detailed

explanation of the ogham script (image courtesy of www. isos. dias. ie )

2.3 Ptolemy's 2nd century AD map of Ireland, reconstructed from co-ordinates (after Lynn 2003)

Chapter 4 Problems with the Iron Age Transition

4.1 Left distribution map of Dowris Phase Late Bronze Age artefacts (after Woodman 1998) and at right, distribution of Halstatt style objects (after Raftery 1994)

4.2 Distribution of La Tene artefacts (after Harding 2007) redrawn by author

4.3 Illustration from Wood-Martin (1884) of Roman type swords spears and axes from Lagore

4.4 Univallate ringfort at Knowth, County Meath 4.5 Published radiocarbon dates modelled using OxCal 4.1, profiles

showing periods of activity at ringforts

Chapter 5 Imported Material in Ireland

5.1 Left, distribution map of accepted finds from Bateson 1973 and at right, distribution map of accepted finds from Raftery 1994

5.2 The Ballinrees Hoard (image courtesy of the British Museum) 5.3 Toilet implements and jewelleryfrom Freestone Hill, not to scale

(after Raftery 1994) 5.4 Plate brooch with incised decoration and conical glass inset,

from Newgrange (after Carson and O'Kelly 1977) 5.5 Plate brooch from Uley with incised decoration and conical glass inset

(after Woodward & Leach 1993) 5.6 The inscribed gold torc end from Newgrange (after Freeman 2001)

5.7 The two imported sherds from Freestone Hill, County Kilkenny 5.8 Decorated strap-end from Rathgall, not to scale (after Raftery 1994) 5.9 The impressive hillfort of Rathgall, County

Wicklow (after Raftery 1994) 5.10 The Stoneyford Roman cremation. Top, the glass vessel used as

a cinerary urn, bottom, the bronze plate with which it was sealed and the glass ungent jar which accompanied the burial

5.11 The collyrium stamp found at Golden in County Tipperary 5.12 Aerial view of Drumanagh, showing rectangular footings of

buildings (after Rafery 1972) 5.13 The 2nd century AD enamelled seal-box found by metal

detectorists at Drumanagh 5.14 The Latin inscribed stone from Lough Dearg (after Macalistair

1945) 5.15 Updated map of the distribution of Roman material in Ireland

(drawn by author) 5.16 Map of key sites mentioned throughout discussion, the greater

Meath environs highlighted with rectange (drawn by author) 5.17 Survey results for the enclosure overlooking the River Boyne,

between Knowth and Newgrange (Warner 2009)

Chapter 6 Death and Burial in the Later Irish Iron Age

6.1 Crouched burial KnH9 from Knowth (after Eogan 1974) 6.2 Double burial KnH27/28 from Knowth, extended unprotected

inhumations (after Eogan 1974) 6.3 Protected extended inhumation in stone cist, BTS II

Betaghstown (after O'Brien 1990) 6.4 Topographical map of the greater Meath environs (including parts

of north County Dublin and County Kildare. Sites with intrusive burials chosen for scientific analysis and 14C dating (drawn by author)

6.5 Plan of Carbury Hill, Site B, showing cremations and inhumations (after Wilmot 1938, reproduced in Raftery 1994)

6.6 An aerial view of the passage tombs of Newgrange and Knowth in the Boyne Valley (Image courtesy of Knowtk com)

6.7 Plan of Iron Age and Early Historic burials around the great mound at Knowth (after Eogan 1974 and Raftery 1994)

6.8 Aerial view of the sandy coastline of County Meath. The site of Betaghstown, the modern city of Drogheda and the passage tombs are shown along with the curve of the River Boyne

Chapter 7 Strontium and Oxygen Isotope analysis on ̀ intrusive' burials

7.1 Measured 6180, converted to meteoric water equivalent 8'8Odw using conversion equations published by Longinelli et al (1984), Levinson et at (1987), Luz et at (1984) and Daux et al (2008)

7.2 Annual mean average 5180 SMOW isotope ratios recorded by the Global Network for Isotopes in Precipitation (GNIP), image courtesy of www. iaea org

7.3 Oxygen isotope values for modern drinking water (51800 left Ireland and the UK and right with continental Europe.

Images courtesy of NERC Isotope Geosciences Laboratory 7.4 Left, Oxygen Isotope Ireland (ä8O ). Redrawn by author

based on survey data published by Diefendorf and Patterson (2005) 7.5 The geological sequences for the greater Meath environs, (includes

parts of north County Dublin and County Kildare). After McConnell et al (2001)

7.6 Topographical map of the greater Meath environs (includes part of north County Dublin and County Kildare). After Newman (2005) redrawn by author

7.7 Carbury Hill, showing 618O plotted against results for 87SrA6Sr 7.8 Knowth, showing 5180 plotted against results for 87Sr186Sr 7.9 Betaghstown showing b180, plotted against results for 87Sr/86Sr 7.10 Rc ith na Senat, showing 6180 plotted against results for 87Sr186Sr 7.11 The results from all sites showing expected parameters for

strontium and oxygen isotope data in County Meath and its environs

Chapter 8 Towards a Materiality of Transcultural Identities in the Later Irish Iron Age

8.1 Updated distribution map of finds, including only those burials for which we can now demonstrate an origin outside Ireland and a radiocarbon date for the period

82 Distribution of natural resources based on Geological Survey of Ireland data (drawn by author)

8.3 5th and 6th century AD imports from the Mediterranean, and 7th century AD imports of E ware

8.4 The imposing site of the Rock of Cashel, County Tipperary, political seat of the Eöganacht in the 4th century AD

8.5 Clonycavan Man (photo by Robert Clark NMI)

Tables

4.1 Radiocarbon dates from ringforts in southern Ireland

5.1 Record of coin finds from Newgrange (Carson and O'Kelly 1977)

5.2 Campbell's classification of later imported wares found in Britain and Ireland (2007): based on Radford (1956), Thomas (1959), Rigoir et al (1973), Kelly 2010

7.1 Details of human and faunal samples collected under Licence to Export and associated Licence to Alter (No. 3983), granted by the National Museum of Ireland (NMI). Additional analyses completed at the request of staff at the NMI and Heritage Council of Ireland and were covered by extensions to existing licences

7.2 Carbury Hill, Site B

7.3 Knowth, County Meath

7.4 Rossnaree, County Meath

7.5 Betaghstown, County Meath

7.6 Räith na Send Hill of Tara, County Meath

7.7 Details of all human samples showing sex and burial type, classification based on relative dating methods, previous classification whether intrusive or local, "C dates where known and suggested origin consistent with the results of the strontium and oxygen isotope analysis

Chapter 1 Introduction

"There is little evidence that the proximity of Rome had any significant effect on the development of Irish society. The intricate social system, the laws, the remarkable

vernacular oral tradition, which survived into the Middle Ages, were a legacy of its

prehistoric past undimmed by the shadow of Rome. " (CunlifJe 2001: 417)

1.1 Introduction

The archaeology of the first five centuries AD in Ireland is characterised by the opening

quotation from Cunliffe's analysis (2001), one of the key sources for which was `Pagan

Celtic Ireland: the enigma of the Irish Iron Age', published by Raftery in 1994. This

accepted culture-historical archaeological narrative of the Irish Iron Age colours even the

most recent studies of this period, within which authors stress the problematic and

seemingly ephemeral nature of the archaeological evidence with an implicit suggestion that one day the ̀ enigma' might be resolved.

This `missing' Later Iron Age is essentially the product of a traditionalist archaeological

approach that has languished within a chronology devised over 70 years ago. The relative

chronology used to date artefacts and sites for this period was devised by Hencken and the Harvard Expeditions' excavation programme in Ireland in the 1930s and 1940s

(discussed in detail in Chapter 2 and 4). In order to understand how the chronology was developed, consideration has been given to the social and political uncertainties that

prevailed at that time when those in Government in Ireland, in the early days of the Irish

Free State, were determined to validate a sense of nationhood fixed firmly within an indigenous ̀Celtic' past. The continued use of this relative chronology has resulted in

whole periods of so-called invisibility within the archaeological record as authors

struggle to make the archaeology fit within the historical fiamework, rather than

challenge the very premise of its continued use. There is no agreed beginning for the Irish Iron Age, nor is there any consensus on the terminology that should be used for discrete

periods within the Iron Age (O'Drisceoil 2008). Existing archaeological approaches

cannot accurately reflect the materiality of either the people or the places of importance

to them as the continued application of the existing chronology renders much material as

pre-dating the later Iron Age or, as in the case of the small homesteads known as ring-

forts, post-dating it by several centuries (these and the dating of other site-types are

discussed in detail in Chapter 4).

Central to any review of this period is a critical assessment of the literary evidence from

both the Irish manuscripts, known as the Annals and the Mythological Cycles, and from

the Classical sources. The first full translation into English from Gaelic of the Annals of the Four Masters in 1846 prompted considerable interest in antiquarian circles, but it was

not until the formation of the Irish Free State Government that excavation would be

directed towards finding an indigenous ̀Celtic' past through the excavation of the historic

sites described within them. An explanation of how the archaeological narrative was

originally formulated allows us to consider the relevance of the existing dating schemes

and to challenge their continued use within the literature, before attempting to redefine

the chronology within the longer Irish Iron Age. There are many references to Ireland

within the Classical sources and they offer much useful information on how Ireland was

perceived at this time by those in the Roman world, all of these will be discussed in detail

in Chapter 2.

According to the historical narrative, the people of the pagan Irish Iron Age were delivered into a Christian golden age by the missionary activities of Patricius, later St.

Patrick. The son of a Christian deacon, Patrick is said to have been captured in Roman

Britain, and held as a slave in Ireland for six years before he escaped to Gaul (Charles-

Edwards 2000). The date of his return to Ireland, traditionally recorded around AD 432,

is the point at which, as scholars, we are expected to move out of the Iron Age into the

next archaeological period, known variously within the archaeological literature as either the Early Christian period, the Early Medieval period or the Early Historic Period (c. AD

400/432 to 800) (the Early Historic period is preferred as a classification for the later

period in this study as explored and explained in detail in Chapter 4). There are two

2

pieces of work that are credited to St Patrick, the Confessio and Letter to the soldiers of

Coroticus, and these are widely used to reference his missionary activities (Charles-

Edwards 2000). Yet the commencement of the proselytizing Patrician missions remains

open to question, not least because we know from the Chronicle of Prosper (d. AD 455)

that a missionary called Palladius was sent to Ireland by an invigorated Roman Imperial

Christian administration in AD 431 (Charles-Edwards 2000). Whilst I do not intend to

rehearse the debate around whether Patrick even existed, or whether he was conflated

with Palladius into an idealised portrait of one individual (following O'Rahilly 1942,

1946), what is central to our consideration within this study is that there is no convincing

archaeological evidence to support this mid 5th century AD date for the wholesale

conversion of communities across Ireland to the new religion of Christianity. Despite the

constant re-affirmation of a narrative founded on the fluidity and homogeneity of an

independent `Celtic' Ireland that moved seamlessly towards Christian enlightenment, this

period is remarkable for its regional heterogeneity. As discussed in detail in chapters 4

and 5 the evidence suggests the continuation of Ireland's `pagan' customs into the 5"

century AD, which at some sites, such as Newgrange in County Meath, includes the

deposition of valuable Roman material.

Since Gudeman and Haverfield's controversial debate in the pages of the Classical

Review at the turn of the 20`h century there have been three attempts to assess Roman

material from Irish contexts (Gudeman & Haverfield 1900; Bateson 1973,1976; Raftery

1994; Warner 1995). These are all discussed in detail in Chapter 2 but it is important to

note here the overtly political reaction to the latest hypothesis proposed by Warner, as the

suggestion of Roman involvement in Ireland still provokes a nationalist response even

today. Warner argued that an exiled prince, Tiiathal Teachtmar, recorded within the

Annals as returning to Ireland with a military force in AD 78, was the same individual

recorded by Tacitus as having been befriended by Agricola in his fifth season of

campaigning in northern Britain. Central to his interpretation is the evidence from a site

called Drumanagh, in north County Dublin, where much Roman material has been found

in recent years (Warner 1995). This article was published within a small, specialised journal in Ireland but was picked up by the newspapers both in Britain and Ireland (see

3

Maas & Byrne 1996). The response to decry Warner's theories was swift and political,

with one author making a tangential link between Roman imperialism and British

imperialism, and suggesting that Warner's hypothesis would serve to damage the

northern Ireland peace negotiations taking place at that time (Cooney 1996). As a

consequence of the reaction to this debate all consideration of Roman influence on Ireland had ceased, until the present study. A critical reassessment of Warner's

hypothesis will be presented in Chapter 5, once all of the material has been reviewed.

A detailed consideration of all of the current theoretical frameworks that are used to discuss non-local material within localised contexts, but especially Roman material in

those areas that remained outside Roman control, will be outlined in Chapter 3, before a

new theoretical framework is proposed, within which all of the material from Ireland may be rehabilitated within an archaeological narrative that reflects contemporary research for

the period. The ripples of Roman cultural influence spread far beyond those areas known

to have been under direct control and administration but it is only in recent years that the dialogues in Roman archaeology have focussed on understanding that impact through the

responses at a local level (Hingley 1997,2001,2005; Gosden 2005; Mattingly 1997,

2004,2006). In the past the outcome of Roman influence was generalised through the

process of `Romanisation', focussed upon the apex of society and modelled on

predictability and homogenity, the innately civilizing influence of which was

characterised by straight roads, thermae and mortared buildings (see Millett 1990). Yet it

is clear from developments within this new post-imperial, post-modern Roman

archaeological discourse that authors have now recognised that Roman-ness does not

always follow a predictable social or economic trajectory. The multicultural nature of the Empire presents a variety of subtle cultural negotiation and social and technological

changes that can be evidenced in areas that never fell under direct control. These debates

are essential to how we view the Later Iron Age in Ireland, as it is the lack of a

predictable outcome in the form of forts, roads and buildings which has been used in the

past to completely reject both the existing evidence for Roman material and any long- term cultural influence on communities and power-brokers in Ireland. Language and literacy are two key elements in the adoption or recreation of cultural habits amongst

4

local people and changes taking place within the Irish language at this time reflect both

the early borrowing of Latin loan words into spoken Irish and the creation of an Irish

epigraphical habit, the ogham script. Scholars continue to debate the probable dating of the ogham with the most recent contributions supporting its use in Ireland centuries before its current dating of the 4th or 5th century AD (Stevenson 1987; Harvey 1989,

1987; McManus 1991; Swift 1997). The most recent debates will be considered in detail

in Chapter 2, as they are an essential part of any consideration of Roman cultural influence on Ireland.

It is necessary to move forward from the academic arguments around culture-contact that have centred on theoretical constructs such as assimilation and acculturation. The use of these terms has been successfully challenged, through the recognition that they reflect an interpretation founded on the premise of fixed or immovable constants, the outcome of

which was, in a sense, predicated upon this constancy (Gosden 2004). Furthermore,

recent work on the circulation of the later imports in Britain and Europe has demonstrated

an association with other aspects of the materiality of the period, opening up the debate to

consider distribution and location as essential to the debate (Campbell 2007); this will be

discussed in detail alongside the material evidence in Chapter 5. We also need to address the historiography around Hibernia in provincial Roman studies, and the persistent

application of a quantitative method for diagnostic ceramics, used to dispute any significant trade between Ireland and the Roman economy (Fulford 2005; Campbell 2007). The types of evidence from Ireland will be discussed in detail as they demonstrate

a complex interplay between the consumption of exotic items, pins beads and brooches

that betray some development or vernacularisation into `Irish' forms. Alongside these, however, appears to be the localised consumption of imported material, ceramics and items of jewellery, used within a local context but not adopted more widely.

In order to demonstrate the importance of an inclusive archaeology, it became necessary to find a theoretical framework that would allow for a positioning of the material, local

and non-local, as meaningfully constituted within the Irish Iron Age. It follows that in

exploring the social and cultural significance of non-local evidence through

5

contemporary theoretical models, attention must be given to the importance of concepts

such as the role of agency, and how this informs adoption, emulation or rejection at a local level. If the material is read through the malleability of identities, through the

manipulation of social signifiers both at the level of the individual, and the community

experience, new insights emerge of those who were living and dying in Ireland at this

time. When new influences are recognised through the consumption of non-local

material, or there is clear adaptation to existing material culture, or there is a transition in

previously long-lived customs such as burial practices, we need to ask questions about

why and how these were taking place. Critical awareness and recognition of previous

political or social constraints within the field of archaeology does not in itself lead to

radical changes to the existing classification; nor should it. But changes and transition in

established cultural reference points do not just emerge; they are ̀ socially constructed'

and ̀ meaningfully constituted' (Hodder 2004: 156). Once this is accepted, only then can

we explore fully whether these changes taking place in Ireland are consequent upon direct influence from Roman-Britain.

Following Horning's work on the archaeology of conflict in Northern Ireland and the

work of scholars such as Featherstone (1990,1999), McDonald (2003) and Welsche

(1999), a theoretical framework structured around the concept of transculturality is

presented in Chapter 3 that offers a new way forward in interpreting the materiality of the

period. The application of a theoretical framework such as this encourages a mechanism by which the materiality of the ancient past can be reconciled within an acceptable

archaeological narrative. It also allows for a recognition that part of the existing narrative is embedded within a nationalist stance, and in order to assess the relevance of non local

material we need to refract that part of its history which has been subsumed within a

grand narrative. As this is explored in detail in Chapter 3, suffice to say for the moment

that in its most transparent form, transculturality as a conceptual and analytical framework allows us to reject the concept of delimited and bounded ̀cultures', which have in the past been used as a separatory construct, and a default for ethnic groupings. Viewed through a lens which accepts that cultural influence is as likely to result in

divergence as coalescence, the localised responses within the archaeology can be

6

recognised as evidential rather than speculative. This framework allows us to explore how cultural influence can impact at the level of both the individuals, and their

communtites, but also to help us understand how aspects of material culture can run in

parallel with each other, occasionally touching or diverging. It also recognises that such

culture contact points did not occur across all of Ireland, and that in many respects in

parts of Ireland, such as in the western counties and much of the south-west, social

change appears to have occurred later to that which was happening elsewhere in Ireland.

The variability in response to change is of equal importance, as it is the trajectory of

change and development, which informs any consideration of regional development in a

temporal sense.

Although some scholars have argued that the stimulus for change and development needs

to be seen as a reflection of Ireland's connections to the economy of the wider Roman

world (Warner 1994; Raftery 1994; Newman 2005), this has not gained universal support in Ireland. Before exploring this further, consideration needs to be given to the more

recent dialogues taking place within Roman studies and the study of Classics (Hales and Hodos 2010). It is now recognised that there were many types of Romans and immense

variability in responses to `Romanisation' across Europe. The use of the term

`Romanisation' has been critically assessed in recent work, and the once common

narrative that equated those who were perceived as ̀ Roman' to be civilized and those

who were not as ̀ barbarian' has been completely discredited. This post-modern, post- imperialist debate has sought to understand the response to Roman cultural influence

through the discrete archaeology of regions of the island of Britannia and beyond her

shores. Even within the civitas capitals, small towns and local homesteads, response to

Roman influence in the provinces does not follow a set linear progression or expected

social trajectory of development. These localised, and eminently variable social patterns

are part of a more widespread discrepant response to Roman cultural influence, and they

reflect the mutable and negotiated nature of identities as reflected within the materiality

of Roman Britain (Mattingly 1997,2004,2006; Hingley 1997,2010). This will be

expanded and explored fully in Chapter 3.

7

Postulated parallels with the material culture and social structures of those beyond the

Roman frontiers of north-western Europe will also be explored in Chapter 3, in order that

we might assess whether Ireland was, as suggested, peripheral to that which was happening elsewhere in Roman Europe. The work of scholars such as Lund Hansen

(2003), Hedeager (1992), Roymans (1990), Fulford (1977 & 1985), Gardner (2001) and Mattingly (2004 & 2006) will be used to explore how the evidence of Roman material has been interpreted in other societies, whose communities lay beyond the frontiers, but

who were still engaged within zones of interaction. Following on from these debates,

consideration will be given to whether there was, at this time, `varieties of Irishness', as

reflected in the variability of the material from across the different provinces (Curtin

1996; Mattingly 2004 & 2006; Hunter 2001; Woolf 1997 & 1998).

In order that an alternative view of social, economic and political development might be

explored, a critical re-evaluation of the actual material evidence for the longer Irish Iron

Age is necessary, before moving to discuss the material of the later period in detail. The

first five centuries AD will be referred to as the Later Irish Iron Age throughout this

study, despite the fact that the greater part of the 5th century AD is traditionally

incorporated within the Early Historic Period. The existing chronology for the Iron Age

includes at its commencement the transition from the Later Bronze Age, and the immense

variability evidenced at a regional level throughout Ireland for the earlier period will be

discussed in detail in Chapter 4. Remarkably, this graduated transition and different

social and economic trajectory is accepted for the commencement of the Iron Age, but

not for its end. This problem is highlighted by the approach taken in two new reports

compiled by staff at University College Dublin and published on-line: Finding an Invisible People (Becker et al: 2008), is a comprehensive listing of newly excavated sites dating to the Iron Age and it includes some new radiocarbon dates; the Early Medieval

Project (O'Sullivan et al: 2007) provides an overview of both the later Iron Age and the

later period. There is significant overlap between both reports, although this is not

addressed in detail within either. Both ignore the relevance of the imported material from

the early centuries and the later Roman ceramics are classified within the `Early

Medieval' report. Both reports follow the existing chronological model despite

8

mentioning new work and new research that directly challenges the basis on which these

were devised. The data contained within both reports will be used throughout this work, but in order to make the material more accessible, some realignment is needed.

There is no clear cessation of the Iron Age that can be demonstrated archaeologically; in

fact there is significant overlap and continuity. Central to our understanding of how and

why different communities in Ireland responded differently to Roman cultural influences

is a consideration of all of the imported material as essential and evidential, rather than as

an adjunct treated separately to a ̀ local' Irish material of the period. A reclassification

series for the Later Iron Age has been devised within this study, and it includes all of the

material from the 5t' century AD. It is simple and transparent and offers a clear guide

mark for scholars on how both material and sites have been dated in the past and the

problems surrounding their relative dating. All of the material evidence for the period,

presently classified as local and ̀ intrusive', will be reviewed in Chapter 4 and 5.

The term ̀ intrusive' presently acts as a form of default term for `foreign' material,

namely Roman, British, Romano-British, Gallo-Roman, Germanic or Saxon imports and

all of this material will be discussed in Chapter 5. The term 'intrusive' has also been

accepted as a form of classification for burials, where migrant population groups or foreigners are suspected during this period. Although the term has been normalised throughout the literature, it is laden with pre-assumptions about the impetus for

consumption of such material, and the origin and identities of both individuals and their

communities in both life and death. Although the period under discussion runs concurrently with the Roman occupation of most of North-West Europe, much of the `Roman' material from Ireland was dismissed as irrelevant to dating sequences at sites

excavated through the 1940s and 1950s (discussed in detail in Chapter 2 and 5). Recent

work by scholars in Ireland, however, suggests that the evidence needs to be fully

reconsidered (Daffy 2002; Newman 1998 & 2005; Raftery 1994; O'Brien 1990,2003,

2006 in press, forthcoming; Kelly 2002; Kelly A. 2007, in press). As the present study

progressed, it became apparent that it was not just the relationship between Ireland and

provincial Roman Europe which was poorly explained, but little had been published in

9

recent years that offered a contemporary assessment of social interaction in the later

period during and after the political demise of the Western Empire. At sites in Ireland,

such as Garranes in County Cork, there appears to have been a flourishing intercourse

through direct engagement with trade from North Africa and the Eastern Empire.

Therefore, the role of imported ceramics (ARSW, PRSW, DSPA, LRAI, LRA2, and E

Ware: discussed in detail in Chapter 5) the classification of which presently falls within

the Early Historic period in Ireland and Late Antiquity (formerly post-Roman) outside Ireland, will be reassessed and a distributional analysis offered of the material circulating in the 5th and 6t` centuries. This will be updated in line with the most recent finds that are

as yet unpublished (Campbell 2007; Doyle 2009 pers. com; A. Kelly 2007,2009

pers. com, in press). It is timely that the relevance of the term `intrusive' is challenged,

given that it continues to be used to classify and typify all that is `Irish' from that which

classified as foreign despite the fact that imports appear to be more common than that

which is classified as of `local' manufacture during the early centuries AD. Material that

has been classified as Roman, Romano-British or British, will be reassessed as a

legitimate indicator of changes in social practices, as read through the materiality of the

period.

In line with questions raised earlier about the applicability and relevance of the continued

use of the term ̀ intrusive' in relation to artefacts, the relevance of its continued use in

relation to transitions in mortuary rites and individual burials will be considered in detail

in Chapter 6. Advances in scientific methods in archaeology presented an opportunity to

offer a new avenue of research into the Irish Iron Age and to test the assumptions that have been made about the origin and ethnicity of those referred to as ̀ intrusive' within the literature (Raftery 1981,1989,1994 & 2005; Newman 2005; O'Brien 1990,1999a,

2006; Warner 1995,1996a & 1996b). Once the existing burial classification has been

discussed in detail, a history of the five sites chosen for scientific analysis through

strontium and oxygen isotope analysis will be presented. The results of the strontium and oxygen isotope analysis, completed as the primary scientific objective of this research will be set out in Chapter 7 along with the presentation of new 14C dates undertaken as part of this thesis. Variations in isotopes in human skeletal material leave distinctive

10

markers, recoverable through a combination of highly specialised laboratory based

techniques, which can be used to interpret likely population mobility and settlement. In

recent years, scientific investigation into skeletal material using strontium and oxygen

isotope analysis has allowed us to map the likely origination and early childhood

residence of past populations (eg. Dupral & Swhwarcz: 2001: 1199 and Schweissing &

Grupe 2003: 1373). Unlike trace element analysis, the isotopic level is not subject to

fractionation through either biological or geological processes and is an excellent

indicator or blueprint of the environment in which an individual was born, lived and died

(Budd et al 2000: 70). It has been demonstrated that the most highly mineralised tissues,

in this case tooth enamel, are most resistant to decay and contamination from diagenesis

(Budd et al 2003: 70).

All of the sites chosen for scientific analysis are within the environs of the plains of County Meath, within the area with the highest concentration of Roman material in

Ireland. Samples were taken from burials at: Knowth (Eogan 1974; 0' Brien 1990;

Raftery 1994), Tara, Räith na Senad (O'Brien 1990; Raftery 1994; Newman 2005),

Betaghstown (O'Brien 1990; Raftery 1994; Kelly unpublished notes NMI), Carbury Hill

Site B (Wilmot 1938) and Rossnaree (unpublished excavation). A history of the

mythology, excavations and archaeological assessments will be presented along with a discussion of existing interpretation offered for them.

The question of migrant populations in Ireland is not a new one, but the somewhat

circular arguments of recent years, such as that from McGarry (2005), which stresses the

problematic nature of the lack of artefactual evidence within burials of this period, ignores the fact that a change in burial custom is, of itself, significant. Cultural change, as

evidenced through artefact or burial remains, does not have to be accounted for,

necessarily, through a form of community replacement. Changes in burial practices may

relate to a shift in political or religious ideology that does not require a parallel change in

the existing cultural milieu, one clear example being that of the later introduction of Christianity (Liversage 1983). That said, where non-local material is evidenced at a site

and there is no archaeological evidence of a pre-existing or contemporary local

11

population, and there is evidence of social change at a regional level, the possibility of a

migrant population group had to be considered more fully. The sites that were chosen for

isotope analysis demonstrate a temporal and contextual transition that appears to be

regionalised, and they are distinguished by the appearance of non-local material at, or

close to, the burial sites. Where possible, samples were collected that included the

(assumed) local burial rite of crouched or flexed inhumation, and the (assumed) later rite

of protected extended inhumation in cists or stone lined graves. Several of the crouched

and flexed inhumations, sampled in order to test for local population groups, do not have

an origin in Ireland, and are likely to have originated in either north-eastern Britain or

south-eastern Scotland. One individual in particular from Betaghstown has an origin in

southern Portugal/Spain or on the North African coast, offering unique evidence of Ireland's engagement with traders in the Late Roman economy. This is an important

finding given the increasing frequency with which imported ceramics are now being

recognised at sites under excavation in Ireland. Interestingly some of the burials that have

been classified as ̀ intrusive' appear to be a mix of local people, and for some we can now

offer an origin in other parts of Ireland. Several important findings will be outlined in

relation to burial morphology alongside a discussion of why the use of relative dating

based on burial types in this period can no longer be used.

A comprehensive re-interpretation of the transcultural nature of the materiality of the

Later Irish Iron Age will be presented in the concluding chapter, rehabilitating the

evidence of non-local material as meaningfully constituted within existing mechanisms of

social reproduction within an Irish context. A new interpretation of the period will be

offered and all of the material contextualised in relation to its probable users. The

transparency of the new classification series allows us to consider evidence previously

treated as irrelevant or outside consideration of social developments within the Later Iron

Age, such as the emergence of ring-forts as homesteads, whilst making explicit the basis

on which the argument is formulated. The inclusion of later material, which up to now has been classified in the Early Historic period, allows for a graduated transition out of

the Iron Age period and into the Early Historic period. The distribution of this later

material correlates with areas of culture contact from centuries before. No cessation is

12

demonstrated archaeologically; rather all of the evidence suggests continuity throughout

the entire period and in some cases development of new nodal entry points, around which later imported material clusters.

The social, cultural and economic significance of archaeologically attested change within

the period will be thoroughly outlined. The social relations involved in improved

technological practices, implied by the construction of different architectural forms, and

the environmental evidence contained within the large multiple ditched and banked

enclosures of this period, does not require us to fix `relational identities' with those

elsewhere (Conlin Casella & Fowler 2004: 8), but neither should we ignore the fact that

these were being developed and reconstructed within the I' century AD, at a time of the

Roman military push forward into Scotland under the command of Agricola. Using the

changes in the landscape, monumentality, burials and newly discovered sites, alongside

the patterning in the distribution of local and non-local artefacts, both the materiality and

the questions relating the identity of these communities will be explored in detail.

In returning to questions raised in Chapter 3, about the extent to which origin needs to

differentiated much more explicitly from the concept of identity, the results of the

strontium and oxygen isotope analysis further the argument that this is a period of immense change in Ireland. People who did not originate in Ireland were being buried at

sites in County Meath, and some, at least, are indistinguishable in death from the local

people buried alongside them. This suggests a complete reconsideration is needed of the

tendency to characterise both people and artefacts as `foreign' within the archaeology of

the ancient past. In this case this `foreign-ness' may simply reflect a modem

predisposition towards characterising all that is seen as ̀ Irish' or `Celtic' as distinct from

that which is not, when in reality these people were part of the communities in which

they lived and died. Then, as now, trade, migration, politics and religion underpinned the

relationship between communities on both sides of the Irish Sea. The rise of kingship and

the parallel development of centres of Christian learning led to a new cultural hegemony

that can be evidenced from the 7s` century AD. The people of Ireland were, in effect, becoming ̀ Irish', in the sense that they are characterised by their portrayals within the

13

later manuscripts. The impetus for major economic, social, political and religious development in Ireland did not happen dramatically in the 6th and 7th century AD; rather its origin lies within the `missing' later Iron Age and part of that development needs to be

accepted as a response to direct contact with the wider Roman world.

14

Chapter 2 Literary Evidence and the History of Scholarship

2.1 Irish manuscript evidence

Ireland has an exceptional body of historical manuscript texts, parts of which are believed to date to as early as the 7t' century AD (Charles-Edwards 2000; McCarthy

2008). Now dispersed across various collections in Ireland, Britain and the Continent,

they were once described as a `window on the Iron Age' and, given the rich detail

recorded within them, they have been used widely in an effort to understand social

construction in the `Celtic' world (Jackson 1964). Scholars in the fields of Celtic

Studies, linguistics, comparative phonology and philology have provided the evidence for the dating of the manuscripts, by separating the texts into those

which are regal or genealogical lists, and those that are linked into

wider world chronicles and can be

dated by reference to other

sources (McCarthy 2008, see Fig.

2.1). By mapping the

development and alterations to

written Gaelic in these

manuscripts, key stages in the

development of spoken Gaelic

have been dated. Existing

conventions dictate that written

and spoken Gaelic are referred to

in linguistic studies as Irish, and

the different forms are classified

as Primitive, Archaic, Early Old,

Old, Middle and Modern Irish

(McManus 1991: 82). Each step

it

r cß+++ä 'ý`ý' CCW

All

ýii" nasxýjocýry ISM

ws

CT Tr

sa: a ýMsw

Fig. 2.1 Extract from 12th century Irish manuscript showing translation of Plato's Timaeus, with Latin annotations (after O'Cröinfn 2005)

change is indicated by the structural changes that affected the vowels and the

consonants, the truncation of word endings and the increasing use of syncopated or

shortened forms (McManus 1991: 83). Although the main body of manuscript work

1S

dates from the 11"' century or later, parts are written in Early Old Irish (c. 7'h century

AD), and it is within these redactions that older passages appear to have been

transcribed in even earlier Primitive and Archaic forms of Irish. Some texts, such as

the lengthy Crith Gablach (tr. Law tracts), also offer an insight into the social and

political organisation of Ireland at the time at which they were written, including

detailed information on landholdings, territorial markers, and individuals' civil rights

(Charles-Edwards 2000; Ö'Cröinin 2005).

A further group of manuscripts are known as the Mythological Cycles, possibly the

most famous of which are those contained within the Ulster Cycle such as the Täin Bö

Cüalinge (tr. The Cattle raid of Cooley). This latter group of manuscripts offers a

narrative history of ancient Ireland, often in prose and verses, and texts such as the

Dinnseanchas, (tr. The Lore of Places), and are thought to retain the mnemonics of

social memory in pre-literate Ireland (Charles-Edwards 2000; 6'Cr6inin 2005). It is

this group of manuscripts that are widely used within archaeological interpretation to

offer details of how people lived in pre-Christian Ireland. All of these manuscripts

offer an origin myth of the `Irish' people, but they are rich in detail of the genealogy

of kings and their royal kin, of warrior-heroes and banquets, of feasting and rituals,

and importantly they record `royal' and sacred sites across Ireland.

The origin of the `Celts' in Ireland is enshrined in a manuscript called Labhair

Gabadla (The Book of The Taking (of Ireland)). Dating to around the 13`h century

AD, it details the many putative migrations of people into Ireland (O'Rahilly 1946).

Amongst the obvious mythological peoples, such as the Tuatha de Danann (people of

the goddess Dana) are references to tribal groups that may offer links with

communities in continental Europe. Amongst these are the Fir Bolg (possibly the

Belgae) and the Gallon and Goideal (from which we take the terms Gael and

Goidelic). These may, however, simply reflect a construction of Irish history by

learned men within the ecclesiastical centres of the 8"' century AD, which were

created in an effort to weave together the different political and geographical

traditions current at the time of their recording, in what was essentially a work of

propaganda (6'Cr6inin 2005: 182). The success of the narrative within these

documents is that it created a sense of a unified history, which is even today reflected

back into prehistory and used widely in archaeological interpretation. Although we

may be cautious about literal translation in relation to facts and history, manuscripts

such as these form part of the wider materiality of the period, and cannot be dismissed

as irrelevant to our considerations.

Before the 2d century AD, according to manuscript evidence, Ireland was divided

into five geographical areas that broadly correlate to the modem provinces today. The

origin of the provinces (Ulster in the North, Leinster in the east and south-east,

Munster in the south and south-west and Connacht in the west and north-west) is

recalled today in the Gaelic word for province, c61c6ad, literally from cüige, the

number five or fifth. The fifth province was the Kingdom of Midhe, meaning `middle', created according to historic tradition, in the first centuries AD by the

confiscation of territory from each of the other four provinces, and credited to a

returning emigre known from the Annals as Tiiathal Techtmar

(O'Rahilly 1946; Warner 1995).

The province of Midhe no longer exists but an etymological reference is remembered

today in the county name, Meath, and it is an area that has one of the highest

concentrations of imported material in the Later Iron Age. Each cöicead is believed to

have had a central tribal capital, and our information about their demise comes, in

part, from a 9th century AD poem by a monk called Aengus the Cele De (ie. servant of

God). This poem is a triumphal record of the replacement of pagan worship with

Christianity at each of the tribal capitals (Brown 1927: 27; Raftery 1994: 64). It

describes the fall of each pagan centre in turn: Emain Macha (present day Navan

fort); Crüaichäin (Rathcroghan) in Connacht; Dun Äilinne (Knockaulin Hillfort) in

Leinster; and Teamhair Brega (Tara). Absent from the list is a capital of Munster,

which is not offered an origin in pre-history. Our knowledge of a political capital for

Munster is associated instead with the emergence of a powerful dynasty sometime in

the 4m century AD, the Eöganacht, and whose history is recorded in a text known as

The Story of the Finding of Cashel (Stevenson 1987: 145; Charles-Edwards 2000).

What we know of the tribal groups, their sub-groups and kin groups by the 8t' century AD is that they were highly complex, and this complexity reflects the shifting power, territories and claims of kingship in Ireland. Although most of the annalistic material

relates to those in power in the north and eastern seaboard of Ireland, it is from the

17

west of the River Shannon that a powerful tribal group known as the Connachta

emerges in the shadowy prehistory of the Iron Age (Charles-Edwards 2000). These

individuals are named for the first time within the Annals as battling for power against

the Ulaid in northern Ireland within the I" century AD. The Annals of Ulster records

the settlements of a people known as the cruithne, a Gaelic term which seems to have

been applied to both a tribal grouping or derbfine within Ireland, and also as a

descriptive term for the indigenous Picts of Scotland. The material culture of the Iron

Age in northern counties of Ireland appears to support a greater cultural affinity with

Scotland than southern Ireland (Warner 1982; Harding 2002,2005). At sites such as

Altagore in County Antrim, and the Giant's Sconce in County Derry, the morphology

much more closely resembles the style of contemporary `dün' or fort sites in

Scotland, rather than enclosed settlement sites in the south (Warner 1982: 179).

Similarly the artefact assemblage, consisting of iron horse trappings, and a diagnostic

type dress fastener, known as a hand-pin, links the material culture across the short

sea crossing to the Scottish coast: the name Argyll is translated as Oirthear Gaedhael,

literally East Ireland (Ö hOgain 2002: 218). It was from Ulster that a powerful tribal

grouping known as the Ulaid are said to have been followed in the historical record

by the emergence of a dynastic family in the later 3'h or 4th century AD, known as the

UI Neil. The most famous son of the UI Neil, was Niall Naoighiallach (Niall of the

Nine Hostages), later High King of Ireland, a warrior-king who is thought to be the

first historical king recorded within the Annals (O'Rahilly 1946). The UI Nill,

according to the Annals, moved south to the plains of Leinster, fought and conquered

Tara and usurped the throne from the Laigini in AD 397 (Line M397.1 Annals of the

Four Masters: CELT). From Tara, Nfall is said to have orchestrated the Irish raids on

Roman Britain and a postulated link has been made between this activity and the

kidnap and enslavement of St Patrick and his two sisters (O'Rahilly 1946). The

Laigini of Leinster, have their own genealogical lists, within which they appear to

have regarded themselves as foreigners, possibly of British origin (6 hUiginn 1990)

Similarly the origins of the immensely powerful Eöganacht are also vague (O'Rahilly

1946; Charles-Edwards 2000). What we do know is that a reorganisation of power

took place some time within the 6'h century AD, with the emergence of a new kingdom of Brega, in modem day County Meath, which encompassed within its

territory some of the most famous `royal' centres, such as Lagore Crannog and Knowth (Charles-Edwards 2000; Bhreathnach 2005).

18

The dating of the emergence of literacy in

Ireland is usually based on the dates of the

manuscripts themselves, as these were being

transcribed at ecclesiastical houses in Ireland

from the 7th century AD (Swift 1997;

Stevenson 1989). The earliest evidence for the

two oldest forms of Irish come from the

translations of the ogham inscribed stones (McManus 1991; Swift 1997,2008). Details

on the origins of the ogham script are recorded in the 12th century `Book of Ballymote' (see

Fig. 2.2. ), within which is a manuscript called

Auraicept na nEcas, translated as, the

Scholar's (or Poet's) Primer (McManus 1991;

147). The ogham script itself is made up of a

series of 20 alphabet characters,

grouped in four groups of five, in Fig. 2.2. Extract from the Book of Ballymote,

Irish known as the In beithe-luis- showing detailed explanation of the ogham script (image courtesty of www. isos. dias. ie

nin', which translates as the `birch- consulted 21/2/2009)

rowan-ash' (McManus 1991;

Newman 2007: 361). Ogham was carved in linear notches across the natural vertical

edge, or arris, of a stone pillar and if a natural edge was not available, a vertical line

was drawn on the face of stone across which the ogham characters were inscribed.

Reference to the types of trees relates to how each tree was perceived in importance in

ancient times, and the properties contained within each was associated with certain

positive human attributes and vice versa. It also relates to how the ogham script is

read, described in the Auraicept as exactly how one would climb a tree, right foot first

from the bottom to the top (McManus 1991; Newman 2007).

In the most comprehensive study of both the orthographic and manuscript evidence,

and likely dating of ogham, it has been argued that its creators were `undoubtedly

familiar with Latin', but they may also have been familiar with Greek (McManus

1991: 149). Furthermore, there is a suggestion in the Auraicept that ogham was seen

19

as a ̀ genesis' of the older language base, and that by the 60' century AD it was no longer in use (McManus 1991: 149). Scholars disagree over the relative dating of the

script, with some arguing that based on the obvious familiarity with Roman literacy,

on linguistic grounds alone they should be dated to around the 2'd century AD

(Harvey 1987,1990). There is a long history of scholarship into this unusual Irish

script, and attempts to offer a relative chronological dating system for it has been

described as one which is `built on sand' (Swift 1997: 55). Most scholars now agree that ogham was devised to record spoken Irish and that it is based directly on both

written and spoken Latin and that the grammatical conventions used in Classical Latin

(McManus 1991; Swift 1997; Harvey 1987). There is one very important issue raised by this last point: the inscriptions reflect Latin orthographic convention and if those

who commissioned the inscription from a lapidary were able to write Irish to reflect the correct Latin, then it follows that they must have been able to pronounce Latin as if it were Irish (Stevenson 1987: 144). A parallel has been drawn between the

emergence of ogham inscriptions in Ireland and the Gaulish vernacular inscriptions

from around the P century AD and it has been argued and that both Latin and literacy must have been ̀naturalised' in Ireland within the Roman, rather than the

post-Roman, period.

Ogham was not just in used in Ireland, however, and within parts of Wales and Cornwall the inscriptions record Irish names in both ogham and Latin (McManus

1991: Charles Edwards 2001: 163). The Silchester ogham found within a well at Insula IX has now been reliably dated to the end of the 4th century AD (Fulford et al 2000). This little ogham is inscribed on a greenstone baluster that once formed part of

a townhouse in late Roman Silchester and it is remarkable because of where it was found and that it suggests a community who were completely familiar with this Irish

medium, in the southeast of late Roman Britain. The inscription itself is also important, as it has been translated as recording a claim of ownership to either the

land or property in Silchester (Fulford et al 2000). In the past most ogham inscriptions

have been interpreted as memorial dedications, yet more recent arguments have

suggested that the oghams had multiple functions and that some, such as those found

around the environs of Tara, formed part of a complex system of territorial land-

markers, set up by those in power in the 4th century AD (Newman 2005). The

distribution of the monumental ogham stones is noteworthy, as the majority of the

20

inscriptions are in the southern counties, mostly in Cork and Kerry, in the same area

which was home to the E6ganacht tribal kingdom who will be discussed in detail later

(Stevenson 1987: 145; McManus 1991: 46). The distribution of ogham inscriptions in

the rest of Ireland appears to correspond with areas that also have a concentration of

Roman material and I will return to this point in Chapter 5 once all of the material has

been reclassified.

The linguistic and philological work of these scholars has added much detail about the

variations in spoken Irish before the widespread borrowings into Irish of ecclesiastical Latin terms (McManus 1983: 32). The insistence on imposing a chronology based on

the dating of the Christian centres of learning and literacy has tended to obscure a

much earlier origin for many secular words borrowed into Irish direct from Classical

Latin (McManus 1983: 32). These words have no obvious liturgical or ecclesiastical

significance, and as they they appear to have been based on how Latin was spoken,

within an Irish context (McManus 1983: 29). Amongst these are many words of a

military nature such as: arm > arms, legion > legion, mil > mile, trebun > tribune,

Gall > Gaul, bärc > ship, Mercüir > Mercury, Saturn > Saturn (1983: 43). Alongside

these, according to McManus, are very early borrowings of terms such as: sesra >

sextarius (a measure), muide > modius (vessel for liquids), esarn > exhibernum (year

old wine), sithal > situla (vessel for drawing water), ingor > ancora (anchor), crioch

> crocus (saffron dye), monad > moneta (money), and dIrna > denarius (weight/coin

denomination) (1983: 43). To these can be added argat > argent (silver), corcora >

purpura (purple dye) and scibar > piper (pepper). In seeking an origin for literacy in

Ireland within the conversion period of the later 6m and 7t' century AD, scholars

appear to have missed the significance of the evidence of writing implements at sites

with early occupation sequences, such as Garryduff, Garranes, Lagore and Knowth. It

is entirely possible that some communities in Ireland were literate long before the

emergence of the monastic centres and that much of the development of the ogham

and early borrowings of Latin terms were as a consequence of their engagement with

the Roman world.

Although we must be sceptical about the historical nature of the individual families

named at this time, by the time that the manuscripts were committed to vellum some 150 tuatha or kingdoms had been established (Newman 2005). Within an immensely

21

complex hierarchical social and political structure, those in charge exercised a form of

political and social cohesion not seen before amongst the disparate tribal groupings (Warner 1995; Newman 1998; O'Rahilly 1946; Raftery 1994). Part of this shift can be evidenced archaeologically in the emergence of large, enclosed centres of political

control, ie. the historic centres mentioned throughout this work, and from within

which much imported material has been found and used to date the construction to

within the first two centuries AD (Raftery 1994; Warner 1995; Charles-Edwards

2000).

Along with the genealogical corpus, poems and flowery prose, are a series of hagiographical texts that record the lives (beatha) and activities of missionaries, most

of whom were later canonised as saints (Charles-Edwards 2000). Within these we

read the idealised portraits of individuals such as St Patrick, St Brigit and St Columba,

their work to establish Christian churches, and the seemingly rapid conversion of the

Irish to the new religion. These too, however, present a problem as although there are

two works believed to be attributable to St Patrick, there are no dates recorded within

these texts. The highly interpretative accounts of the Christian missions presented as

taking place within the 5t' century AD often rely on a generalised reading from 7th

century AD texts such as that of the hagiographers Tireachäin and Muirchiü on the

life and work of St Patrick, or even later 8'h century AD writings such as the Life of Admonnän and the writings of Gildas and Bede (Charles-Edwards 2005). As with the

Annals, caution is needed when extrapolating underlying social constructs, centuries before these documents were even written.

The historiographical basis of archaeological interpretation and the persistent use of a

classificatory system that continues to rely on old chronologies has come under

academic scrutiny from scholars in Ireland (e. g. Tierney 1998: 193; O'Sullivan 1998:

179; Horning 2006; Newman 2005; 2007). The most thought provoking work has

come from University College Cork, and the National University of Ireland in

Galway, where scholars such as O'Connor, Tierney and O'Sullivan have questioned the resistance to the adoption of theoretical approaches in archaeology in Ireland

(O'Connor 1998; O'Sullivan 1998; Tierney 1998; Newman 2005 & 2008). Their

work, although insightful, and very much in keeping with the critical dialogue taking

place in Iron Age and Roman studies in Britain, has not been widely accepted.

22

Scholars working in disciplines other than archaeology appear to have taken the lead

in offering a new critical discourse in Ireland, and they have questioned the creation

of an isolationist position in Irish pre-history and its impact on Ireland's national

psyche (McKee 2000; Garvin 2004). The continued use of the manuscript sources as a

reference tool for the dating and classification of archaeological sites in southern Ireland has similarly been challenged (McKone 1990; Hutchinson 1987,1999 &

2001; Curtin 1996; Howe 2000). Arguing against the relevance of texts that are known to post-date actual events in prehistory, McKone has drawn attention to the fact that the manuscripts are rife with biblical representations, and are concerned

primarily with the oppositional contexts of pagan ways versus Christian morals. As

such, he has argued they have to been viewed critically, as representing both the

audience for which they were written, and the environment in which they were (re)created, within the precincts of ecclesiastical centres (McKone 1990: 8).

2.2 The Classical sources

4 si quidem Hibernia medio inter Brittaniam atque Hispanium sita et Gallico quoque marl opportune valentissimam imperil partem magnis in vicem usibus miscuerit. '

`For Hibernia, lying between Britain and Hispania and placed strategically in the Gallic Sea, would unite the most robust parts of the empire to the great advantage of

both. '

Tacitus, The Agricola, c. 25

McKone is right to suggest that we should be cautious of the way in which the Annals

and Cycles are used within archaeological, and historical, interpretation. We should be similarly cautious, however, when using the classical histories. An uncritical reliance on references to Ireland in the classical accounts could lead to a distorted

normative account of the period, filtered through Roman ideology and perpetuated through an essentially British Victorian imperialism that saw all that was not Roman

as ̀ uncivilized' (Hingley 1994; 1997). This is especially important when we consider how problematic the use of the annalistic and mythological material has been within archaeological interpretation for sites in the Later Irish Iron Age. Although we think

23

of Roman historical accounts as an unbroken linear record, it is clear that there are

significant periods and significant events that are not even recorded for Britain. There

is a complete lack of any literary records for the period AD 84 to 117, an important

period after the Agricolan campaigns took forces into Scotland; and we have, at best,

patchy evidence for the entire P century. With regard to Ireland, as pointed out by

Raftery (1994), Warner (1976; 1995 & 1996), Killeen (1976), Robinson (2000) and Freeman (2001), a short, unofficial or unsuccessful attempt to take Ireland, given that Agricola thought it an easy affair, would not have suited propagandists and

panegyrists concerned with recording Roman military glories for posterity.

Both Ptolemy's map from his Geographia and Tacitus' Agricola are the most detailed, and earliest, classical accounts of Ireland and they are, unsurprisingly, the

most frequently cited in any preface to discussion of Roman material in Ireland

(Warner 1995 & 1996; Raftery B 1994,1995,2005; Di Martino 2003). The quotation

above is part of one of the most widely disputed references to Ireland within the

classical accounts, as it offers a record of Agricola's musings towards an invasion of Ireland. The statement places Ireland within its ancient geographical context: Tacitus

thought it was positioned between Britain and Spain and within the Gallic sea. Such a

position offered the possibility of a strategic move that would serve to unite the islands under the Roman standards (Agricola c. 24; Freeman 2001: 56). References to Ireland within the classical sources were assessed by Killeen (1976) and more

recently by Freeman (2001), who argued that all of the references should be seen through a time-line that reflects the extent of knowledge of the island of Ireland that

was gained once Rome had secured the province of Britannia (Freeman 1995,2001).

In fact it has been argued that Ireland's geographic positioning was so well known to

traders and merchants within the Roman period as to not require any particular

attention and that the notion of Ireland as being a remote and barbarous land was,

actually a product of 19th century thinking that deliberately ignored the many,

unremarkable, references to Ireland in the classical literature (J. Raftery 1980).

It is worth considering the quotation from Tacitus in some detail at this point as it

opens with information on Agricola's fifth season of campaigning in Britain. In a lengthy reference Tacitus discusses an exiled Irish Chieftain, and states that as he was 'expelled, from his home by a rebellion, Agricola.. detained him, nominally as a

24

friend, in the hope of being able to make use of him'. I have often heard Agricola say

that Ireland could be reduced and held by a single legion with a fair sized force of

auxiliaries' (Agricola 24). This particular passage has a long pedigree of creating

animosity between experts as at the turn of the 20th century a heated debate was

played out between Haverfield and Gudeman in the Classical Review (Gudeman and

Haverfield 1900; Killeen 1976: 214; Robinson 2000: 19). Gudeman had suggested

that a military expedition to Ireland took place under Agricola, a position that he

strongly maintained in his later work. Haverfield maintained that such an ̀ epoch

making event' could not have been glossed over in the narrative of Tacitus, and he

followed up his argument with an article published in 1913, in which he described the

Roman evidence from Ireland as ̀ trifling' (Haverfield 1913: 3; Killeen 1976: 214;

Robinson 2000: 19).

Early references amongst the Greek and

Roman writings consist of no more than one

or two lines on Ireland, and they make

reference to some bizarre ethnographic

practices such as endocanibalism; these

musings are anecdotal and have been

described as ̀ fanciful Wandermotiv' based

on earlier Greek stories of `barbarian'

societies and their assumed social practices (Killeen 1976: 212). In Ptolemy's Almagest

Ireland is described as ̀ Little Britain', but in

his later work, the Geographia (from the 2"d

century AD, see Fig. 2.3), Ptolemy records

Mumm F-Iadacam WANommdem In

w. 001. A

" pn IIOýOBOý

0 yr" . ww .... ot

SON ý "ý mum rwm

rýýrwo+

. wr . summ ýý OIMIO. CI

*MM amm

vom* *�um .... «. m @ am .. OU07.. N'

SL. M "... Con»

IY[SMIý IIi01N

IýCM. r

E 1101.0.

for the first time the ancient name by Fig. 2.3 Ptolemy's 2nd century AD map of

which Ireland was known, Iwernia Ireland, reconstructed, from co-ordinates (after (Almagest 2; Geographia 1.6). Spoken Lynn 2003)

Greek in this period did not commonly use a `w' sound, and Freeman has argued that

its use here suggests an accurate attempt to record the native name for the island

(Almagest 2; Freeman 2001: 67). References to Ireland within the Geographia are

much more detailed, and in parts more accurate, than those for north-western Britain

25

and Scotland, and Ptolemy corrects an inaccuracy common in earlier sources by

placing Ireland's geographic position closer to mainland Britain (Geographia 1.6,

Strang 1998; Freeman 2001: 66; Mattingly 2006). The map lists coordinates for

coastal towns and places of safe harbourage and also inland tribes and tribal centres.

A recent cartographic assessement of Ptolemy's map of Ireland concluded that these

co-ordinates, unlike those for Britain but especially Scotland, are within a few degrees

of accuracy (Strang 1998). The source information for the map was probably

compiled from information derived from traders and merchants, and it is clear that at

least some of these had visited frequently, as the information is supplemented with a

note on the length of the day in summer at two of the inland regia or royal centres

(Geographia, 1.7-12, Freeman 1995: 12). Two of the sites listed as regia are probably

political centres, and they appear to correspond geographically with historical sites

known from the Annals: Clogher Hill fort in County Tyrone, and Crüachän

(Rathcroghan) in County Roscommon. A further site, Dunon, is listed as a polis by

Ptolemy, and it corresponds to a site known from the Annals as Dind Rig in County

Kildare; this may be Rathgall Hillfort (Raftery 1970: 211). What is notable is that

none of the sites listed are near to the River Boyne, which is given accurately as the

Buwindafluvia, so they cannot be interpreted as relating to the Hill of Tara. Some of

the monuments still visible today on the Hill of Tara would have been extant at the

time the map co-ordinates were prepared, certainly Duma na nGiall (the Mound of

the Hostages) as it dates to the Neolithic (Newman 1997; O'Sullivan 2005). This lack

of recording of Tara offers an interesting possibility that, at the time the information

was compiled, the site had not yet taken on its pre-eminence as a political centre.

We know that by AD 83 Agricola was pushing forward by sea and land into Scotland,

and there remain questions over the details of this part of his fifth season of

campaigning in Britain. The questions centre around Tacitus' accounts which records `hitherto unknown' tribes as being conquered by Agricola; the academic consensus

appears to be that these were Scottish tribes although there remains the possibility that

these were Irish tribes (Agricola 24; Reed 1971: 145; Kamm 2004; Raftery 1994: 205). It is probable that Tacitus also had access to detailed information about Ireland,

as he likened the climate, weather and people to those in Britain and he noted that Ireland's `approaches and harbours have now become better known from merchants who trade there' (Agricola 24). There are other references to Ireland detailed by

26

Freeman in 1995 and 2001. Pliny the Elder (c. AD 77) recorded that `Hibernia same

width as Britain but less in length; shortest crossing fr om Silures tribe in Britain;

islands between Hibernia and Britain' (Natural History 4,102-3) and a line from

Juvenal which states ̀ Roman arms carried beyond the shores of luverna and Orcadia' (Satire 2,159-16), is dated to around AD 116 and therefore is broadly

contemporary (Freeman 1995: 11; Warner 1976: 281).

The later sources are more varied and although Ireland is not always explicitly described, various references to the Scotti, who first appeared from about AD 267,

have been interpreted as relating to Irish raiders (Rance 2001). The writer Ammianus

Marcellinus, whose History was written towards the end of the 0 century AD,

recorded for posterity the active conscription of a `British' tribe, the Attecotti, into the

Roman auxiliary forces at some undetermined date after the `Barbarian Conspiracy'

of AD 367 (History 27.8.5, Rance 2001; Freeman 2001; Shotter 2004). Scholars have

debated the origin of the Attecotti, who although described as engaging in attacks on

the coast of Britain, along with the Picts and the Scotti, are later recorded in the

Notitia Dignatatum as auxiliaries serving on active duty in Italy, Gaul and Illyricum

(Rance 2001). In a convincing hypothesis based on the annalistic corpus of both

Ireland and Wales, and the Roman historical records, Rance identified the Attecotti as

a Latinized version of `äitheachtüatha', an Irish term used to describe a landless,

migratory or subjugated tribal grouping (O'Rahilly 1946: 162, Byrne 1973: 45; Rance

2001: 249; Freeman 2001). The work was based on the evidence from an Irish

manuscript called Tairired na nDesse, (Wanderings or Expulsion of the Deisi), which

records the migration to Wales of kin groupings known as the Ui LIathdin and the Deisi (Charles Edwards 2000; Rance 2001). A reference in the Annals appears to date

the expulsion of these groups to the mid 2nd century, although a mid 4' century AD

date for settlement in Wales can be supported by both the archaeological and literary

evidence (Rance 2001: 258). In a break with more traditional interpretations, it has

been argued that it was possible that some Irish mercenaries were participating in the late Roman military campaigns, and that some settlement had been encouraged with a

grant of federate status in late Roman Britain (Rance 2001: 252; Somerset Fry et al 2002). This hypothesis may be further supported by some of the later Roman material from Ireland, considered in detail in Chapter 5. One further curious reference is in the

writings of the biographer of the Emperor Tacitus in the 4th century AD. In the

27

Historia Augusta it is recorded that the portents had indicated that a proconsul would be sent to the island of Iuvernia (Tacitus 15.2; Freeman 2001: 97). Although we may be sceptical about the accuracy of that which was predicted by divine communicants

what is significant about this entry is that Ireland would even warrant a mention, let

alone be thought worthy to be receive an individual of such seniority and distinction.

Even if the reference was meant merely as a literary topos to demonstrate the extent

of Roman administration and power, it suggests a certain familiarity with the island of Ireland by those in Rome by the 4d' century AD.

2.3 The study of the Roman material in Ireland

At around 200 finds, the entire evidence for Roman material in Ireland is less than the

assemblage from a single Romano-British site and an up-to-date catalogue is attached

as Appendix B. There are distributional patterns in both the types of material being

imported, such as a regional preference for ceramics and smaller pieces of jewellery

and these will be discussed in detail in the following chapters. It is important to note here that the amount of material is less important than the fact that at some of the sites little, or nothing else, was found within the same contexts suggesting that only imports were being used during that phase of occupation.

Since Haverfield's work in 1913, which dismissed as ̀ trifling' the evidence for

Roman influence in Ireland (Haverfield 1913: 3), there have been four attempts to

review and re-classify Roman material found in the Irish archaeological record. Ö'Riordäin reviewed the evidence in 1947, following his excavations at Rdith na Senad and Räith na Rig at Tara in County Meath, from which an array of Roman

material was uncovered. Ö'Riordäin classified not just material of certain Roman

origin, but also material that he believed was of local origin from around Ireland,

inspired by classical styles. He interpreted this local material as evidence of a

continued insular tradition derived from Roman forms, which were subsequently

adopted and adapted into a later Irish style (O'Riordäin 1947: 47). Notably, Ö'Riordain classified pins and brooches, both of which are common on sites across Ireland and Britain, as being of Roman design, and he argued that the penannular brooch, in particular, was derived form the Romano-British type common in the 2d

century AD in northern Britain (6'R1ordAin 1947: 48).

28

O'Riordäin's work was reassessed in 1973 by Bateson who produced a catalogue of finds, separated into coin fords and non-coin finds, and then classified all of these into

three categories: accepted, questionable or rejected. Only those coins, or artefacts, for

which a context could be offered, through `reliable reports of their finding' were

classified as genuine imports into Ireland within the 1st to the 5t' centuries AD

(Bateson 1973: 24). Bateson argued that although most of those objects classified

under the questionable category were certainly of Roman origin, since inconsistencies

remained in their publication or doubts remained over the circumstance of their finding, they could not be included as genuine contemporary imports. Many of the

stray coin finds and fibulae, that O'Riordäin thought were important, were classified

as either `questionable', or `rejected' by Bateson. While questioning the authenticity

of objects is perfectly reasonable, Bateson was inconsistent in his approach, as he

accepted some material that had neither a context, nor a provenance. The Stonyford

burial and its associated material were accepted as genuine finds, as were all of the

gold coins, but notably not those of lower denominations (Bateson 1973). Bateson

did note, however, that it was not possible for him, at that time, to review all of the

material from older excavations but suggested that some of the material from sites

such as Lagore crannog in County Meath, may have been genuine early Roman

imports.

Bateson's work prompted a Colloquium on `Hibemo-Roman Relations and Material

Remains' in 1974, which brought together interested scholars, and the subsequent

papers were published together in 1976. As several of these papers offered a different

perspective on the material evidence, Bateson updated his earlier work with a further

contribution in which several new finds were added to the catalogue. Within this later

paper, he argued that although questions remained about some of the material,

archaeologists needed to keep an open mind when interpreting new finds in Ireland,

as most were likely to have been imported during the Roman period. This important

qualification formed the basis of a more nuanced consideration by Bateson, one that

appears to have been based on his disagreement with the consensus view of

contributors to the colloquium that favoured a later importation date for all of the

Roman ceramics. To counter this, Bateson argued that although it could not be proven that Roman ceramics found during excavation were from a `Roman' level, neither

UNIV RSS TY 29 OF BRISTOL

LIBRARY I

could it be demonstrated that any had been found in a later ̀ proven dark-age level'

(Bateson 1976: 179).

Warner, later Keeper of Antiquities at the Ulster Museum, contributed to the

colloquium and proposed a new classification based on Bateson's work from 1973.

Warner argued that objects that clearly originated in the Roman world but did not have a secure find context had to be separated into an early (1St century BC to 2nd

century AD) and a later group (3`d to 5th century AD) (Warner 1976: 270). These, he

argued, needed to be separated completely from material for which a secure context

was known (1976: 270). A further category of material, he argued, was needed for

objects where there was evidence of `fusion' with the earlier Roman material, but

which were often described as ̀ early Christian' (1976: 271). Warner's classification is

important in relation to the chronology of sites and how they are dated to the later

Early Historic period, a point to which I will return later in this discussion.

In 1994, Raftery reviewed all of the Roman material once more, and concluded that

during the 18t century AD, a `small-scale landing of Roman legions may well have

taken place' (1994: 219). Raftery considered the lack of historical Roman records for

the period after Agricola's campaigns, and concluded that the possibility of military

intervention in Ireland could not be dismissed (1994: 219). The argument he put forward was centred on material found at sites of historical importance, which

included the Hill of Tara, the passage tombs of Newgrange and Knowth, Knockaulinn

hill-fort (Dün Äilinne, the capital of the Laigini tribe of Leinster), Clogher hill-fort

(the tribal centre of the Airgialla), and his own excavations at the hill-forts of

Freestone Hill (1969) and Rathgall (1971,2006), the fords from which will be

discussed in detail later. At all of these sites, both the material and the contexts were

dated using the diagnostic material found, such as the Saurian ware from Tara or the

3'd century strap tag from Rathgall (1971,2006). The imported material, according to

Raftery, was much more closely datable than any contemporary local material, as the

typologies and chronology for these were extensively published and widely accepted.

His approach was such that where an archaeological association between diagnostic

artefacts and a context at these sites could be demonstrated, then the two should be

accepted as contemporary.

30

The terms `residual' and `reliquary' have been used to dismiss much of the evidence

for Roman ceramic material from sites such as Lagore in County Meath, Ballinderry

No. 2 in County Offaly, Island McHugh in County Down, Lough Faughan in County

Down and Cloonfinlough in County Roscommon (Bateson 1973; Cotter 1999;

O'Sullivan 2002; ). The term ̀ reliquary' was used to suggest that pieces of Saurian

ware at these sites were the result of missionaries collecting `Holy Roman' soil in

Italy, and using it to consecrate ecclesiastical settlements in Ireland (Warner 1976:

272). One key element that links many of the sites where Roman material has been

classified in this way is that they are all crannogs and, significantly, the accepted

chronology of crannogs is from the 6t' century AD (see Crone 1993; Campbell 2001;

O'Sullivan 1998; Edwards 1990,2006; Cotter 1999). This will be discussed in detail

in the following chapter but it is important to note here that none of these sites has any

known ecclesiastical associations that might support such an interpretation.

The classification of material as ̀ residual', or more frequently `long-lived' (Hencken

1942: 57) remains problematic for although the term is used to dismiss its relevance

to dating contexts, it does not adequately explain where it is believed to have come

from in Ireland in the first place. Roman material may have been heirloomed by

people living at these sites, although the types of material found in Ireland do not fit

easily within such an interpretation (Warner 1976; Rai%ry 1994). In England, Roman

metalwork and ceramics are common fords in later Anglo-Saxon contexts and the

material is often repaired or re-worked into another form (White 1988; Neuman de

Vegvar 1997). The various interpretations for the re-use range from the pragmatic

reuse of metalwork, to symbolic or cultic relevance, to indicators of status-by-

association with its original makers, owners or users (White 1988). It is clear,

however, that any, or all, of the factors outlined could have led to its re-use in later

contexts (Neuman de Vegvar 1997). More recently, Eckardt and Williams have

argued that these people may not have known the precise past history of these objects,

and their discovery may have created a special sense of association between objects,

people and places (Eckardt & Williams 2003). This is especially true when we

consider the archaeological evidence of Anglo-Saxon settlements and cemeteries sited

near Roman temple sites, and the very likely possibility that they may have found

discarded material within its confines (Eckardt & Williams 2003: 159).

31

In 1995 Warner proposed that the distribution of provincial Roman material in Ireland

might be best explained by an Irish-led unofficial Roman incursion in the 1st century AD (Warner 1995). He suggested a date of AD 76 based on a record within the

Annals relating to a king called Tüathal Techtmar who was exiled before returning to

Ireland with a military force whose battles are also recorded within the manuscripts. Warner suggested that T4athal might be the same exiled Irish regulus, recorded by

Tacitus as having been befriended by Agricola around the same time (Agricola, 24;

Raftery 1996: 205; Warner 1996: 25). Warner proposed that this was an Irish-led

campaign supported by mercenaries and Roman auxiliary troops whose success

against the aitheachtüatha resulted in T6athal reclaiming his kingship at Tara and

reigning for 30 years (Warner 1995: 25). Supporting his argument with evidence of finds of Roman material from sites associated with Tüathal's restoration, Warner

argued that Drumanagh in north County Dublin might have been the site of a

auxiliary fortress. In January 1996 Warner's article was sensationalised in the Sunday

Times newspaper, in an article entitled `Fort discovery proves Romans invaded

Ireland' (Maas & Byrne 1996). The headline created a lengthy debate within

academic circles but it also sparked an immense public debate on radio and television

programmes in Ireland. The criticisms levelled at Warner and his work ranged

between objective academic debate, led by contributions from Raftery (1996a), to less

scholarly and more personalised attacks on the motivation behind its publication. One

in particular, the editorial in the popular magazine Archaeology Ireland, suggested

that such `unfounded' theories might damage the talks on the Northern Ireland peace

process, which were ongoing at the time (Cooney 1996: 3). Warner responded to the

criticisms in an article in which he described the debate as ̀ becoming less than

objective and less than academic' (Warner 1996).

Warner's work was followed up in 2000 with a contribution from Robinson in which he explored Agricola's rationale behind the construction of the large legionary

fortress of Deva Victrix at Chester. Given that Tactitus recorded Agricola's interest in

full scale invasion of Ireland, Robinson argued that Deva was intended as a new

provincial capital for Britannia et Hibernia, a hypothesis which has received support from Mason in his work on Chester (2001). Robinson explored the original

translations from Tacitus' Latin text, and argued that the use of the Latin verb `exceperat', traditionally translated as ̀ had with him' (an exiled Irish prince), is more

32

properly translated as ̀ had taken out' (of Ireland); the tense, he argues, is pluperfect

and has been consistently mistranslated (Robinson 2000: 27). Such a revised

translation makes more sense if it is posited within a background that includes the

possibility of a brief excursion across the Irish Sea. It is highly likely that a short

expedition to Ireland was undertaken to confirm information already gleaned from

traders and merchants traversing the sea lanes of the Irish Sea. Intelligence gathering

and reconnaissance parties were commonplace before any strategic move into foreign

territory. Caesar tells us in his commentaries how he instructed Volusenus to

reconnoitre the British coastline, and observe weather and sailing conditions in

advance of his invasion of Britain in 55 BC (Bello Gallico, 4.21). An excuse to

reinstate an exiled king, to ensure a client tribute and friendly ally was standard

practice from the time of Caesar with recorded accounts of short excursions to

reinstate named local individuals before handing back authority to them to rule in the

name of Rome. For Britain we have clear evidence from the time of Caesar and his

support of the Commius of the Gallic Atrebates (Caesar, Bello Gallico, 4.21) and the

role that Verica played in legitimising the Claudian invasion of Britain in AD 43

(Creighton 2006). There are also contemporary parallels for such interventionist

policies such as the role of the Roman military in the internal affairs of Cartimandua

of the Brigantes and her estranged partner Venutius in AD 69 (Tacitus, Annals, 12.4;

Robinson 2000: 25). Furthermore, we know from the accounts of Tacitus that

Agricola had the Classis Britannica circumnavigate Britain shortly after he defeated

the warriors of Calgacus at `Mont Graupius' (c. AD 83), and the Roman navy appears

to have taken control of Orkney for at least a brief period (Juvenal, Satire 2,152-

169).

Although the hypothesis of Roman intervention has not been accepted within Ireland,

the site and finds from Drumanagh are widely discussed in work published outside Ireland, in any preface to discussions of Roman influence in Ireland. Warner's

hypothesis has been explored, and in some cases developed, by authors such as Hill

(1995), Driscoll (1998), Rankin (1996), Harding (2004) and Mattingly (2006). All of these authors have argued against the persistent use of simplistic archaeological

models applied to the later Iron Age of Ireland and Britain, and all support the

probability of some engagement within the wider Roman economy. An unsuccessful expedition to Ireland may have taken place during the time of Agricola, but there is as

33

yet no definitive archaeological evidence to support official Roman invasion of Ireland (Cunliffe 2001: 416; Shotter 2004: 31; Rankin 1987: 300; Freeman 2001; Kamm 2004: 75; Mattingly 2006: 439) and a critical assessment of Warner's work will be presented once all of the material evidence has been discussed in Chapter 5. What is significant about this debate was the manner in which an archaeological interpretation that suggested an alternative historical narrative for Ireland quickly became both controversial and politicised.

2.4 The historical narrative

The hostile reaction to Warner's hypothesis underlines the fact that archaeology in

Ireland remains fixed within its culture-historical model, the deconstruction of which is tantamount to heresy (Cooney 1996; Warner 1996a & 1996b; Tierney 1998;

O'Sullivan 1998). The rigid adherence to the accepted narrative of an independent

indigenous `Celtic' past is, in part, due to the need to date sites and material by cross-

referencing it against the proto-history or pseudo-history, enshrined within the

Annals. Although few other than specialists in Celtic Studies are qualified to comment

on the original texts, written as they are in Old Irish, the translations have inspired

generations of archaeologists who have sought out the historically important sites

named within these documents. Even with the benefit of verifiable new data, there

remains much academic scepticism of any new interpretive model, especially one from archaeologists working outside Ireland, which seeks to undermine the existing

culture-historical interpretation (Raftery 1996 & 2006). This position becomes more

understandable, when it is placed within a context of cultural nationalism. The need to find an indigenous past was a fundamental tenet of the burgeoning Republic, a past

which was culturally owned by the Irish people, reinvented in the 19th and 20th

century by the Anglo-Irish literati, and reinvigorated by the appropriation of its

ancient literature, language and artistic forms within the later political separatist

movement. In this way a form of closure was found by which the new nation could rid itself of the legacy of British rule. Throughout most of the archaeological reports

published in the journal series for the Proceedings of the Royal Irish Academy and the Royal Society ofAntiquaries of Ireland, the underlying nationalist historical

perspective is evident even today. The important role of the early Christian church is

stressed as the catalyst for change and development and its role in social development

34

is credited with taking Ireland into the light of history and by inference out of the

darkness of the Iron Age. This emphasis, as applied to both artefacts and sites,

appears difficult to remove despite the fact that it was devised many decades ago. This is to such an extent that material that does not fit into a `La Tene' Iron Age

classification, is grouped together under a later, generic `Early Christian' catch-all,

even when there remain questions about stratigraphic sequences at the sites.

Towards the end of the 190 century, a new nationalist fundamentalism emerged in

Ireland, centred within organisations focused on revitalising Irish games, culture and language, such as the Gaelic Athletic Association (GAA) and the Gaelic League. The

engagement and inclusion of the ordinary, and often materially impoverished, rural Roman Catholic communities gradually led to these organisations becoming more

separatist in outlook (O'Sullivan 1998: 179; Kee 2000: 7; O'Brien 1981: 429;

Hutchinson & Aberbach 1999: 5). Cultural nationalism and the movement for

separation from Britain had, until this time, been the domain of the intelligentsia of

the Anglo-Irish Protestant east coast. Playwrights, poets and aristocracy had been

vocal in their condemnation of the continued British occupation of Ireland, but this

role gradually shifted towards the ideology emerging from an increasingly well-

educated Catholic middle class (Kee 2000: 432; Hutchinson & Aberbach 1999: 515;

O'Sullivan J., 1998: 179; Githens-Mazer 2006: xiv).

Between 1899 and 1902, a secretive society known only as the British-Israel

Association of London, began digging on the Hill of Tara in a search of the legendary

Ark of the Covenant. Little was known about the organisation until recent years but it

is clear that at least one leading figure within the Royal Society of Antiquaries of Ireland was a member of this closed group (Carew 2003: 2). The British-Israelites

saw Tara as a new Jerusalem, the ` spiritual capital of the British Empire' but to

nationalist activists, Tara represented the capital of the `future independent Ireland,

free of the shackles of the British Empire' (Carew 2003: 3). Despite opposition,

permissions were given for this group to dig at Rdith na Senad, the multivallated

ringfort, which is one of the key sites chosen for further analysis in this research. It

was recorded at the time that much valuable material was removed from the site, and it remains unclear what was found, but the Ark was not amongst it (Carew 2003;

Grogan 2009: 11). Leading literary figures such as Yeats, Hyde and Griffiths

35

orchestrated a newspaper campaign to stop the destruction of the site and the drama

rose to a new level when, on visiting in person, they were ordered off the site, at the

point of a shotgun. There was, however, no public outcry over the destruction of Tara

and the push for independence from Britain had not yet gained grass-root support (Carew 2003).

The catalyst for independence was triggered on Easter weekend 1916, when a

relatively small group of pro-nationalist radicals, drawn from a wide range of political

organisations, mobilised to attack strategic buildings in Dublin, and declare an Irish

Republic. Their politicisation and increasingly militant stance was directed at not just

the British government, but at those seen as the perpetrators of colonialist

subjugation, the Anglo-Irish, MPs and ordinary Protestant families. In the most recent

work to address these issues, Githens-Mazer has argued that the protracted series of

executions of those involved, which took place in the wake of the rebellion, `constituted a cultural trigger point' in the history of popular support for radical

nationalism (Githens-Mazer 2006: xiv). Although most people would not have even heard of the individuals who led the rebellion, such as Pearse, Clarke or Connolly, the

circumstances surrounding their execution ensured their martyrdom. The mood of the

people shifted towards sympathy for their cause and along with the symbols and imagery of the heroic past of ruined round towers and Celtic high crosses, their deaths

entered the national psyche, and became synonymous with the struggles against the injustice of continuing British rule (Githen-Mazer 2006: xv). The militant action became the catalyst for change and propelled Ireland onto a separatist trajectory that

would eventually result in the end of British rule in 26 of the 32 counties of Ireland.

This became a reality only after a bitter war of independence, and a bloody Civil War,

which resulted in the formalisation of the Irish Free State government in 1921 (Kee

2000; Townshend 2006)

Former Republican revolutionaries that survived the rebellion, such as Eamonn De Valera who became the first Taoiseach (Prime Minister), and Eoin MacNeill who became first Minister for Education, were now responsible for the formation of the

government of the Free State. Enshrined within Article 8 of the Irish constitution was a clause that made Gaelic the first language of the nation state and English its second language, and this remains today. Alongside this was an article of the constitution that

36

made Roman Catholicism the religion of the State a clause that was only amended in

1995. The new government initiated projects that would deliver its ideology, designed

to rebuild Ireland's sense of nationhood and pride (McManus 2007; Kee 2000;

Townshend 2006). A revitalised nationalist idealism of a heroic indigenous past,

offered a way forward throughout the post war years, and in celebrating everything Gaelic the concept of an indigenous, independent ̀Celtic' past was created. The two

terms became inextricably linked to a nationalist agenda, and in offering a new

history, Ireland could be delivered from what de Valera once described as the ̀ seven

centuries... of being held by the right of a sword' (de Valera 1918 cited in Kee 2000:

7). As Kee argues, although de Valera's position is understandable given the

background to the fledgling independence, nothing is so simplistic, and the ardent

nationalist sentiment and rhetoric ignored the important legacy of a culture and a history, which was inextricably linked to that of Britain (Kee 2000).

In recent scholarship on the development of archaeology within Irish universities,

attention has been drawn to the problematic political consequences in the post-Treaty

years, which resulted in an ever widening divide between `Unionist' archaeology in

Ulster, and the increasingly nationalist sentiment within the south, whereby those in

power on both sides were reformulating the historical narrative to suit political

purposes (Waddell 2005: 216). Described as a `replacement bourgeois hegemony', it

has been argued that such a reconstruction was designed at a political level and did

not allow for the possibility of alternative histories (Tierney 1998: 194). The new

Free State government funded the first Chairs of archaeology in Dublin, and the first

wave of Irish archaeology students who later came to hold these, have been described

as ̀ strongly nationalist' (Wallace 2008: 168). Those who held the Chairs in Celtic

Archaeology such as Rüairi De Valera (son of the Taoiseach), and their students were,

however, also strongly anti-British in any, and all consideration of `external

influences on ancient Ireland (Wallace 2008: 168). The position of Keeper, at the

newly formed National Museum of Ireland, was offered to scholars from Germany

and Austria, and no assistance or advice was sought from what has been described as

the `land of the old colonial master' (Wallace 2008: 169). The newly formed historical

narrative, focussed as it was on an idealised retrospective of the non-English past, was

to play an essential role in informing the way in which archaeology developed as an

academic subject within Ireland in the years following Ireland's independence from

37

Britain (Waddell 2005: 216). A Catholic-based, social and cultural engagement in a heroic indigenous past entered the psyche of the masses and the concept of an independent `Celtic' Ireland was firmly established.

2.5 The Harvard Expeditions

In the decade before the outbreak of the Second World War, archaeology in Ireland

was the beneficiary of a scheme, designed and staffed by Harvard University. Funded

by the Rockerfeller Foundation, and the Irish government's Employment Relief

Scheme, which provided the labour force for excavation, fifteen historic sites were

excavated over six years. Two lead archaeologists were appointed, Hugh O'Neill

Hencken and Hallam L. Movius (Waddell 2005; Johnson 1999). Before considering their role in developing a relative chronology, it is important to note that the work

conducted during these excavations was the first of its kind in Ireland, and therefore

represents a significant step forward from anything that had been attempted before.

That said, the interpretation of that which was excavated, also needs to be considered in light of political pressures within the newly formed Free State, which provided the

workforce, partial funding, and the licences needed to excavate. The objectives of the Harvard team, which had been agreed with the Irish government, were clearly stated

within the research framework. The research was to inform an anthropological study based on `sociological, racial and archaeological data', a blueprint as it were for a `single, unified anthropological history of Ireland' (Waddell 2005: 219). The aim was to create a chronology of Early Historic sites, based on the evidence for Lagore

Crannog from the Annals. Lagore would become the principal site, the keystone, from

which the relative chronology could be established. Central to this was Hencken's

emphasis on the typology of the zoomorphic penannular brooch, an object he believed

to be a quintessentially indigenous Irish object, the absence or presence of which

would form the basis of his relative chronology (Johnson 1999).

Hencken specialised on the sociological aspects of historic settlement, while Movius focused on the collection of anthropological biometric data, from the skeletal evidence uncovered during excavations. Movius' work has long been consigned to bookcases and superseded by contemporary scientific methods, but the typologies and classifications devised by Hencken are, somewhat remarkably, still widely used

38

today. The specifics of both the material excavated, and the problems inherent with his chronology, will be discussed in detail in the next chapter.

Hencken's work has not gone unchallenged in the past, but despite the detailed

arguments set out by scholars such as Joseph Raftery (1972 & 1994), Barry Raftery

(1972,1994 & 1996), Caulfield (1977 & 1994), O'Floinn (1999), and Newman (1997

and 2005), the work of the Harvard Expeditions continues to be used uncritically.

Even with the benefit of contemporary theory in material culture studies, and models

that seek to understand and explore a social biography for artefacts (Appedurai 1986;

Gosden & Marshall 1999), authors such as O'Sullivan on Lagore Crannog and

Moynaugh Lough (1998), and Cotter, on Cahercommaun (1999), continue to date

sites using the old chronology and typology devised by Hencken, ignoring whole

centuries of evidence, and making no attempt to explore a social or cultural meaning

for the earlier artefacts found there. By pushing material into a short time frame that

fits what is expected, the relevance of the earlier occupation levels has been lost.

Although I have questioned the use of the Annals as a relative dating mechanism, I do

not dispute their importance as a tool for interpreting the archaeology, once

independently dated sequences have been established. They were, however,

commissioned and written at a time when the Roman church was asserting its

authority in Ireland (from the 7t' or 8"' century AD onwards), and it is likely that

aspects of the social, economic, and political base of that time, were transposed back

through earlier centuries to offer a heroic origination for those in power at that time. Ö hUiginn has highlighted the political positioning evident in the various recensions,

or versions, of the manuscript material passed down to us, and has argued that each

version places emphasis on the importance of different tribes and their leaders,

relocating key aspects of the stories to assert an association between tribal power and legitimate over-kingship at Tara (e. g. Ö hUiginn 1990 on the history of the Laigini).

Just as cultural nationalism can be seen to have played an essential role in providing

the catalyst that led to independence from Britain, so too it played an important role in

determining how antiquarian investigation developed into archaeological excavation,

what sites were excavated (by reference to the Annals), how they were excavated, (stressing the historical dates given in the Annals), and ultimately how they were

39

interpreted. The remarkable adherance to the historical narrative is exemplified by the

the lack of interest shown by leading Irish archaeologists, over the most significant

Viking settlement ever found, at Wood Quay in Dublin (Wallace 2008: 170). The

Vikings were seen as `foreigners' and as such their 9th century AD town was not

thought relevant to an orthodoxy which maintained that the indigenous society in

ancient Ireland, was `familial, rural and hierarchical' (Wallace 2008: 170). There are

some scholars, such as Newman, who have created a theoretical bridge between the

traditional narrative, and the insights from contemporary scientific analysis, to offer

an alternative reading of the material evidence within the period under discussion

(1997,2005 & 2007). Although Newman's work is highly regarded within academia,

it remains outside the overarching works of synthesis produced for publication in

Ireland. The most recent work that engages with archaeological re-interpretation for

Ireland has been produced and published by scholars working in Britain. This work

has opened up Iron Age studies to consider parallel social and economic

developments that can be directly evidenced within the archaeology between

communities on either side of the Atlantic and Irish seas in the period before and after

the coming of Rome (Harding 2004,2007; Hill 1995; Mattingly 2006). A healthy

academic debate has sought to challenge the persistent use of the term `celtic', based

as it is on an artistic medium, which has become archaeological shorthand for both

cultural and ethnic affiliation.

2.6 The 'Celtic' Debate

The persistent use of the terms ̀ Celt' or `Celtic' within the archaeological literature

has itself been questioned in recent years. Although Caesar in his Gallic Wars stated in reference to the tribes in Gaul ̀ in their own language they are called Celts, in our

tongue Gauls' (Bello Gallico, 616), it is from the Greek records, namely Herodotus,

that we know of a tribe or tribes in central Europe who were described as Keltoi as

early as the 6m century BC (or the mid 5m at the time of his writings). The Keltoi are

given two brief mentions, both in relation to the Danube River in central Europe,

although neither offers much in terms of description, just the name (Herodotus, The

Histories, 2.37 & 4.39). The problem with the use of the collective term ̀ Celt' has

long been the subject of debate outside Ireland, and in recent years it has prompted an

40

extraordinary exchange within the academic journals between scholars involved in

Iron Age studies in Britain. In a review article in 1998, Peter Wells summed up the

problematic nature of the Celticism debate, by entitling his article `Who, where and

what were the Celts? ' It is not my intention to rehearse the complete debate here, as

this has been comprehensively summarised elsewhere, but the key points are essential

to this research (see Collis 2003; Raftery 2006 for the opposing position). The main

argument reflected an attempt to consider the `Celtic' question in terms of

contemporary post-colonial discourse, a revisionist position, centred around the

relevance and application of the collective term to describe the widely different

peoples throughout the European Iron Age. Those who sought to challenge its

continued use argued that migration was part of the social organisation and

reorganisation amongst all of the peoples in Europe, in the pre-Roman Iron Age. The

difference of opinion centred on its origin within 19s' and 20th century British

colonialist approaches towards the accounts of Caesar and others, and what was

regarded as the civilizing influences of Roman culture, from which parallels were

drawn with contemporary British Imperial policies. Within the period, the term

`Germanic' was similarly revived and it provided a convenient label to differentiate

between those who were civilized under the Romans from those labelled `barbarians'

(James 1999,2000; Collis 2001,2003). Traditionalists in favour of its continued, use did not totally reject its more recent history, but maintained that it was (and is) an

accurate term that provides an umbrella for similarities in artistic style, technology

and culture (Megaw and Megaw 1995,1998; Raftery 2006). The difficulties in

assuming similar cultural traits by the use of an artistic medium alone, whether that is

Halstaft or La Tene, were highlighted by Collis, who argued that such an assumption is invalidated by the fact that the Germanic speaking peoples of Denmark adopted La

Tene artistic style, whereas the Celtic speaking Celtiberians did not (Collis 2003). In

Ireland the evidence for Halstaft type material is slim, and material of La Tene design

is patchy and separated geographically, and it is all but absent from the eastern, south-

eastern and southern counties. This lack of material becomes even more pronounced from the turn of the first millennium AD, with only a few examples of decorated La

Terse style metalwork from the whole of Ireland (Raftery 1994). If we were to

continue to follow Megaw and Megaw's logic then the lack of material evidence

completely contradicts their description of the people of late Iron Age Ireland as `Celtic'.

41

The other contentious issue within this debate was that of the classification of people

based on the `Celtic' language itself. The term `Celtic' is used frequently and

interchangeably within discussions about the Gaelic language base. Questions remain

about whether it can be realistically reconstructed as a further indicator of cultural

similarity. The `Celtic' language in pre-history has been divided into those who

spoke a form of `p' Celtic and those who spoke `q' Celtic, which is regarded as the

older form. The difference, often buried beneath layers of linguistic and philological

terminology, simply relates to the way in which the language was vocalised or

spoken. For example, the `P' in the name Patrick (or Patricius) in the Irish `q' Celtic

form would have been replaced with a `kw' sound to make, `Cathraige'; the god

Neptune is known in Ireland as Nechtän (Ö hOgäin 2001: 136; Charles-Edwards

2000). The study of the `Celtic' language base has a history in linguistics and Celtic

Studies, going back to Lluyd in the 17`h century, who first classified the `p' Celts as

Brythons from Gaul (this is the form which is believed to have been spoken on the

island of Britain) and `q' Celts as Goidelic from Iberia (spoken in Ireland, Wales and

Scotland) (James 1999; Collis 2003: 50). There are however significant problems

with such a diffusionist approach, not least that Irish colonists are known to have

settled in Dumnonia, modern day Cornwall, but what we know of the Cornish Celtic

language is that it was a form of `p' Celtic. Furthermore, a simplistic model such as

this ignores the evidence from inscriptions found on ogham stones in Wales, which

are also inscribed in Latin confirming that at least part of the population was

bilingual.

In recent work, a curious aspect of the debate seems to centre around whether an

author is qualified to comment on these issues. In well received papers, Megaw and Megaw have made explicit their combined Welsh, Scottish and Irish ancestry, and Collis in his most recent work responded by declaring his 'I 8h century Irish descent'

(Megaw and Megaw 1996,1998; Collis 2003). The Megaws did not only describe

their own ethnicity in these papers, but identified the leading Iron Age scholar in

Ireland as ̀ the Gaelic speaking Raftery', the suggestion being that this gives Raftery

an authority that Collis, described as an ̀ English archaeologist', lacks (Megaw and Megaw 1996: 180). There are, however, few academics outside the specialist (rather

contentiously named) ̀Celtic Studies' discipline in Ireland, who can translate

42

primitive, ancient or old Gaelic; modern Gaelic which is very different to these early

forms and has been compulsory in the Irish curriculum in primary and secondary

schools since it was designated the first language of Ireland under the Irish Free State

constitution in 1922. The `Celtic' issue was reviewed by Raftery in his most recent

paper, published outside the usual archaeological journals and in which he describes

the debates as taking place between `British' scholars (Raftery 2006). Raftery argues

that as an Irish archaeologist he is unsure of what Collis means when he discusses the

general reluctance of Irish archaeologists to engage in a de-Celticisation of the past

(Raftery 2006). Yet within this paper, Raftery stresses ̀Celtic' autonomy and

continuity in Ireland, an isolationist position that does not fit well with much of his

own earlier work, and directly contradicts his interpretation of the material from his

excavation at Rathgall in 1970. At this hill-fort in County Wicklow, Raftery found a

Roman-type, military strap-tag (dating to P century AD), and after drawing parallels

with diagnostic material in Roman Britain, he proposed that the first four centuries

AD should be renamed the Hibemo-Roman Iron Age (1970: 211).

What this debate has demonstrated is the misconception that any challenge to the use

of the collective term `Celt', invariably used as a cultural or ethnic affiliate, is in some

way a denial of the existence of people who may have come to Ireland from Europe.

This point was heavily stressed in 1996 by Megaw and Megaw, who accused

advocates of the revisionist position of being anti-European `little Englanders',

engaged in a politicised debate that rejected direct influence from Central Europe in

ancient times, in order place primary emphasis on Englishness (Megaw and Megaw

1996: 180; James 1999). This point has been utterly and convincingly rejected by

Collis (2003). Megaw and Megaw appear to have missed the essential point in the

debate, which is that one cannot classify people based on an artistic tradition,

especially when the art style was itself defined as such by archaeologists in the more

recent past. Given Raftery's most recent contribution, which was essentially a

historical rather than an archaeological piece, perhaps Collis is right, and there is a

reluctance for Irish scholars to engage with a debate that might undermine notions of

a right to a `Celtic' identity and nationhood.

Reconciling the existing chronological frameworks and typological classifications for

the Iron Age in Ireland alongside those as devised for European material is

43

problematic. The local Irish material was clearly influenced by types from elsewhere,

and much may have been imported directly, but even the dating for the emergence of

iron technology in Ireland is still disputed, and it should be added that the use of iron,

in itself, should not be used as a wholesale indicator of the transition from the Bronze

Age. Added to this is the problematic nature of the dating for Irish hill-forts, ringforts

and even crannogs (discussed fully within the next chapter). Broader inter-European

Iron Age comparative studies have in the past used the Irish annalistic material as a

`window on the Iron Age' (see Jackson's study by this name, 1964), but they appear

to ignore the fact that Ireland's Iron Age is lacking in comparable dates or material

evidence to that of central Europe. There are no princely burials such as those at Vix

or Hochdorf, no chariots or chariot burials, no oppida, no evidence of nucleated

settlements or proto-urbanisation, no coinage, no classic hill-forts such as Bribracte in

France, Maiden Castle in Dorset, and no elaborate precious metalwork, such as the

Gundestrup cauldron, from Denmark. The settlement evidence we do have suggests

dispersed occupation, with regional centres that developed in the Ist century AD that

acted as foci for community activity, whether this was religious or secular matters

less, as any clear distinction between the two is difficult for the period. For most of

the material that has been classified as ̀ Celtic' within the Irish Iron Age, the parallels

for similarly diagnostic material are not in Europe, but they are in northern Britain

(Harding 2007). The reluctance within Irish archaeology to accept a British origin for

much of the material imports, or realistically assess a shared cultural background

between neighbouring islands within the Iron Age, may be a further reflection on past

approaches that sought a prehistory for Ireland, before British rule.

Although I understand that in past approaches to the period the use of the collective

term `Celtic' was commonplace, it is surely our responsibility as archaeologists to test

the validity of historicism; social development differed greatly amongst the

population groups both in Europe, and in the western islands, and while

generalisations draw on similarities, they do little to ponder essential differences

(James 1999). Work by Mattingly on `discrepant identities' within Roman Britain

(2004,2006) has offered an alternative interpretive framework for the period, one that

clearly demonstrates the variability in response to Roman control, and his work offers

a clearer picture of how Ireland may have fitted in within the wider Roman world. Work by Dark (2000), Smith (2005) and Campbell (2007) has demonstrated that the

44

centuries after the collapse of Roman political power in the west some elements of

Roman culture lingered on, and this is certainly the case with the rise of the new

religion of Christianity, with all its trappings of late Imperial administration. Apart

from Raftery, only Ö h'Ögäin has attempted to fully reconcile Irish `Celtic' history

with that of the Greek and Roman classical accounts of the `Celtic' peoples (Ö

h'Ögain 2002). Throughout his work Ö h'Ögäin stresses conflict and difference,

polarizing all that was `Celtic' from all that was Roman, an ethnic and cultural

distinction which is not as simplistic as he maintains, and he makes little use of the

archaeological evidence from the vast geographical areas he discusses. Although his

work is rich in detail from the classical sources, it is limited in terms of any

contemporary theoretical synthesis and he applies what is best described as a cultural

replacement model for areas that fell under Roman control. We need only look to the

writings of Hingley (1997,2005,2010), Creighton (2005) and Mattingly (1997,2004,

2006,2010) to understand that models that serve to separate cultural groupings into

`Celt' `German' or `Roman' do not reflect the variability and localised responses to

Roman occupation or Roman cultural influence both within and beyond the Roman

frontiers (Mattingly 2004,2006). Acknowledging why the materiality of Iron Age

Ireland was classified as it was in the past allows us to turn our attention to the impact

this has had on previous interpretations of those who were living and dying in Ireland

during the Later Iron Age.

45

Chapter 3 Fixing the Past

3.1 Identity, origin and materiality

In order for a theoretical framework to be transparent, it is essential that authors state

explicitly how they intend to treat both the present and the past, and the

interdependency of both, in an interpretation of the material record (Hingley 2005:

11). In this way, the ideological intent and motivation of the author can be tracked

through the work and agreement or disagreement with her/his interpretation can form

part of a robust debate. The previous chapter outlined the problematic nature of any

attempt to deconstruct the reified narrative of independent ̀Celtic' culture in Ireland

within the Iron Age and how society has been interpreted as being unaffected by the

cultural hegemony of Roman imperialism in western Europe at that time. If we are to

alter the way in which the past is viewed and relayed through the archaeology, then it

is important that we consider the key concepts used to determine and fix that

narrative. Consideration of concepts such as identity, embodiment and agency are

essential to our understanding of each and every aspect of archaeological

interpretation, but to date few attempts have been made to explore the essential

meaning of these within the Later Irish Iron Age. As will be outlined later, this was a

time of immense political and social change throughout Roman Europe, the ripples of

which spread far beyond its frontiers. As archaeologists we need to tilt the angle of

our gaze towards an understanding of changing identities, a plurality of affiliation,

and the multiple ways people engage with their materiality, and perhaps most importantly how they emulate that of others.

We cannot wholly discount the original production and manufacture of an object, as it

was conceived of and created for a functional or symbolic purpose, whether or not its

later history parallels the intentionality of its conception. If it has deviated from its

original course, if it is classified as exotica or `intrusive', then this too is part of its

history just as our (re) interpretation of its significance, utility or unimportance places it once more into a field of consideration. But we cannot fix or ascribe a relational identity between an object and the presumed ethnicity of its maker or user and label it,

or them, as foreign or `intrusive', on the contrary, the best we can hope for are indicators that might point in the direction of a geographic or cultural origin through

distributional parallels of similar material elsewhere. Where the archaeology reads

counter intuitively we are obliged to look again at the constructs we have created

46

through our interpretation, and cast the net more widely for alternative explanations. I

agree with those who espouse to a multiple narratives embodied within such objects

(Appadurai 1986; Gosden and Marshall 1999), and would go further to suggest that

the biography continues to be created whilst our cognizance, our prejudices and our

contemporary views consider types of evidence as significant, or dismiss them; we

essentially (re)create the biography of the artefact with all our inherent pre-

assumptions. As archaeologists we must consider how the use of things helped people

construct, reconstruct or deconstruct their engagement with the world around them, an

active process that reflects how a community perceived itself and its place in the

world. Such an approach offers a way to understand how, and why, when the

opportunity presented itself, people in Ireland may have engaged with those outside

their more familiar social reference points (Renfrew 2004: 23).

In seeking to understand the distant past through the fragmentary remains of that past,

we do not seek to explore only the objects themselves, but also the material

engagement of humans with that which they left behind. It is people's physical and

emotional engagement with the conception, production and dispersal of artefacts,

within and between the places in which they lived (and were buried), that brings

meaning to our studies and defines the study of materiality. For the archaeologist,

attempts to understand materiality and immateriality reflect the school of theoretical

thought grounded in materialism, a construct that has a background in neo-Marxist

approaches (Mattingly 2010: 284). This has, however, been succeeded by the term `material engagement theory', a term first adopted by Renfrew, and although not followed explicitly in this work, it is one which has inspired much of the following

discussion (Renfrew 2004: 23).

Dialogues in archaeological research have, in recent years, shifted the focus away from seeing the movement of material culture as driven by a ̀ mysterious emanation of human needs', or simply by the determinism, suggested by models of social

manipulation or availability (Appadurai 1986: 29). The consumption of exotic

material does not require a singular or generalised explanation; it may relate, instead,

to a range of different social practices and social classifications. Even within material

culture studies, the biography of objects can challenge preconceptions about

consumption, use, discard and reuse (Gosden 1999; Eckardt 2004). Multiple readings,

1 UNi MERS4Y

47 OF BRISTOL LIBRARY

set within localised contexts, allow for the manner in which objects themselves

functioned within societies to become part of the narrative (Appadurai 1986: 29). The

acceptance of one small improvement in, say, technology, would not of itself lead to

imbalance or change, as it might be used within existing practices. It might, however,

create the favourable conditions that might lead to the gradual acceptance of others. The use, or reuse, of Samian ware at a site may reflect the emulation of a Roman

lifestyle but it does not follow that this is the only explanation for its consumption or discard in the archaeological record. Cultural response to new influences does not

necessarily always lead to wholesale cultural adaptation, or assimilation. It can result in opposition to the expected outcome, leading to a dominant powerful minority being

acculturated into the customs of the traditional society (Bartel 1980). Joseph Raftery

once proposed that finds of `exotic' material in Ireland suggested that the Irish were

adept, and willing, to incorporate new ideas and techniques without compromising

their cultural integrity (J. Raftery 1940 & 1981). Although there is merit in aspects of

Raftery's argument, as will be explained, it does not adequately explain the variability

in responses evident in parts of Ireland.

The concepts and constructs embedded within the word `identity' have been explored

in recent years within archaeological discourse (Mattingly 2006,2010). It is not my intention to rehearse the history of this scholarship but the shift in intellectual thought,

as applied to archaeological interpretation, is central to this research. In a definitional

sense, the term identity implies a set of characteristics that somebody recognises as belonging uniquely to an individual or group. Such a definition, however, masks the

amorphous nature and incredible variability within the concept, which is bound up

with issues of self-expression, conformism or rejection of existing social mores and

the variety of ways these may be actualised, formalised and symbolised through social

constructs, behaviour and material culture (Hodos 2010). The concept of identity has

been explored through theoretical analysis, using research into the variety of social

mechanisms by which identity is externalised, by both the individual and the local

community in which she/he participates, and the ways in which behavioural repetition

and social reproduction verify and validate cultural identity (Hodder 1982,1997,1999

& 2003; Arnold 1999 & 2001; Shennan 2000 amongst many others). In recognising behavioural conditions and constraints we can also assess contextually the

oppositional variance of `difference' or `inconsistency' in behaviour, and it follows

48

how this is represented in ways of `doing', ways of `seeing', and ways of `being'

within the individual life-course of both objects and people.

The study of `identity' has, however, fallen into a semantic discourse of its own,

driven by a need for scholars to position themselves in relation to the meaning they

ascribe to each of the related concepts of `race', `ethnicity', `gender' and other social

classification categories and the need to explicitly deal with newer approaches such as

individual `agency' (Arnold 2001; Joyce & Lopiparo 2005; ), archaeologies of

`memory' (Bradley 1987 & 2000; Rowlands 1993; Alcock & Van Dyke 2003) and

more recently `emotion' in archaeology (Tarlow 2000). It has also been argued,

however, in relation to scholarship concerned with these areas that the need for `thick

descriptions, of course, implies meanings are subtle' (2000: 642). Often buried within

such papers are very simple concepts that if distinguished at the outset might obviate

the need for such density.

One such distinction that needs to be clarified at the outset of this discussion is the

difference between identity and origin, terms that are often conflated or equated to

mean the same thing. I see the concept of identity as wholly distinct from that of the

identification of someone's origin (following Brubaker and Cooper 2000). The verb to

`identify' is rooted in an ability to recognize somebody or something and to be able to

say who or what he, she or it is (my emphasis). The ascription of identity, which is

based on the subjective interpretation of a pre-existing classification of material

culture, pre-supposes that such a form existed, whereas in reality we have determined

its significance. In a recent paper, Hodos discussed the complexity that surrounds our

concepts of identity and she discussed the evolving nature of culture, its duality as a

means of personal and community self-expression, but also its amorphous nature,

constantly shifting and evolving through its need to modify its representative patterns

over time (Hodos 2010). One of the key elements of Hodos' argument was that

identity is inextricably bound up with our mechanisms of social distinction, the

oppositional nature of which tends to clarify our sense of who we are, and to whom

we are bounded in the past, and also in the present. In developing Hodos' argument

we might also discuss the constructs used by archaeologists to identify ethnicity. The

materiality of self-expression provides an ideal mechanism by which we can alter our

persona, and more importantly the way in which we can control how we are perceived

49

both by those within our own community, and those who remain outside. Yet our

origin can be obscured through the recreation of identity using outwardly visible

symbols, which serve to either challenge the orthodoxy of expected social mores, or

allow us to blend into a new cultural milieu. This is especially relevant where we, as

archaeologists, continue to use `foreign' material culture to suggest the ethnicity of an

individual or groups of people. If we accept that identity is socially constructed, fluid

and capable of cultural polymorphism, the application of a theoretical construct,

whether this is defined as a model of acculturation, hybridisation or creolisation

(Featherstone 1999; Welch 1999; Friedman 1999,2000) is less important, as

ultimately what we are debating centres on the concept of change, and whether it is

identifiable, recoverable and, perhaps most importantly of all, demonstrable through

the archaeology.

Cultural change can be evidenced in many ways in the archaeological record. Often it

is discernible as a marked break with the norm, which is contextually and temporally

specific, or as a more gradual behavioural change that might be described as a `conformant transmission' (Shennan 2000: 812). Such a derivation or adaptation

might be either associative and/or visible in a later temporal context. Original

emulation in function, style or technique can be gradually normalized through behavioural repetition, leading ultimately to a derivative form, but one that is

culturally `owned' by the recipient group and is essentially ̀vernacularised'(Bartel

1980). This ability of local people to effortlessly adopt new ideas or artistic techniques and incorporate them within the existing cultural milieu was mentioned

earlier, and it can be evidenced in cases made throughout the Irish archaeological literature from the Bronze Age onwards. Evidence presented ranges from Bronze Age

type axe heads made in iron at Rathtinaun, to a Halstaft type sword worked in bronze

dredged from the River Shannon (Joseph Raftery 1940 & 1981; Ö'Rfordäin 1947; 0'

Kelly 1982; Raftery 1994 & 2005). The ̀ vernacularisation' is exemplified by one

object in particular, the penannular brooch. The brooch has an agreed early 2"d

century AD origin in the Romano-British world, but was gradually developed in

Ireland and has become synonymous with essentially Irish `Celtic' culture (Kilbride-

Jones 1980; Hawkes 1981; Warner 1995). The early forms of the penannular brooch,

however, such those found at both Tara and Knowth are indistinguishable from early types found in Roman contexts in Britain and yet they are classified within a typology

50

that includes the later highly decorated zoomorphic forms (Kilbridge Jones 1980;

Warner 1995). If our position is such that we recognise identity as malleable, then we

must look again at the rigidity imposed by chronological and typological frameworks

that force material into constructed categories based on artistic `development' and `type'. It is entirely possible that groups of artefacts that have been placed into a

vertical typology (A = earliest, D= latest) were in fact contemporary. The

development of form and function may have been entirely dictated by dress custom, fashion or practicality (on clothes for indoor, and outdoor use). In a recent

contribution on this very issue it was argued for the need to `... develop approaches to

typology which are concerned less with defining types and more with describing

multi-dimensional surfaces of variability on which the `type' can be seen to vary with

context [as archaeologists we] `tend to force material into type, style, etc, preferring to

ignore the random noise of individual variability' (Hodder & Hutson 2003: 208). I do

not deny the value of reliable chronological and typological frameworks but suggest

that they are, in a sense, a form of archaeological meta-narrative. As many were

formulated over half a century ago for Ireland, they were a product of their time, and

it follows the extent of knowledge at that time (see earlier discussion on Hencken).

Seeking parallels in stylistic or typological classifications may skew the temporal

sequencing, unless we acknowledge that essential differences occur over time, and

space, and seek to understand `difference' and its oppositional partner `similarity' in a localised depositional context. We can then edge away from sweeping generalisations

that can force material into unified or `expected' timeframes, especially when these

are a poor fit.

As recently argued in relation to work on Bronze Age burials, we see and react to

objects metaphorically and we liken form, function, and usage within our own sphere

of knowledge: we compare, contrast, differentiate, and assimilate (Bruck 2004). It

would be wrong, however, to suggest that all knowledge is traceable through the

archaeology left to us. Knowledge can be passed down the generations through a highly descriptive oral mechanism whereby images and objects can be described in

such a way as to allow for their recognition on first encounter. Just because a

community does not produce an assemblage through excavation that provides a ready

convenient diagnostic form, such as ceramics, it does not follow that the community did not know how to make, shape or use clay. Its absence in an assemblage makes our

51

interpretation more difficult, but not impossible, especially where there are other indicators that ceramics continued to be used for different functions. One such

example is the plentiful evidence of ceramic crucibles, lamps, tuyeres, and other

objects from the Iron Age, a period which in Ireland is described as ̀ aceramic',

discussed in detail in the following chapter.

The study of materiality embraces not just objects made material by people, but the

places of importance to them within the landscape. The little things that have been so readily discarded as irrelevant, add up to a much greater whole if we accept that the

construction of earthworks may also have played an active role in the production of

social affiliations and the need for a communal sense of belonging at this time in

Ireland (Conlin Casella & Fowler 2004). In considering how landscape changes may

reflect broader social change, it is essential that we remember, that land, houses and

settlements or the absence of these are an intrinsic part of any consideration of the

materiality of the period (Renfrew 2004: 28). Demarcating the landscape into territory

contained within boundaries, whether delineated by ditches and embankments,

strategically sited barrows or enclosures, suggests that an element of social or political transformation was altering people's perception of space and territory within later Iron

Age in Ireland. Although it has been argued that the necessity for physical statements

of authority within the landscape altered, once property and inheritance became

subject to written documentation, it is equally likely that adaptation at, or use of, existing monuments performed the same function in a proto-literate society (Renfrew 2004).

A recent theoretical framework based on phenomenology and performativity mapped

correlative landscapes and viewscapes within the wider environs of the Hill of Tara

(Newman 2005). Following work by scholars such as Edmonds (1999) and Cooney

(1999) for the Neolithic, Bradley for the Bronze Age (2000), and in the spirit of Tilley

(1994) and Thomas (1999), this new approach highlighted the evidence of the use of natural topography alongside man-made landscape changes to demarcate the wider territory within which Tara is sited. A visible `boundedness' had been created through

the monumentality of hill-forts, promontory forts and ringforts, positioned alongside naturally elevated places and rivers which had created a natural borderline around the

area. The association and incorporation of Bronze Age barrows within Iron Age

52

enclosures, into a `figure of 8', suggests that importance was placed on their

inclusion within the delineation of reconstructed Iron Age sites (Newman 1998,

2005). Part of the territoriality is also demonstrated through the strategic placement of

monumental ogham inscribed stones that formed part of this complex system of

territorial markers by the 4t' century AD (Newman 2005). Such an interpretation

offers a different way of viewing the sense of place as a demarcated space within later

Iron Age Ireland. It also offers an alternative way forward by which we can gain

greater understanding of the re-use of older monuments within a specific temporal

phase.

Although phenomenology has its place within contemporary theoretical analysis, we

cannot know the motivation behind the construction of a monument, as ̀ time's arrow

is not reversible' (Barrett & Ko 2009: 288). Similarly, we cannot assume that all older

monuments were adopted as territorial markers or formed part of an inter-related

viewshed, given that the landscape would have altered greatly over 2000 years.

Adopting a single, unifying epistemological or interpretative method can often restrict

what is available for reconsideration by limiting it to only that which is directly

applicable within that conceptual framework. This is why it is essential to apply an

overarching theoretical model within which multiple avenues of research and multiple

theoretical approaches can be reconciled, and through which we will come closer to

an interpretation informed by the actual archaeology. If we are to offer a realistic

reconsideration of non-local material in an essentially localised context, then we need

to look at the `specificity of social practices' within each area before building these

into regional or national overarching assumptions and generalisations (Jones 2003:

84).

3.2 Identity in life and death

We know that community activity and burial took place at these historic sites in

Ireland; this is evident from the human, faunal and environmental deposits. These

performances are likely to have been enacted by the use of rituals and feasting, but we

need to be cautious about assuming a level of commonality on an inter site basis. In

the absence of absolute dates from clear contexts at a site what we may be viewing is

something quite specific to both the social group and the time within which events

53

occurred, if you like a ̀ specific cultural moment' (Meskell 2004: 249). The problem

with adopting one site as an exemplar and using it to stress similarities over long time

periods is that what we are essentially suggesting is a form of cultural descent and, by

inference, population descent and continuity. It has been demonstrated that such a

premise might be false, and that where graduated developments in material culture

occur between phases at a site we should not automatically assume a direct descent

grouping (Shennan 2000). In a case study of the Circum-Alpine Neolithic, an area

where sites can be closely dated by high-resolution dendrochronology, time gaps of between 150 and 50 years were evident at and between sites. Identifying a cultural transition with in-situ change could not be demonstrated and replacement population

groups (from elsewhere) may have driven the transition in material culture (Shennan

2000: 818). We should be very cautious in inferring temporal overlap or cultural

continuity within and between what may be the evidence for single or episodic events (Meskell 2004; 249).

At sites in County Meath, within periods we can date with some certainty (the

evidence for which is discussed in detail in the following chapters), the archaeological

evidence suggests that people were using almost exclusively foreign imports. I

mentioned in the introduction to this work that we need to ensure that we do not try to

fix a relational identity based on aspects of the material culture at these sites. Each

cultural moment uncovered through excavation might actually reflect a divergent

occupation or use, or perhaps occupation at a site by an entirely different group of

people using their own material culture. If, however, a displaced or migrant

community find it necessary to `fit in' with local customs, values and social practices,

then the archaeology may be indistinguishable from nearby local sites. Some aspects

are not recoverable archaeologically as, say, in a non-literate society: how would we know if migrants had engaged with local people by learning their language? On the

other hand, the maintenance of cultural references can be used as a deliberate

mechanism to reinforce a sense of community and shared cultural heritage within a

settlement, or as in the case of a trade-diaspora, a deliberate marginality within a host

community (Dominguez 2002: 73). But there is also a further way in which migrant

communities can accommodate the conflicting need to adapt, but also to retain its

sense of identity and this is through a deliberate flaunting of cultural difference played

out with emotive demonstrations and focussed on a sense of occasion. Such deliberate

54

outward displays of cultural nationalism are tolerated because they are timebound and

are otherwise hidden or obviated through the need to conform to the social practices

of the adopted society (Meskell 2005). It is entirely different for those who have never

been displaced, for they do not need to replay the cultural past as it is embedded

within everything around them. But for later generations born to migrant parents or

grandparents, the inherited knowledge of the ̀ Old Country' is temporally petrified in

the past. What is achieved through these overtly cultural displays can be likened to a

negotiation in life, whereby they allow for a short period in which the sense of self

(local), is allowed to surface in tandem with the sense of belonging (disaporic). A

modern parallel is that of the the Irish diaspora community in America, who on one

day each year we given the is opportunity to turn an otherwise mulitcultural city such

as New York the colour green, while it hosts the St Patrick's Day parades.

Other occasions may prompt such a cultural display such as when emotions are

brought to the fore, perhaps when someone dies, and this can impact directly on the

manner in which an individual is buried, and more importantly for archaeologists, the

objects that the mourners choose to bury with an individual. The grave goods buried

with an individual born into such a community may reflect the culture of the group,

rather than the culture of an adopted homeland, and this is an important distinction to

make here. The very act of construction of a burial site, whether it be a pyramid,

mound, ring ditch or grave stone, is one of remembrance; an active reality played out

through time with different actors and performance rituals, but all ultimately seeking

to demarcate a space in the landscape for those interred; a place of collective

=nmemorstion. In considering this issue one area consistently overlooked in terms

of its importance is the mortuary or burial site itself (Jones 2003). The emotion

associated with grief creates one of the most sensually charged social arenas, one that

serves to crate and recr to collective memories and a collective sense of identity

(Jones 2003: 85). The rituals associated with burial practices have been described as being designed to perpetuate and extend narratives of identity, and rooted in either

change or continuity, both of which were intentional and 'relationally constituted' (Jones 2003: 86). These monuments are the physical realisation of the memories of

those who built tbaa As visible statements of 1, mmmmme don within the Wes e

they at racted the iateaest of poets and antiquarians alike in the recent historical past,

and so too they att e4 individuals and communities in pre-history, drawn to adapt

ss

and often incorporate their own dead within the perceived boundaries of the ancestors.

The ideological or political significance of such re-use has been debated throughout

the archaeological literature, with scholars advocating caution in any attempt to stress

continuity either in religious or secular function (Bradley 1987,2003; Eckardt 2004).

Given the often wide temporal gaps in relative chronology, and clear differences in

material culture evident between earlier and later phases, this is a necessary caution

(Bradley 1987,2003; Eckardt 2004). That said, the perceived importance of thew

monuments within the Irish landscape appears to have been recognised, and at each of

the sites under consideration in this research, re-use, adaptation or burial within the

Iron Age is evident.

Burial archaeology, as the study of the remains found within burial contexts, is often

regarded as a separate sub-set of archaeology, and one of the objectives of this

research is to form a bridge between the scientific skeletal analysis used here as an

interpretive tool (the methodology and analysis is detailed in chapters 6 and 7), and

the broader aspects of material culture studies and landscape studies in Ireland, as

outlined throughout this study. A period of nearly two thousand years has elapsed

since individuals were committed to the earth at the sites under consideration here and

we too are now part of the performance of remembrance. Burial evidence is part of

the material culture of a site, and it is as ideologically and politically malleable, and as

readily open to social manipulation as other aspects of material culture (Hodder 1982,

1997,1999 & 2003; Parker-Pearson 1982,1989 & 2003; Chapman 1981 & 2003;

HArke 1998). In recent years, the focus of papers concerned with the archaeology of

death and burial in pest societies, has moved away from the classification of the

physicality of burial, the placement and positioning of grave goods and the symbolism

of both, to the importance of understanding how the burial ritual itself may have

formed a powerful mechanism for social memory (Jones 2003; Williams 2003;

Eckardt 2004). The t matment of the body post-mortem should be interpreted as put of

the on-going social relationship between the individual and her/his community but

perhaps, more importantly, any deviation from the ̀ norm' can be likened to the

represauation of an --p a of negotiation, and continued interaction, between the

living and the dead, and might also represent are egotiation of the position of one

group within the existing social order (Arnold 2001). Mortuary ritual can be wen as a 'mmum entelly aggrandizing form of group expression', the reproduction of which

56

can serve to underpin social norms and configurations (Arnold 2001: 217).

Conversely, any deviation from this reproduction can be interpreted as social change,

the impact of which is mediated and controlled at a local level; just enough to assert a

new elite but not enough to fragment existing social structure, a point which is very

relevant for Ireland.

Funerary and burial practice is deeply rooted within the religious and social practices

of communities and, as these are not altered lightly, where changes can be

demonstrated, it is likely to reflect a deliberate break in tradition (Parker-Pearson

2003; O' Brien 1990 & 1999; McGarry 2007). Changes or transition in burial practices

have been identified at sites around Ireland, and the presence or absence of a

particular type of burial has been used to create a relative chronology for burial

practices within the later Irish Iron Age as a whole (O'Brien 1990; McGarry 2007).

Where foreign material has been found either within a burial context or in an

occupation sequence at a site it has been used to suggest a relative date for the burials,

but at sites where different funerary practices are also evident, burials have been

grouped into assumed 'local' and ̀ non-local' burial forms, the latter being classified

as ̀ intrusive' at the site (O'Brien 1990; Raftery 1994 & 1996). Although some of

these burials may indeed represent non-local burial practices we should be hesitant in

ascribing such a generalised nomenclature where neither origin nor ethnicity has been

established. A level of cultural change within distinct regions within the

archaeological record has been demonstrated, but was it episodic and part of a discrete

regional cultural influence brought about through trade? Or was it, as has been

suggested by some authors, attributable to the migration of individuals who had spent

at least a short time in the wider Roman world, whether it be in Britain or provincial Europe (O'Brien 1990; Raftery 1994; Warner 1995). Given what has been outlined

about the power of social memory and cultural nationalism in the more recent past,

the question pond within this research is whether our picture of the Irish Iron Age has

been blurred by modern constructs of ethnic differentiation and assumed status.

As recently argued in relation to funerary archaeology, there is a long tradition of

overlooking the actual skeletal evidence, in what has been described as a ̀ privileging'

of the save-goods (Gowland 2006: xi). Lowland stressed this important point by

highlighting the general archaeological approach that treated the individuals as

57

offerings to the grave goods, rather than the other way around (ibid. xii). This leads us back to the assumptions which have been made about ̀ intrusive' material and `intrusive' burials within the Irish Iron Age. As will be discussed in the following

chapters, the absence of diagnostic finds of ostensibly local material coupled with

evidence of `intrusive' material at Rdith na Senad on Tara and at Carbury Hill led

scholars to assume that the burials found at the site must similarly have been of `intrusive' people (O'Brien 1990; Raftery 1994). Warner's hypothesis, discussed

earlier, is controversial, but not for its assumed anti-Irish sentiment; what he did was

ascribe an identity based on the assumption that the consumption of provincial Roman

material, ethnically affiliated its users to `automatic membership' of a Romanised

group (Warner 1995; Conlin-Casella & Fowler 2004: 8). It has been demonstrated that

reconstructing identities based on the interpretation of structural symbolism as

expressed through artefacts and/or grave orientation cannot be used in isolation from

other aspects of practical social relations as the burial itself may have been contrived

and constructed by an individual, or her/his mourners, as a form of idealised self-

expression in death (Parker-Pearson 1982; Chapman 2003). We must remain cautious, however, about our reliance on artefact-centred approaches. Although full scientific

analysis of environmental and biological material is much more common within

archaeological reports, these are still usually positioned within the appendices, and findings are often not fully incorporated, or given enough weight within the overall interpretation (Lowland 2006: )i). Much detailed information is contained within these reports that adds valuable insights to the wider social conditions evidenced through the pathology of the individuals at the time of these burials and in this study,

wherever possible, findings from such analyses have been used.

This research attempted to address some of these more recent assumptions about `intrusive' burials that are believed to date to this period. The analysis of isotopes of

strontium and oxygen is both teeth and bone has provided evidence of the

geographical won of individuals, and alongside the analysis of carbon and

nitrogen isotopes, these studies have provided a new insight into palaeodidaey won and even occupations (eg. Richards & Hedges 1998; Dupras 2001; Pike & Richards 2002; Huray et al 2004; Schweissing & Grupe 2003; Budd et at 2003; Burton el a/ 2003). Furthermore, recent work on molecular genetics has

established likely fission points for the dispersal of genetic links-, one interesting paper

58

supporting an Irish/Iberian connection in pre-history with no evidence of a link

between Ireland and central European ancestry (McEvoy et at 2004). We can now demonstrate through a scientific analsysis conducted on some of the burials using

strontium and oxygen isotope analsysis, as detailed in Chapter 7, that the origin of

these individuals is far more complex and would have been otherwise untraceable

through their material culture alone or classification of the burial practices alone.

Whilst I accept that the variability in both quantitative and qualitative data at

individual sites often conspires against the use of scientific assessment more widely,

or as a start point in the analysis, where suitable material is available greater use could be made of it.

Before any new theoretical framework is proposed that might offer a different

interpretation that challenges the construct of an indigenous ̀Celtic' Iron Age past for

Ireland, I want to look first at how post-modem and post-colonial interpretations have

altered the equally traditionalist view of the concept of `Romanisation' within the

territories that made up the Roman provinces during the same time period.

3.3 Rows aid `Roaa. Isatio0': a critical dialogue

In the past the ten years a new movement has taken place within Roman archaeology,

which is best described as a paradigm shift away from the traditional models of Romanisation espoused by Moim sen (1909) and developed by Haverfield (1912).

Within this new discourse authors have reconsidered the globalising nature of Roman

culture within an imperialist and colonialist framework (Mattingly 1997,2003,2004,

2006 & 2010; Hingley 1997 & 2005; Gosden 2004). In an attempt to understand the

impact and localised responses to the Roman cultural influence, the legacy of more

recent imperialism and colonialism within the archaeological narrative has been

brought under scrutiny. The recognition that our archaeological narratives reflect the

contemporary thinking at the time in which they were created led to the first

systematic theoretical review of how, in the case of Roman studies, Victorian and Edwardian mores were reflected back through the work of ac aeologis<s, classisiats

and historians working within the 19th and 20th centuries. It identified the problems inherent within an interpretation that focussed on binary opposition, Roman or native,

military or civilian, a dichotomous and separatist approach reflected in the regular

59

use of the term ̀ barbarian' for those outside what was regarded as the civilising influence of imperialist policies. In past approaches, primary focus was placed on the

process of Romanisation, inferred through the level or extent of cultural change brought about by Roman Imperial policy and administration. In challenging Haverfield's original model, Millen argued that in order to understand the mechanism

by which Britannia became ̀utterly Roman', it was essential to consider what benefit

might have motivated or enabled the indigenous elite to engage fully with Roman

occupation and administration (Millett 1990: 212). For Millett, this participation was

one of choice by those in a position of power who, in accepting the incoming Roman

administration, sought to use the trappings of the new cultural order to maintain or to

further their own social status (Millett 1990: 212). Millett's revised model of Roman

Britain has been challenged successfully by those who have highlighted the

problematic nature of the continued use of a model of impact that was essentially an

inversion of Haverfield's approach (Hingley 1997,2001,2005; Mattingly 1997; 2004,

2006). Using examples of how the existing narrative had been formulated through our

reliance on Roman material in `native' surroundings, arguments focussed on the

limited archaeological investigation into homesteads that did not conform to a Roman

architectural style, such as at Trethurgy in Cornwall or Din Lligwy in north Wales,

even where the material culture found during excavation was almost exclusively

Roman (Hingley 1997; Quinnell 1996; Longley 2001). In fact the range of

architectural types in Cumbria alone had already been used to challenge the use of a

simplistic model of rectangular equals Roman, or round equals native (Higham 1989).

More problematic, however, is the impact of such models on archaeological research,

based as they are on the expectation of predictablity, standardisation and homogenity

in Roman Britain. As a consequence the lack of systematic research into the km

predictable localised engagement with Roman culture in areas such as Cornwall has

resulted in an intapreI Lion based on either resistance or a lack of social and/or

economic development (Thomas 2009).

It has become increasingly evident that the impact of cultural change in Roman

Britain was wholly dependent on individualised response within local circumstances

and this is especially true for Cornwall. The three Roman forts known in Cornwall at Nanstallon, Restormel and Calstock are sited coact to lode bearing mineral resources

and although Rastomel and Calstock have not yet been fully excavated, all are

60

positioned close to navigable rivers, but notably neither appears to have developed

extra-mural settlements (Fox and Ravenhill 1972; Mattingly 2006; Thorpe 2007;

Thomas 2009). It is likely that this is due to the limited period of time in which these

forts were used or that they were not permanent military bases and recent excavations

at Calstock may yet provide the evidence. The site at Carvossa, which has been

classified as a fort, is equally problematic in that the ceramic assemblage is not typical

of a that which would be expected from a military garrison elsewhere as it is

dominated by local gabbroic wares. Coupled with the atypical enclosure, it is possible

that it did not have any military function (Quinnell 1995,1996; Mattingly 2006;

Thomas 2009). In fact the ceramic evidence from Carvossa is much more in keeping

with the assemblage found at domestic sites of the period, known as ̀ rounds' in

Cornwall (Quinnell 1995,1996; Thomas 2009). In the past this lack of visible Romanness has been interpreted as a result of either constraint, such as lack of access

or lack of development due to a continuous military presence, or as an active

resistence to Roman rule but perhaps none of these are applicable. Recent research

has drawn attention to the likely dating sequences from the sites excavated and based

on the ceramic evidence and the development of the range of gabbroic forms from the

3rd century AD suggesting that what we are seeing in Cornwall is Roman influence

emulated in an entirely local manner (Thomas 2009). These new approaches have

allowed for plurality in response to Roman ̀ imperialism', and it avoids both the top-

down viewpoint as espoused by Haverfield (1912), or the bottom-up as proposed by

Millett (1990), both of which focussed on the apex of society without due regard to

the negotiations taking place throughout the rest of the population (Mattingly 2004,

2006; Hingley 1997).

The recognition that variability exists in the level of response to Roman culture and

that it was not as uniform throughout the province as once thought, has opened up Romaa studies to an entirely new interpretive framework. This shift in intellectual

position is welcomed by those of us who we working with Roman material on the

fringes of the Empire. It impacts directly on a further area of study that is directly

relevant to Ireland, the debates about how we interpret Roman material in non-Roman

areas.

61

3.4 Static boundaries or fluid frontiers

The assessment of Roman material outside Roman controlled areas has, in the past, been viewed from an artefact-centred approach and determined to a large extent by

how we perceive the ̀ borders' of the Roman Empire (Hedeager 1978,1987,1990;

Fulford 1985,1989,2005; Cunliffe 1988,2001; Whittaker 1983). There is no question

that military emplacements served as physical statements of Roman control and the

manifestation of the ideology of Empire, but these did not prevent cultural interaction

and the movement of material and people in and out of the frontier zones. The

consumption of material, as mentioned earlier, cannot simply be explained by either

need, or access, as in proposing a single unifying construct based on types of objects,

such as coins or ceramics, we restrict our assessments to just one element of material

culture. In so doing, we are generalising rather than focussing on the individual

circumstances, within a local or regional setting, which determines variability within

the archaeological record. Roman material beyond areas of Roman control has been

catalogued and classified since the work of Brogan in 1936. The classificatory

catalogue produced by Eggers in `romische Import im freien Germanien' is still

widely consulted today (Eggers 1951). The material as catalogued is remarkable, but

it was not until scholars started to address the relative frequencies and disparities

evident within both the types of material found and its location beyond the Rhine

Iimesvorsland that assumptions about the impact of Roman influence on local

societies could be challenged (Hedeager 1977,1987,1999; Parker-Pearson 1989;

Bloemers 1989; Jobst 1980; Roymans 1983; Wells 1996; Wigg 1999; Storgaard 2005;

Gardner 2007; Galestin 2010).

Archaeologists working at Roman sites in Europe have directly challenged past

assýmnptions relating to Roman material found in `native' contexts, and demonstrated

that the material was being used in a variety of different ways none of which could be

determined in advance by our preconceived ideas about whether a site was ̀ Roman',

or `native' (Roymans 1983; Wells 1996; Wigg 1999). The people who lived in an area for 200 kms beyond the upper and lower lines of Roman defence in Europe appear to have shared a material culture with those under direct Roman control (Headeeger

62

1977,1987,1999). Nevertheless, recent research has demonstrated that even within

some Roman controlled areas there are surprisingly low levels of Roman material (Wigg 1999). Survey and excavations over many seasons in the area known from the

Classical accounts as the agri decumates has demonstrated that both the settlements

and the material culture remained essentially ̀Germanic' (Wigg 1999). It is only beyond this `buffer zone' of interaction that there is a gradual distinction evidenced

through the consumption and deposition of a discretely different material

assemblages within the archaeological record (Hedeager 1999; Wigg 1999). Roman

material from southern Denmark has also been reassessed and it has been argued that

there are different social trajectories evident which can be mapped, not just by the

Roman material, but by changes within the landscape and the burials of the period (Storgaard 2005). Thousands of Roman made silver bronze and glass vessels along

with high value Roman coins, bronze statuary and jewellery, have been found beyond

the frontiers, all of which date from the I" to the 4a century AD (Storgaard 2005).

References within the Classical accounts, such as that in Pliny, which attest to a journey by an eques sent by Nero to source amber for a new theatre in Rome, could

only have been achieved if passage for such a trade caravan had been cleared and

agreed in advance with eommunites living outside Roman control (Storgaard 2005:

112). This suggests a familiarity between the local people, local rulers and the Roman

administration and suggests that they were prepared, probably for suitable

recompense, to enable an Imperial convoy to pass unhindered through their territories.

There is also a major shift in burial practices evident in Denmark by the second half

of the 3i' century AD, whereby the extravagance of lavish grave goods at sites such as Varpelov, Valloby and Himlingoje ceases and in its place, it has been argued that

status was implied through the introduction of burial under barrows. This significant

change has been interepreted as an implicit recognition that these individuals no longer needed to demonstrate their social status through deposition of elaborate grave

goods (Storgaard 2005: 119). Burial monumentality in this instance has been

interpreted as reflecting a shift towards a stable, social power base of an elite who

owed their good fortune to their continued relationship with the Roman economy (Storgaard 2005).

The impact of Roman cultural influence on the social structures of those immediately

beyond the frontiers is immensely variable and not easily reconcilable within a single

63

explanatory framework. This is evidenced by the completely contradictory

conclusions drawn by two scholars working from the same data set of Roman finds

beyond the northern frontier in Friesland (Wells 1996; Erdrich 1996 as cited in

Galestin 2010; see Galestin 2010 for a complete review of Erdrich's work, which is

untranslated and unpublished). Highlighting the partiality of both authors and their

emphasis on those aspects of the data that fitted the model being presented, it has been

argued that neither author offered a completely satisfactory explanation for the

localised differences in both the levels and types of Roman material found within and between sites in a relatively small area (Galestin 2010). In many respects it is this

reliance on artefact-centred approaches that has hampered any realistic assessment

that might seek to explain the consumption of Roman material, as the vagaries of

preservation, even between sites, may skew any attempt at statistical modelling based

on quantitative methods. What is apparent from all the data, the interpretative models

presented and the analysis recently published, is that aspects of both arguments have

merit, but Erdrich's attempts to demonstrate limited and episodic movement of Roman material into Frisia disregarded much material that might have altered his

conclusion, whereas Well's model of continuous engagement provides a better fit

with the archaeological evidence (Wells 1996; Erdrich 1996 as cited in Galestin

2010). A similar model of engagement has also been explored for Roman Britain and its connections with communities who remained outside the official Imperial

networks, and it has been described as the ̀ extra-provincial economy' (Mattingly

2006: 591). The Roman objects found in areas well beyond Roman control are part of

a network of exchange and each brooch, each pot sherd, has to be considered important because what we are trying to reconstruct are the social implications of

culture contact amongst ordinary communities. Studies of materiality must start with the fragmentary archaeological record, which is situational and context specific, but

each context and each situation may present different evidence and it is from that

difference, that variability, that we can start to appreciate the impact on the individual's sense of self, i. e. their identity, and how that may have been expressed. A

single sherd of Saurian at a site may reflect a single bowl that was treasured and

repaired over generations, or the remains of a borrowed prototype in a crafworking

shop, or merely the only piece left at a site due re-use or reconstruction but we can no longer assume that the evidence at one site will be replicated across others, as without excavation we cannot know what may or may not be found.

64

Recent assessments of Roman material in Scotland have also challenged many

previous assumptions and generalisations, which were based (as in Ireland) on the

relative numbers of finds beyond the northern frontier (Hunter 2001: 292). What is

now clear is that the distribution and location of finds is entirely different in the west

and the east of Scotland. In the west single finds are common, dotted around different

sites, whereas in the cast, there are exceptionally large enclosed sites, such as Tragrain

Law, Eldon Hill and Edinburgh Castle, fron which a wide array of material has been

found. Further north many hoards, such as those found at Birnie that have been

interpreted as payment for goods or services, attest that some individuals or

communities in remote areas engaged in this secondary ̀extra-provincial economy' (Hunter 2001: 294; Mattingly 2006: 590). The multiple examples of large enclosed

sites dating to the I" to 5 centuries AD in the south-east, with no comparable

examples within the landscape of the south-west, suggests that social organisation was

very different across Scotland during the Roman period. In those areas in the east of Scotland that had been under Roman control, during the intermittent northern

campaigns in the Antonine and Severan periods (and briefly under Constantius), a hierarchical social structure had emerged that appears to have restricted access to high

status Roman goods through centralised control at large enclosed defendable sites (Hunter 2001: 294). In complete contrast a flatter and wider social system is evident in the so th-west, with access to a range of material with finds dotted across the

region. This recent work, and the arguments presented within it, are essential to our

understanding of how the impact of contact was reflected through different social

trajectories and it is key to underýnding the variability in the distribution of Roman

finds across Ireland. In the eastern counties of Ireland there are large enclosed defendable sites where multiple finds have been found and yet single finds are also dotted around the entire coastline and at inland sites. This offers a direct parallel with that which is evident in Scotland, between those areas that were in direct contact and those who remained outside the trading networks.

In more recant work theoretical frameworks have been developed in order that wie might midersgmd the social relations being forged in contact areas. Attempts to assen the impact of Roman cultural influence on those communities well beyond the immediate frontiers has been explored through the concept of `fluid frontiers'

65

(Gardner 2007). Highlighting that past arguments about culture contact have centred

around the initial temporal point of contact, it has been argued that it is not within the

initial period of flux that insights can be gained into social influence, but in the later

periods where it can be measured through change and development (Gardener 2007:

46). Criticisms of past approaches have highlighted the fact that our discussions have

been structured around pre-conceived ideas of discrete cultural entities within and beyond borders. A more nuanced interpretation allows us to view these as frontier

communities, within which there was a push-pull of action and interaction, which

results in a fluid and often unpredictable archaeological record (Whittaker 1994;

Gosden 2004: 35; Gardner 2007: 45). People would have engaged with Roman

material in different ways and for some this may have been an attempt at emulation of

a Roman lifestyle, but it does not follow that all those who were in regular contact

needed to express themselves through such emulation. For many the contact itself

may have altered their sense of community identity, and may have prompted or instilled a need for protection against the might of Rome. For these communities our

evidence of the impact of Rome will not be obvious or predictable and Roman finds

may be completely lacking in the archaeological record at a site. This is where the

related concepts of materiality and identity converge within an overarching framework of transculturality, whereby multiple strands of research can create or

recreate new narratives of engagement.

One theoretical strand evident throughout all of the work discussed above are the

interrelated concepts of agency and structure (Gardener 2007). If we accept that the

impact of Roman policy, in all its diversity and multiplicity of circumstance , was determined by the manner in which people reacted and responded to it, then it is the

individual who is the agent of change or continuity and who will inevitably determine

the outcome of that impact. The individual will be governed to a great extent by the

structure within which she or he exists, and in turn, individual access may create

change at at wider social level where emulation becomes a catalyst for change. Perhaps the best way to view the impact of Roman policy and administration is

through a sause of refraction, where no two responses will be identical, but each is the

outcome of a single source by which their mutual interdependency is conditioned.

66

3.5 Transcukurality as an analytical tool

The theoretical frameworks favoured within archaeological post-colonial discourse in

relation to both the processes and effects of culture contact are those of acculturation

and hybridisation. In a definitional sense acculturation is both the process of sharing

cultural values between two discrete cultural groupings, and also the consequence of

that process (Kottak 2007). The process by which new values, beliefs or social

practices are socialised may be sudden or graduated, and depend greatly upon the

circumstances which acted as catalyst for that contact, but it is recognised that

although cultural groups may remain discrete, both are inevitably altered by process. Within the new archaeological discourse of globalisation, hybridity has been

redefined as cultural hybridisation, and it has been argued that in colonial

circumstances, either in the present or the past, it is sufficiently robust to incorporate

the previously neglected issues of negotiated localised responses within what Bhaba

defined as a `third space of enunciation', i. e. a cultural outcome which is both similar

and different (Bhaba 1994: 54; Knapp 2008: 60). There remains, however, an

underlying relationship between these constructs and colonialism, and a sense that

both terms have been developed within archaeology to explore the types of processes likely within a given get of colonial circumstances, and that the cultural outcome is in

some sense predicated on there being a dominant group. Although each of these

theoretical approaches have been developed in archaeological research in recent years (Friedman 1999, Featherstone 1999, Mattingly 2004,2006,2010; Gosden 2004;

Hingley 1997,2005,2010; Hodos 2009,2010), to assess localised responses to

culture contact, (whether in colonising situations or less organised population

migration), there remains one further underlying assumption and that is that the

process of culture contact led to the creation of a discrete transition in material culture

or social practices. As outlined throughout the previous chapters it is only in certain

parts of Ireland that contact with the Roman world can be demonstrated through the

archaeology, and even in thew areas, the response was not uniform. If, as suggested by scholars. Ireland remained outside and independent of cultural influence from the Roman world, why are the material assemblages at so many large enclosed rtes with

a historic, royal or religious pedigree, almost exclusively made up of foreign imports?

There are aspects of the materiality both within the landscape and at a social level that

appear to have been directly influenced by events that were happening outside Ireland

67

both before, during and after the Roman period, and although parallels have been

drawn with levels of impact on other societies, who were similarly beyond the frontiers of Rome, there are discrete differences in the evidence from Ireland, this is

best described as discrepant levels of engagement and emulation.

The obvious argument against proposing any new theoretical construct within which

these responses might be reconciled is that there are already so many being applied

across different aspects of Iron Age and Roman studies and their application has

prompted some vigorous dissent amongst scholars. The ̀ pernicious' influence of

postmodernism within archaeological dialogues was recently likened to a `neoconservative ideology'(Trigger 2008: 191). Clearly, replacing one form of

unifying narrative, beloved by culture-historians, with a modern approach which `reifies diversity' may ultimately lead to a circular relativism (Trigger 2008: 191). It

might be argued, however, that acculturation, assimilation, hybridisation or even

creolisation we more appropriate in a study such as this, but as explained earlier in the

discussion each model has its own inherent assumptions some of which simply do not

provide a good fit with the archaeological record. Our cultural trajectory is not

always predetermined, it wavers, shifts and changes as we move through out lives and it can be wholly dependent upon individual circimýstances; just as we apply agency

within our lives. so too, did people in the past (Welshche 1999). One theoretical

approach used in relation to our understanding of the modern global culture may fit

better, and it is that of transculturality. Transculturality, in a definitional sense, relates

to how cultures cross or intersect, but it is an approach which takes us beyond a

predefined cultural group or a predetermined outcome, to reject generalised

assumptions about perceived cultural difference in order that we may seek out the individualism of response to each or any cultural influence (Welch 1999: 206). In

exploring aspects of adaptability, which are affected through circumstance and

occasion, interpretation can allow for a mutiplicity of outcomes, read through artefact,

person or landscape. The acceptance of a transcultural archaeology for the materiality

of ancient Ireland allows for a consideration of the mutable nature of identity, played out not just between individuals and c ommuntites within Ireland, but also through the

transfer of less tangible evidence, the engagement with ideas and emulation of those in the world beyond Ireland's shores. It also allows us to include the probability that

some of those who triggered the engagement were foreigners who lived and died in

68

Ireland but who considered themselves, for want of a better term at this point, `Irish'.

Others may have lived within small diasporic communities, and actively chosen to

separate themselves to maintain their cultural integrity. Such a mechanism offers a

greater understanding by accepting that strands of material culture, some local and

some not, can run in parallel with each other. These may occasionally touch, leading

to vernacularisation or be completely separate and create uniquely foreign

assemblages such as those found at Tara and Dün Ailinne.

More importantly, however, is the inclusive nature of a transcultural theoretical

approach in archaeology, as existing theoretical constructs such as agency and

structuration, phenomonology and performativity, materiality and immateriality, all fit

well within its overarching open framework. So, too, does middle ground theory as

the mechanism by which an existing culture is exposed to new values, and gradually becomes altered by the take up of other values (Young 1991; Gosden 2004: 31). There

is a direct parallel between the archaeology of Iron Age sites in southern Britain,

before the Claudian invasion, and the archaeology of communities in the eastern

seaboard counties of Ireland, from the latter part of the I' century AD. In both areas,

the archaeology suggests an interrelationship that would have created the environment by which mutual values were engendered (Gosden 2004: 31). An approach to

interpretation that recognises the diversity of the archaeological record can mediate between past culture-historical approaches, which rely on a sense of national identity,

and the need to offer a new path through what Meskell has eruditely described as ̀ the

unstable terrain of interrelationships between sociality, temporality, speciality and

materiality' (Meskell 2005: 2). Multiple discourse, multiple hypotheses, and

multivocality generate the intellectual conditions to push forward and challenge our

understanding of the individual, the community and perhaps more contentiously the

pre-defined, pre-determined cultural norms we have constructed for them. Before we

can deconstruct pest interpretations and offer an alternative view of the past, it is

essential to review the material evidence for both local and imported material in Iron

Age Ireland.

69

Chapter 4 Problems with the Iron Age transition

4.1 The Loeser Irish Ire. Age: problems with dating temporal transition

The previous chapters have been concerned with outlining the background to why a new

overall theoretical framework is needed for the Later Iron Age in Ireland. I want to turn

now to the actual material evidence and the exist classification for both artefacts and

sites before offering a way forward out of the circularity of interpretation that has

hampered past analyses.

The archaeological evidence for the Iron Age in toto, (Fig. 4.2) compared to what we have for the Bronze Age, is limited. The lack of known settlements sites led Raftery to describe the people of Iron Age Ireland as invisible (Raftery 1994: 112). What little

material evidence we have appears to cluster in the nom, eastern and south-

eastern counties. Fuwthe more, as mentioned within the introduction, there is at present no

academic consensus on the absolute dates that mark the commencement or a realistic

cessation of the Irish Iron Age (Raftery 2005; O'Drisceoil 2008). Scholars have struggled

to set a meaningful unencement as early as the 8d' or 7' century BC, to allow for

comparability with datable material from Europe and Britain (Raftery 1994,2005).

Traditionally, the date most frequently used to mark the end of the Iron Age was AD 432,

the date of the missionary activities of St Patrick, which allowed for everything thought

to date from the mid 5' century or later to be classified as ̀ Early Christian'. This term,

which remains in use even within contemporary analysis, is problematic, given that it is

as best inaccurate (there is simply no archaeological evidence to support wholesale

conversion to Christianity at this early date), or at worse somewhat pejorative (reflecting

a 19' century interpretation which stressed the importance of the Christian mission and a Catholic hegemony). In more recent work, the term Early Historic appears to be more

commonplace, and with some qualification it is preferred throughout this work (O'Sullivan 1998; Becker et al 2009). A third descriptive term, which sears to be favoured amongst those writing outside Ireland, is the Early Medieval period and their

research relating to the 4' and Sa centuries is usually published in the Journal of

70

Medieval Archaeology. Finally, one American publication has added to the confusion by

classifying the Later Irish Iron Age as the ̀ Later (Roman) Iron Age' in the introduction to

a recent monograph on the site of Dein Ailinne, County Kildare (Johnston 2007). Added

to the difficulty with the lack of consistency in terminology is the reluctance to consider the possibility that some archaeological sites, namely crannogs and ringforts, may date to

the period of transition in the Ss' and early 6* centuries AD.

What is apparent, however, is that although the names given to the periods are currently interchangeable, the temporal sequence has remained the same as that for the original `early Christian period', with a start point from the mid 5 century AD. We have simply

changed the name and retained the emphasis on Christianity acting as a catalyst for

change. The following chapter will firstly outline the existing relative chronology for

artefacts and sites and highlight some of the inadequacies inherent within traditional

interpretation, before turning to assess how dates from recent excavations around Ireland

are challenging some long held assumptions.

4.2 Reladve Crro. olopº: artefact ebwilieatiom and h'Polo r

In continental Europe during the 86 to e century BC archaeological evidence suggests a flourishing iron working tradition, known as the Hallstatt, yet in Ireland there is little

more than a handful of objects that are of Halstatt type (Raftery J 1972; Rafery B 2005;

Lynn 2003; O'Sullivan 2003). A few examples of sword chapes and brooches have been

found (Fig. 4.1), and although these carry decorative work suggestive of a Continental

Halstau influence, their find contexts are shared with Irish Later Bronze Age material of the Dowris Phase (c. 700 BC).

71

 . U r  sý'ý Y  d  . " Is"

 "" o Q, O

X

19

Fig. 4.1 Left, distribution of Dowris Phase Late Bronze Age artefacts (after Woodman

1998) and at right, distribution of Halstatt style objects (after Raftery 1994)

The site that has produced significant evidence for this early period is Emain Macha,

Navan Fort, in the north-east of Ireland, which has been dated to the 7th century BC

(Lynn 2003; Raftery 2005: 140). Although the Bronze Age to Iron Age transition appears

to be attested at sites such as Rathtinaun in County Sligo, the material assemblage can be

problematic. At Rathtinaun, J. Raftery suggested some of the objects were of Hallstatt

style, but they were rendered in bronze, rather than iron. Iron working was, however,

evidenced at the site also, and radiocarbon dates suggested this activity took place in the

2°a century BC, which would suggest a very late date for metalworking in a Bronze Age

tradition (Raftery J 1972; Raftery B 1994,2005: 178; O'Sullivan 2003). Of the known

Halstatt type pieces of metal-work from Ireland, most have been recovered in or close to

rivers suggestive of a practise of deliberate deposition, again reflecting earlier Bronze

Age practices that appear to have continued through the Iron Age (Raftery 2005). At

crannog sites such as Moynaugh Lough and Lough Gara, there is evidence of later

Bronze Age occupation followed in sequence by Iron Age occupation and craftworking a

chronology verified through radiocarbon dates (O'Sullivan 1998; Frendengren 2003)

72

The technology of iron working has been recognised at newly excavated sites around Ireland and some of these have produced quite early dates, however, no diagnostic

artefacts have been recovered (Raftery 1994: 227; Warner 1995: 29; Newman 1998: 130;

Becker et al 2008). Metalwork in the curvilinear later decorative style known as La Tene

has become synonymous with the ̀ Celtic' Iron Age (traditionally dated from the 3`'

century BC to the turn of the 1" millennium AD) in Ireland and yet, with the exception of

the more famous pieces of metalwork such as the Monasterevin or Bann discs, the Petrie

crown and the Attymon horsebit, decorated metalwork is rare within the Irish

archaeological record. Apart from the occasional decorated horse-bit, such as those found

in Ulster, the small pieces of horse-gear, bridle bits and curious Y shaped pendants found

across the northern half of Ireland might fall anywhere within the entire Iron Age. In fact

arguments have persisted over the dating of one such decorated specimen, the Attymon

Horse bit, with a range of dates offered between the 1' century AD and the 6t' century AD (Megaw & Megaw 1977; Mitchell 1977; Kelly 1983; Raftery 1994). More recently it has been suggested that much of the decorated metalwork that was classified as La

T6ne in the past actually betrays Roman influence in its design (Harding 2007). Although

past analyses have suggested Ireland had a thriving indigenous metal-working tradition

throughout the Iron Age, many of the examples used to support this wholesale Irish

tradition come from Ulster and link well into the material culture of Scotland in the Pre

Roman Iron Age (PRIA) (Warner 1976; Harding 2004). There are, of course, exceptional

sites in the north of Ireland such as Lisnacrogher in County Antrim, where antiquarians

recovered an important collection of La T6ne swords, scabbards and spear butts (Raftery

1994,2005; Lynn 2003). Although there is little in the way of recorded details of the site itself, the mention of underlying brushwood piles and the preservation of the material

suggests it was water-logged, and it is likely to have been a crannog. The link between

Scotland and Northern Ireland is strengthened by finds such as those from Beirgh in West

Lewis, from where the only known moulds for spear butts of the Lisnacrogher type were found, within a dated 3`d century AD context. The site of Beirgh offers an important link

between the two islands at this time, as within the same dated contexts were finds of decorated hand-pins which offer close parallels with those found within a northerly

73

distribution in Ireland

(Harding 2004). In Ireland,

however similar hand-pins are

generally dated to a much

later period, the 7t' century

AD, but this is based on

relative dating methods, and it

may be that they were in use

in Ireland much earlier than

previously thought (O'Floinn

1999; Harding 2004,2007).

Antiquarian investigations

were also responsible for the

finds in 1798 of four sheet

bronze trumpets discovered in

a small lake close to the

summit of Navan Fort (Emain

Macha). The name of the lake, Loughnashade, translates as the `lake of jewels' and the

discovery of human skulls alongside valuable metalwork suggests that this was a place of

ritual deposition (Lynn 2003; Raftery 1994). The trumpets are similar to those depicted

on the famous Gundestrup cauldron from Denmark, which is likely to be of Thracian

manufacture dating to the 0 century BC (Cunliffe 2001). Parallels have been drawn

between two gold buffer torcs found at different sites in Ireland and stylistically similar

torcs from continental Europe, which have been dated to c. 300 BC. The Clonmacnoise

Torc is a large highly decorated gold buffer torc and parallels have been offered with

gold torcs in the Musingdon-Dux Horizon of central Europe dating to this time (Raftery

1994 & 2005; Harding 2007). A gold hoard, found at Broighter in County Derry,

consisted of a large buffer and twisted torc and a miniature boat (complete with tiny

oars), a small dish or bowl and two woven necklaces, all wrought in gold and, again, a

relative date based on parallels found in the Classical World places it around the Is'

74

Fig. 4.2 Distribution of La Tene artefacts (after Harding 2007, redrawn by author)

century BC. Though the buffer torc is decorated in La Töne style and it can be paralleled

with a torc from the Snettisham hoard found in Britain, a Mediterranean origin has been

confidently offered for the two smaller, rod-twisted torcs, while the small gold model of a

sailing boat has been paralleled with the descriptions of the crafts of the Veneti tribe of Northern Gaul (O'Kelly 1989; Raftery 1994 & 2005; Warner 1982; Harding 2007).

Imported material was also found at the excavations at Ballydavis, in County Laois,

which revealed a large cremation cemetery with `exotic' grave goods. Amongst these

were glass beads of Guido's `Oldbury type', a brooch of Nauheim classification, and an

exceptional find of a highly decorated tinned bronze box. The closest parallel for the

decorated box is amongst the finds from the middle Iron Age site of Wetwang Slack in

East Yorkshire, dated to c. 300 BC (O'Brien 2003; McGarry 2007: 219). The cremations

at Ballydavis were radiocarbon dated to 360-46 Cal BC (GrA 13594,2140 ±50), and are

comparable with Wetwang (Brindley and Lanting 1998: 6). Similar brooches and beads

were also found during the excavations at the satellite burials at Kiltierney in Fermanagh,

and Oranbeg in County Galway, and have been given a relative date of the l' century BC

to the 1st century AD, based on these parallels (Raftery 1994,2005).

In the west, in County Galway, the highly decorated Turoe stone exemplifies the La Tene

artistic medium on a standing stone, and although parallels have been made with decorated stones of the same period in Brittany (Cunliffe 2001), these are not convincing. The decoration on the Turoe Stone is of superior La Tene form and it has been offered a better parallel with the complex designs found on metalwork such as that on the Cork

Horns, thought to date to the 2nd century BC (Koch 2006). From the northern counties of Fermanagh, Tyrone and Cavan are examples of `Celtic' stone heads or idols that have

been more convincingly paralleled with stone heads from Gaul and Galicia and for which

a comparable date of c. 1 ' century AD has been suggested (Rynne 1972; Cunliffe 2001).

Bronze lathe-turned bowls, such as that from Keshcarrigan, and wooden bowls and

cauldrons, such as those from Fore and Toar Bog in County Westmeath, have also been

ascribed a date between the I" century BC and the I 't century AD, and the stylistic

parallels offered are from Stanwick in Yorkshire, Lamberton Moor and Di mfriesshire, all

75

northern British and dated to the end of the PRIA (Raftery 1994; ADS 2005; McGarry

2007; Harding 2007).

Evidence of material that is classified as of La Töne design is found much less frequently

in the southern counties of Ireland (Raftery 1994: 227; 2003: 145; 2005). Access to high-

status material appears to have been geographically limited to certain areas, and this

pattern becomes even more pronounced from the turn of the first millennium. Although

material is restricted in distribution, and chronological certainties based on diagnostically

sensitive items are rare, the period is widely cited as a ̀ La Tene cultural horizon' for

Ireland. Despite the fact that many of the artefacts just mentioned appear to have been

imported, the decorative style has become synonymous with the ̀ Celtic' Irish Iron Age

(Raftery 1994: 219).

To date, highly decorated objects such as ring pins, hand pins and penannular brooches

have been treated as distinctively `Irish' objects and classified into a typology based on

style and decoration. These are useful as indicators of chronological development, but

less useful for dating purposes, given that are rarely based on independently derived

datable contexts. The zoomorphic penannular brooch is one such example and as discussed in Chapter 2, the existing chronology and typology was devised by Hencken

based on his excavations at Lagore and it has been refined over many generations of

scholarship, most notably by Fowler (1960,1964), Kildbride-Jones (1980), Newman

(1989) and Johns (1996). Classified by the development of the zoomprophism on their

terminals and the development of the free moving pin, the typology is based around main

groups (original form) and subgroups (developed forms). There are many problems with

the existing typology, not least that if we accept Fowler's extended and immensely

complex sub divisions, they would appear to have both a contested origin either amongst

the Iron Age communities of Iberia or in the area of Yorkshire inhabited by the people known to us as the Brigantes, and an equally contested development (either Irish, British

or Romano-British: Fowler 1964). In a more recent contribution to the classification

series for some of the later types of penannular, it has been argued that those Fowler

dated to the later period of the 6m or 7t' century AD should be assigned a 5t' century AD

76

date (Newman 1989). Furthermore, both the origin and the development of the

classificiation has been challenged, based on finds from secure contexts in Ireland

(Newman 1989). These functional little items could have had many and varied forms

across a similar time period, reflected in the similarity of decorated zoomorphism across different types of pins. We need to explore the structure and accuracy of the existing

classification and typologies because even with the contributions of scholars such as Newman (1989) and more recently O'Floinn (1999), the older classifications are still being used as a form of relative dating.

4.3. Hencken's chronology and typology

The chronology and typology used to separate Iron Age material from that of the ̀ early

Christian' period was devised during the excavations which took place between 1930 and 1936, under the Harvard Expeditions to Ireland, outlined in the previous chapter. Of the

four important sites which were excavated by this team, three were crannogs (Ballinderry

1 and Ballinderry 2 and Lagore Crannog), while the fourth site, Cahercummaun, was a

massive stone fortress, or cashel, in County Clare. Crannogs are artificially-built

palisaded lake enclosures which are common across Ireland and are traditionally dated to

the 7m century AD or later. Although crannogs are notoriously difficult to excavate, the

anaerobic conditions create an ideal environment for the preservation of organic material. Within the introduction to his report on the excavations at Lagore Crannog, published in

1950, Hencken noted that the manuscript dates did not appear to fit with many of the

objects he found in the earliest phases of the site. Despite this important fact, he stated

that he had felt it appropriate to date the artefacts in accordance with the ̀ known' history

of the site (Hencken 1950: 6). The site had been identified from the Annals as ̀ finis Locha

Gabhar', a royal site of great importance and mentioned in various entries between AD

676 and its recorded destruction in AD 934 (Hencken 1950: 24).

The site was exceptionally rich in finds, the majority of which can be said to fall into the horizon of the 7t` to 9m centuries AD. There was also an assemblage of much earlier

material including material of likely Roman origin, the existence for which was either

77

explained as heirloomed or treated as irrelevant to the dating sequences overall. For

example, the remarkable assemblage of glass beads, many of which were exotic and non- local, were described as ̀ such common-place things on similar Irish sites and are so

numerous on European sites in general from the Bronze Age onwards that they have no

special significance' (Hencken 1950: 134); given that Lagore was one of the first

comprehensively excavated sites dating to this period, it is unclear what other `similar'

Irish sites he was actually referring to. Alongside many artefacts for which the closest

parallels are in the Roman world, such as swords, spears and axes (Fig. 4.3), was a

collection of material of distinctively northern British origin. These included a collection

of complete glass armlets, which Hencken acknowledged as Kilbride-Jones Type 2, dated

to the 2nd century AD. These `D' shaped

bangles are commonplace on sites in

northern Britain, and are found within the

same contexts as early Roman material

(Harding 2004,2007). Hencken recognized

that much of the early material such as

weapons, bone hair pins, tools, Samian ware,

mortaria and domestic glass had `Roman

connections', but by separating material into

either `Roman' or `British', he ignored the

fact that much of this material was

contemporary, and when taken as a

collective it adds greatly to the weight of

evidence for a period of earlier occupation at

the site. (Hencken 1950: 161,191,193;

Kilbride-Jones 1934). Several Roman-type

copper alloy toilet instruments were also

Fig. 4.3 Illustration from Wood- Martin (1884) of Roman type swords, spears and axes from Lagore

found along with Roman type bronze barrel locks and heavy keys. Parallels for these

within Ireland are found amongst the Roman material from Räith na Senad and at the site

of Uisneach in County Westmeath, where a 4`h century AD Roman coin was found within

the same context as a Roman-type key (Macalistair and Praeger 1921; Warner 1995;

78

Halpin and Newman 2006). In the published report Hencken noted that it was

problematic to dismiss so much material of Roman or British character but he remained

convinced of the relevance of the history of the site within the Annals: the site as a whole he dated to between the 7th and the 10th centuries AD. This is the accepted dating for

Lagore and it follows the material found at the site, which has become the principal

reference against which all similar material and sites are dated.

Hencken's excavation notes (topographical files NMI: consulted 2008) are in the

National Museum of Ireland and they differ somewhat from the published work in

several ways. In his notes, Hencken described the crannog as a ̀ layer cake', and states

quite categorically that the earliest phases he recognized were at the edges of the area

under his excavation and stratigraphically related to the centre of the crannog, which was

not excavated. Hencken also recorded that ̀ an older structure appeared at the bottom of

the island, beneath the brushwood layers' (Hencken 1935: 157). A reassessment of the

stratigraphy of the crannog that was reconstructed from published drawings found that

there was significant evidence to suggest that this structure was in fact an early

occupation sequence that Hencken missed (Lynn 1983: 52; 1985-86: 70). If the

brushwood piles noted by Henken are the evidence of occupation in the earliest phases,

then the stratigraphical relationship to the centre, and to the earlier material, becomes

much clearer. In his notes after the second season of excavation in 1935, Henken added

that he found evidence of several overlapping houses with hearths at different levels, and tellingly states that ̀ the greater part of them lies in the unexcavated centre of the

crannog' (Hencken 1935: 158). This led him to conclude, in his own words, ̀ it is quite

possible that the crannog was in use for some time before it was adopted as a Royal site' (Hencken 1935: 158).

Lagore had been the subject of intense antiquarian interest to the extent that back in 1884

Wood Martin had reported on the collections of old finds, and his work is instructive in

highlighting the sheer level of looting which took place before Hencken's excavation. Wood Martin also reported that at the time of his explorations, two extended skeletons had been found buried within the centre of the crannog. Although the local people would

79

not allow the burials to be removed, he noted that ̀ one specimen was however secured

and deposited in the museum of the Royal Irish Academy' (Wood-Martin 1884: 204).

This suggests that at least two extended unprotected inhumations were buried in the

unexcavated centre of the crannog. If Hencken's notes, rather than his published report,

are used then the areas he was excavating can be explained as the extensions to an older

crannog in the middle of the site. The early Roman and British material would appear to

be related to an area which remains unexcavated, and given the sheer scale of looting and

clearance of human skeletal material described by Wood-Martin, the stratigraphy in the

area Hencken did excavate is likely to have been significantly compromised. There are

parallels for more than just the Roman and British material that was excavated at Lagore,

not least amongst the material Hencken found so difficult to date which he classified as `Teutonic'. Most notable of these was the anthropomorphic carved wooden idol, which

can be paralleled with similar idols from sites such as Oberdorla and Hedeliskaer

(Hencken 1950; Kaul 2003: 34). Lathe turned wooden vessels, such as those found at

Lagore, are commonplace at sites traditionally dated to the ̀ Early Christian' period, but

as they are often un-datable, it seems extremely probable that they served the same

purpose, or function, within the Later Iron Age. This is worth considering further, as the

closest parallels for many of the bronze banded wooden bowls and lids from Lagore, are from sites such as Hjortspring in Denmark (Kaul 2003). In Denmark these have been

dated securely to the I 't or 2od century AD, in some cases by associated Roman coin fords, and interpreted as locally produced, wooden imitations, of ceramic fine wares. Many of the swords from Lagore are similar to the Roman gladius and spatha (see Fig.

4.3) but three have a shallow groove down the centre of the blade, the parallels for which

are with swords in use in the ̀ Germanic' world during the Roman Iron Age (Edwards

1996: 89). Although Edwards has argued for an Anglo-Saxon parallel for the ̀ slave' or dog collars found at Lagore, it is clear that the design of the chain still attached to the

collar, which is of figure of eight chain link (clearly evident from Hencken's drawings),

also had an origin in the Roman world (Johns 1996; Edwards 1996: 88).

Despite Hencken's confident assurances of the integrity of his contexts at the site, a river had to be diverted to excavate, and it is extremely likely that the saturated zone and

80

constant river irrigation would have affected greatly the contexts in which these objects

were found. In keeping with Lynn's suggestion of earlier occupation sequences at the

site, it seems likely that the earlier material was contemporary with the period of

occupation within the centre and west sections that remain unexcavated.

In 1934 the Harvard Expeditions excavated another important site, a large well-preserved tri-vallate stone fort, or cashel, at Cahercommaun in County Clare. The results of the

excavations were published in the Journal of the Royal Society ofAntiquaries in 1938.

The site sits in a prominent position on a limestone plateau overlooking the bleakly

beautiful landscape of the Burren. There are at least 20 such cashels known in the area

and the stone fort sits within a landscape of prehistoric monuments, which attest landscape occupation over millennia. The excavations over a six week period produced a

range of material, some of it quite early: worked flint, saddle-quern stones and stone

axes. There was an array of metal artefacts including an iron shears, knives, and socketed

and pronged tools along with other objects that are not closely datable. Some decorated

metalwork was found, including a zoomorphic penannular brooch, the head of an

enamelled brooch pin and an iron penannular brooch. In 1999, Ö' Floinn reviewed the

decorated metalwork from Cahercommaun as part of the Western Stone Forts Project.

Much of the material was un-datable, but some of the metalwork was discussed in detail,

and Ö'Floinn argued that although a great deal dated between the 7'h and I& century,

several pieces clearly pre-dated this and were more likely to be 5th century in date and

suggesting an earlier occupation sequence at the site (Ö'Floinn 1999: 74).

As long ago as 1972, the dating sequences of sites excavated by Hencken and his team

were challenged by Raftery who, following his own excavations at the hill-fort of Freestone Hill in County Kilkenny, argued strongly against Hencken's dating sequences for Cahercommaun, and a further site excavated by the Harvard team at Ballinderry No. 2 in County Offaly (Raftery 1972: 53). The results of the excavations at Ballinderry No. 2

were published in 1942 and the site was dated to the 8"' century AD (Hencken 1942; Newman 1997). In keeping with his approach at Lagore, Hencken placed primary

emphasis on the ̀ known' historic dating of the site, and material that could not be

81

reconciled was treated as long-lived. Romano-British coarse ware (probably late 2nd

century) and a sherd of Arretine ware (c AD 20) were recovered within the lower phase

of the site but these were dismissed by Hencken, who suggested instead that Roman

ceramic styles had continued to a very late date, something which cannot be

demonstrated or supported archaeologically (Hencken 1942: 56). Amongst the recorded finds for Ballinderry was an assemblage of early material clearly inspired by Roman

types such as ball-headed hair-pins. An earlier date is also suggested by a gaming piece

of Irish rectangular style, marked by the ogham character for `v', in place of the Roman

numeral of same character (Hencken 1942: 57). Despite O'Floinn's work to correct the dating sequence at Cahercommaun, in the summation of the report to which he

contributed, the editor appears to completely ignore both his findings and their relevance

to this important site. The editor concludes by stating that Hencken's chronology was not `too far off the mark' implying that the five centuries in between Hencken and O'Floinn's

conclusions are obviously not significant (O'Floinn 1999; Cotter 1999: 82). Yet

O'Floinn's work appears to give added weight to Raftery's arguments: that an origin for

some of the material that has been dismissed at these sites may indicate earlier occupation

sequences, and that a date within the Later Irish Iron Age seems likely.

Within and between sites, the absence or presence of a single class of brooch continues to

be used to argue against the re-dating of sites, in line with Hencken's original

chronology, especially where earlier material is thought irrelevant to the principal period

of occupation (such as the absence of a brooch of the developed zoomorphic class from

Lagore by Warner 1983). Although hand pins, as previously mentioned are dated to the

6'h or more usually the 7'h century AD in Ireland, a well worn silver hand-pin was

recovered from a 4th century AD context at the Roman mansio site at Tripontium

(Limbert 1996: 264). Hand pins known from northern Britain reflect a similarly defined,

or discrete, localised distribution that appears to link well into the distribution in Ulster,

and it is reasonable to suggest that they reflect a different form of dress within these

areas. At the site of Carraig Aille II, mentioned previously, Ö`Riordäin recovered a bronze and iron hand-pin from the same context as an imitation coin of Constantius (AD

82

324-61). These examples suggest that these pins are likely to have been in circulation

centuries before their traditional dating of the 76 century AD.

Separating contemporary material into distinctive types, either British, Roman, or Germanic, obscures the evidence that in Roman Britain local material continued in use,

even in highly Roman contexts, and that by the 5m century the archaeology of Late

Roman Britain reflects immense influence from continental Europe (Dark 2000). It is

worth remembering also that the demographic make up of the Roman auxiliary force,

included men who originated well beyond the Rhine limes, and who were posted in

Britain along the line of Hadrian's wall (Shotter 2004; Kamm 2004; Jones and Mattingly

1990). Several large, single edged iron knives, known as scaramaxes, were also amongst

the finds at Lagore and these are also of continental type. The reason why it is important

to be sure, or at least explicit, about how the dating schemes are applied across sites, is

that the chronology as devised by Hencken is still used in the overarching works of

synthesis such as that from Edwards (1990 and republished 1996), and used as a principal

reference for Ireland in the Early Historic period. Hencken's conclusions also feed into

other work, which is still used as primary reference material such as Guido's

classification of glass beads from British and Irish sites which is based directly on

Hencken's classification of fords from these sites (Guido 1978). Guido's work is still

referenced by authors such as Johnston on Dun Minne (2007) and Roche on Räith na Rig at Tara (Roche 2002: 97 citing Guido 1978: 40).

In a recent assessment of jewellery in the Romano British period, it has been argued that

it is the function of objects such as brooches, not the decoration, which should be the

primary focus of research (Johns 1996). Any classification or typology needs to consider

the development of the fastener itself, such as the amount of cloth each pin was capable

of holding, in order to inform a discussion on the social considerations of types of

clothing, fabrics and fashions (Johns 1996). The smaller brooches with a straight pin

could hold fine linen or fine wool, but those brooches with a humped profile were designed to gather either thicker cloth such as wool, or to hold together a cloak. By

looking at how the functionality can dictate design and development, greater social

83

insights can be gleaned about what type of clothes were being worn: it follows that brooches which may be different in style may be contemporary but designed to hold

together clothes used for outdoors or indoors.

The plentiful material which has been dated using relative methods, to either the Later

Irish Iron Age, or the ̀ early Christian period', from the eastern and south-eastern

seaboard, contrasts sharply with the relative lack of high status material from the west

and southwest of Ireland. Although the work of Becker et al (2009) adds many ephemeral

sites, some of which show evidence of metalworking, no significant diagnostic material

or metal work has been recovered that might alter the existing distribution. Although

metalwork associated with horse trappings has been found in the northern half of Ireland,

there is little artefact evidence from either southern Leinster or Munster for this period. Given the extent of excavation carried out in Leinster and Munster in recent years, it is

reasonable to suggest that this reflects the true picture (Becker et al 2008).

The weaponry used during the Later Irish Iron Age offers no clear parallels with the rest

of the ̀ Celtic' world. The swords from Ireland are short, light weapons, in contrast to the

long heavy slashing swords more commonly associated with `Celtic' warriors in Europe:

although some can be classified as an Irish type with distinctive curved quillion, others,

notably most of those from of Lagore mentioned earlier resemble the Roman spatha and

gladius with single raised rib and straight quillion (Edwards 1996: 88, Fig. 4.3). A leather

shield found in a bog at Clonoura in County Tipperary which is thought to date to the first centuries AD is a beautifully made wooden object covered in hide, with a central iron boss, and it still bears scars across its front. But it too strays from what would be

thought to be the ̀ Celtic' shape: it is playing card shape (not round) much more in

keeping with a Roman style, but it is very small in comparison to those found elsewhere,

or as depicted on Classical imagery such as Trajan's column in Rome. Although the

warrior society of Cüchulainn is recorded in myth and legend as using chariots, there is

no word in Old Irish for chariot: the nearest we can get to it is carpat, which translates as

a cart (Greene 1970: 62). Recent archaeological finds support the more mundane domestic description, with finds of heavily constructed wooden wheels from bogs on

84

both sides of the River Shannon in the Midland counties. Archaeological reconstruction

suggests that these are from flat-bottom, low-sided vehicles, pulled by two horses: in

keeping with a description of a domestic cart rather than chariot (Greene 1970: 63).

By the end of the I" century AD material of La Töne style all but disappears from the

archaeological record, and in its place we have material of northern British character and

contemporary, early Romano-British, material in the archaeological record. This material,

where recognised, has in the past been classified as ̀ intrusive' (Raftery 1994; O'Brien

1990) and yet given the proximity of Ireland to Rome's northernmost province, its

appearance should hold no surprise. Although in the past it has been catalogued and

classified, little attempt has been made to explain the social implications of its

occurrence, especially at sites that are attested to later as having either ̀ sacred' or `royal'

importance in the annalistic writings.

4.4. Refining the chronology with new absolute dates

In the past, as in modem times, the topography dictated the provincial boundaries and

restricted the extent of the trade and communication routes, through un-fordable rivers

and lakes, swathes of bog land and treacherous mountain passes. Although its floodplain

has altered over time, the immense River Shannon continues to split the country in an arc from the north-west, through the midlands, to the south-west. It has acted as an artery,

which has through the millennia allowed for the movement of material into the midlands

of Ireland. The drumlin belt, alongside the hill-range known as the Sperrins, rings the

northern counties and joins the wide lakes of Upper and Lower Lough Erne to the north-

west, thus separating north from south, geographically and topographically. The high

mountains of the south Tipperary/north Cork area, similarly, separates the modem

province of Leinster from the counties of Munster.

Recent excavations have produced some important findings with regard to unenclosed

settlements across Ireland (Becker et al 2008). It remains uncertain, however, whether

85

this is evidence of episodic use or more permanent activities. Many such sites discussed

within Becker et al 2008 are categorised based on metalworking evidence, such as bowl

furnaces or smithing hearths, but not all appear to have any associated evidence of

permanent occupation (Becker et al 2008: 35). Although furnaces were identified at sites

such as Rossan in County Meath (820-780 cal BC; 370- 50 cal BC), Kinnegad County

Westmeath (800-400 cal BC; 400-340 cal BC), and Reask County Kerry (758-404 cal BC; 404-54 cal BC), it is evident from the range of radiocarbon dates from each site that

multiple periods of metalworking activity have taken place, with the earlier dates

suggesting a transition from the later Bronze Age to Early Iron Age, followed by some later Iron Age activity. Apart from the evidence of metalworking, there are few features

that might be interpreted as evidence for settlement and there is no domestic debris.

There are only three sites in Ireland that have been classified as certain unenclosed Iron

Age settlements through absolute dates derived from closed contexts. At Killoran, in

County Tipperary, an unenclosed roundhouse was excavated, with little artefactual

evidence, none of which was diagnostic, dated to 200-450 cal AD (1890±130) (Gowan

2005: 297; O'Drisceoil 2007). A somewhat unsubstantial circular house or hut with no

occupation evidence, was excavated at Magheraboy in County Sligo and dated to 380-40

cal BC (2140±70) (O'Drisceoil 2007: 24). At Carrickmines Great, County Dublin, a large unenclosed settlement site was excavated in advance of the new road system. Although mentioned by Becker et al (2008), the site was published in full by O'Drisceoil

in 2007, and it is probably the most important Iron Age site found to date.

The Iron Age settlement site at Carrickmines Great contained evidence of multiple

periods of occupation, with stratified contexts dating from the Neolithic and the Bronze

Age as well (O'Drisceoil 2007: 12). The Iron Age evidence was concentrated in the

south-east comer of the site, where two house structures were found along with further

post and pit-holes, evidence of an iron-smelting furnace and cereal processing, suggesting local cultivation of cereals. Radiocarbon dating was used widely across the site, and all

of the dates (shown at Appendix A, Table 1) point to an occupation sequence from 380

cal BC - 70 cal AD (O'Drisceoil 2007). A cremation burial was also found at the site that

86

has returned a date of 340-540 cal AD, the significance of such a late date for the rite of

cremation is discussed in detail in Chapter 6 but it is important to note here that this is

four centuries later in date than expected for the rite of cremation (O'Brien 1990;

McGarry 2007)

There are a few sites that may offer better evidence of settlement, located through their

association with a field monument known as the ̀ fulachta fiadh', the majority of which

appear to date to the later Bronze Age, but given the evidence from Rathtinaun, may fall

within the transition to the Iron Age. The fulacht are horseshoe shaped spreads of burnt or fire cracked stones, within which the remains of wooden troughs are often found.

Although interpretation of their function varies from cooking sites to `steam baths', they

are invariably sited close to a water source or in marshy ground where a central pit would

quickly fill with water (Edwards 2005: 273). Although the Becker report lists sites in

many counties where it is stated that multiple phases of activity have been established,

some of which date to the Iron Age, only two radiocarbon dates were published and these

are from the same site (Becker et al 2008: 39,40). The site is at Parksgove in County

Kilkenny, and it produced dates of 1109-843 cal BC and 856-855 cal BC, which appear

to confirm a later Bronze Age date rather than Iron Age `occupation' (Becker et al 2008:

40). At Colp West, a site near the mouth of the River Boyne in County Meath, a fulacht

on the site returned an Iron Age radiocarbon of 780-390 cal BC, along with a sequence

of dates starting in the Bronze Age through the Iron Age to the 8`h century AD.

There is only one other enclosed site, at Lislackagh in County Mayo, where both the

house structures and a range of domestic evidence suggest a permanent settlement. Radiocarbon dates derived from secure contexts returned a series of Iron Age dates (UB-

3764,2071±81 BP, 200 cal BC - 10 cal AD, UB 3765 1996f8 1 BP, 200 cal BC - 200 cal AD, and UB-3766,2050f49BP, 200 cal BC - 60 cal AD) (O'Drisceoil 2007: 24).

Lislackagh produced a wide material assemblage, from personal items such as decorated

blue glass beads, a polished lignite bracelet, and bronze pins, to domestic objects such as

nails, hooks and knives, along with evidence of smithing at the site (O'Drisceoil 2007).

The site was, however, classified as dating to the Early Historic period, based on the

87

similarity of the morphology of the enclosing ditch to other enclosures dating to the later

period, despite the fact that the absolute dates confirm that this was an iron Age site a

point to which I will return later in the discussion.

The existing classification of hill-forts in Ireland is based on the work of Raftery from

1972, and this remains the most comprehensive analysis of these sites to date (Raftery

1972: 39). Although it is now generally accepted that most of the hill-forts known in

Ireland, as in Britain, have an origin in the Late Bronze Age, few hill-forts have been

fully excavated and even fewer have been securely dated. Of those that have been

excavated, all have evidence of Later Iron Age activity, but whether this is permanent or temporary occupation remains a matter for debate. The hill-forts have a wide distribution

in Ireland and vary greatly in construction, from small univallate sites enclosing only a

couple of hectares, such as Dunbeg in County Down, to those in commanding positions at

precipitous heights, such as Mooghaun in County Clare or Caherconree in County Kerry.

Within the classification are a series of sites known as inland promontory forts and

coastal promontory forts, which again demonstrate huge variability in both position and

construction, to the extent that it is questioned here whether they should be classified as hill-forts (Raftery 1972,1994,2005).

Although it has been argued that multivallation, or multiple embanked construction, is

unusual in Ireland, many of the coastal promontory forts such as Dunmore in County

Kerry, the site of Drumanagh in North County Dublin, and inland promontory forts such

as Carrickdexter in County Meath, were constructed using this form of raised bank and ditch enclosure (Doyle et al 1987; Newman 2005). The construction technique is not limited to hill-forts, however, as it is evident at many ringforts across Ireland (discussed

later). At Garranes in County Cork, Räith na Senad at Tara in County Meath and Rathgurreen in County Galway, excavation has revealed occupation and non-local

material dating between the 4m century AD to the 6'h century AD (O'Riordain 1942;

Raftery 1994,2005; Newman 1998,2005; Grogan 2009). Of the hill-forts excavated in

recent years some appear to have been significantly altered during the late 1 °` or 2°d

century AD during which time the older monuments on the sites were effectively

88

enclosed to form a distinctive ̀ figure of 8' enclosure (Newman 1998: 138). At Din Ailinne in County Kildare, Freestone Hill County Kilkenny, Clogher in County Tyrone

and Räith na Senad on Tara, this rebuilding or re-construction has in most cases been

based on the dating of evidence of early Roman or Romano-British material found either

within the occupation site or within the ditches (Raftery 1969; Raftery 1994: 206; Warner

1995: 29; Wailes 1990: 16). This has led some of the leading authorities on this period to

suggest a link with a homogenous non-local population at these sites (Raftery 1972: 48;

Warner 1995: 29)

One further Iron Age settlement site, at Scrabo in County Down, has been identified

within a hilltop enclosure and radiocarbon dated. Given its positioning and encircling

enclosure it has been classified as a small hill-fort, but it does not fit well within such a

classification. At Scrabo several groups of hut-circles were found, along with associated

pits and hearths and some coarse pottery (O'Kelly 1989: 309; Ö'Drisceoil 2007: 24). The

two radiocarbon dates, taken from charcoal from wall slots around the huts or houses,

overlap: (UB-414E 2305±70) 550-150 cal BC and (UB-414A 1925±100) 250 cal BC to

350 cal AD (O'Drisceoil 2007: 24; O'Kelly 1989: 309; Radiocarbon 1971: 452). In the

general comment section underneath the publication of the radiocarbon dates, the

excavator noted that the occupation at the site should be interpreted as within the Bronze

Age to Early Iron Age tradition, and suggested that the large error and late dates were due

to contamination of soil samples (Owens 1971: 452). It is entirely possible, however, that

the later dates reflect re-occupation at the site, in keeping with datable evidence of similar

reuse at sites, large and small, around Ireland.

Small lacustrine sites, known as crannogs, are dotted over most of the northern and

central parts of Ireland, and these were built either just inside a lake, or adjacent to the

lakeshore. Raised by building on an existing island, or artificially built up with brushwood and soil, a platform was created upon which small palisaded enclosures were

erected. They are consistently referred to within the academic literature as dating to the

7* century AD and later, yet many of those excavated appear to have evidence of

occupation centuries, and in some cases millennia before their ascribed date. Despite the

89

generalized dating of these sites, there is growing evidence of Mesolithic, Neolithic, and Bronze Age activity, such as that recovered at Moynagh Lough, County Meath, Lough

Iron, County Westmeath and Lough Kinale, County Longford (O'Sullivan 1998).

Worked flint is regularly recovered from lower levels at the crannog sites, and in the past, this was ̀ dismissed due to its irrelevance to the later Early Historic occupation phases' (O'Sullivan 1998: 62). At Cullyhanna Lough, County Antrim, excavated by Hodges in

1958, the evidence for woodworking alongside the lack of pottery and ̀ poor quality flints' was interpreted as evidence for an Early Historic date beginning around the 6ý'

century AD. Timbers from this site were dated in 1976, however, and a Bronze Age date

was returned at 1526 cal BC (Hodges 1958; Hillam 1976; O'Sullivan 1998: 73). At the

site of the crannog of Knocknalappa, County Clare, the plentiful Bronze Age ceramics,

metalwork and amber beads found through excavation led the site to be interpretated as a Bronze Age midden, which had been dug to provide a base for the Early Historic

homestead.

A series of radiocarbon dates, however, have now confirmed significant occupation on

the crannog between the Early Bronze Age (1887-1671 cal BC), and a further extensive later Bronze Age occupation (1033-848 cal BC) (O'Sullivan 1998: 87). The site of Rathtinaun in County Sligo was used earlier to demonstrate the Bronze Age to Iron Age

transition in Ireland, and it is one of the few sites which was radiocarbon dated to the 2°d

century BC (the original radiocarbon date result could not be found and so cannot be

calibrated here). O'Sullivan reviewed the site in 1998, and he dismissed the relevance of the absolute dates, suggesting that these were derived from intrusive material at the site,

and then classified it as Early Historic (O'Sullivan 1998: 90). Recent work by a team of

archaeologists excavating the environs of Rathtinaun at Lough Gara on the

Sligo/Roscommon border, have unequivocally dated the occupation of at least two

crannogs here to the later Bronze Age (Fredengren 2003). Radiocarbon and dendrochronology have provided a series of dated contexts for the nearby crannogs on the islands in Lough Gur, and it appears likely that the original radiocarbon dates for

Rathtinanaun are correct and they form part of series of sites that cover the Later Bronze

Age to Iron Age transitional horizon. By pushing the overall dating scheme into the Early

90

Historic period the significance of its development within much earlier archaeological

periods has been lost, as has any study into their importance as centres for craftworking

within these earlier periods.

Consistent with the problems around revisiting the dating schemes for crannogs is

another type of settlement site, the ringfort. The term ringfort, (also known as räths or

raths) is used in Irish archaeology to describe a small ditch and bank enclosure (Edwards

s"`,:: ,.

y

Fig. 4.4 Univallate ringfort at Knowth, County Meath (image courtesy of

Knowth. com: 2009)

1996; Raftery 1994; Stout 1997; O'Sullivan 1998). Our knowledge of the social status

attached to these sites is based on a series of annalistic manuscripts known as the

Seanchas Mär. Within this collection are codices, or law tracts, and some, such as the

Cain Fhuithirbe, have been dated on historical grounds to within a few years of AD 680

(Breathanach 1986). These texts provide a reference to the social status and obligations of

clerics, nobles, and their dependents at the time at which they were written, and one in

particular, the Crith Gablach, classifies and identifies the legal, social and economic

status of the occupants of each type of ringfort (Charles-Edwards 1986 & 2000). The

most common ringfort is the simple univallate type (Fig. 4.4), and the more complex

91

s V

multivallated types, or those with raised interior platforms, are traditionally interpreted as

the homesteads of the elite. These enclosures vary greatly in size and morphology and

they also appear to have varied in function, with evidence of homesteads, livestock

enclosures and in some cases, such as at Garranes, craftworking centres. The term ̀ fort'

probably relates to a later recognition of the use of ditches and banks to protect the site,

although these too vary from site to site, as most are not of defensive construction, but

merely delineate the extent of the enclosed space. Other terms commonly occurring in the

texts are lios, cathair or caher, and cashel all of which are used to identify similar

enclosures. The terms all relate to the type of enclosures and reflect the construction in a descriptive sense; lios usually relates to the interior of the enclosure eg. Lismore (lios

mdr = big lios) and Lisbeg (lios beag = small lios) are two of the most commonly

occurring townland names across eastern Ireland and the midland counties. The term

`rath' or raith, tends to refer to the exterior of the enclosure as one of earthen

construction, and cathair, caiseal, or cashel seems to imply the use of stone in its

construction (although these often appear to have been constructed with both earth and

stone) (O' Kelly 1970: 51). Cashel may be of later origin as it appears to be a term taken

from the Latin `castellum' and the question of its origin is much debated (Raftery 1994;

O'Kelly 1970; Charles-Edwards 2000).

The classification, dating and terminology for ringforts across Ireland has been the

subject of considerable debate since the 1970s, and the present dating scheme is greatly disputed (0' Kelly 1970; Ö'Riordäin & de Valera 1978; Raftery J 1972; Edwards 1990;

Stout 1997). Many scholars continue to argue valiantly for their continued classification

as ̀ Early Christian' (Edwards 1990; Ö'Cröinin 1995; Stout 1997; Kerr 2007) despite the

fact that the most traditionalist scholars such as Raftery (1972), Ö'Rlordäin and de Valera

(1978) and Caulfield (1994) suggested many years ago that the concept of enclosed homesteads, lay in the Iron Age, or perhaps even as early as the Later Bronze Age. The

existing dating scheme is based on the radiocarbon dates of around 150 ringforts, many

of which were multiple dates from the same sites, based on excavations in the northern

counties (O'Kelly 1970: 50; Limbert 1996; Stout 1997: 27: Kerr 2007). There are, however, over 50,000 ringforts known throughout Ireland and in recent work some

92

authors have questioned the validity of a dating sequence based on a very small sample from within a restricted geographical area (Limbert 1996; Mattingly 2006: 447). Just as many may have a principal period of occupation with the Early Historic period, many

more are likely to have an origin in the Later Iron Age.

The chronology of the ringfort was recently reviewed and in a wide-ranging paper both

the orthodox dating sequence and the existing typology for the ̀ diagnostic' material, or lack of it, used for dating of these sites was successfully challenged (Limbert 1996). It

was argued that the few radiocarbon dates available for the south had created a sample

which was regionally biased towards sites from the northern counties and it was wholly inadequate for sites in the south of Ireland. Questioning the continued use of the existing dating schemes based on the typology of metalwork such as the decorated hand-pins, or lack of them, it was argued that these objects were consistently benchmarked against the

relative dating derived from older excavations, which was itself questionable. The

precedent for building `in the round' must be sought within a prehistoric context, and at least a significant number of ringforts are likely to have been in use in the Later Iron Age

(Limbert 1996: 285). Although Limbert's work is convincing, it has not been widely

accepted. Furthermore, a recent re-evaluation of the artefacts found during an excavation in the 1940s of the ringfort at Rathgureen in County Galway has produced a previously

unrecognized terracotta oil lamp, which has now been classified as a Roman import dated

to the early 4t' century AD (Comber 2002).

The origin and morphology of the ringforts in Ireland was recently reassessed through a

statistical analysis of the topographical relationships between the homesteads and the

agricultural potential of the land within which they are found (Kerr 2007). Placing

primary assessment on the data available from the north-east counties in Ulster, the sites

were classified as ̀ Early Christian raths', and as explained we cannot know if their

occupants were Christian, and the term rath is a quite specific descriptive term for types

of ringforts in the south of Ireland (Kerr 2007; O'Kelly 1970). Despite this, there are two important findings within this analysis and I want to discuss these here. The first is that it has been demonstrated that although ringforts appear morphologically similar, and have

93

up to now been treated as a discrete group of monuments, there is immense variation

within the radiocarbon dates that have been collated for them. Despite the fact that only three or four new radiocarbon dates have been added for the southern Irish counties,

when the data is modelled on the radiocarbon curves, it becomes apparent that some

univallate types are signficantly earlier in the north; furthermore, in the south these early dates included some of multivallate type (Kerr 2007: 98-99). Through the assessement of

the animal bones assemblage at these ringfort sites we can now say with some certainty

that these are predominantly of cattle, sheep, and pig in this order of relative frequency.

This prevalence has been confirmed in more recent work published by McCormick

(2007) based on additional data provided by his assessment of published and

unpublished accounts of faunal remains from older excavations throughout Ireland, and

which included crannog sites. There is still a significant bias within the data towards

ringfort sites in the northern counties of Ireland and although the analysis referenced

some data from sites in the south, we cannot securely base conclusions on an Ireland-

wide basis without signficantly more data. One interesting outcome of his analysis of the

northern ringforts is that the type known as platform ringforts, which have a raised internal area, appear to date significantly later than those of the univallate or multivallate

type. This would need further work in the rest of Ireland in order to establish a

chronology but it is interesting that the pattern appears to be confirmed by the few that

were included for the south (see Table 4.1, and Fig. 4.5 below). There are now

radiocarbon dates from both crannogs and ringforts that directly challenge the continued

classification of these as exclusively Early Historic in date (see Fig 4.5). I recognise that

for the majority of crannogs the principal period of occupation may indeed fall within the

6m century or later, but we need to move forward and systematically collate the

information from sites where an occupation in the earlier periods is suspected or known.

For instance, despite the evidence of immense Iron Age occupation at Lislackagh, as discussed earlier, demonstrated through a series of Iron Age radiocarbon dates, it is still

classified as an Early Historic ditched enclosure (see O'Drisceoil 2007 for a full

discussion). It has become increasingly evident that Limbert was correct and many of these sites need to be considered as transitional sites within the Later Irish Iron Age.

94

Site sad

Townlaad

Lab Ref Date 14C calibrated Type References

Aughadegnan

Longford

Ub3454 1551±33 425-580 cal AD univallate Carroll 1993; Kerr 2007

Ub3459 1552±49 410-606 cal AD

Ub3461 1597±71 259-607 cal AD

Balrothery,

Dublin

Ub4807 1689*15 261-411 cal AD multivallate Carroll unpub, Ken: 2007: 89

Ub4808 1693±22 259-411 cal AD

Ub4809 1639±21 342-532 cal AD

Conva, Cork Ub3635 1389t37 584-687 cal AD Univallate Doody 1995; Kerr 2007

Ub3638 1525143 426.619 cal AD

Clogher, Co.

Tyrone

Ub2243 R 1795±120 41 cal BC - 534 cal AD `cashel' Warner, Mallory & Bailie (1990)

Narreghmore, Co.

Kildare

P2413 R 1940+60 88 cal BC - 230 cal AD `ringfort' Warner, Mallory & Bailie (1990)

Rabeenamadra,

Co. Limerick

U 248 R 1840+110 92 cal BC - 428 cal AD Univallate Stenberger 1966; Kerr 2007;

Warner, Mallory & Bailie (1990)

U241 R 1430±130 268-890 cal AD Univallate

Lisdoo,

Fermanagh

Ub2202 1655+45 258-534 cal AD Bivallate Brannon 1982; Kerr 2007

Millockstown

County Louth

(not

published)

1595+70 259-608 cal AD Univallate Manning 1986; Kerr 2007

Naeaghmore,

County Kildare

P2412 1580±170 66 cal BC - 775 cal AD Multivallate Fanning 1972; Kerr 2007

P2413 1945±60 90 cal BC - 226 cal AD

Raffen, County

Meath

Ub3435 1676±168 37 cal BC - 661 cal AD Univallate Newman 1997

Ub3437 1506185 355-670 cal AD

Rathgureen,

County Galway

Ub4331 1925130 2 cal BC - 134 cal AD Multivallate Comber 2002

Tuam County

Galway

Ub4324 1410±65 443-771 cal AD Crumlish 1998; Kerr 2007

Table 4.1. Early radiocarbon dates from ringforts in southern Ireland. Based on

radiocarbon dates published in Kerr 2007, Becker et al 2008, O'Sullivan 2007, Warner,

Mallory & Bailie 1990.

95

R_DaW

R_Dale Narr plnwre

R-Dab Nwn gim-Nm ---A

R-Dab

R_Dale

R-Dab RON _

R-Dale T -

R_Data T

R_Dab

R-Dab Solm hey

R_Dats

R Dai

R_DTh Cam a

R-Dab

R-Dab

FLDab

R_Dri Phita a

R -Dab

Ueda o

R-DaM

R Date

I Al 12u u 147Rd -

UKJ 1IIWI

Calbt d dw (CmMtmA }

Dam AuglmidW an

lcaSCrIcWAD t 401 t 001

Calbrid do* (a19CAML&A

Fig. 4.5 Published radiocarbon dates modelled using OxCal 4.1, profiles showing

periods of activity at ringforts some predating AD 600. Based on radiocarbon dates

published in Kerr 2007, Becker et al 2008, O'Sullivan 2007, Warner, Mallory & Bailie

1990.

96

4.5 Circularity in interpretation

I mentioned in the introduction to this work two new projects funded by University

College Dublin and I want to turn now to discuss the recently published Iron Age project `Finding an Invisible People' authored by Becker et al (2008). Although the stated aim of the project was to assess cultural change within the Iron Age in Ireland, there is no discussion of this anywhere within the published document. The Iron Age in Ireland is

classified as aceramic and yet there are references throughout to sites which have been

radiocarbon dated to within the Iron Age, but which also appear to have evidence of

ceramics (Becker et al 2008: 54). There is no classification or assessment offered for

these finds and yet any or all ceramic evidence is crucial if we are to realistically reassess the archaeology for the period. There is also a consistent use of inferential statements

regarding areas which have not been excavated, and the assumed likelihood that on

excavation these will result in similar data to that from elsewhere (Becker et al 2008: 24).

New radiocarbon dates are dotted throughout the report, but they were not brought

together in one table, they are not consistently cited, and no laboratory references were

given. The result is that many of the dates cannot be calibrated to check their validity and

can only be used by others in a similarly inconsistent way. Many of the sites and their

material assemblages were already widely published before the project's inception, but

the distribution analyses are based on an outdated map produced by Raftery in 1984.

Using older distribution maps creates the impression that new sites have altered the distribution significantly, but in reality there are very few finds of diagnostic material or

actual evidence of settlement occupation in the south. Although transition into the Iron

Age was considered in in the report, there was no corresponding work completed for the

transition to the Early Historic period. Perhaps more worryingly no consideration was

given to the existing assemblage of material dating to the Later Iron Age that has been

classified as ̀ intrusive'. Here is an entire body of data which has been excluded from any

consideration within the report on what I can only assume is the basis that its authors thought it was irrelevant to our understanding of `Irish' Iron Age.

97

In order to create a more robust chronology for the Iron Age we need to separate that

which we know, either through absolute or reliable but still relative dating, from that

which is open to further scrutiny.

4.6 A classificatory system for the Iron Age

The classification scheme proposed here does not preclude the use of any, or all, of the

material and the sites detailed within the tables under each heading. The rationale behind

the use of such a scheme is that it will identify that which is securely dated from that for

which there are still questions, and it will highlight the potential for reassessement or

reinterpretation. It is also explicit in its recognition that some ringforts and crannogs have

much earlier sequences of occupation that need to be considered relevant to the Iron Age

as well as the Early Historic period. The following classification series should be added

as a search function to the two databases now on-line.

IA1 - Material, or sites, securely dated through scientific dating (RC, AMS or

dendrochronology): Appendix A: Table 1

Appendix A, Table 1, is based on the work produced by Warner, Mallory and Bailie

(1990), Raftery (1994), O'Sullivan et al (2008), Becker et al (2009) and lists all those

sites which are known to have absolute dates, and that can be used as anchor points

within the longer Iron Age.

IA2 - Material found in secure contexts or dated through reliable diagnostic parallels

elsewhere (ceramics and metalwork).

All finds for which we can offer secure dating based on diagnostic material which can be

closely paralleled.

98

IA3 - Material or sites for which only relative dating has been proposed and which require reassessment in light of new dates elsewhere

The new '4C dates for Räith na Senad on Tara confirm that some of the early form

penannular brooches were in use between the 2nd and 4th century AD and the parallels

offered for these were with those from Cahercummaun and Ballinderry No. 1. This

confirms that the relative dating at both of these important sites needs to be revised to include an earlier occupation sequence. A new classification series and typology needs to be compiled based on the work of Newman (1989) and O'Floinn (1999), which would involve a systematic review of all of the material presently housed in the National

Museum of Ireland, and collections elsewhere, although this lies beyond the scope of the

present work.

IA4 - Material from sites where the stratigraphical relationship between contexts

remains unclear, even where some have been securely dates, under either L41 or IA2 classification: Appendix A: Table 2

Appendix A, Table 3, details sites where imported material was found during excavation but treated as irrelevant to the dating sequences. The stratigraphy at Lagore crannog was

greatly compromised and despite the likelihood that we can recreate a reasonable interpretation based on Lynn's work from 1995, whilst the stratigraphical contexts of the finds remain uncertain and in the absence of absolute dates, it needs to remain as IA4.

The Iron Age itself should be reclassified into three discrete temporal periods:

70OBC to 300 BC = Early Iron Age

30OBC to AD100 = Middle Iron Age

IOOAD to AD500 = Later Iron Age

The first period, the Early Iron Age, includes the evidence presented in this chapter that

suggests a gradual transition out of the later Bronze Age Dowris Phase, into what is

99

known elsewhere as the Early Iron Age. This classification respects the fact that there is

still much to be learned about the regional distribution of the take up of even iron

working technology and that for some parts of southern Ireland, the Bronze Age tradition

continued into later centuries. This variability in response to new technology shows a

marked difference between the lack of sites in Munster and the relatively high quality

material found in sites in the northern counties.

The second period, the Middle Iron Age, is dated in accordance with the relative dating

used for material of La Tene style but reflects the fact that much of the later material may betray early Roman or British influences (Harding 2004).

The later Irish Iron Age has been dated to bring it broadly into line with the Roman

control of north-west Europe, as will be discussed more fully within the next chapter. Pulling the end date back to the start of the 6th century allows a more realistic

consideration of the growing body of evidence of later imported ceramics from the

Mediterranean. The pattern of distribution of the later ceramics, suggests the continuation

of nodal entry points used during the entire Later Irish Iron Age.

A prefix before the series will identify the current temporal period into which the material

or site is believed to fall (IA = Iron Age; EH = Early Historic). A marker is needed within the classification series that identifies sites where there is clear evidence that they cannot be regarded as exclusively either Later Bronze Age, Iron Age or Early Historic, and this

needs to be transparent within the archaeological overviews. The term ̀ multi-period', in

itself, does not indicate which periods a site moves through in terms of the agreed

chronology. It is suggested here that the series be extended to include the use of the term

`transitional' much more frequently within assessements, alongside the explicit use of the

overall dating sequences established at each site. This will enable future researchers to identify much more easily those sites that can be accurately dated to within each period

of the Iron Age, and those where questions remain about generalized dating based on

older chronologies and for which further analysis is needed. All of the radiocarbon dates

from sites around Ireland need to be collated and published within a searchable database.

100

A standardised or at least a consistent approach is needed to how the data is presented but

in its most basic form it needs to include all of the following: site name, townland name (and barony name if different), the laboratory reference number, the uncalibrated date,

the calibrated date and all the references used, including references to unpublished

reports from contracted archaeology consultancy firms or individual archaeologists.

What becomes very clear once the data is restructured is that many of the sites where

there is material that has clear parallels outside Ireland fall within the IA4 classification. Amongst the many sites are those where Roman, British, Romano-British, Late and post- Roman, Germanic or Saxon material was revealed during excavation, and it is these

material imports which are discussed in detail in the following chapter.

101

Chapter 5 Imported Material in the Later Iron Age

5.1 Reclassifying Roman material from Irish sites

In the past the lack of a clear chronology for the later Irish Iron Age, as discussed in the

previous chapter, rendered many of the existing interpretations as loosely based around

relative dating methods. We can now offer a more realistic interpretation by including all the material that dates to the first five centuries AD as relevant to this discussion along

with new finds and new absolute dates that have not yet been incorporated within

overarching works of synthesis. In the past, material evidence that did not support the historical dates given in the Annals, especially the evidence for British, Roman, Romano-

British and Roman derived objects, was consistently rejected as ̀ residual' or in too many

cases simply `long-lived' (Hencken 1942: 56 & 57; Warner 1976: 267; O'Sullivan 1998:

140). Not every object found will be discussed in detail (an updated catalogue is at Appendix B, Table 1) instead discussion will focus on those objects or sites where new information or new interpretation of the finds can be offered. The hypothesis of direct

Roman military engagement in Ireland in the I" century AD, as discussed in Chapter 2,

will be reconsidered once the material has been reassessed here. The two existing distribution maps of accepted Roman finds are shown at Fig. 5.1 below.

102

U_c

i. 'i

x. ft's 5"""

aMY

i

v jj

Fig. 5.1. Left, distribution map of accepted finds ftom Bateson 1973 and at right,

distribution map of accepted finds fr om Raftery 1994

The most important qualification that needs to be restated at the outset of this analysis is

that all of the imported material should be considered as part of the materiality of the

Later Irish Iron Age. Initially, for the purposes of transparency within this review, the

material is classified together under imported items, but utilising the same classification

system as outlined in the previous chapter (with the addition of coinage as a further

dating indicator).

IAI - Material, or sites, securely dated through scientific dating (RC, AMS or dendrochronology)

Radiocarbon dates have now demonstrated that the material in the Rose/Mauve phase at Dün Aillinne dates to the 1 s' century AD. It is best paralleled in the LPRIA of northern Britain and it should not be classified as exclusively Roman (Harding 2004,2007). The

material found within the grave contexts at Betaghstown should not be regarded as

103

exclusively Roman either, as although the earliest radiocarbon date lies within the 4th

century AD, many fall within a range of 5' to 6'h century AD and parallels for some of the material can be found in early Saxon contexts, both in Britain and the continental Europe (Owen-Crocker 2004). The parallels for all of the artefacts found within burial

contexts at Betaghstown are from outside Ireland (these will be discussed in detail in

relation to the burial evidence in the following chapters). Excavations at Rathgall hill-

fort, led to the discovery of aP century AD decorated Roman strap tag (Raftery 1969:

98,1970). This object was the subject of a paper in which the decorative style of a triskele, lyre and triple circle motifs were paralleled with similar objects in the Roman

world and the form and function were classified as part of a belt of cingulum type, an

essential part of a legionary soldier's uniform (Raftery 1970: 2(4). The date is consistent

with the absolute dates from the site (SI-14801685±70), 144-541 cal AD, taken from

charcoal found within a smelting pit in the occupation area (Raftery 1994: 230).

Directly across the river from the Great Mound at Knowth is the site of Rossnaree where burials were found in the 1940s. A small Roman-style earring was found with one of the

burials: it is intricately incised with a herringbone decoration and has evidence of tinning,

a technique which is not known on silver within the Irish Iron Age. The burial has been

dated to 250-520 cal AD (discussed in chapters 6 and 7; Sikora 2007: unpublished

report). A series of radiocarbon dates have now confirmed that the principal period of

activity at Rdith na Senad, on Tara, is between the 2d and 4`h century AD which suggests that the Roman material is contemporary. Although there remains a problem over the

extent to which the stratigraphy was compromised due to the illegal digging (as explained in Chapter 7) the overall dating sequences were derived from samples carefully excavated in the 1950s and we can now accept these as reliable. Although the Roman material from

Newgrange has been classified based on the coins found, there is also an Iron Age

radiocarbon date from a sod layer within the mound (UB-360 1650±45), 259-536 cal AD

(Radiocarbon: 1971) that confirms contemporary activity at the site. The Roman material found at Clogher hill-fort in County Tyrone is confirmed as contemporary with the earlier

phases of settlement and a range of dates from occupation within the ringfort confirm dating to the Later Iron Age (see Appendix A, Table 1)

104

IA2 - Material found in either secure contexts or dated through reliable diagnostic

parallels elsewhere (ceramics, metalwork or coins)

Most of the imported material falls within this classification (see Appendix B, Table 1)

and for ease of reference it has been divided into group headings.

Coins and Coin Hoards

Bateson's catalogue listed 86 sites where either single or multiple coins have been found

in Ireland, and of these he accepted only 13 as genuine (all 86 sites are included with Appendix B, Table 1). His justification for this was due to the lack of information on the

circumstances of their recovery, or the lack of context detail. Yet as mentioned earlier he

was inconsistent in his approach, as he accepted high denomination coins but not the

lesser value copper coins. One important coin, rejected by Bateson, was found in County

Kildare and it is a denarius of the Roman usurper Carausius (287-293 AD), but it has

been lost and no further details are known. Copper coins of Trajan (97-117) and Hadrian

(117-138) were found with extended inhumations at Bray Head in Wicklow in the 19t

century. The burials appear to be in correct Roman style and were protected within stone lined graves; these will be discussed in detail with the ̀ intrusive' burials in the following

chapter. A coin of the Younger Faustina (161-175) was picked up as a single find from

the Navan area, close to Newgrange and Knowth in County Meath (Bateson 1973: 53). A

4m century AD coin was found alongside a large Roman type key at the site of Uisneach

in County Westmeath and a radiate coin of Constantius II (324-361) was amongst the finds from the excavations at Carraig Aille in County Limerick. Four Ist century AD

Roman coins have been found at different sites around the Dublin area and there are four

4'h century AD Roman coins from the same geographical area. The gold solidus of Valens (364-378 AD) was found at Templeogue in County Dublin, and a further solidus

of Valentinian (364-375 AD) was found at Ballintoy in County Antrim (Bateson 1973:

43).

105

There are four silver hoards from Ireland that were found in the 19th and 20th centuries

and a further two hoards of mixed denomination coins. Two of the silver hoards date to

the 2d century AD, and these were found at Flower Hill and Feigh Mountain on the north Antrim coast. The find at Feigh Mountain consisted of some 500 silver coins, in a poor

state of preservation, but they appear to have run from Vespasian (AD 69-79) to

Commodus (AD 176-192). The hoard at Flower Hill was found close to a standing stone

and comprised of some 300 silver coins. Little detail of the coin issues is known, but one

was attributed to Matidia, the niece of Trajan (AD 97-117) (Bateson 1973: 45). Hoards of

similar date and denomination have been found in Scotland, such as those found at Birnie, and these have been interpreted as bribes or payments (Hunter 2000). The

similarity suggests that the Irish hoards should be considered in the same way. The

Roman authorities are known to have secured a temporary respite from otherwise hostile

tribes on the northern frontier of Britain through large payments in silver, and it is

possible that this was extended to perceived threats from their kin in Northern Ireland

(Kamm 2004; Shotter 2004).

At Newgrange passage tomb in the Boyne Valley, the earliest coins found can be dated to

Domitian. Along with two gold aurei were six solidi and two solidi uniface pendants. These solidi pendants are coins that have had a hole punched above the obverse imperial

image through which a gold suspension loop was fastened. The parallel for Roman gold

coins used as in this way, as jewellery pendants, is within the Germanic kingdoms, where the type of suspension loop on E56.476, moulded with striations, is also common

amongst the later bracteates (which were based on these earlier types). It is probably also

notable that they were both minted in Trier (see below: Carson and O'Kelly 1977).

106

Emperor Date of Reign

Type Mint and probable date

Inscription

Legible or illegible

Museum

No.

Domitian AD 81-96 As Rome c AD 81 Illegible E56.995

As Rome c AD85 Legible E56.978

Denarius Rome c92-93 Legible E56.561

Geta AD 198-212 Denarius Not recorded Not recorded

Gallienus AD 253-268 Antoninianus Not recorded Not recorded

Postumus AD 260-268 Antoninianus Cologne AD 260 Legible E56.961

Probus AD 276-282 Antoninianus Rome AD 276-

282

Legible E56.560

Maximian AD 286-305 Aureus Trier AD 295-305 Legible E56.1736

Aureus Trier AD 295-305 Legible E56.577

Constantine I AD 308-337 Gold uniface pendent Trier AD 330-337 Legible E56.657

Follis London AD 310 Legible E56.558

Follis London AD 310 Legible E56.650

Solidus Trier AD 335-6 Legible E56.649

Constantine H AD 317-340 Solidus Trier AD 322-3 Legible E56.994

Gold uniface pendent Trier AD 320-30 Legible E56.476

Magnentius AD 350-353 Maiorina (bronze) Amiens 350-53 Legible E56.1735

Maiorina (bronze) Trier AD 350-353 Legible E56.1735

Maiorina (bronze) Trier AD 352-353 Legible E56.960

Valentinian I AD 304-375 Solidus Trier AD 367-375 Legible None

Gratian AD 367-383 Solidus Trier AD 367-375 Legible E56.647

Theodosius I AD 379-395 Solidus Trier AD 378-395 Legible None

House of

Valentinian or

Theodosius

- Siliqua (Clipped) Uncertain but

clipping of siliqua

is more common

in R -B than other

province

Legible E56.658

Arcadius AD 383-408 Solidus Milan Legible E56.963

Uncertain Two small brass - types unrecorded

Table 5.1. Record of coins finds from Newgrange (Carson and O'Kelly 1977)

107

The two later silver hoards from Ireland were found at Balline in County Limerick in the

south west of Ireland, and the more famous Coleraine or Ballinrees Hoard, (Ballinrees is

the townland in which it was found, Coleraine the nearest town), again from the northern

coast.

The Ballinrees hoard (Fig. 5.2)

contained 1,506 siliquae that ranged

in date from Constantine II (324-

361) to the volatile reign of Honorius

(395-423), along with hack silver

and silver ingots. One of the ingots

was stamped EX OFF CVRMISSI

and it has been paralleled with a

similarly stamped ingot from Kent.

The stamped ingots, even those from

Ireland, have been classified as

imperial donatives to troops serving

in the early 5`h century AD (Painter 1972; Raftery 1994: 216). The coins from Ballinrees

were clipped and, although clipping is known to have been used throughout antiquity to

test for the quality of the coinage, it is especially common on silver coin finds from

Roman Britain, and it may suggest that the ingots and the coins originated here, rather

than the continent (Fulford 2005: pers. comm). The hoard from Balline did not contain

any coins, but a collection of silver ingots and several pieces of silver dishes. The dishes

originated in the Mediterranean world, and the small ox-hide-shaped ingots were stamped

with the originators marks; one marked ISATIS can be paralleled with a similarly

stamped ingot from the Roman fort at Richborough (Painter 1972). Several of the ingots,

from both hoards, had additional pieces soldered to them in silver in all probability to

ensure that the total weight made a standard Roman pound.

108

Fig. 5.2. The Ballinrees Hoard (image courtesy of the British Museum)

A further hoard was found at Cushkinny, in the area of Cork Harbour in 1878. The coins

range in date from Claudius Gothicus (268-269) to Constantine II (317-340), and as the hoard covered two periods of new coin issues Bateson rejected it (Bateson 1973). I would

argue that we should use the latest coin as a terminus ante quem for the hoard as a date of the 4th century AD fits well with the predominately 4th and 5t' century AD material

clustering in the area of County Cork.

In work just published the evidence for a further coin hoard from Räith na Senad on the

Hill of Tara was reconsidered. The hoard was found during a controversial excavation

that took place at the site at the turn of the 20th century, discussed in detail in the

following chapter (Bateson 1973; Carew 2003; Grogan 2008: 119). The hoard consisted

of 15 or so ̀ copper' coins of Constantine I (AD 306-337), and it was found buried within

a Romano-British ceramic container. Bateson dismissed this hoard as a hoax and he

stated had it had been deliberately planted for those taking part in the dig to find. After a detailed analysis of the location of the hoard, the recorded depth at which it was found,

and the other Roman material that came from that area of the site, no evidence was found

that would suggest the hoard was planted and it has been rehabilitated as contemporary

find with the rest of the Roman material from Tara (Grogan 2008: 119-120).

Oil Lamps

One small Roman ceramic lamp was recently recognised in a reassessment of old finds

from an elaborate, bivallate ring-fort, at Rathgurreen in County Galway. This has been

dated to the mid 4th century AD, and is a Roman import (Comber 2002). A similar lamp

has been recorded amongst the ceramic evidence from Räith na Senad (Grogan 2008). A

further example of a terracotta oil lamp was found at Drumcliffe in County Sligo in the

19th century (Bateson 1973: 76).

109

Toilet Implements

Small bronze Roman toilet implements were found in association with extended

inhumations at the coastal sites of Dooey and Ballyness, both in County Donegal. At

Ballyness a Roman-style tooth pick was also found and can be directly paralleled with

some in the Hoxne treasure hoard, dating to the 4`h century AD. Other toilet implements

have been found at Knowth, Freestone Hill and Clogher hill-fort. Bateson's descriptor of

these as a two-pronged instrument can now be replaced by Eckardt and Crummy term

`bifurcate toilet implements' (Eckardt and Crummy 2006). The object from Clogher was

found within a dated context of the 4`h century AD. It is a well-made copper alloy

example (viewed in 2005 courtesy of Richard Warner), formed with careful

ballustrading, and an incised dot and circle decoration, and it can be directly paralleled in

3`d or 4`h century contexts in Roman Britain (Eckardt and Crummy 2006). The toilet

implements from Freestone Hill are highly decorated and dated by the 4 ̀h century coin

found within the same context (Raftery 1994,2005; O'Floinn 2000: 20). At Knowth three

toilet implements were found and have been classified as a tinned and gilded ligula

(4031), a tinned ligula, or possibly stylus (4032), and an `ear-scoop' (4033) (Mullarky

forthcoming; Bateson 1973). The interpretation of one of these objects as a stylus is

interesting, as others are known from the sites of Garryduff and Lagore Crannog.

Jewellery

There are eight Roman style fibulae

of the I to 3d century AD that

have been found as single finds

around Ireland (Bateson 1973;

Raftery 1994). Apart from an

unusual Pannonian style brooch

from Galway the majority are

concentrated in the northern counties and in

r/// r INNW-

? ýý.. ý "1

Fig. 5.3. Toilet implements and jewelleryfrom the area around County Cork and most fit Freestone Hill, not to scale (after Raftery 1994)

110

well into the classification of Roman types, common in Roman Britain at the same

period. AI S` century Roman fibula and a piece of Roman-style bracelet were found

during the excavation of a well at Randalstown, in County Meath, along with sherds of Samian ware (Kelly 2002: 25).

Although Roman-style melon and glass beads

have been found with some frequency at

sites around Ireland, there are several sites where

multiple pieces of jewellery have been found.

These were not within burial contexts and no burials were recovered from the sites, which has

led to an interpretation of likely religious deposition

(O'Floinn 2000). At Freestone Hill several bronze

bracelets, some plain but some with incised herring

bone decoration (shown at Fig. 5.3), were found within

the same context as a 4`h century coin, the toilet

implements mentioned above, and the two sherds

discussed later under ceramics (Raftery 1994,2005;

O'Floinn 2000: 20).

The most important collection of Roman jewellery ever

found in Ireland comes from around the passage tomb of Newgrange. The Roman material from Newgrange is

remarkable for its high quality gold and silver

jewellery and coins and that it also offers an

interesting insight into religious practices in Ireland

from the date of the first coins in the 15` century AD to

Fig. 5.4 Plate brooch from Newgrange with incised decoration and conical glass insert (after Carson and O'Kelly 1977)

Fig. 5.5. Plate brooch from Uley with incised decoration and conical glass inset (after Woodward & Leach 1993)

as late as the turn of 5`h century AD. Direct parallels can be made between the gold and

conical glass plate brooches (see Fig. 5.4 and 5.5) found at Newgrange and Uley in

Gloucestershire, and identical little brooches that have been found at Romano-British

111

Il cm=

I lcm

temple sites in Britain such as at Nettleton, Bath and Wroxeter (Woodward & Leach

1993; Aitchison 1988).

One very interesting find from the area in front of the megaliths at Newgrange, is a

terminal of a Bronze Age gold torc (clearly predating its re-use), inscribed in pontille

(Fig. 5.6) (Raftery 1994: 211, Bateson 1973: 47). Although scholars have struggled to

translate the inscription which reads SCBNS. MB, the name Scribonius, or its female

equivalent 5cribonia are commonplace in the Roman world, and a similarly inscribed

torc end is known from a burial with a cinerary urn found close to Birdoswald Roman

fort (Carson and O'Kelly 1977). The dedication MB is like to be to Mars Belenos or

Belatucadrias, both commonly abbreviated as such in Roman contexts elsewhere, or an

alternative reading may suggest a dedication to the goddess of the river Boyne, Mater

Buvinda (Kurzmann, forthcoming). The reading of the inscription itself is in many

respects less important than the fact that this small inscription supports the possibility of

literacy in Ireland at this time, a point to which I will return on my discussion of literacy

in Chapter 8.

I Icm

Fig. 5.6 The inscribed gold torc end ftom Newgrange (after Freeman 2001)

112

One other object from Knowth that should be considered amongst the 2d century AD

material is an early Romano-British ̀ proto-zoomorphic' penannular brooch, similar to

one found at Tara (Kilbridge-Jones 1980; Eogan 1977; Mulvin forthcoming). Given the

proximity of the two passage tombs at Knowth and Newgrange, it is remarkable that

there does not seem to be any overlap between the types of material at each site. No gold

or silver jewellery, or Roman coins, were uncovered at Knowth and it has been argued in

the past that the assemblage is domestic in character, albeit high status given the Samian

ware, and that it might be the settlement for those who were worshipping at Newgrange

(Raftery 1994). One other object from Knowth included by Mulvin, which was not included in Bateson's catalogue in 1973, was a large leaded bronze spiral ring, which is

tapered into a serpentine head, and she has paralleled it with similar rings from Garranes

and Dalkey, both sites with much imported material (Mulvin, forthcoming).

Ceramics

A small rim sherd of a Romano-British copy of Samian form Dr 37 was found at Finaghy, a large piece of Samian ware (undecorated) was found at Dundrum and at Sydenham Avenue in Belfast, two large pieces of a Dr 18/31 vessel and a basal piece of a Dr 18 were also found (these were viewed by the author in 2006 by kind permission of Richard Warner). Sherds of Samian ware were found with Roman jewellery at St. Anne's

`holy' well in Randalstown, not far from Tara in County Meath. Two sherds, one of

Severn Valley ware and one described as ̀ East Gaulish' were found at Drogheda in

County Louth, the point at which the River Boyne enters the Irish Sea (Kelly 2002: 25).

Saurian ware sherds have been recovered as accidental finds from the promontory fort of Drumanagh in County Dublin, and sherds of Dr 30 and Dr 37 along with Roman glass

were excavated at the site of Dalkey Island, just off the Dublin coast. A large storage

vessel, described as an olla, was found off the coast of County Sligo in the early 20th

century. It is marked with a graffito in Roman cursive and is now in the National

Museum of Wales (Ö'Riordäin 1947: 65; Bateson 1973: 77). A section of Roman brick

113

or tile was found at Baronscourt in County Tyrone but its location was unknown to

Bateson at the time of his review (1973: 64).

Two sherds of imported pottery were found at Freestone Hill in the same context as the

Roman jewellery, and these were described as of `hard wheel made pottery', and

classified by Bateson as `sub-Roman' although he, like Raftery and O'Floinn, could offer

no parallel amongst known imports from sub-Roman Britain (Bateson 1973; Raftery

1994; O'Floinn 2000: 22). These two sherds were viewed by the author as part of this

research, at the NMI in 2008, and that they proved difficult to classify is understandable

(shown at Fig. 5.7).

=- MIM

rwr a

Fig. 5.7. The two imported sherds from Freestone Hill, County Kilkenny.

Images by author, reproduced with permission of the NMI

One is a small fragment of the lip of a wheel-thrown orange ceramic with a grey core,

probably a small closed form of Severn Valley ware and a similar example, found at Räith na Senad, was recently classified as Severn Valley ware (Evans 2008). The second

114

piece is a brown upper body sherd of creamy white fabric with no discernible inclusions

and a regular clean fracture. The lack of decoration on an upper body sherd, its colour, fabric and fracture, all suggest that this piece is likely to be from a small vessel, probably a beaker of Nene Valley colour coated ware. Its date is in keeping with the Severn Valley

ware and is consistent with the 4th century AD date attributed to the rest of the Roman

material from the site.

All of the Roman material found at Knowth has been reassessed in preparation for a forthcoming monograph (Knowth, RIA Monograph No. 5, forthcoming). I am grateful to Helena King at the RIA, for making this available to me, in advance of publication. The

finds were concentrated at the east side of the site, in the area with the most dense

occupation evidence, although the Samian ware sherds came from two different

quadrants of the site, one in the west and one in the south. The Samian ware sherds are from two highly decorated bowls of Dr 37, from Lezoux, and can be reliably dated to the AD 120-150. There is no evidence that either sherd has been reworked, and Mulvin has

argued that they are contemporary with the rest of the domestic evidence from the site (discussed under toilet implements) (Mulvin forthcoming).

The assemblage of Roman ceramics from Raith na Senad has now been fully published

and Evans has classified approximately 25 sherds of possibly Gallo-Roman, or Romano-

British ceramic vessels from the site. A further two vessel sherds and a mortaria sherd

were sent to various Roman specialists in England for classification after the initial

excavation, and these are all now lost (Evans 2008: 123). About one third of the sherds

are of Samian Ware but the rest are a mix of probably Severn Valley Ware and various

other oxidised forms. Three sherds are from a Dechelette type 72 cup, a type that is not

common even in Roman Britain (Evans 2008: 124). Notable also is that many of the

ceramic forms now classified as both Samian ware and Severn Valley ware are drinking

vessels.

115

Metalwork

At Lyles Hill, in County Antrim, there are

two sites that appear to be linked by the

evidence of non-local material. One of these

contained a small cache of objects that

included a tiny bronze bell, jet bracelet,

amber beads and a glass bead. The glass bead

was identified as a Guido Class 9 Roman

-y

. -.. .. J

-. 1. -'

type and the closest parallel for the bell found Fig. 5.8 Decorated strap-tag from Rathgall,

in the same context is a Roman-type horse not to scale (after Raftery 1994)

harness bell (Warner 1995: 28). A 3ia century

AD military strap tag (Fig. 5.8) was found during excavations of Rathgall hill-fort (Fig.

5.9) and classified as the end of a cingulum belt (Raftery 1994).

Two copper `bun' ingots, identical in both manufacturing style and weight to those

produced under in Roman Britain have been found in north County Dublin, one at

Damastown and the other at Drumanagh (topographical files: NMI: consulted 2008;

Raftery 1994; Mitchell and Ryan 2003). This area is rich in natural deposits of copper so it seems unlikely that there would be a demand as such for imports. Much more likely is

the techniques of processing and manufacturing which were common in Roman Britain

has been developed at a site local to the area.

116

Fig. 5.9. The impressive hillfort of Rathgall in County Wicklow (after Raftery 1994)

At the site of Dün Ailinne, in County Kildare, a series of brooches was found within an

Iron Age context along with several pieces of metalwork (non-diagnostic), a lance head, a

button and loop fastener, and a decorated bracelet (Wailes and Johnston 2007). The

relative date and the parallels for the material found fit with the relative date for burials

with grave goods at Lambay Island, off the coast of Dublin (Whimster 1981). Although

perhaps pre-dating the 1 s` century AD Roman military campaigns in the north of England,

the individuals at Lambay were buried with weapons, which suggest some Roman

influence. The jewellery and brooches included fibulae of Dolphin, Langton Down and

Thistle types. The scabbard mounts and shield boss can be paralleled at Silchester (Rynne

1976: 242); the triangular plaques are closely paralleled in north-east Gaul and associated

with the tribe the Menapii (op cit. 242); the swords were of non-La Tene type with a

straight quillion and like some found at Lagore crannog they are of Roman style. The

heavy beaded iron torc can be paralleled from sites in Yorkshire, and as these have a very

117

restricted distribution, they are usually associated directly with the Brigantes (Rynne

1976: 239). Although a Roman coin of Trajan or Hadrian was found in the area in the 1840s, it is has been lost, and in any event we cannot be sure whether it relates in any

way to these burials; on parallels alone the burials appear to be over a century earlier (Bateson: 1973: 52; Rynne: 1976: 231; Di Martino: 2003: 151; O'Brien 1999: 26). A

decorated gold band was found in the same area, which Rynne dated to the second half of the 1" century AD, parallels for which have been established at Rodborough,

Gloucestershire, but again, this may reflect later activity on the site, and there is no

evidence that it was associated with the burials (Rynne 1976: 232).

At Raith na Senad an array of copper alloy, iron and lead objects have now been

classified and the range and quality has been described as being consistent with what

would be expected from a Romano-British site (Allason-Jones 2008: 107). Only one

piece of a fibula (E615: 220) was identified from the small finds but several early

penannular brooches (both whole and in pieces) were identified. This brooch form is

equally common amongst fords from Roman contexts in Britain as the bow brooch, and

good parallels are found at Romano-British temple sites such as Uley and Lydney Park in

Gloucestershire. Many iron objects were found throughout the site the function of which is now difficult to identify with any certainty but several examples of nails with square

sectioned heads were classfied as of Roman type. A lead seal with an official Roman

insignia was also found at the site and it provides a direct link between the activities at Tara and engagement within the Roman economy (Allason-Jones 2008: 109). Amongst

the many metal fords at the site were craftworking tools, including a set of dividers, and

the remains of a Roman type barrel lock. Interestingly a key designed to open such a lock

was found at the site of Uisneach in County Westmeath (Macalistair and Praeger 1928-9:

121; Bateson 1973; Raftery 1994: 209). The key was found within the same context as a 0 century AD Roman coin during an excavation in the 1920s and more recent landscape

survey has suggested that the site of Uisneach should be reclassified as an Iron Age hill-

top enclosure of c. 100 BC to AD 200 (Warner 1995: 27; Donaghy and Grogan 1997;

Donaghy 2008: 113).

118

A Roman-style patera and saucepan handle were found in a field in Bohermeen in County

Meath and the unusual handles on these can be paralleled with a similar example from

Glenshee in Scotland (Robertson 1970; Bateson 1973; ). Athough no explanation for the

circumstance of their finding could be offered for them at the time of Bateson's work,

Bohermeen is the townland within which the site of Faughan Hill is located. Faughan is

known throughout the historical records as the burial place of Niall NoighIallach, who is

thought to have led raiding parties into Roman Britain (see Chapter 2 for discussion,

Raftery 1994: 218; Daffy 2002; Newman 2005: 386). Patarae were also found during

excavations on Iniskea Island County Mayo at a site with industrial scale purple dye

extraction, and further examples were found on Rathlin Island, off the coast of County

Down (Henry 1945; Raftery 1994). At Rathlin, a range of small metal objects including

knobs and rings were also found and described as being of Roman style (Bateson 1973:

66). A small votive bronze of Mercury was found at Caugh Hill in County Derry, close

to the River Bann and to where the Broighter Hoard (believed to be of 1st century AD

date) was found in the last century. The figurine from Caugh Hill can be paralleled with

one from Bristol and one from Perth in Scotland (Warner 1983: 179).

Glass

At a site near to Lyles Hill, an extended inhumation burial was found that appears to have

been buried with a stone urn and a glass urn, a jet bracelet and two glass bracelets. The

glass urn was sealed with a bronze lid, in a Roman style (Warner 1995: 28). It is unclear

whether this is the same glass urn of Isings form 67(a) that came from the townland of Dunadry on the Antrim coast (Bourke 1994: 164).

A Roman glass cinerary vessel and associated gravegoods were recovered from a ringfort

at Stoneyford in County Kilkenny around 1850. The cremation burial can be paralleled

with a glass cremation urn from Colchester dated to the middle of the 1st century AD

(Bourke 1994: 164; O'Brien 1990: 37). When found, the intact green glass urn also of Isings Form 67(a), (Fig. 5.10) still contained the remains of the cremation. The damage

119

visible on the image at Fig. 5.10 is as a result of it

being broken accidentally at the NMI in recent years,

at which point the cremated remains were lost.

Amongst the associated finds at Stoneyford was a

toilet implement decorated with dot and circle motif, a

glass unguent bottle and the urn itself was sealed with

a decorated bronze lid (Fig. 5.10) (Bourke 1994: 164;

O'Brien 1990: 37; Raftery 1994).

The Roman glass from Raith na Senad was

originally classifed by Bourke in 1994, and he was

invited to review and catalogue of the material

once more for the recent monograph. Five glass

sherds were recovered from the site and two of _b u

these form part of the same vessel. All of the Fig. 5.10. The Stoneyford Roman cremation. Top the glass vessel used as a cinerary urn,

glass vessels found at the site are of highest bottom the bronze plate with which it was sealed and the glass ungent jar which

quality soda glass and they are uniquely Roman accompanied the burial (after Daffy 2002,

in character and were imported to the site Rafterv 1994)

(Bourke 2008; 115). The forms catalogued are one mould blown drinking cup, (2"d

century AD), three unclassified sherds found in association with an early form

zoomorphic penannular brooch (datable by similar brooches found in Roman contexts at King's Weston Roman villa and Lydney Park to the 3`d century AD), and two conjoining

sherds from a cone beaker with wheel-engraved decoration, Isings form 106(a) (4th

century AD). As with all of the most common form of ceramics at the site, it is notable

that these are all of the highest quality glass drinking vessels (Bourke 2008: 116).

Approximately 15 other glass beads of varying colours, glass pin-heads and one example

of window glass, were classified within the small finds from the site (Allason-Jones:

111).

120

Glass beads of many different forms and styles, including melon beads, are common from finds across Ireland and those which have been classified as imported are detailed at Appendix B (Table 1). Several pieces of glass that were described as Roman were found

during the excavations at Knowth, but these await classification (Eogan 1974). A large

lump of red glass was found at Kilmessan, near the Hill of Tara, and it appears to show

the evidence of knapping (viewed at the National Museum of Ireland 2008). A chemical

analysis was completed on the glass from Kilmessan and it established that the level of

antimony present in it is consistent with a Roman origin (Bourke 1995,2008: Stapleton et

al 1999: 919).

The availability of scientific techniques I

and shared analytical and theoretical

research models, all offer the possibility

that some context may be recovered for

these objects, but even without these,

new contexts can be offered through

fieldwork investigation. This is

demonstrated by recent landscape

investigation into the area in which the

small collyrium stamp was found at

Golden in County Tipperary (illustrated

at Fig. 5.11). The little stamp, now

housed in the British Museum, is made of

/7' //////t,

" %/

"Z 1.7 "/

11NRtTVq'u3\vMM l CLAZV 2Y ic Fig. 5.11 The collyriwn stamp from Golden in Tipperary, not to scale (after Daffy 2002)

hard green schist and is inscribed M(arci)

Iuvent(i) I Tutiani/Diamysus Ad Vet(eres) Cic(atrices) (Raftery 1994; Daffy 2002; RIB

II. 4: 2446.28). The fieldwork found that there was a small holy or sacred well at the site,

and that human remains had also been uncovered at the same site where the stamp was found. Given the parallels with similar finds of collyrium or occulist's stamps at the cult

site at Wroxeter and Lydney, it was argued that there may have been a small Romano-

British curative cult on the site (Daffy 2002).

121

L43 - Material or sites for which only relative dating has been proposed and which

require reassessment

The diagnostic nature of much of the imported material of the period has resulted in it

being used to offer a relative date for the context in which local non-diagnostic material has also been found. Close typologies and chronologies for ceramics of this period

alongside clear parallels for their use at sites in the provincial Roman world, with the

additional evidence provided by coinage, make it one of the more reliable relative

methods in use for imported material. This classification was designed for sites within the

Later Iron Age in Ireland that have been ascribed a relative date based on assumptions

about similarity in non-diagnostic ̀local' material but it does not apply to imported

material.

IA4 - Material from sites where the stratigraphical relationship between contexts

remains unclear, even where some have been securely dated under either L41 or IA2 classification

Within this final' classification are those sites where Roman material was thought

irrelevant to the dating sequences overall or where there has been no investigation of a

site where Roman material has been found. The material of Roman and Germanic style from Lagore Crannog (discussed in detail in the previous chapter) was not considered

relevant to the dating of the site overall. Roman style toilet implements and Samian ware (Curie Type 23) were amongst the finds at the crannog at Lough Faughan in County

Down and several sherds of New Forest ware were found just off the coast of County

Antrim on Rathlin Island. Other crannog sites have also produced finds of imported

ceramics including Island McHugh, County Tyrone (rim sherd of Dr 35 cup), Ballinderry

No. 1, County Westmeath (sherd of New Forest ware), Ballinderry No. 2, County Offaly

(sherd of Arretine ware and melon beads) and Cloonfinlough, County Roscommon (blue

glass beads and a copper coin of Hadrian).

122

Another very important site discussed in

Chapter 2 is the promontory fort of Drumanagh

in North County Dublin. The fort (Fig. 5.12) is

located fifteen miles north of Dublin, beside a

sheltered harbour, known locally as

Loughshinny. Reaching out into the sea, its

interior is defended by high sea cliffs on three

sides and a form of close-set multivallation on

its landward access (Raftery 1994: 204;

Cooney 1993: 26). The extent of Roman

material found over the years at Drumanagh

remains uncertain, and it is unlikely to be

published until a legal dispute between the

landowner and the Irish government is resolved

Fig. 5.13. The 2"d century AD enamelled seal-box found by metal detectorists at Drumanagh. Image reproduced by kind permission of EP Kelly, NMI

123

Fig. 5.12 Aerial view of Drumanagh, showing rectangular footings (after Raftery 1972)

Given the size, location and structure of Drumanagh it is best paralleled with other Iron

Age sites in Britain, such as Hengistbury Head, which served as a trading centre or

entrepot (Warner 1995; Cunliffe 2001). Opposite Drumanagh on the Wirral are the

remains of the now denuded port of Meols, from which an array of Iron Age, Roman and later Saxon material has been found over the last centuries (Griffiths et al 2007).

Detailed analysis of both the site and the finds from Meols were recently published, and

the authors concluded that, although the site has now been almost entirely lost through

coastal erosion, the evidence collected over the centuries suggests a major entrepot

existed here from the Iron Age onwards. Its position across the Irish sea from

Drumanagh is noteworthy, as is its position in relation to the legionary fortress of Deva

Victrix, the naval base for the Classis Britannica during Agricola's campaigns (Mason

2001).

The material from Drumanagh is of local types and imported objects the date range for

which appears to be from the 1' century AD to the 6'h century AD (Mitchell and Ryan

2001; Raftery 1994; Warner 1994; Newman and Halpin 2006). Samian sherds have been

collected from the site and some unclassified amphorae sherds have been mentioned in

publications (Raftery 1994; Mitchell and Ryan 2001). Two unclassified Roman coins (no

further details of these are known) were found in a ploughed field close to the site and a Romano-British type of copper ̀ bun' ingot has also been found (Raftery 1994: 208). The

intervention of the Keeper of Irish Antiquities at the National Museum of Ireland (NMI)

prevented a sale at Sotheby's in London in 1993 of a range of looted items from the site. Although these are not available to study, they have been referenced within other work by those who have viewed them and appear to consist of a range of votive pieces and

metal-work of both local and Roman origin (Mitchell and Ryan 2001). An enamelled

rhomboid Roman seal box, which on typological parallels dates to the 2nd century AD

(perks & Roymans 2002) was found by metal detectors in April 2005 (shown at Fig.

5.13), and a Roman military strap tag of unknown date was recently found at the site (Kelly, pers. comm, 2005 & 2006). An archaeological survey has been carried out on the

site but this has not been published. The site remains at the centre of an on-going legal

124

debate between the owner and the Irish government as it has been scheduled under Irish

Law (to avoid further deep ploughing). The State does not want to pay the now, much inflated, price the owner wants for his unusable land and all of the material will remain

sub judicae (and unavailable for viewing) until the dispute is resolved (Kelly pers. comm: 2005). Although we can speculate about the date range of the material found, without the

results of the survey conducted by Kelly and a clearer idea of the stratigraphy of the site,

the material remains within an IA4 class series.

The only known Latin inscribed stone from Ireland

needs to be considered within this classification as

although we know it was found on Station Island in

Lough Derg, County Donegal, the circumstances of its

finding are vague (Macalistair 1945: 302). The name

`lough dearg' translates as the `red lake', and the

hagiographical record relates this to a battle between St

Patrick and the last great serpent of Ireland, believed

'. 0 R ýý E (7, NVS T

EL! SASOrl ! Vri

to be a veiled reference to a surviving centre of pagan Fig. 5.14 The Latin inscribed stone

worshippers, converted by the saint in AD 445 from Lough Derg (after Macalistair 1945)

(Macalistair 1945).

The stone (shown at Fig. 5.14) measures approx. 30cm by 28cm, and it is inscribed in

Roman capitals and illustrated within Macalistair's Corpus Inscriptionum Insularum

Celticarum (CIIC) from 1945. The reading by Macalistair is as follows: OR

ORNECNVS ET ELISA SOMNIVM. The lapidary appears to have made two mistakes

in his work: the first is the N of ORN was written with the diagonal stroke the wrong way

round, and subsequently corrected; the second mistake appears to have necessitated the

later carving of the letter E of E(LISA), which although also in capitals is smaller than

the other letters. The illustration in CIIC, shows the C at the end of the first line is

indistinct and oddly aligned, and all of the right hand side and the last letter of the fourth

line appear to have suffered some damage (read as N by Macalistair 1945: 302). I would

suggest that the C might be better read as an 0 which would read from the second line

125

ORN[AMENTUM] EONVS ET ELISA SOMNIVM, which offers a dedication and a Roman praenomen for both the male and female visitors. The use of the term somnium is

especially interesting, given that role of the oracle and oracular visions are more familiar

to us from the Classical World within works such as that of the ̀ Somnium Scipionus'

written by Cicero in the ist century BC (Hanson 1983). The style of epigraphy, in Roman

capitals rather than insular majuscule along with one, or as I have suggested two Roman

names all point to an early date, but without further details this remains speculation.

5.2 The Later Imported Ceramics

I want to discuss the later ceramic material separately, because our understanding of the

imports within the later period has altered greatly since Bateson's work. In Britain the

period after the collapse of Roman control and administration is known as Late Antiquity

and it clear from the debates taking place that there is still much to be learned about the

shift in control and the emergence of kingdoms and discrete political centres in both

Britain and Ireland (Dark 1996,2002; Newman 2005; Campbell 2007). The recent

excavations in Ireland have revealed many new sites with imported ceramics, and

although most are as yet unpublished, summary details of the excavations and finds are

available at www. excavations. ie. In the absence of published work, references to the

material found at these sites are given as the date of the excavation and the licence

number. The ceramics for this period have a long history of scholarship, and were first

recognised as imported wares at Tintagel, by Radford (1956). Thomas (1959) added to

the classification and his work and this was refined by Campbell (2007) but most notably

by Kelly (2010). The most frequently found forms of later imported wares in Ireland are

as follows:

126

Original Class Revised Class Abbreviation Origin

A Ware

Ai African Red Slipped ware ARSW Tunisia

Aii Phocaean Red Slipped ware PRSW Phocaea, Turkey

B Ware Late Roman Amphora LRA (+ sub class)

Bi LRA2 Peloponnese, or possibly Kos or Bii LRAI Chios

Cilicia, Turkey

D Ware derivee des sigillees paleochretiennes DSPA Bordeaux

E Ware E ware E ware Origin unknown but probably

north-west France

Table 5.2. Campbell's classification of later imported wares found in Britain and Ireland

(2007): based on Radford (1956), Thomas (1959), Rigoir et al (1973), Kelly 2010

One of the most important sites excavated in Ireland with evidence of later imported

material is Lisnacaheragh Ringfort in County Cork, known from the name of its

townland, as Garranes. Garranes is an impressively large ringfort, with a form of close set

multivallation and has the only recorded in-turned gateway in Ireland, suggesting a highly developed defensive structure (Ö'Rfordäin 1942: 140). The site was excavated in

1942, and tentatively dated to the year AD 500 based on the evidence for Thomas' A

ware (ARSW) and Bii ware (LRAI). The report, however, drew attention to the

significant evidence for much earlier material at the site (Ö'Riordain 1942: 142). Recent

work on the Iverni project in the environs of Garranes, by archaeologists at NUI Galway,

has recovered more LRAI, during fieldwalking in the adjacent area (Comber 2009).

Garranes had as many as eight massive timber gateways over the period of its use, and

the sheer quantity and wide range of Roman pottery, glass and jewellery from the site is

unparalleled in Ireland. Until recently, it was also the only site in Ireland that had

produced evidence for early craflworking in gold and glass. Two more ringforts were

excavated at Garryduff, close to Garranes, and they too held some exceptional finds

(O'Kelly 1962). Garryduff produced the earliest evidence of a furnace in Ireland, and

along with extensive tools, and it is clear that craftworking was one of the principal

127

activities at the site (O' Kelly 1962; Edwards 1996: 88). The pottery from Garryduff was

recently reassessed and as with Garranes, it dates from the later 5t` century AD (Cambell

2007). The history of these sites is linked to the dynastic group previously mentioned, the E6ganacht Rathlinn, whose tribal capital is said to have been established at Cashel in

County Tipperary, not before the beginning of the 4th century AD. Recent excavations at Cormac's chapel on the Rock of Cashel appears to confirm an early sequence of

occupation, with a find of a fine 1 s` or 2°d century AD Roman fibula, Romano-British

coarse-ware and a later stratified series of LRA1 (Hodkinson 1994).

Another coastal promontory fort on Dalkey Island, near Dublin, was excavated between

1956 and 1959 (Liversage 1968). The site was occupied over several millennia with clear

evidence for Neolithic and Bronze Age settlement on this rocky outcrop. The report followed the norm by classifying the ceramic assemblage from the site into that which

was prehistoric and the rest under the `early Christian' catch-all. Amongst the later Iron

Age material were two pieces of Samian ware, both body sherds of Dr 30 and Dr 37

respectively. One sherd was marked by the potter DOECCUS whose work originated in

Lezoux in the 2°d century AD. These sherds were found in the same context as several

pieces of green `Roman' glass and one piece described as Roman `window glass'

(Liversage 1968: 193). In the specialist section of the published excavation there was a

record of a disagreement between two Roman glass specialists on whether some of the

pieces were etched Roman forms or whether striations evident on the sherds were due to

taphonomic processes. Harden argued that etched forms of decoration had ceased to be

used long before the date given for the site, that is, by the 5'h century AD (Liversage

1968). The relevance of the Samian ware does not appear to have been taken into

consideration as regards a possible dating context for the glass; and yet the two might be

contemporary. The ceramic assemblage from Dalkey was reconsidered in a paper in 1998

and reclassified as LRA1 and LRA2 ceramics (Doyle 1998). The construction of parts of

the promontory fort appears to be much later than would have been previously thought as

sherds of LRA1 and E ware were found stratified below a section of bank. This suggests

that at least one section of the ramparts was reconstructed and rebuilt at a significantly

later date than previously thought (Doyle 1988: 97).

128

IRELAND: TRIBAL TERRITORY BASED ON MANUSCRIPTS

`ýp

0 //® 8ANTRIM DONEGAL

'ULAID (LATER Ui NEIL) f"%

W* 40 TYRONE

�0% 0

0wr o

DOWN O i4RMANAGH

ýýNA 1 SLIGO 1" LII RIM

MAYO CONNACHTA CAVAN Lo, J_

FOSCO'. ' ON MEAT

moo

° LONGr ORL) r0"

J

DO 0 rJ

&0

WESTME*H DO 0 GALWAY

0EL ELIN . OFFALY 0

^_ Oo

O KILUnN:

LAIGINI WVLOY:

OIS

/

CLARE 0

XLOW TIPP RY

LIMERICK "T. XFORD

®E ANACHTN

KERRY WATERFORD

ORK KEY. OO BURIAL

O FIND; istto3(dCADI ® HOARD 0 FIND 14th to 5th C AD)

Fig. 5.15 Updated map of the distribution of early and later imported material in

Ireland (drawn by author) showing the nodal points of entry of earlier and later

material: note the overlap for the entire period evident in the greater Meath/Boyne

environs. Evidence for later E ware is not shown on this map but is shown at Fig. 8.1

129

Any study of ceramics from the entire Roman period in Ireland runs foul of accusations

of the ̀ maximalisation' of the evidence, due to the small number of total sherds

recovered at these sites (Hodges 1982: 51; Campbell 2007: 7). This problem was discussed in detail recently and it was argued that it is not just the amount of evidence in

a quantitative sense which is important, it is what the occurrence of these types of imports, alongside other material such as glassware, tells us about the places where the

social interaction was taking place (Campbell 2007). The relatively small number of

ceramics found in Ireland has been used in the past to argue against any significant

engagement with the Roman economic markets, and more recently against any regular

trade between Ireland and provincial Roman Europe (Fulford 2005). Yet, there appears

to be a direct relationship between what we know of the origins of the sites where the

later material has been found and the emergence of the political tuatha, or kingdoms in

Ireland and a similar pattern is evident in the emergence of powerful kingdoms in western Britain and Scotland (Newman 2005; Campbell 2007: 140; Kelly 2010). Furthermore the

evidence for the later material follows exactly the same distribution patterns in Ireland as it did in the earlier centuries (see Fig. 5.15).

We now have evidence PRSW and LRAI at Collierstown in County Meath, stratified

within the later Iron Age contexts (2004: 04E0422). At Colp West, at the mouth of the

River Boyne, Gowan recorded that finds of LRAI were stratified within contexts that

appear to predate the many inhumation burials at the site (1988: 0122746). More recent

excavations at Colp West have revealed a complex of enclosed settlements and burials

adjacent site to that excavated by Gowan (2001: 99E0472). A total of 17 corn drying

kilns were excavated and the site has been classified as transitional with what appears to

be datable sequences of occupation from the Bronze Age, through the Iron Age to the

Early Historic period. At Cherrywood, again in County Dublin, `B ware' (unclassified)

was found in association with an series of inhumation burials, which are as yet undated (1998-1999: 98E0526,99E0517,99E0518,99E0519,99E0523). At Cabinteely in Dublin

sherds of PRSW, LR1 and DSPA bowls and their lids were found within stratified

130

contexts (1998: 98EO035), and at Ninch DSPA was found with mortaria bowls, and later

contexts at the same site contained sherds of E ware. Sherds of `B ware' (unclassified,

but probably LRA 1) were found in Tipperary, at the site known from the famous

Derrynaflan hoard (1987: S 108490). A find of ARSW has now been confirmed from a large ringfort, or possibleferta, between Bennettsbridge and Kilree on the banks of the River Nore in County Kilkenny (Kelly pers. comm 2009). This site is only four and half

miles from Stoneyford where the Roman cremation burial was found.

The concentration of finds are in both the early and later centuries are within areas which

are today some of the most fertile agricultural land in Ireland and it likely that then, as

now, these areas were capable of producing a surplus which enabled trade for these

imported goods (discussed in detail in Chapter 8). Along with the continued importation

on the eastern seaboard, there is a new distribution entry point through County Cork, in

the South. At Caherlehillaun in County Cork, Sheehan has identified what he believes to be the earliest Christian church in Ireland, and at this site there is evidence of LRA1, and

some E ware which was found within later contexts (Sheehan 1998).

As discussed in Chapter 2 Bateson excluded most of the material for which no context

could be confirmed, although as also mentioned, he appears to have been inconsistent and included some within his accepted category (most notably the Stoneyford cremation burial and its associated grave goods). Some of the material he placed into his `rejected'

category is not disputed here, such as finds from excavations where it can be confirmed

that significantly later material was found stratified within the same context. There is a

problem, however, in establishing whether single finds from older excavations were

actually found within the same stratigraphical context, or simply at the same site, but it

should be said that this is a common problem in older excavations where recording

techniques were somewhat inconsistent. This distribution of imported material supports

the impression that there were key importation points in the earlier centuries: around Down and Antrim (the shortest crossing points between Scotland and Northern Ireland),

and the greater Dublin area, including the outlying islands. It also suggests that some of the material rejected by Bateson may reflect genuine imports. If we accept these as the

131

principal areas of contact, then it is perhaps no coincidence that they are strategically

opposite the legionary strongholds of Deva Victrix (Chester) and Segontium (Caernarfon)

and the important contemporary entrepot at Meols, on the Wirral (Griffiths et a! 2007).

Within the 3`d and 4th centuries AD the relative peace and prosperity of the villa estate

economy in the south and west of Roman Britain would have created ample opportunity for interaction between those on either side of the Irish Sea. In the later 4th century AD,

although imports continue to cluster around the greater Dublin area and into County

Meath, a new point of entry is discernible around the area of Cork harbour on the south

coast (shown at Fig 5.14). Although Bateson dismissed two coins hoards, one from

Cushkinny in County Cork, and another from Rdith na Senad at Tara, these are both

included here as contemporary imports. It seems clear that despite the fall off of central

administration throughout the Roman provinces, some communities in Ireland appear to have been well placed to exploit a new route for the direct importation of material from

the Mediterranean.

5.3 A reconsideration of the `Hibernia-Romana' Debate

This chapter has outlined and discussed the evidence for imported material in Ireland and

an updated catalogue is presented at Appendix B, Table 1 and at Fig. 5.16 is a map of all

the key sites mentioned throughout the discussion. Given that there is growing evidence

of later imported ceramics, I want to offer a reconsideration of the controversial debate

that was published within the archaeological literature in the 1990s, discussed in Chapter

2. The central tenet of the hypothesis contained within the article by Warner was the

suggestion that the Roman military may have taken an active participation in local

politics in Ireland, to secure a friendly ally at the time of Agricola's 1st century AD

military campaigns into northern Britain. This part of Warner's hypothesis is

supportable as good parallels can be offered for similar action in the Roman provinces, but there is a problem with reconciling the material found at the sites he used to support his theory with any explicit military function. Warner used the material found at Dün Ailinne and Rdith na Senad on Tara, along with the exceptional finds of precious

132

jewellery and Roman gold coins from Newgrange and material from Freestone Hill in

County Kilkenny.

The material from Tara and Newgrange is contemporary (IA1) with activity at these sites between the 2nd and 4th centuries AD (Grogan 2008). The material from Newgrange is

datable not only by the finds of Roman coins within the same contexts as the jewellery,

but also an Iron Age radiocarbon date from a sod layer within the mound (UB-360

1650±45), 259-536 cal AD (Radiocarbon: 1971). The material from Freestone Hill is

datable by the find of a Roman coin (IA2) within the same contexts as Roman jewellery

and ceramics, and it is broadly contemporary with that from Tara and Newgrange. The

material from Dan Äilinne, however, is considerably earlier than that from the previous

sites mentioned, and therefore it is more appropriate to describe it as British, and of PRIA

date, or of the period directly after the conquest, before more Roman forms became

commonplace amongst the material assemblage of northern Britain (Wailes 199: 10). One

of the brooches, a Nauheim derivative had been repaired and a new pin fitted to it, which

suggests that it had seen some wear before it was lost (ibid; 10). The Rose/Mauve phase

of the site, within which the material was found, has a wide radiocarbon dating span, from (S1984 2200±50), 391-116 cal BC, to (S1980 1900-+85), 96-336 cal AD (Johnston

and Wailes 2007: 180). The date range of the material found at the site is comparable

with the material from the burials at Lambay Island, also used by Warner to support his

hypothesis. The material from Lambay falls within IA2 classification, as it has not been

scientifically dated, but close parallels of the material can be offered from contexts in

northern Britain that fall into the immediate post-conquest period (Whimster 1981). The

material from D6n Ailinne, Lambay, Tara Freestone Hill and Newgrange is not directly

comparable to each other, nor is it contemporary with each other.

133

KEY TO SiTES WITH IMPORTED MATERIAL

1. UfJil? ýff1 ... h. lJ 1__ ""...

2 L. «+WeyISlm. d 22 L011r ^" 3 brýlr Heed 21 D. - 4 Lagorc

14 Garar, c;

S Tata 26 Crwc--It I4am,, 111-11 6 Pwlkryy Mord Z7 Bd4r. r. r. 7 Nowgran9m & Kýn. ýrth 18 Balknc 9 Bet ;9 Cushkmny

. ct. n 9. Freestone Hill 10 Flower H111 II detgh Min 12 BelIndwry t 13 8alhndcfry 2 14 Wand WHQ,, 15 Ciogher 16 Gold- 17 Utsnexh 18 Krihgall 14 Csrbury H111 20 Stetxyrar f

Fig. 5.16 Map of key sites mentioned throughout discussions, the greater Meath environs

highlighted with rectangle (drawn by author)

134

The most significant aspect of Warner's hypothesis was that these sites were the subject of a military campaign. It follows that the material should be comparable from that which we would find within forts and camps in Roman Britain, and herein lies the first problem. In military contexts in Roman Britain, weaponry, lances, shields, spears, swords, strap. tags, and button and loop fasteners are as likely to be found through excavation in extra- mural settlements or sites not associated with the forts, just as items of women's dress or children's shoes are as likely to be found within what was once believed to be an

exlusively male domain (Bowman 1994; Alcock 2006). It is equally problematic to infer

the ethnicity of an object's user, based on the assumed origin of the material culture at a

site alone, as discussed in detail in Chapter 3. Furthermore, such items are common finds

beyond the line of even the Antonine Wall, such as amongst settlements and homesteads

which remained outside the direct control of the Roman military or civil administration (Hunter 2001; Shotter 2006; Kamm 2004). With the exception of the high value finds

from Newgrange (IA1/2) and the Roman-type swords from Lagore (IA4), most of the

material from Ireland is domestic and somewhat unremarkable in character. This does

not, of itself, prove beyond reasonable doubt that the Roman military was not in Ireland

in some form, but it does leave open to question the classification of such material as the

evidence of a Roman campaign. There is one very important diagnostic material, black

burnished ware, the question of which cannot be ignored. Black Burnished ware (or more

correctly SEDBB 1) was the quintessential, utilitarian Romano-British ceramic, and its

ubiquitous presence at virtually all military sites on the western seaboard in Britannia,

and its distribution around Hadrian's wall from the time of its building in the 2nd century AD, suggests that it is this, rather than small items of metalwork, that might be indicative

of a Roman military presence (Allen and Fulford 1996). Given the extent of excavation throughout Ireland in recent years, we can no longer argue that its presence has either been missed, misclassified or discarded. If, however, a small incursion took place, cented

at a single marching camp or temporary fort, and intended as a warning, then it may yet be found (Raftery 1994,2005). The site of Drumanagh, as discussed earlier, may prove to have been such a temporary fort for a Roman campaign as it certainly appears to have

a grid pattern which is uncharacteristic of other Irish sites (see Fig. 5.12). Until the

material and the survey results can be examined however, any definitive conclusions will

135

have to wait.

The most important aspect of Warner's

hypothesis is that he highlighted the

correlation between British and Romano-

British material and sites where

monumentality in the form of

. r+ +}+

" .,

" My 'ýý

e' ý y; ý'ýý

ý"1ý"

-; considerable earthworks and structures

are contemporary (based on the IAI and

IA2 evidence). Although Warner argued

that no local material was recovered at

the sites he used within his hypothesis,

this needs to be corrected, as the latest

publication of the material found at Räith

na Senad on Tara catalogues non-

diagnostic material which may be of

local origin.

_4 )

i ýºý! iiiFr

r ý_

wý

4

t.

. 44

Fig. 5.17. Survey results for the enclosure overlooking the River Boyne, between Knowth and Newgrange (Warner 2009)

Recent geophysical surveying in an area directly overlooking the River Boyne between

Newgrange and Knowth has revealed the details of a previously unrecognized rectangular

embanked enclosure. The image at Fig. 5.17 shows the results of the combined

magnatometry and resistivity surveys, which revealed a circular or polygonal structure

within the enclosure (I am very grateful to Dr Warner for permission to include his

unpublished work within this research). The site has not been the subject of any

radiocarbon dating but its discovery suggests that even at sites such as Newgrange and

Knowth, both of which have been the subject of immense levels of research over many

decades, there remains much to learn. Warner's research has demonstrated that we need

to look beyond what we think we know about these sites and in so doing we may discover aspects that offer an alternative history for them. The growing evidence of direct

Roman imports into Ireland needs to be assessed within a framework which reflects both

the wider Roman world and the specifics of what is known about the Irish Iron Age.

136

Chapter 6 Death and Burial in the Later Irish Iron Age

`Though grave diggers' toil is long, sharp their spades, their muscle strong, They but thrust their buried men, back in the human mind again. '

(Under Ben Bulben, WB Yeats 1939)

6.1 Background to the research

In the last years of his life, the poet William Butler Yeats found himself writing missives

to his Irish audience, cautioning them not to forget their ancestors, their history and their

past, at a time of great social and political change in Ireland. He appears to have believed

that if not in life then perhaps in death the souls of the Irish might find peace and cast off the yoke of British subjugation adding ̀ that we in coming days may be still the

indomitable Irishry' (Under Ben Bulben, iv. 1.15). As recently pointed out by Seamus

Heaney, Yeats was part of the Protestant Ascendancy in Ireland and born into landed

Irish gentry, descended from Williamite settlers, and educated in part in England. Yet, for

many, the quintessential portraits of `Irishness' are those associated with the poetry of Yeats, who brought the ancient manuscripts to life in extended colourful verse, notably in

English, as he could not speak Gaelic. The barrows and cairns that dot the monumental landscapes of what became Yeats' home county of Sligo were the familiar subjects of his

poetry, through which he brought to life the heroes of myths and legends of the ancient Irish past. Perhaps, in these last poems, Yeats recognized his own mortality and that he

would soon form part of the misty history of `Celtic' Ireland that for so long had been

the subject of his work. That a life should not be forgotten and that a name should be

spoken as an act of remembrance are concepts that cross the boundaries of both modern

organised religions and older belief systems, and they bring us back into the realm of the

archaeologist.

The geographic region under discussion, the plains of the County Meath, is possibly the

most important in all Ireland in terms of the density of historically attested ̀royal' sites

and sites of ritual significance. Sites such as Teamhair (Tara), the passage tombs of

137

Newgrange, Knowth and Dowth; the Genach sites of Tailtiu (Teltown) and Tiagta, (The

Hill of Ward) are steeped in history, myth and legend (Newman 2005). In Chapter 2, I

discussed how the culture-historical narrative was created for these historic sites and highlighted the problems inherent in any attempt to separate the archaeological facts from

a nationalist myth-history. In proposing a new framework for how all of the material

might be reconciled within a broader archaeological interpretation it is essential that we

unlock the circularity of exiting interpretation, based on the old chronology and typology,

and a new direction through the application of new research methods. Although the

records pertaining to Roman material allow us to document specific finds, much of the

material detailed in the previous chapter has been lost over the years and is now

unavailable to study further. The advances in scientific methods in palaeopathology and

osteo-archaeology, however, presented an opportunity to conduct an analysis on human

and faunal skeletal material from those burials that have also been classified as ̀ intrusive'

within the literature (Raftery 1981,1989,1994 & 2005; Warner 1995,1996a & 1996b;

Newman 2005; O'Brien 1990,1999a, 2006; McGarry 2005,2007). These burials have

been classified on the basis of the transitions in the burial practices used, this in turn has

been developed into a relative dating method applied widely across burial sites in Ireland.

6.2 Burial in the Later Irish Iron Age

Before turning to discuss the distribution of burials and how these have been interpreted,

it is important to clarify the approach used here. In Iron Age burial contexts the relative

positioning of the head, arms, hands and legs are used to infer deliberate placement at the

time of interment, and this positioning has in some instances been used to infer Christian

burial (O'Brien 1999; McGarry 2007). Where there is uncertainty through disturbance of

the grave context it is best to avoid categoric classifications as taphonomic processes have to be assessed at the point of excavation, as even voids within a grave will create

additional pressure during the process of sedimentation. These can alter the position of

the skeleton and observations on orientation, such as head position, might be due to the

natural decay process rather than a deliberate or symbolic positioning of the corpse (Duday 2006: 35). Although it has been argued that intentionality has to be established

138

through excavation in order that we can confirm a funerary context, in some cases

especially older excavations it is not always possible to establish this with any certainty

(Duday 2006). At some Iron Age sites in Ireland, notably the King's Stables in Armagh

and at Raffin Hill-fort, parts of human skeletons, skulls, occipitals or mandibles have

been found, and in the past these have been included within discussions about inhumation

practices (McGarry 2007: 221; Newman et al 2007: 351). It is argued here that these do

not fit within a discussion of burial practices as these human body parts were not buried

as individuals as such, but appear to have been selected for a symbolic purpose as part of

a structured deposition (see Newman et al 2007) and it is worth remembering that

`certain intentional deposits of human remains have nothing to do with burial' (Duday

2006: 30). The circumstances surrounding the deposition of the partial skull at Raffen

reflects little that suggests concern for the individual passing into the next world, the

skull itself had been objectified or tokenised within a ritual (Newman et a! 2007).

Furthermore the deliberate overkill (with multiple wounds where any one would have

been fatal) evidenced on most of the Iron Age bog bodies found in Ireland suggests that

these, too, do not reflect wider burial practices, but rather discrete episodes of ritual

activity (see also Chapter 8).

One further qualification should be made to the existing classification and interpretation

of burial customs within the published literature. The terms `primary' or `secondary' used

in relation to burial practices in the Irish publications appear to be used in a temporal

sense to denote how the inclusion of a burial relates to the dating of a context of the site

or monument under discussion. It is rarely, if ever, used to identify or separate the act of

the deliberate internment of a corpse or a cremation in-situ. In the literature outside

Ireland, the terms primary burial and secondary burial have a very different meaning

(Duday 2004: 35). A primary cremation burial is used to denote the place of cremation

where the remains are then interred (primary cremation or bustum); it becomes a

secondary cremation if the remains have been gathered up and buried elsewhere usually

within a cinerary vessel. A primary inhumation burial relates to a body that has been

brought to a site and interred. Secondary burial in this sense should reflect a purposeful

treatment of a body, such as excarnation, followed by the deliberate collection of the

139

remains and interment at a different place. Therefore for Ireland, burials found inserted

into older monuments should still be classified as primary burials, not as secondary burials as currently described, if the site or the context is the individual's original point of deposition.

Cremations

The earliest form of indigenous burial practice in the Iron Age in Ireland appears to be

that of unprotected cremation deposits in pits; often these are a central primary burial

over which a small shallow ring-barrow or embanked ditch was raised (McGarry 2007;

O'Brien forthcoming). As none have been reliably dated before the 1" century BC

archaeologists are divided as to whether this custom reflects the continuation of the

practice of cremation from the Later Bronze Age or its re-introduction under influence

from Roman Britain, given that the closest parallels for the grave goods are found

amongst cremation burials in Britain spanning V century BC to 0 century AD (Raftery

1994; O'Brien 2009: forthcoming; McGarry 2007). There are a few sites where multiple

cremations have been located and which might suggest a permanent population group

nearby. Such community cremation burials have been found at Kiltierney in County

Fermanagh, Ballydavis in County Laois, Oranbeg in County Galway, Carbury Hill Site A

in County Kildare, and within the flat cemetery at Rdith na Senad at Tara, which predates

the activity at the site considered in Chapter 5.

There are only two sites, to my knowledge, where cremation pyres have been identified:

at Rockfield in County Kerry (dated 390-60 cal BC) and at Cush in County Limerick. At

the site of Cush, the excavator believed he had located the cremation pyres of two

unprotected cremation deposits buried in pits below individual burial mounds (Ö'Riordäin 1940; Raftery 1994: 181; Aegis Archaeology: consulted 10/06/2009). There

are little details of what was found at Rockfield as this is as yet unpublished, but

associated with one of the cremations at Cush was a small bone plaque, ornamented with

a dot and circle motif. This plaque is closely paralleled with the bone plaques found at

140

Lough Crew, Cairn H, and recently radiocarbon dated to the 1'd century BC (McMann

1993: 37; Raftery 1994; Kelly 2007: pers. comm).

At some sites in Ireland there is evidence of a layer of charcoal and burnt bone dispersed

widely across the site such as at Grannagh in County Galway, where burnt bone and

charcoal was described as ̀ strewn in handfuls on the ground' (Macalistair 1916: 509;

O'Kelly 1989; 331; O'Riordain 1940; Raftery 1994: 180). At Ballymackaward in

Donegal, the charcoal spread and cremated material found during excavation was dated

to the 2' century AD and identified as a separate burial sequence at the site (O'Brien

1999a). A blackened layer of bone and charcoal was also noted at Räith na Senad at Tara,

but it was not dated separately. It is possible that what we are seeing at these sites is a further burial practice, the deliberate dispersal of cremated remains over an ancestral burial ground.

There are some exceptional burial sites, but in all but a few cases, these were excavated before modem techniques, and details about them are, at best, patchy. One such burial is

that from Loughey, a townland near Donaghadee in County Down, which has provided

much debate amongst scholars within the literature covering this period. There are few

details of the find spot or the nature of the burial, although it was probably a simple pit

cremation (Raftery 1994,2005). The grave goods were somewhat exotic and consisted of

an array of glass beads, bronze finger rings, glass bracelets, a 1'` century AD fibula and

miscellaneous other objects in bronze (Raftery B 1981,1994 & 2005; O'Brien 1999 &

2006 forthcoming). An analysis was completed on the glass beads from the burial in 1987

and it was established that the chemical composition was of continental origin. It was

speculated that the glass was imported into Ireland between the V` century BC and the 1"

century AD and then worked into a finished form in Ireland (Henderson 1987). It is also

possible that the beads were exotic finished imports, given the frequency with which

similar artefacts, some local and some imported, have been found alongside burials which have been securely dated (IAI).

141

Until recently it was generally accepted that cremation practices ceased in Ireland around

the 1S` century AD and were replaced by crouched inhumations within pits (O'Brien

1990,1999: Raftery 1994; McGarry 2007). However, more systematic radiocarbon

dating of burials has revealed that cremation continued alongside the new rite with

examples from Furness (430-600 cal AD) and Carrickmines, (340-540 cal AD) (Grogan

1984; Ö'Drisceoil 2007) both securely dated to the later period.

Crouched Inhumation

Burials that take the form of unprotected

crouched inhumation in pits along with a few

personal grave goods appear to overlap

chronologically with unprotected cremation

deposits, and both are found either within low

mounds or in the ditches of earlier monuments

(both are evidenced at Räith na Senad (O'Brien

1990,2003; McGarry 2007). The crouched or

foetal position (Fig. 6.1) is characterised by the

knees being pulled tightly up to the chest, but it

remains unclear whether the flexed position,

with knees bent less tightly, is in fact a separate

burial practice. The variation may relate to the time it took to bury the individuals, as we

know that although rigour mortis is likely to set in between four and six hours after death,

it will dissipate after 36 hours, allowing the corpse to be more easily manipulated into

position (Janaway 1997: 65). It is possible that the anatomical positioning of the corpse

into a foetal position was part of the funerary ritual in the immediate hours after death,

and it would not have been possible to change this whilst the body remained stiffened

through rigour mortis. This would also offer an explanation for the variation in

anatomical position between those tightly crouched and those that are less so; again this

may be due to the time it took to carry out the burial itself. This is suggested by a burial,

from Laytown in County Meath, where one male (F 181) had his legs drawn tightly up

142

over the left side of his abdomen despite the fact that he was buried in a long stone cist,

which would have accommodated him in an extended position (Eogan forthcoming).

Crouched and flexed inhumations appear to fall within a date range of the 1st century BC

to the 2nd century AD. Although the quantities vary amongst the burials there is a

commonality of artefact type between cremations and inhumations, these are often small

personal items such as an iron or bronze knife, bronze or iron spiral rings, objects of

worked stone and occasionally `exotic' material such as glass beads or continental-type

fibulae (Raftery 1994; O'Brien 1990,2003; Keeley 1996; McGarry 2007). There is also

commonality in location of both cremations and

crouched (and flexed) inhumations, as often they are

found as a single burial inserted into the banks or

within the ditches of much older mounds or

monuments.

6.3 Transition in burial practices

The first transition is, as previously mentioned,

cremation to crouched inhumation in shallow dug

graves (Fig. 6.1). This form of burial rite is evidenced

from the 1 s' century AD in Britain; as only a few have

been radiocarbon dated in Ireland, it has not yet been

established whether what we are seeing is the

evidence of a single episode (the burial of one or

two individuals), or a wider shift in burial customs.

The second transition is believed to have taken place by the 4`h century AD (Fig. 6.2) and it is characterised by extending the body in a supine position (O'Brien 1990; McGarry

2005). It has been argued that this change was due to direct cultural influence from the

Roman world where extended inhumation had replaced the earlier cremation practices by

the 3rd century AD (O'Brien 1990; Raftery 1994). By the 5`h century AD, extended

inhumations are virtually all protected in graves with stone linings (Fig. 6.3), and these

143

Fig. 6.2. Double burial KnH 27/28 from Knowth, of extended unprotected inhumations (after Eogan 1974

are known as long cist burials to differentiate them from Bronze Age short cists (O'Brien

1990,2003; Raftery 1994; McGarry 2007). It has been widely accepted within the

literature that the move to create a stone lining within the burial is a discrete burial

practice and as such reflects a further transition in burial rite (O'Brien 1990,2003;

Raftery 1994; McGarry 2007). The wider use of radiocarbon dating places most of the

cist burials around the mid 5t' century AD, and by the 6`h century AD protected extended

supine inhumation is the most frequently identified burial practice in Ireland. Recent

excavations have revealed multiple burials of this kind at sites such as Laytown and Colp

West in County Meath, and in the far west at Owenbristy in Galway (Eogan forthcoming;

O'Brien forthcoming). A burial from Feerwore found within a mound was recently

radiocarbon dated by O'Brien (pers. comm 2009). Once assumed to be earlier Iron Age,

the burial has now been dated to the Late Iron Age to 429 - 591 cal AD (UB-7154

1541±32). (I am grateful to Dr O'Brien for permission to publish this ahead of her own forthcoming publication).

There is immense variability, however, in both

orientation and presentation of cist

inhumations, and it is unlikely that all date to

the same period. At Ninch in County Meath,

close to the villages of Betaghstown and

Laytown, two extended inhumations were

excavated where one was placed in a partial

cist and the other unprotected, but careful

excavation suggested contemporary burial

beneath a raised mound (Sweetman 1982).

Although the radiocarbon date for the

unprotected grave was earlier (UB 2425,

1820±115), 88-531 cal AD, than that for the

cist burial (GU 1453,1510± 65), 420-650 cal

AD, these were reconciled as charcoal samples

had been used for dating and they are likely to

.. ý ýr

ý. ý ,

Fig. 6.3 Protected extended inhumation in stone cist, BTS 11 Betaghstown (after O'Brien 1990)

144

be contemporary (Sweetman 1982: 68).

The wider use of radiocarbon dating is highlighting problems with the existing relative

method as more examples can be found of different forms of burial practice being used

contemporaneously. The burial sequences evidenced through radiocarbon dating at Ardnagross in County Westmeath demonstrated that a crouched female inhumation post- dated an extended unprotected inhumation, which in turn, post-dated other burials at the

site, notably one of which was within a stone lined cist (Archaeometry: datelist 22; 407).

At Clara in County Offaly and at Sonnagh Demesne in County Westmeath radiocarbon dated burials also suggest the continuing use of crouched burial practices which post-date

extended inhumations (Sikora pers. comm 2008), with the crouched burial dated to 420-

680 cal AD, far later than would have been suggested by the burial morphology alone. The earliest date received for burials from Betaghstown, was for an unprotected extended inhumation (BTS)UV 256-576 cal AD); the two crouched burials nearby (BTSXI 397-

635 cal AD and BTS XII 356-622 cal AD) were assumed to be the earliest at the site, but

they actually post-dated the extended burial, BTS XIV, and were contemporary with the

nearby cist burials (BTSXV 385-632 cal AD and BTSII 430-670 cal AD, Fig. 6.3), see Chapter 7. At Drumabarra a `lintel' grave believed to date to the 7t' century AD or later

(and therefore outside the remit of this work) returned a 5t' century AD date (430-470 cal AD) (Sikora and Cahill: forthcoming).

It has been suggested that the conversion to Christianity in Ireland led to protecting the

body within a stone lined grave (McGarry 2007). The absence of grave-goods and the frequency of an alignment in an east-west direction has been used in recent work to

support this (O'Brien 1990,2003, forthcoming; Raftery 1994; McGarry 2007). The

orientation of Irish cists varies greatly, however, and it cannot be said that they are all

aligned east-west; in any event we know that such an alignment was used in Roman

burials (Toynbee 1971; Philpott 1991). The orientation of the body east-west was a

practice that was adopted by the Romans from an earlier Greek tradition (although this

too was immensely variable, see Kurtz and Boardman 1971).

145

A passage from Tireachan's Life of St. Patrick highlights the difficulty with any

assumption that Christianity led to a change in the burial rite. The passage describes how

St. Patrick was walking past two burials when a voice called out to him and asked that he

move a cross, which had been inadvertently erected over the grave of a pagan. Christian

parables aside, what this passage suggests is that even by the 7th century AD, the time in

which it was written, it was not easy to distinguish between pagan and Christian graves

(O'Brien 1999, forthcoming).

6.4 `Intrusive' burials

The earliest example of `intrusive' burial in Iron Age Ireland is the cremation burial from

Fore in County Westmeath, now dated to 350-30 cal BC (McGarry 2007: 219). The

cremated remains were found within a finely decorated bronze bowl and the bowl has

been paralleled with one from Spettisbury Rings in Dorset (McGarry 2007: 219). The

bronze bowl from Spettisbury Rings is dated on typological grounds to the II century

AD, however, it is an exceptional find in this part of Britain where pre-Conquest

cauldrons of local type are more usually significantly larger (James & Rigby 1997). The

Roman cremation burial from Stoneyford, discussed in Chapter 5 (Fig 5.10) was found in

a small townland in County Kilkenny on the river Barrow, a large navigable river which

enters the sea at Waterford harbour. The urn itself was within a ̀ slab lined cist' within an

undisturbed ringfort (Bateson 1973; Warner 1976; Raftery 1994,2005; Bourke 1994).

The use of a box cist such as this is common in early post-conquest Britain, where such

vessels are often associated with pipes for libations (Toynbee 1971; Alcock 2006). The

Stoneyford cremation burial can be paralleled with a glass cremation urn from

Colchester, dated to the middle of the l century AD (Bourke 1994: 164; O'Brien 1990:

37). The burial from Dunadry on the Antrim coast contained a second glass cinerary urn,

also of Isings 67(a) and although the vessel is now lost it was recorded as being found

beneath ̀an exceedingly large mound' (Bourke 1994: 164; Raftery 1994; Bateson 1973).

146

Apart from the Stoneyford cremation there are two other reports of Roman burials found

in Ireland. At Bray Head, along the coast just south of Dublin, labourers working on an

access to a local demesne in the 1930s, accidentally uncovered some burials in stone

lined graves, which were removed so as to not hold up work (Bateson 1973). The

numbers found are uncertain, but at least two were buried in a correct Roman manner

with coins of Trajan and Hadrian to pay Charon's fee (O'Brien 1990; Raftery 1994). An

article in the Irish Times in 1966 suggested a further group of Roman burials were

uncovered when the railway cutting was extended through Bray Head. The article

reported that the additional burials contained two Roman soldiers complete with helmets,

armour and swords, and that the finds were sent to the British Museum in London

(Bateson 1973: 78). Although Bateson dismissed this last account as fanciful, we know

that the collyrium stamp from Golden was also sent to the British Museum and the `olla'

found off the coast in Sligo was sent to the Museum of Wales (Bateson 1973); further

research at the British Muesum may yet locate these fords.

Whilst changes clearly took place within burial practices in Ireland during the Later Iron

Age we need to be cautious about assuming that all of these date to the same period or

that all were directly influenced by those changes taking place in burial practices in the

Roman world (Toynbee 1971; O'Brien 1990, forthcoming; Raftery 1994). Cremation

burials in glass cinerary urns are rare even in Roman Britain which makes the burial from

Stoneyford in Ireland an exceptional find. There are a few glass cinerary urns that have

been reliably dated to the immediate post-conquest period and two from the site of Bishopsgate in London offer a parallel with that found at Stoneyford. The cinerary

vessels from Bishopsgate are of Isings form 67(a) and 50(a) respectively and they also

were accompanied by glass phials and both urns contained cremated remains. 1' century AD examples are clustered within a distribution in the south east of England and although

glass cinerary urns have been excavated at small rural settlements and small towns they

are much less frequently found at major urban centres (Philpott 1991: 27). There is

evidence for localised survival of cremation as the main burial rite at sites such as Kelvedon, Ipswich and High Rochester all of which may be as late as the 0 century AD.

One example of aP or 4th century burial with glass cinerary um is known from the site

147

of Attingham in Shropshire, dated by the coins described as being of the `lower Empire'

(Philpott 1991: 51 & 26). In parts of northern and western Britain, cremation has been

evidenced as late as the 4"' century AD, at sites such as Winterbourne Down, Winchester

and Owslebury, and these are likely to reflect a local tradition of cremation, rather than

an ̀ intrusive practice' (Philpott 1991: 52).

Although there is evidence of extended inhumation as early as the 1s` Century AD in the

Cotswolds in Britain, these reflect the local practices of the pre-conquest period in the

territory of the Dobunni (Whimster 1989 21; Philpott 1991: 6). By way of contrast to

Ireland, however, items of small personal grave goods appear to have been the more

common accompaniment for those buried in an extended position rather than in a

crouched form. Although the transition to the practice of extended inhumation took the best part of a century in Roman Britain, it had become the preferred burial rite within all

of the Roman Provinces by the mid P century (Toynbee (1971). Transition in Romano-

British burial practices is attested at the site of Mucking, in Essex where cemeteries

spanning the pre-conquest Iron Age through to the 4m or 5th century AD have been

excavated. At cemetery III PRIA flexed burials were excavated with small personal grave

goods and some imported Roman pottery, and at cemetery I the change in burial practice

to that of cremation is dated by associated diagnostic material to the late 1" century AD.

The site also had later burials with military belt fittings and these were classified as Germanic and dated to the 5th century AD (Todd 1998: 475). Crouched or flexed burials,

although less common by the 2d century AD continue to reflect a preference for local

burial practices in some areas and show similar variability in the positioning of the corpse

as they do in Ireland. At Allington Avenue in Dorchester and Poundbury cemeteries I and II, the local practice of crouched inhumation in the I" century AD appears to alter to that

of an extended position at the end of the 2"d century AD, but without evidence for the use

of cremation in the intermediate period (Philpott 1991: 54). More recently a Roman

cemetery was excavated just outside Dunstable in Bedforshire and although the burials

have not yet been radiocarbon dated or fully published they are believed to date to the 2°d

to 3'd century AD (bedfordshire. gov. uk/excavations/Dunstable: consulted 18/11/09) The

burials consisted of 24 cremations and seven inhumations, one crouched and one in a

148

prone position, and none of the inhumations had any grave goods within the burials.

Burial in the crouched position continues well into the 3d century in the south central and

south-western Britain, at sites such as Radspole Lane at Weymouth (3'' or 4t' century), Gambier Parry Lodge, Kingsholm (1" to 4th century) and Hinchwick in Gloucestershire

(buried in a cist with 3'' century pottery). In northern Britain crouched burial continued into the 3'' century AD even at sites situated next to Roman forts or towns. At Norton

nine crouched burials were excavated and are contemporary with six cremations all of

which were dated as late as the 3'' century AD. Nearby at Langton Villa two crouched burials were found to be contemporary with the period of intense occupation at the villa dated to 320-395 cal AD. Examples of 3`' and 4th century AD crouched burials are known from Welton Wold and Winterton on Humberside, (Philpott 1991: 57).

Parallels for burial in stone-lined cists are also plentiful in the Roman period especially in

northern Britain and in southern Scotland, only here these appear to be significantly

earlier than those in Ireland. There is also plentiful evidence of a local tradition of cist burial from Avon and Gloucestershire dating to the I' century AD (Tait & Jobey 1971:

63; Whimster 1989: 18 & 47; Philpott 1991: 61). 'Native' burial practice in north-eastern Britain from the Neolithic to the post-Roman period has been classified as crouched or flexed inhumation within stone lined cists, with no wholesale change consequent upon

the Roman occupation (Philpott 1991 citing Faull 1977). In Britain, crouched and flexed

inhumations are often accompanied by only a few personal items, and offer good diagnostic parallels with those found in Ireland. But these also occur with no

accompanying diagnostic grave goods, leaving open the possibility that in the absence of

absolute dates, these may fall anywhere within an Iron Age or Roman context. In his

assessment of the impact of Christianity in the province of Britannia, Philpott was

categoric in his assertion that the onset of Christianity left `no distinctive material

expression', and that burial practices throughout the 4th century AD appear to have

continued in line with existing local burial practices (Philpott 1991: 240).

The immense variability in body positioning, orientation, level of grave furnishing and morphology, militates against any temporal or chronological assumptions or

149

generalisation about burial practices in Roman Britain (Philpott 1991: 53). Our

assumptions about the formality of grave furnishings is similarly called into question

given the evidence from later Roman cemeteries such as Lankhills, Poundbury or Kelvedon (Alcock 2006; Jones and Mattingly 2002). Some of these burials lack even basic coffins and at Poundbury burials with stone lined cists post-dated earlier coffined burials (Toynbee 1971; Philpott 1982 & 1991; Baldwin 1995). Given this, caution is

needed when suggesting that ̀ Roman' influence can be inferred, on any, or all of these

changes in Ireland.

6.5 Sites chosen for scientific analysis

The classification of some these burials as ̀ intrusive' prompted the completion of a

survey of burial evidence and likely periods of migration into Ireland in prehistory

(Mallory and Ö'Donnabhiin 1998). Much of the skeletal material from older excavations

could not be located at the time of the survey and even though some of these burials were from historic sites of national importance, such as Tara and Knowth, these had not been

radiocarbon dated (O'Brien 1990; Mallory and Ö'Donnabhäin 1998). A successful licence application was made to the Keeper of the National Museum of Ireland to conduct

a scientific analysis using strontium and oxygen isotope analysis on five of the sites with

150

Bog of Allen

11-1 10 0

r RhW

k *Tara

p

Carburv Hill

r kiwi

and Rossnaree ý` Betaghstown

Over Delvin

Wd

RF4w LNky

Fig. 6.4 Topographical map of the greater County Meath environs (including parts of

north County Dublin and north County Kildare. Sites with `intrusive' burials chosen for

scientific analysis and 14C dating (map drawn by author)

`intrusive' burials. The missing skeletal material was located at various sites in Dublin

and the following were chosen for analysis: Knowth (Eogan 1974; 0' Brien 1990;

Raftery 1994), Tara, Räith na Senad (O'Brien 1990; Raftery 1994; Newman 2005),

Betaghstown (O'Brien 1990; Raftery 1994; Kelly unpublished notes NMI); Carbury Hill

Site B (Wilmot 1938) and Rossnaree (unpublished excavation). To ensure maximum

parity in relation to the underlying geology, the five sites were within a 20 mile radius of

the Hill of Tara (see Fig. 6.4). Following preliminary strontium and oxygen isotope

results for burials at Knowth, Tara and Carbury Hill, an application for funding was made

151

to the Royal Irish Academy Radiocarbon Dating programme and an award was granted

for four samples across these sites. The analysis and results are presented in the following

chapter and there follows an overview of what we know about the history and

excavations at each of the sites

6.6 Site B, Carbury Hill, County Kildare

Although Carbury is a relatively small hill its position within an otherwise flat rolling

landscape creates a sense of commanding importance and visibility. The geography of the

area consists of some cultivatable land but also many areas of bogland, and it is

positioned at the edge of a geographical area known as the Bog of Allen (see Fig. 6.5).

The Bog of Allen extends over a considerable distance from this point towards the River

Shannon, through the western part of County Kildare and includes most of County

Offaly.

Excavated in 1938, the sites at Carbury

Hill are important as they are

positioned close to the point at which

the River Boyne rises (known locally

as Trinity Well). The site has a folk

history associated with Nechtän

(Neptune), perhaps because of its link

to the River Boyne, but it is not

coloured by royal associations and

there does not appear to have been

settlement at the site before Norman

times (Wilmot 1938). At the site today

there are the remains of a Norman

motte and bailey, within which was

built a later Tower house, dating to

11ýýý111iUiü .........

... r ....,.,. iý: ý .,: ýý., ",,, .,,:.,. ,, ý,.. ,.,.

Fig. 6.5 Plan of Site B, Carbury Hill, showing position of cremations and inhumations (after Wilmot 1938, reproduced in Raftery 1994)

152

around the 15th century AD.

Along the top of the hill are three earthworks, (Site A, B and C) described within the

original excavation report as earthen circles (Wilmot 1938: 130; Raftery 1981 & 1994;

O'Brien 1990). The visibility of the burial mounds, placed as they are in a commanding

position on the brow of the hill, appears to have attracted others to use the sites as burial

grounds and each of burials appears to have been disturbed by later burials, suggesting

episodic use. Site B (see plan at Fig. 6.5. ) is significantly larger than the other two

earthworks with no visible openings (Wilmot 1938: 130,136,140).

An unprotected cremation deposit was found buried in a small pit within the centre of Site A and a further cremation had been inserted into the ditch. An iron penannular brooch (greatly eroded) along with a small jet spoon were also found close to the

cremation that had been inserted into the ditch and a parallel for the jet spoon has been

offered amongst similar spoons of a late Roman type (Wilmot 1938; Raftery B, 1981,

1994 & 2005). In the absence of modern excavation methods and as the site had been

disturbed, it remains unclear whether these objects and burials were within the same

contexts (Wilmot 1938: topographical files NMI; Raftery 1994). Most of the objects found were undiagnostic, such as a fragment of fused blue glass, some worked flint, an

unfinished stone spindle whorl, and an iron file.

In the centre of the enclosure at Site B (Fig. 6.5) four unprotected cremation deposits

were uncovered which appeared to have been disturbed by the burial of up to 14

unprotected extended inhumations. Within one of the cremation burials were fragments

of an iron pin and two iron rings. Site B also appears to have been used for episodic burials as the inhumations were themselves disturbed by the insertion of more extended burials. Small items were recovered as surface finds including flints, a circular stone disc,

a sherd of lead glazed medieval pottery (viewed in 2007). The only potentially diagnostic

item deposited with a burial was a small iron shears, which appeared to have been placed

under the left femur of a male burial (Skel I in Wilmot 1938). Parallels for such shears

are commonplace throughout the Iron Age, Roman and Anglo-Saxon periods in Britain.

153

All of the burials sampled for the strontium and oxygen isotope analysis were extended

inhumations and there was no evidence of an intermediate burial rite between the

cremation and extended inhumations. We can now offer a new radiocarbon date for CHI

(Skel 1 Wilmot 1938) found in Site B and this is 471-653 cal AD (UBA 11696:

1491±28). Given that only one burial has been dated, and that this burial appears to be

have disturbed previous burials, we should not assume that all of the rest are

contemporary and some could be earlier within the Later Iron Age.

6.7 Knowth, County Meath

ýý. ý.. I

,ý gis, , . ý..

ý_

5'-. ,,, r"" -- �ý mi

a

Fig. 6.6. An aerial view of the passage tombs of Newgrange and Knowth in the Boyne

Valley (Image courtesy of Knowth. corn)

Knowth is one of the three Neolithic passage tombs in the area known as Brugh na

Boinne, or the Boyne Valley; the others are Dowth and Newgrange. Knowth is similar in

154

size to its sister site of Newgrange (both are shown at Fig. 6.6. above) at around 80

metres in diameter, but it is distinguished by a series of 17 satellite tombs around the

circumference of the great mound (Eogan 1974). The passage tomb is unusual in having

an east and west entrance and it is aligned, as at Newgrange, with the winter solstice (Eogan 1974). The passage tomb complex at Knowth has been the subject of extensive

excavation and publication since the 1960s and the period under consideration here is

classified as its first phase. Both the material and burials were found within a double

ditched enclosure of the later Iron Age that was reconstructed and occupied in the Early

Historic period (Eogan 1974,1977: 70). Over 30 inhumations were found and fully

excavated, the majority of these were flexed or crouched inhumations, but some extended

unprotected inhumations were also excavated. The burials were found in discrete clusters

(see plan at Fig 6.7 below), with the majority of the Iron Age burials to the south side of

the enclosure. On the west side of the passage tomb, some remove from the rest, three

extended inhumations in cist lined graves (Burials 32-35) were excavated.

The orientation and presentation of the Iron Age inhumations is variable; some of the

inhumations are unprotected and flexed, others crouched, while others were found in

disturbed contexts due to the insertion of later burials. Burials KnH7, Burial KnH9,

Burial KnH 10 and Burial KnH21 have recently been radiocarbon dated and these are

flexed inhumations, and all but one (Burial KnH10) have been dated to the 1" and 2d

century AD; the outlier is earlier 170-90 cal BC (O'Brien 2006,2007; forthcoming).

155

T za

N ,, t9

27%21 20X

1,22- 24 X26

e 21i ýIIIIIITtT

ýý ý\,, , 111 11; / . `N\ffý \ý'. ̀ , \\1'1111, x;

1!,,,,, '/ /iý

/ý

,

iQ 1 `-\\"ý ýý\\ý.;. \"ý,, 111 _ýýilli 'ýý: '; ýýi///

ýiýj/-''ý

sit, .

13

EXTENT OF OISTURnEO AREA

ý- ý� / (11 ti \ ý-- Qxtandad burial -"-Crouched burial X 01 turbad burial

01a burial 'iii/iý ýýll llIffIt Ill l! ýý ýýýýý`f

11111111& 5 10 15 mI

I 1_04 7 X10 %e 4s OS

ax xI

Fig. 6.7. Plan of Iron Age and Early Historic burials around the great mound at Knowth

(a)? er Eogan 1974 and Raftery 1994) Burials chosen for sampling indicated with arrows

The extended unprotected inhumations (Burials KnH26-28) have not been radiocarbon dated and although they may represent a discrete group, a relative date has been proposed for them of the 7th or 8'h centuries based on the fact that one is a double burial (Burial

KnH27/28), a practice that was common in this period in Ireland (O'Brien forthcoming).

One decapitated extended inhumation was excavated from the ditch (Burial KnH6) and a further erratically positioned extended inhumation (Burial KnH5) was found to the north. Grave goods were found with the crouched and flexed inhumations and although some

can be paralleled with material from other Irish Iron Age sites, some are clearly non- local. There are no cremations from this site and the burial rite appears to move from

156

crouched/flexed positioning to extended supine. The extended inhumations found at Knowth have all been classified as ̀ intrusive' (Raftery 1994,2005: 39; O'Brien 1990).

Raftery's contribution to the fifth monograph on Knowth, which is currently in

preparation (I am grateful to the RIA and to Professor Raftery for making this available

to me in advance of publication), discusses a series of 14 crouched inhumations, which

were found outside the kerb of Passage Tomb 1. Ten of the crouched burials were buried

with an extraordinary number of bangles, beads and necklaces and it has proved difficult

to parallel the sheer number and positioning of the grave goods with any other examples

from Ireland (Raftery: forthcoming). Many of the smaller blue glass beads such as those

found around these examples are difficult to identify as specifically Irish or `exotic',

although it has been suggested that these may also represent non local individuals. Only

one of these burials No. 4 has been radiocarbon dated to 70 - 240 cal AD (GrN15,369

1830±30BP) and this date suggests that it is contemporary with the earliest burials

sampled in my scientific analysis detailed in the following chapter. A further programme

of strontium and oxygen isotope analysis would do much to establish if these are part of

the same population group as individuals sampled for my research. The sheer level of

adornment on these burials is exceptional in Ireland and there are no direct parallels for

the nature of such grave assemblages. The symbolic positioning of bangles and necklaces

of different materials within burials should be explored for parallels amongst burials

outside Ireland. Burials from sites such as Somozgyzil, in what was the Roman province

of Pannonia, are characterized by just such a practice of the symbolic positioning of

bangles, beads and necklaces on the arms of those buried (Swift 2000). There is at least

one Pannonian type fibula from County Galway, which is believed to be a contemporary

import (Raftery 2000; Bateson 1973). As has been argued throughout this work, it is

entirely possible that trade connections were established, before and during Roman times,

and direct contacts with even Eastern Europe in this period are not as unlikely as once

thought, given a recent find of an Irish type horse-bit in Romania (Raftery 2000).

Over the course of the excavations at Knowth, many inscriptions have been found within

the passage tombs themselves and although they are not of Iron Age date they provide an

157

interesting link with some of the scientific results. Five of these are ogham inscriptions,

in the later scholastic tradition than those discussed in Chapter 2, and they have proven

difficult to translate. The other inscriptions are ancient graffiti in majuscule insular script

(Swift 2008: 120; Byrne 2008: 90). One of the ogham inscriptions from the eastern tomb

records the name Tallurgg, a name which is very rare in any of the ancient manuscripts

from Ireland but is known from amongst the king lists of Pictland, an area where many

such scholastic oghams have also been found (Swift 2008: 114). Both types of

inscriptions found at Knowth are contemporary and have been dated on stylistic grounds

to the first part of the 8`h century AD (Swift 2008: 119). This link with Scotland is

important as some of the results of the isotope analysis suggest an origin for individuals

in either Yorkshire or the area of south-east Scotland, which was Pictland at that time,

and a new 14C date suggests that the inscriptions and this individual are contemporary.

6.8 Rossnaree, County Meath

Rossnaree translates as ̀ headland of the kings' and it is generally referred to as a ̀ royal

outpost' of Tara (Newman 2005: 371). The burial site sits directly opposite the Great

Mound of Knowth, on the south side of the River Boyne. The burials were uncovered

during excavations for machine gun-posts in 1942 and were brought to the attention of

the then Keeper, Joseph Raftery (Sikora and Cahill: forthcoming). Three adult females

were buried at the site, along with the remains of a baby of less than six months old. The

burials were greatly disturbed before Raftery could inspect them but one small item was

found close to the shoulder of Burial 1. This was a small piece of silver which had been

tinned and intricately engraved in an incised herringbone pattern. Given where it was

located on the burial, it is most likely to be a fragment of an earring. Items of jewellery

made from precious metals are very rare in the Irish Iron Age, as is tin-plating, but the

manufacture, design and quality of the piece are suggestive of an origin in the Roman

world. Despite a careful search of the archival material at the NM! it appears that no

faunal material was retained from the site at the time of excavation so a human tooth

from Burial 1 was used for analysis and given the site's proximity to Knowth it was

decided that an inter-site comparative analysis would be applied to the data from both.

158

The burial has been radiocarbon dated to 257-533 cal AD (Sikora and Cahill;

forthcoming)

6.8 Betaghstown, County Meath

The site of Betaghstown (in its anglicised form also known as Bettystown) lies just inland

from a stretch of coastline that has both shallow waters and sandy beaches, and it is

ideally positioned for access to the interior of Ireland through its proximity to the

immense River Boyne (Fig. 6.8). Betaghstown is just 12 miles north of the promontory

fort of Drumanagh discussed in detail in chapters 2 and 5, and it is within a hinterland of

all of the important sites mentioned throughout this study.

.ý wJ sJ

TM

ý/ "1ltel ý. KýJ

ýý

Fig. 6.8. Aerial view of the sandy coastline of County Meath. The site of Betaghstown, the

modern city of Drogheda, and the passage tombs are shown along with the curve of the

River Boyne: Newgrange and Knowth (marked with blue square). Image courtesy of Google Earth 2010

159

A Bronze Age barrow and cist were uncovered at Betaghstown in the 1970s and removed before the construction of the local housing estates. Human burials were uncovered

within a garden of one of the new houses and on excavation it became apparent that

several of the burials had been truncated with pipes for utilities (I am grateful to Mr Kelly for permission to use both his unpublished excavation notes and the files at the NMI). and Around 20 complete individuals were excavated overall, two were crouched burials, two

extended inhumations in cist graves and the rest were extended unprotected supine inhumations. Several of the burials were recorded as having personal grave goods,

consisting of an assortment of knives and belt buckles, but the notes and files do not identify items with specific burials. The only burial for which we have detailed

information is that of the female crouched inhumation (BTS )(I). This burial had an array

of diagnostic grave goods, good parallels for which can be found from Late Roman or

early Saxon contexts in Britain or continental Europe. These consisted of an iron belt

buckle (found on the middle of the body) and a pair of omega type penannular brooches

with an oval bronze disk (perforated at each end) found in the area of the upper chest. Along with the bronze disk was a small pendent also found in the neck area and described

as a miniature `stone axe' (Kelly: notes topographical files NMI). In Britain, although

examples of such miniature axes are known from pre-conquest sites, they are generally found within late Roman contexts but there also good parallels for them amongst Saxon

and Anglo-Saxon burials (Eckardt & Williams 2003). The penannular brooches are

simple forms with rolled back terminals and are known as either ̀ omega' brooches due to

their shape, or Fowler Type E (Fowler 1964). The British parallels for these brooches are, however, in Late Roman contexts, and yet on the Betaghstown burial the brooches appear

to have been worn as a pair with a plain brooch suspended between them, in a continental European style of the 5th century AD (Owen-Crocker 2002).

Traces of textiles were preserved on the back of all of the metalwork found with this

female burial and there were also traces of a fine `hair-net' on the lady's skull. All of the

textiles were analysed using high resolution scanning by a specialist in this area and it

was established that those on the back of the items from the upper body differed from that

on the belt buckle (Jorgensen 1984). The textile on the ̀ hair-net', the back of the brooch

160

and the pendent was a fine, gauze-like plaiting of horse hair, which can be paralleled with

similar use of a horse hair woven fabrics from two Dutch terpen sites, near Groningan

(Jorgensen 1984). The example from Betaghstown was, however, much finer than the

continental examples. The fabric on the back of the belt buckle was coarse in comparison

and it had been made using a Z/Z twill, which points to its manufacture in west and

central Europe, with the clearest parallels again from Dutch terpen settlements within the

5th century AD (Jorgensen 1984). It is clear from the work on the textiles that the

variation in thread and weaving reflects the lady's costume, which we know would have

consisted of a fine undergarment, over which a coarser peplos-style tunic would have

been worn (Owen-Crocker 2002). New research using strontium analysis on wool from

ancient textiles in Denmark has revealed that there are clear differences between the

strontium signatures on coarser thread and those of finer quality (Frei et al 2009).

Culturally or socially specific weaving techniques were used within migrant communities

who had settled in Late Roman Britain (Owen-Crocker 2004: 22, Dark 2002: 25) but

finer quality thread was being imported from outside local areas in Europe and demand

for these would have continued after communities had migrated to Britain and elsewhere (Frei et al 2009).

The female skeleton BTS XI whose grave goods were just discussed could not be traced

in the National Museum of Ireland, although it was the subject of an unpublished

osteological report in the early 1990s. The report was available to consult within the

topographical files at the NMI and it is referenced here, and in the discussion on

pathology in the following chapter, as Power 1990. The crouched male skeleton BTS XII

does not appear to have had any grave goods and it has provided one of the most exciting

results of the analysis, discussed in detail in the following chapter. Of the complete

skeletons, eleven were chosen for scientific analysis and seven have been radiocarbon dated to between the 3d and 6th century AD, and these are discussed in detail in the

following chapter.

161

6.10 Rditk na Senad, Hill of Tara, County Meath

Positioned on a limestone ridge with commanding views of up to 20 of the counties of Ireland, the Hill of Tara holds a special place in the hearts and minds of Irish people. It is

seen as the cultural capital of `Celtic' Ireland and is steeped in myths and legends. There

are over 50 monuments along the escarpment of the hill and these have been the focus of immense survey and study over many years. According to historic tradition, Tara was the

seat of sacral kingship in Ireland. The king of Tara held not just the principal kingship in

Ireland, but by inauguration through a ritual event known as the Fels Teamhra, he

became, essentially, the king of the world. The origin of the Gaelic name for Tara,

Teamhair, has recently been comprehensively reviewed and although in the past it was

translated as simply `place with a view', the origin of the name lies in the Greek temenus

and the Latin templum and is more correctly translated as a temple (Bhreathnach 2005).

An investigation of the occurrence of the word at other sites around Ireland confirmed

that all were associated with sacred enclosures or religious festivals (Bhreathnach 2005:

306).

After the destructive and futile search for the Ark of the Covenant, discussed in Chapter

2, the Hill of Tara was scheduled as a national monument under the new government of

the Republic, to prevent any further destruction. A remedial excavation was carried out at Rdith na Senad by Sean P Ö'Rfordäin in the 1950s in an attempt to assess the level of damage done to the site and in an attempt to map and record the undisturbed stratigraphy. Ö'Rfordäin had published his excavation of Lisnacaheragh (Garranes) ringfort in 1942

from which a remarkable assemblage of late Roman amphorae and slip wares were

recognised and correctly classified by him, and he had also published an important paper in 1947 in which he had catalogued all of the Roman material found in Ireland before this

date. His experience with Roman material was fortuitous as the excavation produced one

of the most important assemblages of Roman material ever found in Ireland (discussed in

the previous chapter).

162

The stratigraphical relationship of the extant monuments on Tara were explored as part of

an immense survey of the Hill of Tara published in 1997 to establish a relative

chronology for the monuments. Räith na Rig is the most visible monument in the

complex and it encloses within its ditch and bank the coinjoined ringforts known and Teach Cormac and an Forad. Rdith na Rig was excavated in 2002 and trenches were

placed in those areas that overlapped Ö'Riordäin's trenches from 1950. The proximity to

the original excavation was deliberate to allow the modern excavation team to assess and

validate Ö'Riordäin's work, which they did in fact confirm. A series of radiocarbon dates

confirmed that the most visible earthworks were all constructed in the 1st century BC to

1'` century AD as part of what appears to be a period of rebuilding and reoccupation at Tara (Newman 1997 & 2005). Survey also revealed a previously unrecognised enclosure

not visible on the surface. This was a massive timber structure that enclosed Räith na Senad, and overlay parts of Räith na Rig; a parallel for its construction can be found in

the previously mentioned site of Dün Äilinne in County Kildare and at Crüchäin in

Roscommon. There is now a series of new radiocarbon dates for Räith na Senad which

confirm that the Roman material was being used during the principal period of activity at

the site between the 2°d and 4th centuries AD (Grogan 2008).

The burials at Raith na Senad were clustered in two areas of the site, around a barrow in

the north-west and a further group towards the south (Grogan 2008: 95). A second ring ditch had been created to enclose the barrow and this contained five cremation deposits,

and one of these appeared to have disturbed a crouched inhumation. This was interpreted

as evidence that the rite of cremation has succeeded the (assumed) earlier rite of crouched inhumation at the site (O'Brien 1990). Furthermore it was argued that the burial evidence followed the development in burial customs in the Roman world, whereby the rite of

cremation was followed by extended supine inhumation (Toynbee 1971).

The most important aspect of the recently published work on Rdith na Senad is not

necessarily the catalogue of Roman material, which was to a great extent already known

from previous work (see Newman 2005), but the specialised reports contained within the

appendices. We now have an analysis of the wider landscape environment conducted on the charcoal samples from the site; a sequence of radiocarbon dates from a range of

163

contexts, although the inter-context stratigraphy is less certain; and importantly, a record

of the faunal evidence from the site, although the faunal material was destroyed after it

was recorded (Grogan 2008). The small finds from the site were assessed by Allason-

Jones and these were covered in detail in Chapter 5. Allason-Jones argued that the lack of

any items of precious jewellery was indicative of a ̀ settled, limited-income, domestic

group' occupying Tara at this time (Allason-Jones 2008: 107). What Allason-Jones did

not consider, or was perhaps unaware of, is that the total artefact assemblage (for both

local and non-local material but especially ceramics and glass) in the same period in

Ireland, is so limited that a collection such as this is significant in its own right and is

indicative of high-status (Grogan 2008: 96). There is evidence for a small rectangular

post-built structure on the site and a rather less distinct additional structure, which has

been described as a ̀ residential enclosure' (Grogan 2009: 55). There is little in the

material assemblage, however, which would point to permanent occupation at the site and it seems likely that these are temporary structures which were built for episodic activity at

the site.

The ceramics assemblage, as discussed in Chapter 5, had a very high proportion of Sainian ware vessels and also closed forms of Severn Valley ware and there are two

Dechelette 72 beakers, which are rare even in Roman Britain (Evans 2008: 123). The

high proportion of ceramic drinking vessels appears to correspond with a relatively high

number of very high quality glass drinking vessels from the site. Furthermore, a

considerable number of the metal objects were classified as brooches or parts of brooches, and although only one part of a fibula type brooch was recorded from the site,

there are multiple examples of early form penannular brooches, or parts thereof, the

parallels for which are in 3rd and 4th century Roman contexts in Britain. We need to be

c tious about using ̀ ritual' to distinguish between occupation sequences at sites, given

that the distinction in pre-history between sacred and profane often reflects more of an

archaeological construct than can be demonstrated by actual evidence (Bradley 2003).

yet the remarkable assemblage of drinking vessels is in keeping with what has been

described as ̀ power drinking' in Iron Age Europe. It has been argued that this was an

essential component of ritual feasting, inextricably linked to publically demonstrating the

164

right of an individual to rule (Arnold 2001). Within the old Irish tale, Baile in Scäil,

Ireland is personified as a young woman, who is asked to choose a new sovereign from

the various contenders for Kingship by offering him a gold cup of ale (Arnold 2001;

CELT consulted 20/4/2007). There is also evidence through the results of the isotopic

analysis to suggest that pigs may have been imported from Europe, as discussed in the following chapter. Parables and mythology aside, a religious function at Raith na Senad

is supported further by the report on the faunal assemblage, which is dominated by pig

(Grogan 2008: 12 1). Cattle and sheep, as explained in Chapter 4, dominate the faunal

remains at secular domestic sites, whereas pig is much less frequently evidenced

(McCormick 2002; Kerr 2007). All of the evidence at Raith na Senad points towards a `specialized' function and a religious activity is in keeping with both the revised

etymology of its name from the Latin Templum, and the later history of the site as the

sacred ritual inauguration centre for the High Kings of Ireland (Bhreathnach 2005).

All of the extended inhumations from the site were classified as ̀ intrusive' (O'Brien

1990; Raftery 1994). It was important for this study that, as at Knowth and Betaghstown,

samples of human teeth were taken from individuals buried using each type of burial

practice evident at the site in order that we could test for variation in the strontium and

oxygen istotope levels of both ̀ local' and ̀ intrusive' population groups (the rationale and

methodology are outlined in detail in the following chapter). As no prior scientific

analysis, nor radiocarbon dating, had been completed for the burials two extended

inhumations were selected for dating (Burial D and Burial E), the results of which are

shown in full in the following chapter.

165

Chapter 7 Strontium and Oxygen Isotope Analysis on `Intrusive' Burials

7.1 The archaeological background to the sampling strategy

The objective of the scientific analysis in this study was to test previous assumptions

about the ethnicity, identity and the origin of burials identified as ̀ intrusive' by O'Brien (1990), Raftery (1994), catalogued by Mallory and Ö'Donnabhäin (1998). The existing

classification and chronology was discussed in detail in the previous chapter and from

that discussion certain underlying assumptions became apparent. Within current

archaeological scholarship for the later Irish Iron Age in Ireland burials have been

classified as intrusive if they have been located at sites where ̀ intrusive' material, (i. e.

artefacts classified as exotica and imported into Ireland) have also been recovered, either from the burials themselves or from contexts at the same site. Similarly, where extended inhumations have been found alongside the assumed earlier burial practices of either

cremation or crouched inhumation these too have been classified as ̀ intrusive'. A further

problem with the assumptions inherent within the existing classification scheme is the

continued use of a relative dating system that is based on the burial practice itself, despite

the fact that the wider use of14C in recent years is directly challenging such

generalisations, as outlined in the previous chapter. This research was designed to

critically assess the basis around which the existing classification has been framed

through the application of strontium and oxygen isotope analysis on samples of human

and faunal skeletal material from `intrusive' burials, and to test the reliability of the

existing relative dating methods for burials through a series of new 14C dates on some of those chosen for additional scientific analysis. A successful application was made to the Royal Irish Academy Radiocarbon Dating Awards scheme and four important new dates

are presented later in the discussion.

As outlined in the previous chapter, samples were taken under licence from human and faunal remains at the five sites chosen for analysis using strontium and oxygen isotopes.

In order to establish a comparative data set, the research was focussed on sites from an

area of similar geology within a 20 mile radius of the Hill of Tara, which allowed

strontium and oxygen isotope values to be analysed on an inter and infra site basis.

166

Samples were collected, where possible, from each type of burial practice evident at each

site. The rationale for this was to test, on an intra and inter site basis, the validity of the

suggestion that local and non-local population groups were using discrete burial rites and to see if `intrusive' burial types were represented by non-local strontium (Sr) and oxygen (0) signatures. In larger Sr and 0 isotope studies, it is often apparent that there is a `local' group within a narrow range of Sr isotope values, representing the majority of individuals, with smaller numbers of immigrants with Sr isotope ratios that fall outside

this range. With a small study such as this, based as it is on a selected sample of burials

that were believed to be foreigners, and without the surety of faunal remains for each and

every burial context, a level of pragmatism is needed in extrapolating any generalisations

across Ireland as a whole. The data analysis is presented as a regional study, in an area of Ireland with the highest concentration of both imported material and burials which have

up to now been classified as ̀ intrusive'.

7.2 Overview of strontium and oxygen isotope research

Tooth enamel is the most highly mineralized tissue in our bodies and whilst it is forming

it records the chemical signatures of the geological and climatic environment in which we

were living at that time. Once mineralization is complete (usually around 12 years of age) the enamel does not alter or change during life or through diagenesis in a depositional

context (Budd et al 2003; Evans et a12004; Montgomery et a! 2005). Measuring the

ratios of the various isotopes present in the tooth enamel can offer an indicator of an individual's place of origin and has been used as a tool within archaeological science to

track population movement or migration (e. g. Dupras 2001; Schweissing & Grupe 2003;

Budd et al 2003; Bentley et al 2005; Montgomery et al 2005; Evans et al 2006; Daux et al 2008; Chenery et at 2010). The strontium in teeth is derived from the geological

environment taken up through everything we eat and drink (water, flora and fauna) and it

consists of three stable isotopes and one radiogenic isotope. It is the ratio of the

radiogenic strontium 87Sr assessed in relation to the stable 86Sr (shown as 87Sr/86 Sr),

which varies with changing geology. Expected values for strontium isotopes can be

estimated from knowledge of the geology. Young rock, or rock that has less rubidium,

167

tends to give lower 87Sr/86Sr from around 87Sr/86 Sr c. 0.706 whereas older rubidium rich,

and it follows more radiogenic rock tends to give much higher ratios of 87Sr/"Sr > 0.715

(Evans et al 2006). There is a difficulty in characterising a single geology even within a

relatively small geographical area, given that the soil may be derived from various

sources or have been moved through natural geological/geomorphological events or even

man-made processes. Methods are being developed to get around this problem and in

recent research a parallel chemical analysis of soil and vegetative samples taken at the

time of excavation has proved immensely useful (Chenery et al 2010). The research for

this study was, however, dependent on the extent to which both skeletal and faunal

remains were conserved within the archives at the museum in Dublin and limited to that

which was still available. At the time of the original publication of these sites faunal

material was identified as having come from the burial contexts and it was interpreted as forming the basis of funerary feasting or food offerings (O'Brien 1990). It proved difficult to locate suitable animal samples for some of the sites, as much material was not

retained post excavation or simply could not be located, and that which was left was from

larger animals such as cattle or horse. As there can be considerable variation in even localised geology, and as enamel mineralization is an incremental process, larger animals

grazed over greater distances from settlement or burial sites will record the variation in

geological and chemical composition of the land in which they have been raised or

through which they have been moved. As a result cattle and horse remains may not offer

an exclusively localised signature (Bentley et al 2005). This is, however, a problem inherent in studies from sites excavated before more modem excavation techniques. A

localised signature can be derived from small animals, preferably rodents, as these are

unlikely to have moved far beyond the immediate site area, and where available sheep,

pig and a cat sample were used (Bentley et al 2005). As this was the first analysis

conducted on samples from multiple sites in Ireland it was decided that the larger animals

would also be sampled to collate as much data as possible from the available material as a

start point from which a new comparative data set could be created.

The oxygen in teeth is derived from drinking water sources and the mass differences of the isotopes (180,170 and 160) leads to relative enrichment or depletion (fractionation) by

168

processes such as evaporation and precipitation that will vary according to climate, latitude, longitude, and altitude. Although in warmer regions there are greater levels of 180 in rainwater (referred to as enrichment) and it follows colder climates or higher

altitudes have significantly less (referred to as depletion), the ratio of 180 is primarily dictated by precipitation levels in any geographical area, (see Budd et al 2003 for a

comprehensive overview). Conventionally, the differences in oxygen isotope ratios are

expressed relative to a standard as 5180. The 0 isotope ratios in meteoric (ground) water

have been mapped across Europe and are available to consult on-line at

b=: //wateriso. eas. 12urdue. edu and httn: //www. iaea. org.

The 180 isotope ratio within the carbonate of tooth enamel is derived from the local

meteoric water sources but the processes of metabolic fractionation is, however, species

dependent and will vary across samples. The mineralization of enamel is also, as

previously mentioned, susceptible to the seasonality of precipitation, which may result in

higher or lower ratios than could otherwise have been predicted (Bentley et al 2005).

Scholars working within this field of scientific research have devised different conversion

equations by which the human and faunal samples (5180c) are converted back to drinking

water equivalents (S'gOd,, ) and these have been published in full by Longinelli et al

(1984), Levinson et al (1987), Luz et al (1984), and Daux et al (2008). The 5180 on the

simples used in this research was measured on the carbonate of the tooth enamel (5180 )

relative to PDB corrected to 818Op in line with lacumin et al (1996) relative to standard

mean ocean water (SMOW), and its meteoric water equivalent was converted to modem drinking water values (818Od,, ) using the equations proposed by Longinelli (1984) for

humans and cattle. The pig sample from Tara was modeled using Bentley's (2005)

variation of Longinelli (1984). In a recent article Chenery et al (2010) have argued that

the equation proposed by Levinson et al (1987) provides the best fit between expected 8180, , values and actual results.

169

* Daux O Longinelli 3 x Levinson e Luz et al

1 0

-1 -2

9 -3 -4 -5 -6 -7 -8 -9

-10 -11

X

X

X

xxO

XOp

x Xx

ox xXXXXX e

XO pxx XX

XXXX Ox XXX

Xx 28 x26xx x! X00 0

0Ö e s!

öi e 60 Ax

xe $ 6806o 0o0 00° aoog 0.0 6o 00 0 °o ö

x8og

Human samples from Ireland (n=40)

Expected range Ireland S1e0e� _ -4.4 to -7.3°/°0

Fig. 7.1 Measured 6180c converted to meteoric water equivalent 618Odw using conversion

equations published by Longinelli et al (1984), Levinson et al (1987), Luz eta! (1984)

and Daux et al (2008). Expected 6180iw values for Ireland are based on oxygen isotope

values for drinking water published by Diefendorf and Patterson (2005).

In order to test the validity of this assertion all of the data collected for this research were

recalculated using each of the equations now published. When the equation published by

Levinson et al (1987) is applied to the data from sites in Ireland the results produce

significantly more enriched values than any of the other models (see Fig 7.1). This

finding is in keeping with the work of Daux et al (2008) who also modeled his own data

using each of the published equations and found the best fit to be with that of Longinelli

(1984). As is clear from Fig. 7.1, the enriched values generated by the application of the

Levinson equation results in very few of samples having oxygen isotope signatures from

Ireland or Britain, or indeed most of continental Europe. Using the survey data compiled

by Diefendorf and Patterson (2005) for the spatial distribution of 8I8Odw isotopes in

170

drinking water sources across Ireland, the generalised range is expected to fall (south-

west to north-east) between ö180d� = -4.4 and 8180, � = -7.3 with localised isotope

ratios for counties Meath and Kildare varying respectively between 8'8Od,,,, = -6.3 and

6180d� = -6.9. These figures are qualified by the variation that may occur across a range

of water sources, including bore-holes, lacustrine and surface water (including variances

for differential altitude). Although all of the samples analysed in this research are from

burials that have been classified as intrusive in the past, the application of the Longenelli

et al (1984) equation results in some of the human samples fitting with results derived

from the faunal samples at the same site, and these also correlate with the predicted local

6180 based on Diefendorf and Patterson (2005). This suggests that the Levinson et al

(1987) equation provides a poor fit for the data from Ireland overall and this is especially

true once the data is interpreted through the archaeological evidence, as discussed in

detail later.

The application of both oxygen and strontium analysis on selected human and faunal

samples has been used to successfully demonstrate population mobility in prehistory (e. g.

Schweissing & Grupe 2003; Budd et al 2003; Bentley et al 2005; Montgomery et al

2005; Evans et al 2006). As will be discussed in detail there are some results from the

Irish samples of which we can say with certainty that they fit neither a strontium nor an

oxygen ratio that accords with this region of Ireland, and some that are certainly not from

Ireland. Many of the b'80dw results for the sites sampled are, however, similar to those

recorded in mainland Europe and the Mediterranean. The map of Europe and north Africa

(shown at Fig. 7.2) details the annual mean average 6180 from precipitation recorded at

water isotope research stations (source: www. iaea. org). In the absence of a

comprehensive data set for both strontium and oxygen isotope ratios across Ireland,

Britain and Europe, a lex parsimoniae explanation and interpretation is given for those

individuals whose strontium and oxygen isotope ratios is consistent with an origin in

Ireland. As methods will continue to be refined in future research and for the purposes of

transparency, all of the raw data for both the oxygen and strontium isotope analysis are detailed in Appendix C. The methods used to conduct the analysis at the University of

171

Bristol and the Research Laboratory for Archaeology, University of Oxford, are detailed

in full below.

-.

°o .8_ Y

y-A

10

r

Fig. 7.2 Annual mean average 6180 SMOW isotope ratios recorded by the Global

Network for Isotopes in Precipitation (GNIP), image courtesy of tinv . iaea. org. Note: as

this is an average there will be some variability within and between yearly ratios

Known localized 6'80 variation added by author: 04518

Odw = -8 to -9 0818

Odw - -7 to -8

0 818Odw = -6 to -7

Methodology: strontium isotopes

The enamel surface of an intact molar was first cleaned using a dental burr and hand drill.

A wedge of enamel (ca. 0.5mm wide, 1 mm deep) representing the complete growth axis

of the tooth enamel was removed using a flexible diamond-impregnated dental disk.

172

Thus, the strontium isotopes represent the mean value of the whole enamel growth

period. Any dentine adhering to the enamel section was then removed using a dental burr

and the remaining enamel sample cleaned in an ultrasonic bath. The whole tooth enamel

section was dissolved in 3ml 7N HNO3. The supernatant was dried and redissolved in 3N

HNO3. An aliquot of this solution was removed, representing c. 3 mg of solid enamel and

made up to 0.5 ml with 3N HNO3 to be loaded onto ion exchange columns. The strontium

was separated using standard ion exchange chromatography using 70 µl of Eichrom Sr

Spec resin (50-100 µm). Samples were loaded in to 0.5ml 3N HNO3 and washed with 4N

HNO3. Strontium was eluted in 1.5m1 MilliQ water. The elutant was dried down and

loaded using a few W 10% HNO3 onto rhenium filaments preconditioned with 1- pl

TaC15 solution and 10 10% H3PO4. Isotope ratios were measured using a

ThermoFinnegan Triton Thermal Ionization Mass Spectrometer. All strontium ratios have

been corrected to a value for NBS 987 of 0.710248 and the reproducibility did not exceed

(0.000013 (2a)

Methodology: oxygen isotopes

Each tooth was abraded with a tungsten carbide dental burr to remove any calculus and

particulates adhering to the surface. A longitudinal sample, representing the whole

enamel growth period, was removed as a V-shaped wedge using a flexible diamond

edged rotary dental saw. Enamel was cleaned of all adhering dentine using a burr to

produce a sample of core enamel tissue. The samples were then transferred to clean

labeled glass vials, weighed, rinsed with MilliQ water, ultrasonicated for 3 minutes and

liquid pipetted off. The washing was repeated three times and the samples dried on a

hotplate. Each sample was transferred to a clean agate mortar and ground to a fine

powder under methanol. After evaporation of the methanol, the powder was collected,

weighed and transferred to prepared mini-spin tubes and sent to the Research Laboratory

for Archaeology, Oxford University, for analysis. Samples for analysis were treated with 0.5m1 of 3% sodium hypochlorite (NaOCI) solution in 1.5 ml centrifuge tubes for 24

hours to oxidise any organic matter then rinsed 5 times in deionised water in a centrifuge. Samples were then treated for 24 hours with 0.5m1 of 1M calcium acetate buffered with

173

acetic acid to pH 4.9 to remove carbonate contaminants in non-carbonate lattice sites (e. g.

pore filling cements), rinsed 5 times with distilled water and oven dried at 60°C overnight

(as per Koch et al, 1997). Oxygen and carbon stable isotopic results were obtained in the

Department of Earth Sciences, University of Oxford, using a VG Isogas Prism II mass

spectrometer with an on line VG Isocarb common acid bath preparation system. Each

sample was reacted with purified phosphoric acid (H3PO4) at 90°C, with the liberated

carbon dioxide being frozen into a liquid nitrogen cooled `cold finger' apparatus prior to

admission to the mass spectrometer. Both oxygen and carbon isotopic ratios are reported

relative to the VPDB international standard. Calibration was against the in-house NOCZ

Carrara Marble standard with a reproducibility of better than 0.2 per mil (0.02%).

.. /...

kL

I

w. ric aw

ýý

__J oxygen Isotope Values for Modern UK Drinking Water

Urgen Fotopcs Values for Modern European Drinking '. Vater

e. Kwn

u ýOOOr

Fig. 7.3 Oxygen isotope values for modern drinking water (6180? %V): left, Ireland and the

UK and right with continental Europe. Images courtesy of NERC Isotope Geosciences

Laboratory

174

III LU

Fig. 7.4 Oxygen isotope ratios fr om drinking water sources in Ireland (8'8Odt, ). Redrawn

by author, based on survey data published by Diefendorf and Patterson (2005)

175

Generalized 6'-0,,,

Sites Lab. No. No. Faunal Note

Ref Human Samples

Samples

Carbury Hill, Co. Kildare CH 6 1 CHI sent for AMS dating

under RIA award

Knowth, Co. Meath KnH 9 1 Cist No. KnH32 sent for

AMS dating under RIA

award

Rossnaree, Co. Meath RnR 1 0 (none Already RC dated

available)

Räith na Senad, Tara, Co. RnS 7 1 RnS B and RnS D sent for

Meath AMS dating under RIA

award

Betaghstown, Co. Meath BtS 11 2 Already RC dated

High Island, Co. Galway HIs 2 1 11 Century AD

For Heritage Council Ireland

Truska, Co. Mayo TsK 2 1 11 Century AD

For Keeper of NMI

Ballymacaward, Co. BMc 2 1 5 Century AD

Donegal For Dr E O'Brien

Table 7.1 Details of Human and Faunal Samples Collected under Licence to Export

and associated Licence to Alter (No. 3983), granted by the National Museum of Ireland

(NI W. Note: additional analyses were completed at the request of staff at the NMI and Heritage Council of Ireland the reports for which are not yet published These did not form part of this research project but the additional data are included here for

comparative purposes.

176

'000i R. Bo

am tow

loossna'ec

0

.. i, ý rw fýr. S "

5 IL. I. urt

Oi do'; icia-

ordovician-Silurian

Silurian Overstep Sequence

Carboniferous iMilveron G! oupl

Carboniferous (Silesian:

Carboniferous (Fingal Group:

Fig. 7.5 The geological sequences for the greater County Meath environs, (includes

parts of north County Dublin and north County Kildare). After McConnell et al (2001)

Geological Survey of Ireland, Map no. 13, redrawn by author.

177

Fig. 7.6 Topographical map of the greater County Meath environs (includes parts of

north County Dublin and north County Kildare). After Newman et al (2005), redrawn by

author.

7.3 Carbury Hill, County Kildare

Geological Overview

The geological conditions were assessed with reference to the Geological Survey of Ireland Map no. 16, Kildare-Wicklow. The geology of North West Kildare within which Carbury Hill sits is predominantly Silurian-Ordivician geology (Fig. 7.5). The hill itself

178

sits in a commanding position within an otherwise flat countryside (Fig. 7.6). We would

expect 87Sr/6Sr = 0.7083 for the geology of the Carboniferous period; 87Sr/16 Sr = 0.7088

for Silurian and 87Sr/86Sr = 0.7091 for Ordovician geology (McArthur et x12001). A

cow's tooth from amongst the animal bones stored with the human skeleton (CH2) was

used for the baseline analysis.

Pathology of remains and teeth

Wilmot uncovered four cremations, at Site B and two of these were disturbed by the

inhumation burials. His published account suggests that cremation material was scattered

throughout the site, and he explained that this was probably due to disturbance caused by

the burial of the inhumations at a later date (1938: 136). Once the skeletal material was

located, it proved to be in a very poor state of preservation and this may be in part due its

storage, unwashed and unsorted in the habitat drawers of the National Museum of

Ireland. Some of the skeletal material appeared to be spongy, suggestive of exposure to

high levels of moisture in its current position. No previous assessment of the material

appears to have taken place as the finds tins, clearly dating to the 1930s, were found in

association with some of the human skeletal material. It was apparent that although the

boxes of skeletal material were labelled as single individuals, more than one individual

was contained within some of them. Samples were taken from boxes that contained

individual named burials that could be cross-referenced and identified within Wilmot's

assessment. All of the burials sampled were found within the central section of the

enclosure of Site B. Six human teeth, five male and one female, and one cow's tooth

were sampled and the burials were given new laboratory numbers and these, along with

Wilmot's original burial numbers, are shown at table 7.2.

179

Lab Ref NMI Ref Burial Type 57Srfl6Sr b'"OO 6 Odw pa N '4C Dates

show Calibrated to 20

CHI Skel. I Ext. Adult Male 0.708435±. 000003 27.64 -5.87 471-653 cal AD

CH2 Skel. IV Ext. Adult Male 0.709478±. 000002 28.17 -5.06

CH3 Skel. VI Ext. Adult Female 0.708862±. 000003 28.59 -4.43

CH4 Skel. VII Ext. Adult Male 0.708481±. 000003 27.60 -5.91

CH5 Skel. VIII ExtAdult Male 0.708497±. 000008 27.34 -6.31

CH6 Skel. X Ext. Elderly Female 0.709518±. 000003 27.90 -5.47

CHCow CH Cow 0.709318

Table 7.2 Results for Carbury Hill, Site B.

Errors are given at 2a oxygen results reported as drinking water equivalents, typical error 0.2

per mil. Radiocarbon dates have been calibrated using OxCal 4.1 and the IntCatO9 calibration. . * This date was funded under an award from the RIA and is shown in full within the discussion.

Analysis

The a7Sr/*Sr for these individuals appears to fall into three groups. Four of these are

within a range of "Sr/"Sr = 0.7084 to 0.7088, and two individuals, CH2 and CH6, along

with the faunal sample are within the range 97Sr/86Sr = 0.7094 to 0.7095. The results for

four of the burials are within 8'8O jw = -5- to -6 range which is much more enriched than

would be expected for this part of Ireland which is predicted to be 8'8Od� = -6 to -7.

Although the cow's tooth came from the same drawer as skeleton CH4, there was no

record of its finding nor any mention of other faunal material found in Wilmot's

published report or in his original notes (consulted at the NMI: April 2008). One female

burial with a slightly higher result, CH3 87Srfl Sr = 0.7088, has the most enriched S18OdW

value = -4.43 and this suggests an origin which is different to that of the rest of the main

group. The 87R/ 6R results for burials CH1, CH4 and CH5, are identical, 87Sr/6Sr =

(0.7084,0.7084,0.7084). There is a variation in the 8180� for CH5 which is more

180

depleted than the rest of the group 8' 8Odw = -6.31, but the 8' 80d,,, for CH 1 and CH4 is

more enriched at 818Od� = (-5.87, -5.91). The Sr results for four of the six samples are

consistent with, or just above, the expected Sr for the local geology and the oxygen

results fall just short of the expected range for the region. Although Carbury hill is only 110m above sea level it is the only elevated land in an otherwise flat landscape, bordered

to the north and east by the plains of Meath and to the west and south-west by the

immense Bog of Allen. Given its topographical positioning, precipitation may have been

greater within any occupation area local to the site with a resultant increase in levels of 8"0 leading to the enrichment of the oxygen results. Similar oxygen levels are, however,

most frequently found in the west of Ireland (Diefendorf and Patterson 2005) and are

common outside Ireland in the area around the Atlantic sea board. We have now

comparative data for both strontium and oxygen from burial sites in the west of Ireland

derived from four burials at two sites in counties Galway and Mayo completed for a

separate project at the request of the Department of Heritage and National Museum of Ireland. The complex geology of the west of Ireland produced more elevated Sr with the

results that burial and faunal samples fall between 87Sr/ Sr = 0.7092 to 0.7094. Two of

the individuals at Carbury Hill (CH2 and CH6) have comparable oxygen and strontium

and it is very likely that they originated in the west of Ireland. Burial CH5 has a Sr result

consistent with an origin on carboniferous limestone and the oxygen result is in keeping

with the nearby local range for County Meath, suggesting a likely origin in the area. CH3

is likely to have had an origin in an area with significantly more enriched oxygen values

and there is only one small area of Ireland where similar results were obtained through

survey samples and this is in the extreme south-west of Ireland, on the coastline of County Kerry (see Fig. 7.3). No comparative data for Sr is available for the coast of Kerry, although a geological assessment based on published work indicates that the

primary geology of both the Iveragh and Beara peninsulas in County Kerry is old red

sandstone of the Devonian/Silurian age (Morris 1980; 151) and the result for CH3 at 87Sr/"Sr = 0.7088 is in keeping with the predicted range for the local geology.

181

Summary of Results

-4.00 Lnz ZZ

LLl > -4.50 N zf

J OC Co O

-5.00 A

B SOUTH-WEST COAST OF IRELAND

-5.50

0-6.00 00 V. i

-6.50

-7.00

-7.50

-8.00 ''1 0.7080 0.7085 0.7090 0.7095 0.7100

87, o_ 186, &_

IRELAND AND THE UK

mir/ -Or

Fig 7.7. Carbury Hill, showing 8'8O j plotted against results for 87Sr? 6Sr: expected local

range for both strontium and oxygen plotted as highlighted rectangle. Arrows indicate

geographic region consistent with depletion level of 61804,,,.

The analysis suggests that we have burials from at least two population groups at Carbury

Hill (Group A and B, see Fig. 7.7), during the period in which Site B was used for

extended inhumations. This fits well with the observations noted by Wilmot during his

excavation that several of the extended burials had disturbed earlier extended

inhumations and one of the cremations, but that no cremations appeared to have been

inserted after the inhumations (Wilmot 1938: 136). All of the burials sampled were

extended unprotected inhumations, there were no crouched burials found at any of the

182

three sites excavated by Wilmot in 1938. All of the burials were in simple shallow graves

with no evidence of any protection or lining within the graves themselves, and all appear

to have been orientated northeast-southwest, with heads to the south-west. The burials

were classified as intrusive for two reasons: the transition from the (presumed) earlier use

of cremation at the same site and the finds which were paralleled in burial contexts

outside Ireland (Wilmot 1938; O'Brien 1990; Raftery 1994). The traditional relative

burial chronology suggested that these burials dated to the 4th century AD, before the

developed practice of burial in stone lined cists, believed to have been introduced into

Ireland in the 5th century AD.

The strontium and oxygen isotope analysis has demonstrated two or three of the

individuals (CH4, CH5 and probably CH1) are consistent with an origin close to the area

in which they died. At least two individuals (CH2, CH6) are best explained by an origin

in the west of Ireland and one individual (CH3) originated near the coastline of Kerry.

Although these oxygen ratios are recorded in areas outside Ireland around the Atlantic

seaboard (see Fig. 7. x) all of the burials at Carbury Hill can be offered an origin within

Ireland and they should no longer be treated as intrusive within the archaeological

literature. The radiocarbon date for Burial CH1, buried with a small iron shears, is 471-

653 cal AD (2a) (UBA 11696: 1491±28). The date falls within the expected range for

cist burials rather than the (presumed) earlier unprotected extended inhumations and this

date supports the argument set out in Chapter 6 that both protected and unprotected

inhumations may not represent discrete burial rites but are more likely to have been

contemporary practices.

7.4 Knowth, County Meath

Geological Overview

The age of the underlying geology was assessed using published data from the

Geological Survey of Ireland Sheet No. 13 Meath, and referenced against the expected

maxima and minima for distinctive geological conditions based on MacArthur et al

183

(2001). The Boyne valley lies within an area of predominantly carboniferous limestone

that is bounded by areas of Lower Palaeozoic rock. These lie to the North of the site (the

Louth Massif) and to the south (the Balbriggan Massif) and intersect with the

carboniferous belt north of the Boyne Valley and the site of Knowth in the region of the

River Mattock (Geological Survey of Ireland; Stout 2002; Fig. 7.5 & 7.6). As pointed

out in recent work the area under review is essentially a carboniferous syncline which is

clearly defined topographically, the principal core of which is formed of Namurian shales

and sandstones which lie stratigraphically above the Dinantian limestone (Stout 2002: 3).

Using the data from the Geological survey we can be more specific about the geological

age of the nearby Lower Palaeozoic rock; these are two belts of Silurian/Ordovician age

rock consisting of black mudstone and quartzose greywacke and thinly bedded siltstone

and sandstone. There is also one known igneous intrusion, known as the Drogheda

granite, although it is appears to be very localized and some distance from the site of Knowth. We would expect 87Srfl Sr =<0.7083 for the geology of the carboniferous

period; 87Sr/86Sr = 0.7088 for Silurian and 87Sr/86Sr = 0.7091 for Ordovician geology

(McArthur et al 2001).

Lab Ref Ref Burial Type 'S ^r 8150C b Od� 14C Dates

smow Calibrated to 2a

per (5 PRIA No. 5 Ext (Female) 0.711309±. 000005 26.99 -6.83 Undated

KnH6 PRIA No. 6 Ext (Male) 0.713901±. 000005 25.69 -8.77 660 -788 cal AD (new no. 14)

j9 PRIA No. 9 Crouched 0.709489&. 000004 26.88 -6.99 175-50 cal BC (new no. 10) (Female)

KnH 10 PRIA No. 10 Burial (10 & 11): 0.709434±000004 26.55 -7.48 40 cal BC - cal AD 121 (new no. 8) Flexed (male)

KnH11 PRIA No. 11 Burial (10 & 11): 0.709369±. 000003 26.74 -7.20 40 cal BC - cal AD 121 (new no. 9) Flexed (male)

KnH27 PRIA No. 27 Extended 0.709072±. 000003 26.75 -7.18 Undated

KnH32 PRIA No. 32 Cist Burial 0.710747±. 000005 26.71 -7.25 710-891 cal AD* KnH33 PRIA No. 33 Cist Burial 0.710847±. 000003 27.21 -6.50 Undated KnH Cat KnH Cat 0.710491±. 000003 27.87 -6.64

Table 7.3: Detail of samples and laboratory references for Knowth. Errors are given at 2a; oxygen results reported as drinking water equivalents, typical error 0.2 per mil;

184

Radiocarbon dates calibrated using OxCal 4.1 and the IntCatO9 calibration. 1 am

grateful to Dr Clearyfor permission to use the radiocarbon dates presented here. *This

date was funded under an award from the RIA and is shown in full within the discussion.

Pathology of remains and teeth

The skeletal and faunal remains from some of the burials are located at University

College Dublin at Earlsfort Terrace. All of the material had been sorted and classified during post-excavational assessment. Access and permission to take samples, was granted by Professor Eogan and his colleague Dr Kerri Cleary, who acted as curatorial supervisor during my visits. The preservation of the material was good, although some had been

removed for osteological analysis in advance of a new monograph. Eight human samples

and one faunal sample were analysed for Knowth: three crouched or flexed (one a double

burial), three extended unprotected inhumations and two cist burials. The traditional

chronology would place the crouched burials as earliest (1' century AD), followed by

unprotected extended inhumation (3`h century AD) and finally cist burials (from the 4t'

century AD). Burial KnH10/11 a double flexed/crouched burial had an unusually rich

assemblage of grave goods and was thought intrusive. Burial KnH5, Burial KnH6 and Burial KnH27 were also considered to be intrusive as they were unprotected extended inhumations and were evidenced at a site where earlier (crouched) and later (cist) burials

were also excavated (O'Brien 1990). Burial KnH9, a crouched burial was thought to be

of local origin as were the cist burials KnH32 and KnH33 and these were sampled to test if the strontium and oxygen isotope analysis could identify a local range. The original

published PRIA burial references were retained (the new references are also shown) but

the samples were given a Bristol laboratory prefix and all of these are shown at Table 7.3.

Analysis

The results are shown in Fig. 7.8 and these fall into three distinct groups so it is highly

unlikely that the burials represent only locals. Burial KnH5 and Burial KnH6 were

chosen for sampling as they were extended inhumations thought to date to the Iron Age

185

or Early Historic period (O'Brien 1990; Raftery 1994). At the point at which this

research was started no radiocarbon dates had been published for Knowth and although Burial KnH5 is undated, the dates for Burial KnH6 were forwarded to me by the editor at the Royal Irish Academy in advance of a new monograph. The date for Burial KnH6 is

660-778 cal AD (2a) and although this puts it outside the remit of this research there are

some relevant points to note in relation to the rest of the analysis. The 87Sr/86Sr = 0.7139

for this male was the highest for any sample in this research and the most depleted of the

results at Knowth at 618Odw = -8.77. The Sr is indicative of igneous geology (old

radiogenic rock) and both the Sr and the 0 results suggest that this individual originated in Scotland, in what would have been Pictland at that time. The original publication

report for the site and burials acknowledged that KnH6 appeared to have been

decapitated, but suggested that the separation of the skull from the rest of the skeleton

may have been due to taphonomic disturbance (Eogan 1974). Given that the skull was

approximately 40cros from the skeleton, face-upright, and appeared to have been

undisturbed, this latter suggestion seems unlikely.

Burial KnH5 (female) was sampled because it was a supine extended inhumation again thought to date to the Iron Age (O'Brien 1990; Raftery 1994). The S18O" _ -6.83 is

within the range that would be expected for Meath but the Sr at 97Sr/86Sr = 0.7113 is too

radiogenic for the local geology. The positioning of the body with the right arm raised high beyond the shoulder and the marked positioning of her head, twisted as it is to the

right, cannot be explained by taphonomic disturbance and it seems that little care was taken to position the corpse once it was placed in the shallow grave. As with KnH6

neither the strontium nor the oxygen result for KnH5 is consistent with an origin in this

part of Ireland.

Burial KnH9 (female) was a crouched inhumation and was interpreted as reflecting the

earlier `local' Iron Age burial rite and it was sampled for comparative data against which the `intrusive' burials could be tested for possible population groups. Although the

skeleton was recorded as in a poor state of preservation (Eogan 1974) this was not

apparent on inspection and the teeth were well preserved. The strontium result is

186

consistent with the nearby belt of Silurian/ Ordivician geology at 87Sr/86Sr = 0.7094, and

her 818Od,,,, result is more depleted than Burial KnH5 at 81804,, = -6.99. The pathology

report recorded that this lady was unusually tall, a comment which was also made about

KnH6, and that her upper and lower incisors were worn in such a way consistent with a

fibrous material being pulled through her teeth. This repetitive action is consistent with

the over-development of the musculature of her jaws, also noted in the osteology report

(Eogan 1974). A similar wear pattern was noted on the teeth of the female burial sampled

from Rossnaree and one explanation offered for this is that the constant abrasion of a

woollen or flax thread, being held between the teeth, gradually left a wear pattern on the

incisors (Buckley, unpublished pathology report). The 14C date for KnH9 is within the

Middle Iron Age at 175-50 cal BC and this is the earliest crouched burial to be dated to

the Iron Age in Ireland. The more common rite evidenced for this period in Ireland is that

of cremation with good examples from Kiltierney and Ballydavis (see Chapter 6 for full

discussion).

The two skeletons that formed a double burial, Burial KnH 10/11, were placed in a flexed

position and decapitated (Eogan 1974). They are often referred to in the literature with

the soubriquet ̀ the gamblers' because of the unusual assemblage of gaming counters

buried in and around the bodies (Raftery 1994; b högäin 2001). The two individuals are

recorded as the same age and approximate height, and both have evidence of posterior

platymeria on the femora (Eogan 1974: 75). Although they were subject to decapitation,

it is unclear from the report whether this was pre- or post-mortem. Both skulls are evident

within the burial and these appear to have been placed in opposite directions before an

array of grave goods were positioned around the bodies. Unlike KnH5 and KnH6, the

arrangements in and around the burials suggest some care was taken to accord these

individuals a formal burial.

The 87Sr/86Sr for Burials KnH 10 and KnH 11 are virtually identical (0.7094,0.7093) but

there is a slight variation in the 5180 (-7.48, -7.20). The strontium suggests geology

older than that which is present locally, and the 818Od,,,, result is not consistent with an

origin in the greater Meath area, but is suggestive of an origin in the northern counties of

187

Ireland. Although in contemporary society we view the evidence of decapitation as

punitive, there is no additional osteological evidence for other injuries inflicted pre or

post-mortem upon the individuals in burial KnH10/11. Given the similarity in both the

strontium and oxygen isotope results, the pathology of their skeletal remains, and the

number of grave goods buried with them, we might speculate as to whether they were brothers, or perhaps even twins, who were marked out as special in life for this reason.

Three other human burials were sampled and these were KnH 27, an extended

unprotected inhumation again thought to be intrusive, and cist burials KnH32 and KnH33

thought to be local. The strontium isotope ratios for the two cist burials are virtually identical at 87Sr/96Sr = (0.7107,0.7108) and are significantly higher than that for the

extended inhumation which is KnH27 87Sr/86 Sr = 0.7090. The oxygen results for the cist burials KnH32 and KnH 33 are, however, quite different, with & Odw = (-7.25, -6.50

respectively) and so it is very unlikely that these two individuals originated in the same

area. The oxygen result for Burial KnH27 (8I8Od,,,, = -7.18) is outside the range we

would expect for the Meath area and is consistent with an origin in the northern counties.

Observations on the individual burials were based, in part, on the result of the analysis of

a cat sample (KnHCat), which came from secure, but later, 10"' century context. The

result from the cat is in keeping with the expected oxygen range for this part of Ireland at 818Od,,, = -6.49, but the strontium at 97Sr/86Sr = 0.7104 is higher than predicted for the

underlying geology. As the cat has striations on the mandible consistent with it being

used for pelt (see McComick for detailed discussion 2005) it is entirely possible that the

cat was brought alive, or dead, to the site for this purpose. Burial KnH33 has, however, a

virtually identical signature to the cat sample for both Sr and 0, it is also extremely likely

that the 87Sr/86Sr for the area around Knowth reflects a significant input of people,

animals and food raised and grown on the older geology to the north of the site.

188

-5. UU

-5.50

-6.00

-6.50

47.00 0

00 -7.50 00

-8.00

-8.50

-9.00

-9.50

-10.00

IRELAND AND THE UK

YORKSHIRE AND SCOTLAND

SCANDINAVIA AND CONTINENTAL EUROPE

0.7083 0.7091 0.7099 0.7107 0.7115 0.7123 0.7131 0.7139

87Sr/86Sr

Fig 7.8. Knowth, showing S'8O plotted against results for 87Sr/&Sr. Expected local

level for both strontium and oxygen plotted as highlighted rectangle. Arrows indicate

geographic region consistent with level of depletion of 5180

Summary of results

Burials KnH 10/11, KnH 5, KnH27 and KnH33 have oxygen signatures suggestive of an

origin in the northern Irish counties and the elevated strontium of KnH5 and KnH33

narrows potential areas for these two individuals to the older geology of counties Down

and Antrim. Cist KnH32 was radiocarbon dated as part of this research and the result also

challenges many of the assumptions about burial morphology and relative dating. The

radiocarbon date is (UBA 11697; 1210±28) 709-747,766-891 cal AD (2ß) and places it

well outside the Iron Age and within a later period of occupation at the site during the 8th

or 9' century AD. It is interesting to note that KnH6 and KnH 32 both date to a later

period of activity at Knowth during which epigraphical inscriptions in both scholastic

ogham and insular majuscule script were inserted into the eastern side of passage tomb,

discussed in detail in the previous chapter. One of the two areas highlighted throughout

this analysis as a possible origination for individuals at the site is eastern Scotland, and as

189

discussed in the previous chapter, one of the ogham inscriptions records the name of the

king Tallurgg known from the Pictish kinglists which are contemporary with the dates of

KnH6 and now KnH32 (Swift 2008: 120; Byrne 2008: 90). The new dates and the new

strontium and oxygen isotope results suggest that Knowth, like its sister site of nearby

Newgrange, attracted both local and non local people during the Later Iron Age and

through to the 9th century AD. The results from Knowth suggest that individuals from

Northern Ireland, southern Scotland or Yorkshire were accorded burial alongside local

people over the many centuries that the site was in use (Group A and B shown at Fig.

7.8). More importantly there is little actual correlation between the existing relative

dating method and the absolute dates received for the individual burials at Knowth. Nor

is there any patterning in the isotope results that might suggest a correlation between a

likely origin for individuals and the burial rites being used.

7.5 Rossnaree, County Meath

Geological overview

The site of Rossnaree is on the south side of the River Boyne, positioned directly

opposite the great mound of Knowth (Figs 7.5 and 7.6). No faunal material was kept from

the site but given its proximity to Knowth the results can serve as comparative data.

Pathology of remains and teeth

Only one individual female was sampled from Rossnaree and no faunal material was kept

post-excavation. Burial RnR1 was a female who had been overlying two other females

burials and the remains of a small child. RnR1 suffered from a degenerative disease in all

of her major joints including, unusually for this type of disease, her shoulders (Buckley,

unpublished pathology report. I am grateful to Ms Buckley for her permission to

reference her report here). The extent of compression of the vertebrae, along with the

shoulder injuries, suggests that this individual carried great weights across her back for

long periods. Furthermore the soleal ridge is markedly pronounced, an indicator that is

190

common amongst individuals whose leg musculature is very developed, and common

amongst individuals who walk long distances over prolonged periods. Although her

dentition was good overall there was evidence for enamel hypoplasia (striations along the

teeth), which suggests that she suffered from periods of trauma at a young age such as

severe malnutrition. The acute abrasion on her lower incisors is indicative of the teeth being used as a tool, and similar dental attrition was noted on another crouched female

burial from Knowth (KnH 9), although the burial at Knowth is considerably earlier than

RnR1

Lab Ref Ref Burial Type Sr Sr 8180, a Od� 14C Dates smow Calibrated

p. la, to 2a

RnR Burial No. 1 Crouched female . 7089798±. 000004 26.83 -7.06 250-520

cal AD

Table 7.4: Rossnaree, County Meath

Errors are given at 2o, oxygen results reported as rain water equivalents, typical error 0.2 per

mil. Radiocarbon dates have been calibrated using OxCal 4.1 and the IntCatO9 calibration

Analysis

No firm conclusions can be drawn from this single sample but an interpretation can be

offered based on the comparative data now available from Knowth, directly across the

river from this site. The strontium VSr/86Sr = 0.7089 is in keeping with the predictive

range for Silurian-Ordovian geology, which, as noted earlier is local to Rossnaree and Knowth, and the 8'8Od,,,, = -7.06 is also consistent with an origin in the local area. The

female burial was radiocarbon dated to 250 to 540 cal AD (2(y) (1660f40BP). Despite

the fact that she was buried at a site directly associated with the Kingship of Tara, and

appears to have been wearing Roman style earrings (Newman 2005), the osteological

report suggests that she led a life of hard manual labour with periods of severe

malnutrition as a child. This raises important issues about the assumptions that we make

as archaeologists about the status and identity of individuals accorded burial at historic

sites.

191

7.6 Betaghstown, Co Meath

Geological Overview

The site of Betaghstown sits within a belt of pale microtised, grainstone-wackstone, of

Dinantian formation. It is flanked to the north and west by an area of dark calcareous

limestone and to the south by an older formation of Silurian (consisting of thinly bedded

sandstone) and Ordovician (lippilli tuff and mudrock). The geology is similar to that

from Knowth and Rossnaree (see Figs. 7.5 & 7.6) in that we would expect 87Sr/"Sr =<

0.7083 for the geology of the carboniferous period; 87Sr/86Sr = 0.7088 for Silurian and 7Sr/6Sr = 0.7091 for Ordovician geology (McArthur et al 2001).

Lab Ref Ref Burial Type Sr Sr b Oc 8 Oa,. 14C Dates

smow Calibrated to

p, =I, 2a

BTS II B. 2 Male Cist 0.710132±. 000003 27.13 -6.62 430-670 cal AD

BTS IV B. 4 Juv. Male Extended 0.709853±. 000004 27.60 -5.91

BTS VII B. 7 Female Extended 0.709559±. 000003 27.86 -5.52 340-604 cal AD

BTS X B. 10 Male Extended 0.709660±000003 27.18 -6.54

BTS XI B. 11 Female Crouched Not located 397-635 cal AD

BTS XII B. 12 Male Crouched pending -2.96 356-622 cal AD BTS XIII B. 13 Male Extended 0.709923±. 000003 26.62 -7.93 BTS XIIIa B. 13a Male Extended 0.7100781.000003 26.37 -7.76 416-641 cal AD

BTS XIV B. 14 Female Extended pending 25.80 -8.61 256-576 cal AD BTS XV B. 15 Male Cist 0.709990&. 000003 26.39 -7.72 385-632 cal AD

BTS XVI B. 16 Female Extended 0.709908±. 000003 28.44 -4.66

BTS_Sheep 1 0.710121±. 000003

BTS_Shoep 2 0.708784±. 000003

Table 7.5: Betaghstown, Co. Meath

Errors are given at 2a; oxygen results are reported as rain water equivalents, typical error 0.2

per mil; Radiocarbon dates have been calibrated using OxCal 4.1 and the IntCatO9 calibration

based on results first published in Archaeometry No. 35 (1993)

192

Pathology of remains and teeth

Eleven human samples, six males and four females and one juvenile male, and two faunal

samples (both sheep) were chosen for strontium and oxygen isotope analysis. Seven

burials were extended unprotected supine inhumations, two were crouched and two were

cist burials (Table 7.5). There are significant indicators of severe dietery deficiency

amongst the whole population group from Betaghstown identified by Power (1990) in her

unpublished osteology report (Ms Power's report was available to consult on the

topographical files at the National Museum of Ireland and I am grateful to Mr Kelly for

granting access to reference her work here). Common conditions found were anaemia,

osteoarthritis and enamel hypoplasia. The teeth sampled for this research were in a very

poor state, with a high level of calculus present along with severe dental attrition and

evidence of caries. This is suggestive of either poor dental hygiene, a diet that included

poorly processed cereals, or a high proportion of grit from quernstones that had

contaminated cereal based foods (or a combination of all of these) (Power (1990)

unpublished osteology report)

Analysis

Betaghstown is one of the few sites for which there is a series of radiocarbon dates, and

these are shown at Table 7.5 above. The earliest burials from the site are burials BTS

XIV and BTS VII, both female extended unprotected inhumations, which date to 256-576

cal AD and 340-604 cal AD, respectively. The crouched burials (BTS XI and XII) fall

within a group of burials all dating between 356 - 632 cal AD, one of which (BTS XV) is

notably a cist burial and as explained in Chapter 6, these are assumed to date significantly later than the rest based on the existing relative chronology of burial `types'.

The results are shown at Fig 7.9 and all demonstrate elevated ratios that are not

comparable with an origin on the sedimentary carboniferous limestone of the area. The

193

three burials that stand out in terms of the results are the male burials BTS XIII, BTS

XIIIa, BTS XV, which have almost identical 87Sr/lSr results and oxygen results at (0.7099,0.7100,0.7099) and 618Od,, (-7.93, -7.76, -7.72) and all within the same 14C date

range (Group A, see Fig. 7.9). Despite the fact that they are buried using different burial

practices (BTS XV is a cist burial), the data suggests that these three males originated in

the same geographic area. The oxygen range is significantly more depleted than any that has been recorded in Ireland before, the most depleted sample being 8180,, = -7.4 (Diefendorf and Patterson 2005) and it is close to the next range of 618Od� = -8 to -9, found in only south-eastern Scotland and the greater Yorkshire area and in Continental

Europe. Burial BTS XVI is a female extended burial and although her 87Sr/6 Sr strontium is similar to the males just mentioned at 87Sr/6Sr =0.7099, her oxygen result is the most

enriched for the site as a whole at 818Odw = -4.66 and she can not have come from this

area of Ireland. As with one of the results from Carbury Hill, there is only one area of Ireland where these levels have been recorded and this is on the Kerry coastline where

similar oxygen levels have been recorded and the predicted Sr from the known geology

of the area would fit with her result (see earlier discussion). It should be noted again, however, that these results are similar to oxygen ratios recorded outside Ireland around

the Atlantic seaboard (see Fig. 7.2).

The earliest burial from the site is burial BTS XIV a female extended inhumation, which has been dated to 256-576 cal AD. The oxygen isotope result for BTS XIV is very

depleted at 818Od� _ -8.61 and falls within a range that, as mentioned earlier, is not found

in Ireland but in only two parts of Britain, continental Europe or Scandinavia. Burial BTS

XIV is, to my knowledge, the earliest example of an extended unprotected inhumation

excavated in Ireland and although there were no grave goods with the burial, the isotope

results suggest an origin outside Ireland.

The most important result from Betaghstown is that for the crouched Burial BTS XII,

found close to the crouched female Burial XI. The oxygen isotope ratio for Burial BTS

XII was the most enriched from all of the samples analysed in this research. At

194

8180dw = -2.96 and there is no possibility that this male originated in any of the

geographical areas previously discussed in this analysis. The oxygen isotope ratio is

consistent with an origin on the coast of southern Spain or Portugal or in north Africa

(see Fig. 7.2).

Summary of Results

-2

-3

-4

-5

V 06

GD 94 (0O

-7

-8

-9

_in

V) im z

öZ z 0

ni W

Zýý0

0

0.7082 0.7086 0.7090 0.7094 0.7098 0.7102 0.7106 87Sr/86Sr

SOUTHERN SPAIN/PORTUGAL AND NORTH AFRICA

IRELAND AND THE UK

YORKSHIRE SCOTLAND EUROPE

Fig. 7.9 Betaghstown, showing 818Odt, plotted against results for 87SrA6Sr. Expected

local range for both strontium and oxygen plotted as highlighted rectangle. Arrows

indicate geographic region consistent with level of depletion of 6180

Note: BTS XII and BTS XIV do not appear here as the Sr analysis is in progress

195

The most depleted oxygen isotope signature was from the extended burial BTS XIV and

the result suggests an origin outside Ireland, either in Britain or in continental Europe.

Two burials (BTS II and BTS X) have an oxygen result which is in keeping with the

predicted local level but the rest of the burials appear to have originated outside the area

and as the range is also common across areas of Britain it is equally possible that they

originated outside Ireland. The radiocarbon dates from Betaghstown support the fact that,

as at Knowth, we can no longer use the burial type to infer relative dating. At

Betaghstown one of the extended unprotected inhumations is earlier than all of the rest of

the burials including the (assumed) earlier practice of crouched inhumation. The

crouched burials, the cist burial BTS XIV and some of the unprotected extended

inhumations all appear to date to the same time period. It could be argued that this

simply reflects the burial practices of distinctive socio-cultural groups using the site for

burial, or perhaps discrete phases of use for burial. If this were the case we would expect

the strontium and oxygen results to correlate with types of burial practices, and this is not

the case. Variability in the isotopes was evidenced across all of the different burial

practices at the site. The most striking mismatch is between the relative and absolute

dates for the crouched burials, which are probably three or four centuries later than have

been suggested based on the form of burial rite. The artefacts found on the female

crouched burial at the site, BTS XI (which sadly could not be located at the museum and

so did not form part of the analysis), were discussed at length in Chapter 5 and the

parallels for most of her jewellery and costume fitments are in Late Roman or Early

Saxon contexts. BTS XI was 14C dated to 397 to 635 cal AD. We know from the classical

sources that the 4th century was a period of immense population movement in Late

Roman Europe and that by this time migrants from mainland Europe had settled in

Britain. By the end of the 4th century some of these communities were forced to move

once more after the Late Roman foederati were sent to Britain to help the locals deal with

the influx of settlers (Dark 2002). Given that these people could not return to mainland

Europe for the same environmental reasons that they had left in the first place, it does not

seem unreasonable to suggest that some may have sailed to Ireland to find a new home.

Burial BTS XII was 14C dated to 356 to 622 cal AD and this result at b'80d� = -2.96

provides clear evidence that this individual originated in either Southern Spain/Portugal

196

or north Africa. The parity demonstrated between the known geographical area where

such enriched oxygen isotope ratios have been recorded and the known distribution of

finds of the Late Roman ceramics, LRAI, LRA2, ARSW and PRSW, in the same

geographical areas, provides clear evidence that links Ireland directly into the Late

Roman world. As discussed in Chapter 4, the relative frequency of fords of these

ceramics in newly excavated sites in the greater Meath area in Ireland suggests that the

role communities in Ireland had within the later trading networks has been significantly

underestimated (Kelly 2010).

7.7 Rdith na Senad, Hill of Tara, County Meath

Geological Overview

The hill of Tara is an escarpment, which rises to a height of about 155m above sea level

and from which on a clear day a commanding view of 26 of the 32 counties of Ireland

can be viewed (Newman 1997). The geology is sedimentary, carboniferous limestone,

consisting of dark laminated calsilites, calcareous shales and some sandstone. The River

Boyne is within 3km of the site and it curves towards the hill on its route between the

towns of Navan and Trim. The nearest older geology to the area of Tara is the clearly

defined area of Silurian/Orodvian age, north of the site of Knowth and discussed earlier,

and as the crow flies this is some 15 kin from Tara (Newman 2005; see Figs. 7.5 & 7.6).

We would expect 87Sr/86Sr =<0.7083 for the geology in the immediate environs of Tara

and also, given its proximity to Knowth 87Sr/Sr = 0.7088 for Silurian and 87Sr/86Sr = 0.7091 for Ordovician geology (McArthur et al 2001).

Pathology of the skeletal remains and teeth

The preservation of the skeletons was excellent, and seven human samples (all male) and

one faunal sample, a pig found with Burial B, were chosen for analysis. Three crouched burials, three extended burials and one skull were sampled; the skull came from within

the ditch that connects Räith na Senad and Rdith na Rig and although not strictly

197

speaking a burial, it was sampled at the request of Dr Grogan. The lack of obvious wear

attrition or calculus on the teeth sampled from Tara was in marked contrast to all the

other sites analysed in this research project. The lack of wear on the teeth can be

interpreted as the result of a better cereal processing techniques, perhaps using a finer

grained stone, which did not leave gritty residue within finished bread. But it might also be explained by a good knowledge of dental hygiene (Buckley 2008 pers. comm). The

pathology of the skeletons was discussed with Ms Buckley, as each sample was chosen,

and she noted the severe bone porosity on Burial J, which she suggested could have led to

his demise. Severe porotic hyperostosis on skeletal evidence is indicative of iron

deficiency anaemia (Ortner 2003: 373). Although the teeth were free of any calculus,

there was evidence to suggest gum disease, and one individual, Burial E, had an abscess

severe enough to have left a gaping hole in the mandible.

Lab Ref Burial Type Sr Sr 8 OC S O, M, 14C Dates smm

Calibrated

- to 2a

RnS B Juv Male Crouched . 708641±. 000003 27.23 -6.46 RnS D Adult Male Extended . 708404±. 000003 26.47 -7.61 1252-1290 cal AD*

RnS E Adult Male Crouched . 709104±. 000003 26.54 -7.51 721-890 cal AD*

RnS H Adult Male Extended . 708600±. 000003 27.26 -6.42 898-1024 cal AD

RnS I Adult Male Crouched . 709821±. 000004 27.31 -6.35 RnS J Adult ? Extended . 709567±. 000003 26.52 -7.54 RnS K Adult Skull . 708862±. 000003 27.55 -5.99 RnS Pig (near Burial B) . 709625±. 000003 24.80 -10.22

Table 7.6. Rdith na Senad, Hill of Tara, County Meath

Errors are given at 20; oxygen results are reported as drinking water equivalents, typical error 0.2 per mil. Radiocarbon dates have been calibrated using OxCal 4.1 and the IntCat09

calibration.. * These dates were funded under an award from the MA and are shown in full

within the discussion.

198

Analysis

There are several burials from Räith na Senad (see Fig. 7.10) for which the strontium

result is consistent with, or slightly elevated, above the underlying geology of

carboniferous limestone (Burial B 87Sr/86Sr = 0.7086, Burial D 97Sr/96 Sr = 0.7084, Burial

H 87Sr/86Sr = 0.7086, Burial K 87Sr/lSr =0 . 7088). The oxygen for Burial B and Burial

H are also consistent with a local origin at 818Od,,, = (-6.46, -6.42). Burial K at &8Odw =-

5.99 is just outside the local range but it is close enough to suggest that he is also from

the immediate area around Tara are (see Fig. 7.3). The oxygen results for Burial D,

8180 ,,, = -7.61, Burial E 818Od� = -7.51, and Burial J 8I80dw = -7.54 are not consistent

with an origin in this part of Ireland and it is likely, given the similarity in both 0 and Sr

results, that Burial E and Burial J originated in the same area. Although Burial I has an

oxygen signature at 81804 w= -6.35 that suggests a local origin, his strontium results at 87Sr/6Sr = 0.7098, is similar to some of the results from Knowth and is consistent with an

origin in that area. Although Burial K was found near Burial J, there is a significant

difference in both the strontium (87Sr/86Sr = 0.7088,0.7096) and the oxygen (818Od�

5.99, -7.54) results for both.

The most depleted 8180 , was recorded was for the pig's tooth found in association with

the human Burial B, and at 81804, = -10.22 it is one of the most interesting results of this

research. Given the absence of a14C date for Burial B from the skelton and the

uncertainties that remain about stratigraphical associations at the site some caution is

needed in interpreting the result. Charcoal found near to Burial B returned a date of 414-

540 cal AD (UBA 8533 1691±20 BP) which suggests that in keeping with some of the

other burials discussed earlier that we may have yet another late date for a crouched

inhumation (Grogan 2009: 146). The 8'804. recorded for the pig's tooth is not local to

the islands of Ireland or Britain, or indeed most of Europe, but it corresponds to values

recorded in eastern Germany, Hungary and Romania and also parts of Norway

(http: //wateriso. eas. purdue. edu: consulted 17/12/09; IAEA/WISER (2008) Global

Network Isotope Project: consulted 17/12/09). Colleagues at Bristol University

Veterinary School studied the pig's tooth (I am grateful to Kate Robson-Brown for

199

initiating the analysis) and established that it is an upper molar from a female pig, around

5 years old. McCormick has argued that animal husbandry skills were being practiced

throughout the Iron Age in Ireland (McCormick 2002,2007). It is entirely possible that

this pig at Räith na Senad provides evidence of the importation of stock for breeding

purposes: this will be discussed in full in the following chapter.

Summary of Results

-5.00

-6.00

7.00

ß 0

0 -8.00 00

-9.00

-10.00

-ii on

IRELAND AND THE UK

YORKSHIRE AND SCOTLAND

SCANDINAVIA AND CONTINENTAL EUROPE

0.7082 0.7086 0.7090 0.7094 0.7098 0.7102 1 87Sr/86Sr

Fig. 7.10 Räith na Senad, showing 6'8Od,,, plotted against results for 87Sr/86Sr. Expected

local range for both strontium and oxygen plotted as highlighted rectangle. Arrows

indicate geographic region consistent with level of depletion of 5180d"

The strontium and oxygen results for the human skull (RnS_K) are consistent with an

origin at, or near, Tara. Only three of the burials are likely to have originated outside the

local area probably in the northern counties, but not all on the same geology (see Fig.

7.10). More importantly there is no correlation between the strontium and oxygen

200

analysis and the type of burial rite. Since the start of this research, two of the extended

inhumations have been radiocarbon dated and published in full: Burial D has been dated

to 1252-1290 cal AD, Burial H to 898-1024 cal AD. It is unfortunate that both Grogan

and I chose the same extended inhumation (Burial D) for analysis and we can now

confirm that burials D and H are separated by at least three centuries, and both post-date

the activity at the site in the 2' to 4th century AD. Burial D was found some distance

from the rest of the group, closer to where the curtain wall of St. Patrick's church intrudes

into the ringfort bank and it seems likely that this burial is associated with the cemetery

for the church, as the dates are comparable. One clue to the late date of Burial D is given

in the recent illustration of the burial (Grogan 2008: 44), in which the skeleton is drawn

with its ankles very close together and this is suggestive of a shroud being used (O'Brien

forthcoming). Burial H dates to the 10th century AD and appears to have been placed

extended supine but un-shrouded into an unprotected shallow grave. Burial E, a crouched

burial, has been dated as part of this research and the result is, yet again, completely at

odds with the relative dating which had placed it in the 1" or 2'd century AD (O'Brien

1990; Raftery 1994; McGarry 2005; Grogan 2008): Burial E dates to 721 - 890 cal AD

(1205±24 BP): UBA 11698. In the absence of a complete set of radiocarbon dates for all

of the burials it is best to be cautious about extrapolating any generalisations but of those

now dated we can say with some certainty that the absolute dates do not correlate with

the relative dates suggested for them in the past (O'Brien 1990). Nor is there any

correlation between the isotope results and the type of burial practice used. Most

importantly, an origin for all of these burials can be offered within Ireland, and they

should no longer be classified as intrusive.

7.8 Review of the results of the analysis and discussion

The most important aspect to record here is that the analysis was conducted on a specific

set of samples all classified as ̀ intrusive' within the archaeological literature. This

classification was made due to the relative frequency of the occurrence of non-local

material either in the burials themselves or nearby at the site. We can now make some

201

observations based on the new absolute dates and the oxygen and strontium analysis as

outlined in Table 7.7 below.

All Sites

4. UU

-4.50 -5.00

-5.50 -6.00

-6.50 -7.00

0 -7.50

-8.00 -8.50 -9.00 -9.50

-10.00 -10.50 -11.00

ACH O Bts XRn S O Knowth Q Rn Ree

0.7080 0.7090 0.7100 0.7110 0.7120 0.7130 0.7140 0.7150

87Sr/86Sr

Fig. 7.11. The results from sites showing expected parameters for strontium and oxygen

isotope data in County Meath and its environs (shaded rectangle) and expected oxygen

drinking water value range for all Ireland based on Diefendorf and Patterson (2005)

(open square). BTS X11 and BTS XIV do not appear here as the Sr analysis is in progress

Lab Ref S e x

Burial Type

Classification based on Relative Dating method

Prior Classification

14 C Date (where known)

Origin based on isotope analysis

CHI M Extended 3 or 4 CAD Intrusive 471-653 cal AD

Local

CH2 M Extended 3 or 4 CAD Intrusive West Ireland CH3 F Extended 3 or 4C AD Intrusive SW Ireland CH4 M Extended 3 or 4` C AD Intrusive Local CH5 M Extended 3 or 4C AD Intrusive Local CH6 F Extended 3 or 4C AD Intrusive West Ireland KnH5 F Extended 3 or 4C AD Intrusive North Ireland KnH6 M Extended 3 or 4C AD Intrusive 660-788 cal

AD Scotland, Yorkshire or Europe

KnH9 F Crouched 1' or 2C AD Local 175-50 cal North Ireland

202

BC KnH10 M Flexed 1 °' or 2 CAD Intrusive 40 cal BC - Southern Midlands

AD cal 121 Ireland KnHI I M Flexed 1' or 2C AD Intrusive 40 cal BC - Southern Midlands

AD cal 121 Ireland KnH27 F Extended 3 or 4C AD Intrusive Southern Midlands

Ireland KnH32 M Cist Ext 5C AD Local 710-891 cal North Ireland

AD KnH33 M Cist Ext 5C AD Local North Ireland RnR1 F Extended 3 or 4C AD Intrusive 250-520 cal Local

AD BTS II M Cist Ext 5C AD Intrusive 430-670 cal Local

AD BTS N M Extended 3 or 4 CAD Intrusive Local BTS VII F Extended 3 or 4C AD Intrusive 340-604 cal West Ireland

AD BTS X M Extended 3 or 4C AD Intrusive Local BTS XI F Crouched 1 or 2 CAD Intrusive 397-635 cal Not sampled

AD BTS XII M Crouched I" or 2C AD Intrusive 356-622 cal Southern

AD Spain/Portugal N. Africa

BTS XID M Extended 3 or 4C AD Intrusive Scotland, Yorkshire or Europe

BTS XIIIa M Extended 3 or 40'C AD Intrusive 416-641 cal Scotland, Yorkshire AD or Europe

BTS XIV F Extended 3 or 4C AD Intrusive 256-576 cal Scotland/Yorkshire AD or Europe

BTS XV M Cist Ext 5 *CAD Intrusive 385-632 cal Scotland/Yorkshire AD or Europe

BTS XVI F Extended 3 or 4C AD Intrusive SW Ireland RnS B M Crouched I or 2C AD

- - -- Local Local

RnS D M Extended 4w C r4C AD 3 or Intrusive 1252- 1290 Southern Midlands cal AD Ireland

RnS E M Crouched 1 or 2C AD Local 721 - 890 North Ireland cal AD

RnS H M Extended 3d or 4C AD Intrusive 898-1024 Local cal AD

RnS I M Crouched 1 or 2C AD Local North Ireland RnS J M Extended 3 or 4C AD Intrusive North Ireland RnS K Skull - Intrusive Local

Table 7.7 Details of all human samples showing sex and burial type, classification based

on relative dating methods, previous classification whether intrusive or local, '4C dates

where known and suggested origin consistent with the results of the strontium and oxygen

isotope analysis.

203

As shown in Table 7.7 there is little correlation between prior classification based on the

relative dating methods and the absolute dates now published for these burials. Not all of

the burials associated with foreign material are non-local, so they cannot be classified as

`intrusive' based on the existing parameters used for such a classification. Some of the

extended burials appear to be of local people, while some of the crouched burials,

traditionally believed to reflect local burial practice, are in fact non-local. It is clear,

however, that many of the burials are of people who did not originate either in this part of

Ireland, or even Ireland as a whole, and yet the burial customs are indistinguishable from

those previously classified as local. Although samples for strontium and oxygen isotope

analysis were taken from each of the burial practices known in Ireland during the Later

Iron Age, there is no correlation between the strontium and oxygen isotope results and

the types of burial rite used (shown in Table 7.7). The results of the analysis suggest that

we cannot identify non-local people by either the morphology of the burial, the burial rite

itself, or the association of `intrusive' material at these sites. Some of these individuals

may have been born elsewhere but in death they were indistinguishable from everyone

else, until now.

The results of the stontium and oxygen isotope analysis provide evidence of some individuals who may not have originated in Ireland, the significance of which is

demonstrated by the Betaghstown results. As discussed earlier, one individual, Burial

BTS XIV, appears to have an origin in Late Roman Britain or Europe, and BTS XII

originated in Southern Spain/Portugal or north Africa. These results, when placed within

a context of mid 5th century AD Europe can be interpreted as a migrant population group

who having settled in post-Roman Britain may have been forced to move once more to

Ireland, before settling and raising children. But they may have migrated directly to

Ireland, given that the oxygen range evident in Yorkshire and southern Scotland is also

common across the parts of continental Europe from which historic migrations took

place. Some of the grave goods from Burial XI, especially the omega brooches can be

paralleled in late Roman contexts in Britain (O'Brien 1990; Raftery 1994). Overall the

variability in the strontium and oxygen isotope results on an intra and inter-site basis but

also even within the local and non-local results suggests that then, as now, historic sites

204

such as Knowth and Newgrange were attracting visitors in the Later Iron Age and some

at least were allowed to be buried there. Burial practices that have in the past been treated

as exclusively local rites (crouched and flexed forms) are common outside Ireland and I

have argued throughout this research that we cannot fix a relational identity based on the burial practice alone nor can we generalize about either the origin in life or assumed identity in death based on the presence or absence of imported items at a site; there are

now simply too many examples that directly challenge these. In the case of BTS XII, a

crouched burial, not only does it date to three centuries later than the relative chronology

would suggest, but this individual was not a local person. At Rdith na Senad, the relative date of the 1st or 2nd century AD given to the crouched Burial E (O'Brien 1990; Raftery

1994; Grogan 2008) can now be corrected by the new dating conducted as part of this

analysis of 721 - 890 cal AD. The dates from the Betaghstown burials demonstrate that

some of the unprotected extended inhumations are contemporary with cist burials and

although in the past these have been treated as a discrete burial rites, all the evidence

suggests that these forms are contemporary. The 14C dates for Betaghstown and the new dates for burials at Knowth, Räith na Senad and Carbury Hill confirm, as expected and

explained in Chapter 5, that we can no longer use burial type as a form of relative dating,

either on an inter- or intra-site basis. Most importantly of all, this study has demonstrated

that individuals who did not originate in Ireland were indistinguishable in death, until

now, from those who were regarded as local people. It supports the arguments set out in

Chapters 3 and 5 that our propensity to see foreign-ness, either within material culture or

as in this case within burials, may simply reflect a modern ethnogenic construct. The

need to characterise all that is seen as ̀ Irish' or `Celtic' as distinct from that which is not belies the reality that these people were part of the communities in which they lived and died, regardless of their origin.

205

Chapter 8 Towards a materiality of transcultural identities within the Later Irish Iron Age

S. 1 Overview

The archaeology of Iron Age in Ireland has been labelled as enigmatic and

problematic, and its people as essentially ̀ invisible' (Raftery 1994). As discussed in

Chapter 4, two recent projects by archaeologists at University College Dublin `Early

Medieval Project' (O'Sullivan et al 2008) and ̀ The Iron Age: Finding an Invisible

People' (Becker et al 2008) aimed to contextualise all the findings with a view to

creating a more realistic social archaeology for the periods. Although the Iron Age

project revealed many new sites where contexts returned Iron Age dates, little

diagnostic material or settlement sites has been found to alter previous distributional

analyses for both periods (Raftery 1994). Central to both reports is the traditional

relative chronology, which was not critically assessed in light of recent discoveries

and new absolute dates. Both reports reiterate previous circular arguments and

perpetuate an existing model which excludes all consideration of material classified as `foreign' or `intrusive', from that which is deemed to be ̀ local' or `Irish'. My

research was designed to deconstruct this existing historical narrative of `independent

Celtic' Ireland (O'hOgäin 2002: 193) by challenging the continued use of a relative

chronology for the Iron Age around which an isolationist construct has been framed.

8.2 Challenging the underlying constructs and assumptions

Chapter 2 reviewed the literary evidence from the Classical World and the Irish

Annals and although some of the classical references are brief, others, such as that

from Tacitus and Ptolemy, are lengthy and provide valuable information about how

Ireland was viewed in the ancient world. Although historical revisionists have argued

against the continued use of the Annals and Cycles as a proto-history for Ireland,

these form an essential part of the materiality of the period (McKone 1991). Social

constructs, the belief system and community behaviour had changed fundamentally

by the time the Annals and Cycles began to be written down, during the 7th century AD (O'Croinin 2005). But through the redactions and scribal annotations on these

manuscripts we are made aware of the lingering conflict for those recording a heroic

pagan past whilst sitting within the confines of the new religious houses. What we

206

cannot do, however, is use the references within the Annals before the 7`h or 8t'

century AD as historical records, or use them to anchor a relative date for a site (see

discussion on Hencken in Chapters 2 and 4) or allow a presumed history of a site to

dictate dating sequences overall.

The assessment in previous chapters of how the existing narrative was formulated

highlighted the various underlying assumptions and constructs around which a

relative chronology has been built for both artefacts and burials in the period. In the

past archaeologists had to work without the benefit of absolute dates for the contexts

within which these were found, but the wider use of radiocarbon dating in recent

years has now produced a series of dates for material, burials and dated contexts at

sites from which we can anchor absolute dates within the Iron Age. The various

assumptions within traditional archaeological interpretation in Ireland can be

summarized as follows:

Assumption 1: Foreign or intrusive material is not relevant to the overall

relative chronology for the Irish Iron Age and it can be readily distinguished

from the typologies of local material (eg. Henken 1938,1942,1950; Warner

1976; Raftery 1994,2006; O'Sullivan 1998; Cotter 1999)

Assumption 2: A fixed relational identity can be inferred between the likely

origin of an object and the ethnicity of its user (eg Raftery 1994; Warner 1995; Ö'hOgäin 2002; Grogan 2008)

Assumption 3: The change in burial rite from crouched to extended

unprotected supine inhumation took place by the 3'd century AD, and where burials of this type are identified alongside ̀local' burial practices they can be

classified as intrusive as they are likely to reflect the burials of foreigners (eg.

O'Brien 1990,1999a, 2003, in press; Raftery 1994; McGarry 2005,2007).

Assumption 4: The presently used relative chronology for burial in the Iron

Age, which is based on the ̀ type' of burial rite and a similarity in grave

morphology remains a valid method by which sequences can be dated at sites (O'Brien 1990,1999a, 2003, in press; Raftery 1994; McGarry 2005,2007).

207

The existing relative dating scheme has, as its central tenet, a typology of artefacts

that was constructed around material assemblages found during the excavations of historic sites in the 1930s and 1940s. It was formulated around the extent of knowledge at that time and one object in particular, the penannular brooch, highlights

the problematic nature of the continued reliance upon it. At many sites the presence

or absence of a particular type of penannular was used to suggest a dating period for a

site as a whole (see Warner on Lagore 1985: 76). It is now recognized that the

difficulty in such an approach lies within the assumptions that we have made, not just

about the development of objects (using typology to date sequences), but in the long

and varied histories or biographies of even a single brooch (Gosden and Marshall

1999). The penannular brooch has a contested origin, either in the territory of the

Brigantes in northern Britain or in Iberia, but it is equally common in Roman contexts in Britain (Fowler 1963,1965; Kilbride-Jones 1980; Newman 1989). It was adopted,

adapted and developed by different communities in different ways, both in Britain in

the Roman period and in Ireland, and to argue that this is essentially an Irish brooch is

to ignore the plentiful parallels of its early use outside Ireland.

The dating sequence at Lagore crannog was used to date the sites of Ballindery No. 2

and Cahercummaun, (although this latter site was published first in 1938). The

metalwork and small finds from Hencken's excavations at Cahercummaun were

recently completely reviewed and it was argued that the penannular brooch found at

the site should be reclassified and dated to no later than the 5t` century AD (O'Floinn

1999: 71). As discussed above it seems clear that the brooch, which was one of the

few diagnostic finds from this site, can be paralleled in 4th century contexts in Ireland

suggesting that the dating for Cahercummaun, as at Lagore, is out be several centuries

(as argued by B Raftery 1972; J Raftery 1994; Caulfied 1995). The recent assessment

of the small fords from Rdith na Senad at Tara has identified an early form penannular

within a phase of activity dated against secure parallels of the 3'h or 4th century AD

(Allason-Jones 2008: 107). The parallels offered for similar brooches found within Irish contexts were at Cahercummaun and Ballinderry 2 (Allason-Jones 2008: 107).

The revised date for the brooch at Tara fits with the probable date of the Roman

material found at Knowth, where another example of an early form penannular was found (Kilbride Jones 1989; Eogan 1974; Allason-Jones 2008: 107). Earlier dating

208

sequences are likely at all three of Hencken's sites and it follows that we can state that

the Roman material which was found at Lagore and Ballinderry 2 was relevant to the

dating sequences as it indicated an earlier phase of activity at these sites. If, as I have

argued in Chapter 4, there is an earlier level of occupation that Hencken missed at Lagore then the relative dating system starts to unravel (Lynn 1985: 73).

Based on the strontium and oxygen isotope analysis and new radiocarbon dates we

can now challenge the basic premise on which assumptions 3 and 4 above were

applied to burials, as outlined in detail in the previous chapter. The findings indicate

that the existing relative dating system using burial morphology to suggest discrete

temporal periods of activity, on an inter- and intra-site basis, should no longer be used

to ascribe dates to burials in the later Irish Iron Age. Furthermore, the presence of `intrusive' material at a site where transitional burials are also found cannot be used to

infer that those individuals were of non-local origin. These findings also call into

question the first two assumptions above, which underline the way in which

`intrusive' material has been classified separately to that which is thought to be of local manufacture. It also challenges the more fundamental assumption that the origin

of an object is indicative of the ethnicity of its user.

Transition in material culture and social practices in Later Iron Age Ireland appears to

have been driven by the purpose of the contact, either economic or religious, and the

level at which this was happening varies between temporal periods and across

geographic areas within the archaeological record. Where cultural influences from the

Roman world can be evidenced, however, is through the development of a new social hierarchy and the establishment of the large enclosed sites from which an array of

Roman imports has been found. The initial impetus, ideas and influences from Roman

Britain that acted as a catalyst for change would have been lost over the centuries, as

the gradual pace of transition reached communities in the west and south of Ireland.

The complex nature of what we have in the past regarded as being Roman, even

within the Roman province of Britannia, adds to the difficulty of reading an

archaeology for this period in Ireland, as here the archaeology also reflects the diverse

cultural influences bounded together under the umbrella of a `Roman' cultural

package of the period, including that which is regarded as ̀ Germanic'. I have argued

earlier that applying a single epistomological position may simply replace one

209

narrative with a new dominant narrative, and in the move towards accepting the latter,

useful and relevant parts of the original may be lost. In order to assess the likely level

and extent of Roman cultural influence on parts of Ireland, it is necessary to underline

how both local people, and foreigners, appear to have engaged with a variety of social

practices in Ireland. Embedded within the emerging social structure was a familiarity

based on contacts established within the PRIA, evidenced through the wide array of

imported artefacts whose origin and distribution correlates exactly with that of the

Roman material. Using the new evidence of sites and fords, along with a scientific

analysis, this work is set within a contemporary theoretical framework and what is

offered is transcultural social history for the later Irish Iron Age. Embedded within the

culture of the people living on the eastern sea board counties of Ireland are aspects of

social, political and economic development that reflect their engagement through a

shared value system with a world which was at this time essentially Roman.

210

IRELAND: TRIBAL TERRITORY BASED ON MANUSCRIPTS

`I

I, 1® 8ANTRIM DONEGAL

OULAID (LATER Ui NEIL)

TYRONE opl/

0

DOWN 014RMANAGH

} 10

a O

SLIGO ,1

EITRIM MAYO CONNACHTA CAVAN N LOU

FOSCO : MON C_ý

0 0 10 R

WESTMEFH GALWAY

ýOFfALY - cll: )-Io0

ine 8 00

DL LLIM "

0 KILDARE

LA/GiNi WISKLOW

IO CLARE

TIPP RY KI

LIMERICK

®E ANA( , AT -, F OPD

KERRY WATENFCMU

©9CORK KEY: O BURIAL

o FINDIsstto3rdCAD) O HOARD 0 FIND 14th to 5th C AD)

Fig. 8.1. Updated distribution map offinds, including only those burials for which we

can now demonstrate an orgin outside Ireland, and a radiocarbon date for the period

211

8.3 The environment and the economy

The analysis on environmental and faunal data from the excavation at the historic site

of Dün Ailinne has convincingly demonstrated that activity was episodic, with

multiple periods of rebuilding centred around late Autumn and the interpretation fits

well with what we know of the festival of Samhain (Johnston 2008). A transcultural

approach to the archaeological interpretation allows for the deposition of foreign

material at sites such as Freestone Hill, Newgrange and Rkith na Senad to reflect a

collective activity that is likely to have included some people whose material culture is more obvious within the archaeological record. But it does not follow that theirs

was of higher status or of greater importance. The faunal remains from these sites are

particularly significant, as the quantities are so large as to have been interpreted as

community feasting. At DOn Ailinne, within the Rose/Mauve phase (dated IA1),

choice cuts of meat brought to the site reflect the butchery of an extraordinary number

of calves (Johnston 2007). The faunal remains from Räith na Senad suggest that pig

and oxen were eaten, but not cattle or sheep, and again we can interpret these as the

selection of particular types of meat for the purposes of community feasting. It is

worth remembering, as explored earlier, that it is our preconceptions that have

elevated artefacts to indicators of high status and relegated organic remains to

domestic detritus. In reality, an individual's status may have relied on the ability to

demonstrate wealth through the provisioning of food on such an occasion.

There is also a growing amount of faunal evidence from the larger enclosed sites that

suggests specialised animal husbandry was taking place within the Iron Age

(McCormick 2002,2007). The results of the isotope analysis conducted in the present

study on a pig's tooth found with Burial B at RdIth na Senad suggests an origin for

the pig in continental Europe, and this may be evidence of the importation of breeding

stock. A 14C date of 414-540 cal AD was derived from a sample of charcoal taken from a context near the burial, but as there remain questions over the stratigraphical

relationships in this area of the site, it is best to be cautious until the human burial is

dated. There is a clear difference between species frequency at Räith na Rig and at Rbith na Senad, two sites which we know through radiocarbon dating overlap in

terms of activity (Roche et al 2002; Grogan 2008). At Rdith na RIg the percentage

212

MNI for faunal remains sits firmly within a domestic secular profile of predominantly

cattle, yet at the adjacent ringfort at Räith na Senad the faunal assemblage is

dominated by pig, with virtually no evidence of domestic cattle. The interpretation

offered earlier of a ritualised activity taking place at the site, suggests that the faunal

evidence is distinctive and significant as pigs and pig-keepers are associated with the

otherworld in Irish mythology (Ö'hOgäin 1999; Koch 2006: 219).

In the past the reliance on a small group of pollen samples taken from bogs in

northern Leinster (Red Bog County Louth) and north Munster (Littleton Bog County

Tipperary) have been used to infer a downturn in agriculture after the 1st century AD,

known as the ̀ Iron Age lull', with recovery suggested from the 50` century AD. This

has helped to create an impression that communities in Ireland were not in a position

to create the surplus production needed to engage in trade at any significant level

(Mitchell & Ryan 2001; Newman et al 2007). On closer inspection of the pollen

diagrams published by Michell and Ryan (2001), and Newman et al (2007), it is clear

that agricultural production was stable before the mid P century AD, with evidence

of poaceae (graminae) alongside an increase in the indicator species most commonly

used for cereal cultivation, plantago lanceolata. Cereal cultivation and processing

were taking place at the site of Carickmines in the later 1st century AD, and this is

supported by the most recent analysis of the landscape at Rdith na Senad, where the

range of charcoal evidence was interpreted as indicative of an open mixed landscape,

arable and pastoral, in the period of activity dated to the 2d to 4m century AD

(O'Drisceoil 2007; Grogan 2008). The data from the pollen records at Emlagh does

not correspond with previous interpretation, which appears to have equated an

increase in arboreal pollen with an increase in woodland (Mitchell and Ryan 2001).

The pollen data from Redbog, Littleton Bog and Emlagh indicate a marked increase

in hazel and ash, both trees which are suitable for coppicing and cultivation, as part of

a mixed agricultural economy (Dincauze 2000: 357); there is no correponding

increase in pine or oak at Emlagh, which would suggest woodland resurgence.

Furthermore, although the evidence for ceralia and plantago lanceolata are less

within the earlier periods, they are still both evident within the pollen records

throughout the entire period. The environmental evidence based on the analysis of

charcoal from Räith na Senad and Räith na Rig on Tara and Raffen hillfort all

confirms that the plains of Meath by the 2°d century AD, were within an open mixed

213

environment, with pasture and arable cultivation taking place. We can now add these

to a growing list of sites around Ireland where intense arable activity has been dated

from the 2°d to the 5t' centuries AD. Sites such as Long Lough in County Down, dated

to between P and 5t' centuries (Hall 1990: 380; Kerr 2007; 44); Lough Doo County

Mayo, dated between P and 5t' centuries AD (O'Connell 1987: 52; Kerr 2007; 44);

Lough Sheeauns in County Galway, before AD 500 (Malloy and O'Connell 1991:

102; Kerr 2007; 44); Loughnashade in County Antrim, intense arable activity between

AD 220 and AD 550 (Weir 1993: 25; Kerr 2007; 44).

Agricultural surplus must have formed a significant part in any exchange system as

material imports within a non-monetized economy had to be paid for through suplus

production (McCormick 2002). The biometrics of animals found at historic centres in

Ireland are far larger than the ordinary domestic livestock (McCormick 2002,2007).

The skeletons of domesticated dogs found at sites such as Haughey's Fort in Antrim,

and at Rdith na Rig at Tara are far larger than those found elsewhere, and again it

suggests a deliberate breeding programme of large hounds (McCormick 2002;

Bhreathnach 2002). In the 4th century AD the Roman consul Symmachus wrote a

letter his friend in which he discussed the arrival in Rome of seven Irish dogs that

were presented to the people of Rome on the day before the public games were to

commence (Epistle 2.77; Freeman 2001: 102). Symmachus also remarked that they

were so large that the people thought that they had been brought in iron cages (op. cit.

102).

8.4 Trade and imports

That Irish dogs were mentioned in a 4th century AD Roman text may suggest that they

were a diplomatic gift, but they may also have formed part of an earlier and more

regular trade between Ireland and Roman Britain. The I' century BC geographer Strabo listed goods traded between Gaul and Britain many of which were organic in

nature, such as hunting dogs, hides, wool and slaves (The Geography, Book IV, 2-4).

In the text of the Confessio, accredited to St Patrick, it is recorded that hundreds of

people were captured in western Britain by raiding parties, and it is possible that Ireland was involved in the human slave trade markets, as this was certainly the case in 10th century Viking Dublin (Wallace 2008).

214

-11 There is one other product that, to my 0

o. 08

knowledge, has not been explored ; UUID (LATER Ul NE

previously as a possible item of trade and ."9 G II? and sli er

that is raw metal. An analysis of the -'" CONNACHTA' ,

distribution of sites where Roman ° Lnd or

material has been found in Ireland, cross bra

referenced with the geological maps 9°"" ' `A'IM

assessed during the isotope analysis,

suggests that there is a distributional

overlap with natural deposits of iron, KEY " O.

copper, lead and silver (see Fig. 8.2). At 0 A.. S. CADI

least one copper bun ingot, identical

in both manufacturing style and Fig. 8.2 Distribution of natural resources based on Geological Survey of Ireland data (drawn by

weight to those produced under author)

Imperial direction at sites such as

Parys Mountain in north Wales, has been found in County Dublin (topographical

files: NMI: consulted 2008; Raftery 1994). Given that there are plentiful copper

deposits in this region, it seems very unlikely that this item was imported into Ireland.

A much more likely explanation is that techniques of extraction and processing that

had originated in Roman Britain had been adopted at a local copper processing site in

Ireland. In the initial post-Conquest period the coordination and extraction of natural

ores was under direct control of the military, evidenced by the standardisation of

imperial hallmarks found on lead pigs (Frere et al 1990: 39). The association of the

Roman military with exploitation of metals is especially significant for Cornwall,

where all of the known Roman forts, at Nanstallon (Roman tin extraction), Restormel

and Caistock (Roman silver mining), appear to have been strategically positioned on

natural ore bodies (Jones and Mattingly 1990; Thomas 2009). Military control

appears to have been handed over quite quickly at the mining site of Charterhouse-on-

Mendip, where the name of a local contractor appears on the lead pigs produced at the

site by AD 49 (Todd 1995: 65). These contractors appear to have been peripatetic, as

the same names appear on lead pigs produced in Mendip and as far afield as

Derbyshire (Frere et al 1990: 40). It is not unreasonable to suggest that contractors

215

involved in copper industry in north Wales may have extended their business interests

to County Dublin (Kelly 1976).

The extraordinarily rich artefactual assemblage from a small excavation at Rd ith na

Rig, suggests that craftworking was carried out at some sites in the Iron Age on an intensive scale (Roche et al 2002). Along with the evidence of multiple hearths being

used, moulds for bronze working were found along with glass bangles and beads.

The equipment for specialist craftworking, joiners dogs and associated tools are the

earliest found in Ireland (Roche et al 2002). Although the authors noted that there was

no definitive evidence that glass was being made at the site, the finds catalogues

included worked and wasted glass with tool marks clearly evident, suggesting that

glass was being worked into finished objects here. Furthermore, recent chemical

analysis of the glass found here and at the adjacent site of Räith na Senad matches

that of the earlier chemical analysis carried out on lumps of red glass found nearby at

Kilmessan, with the levels of antimony in both matching those of Roman glass

(Bourke 1995,2008: Stapleton et al 1999). The chemical composition of this glass is

completely different to that which was used for enamelling in the Early Historic

period in Ireland (Stapleton et al 1999: 919). Therefore, it is entirely likely that raw

glass was also being imported from the Roman world along with the more obvious

consumables.

The scale of the activity uncovered at Rdith na Rig, the radiocarbon dates for which

support some contemporary activity with Rdith na Senad, suggests a major centre of

craftworking at Tara. Similar activity was demonstrated at Dun Aillinne, Emain

Macha and Clogher hillfort, the latter containing evidence of so many moulds that it

has now been classified as a `brooch making' site (Warner 1985-6: 76). Other centres

of mass production are known from the later Iron Age such the island of Inishkea, off

the west coast of Mayo. Here, a site was excavated where the scale of the activity led

it be to classified as industrial. The site was for the extraction of purple dye,

evidenced from the extraordinary amounts of broken murex shells, or dog-whelks,

amongst which was found a Roman patera of probable 3rd century AD date (Henry

1945). At Simonstown, again in County Meath, excavated recently in advance of a

new motorway, over two tonnes of metallurgical scrap waste was removed and the

216

extraction and processing has been securely dated from as early as the 4th century AD

(O'Sullivan 2008).

The great majority of imports classified as IA2 within this research date from the 2"d

century AD through to the 5t' century AD and it likely that Britain's role in relation to

Ireland can be paralleled with the pivotal role that Denmark held in relation to the

Baltic in the Roman period. In the 2nd century AD Denmark acted as a conduit

through which those involved in the amber trade in the Baltic area, which remained

outside direct Roman control, were linked into the greater Roman economy (Lund

Hansen 2003; Storgaard 2003). Up to now, Ireland has always been considered a

recipient of exotic goods, but there is growing evidence to suggest that it may also

have been an exporter. Hounds, dye, metalworking, glass making, cattle, horses and

possible slaves may have been traded with the neighbouring island of Britannia in

return for material that might not leave obviously `Roman' evidence, such as animal

stock for breeding purposes. Entrepreneurs and their retinue could have filled the port

of Drumanagh on Romano-British ships but likewise Irish ships and traders would

likely have visited ports in Roman Britain. People travelling to Britannia would have

seen great sights, used the straight metalled roads, visited hostelries towns and

temples designed and built by the new Romano-British administrators. The interplay

and commerce would have ensured that ideas came back to Ireland not just in terms of

the new political order, but about the military might positioned on its doorstep. The

fact that exotica has been found alongside local material at Drumanagh suggests that

this was a nodal trading point, and its construction and situation are remarkably

similar to British Iron Age sites, such as Hengistbury Head, and the now ploughed out

site of Portland in Dorset (Cunliffe 2001; Taylor 2001). All three sites are

characterised by their positioning, with steep cliffs forming natural sea defences on

three sides, and a double or triple embanked landward facing defence. Furthermore all

have a nearby harbour of shale, or sand, where boats could be off-loaded. Although

we can only speculate about what will be found at Drumanagh once permissions are

granted to access the site, it is probable that the range of imports will be comparable

to that found at a site such as Hengistbury.

In Ireland, imports continue through to the 6th century AD and if we include the

evidence of the later ceramic E ware, importation from Merovingian Gaul continues

217

through to the 7th century AD. Although these ceramics are no longer Roman they

cluster at sites around the same distribution networks, between the Boyne and the

Liffey, the River Bann in the north, and with a further access point in the south in

County Cork (Fig. 8.1 and 8.3). The Late Roman amphorae, Phocaean Red Slip ware

and more recently African Red Slip ware which has been recognised with increasing

frequency at sites in Leinster is all but absent from the northern counties.

The central point of

disagreement for those who

have attempted to assess trade

and exchange along the western

and northern seaboards during

the Late Roman period through

to Late Antiquity has focused

on whether finds of imported

ceramics are as a consequence

of wholesale

commercialisation, or simply

short-haul trading (for a

summation of the lengthy

debates see Campbell 2007).

Most recently it has been

argued that there is little or no

evidence of trade along the

I-

western sea-lanes and that Fig. 8.3. S'h and 61h century AD imports from the these were essentially closed to Mediterranean, and 7's" century AD imports of E ware

trade at that time (Fulford 2007). Hollinrake 2007, with additions by author)

.

Notable within this latest paper are the citations from Bateson's catalogue from 1973,

and conspicuous by its absence is any consideration of Bateson's later work from

1976, in which he stated the need for traditional Roman scholars to be open to the

probability that much that he originally questioned may in fact have been

contemporary imports (Bateson 1976). Also absent is any reference to scholars such

as Newman (2005), O'Floinn (2000), Kelly (2001), Hunter (2001), Harding (2004 &

218

+ E-wars 0 Aegean imports

2006), Doyle (1998), Rankin (1999) Raftery (2005), Warner (1995 & 1996),

Mattingly (2004 & 2006), or Campbell (2007), all of whom have put forward

explanations that better account for the material culture patterning of Roman objects

within Ireland in both the earlier and later periods.

Given the relative frequency with which the later imported ceramics are now being

recognised at newly excavated sites in Ireland, it is reasonable to argue that these

reflect regular commercial trade, rather than occasional or secondary haulage. This

trade may have formed part of a network of commercial activities associated with the

exploitation of tin resources in Cornwall, but it does not necessarily follow that it had

to be part of it. A recent survey of nautical activity for the Late Roman period, based

on the distribution of imported ceramics in the south-west of Britain, assessed the

probability of commercial shipping networks from the Mediterranean to Britain and Ireland (Hollinrake 2007). Using the example of the two Roman period boats found in

the British Isles (one from Blackfriars dated to AD 130, and one from Guernsey dated

AD 280) it was argued that the tonnage, or weight bearing capacity, of these boats

could be established by using the line on the boat's hull at which damage from the

teredo or ship-worm was evident (Hollinrake 2007: 7). Estimated to be of around 35-

40 tons and at over 25 metres in length, these ships were large enough to carry

significant amounts of cargo (Hollinrake 2007: 8). The distribution of late Roman

ceramics in north Africa, but especially from the coast of southern Spain and Portugal

suggests there were major ports along the sea route from the eastern Mediterranean at

which supplies and specific cargos were refreshed for on-going journeys. The oxygen isotope results from a burial at the coastal site of Betaghstown in County Meath are

consistent with an individual who originated where this trade was based.

219

Fig. 8.4. The imposing site of the Rock of Cashel, County Tipperary, political seat of

the Eöganacht (photo by author)

In the 4"' century AD a powerful elite group known as the E6ganacht are recorded for

the first time in the Annals as holding a political seat at the stronghold of the Rock of Cashel, in Munster (Fig. 8.4). Recent excavations at Cormac's Chapel on Cashel have

confirmed 5`h and 6`h century AD activity at the site but a fine 2"d century AD Roman

fibula was also found during the excavations. The sites of Garranes and Garryduff in

County Cork are both historically associated with the E6ganacht Rathlinne and it is

notable that an array of imported material is evident at both. Unlike other power

brokers whose geneological lines of successions are outlined in some detail within the

Annals to demonstrate their legitimacy as kings, the E6ganacht are not offered any

such pseudo-history (Charles-Edwards 2000). In the 3`d century AD, however, a tribe

known as the Venii are recorded as settled in southern Ireland in the area of Cork, a

town whose prosperity through to recent times has relied on its safe natural harbour

(Ö h'Ögäin 2002). The similarity in the etymology of the name Venii and that of the

Gaulish tribe the Venetii may suggest a link or even an origin for the E6ganacht in

Gaul itself. More importantly, the level of imported material at both Garranes and

220

Garryduff demonstrates clear links with the Mediterranean from the 4' and 5'

century AD (Ö'Riordäin 1945; 6 h'Ögäin 2002; Campbell 2007; Kelly 2010).

8.5 Settlement and society

The dating of all the major type sites, hill-forts, promontory forts, ringforts and

crannogs needs to be comprehensively reviewed. Despite around 150 radiocarbon

dates from ringforts, many are mutiple samples from the same site, and even

collectively they still cannot be accepted as a representative sample of the 50,000 +

ringforts that are known throughout Ireland (Stout 1997). There is also a problematic

regional bias as most of the dates are from sites in the northern counties.

The modelling of recently published radiocarbon dates in the present study supports a

date for the emergence of the ringfort as a domestic homestead before the 5th century

AD (see Chapter 4, Fig. 4.5). Although there are now many Iron Age radiocarbon

dates for ephemeral posts and pits found during the recent commercial excavations

these contexts need to be explored much more fully before we can regard them as

settlements (contra Becker et al 2008). It is essential that we now move away from

dating sites based on similarity in general morphology and their assumed

contemporaenity, as we know that even simple ringforts had multiple functions and

offer widely different absolute dates (Stout 1997; Kerr 2007; O'Sullivan 2008). We

need to differentiate between homesteads which are defendable (in the sense that they

provide defense if needed), and those ringforts, hillforts and promontory forts which

are overtly defensive in nature. These latter sites, such as Rathgall, were built or

reconstructed with the deliberate intent of protecting either the people or the activities

taking place within their vast banks and ditches.

Over time the social, economic and political changes taking place in the eastern

counties would have impacted on the rest of Ireland and this certainly appears to be

the case from the evidence of later imports in the Cork area. The immense metalled

roadway known by its Irish name an Claidh Dubh has been dated to the 21d century

AD (Doody 2008). A knowledge of engineering, planning and construction was

needed to create the road itself and its flanking ditches, the remains of which stretch

221

for over 20 kms across the north Cork landscape and yet it appears anachronistic

today, positioned as it is seemingly coming from and going to nowhere in the

landscape. This point was made by Raftery in relation to his excavation of another impressive Iron Age road at Corlea in County Longford dated to the Middle Iron Age

(Raftery 1994). Within the wider landscape, however, it is clear that the Corlea Road

travels directly towards Atha Liag, modem day Lanesboro in County Longford. This

ancient fording point of the River Shannon, north of Lough Ree leads to the site of Cruchain Al, west of the River in Connacht (Raftery 1994). Cruchain Al was the seat

of those power in Connacht and today it is a complex of ringforts, barrows and

mounds of undetermined date, but in status and sheer scale it equals its better known

rival of Tara. Cruchain has never attracted the interest that Tara has, and yet I would

argue that it is of equal, if not greater importance, given that a site labelled as a regia is noted west of the Shannon on the map by Ptolemy of the 2d century AD.

Much of the material which is relevant to the Later Iron Age activity at sites around

Ireland has been orphaned in the past, relegated to intrusive, or pulled forward into a later period. As material of likely Romano-British origin had little other than a

townland recorded as a find spot, previous scholars were reluctant to regard these as

contemporary imports (Bateson 1973,1976). Given the parallels between material at

sites in Scotland and along the Antrim coast in the earlier Iron Age (chapters 2 and 3),

it is possible that a continued familial or kin based social structure was maintained

throughout the Iron Age with close contacts across the short sea crossing. The area of Argyll on the western coast of Scotland owes its name to the Gaelic oirthir gaedhael,

which translates as ̀ eastern Ireland', and we know that this close community

relationship resulted in the formation of the kingdom of Dill Riada in Scotland from

the 6m century AD (Lane and Campbell 2000).

The first pattern evident within the analysis of imported material is that virtually all of the `exotic' imports found with I "century BC to 1 s` century AD burials, both

cremations and inhumations, have close parallels in northern Britain in the pre- and

post-Conquest period, as do the burial rites themselves. Furthermore, the ealiest Roman material follows the same distribution in counties in the north and east of Ireland as material which has been classed as British of PRIA date. There is a clear

regional preference for the types of material coming into the northern counties which

222

is different to that from the eastern and south-eastern counties. The Roman material in

the north of Ireland is predominantly Samian ware, with a wide range of forms

evidenced but notably examples of the highly decorated Dr 37 (Bateson 1973,1976).

Along with two early silver coin hoards the material is classified as IA2 and a parallel

offered with that found beyond the frontier in Scotland. Hoards from Scotland are

usually explained as bribes of high quality material to those power brokers most likely

to cause trouble along the north-west frontier (Kamm 2004; Hunter 1997,2001).

Roman material of 2°d century AD clusters in the north-eastern counties, around the

River Boyne in the east and the River Barrow in the south-east, and the latter includes

the only known Roman cremation burial in Ireland. The Stoneyford cremation vessel,

a thinly walled green glass vase, suggests not just familiarity with Roman objects but

a community who had access to such material before the burial, and who were familiar with correct Roman burial customs (Warner 1995; Raftery 1994,2005;

Bourke 1994,2008). There are at least six ogham inscriptions from the area of the

Roman cremation burial that record Roman names (McManus 1992: 113; Thomas

1998: 11). Along with the collyrium stamp found nearby at Golden in County

Tipperary all of the evidence supports Warner's suggestion that by the 2°d century AD

there was a settled community living in a Roman style at an as yet unknown site in

Tipperary (Warner 1995).

There are no hoards of precious metal in the the eastern counties but there is a

remarkable series of fords of gold jewellery and coins from around the passage tomb

at Newgrange. But here, there is a difference between the material from the earlier

centuries, which comprises some coins dating to Flavian period, and the later 4th and 5th century evidence for gold solidi, gold pendents and an array of precious jewellery

(Carson and O'Kelly 1977). If the jewellery and coins had been deposited as a hoard,

or within a burial context, we might be tempted to parallel the items with similar gold

pendents found within Germani Libera or Scandinavia (Hedeager 1990; Lund Hanson

2003). We cannot, however, neglect the context of these finds, as they were not deposited as a single hoard but appear to have been made as separate deposits over a

period of several centuries. The closest parallels for the jewellery are from cult or

ritual sites in Roman Britain, such as Nettleton in Wiltshire, Aqua Sulis in Somerset

or Lydney in Gloucestershire (Bathurst 1972; Wedlake 1984; Cwiliffe 1988a). The

material from Newgrange is contemporary with some of the more domestic evidence

223

from the nearby passage tomb of Knowth. At Knowth, toilet implements, a stylus, bowls of Dr 37 and an earlier penannular brooch were found within the first phase of

settlement at the site and it all dates, on stylistic grounds, to the 2"d to 4t' century AD.

The activity at the site of Räith na Senad has now been securely dated to between the 2°d and 46' century AD and although Grogan (2008) has suggested that this was a domestic settlement, the remarkable evidence suggests otherwise. The highest quality Roman drinking vessels, along with environmental and faunal deposits all suggest a

religious festival was held at this ringfort Furthermore, such an intepretation fits

better with the origin of the site's name: Teamhair, from the Latin templum (Newman

2005; Bhreathanch 2005).

There is no certain evidence for the importation of material in the 3`h century AD,

although a silver coin of the usurper Carausius (286-293 AD) was found in County

Kildare and there is a strap-tag from Rathgall which may date to the same period (IA2). The P century was a period of disarray within the politics and administration

of the Empire, with Diocletian only salvaging order from the chaos after his

succession in AD 284. We might wish to speculate on the origins of the tribe known

as the Laigini, whose geneological lists recorded within the Annals refer to their

origin as ̀ British' and who appear to have taken control of Tara in the 2°d or P

century AD (0' hUiginn 1990). The emergence of named dynasties, the Airgialla, the

Three Collas, the Ui NO, in the 4th or 5th century whose rise to power results in the

usurption of the Kingship of Tara from the Laigini (O'hUiginn 1990; Warner 1995)

provides a tantalising glimpse of the political changes taking place, even if these have

been obscured by the redactions that offer alternate origins for them (see Ö'CÖrrain

2005, for a detailed overview).

There are differences even in the types of hoards evident from the north and south

throughout the period in Ireland. Three of the silver hoards are from the northern

counties, not the eastern counties facing Roman Britain and the only later silver hoard

in the south of Ireland is from Balline and is dated to the 5th century (IA2). The

Balline hoard has been classified as a payment, rather than the result of random

plunder in Roman Britain, and it is parallelled with silver found at Richborough in

gent, stamped with identical hallmarks (Painter 1978).

224

A hoard of low value coins dating to the reign of Constantine were found within a Romano-British ceramic pot at Rdith na Senad at Tara (Bateson 1973; Carew 2000;

Grogan 2008). Many such hoards, containing low value coins dating to the reign of Constantine, have been found in Roman Britain, with a concentration in distribution

around the Severn Valley in what was an area of rich agricultural estates in the P and 4s' centuries AD (Jones and Mattingly 1990). The explanations offered by

archaeologists for the burial of coins in such quantities are varied, and they range from votive or religious reasons to the removal of coins from circulation for economic

reasons. The reinterpretation of the site of Rdith na Senad as one of religious activity,

rather than domestic occupation, might clarify the motivation for similar finds and

their contexts in the south-west of Roman Britain, but more importantly it offers clear

parallels between hoarding practices in Roman Britain and in eastern Ireland.

The growing number of recognised later imported ceramics (shown at Fig. 8.1) which date to the early 5t' century AD at sites in County Meath undermines any suggestion

that Ireland was outside the Late Roman trading networks (contra Fulford 2005) and

supports those arguments in favour of direct involvement with the Roman economy:

views which up until now have remained outside mainstream scholarship, (eg. Raflery

1994, Warner 1995 and 1996, O'Brien 1990 and 1999 and Mattingly 2006). Along

with sites such as those mentioned in County Meath is the extraordinary level of

exotic imports from the ringfort at Lisnacaheragh (Garranes) in County Cork. The

recent reassessment of the ceramics from its neighbouring site at Garryduff places

activity for both sites firmly within the later 4t' or 5t` century AD (Kelly 2010).

Material of later date has been picked up from field walking at the promontory fort of Drumanagh and it is likely that this is the trading centre from which much of the

material in the eastern counties was distributed. The later imported ceramics, as

classified and detailed in Chapter 5, have been recognised with increasing frequency

in recent years at sites in the eastern counties. The range and quality of ceramics now include sites where Phocaean Red Slip and more recently African Red Slip ware has

been found along with Late Roman amphorae and glassware and these have pushed

the dates for importation of ceramics back to the early 5s` century AD. Late Roman

material from sites in County Meath, classified and detailed in Chapter 5, follows the

some distribution of that of earlier Romano-British material.

225

Many of the sites where Roman material has been found are historic sites that have

caught the imagination of both antiquarians and archaeologists because of their sheer

monumentality and visibility within the landscape. They were designed to impress,

whether they were secular, religious and/or political, and they represent the deliberate

manipulation of the landscape by those at the apex of social constructs within Ireland

at this time. More recent interpretations of sites where evidence of imported ceramics is limited to one or two sherds have been criticised for over emphasising or the

`maximalisation' of the relevance of such material in relation to wider activity at the

site itself (Campbell 2007 citing Hodges 1988). It is equally likely that authors have

minimalised the importance of such finds in Irish contexts and we should also be open

to the very likely possibility that a national approach to investigation has obscured

regional development and variability in Ireland during the Iron Age. A period of

transition and development is evident in Ireland in the 5`h and 6'h centuries AD and we

need to explore why and how these changes were happening on a local and regional basis. Material that dates to this period has been squeezed out of the Iron Age into a

relatively short timeframe at the start of the 6th century AD, making the impact of

gradual change appear sudden and dramatic. Although wholesale conversion to

Christianity has been used in the past as a catalyst for such a sharp development

trajectory, it is clear that much that is classified as new within the Early Historic

period, such as settlement types and agricultural technology, have less to do with

religious belief than secular development within the later Iron Age.

8.6 Ritual and religion

It is likely, however, that pre-Christian religious ritual was an essential part of the

performance played out on a seasonal or cosmological basis at sites that today are

classified as ̀ historic' centres. The assessment of the environmental and faunal

material found at Dün Ailinne has demonstrated that the Ist century AD activity took

place in late Autumn (Johnston and Wailes 2007). The autumn festival was Samhain

and it is remembered even today in Ireland through the name for Halloween, Oiche

Shamhna. Archaeology has demonstrated that the rituals which took place at Rdith na Rig on Tara, and its associated site of Räith na Serer Raffin and Emain Macha were

performed within the confines of massive wooden structures (Johnston and Wailes

226

2007; Newman 2005; Grogan 2008). Diin Ailinne appears to have been one of the

earliest sites to be used in this way and its repeated rebuilding supports an interpretation that preparations were made in advance of specific ceremonies, which included the task of rebuilding an encircling timber structure each time. Emain Macha

was systematically destroyed through a deliberate fire in the first century BC, yet

there is later activity at the site suggestive of a continued link after its heyday (Lynn

2003). The context within which partial human remains have been found at some of

the historic sites suggests that these may have formed part of the ritual enactment, and

it is clear that at Haughey's fort and Raffia the remains were objectified, or tokenised,

in some way. This is especially true of the partial human skull found deposited

directly below a squat pillar stone at Raffin Fort in County Meath. When dated the

skull proved to be significantly earlier than the dates derived from the charcoal within

which it was stratified, suggesting this was a structured deposit which was then

marked by the pillar stone (Newman et al 2007).

The level and extent of mutilations on bodies found within the bogs in Ireland, which

are often preserved with ligatures still binding them around the neck, wrists and feet,

suggests that these individuals had become vessels upon which a ritual was enacted

(Kelly 2006). The majority of the bog bodies date to the Middle and Later Iron Age

and so are broadly contemporary in date with the sacrificial deposits at the King's

Stable and Raffia. The exceptional anaerobic conditions of the bogs preserves flesh.

wounds, but in the absence of skeletal evidence, no meaningful comparison can be

made between these and burials from dry land. If, however a body had been bound or

trussed, as with some of the examples from the bogs, the skeletal positioning would

reflect this. The only burial site where many individuals appear to have died from

multiple weapon injuries is that of Owenbristy in Galway, where the burials have

been dated from the 6t' to the 12"' century AD (O'Brien 2009: pers. comm). The

practice of placing people into the bog pools is believed to have some form of

shamanist characteristic and it has recently been argued that the deaths of these

individuals were performed at key boundaries in order to protect tribal territories

(Kelly 2006).

227

One of the most recent bog bodies found

in Ireland is known as Clonycavan Man

(Fig. 8.5), and he has been dated to 392-

201 cal BC (Kelly 2006). Analysis

revealed that he was unusually short in

height, standing at just over five feet and

that he had used a form of hair gel made

from a species of pine tree that is not

found in Ireland, but is found in southern

Spain. Kelly proposed that the use of the

gel to position the hair high up on the

crown of the head, was remarkably

similar to information recorded by

Tacitus about the Germanic tribe the Suebi

and their distinctive hairstyle, the Suebian Fig. 8.5 Cloncavan Man (photo by Robert Clark, National Museum of Ireland)

knot (Germania. 38; Kelly 2006). The term Suebi, like the term `Celt', is a collective one, which in reality encompassed many

peoples from this part of Germany. A link, however, is strengthened by another

reference in Tactius, (discussed in Chapter 4) in which he records that one population

group who formed part of the Suebi were the Chauci, and as discussed earlier, a

similarly named group were listed as settled in the area in which this bog body was

found on Ptolemy's map (Geographia H.. 1; Germania. 38). The closest parallels for

the massive oak road excavated from the bog in the Midlands at Corlea (dated to 148

cal BC), were the similarly built roads from within the territory of the Chauci, in

Lower Saxony in Germany (Raftery 1996c: 421). It seems reasonable to suggest that

that there is a link between the communities recorded by Ptolemy as living in the

Midlands of Ireland in the 2°d century AD and similarly named people in Europe

known to Tacitus in the Is` century AD.

Although the small rectangular structures found at Räith na Senad have been

interpreted as residential buildings, this seems unlikely. Both the Roman artefacts and

the faunal assemblage indicate that feasting was taking place at Räith na Senad and

we can now date the activity to between the 2nd and 4t' centuries AD (Grogan 2008;

228

Evans 2008). In line with Bhreathnach's reconsideration of the etymology of its

name, Teamhair, (from templum) the published faunal assemblage was dominated by

pig, the consumption of which is highlighted throughout the Irish texts as associated

with feasting. Rdith na Senad was not used as a place of the dead during the later Iron Age; it was a place of the living, and high living at that (Bhreathnach 2005).

During the 4t' century AD at a time when the rest of the provinces appear to have

gone into decline, Britannia had a thriving economy, with the creation of great farming estates and displays of civic pride through the building of walls around its

towns. The Roman ideal of the civitas administration appears to have flourished in

parts of the British midlands and southwest, and temples would have teemed with

cosmopolitan visitors. This is evident through the votive deposits left behind at these

shrines, amongst which are objects and items which can be paralleled in Ireland, such

as the penannular brooch described as of Irish-style found at Bath and the small silver handpin found at the mansio in Tripontium and dated to the 4m century AD (Wedlake

1984; Bathurst 1972; Cunliffe 1988a; Johns 1996; Roche 1993). The devotees who

came to these centres were not just practicing religious ideology, but they were also

subject to the commerical opportunism that visitors present, attested by the sheer scale

of the ancillary buildings at Lydney. In recent work Kurzmann has explored whether

there is evidence for cult practices that might link Ireland and the Roman provinces (I

am grateful to Dr Kurzmann for making her paper available to me, before

publication). The interpretation offered for the small collyrium stamp from Golden is

consistent with interpretations offered for finds of similar stamps common from

shrines in the Roman provinces (Daffy 2002; Kurnnann, forthcoming). These stamps

reflect the problems associated with eye conditions in the northern provinces (Boon

1983 Birley 1992: 115,117) and they are often found within the same contexts as

miniature body parts, at sites where the sculptural or dedicatory inscriptions have

identified deities who had a healing role. In Roman Britain, examples of both have

been found at religious centres such as Wroxeter, Lydney, Uley, Aqua Sulis, and Nettleton (Bathurst 1972; Cunliffe 1988; Wedlake 1982).

Clustered as these temples are along the length of the River Severn, they are dedicated

to the god of healing (Apollo) and water (Neptune). This links well into what we know of dedications and mythical links to Nechtan/Ni(adü in Ireland and the

229

deposition at Newgrange on the banks of the great sacred River Boyne. There are

more direct parallels between the material found at Newgrange and sites in Roman

Britain, the most obvious being with Nettleton Scrubb in Wiltshire. Here, not only is

the type of material deposited virtually identical to that from Newgrange, small gold

plate brooches with a conical glass inset and gold bracelets, but the links stretch further with the dedications to Mars Nodens, alongside the principal cult deity of Apollo (Wedlake 1984). In historic tradition Newgrange was the centre of cult focus

for the Irish god, An Daghda Mör (the great God), and it is linked to the cult of Nechtän, believed to be an early Irish borrowing of Neptune (Ö h' Ögäin 2001). It is

also associated with Lugh (Lugg in the Roman provinces), who in Irish mythology

was the son of the Irish deity Nüadü Airgetldmha (tr. Nüad4 silver-hand) (Ö h'Ögäin

2001). Nüadü is cognate with Nodens and was the son of Belenos (sun god) who is

syncretically linked through interpretatio roman with Apollo and Mars in Romano-

British and Gallo-Roman cults (Green 2005: 119). I discussed the inscription

SCBNS. MB, on a gold tort from Newgrange in Chapter 5. The first part is likely to

reflect the name Scribonius/ Scribonia and the second part may reflect a dedication to

Mars Belenos/Belatucadrius, as similar abbreviations are known from Roman

contexts in Gaul (Kurzmann forthcoming). At the cult site of Lydney in

Gloucestershire where dedications to Nodens are known, worshippers offered similar jewellery to that deposited at Newgrange, along with little votive body parts, including arms and hands which may link into the Irish name for Nodens ie. silver- hand. Although the association of a hound or a dog as the familial of Nodens is

known in Roman Britain, such as at Lydney where the finding of a small votive hound was interpreted as such, there are no existing depictions or representations of dogs in the the Irish archaeological record.

The Roman material from Freestone Hill also suggests a ritual or religious function

for the activity at the site (O'Floinn 2000). The limited type of material found,

bracelets and a 4th century Roman coin, is similar to religious deposition at Lydney

where parallels can also be offered for the types of contemporary bracelets found at Freestone Hill (Bathurst 1972). There is, however, a question of scale between the two sites although the quantity found at Lydney may simply relate to the length of time the site was in use and the number of visitors. Deposition at Newgrange also offers a clear comparison with many of the temple shrines along the length of the

230

Severn Valley, although the quality of the material at Newgrange remains exceptional

even within the Roman world.

Although votive coin deposition is known across ritual or religious sites in Roman

Britain, there are clear differences between the high value coin and jewellery

deposition at Newgrange and that from, say, Uley in Gloucestershire. The coins found

at prehistoric sites around Britain such as the West Kennet Long Barrow are often low

value potins (Woodward & Leach 1993; Aitchison 1988). Not everyone would have

had access to high status objects, however, and importance may have been placed on

the enactment of the votive offering regardless of the value of the object deposited. At

the temple site of Uley there is further evidence that links Romano-British worship to

Ireland through what was interpreted as the remains of a centrally placed sacred tree,

around which the depositions were made. The sacred tree in Ireland is known as the

bile and it is remembered, even today, in toponymic association in townland names

such as Bellow (Bile) and Moyville (Maigh Bile), (tr. the plain of the sacred tree) in

County Meath (Charles-Edwards 2000; Flanagan & Flanagan 2002: 30). Given the

sacred nature of Newgrange, and the links to an Daghda Mör and Nechtän (Neptune),

the nature of the deposition at the site and its dating, the parallels in the Roman world

are substantive (b h'Ögäin 2001).

S. 7 Language and literacy

In Chapter 2, I outlined in detail the origin of many words borrowed into Irish from

Classical Latin and many appear to have an association with either the military or

mercantile trade (McManus 1983: 42). They cannot be explained as later borrowings,

as the changes happening in spoken and written Latin itself dictated how these were

tranliterated into spoken and written Irish within a different form (McManus 1983).

There are several sites mentioned throughout the previous chapters with evidence of

earlier occupation sequences within which writing implements, notably stili, have

been found, such as those from Garranes, Garryduff, Lagore and Knowth. These may have been more widespread but as a stylus is characterised by its flattened end, once broken, it may have been classified as a pin or brooch end. Further evidence of literacy may be inferred from the inscribed tore end from Newgrange, dated by its

association with Roman coins to the 4"' century AD (IA2), and there is a small Latin

231 'yam ` , rs'

inscribed stone from Lough Derg, discussed in Chapter 4 (IA4). The evidence from

ogham inscriptions suggests that this unusual form of Irish epigraphy had a floruit

well before the Christian centres of learning and manuscript tradition were

established, as the small example from Rabin fort is dated to around the 3rd century AD by the context in which it was found and the ogham found at Silchester dates to

the end of the 4m century AD (Fulford et al 2000; Newman 2007). Whilst scholars

may continue to argue about an absolute chronology for the ogham inscriptions, it is

clear that early literacy in Ireland was based directly on both written and spoken Latin

and the grammatical conventions used in Classical Latin in the early centuries AD

(McManus 1991; Swift 1997; Harvey 1987). There need not be an exclusive link

between the origin of literacy in Ireland and the establishment of religious houses by

the literate class in 7'h century AD. The need to demontrate this essential Christianised origin may have led to the misdating of both the ogham inscriptions and the evidence for early borrowings from Latin into spoken Irish, both of which predate

the Christian centres by several centuries (McManus 1983: 24).

8.8 At the edge of the known world

The ripples of the Roman economy stretched far beyond the Italian peninsula, to the

dusty streets of Dura Europus in Syria, to Koihapur and Arikamedu in India, and

across the wild green seas to Britannia (James 2004; Pearson 2003). Amidst the

classical histories we have glimpses of tribal leaders who found themselves in the

front line of Roman expansionist policies. In order to preempt full scale attack they

made conciliatory moves through offers of tributes, although Roman pragmatism in

such cases did not always act as a deterrant to annexation. We know from the

writings of Tacitus that Agricola had in his company a `regulus' or prince of a kingdom in Ireland, from which he had been exiled due to political problems (Agricola 24). In Chapter 2, I mentioned the disparity that is evident in the English

translation from the original Latin and which may have altered the way in which the

passage has been interpreted in the past (Robinson 2000). The use of the Latin verb `exceperat', has traditionally been translated as ̀ had with him' (an exiled Irish

prince), but it is more correctly translated as ̀ had taken out' (of Ireland); the tense, is

apparently pluperfect and it has been consistently mistranslated (Robinson 2000: 27)

232

In recent years much has been written about the possibility of a small unofficial incursion into Ireland around the time of Agricola's campaigns on the northern front

in Britain. In Chapter 5, I reconsidered the basis on which a hypothesis for direct

military intervention by an Irish-led Roman auxiliary force has been proposed (Warner 1995). The sites and finds used to support this theory cannot be classified as

militaristic, and the finds are not comparable with each other nor are they

contemporary with a postulated 1st century AD date. Although the principal evidence

on which this hypothesis was based, i. e. that the Roman material at these sites is

military in character, is no longer supportable, what has been successfully demonstrated is that a political power base was already established within these

historic centres by the Ist century AD. The continuation of imports from the Roman

world into these sites suggests that this was an on-going relationship. It is entirely likely that those communities in Ireland, closest to Britain, who Tacitus recorded as

trading with merchants in Britannia, had been warned of Agricola's intentions and

came to an agreement with him. Tacitus records that ongoing trade between

merchants was a useful way of gathering intelligence for the Roman military, and it

would have proved useful for those in power in Ireland also. The sea did not act as a barrier in the period under discussion here, but much more likely is that the short

crossing between Drumanagh on the coast of County Dublin and Meols on the

Wirral, near the vast Roman naval military base of Deva Victrix, created the ideal

circumstances for diplomatic interaction, albeit veiled with the possibility of a

military advance should those in Ireland become uncooperative.

The material culture of those interacting in frontier communties or communities that

lie within a zone of influence but not under direct control of a foreign power, should

not be seen as dichotomous (Gardner 2007). Nor should differences in material

assemblages be used to assert a community's acceptance or resistance to foreign

influence in a dichotomous sense (Gardner 2007; Gosden 2004). Material moves even beyond a physical barrier and it can serve different functions inside, within and beyond the barricades. One modem ethnographic parallel within the island of Ireland

itself is the political and physical frontier devised to surround six of the 32 counties to

create ̀Northern' Ireland The barricades and armed British troops may have been

resented, and in some cases violently resisted, by people on the southern side of the border but it did not stop families such as mine driving through check points at the

233

weekend to shop in Marks and Spencers and filling up with cheap petrol. Frontier

communities are engaged in a dynamic push/pull of interaction, and takeup of aspects

of social cultural influence and these will inevitably vary according to local

circumstance and needs (Gardner 2001; Gosden 2004).

The line that the linear earthwork known as the Black Pig's Dyke takes across the

northern counties suggests that a frontier already existed between the communities in

the Ulster and Leinster. These massive earthwork constructions have been dated to the

ist century AD and they reflect a collective community activity of shaping the

landscape to protect people and their livestock at a time of political conflict both

within Ireland and beyond her shores.

8.9 Becoming `Irish': it transcultural process

Looking beyond Roman Britain to its closest neighbour Ireland should not require a

significant leap in archaeological interpretation, but the blank maps of Ireland which

accompany published work on Roman Britain suggest it is still outside any serious field of consideration and scholarship. In past considerations of Roman influence on Ireland authors have based their arguments on the essential difference of `exotic'

material from that of local origin but in seeking to assess the material evidence in this

way, authors have completely ignored the similarities between the consumption of

earlier exotica of say Halstaft or La Tbne derivation (Raftery 1994,2006). The

theoretical framework preferred throughout this work was one of transcultural

identities, specifically designed to avoid the juxtaposition of one discrete cultural

collective against another. The classification and separation of artefacts based on

parameters of cultural difference has led to exclusion or inclusion into one category or

another, rather than the material being assessed as holistic and contemporary in use. Regionalised approaches are necessary, as the overarching nationalist interpretation

for Ireland as a whole has led to further gaps in the chronology. In reality, the iron

Age of Munster follows a different social development trajectory to that of the east

and southeast, which is in turn different to that from the north of Ireland. In the early Iron Age, there is little evidence of centralised activity or occupation in the south and west of Ireland, suggesting that transition came gradually to these parts of Ireland.

234

The distribution of the earlier Halstatt and La Tene style objects in the early and

middle Iron Age is dotted along river routes, like the Shannon from the southwest and into the Midlands, and the Bann in the north. Whatever the circumstances of their loss

or deposition it suggests that imports were being moved inland along the river routes. Although the La Tene style objects have a northerly distribution, with only a few

found in Munster or southern Leinster, the limited evidence has been interpreted as a discrete cultural phase in Ireland and classified as an overarching national `La Tene

period' (Raftery 1994,2005). From the turn of the first millennium AD, imports

continue to cluster along other riverine routes. Romano-British material clusters in

the region of the River Bann and Lough Neagh in the north (notably this is mostly Saurian ware), along the River Boyne to Knowth and Newgrange, up the River

Gabhra to Tara, around the River Barrow in southeast and well inland into Tipperary.

It is also dotted at sites in the Midlands within easy reach of the Shannon and there is

a small Roman-style boat from the inland site of Lough Lene, in County Westmeath,

dated to the 2d or P century AD (Brindley and Lanting 1990). I am not suggesting

population continuity over such a long period of time, but rather a practical long-term

knowledge of both the safest seafaring routes and the estuaries from which access

could be gained to the interior of Ireland. The relative scarcity of diagnostic Iron Age

finds from dry-land sites in Munster, compared to those recovered from watery or

ethereal contexts, may reflect the continuity of established and well evidenced later

Bronze Age and Iron Age depositional practices (Grogan 1999). The regionalised distributional patterning of imported material from both before and after the Roman

period in Britain has been, in the past, relegated to `intrusive, and treated not only as

separate to each other, but separate to the rest of the archaeological evidence

classified as of `local' Iron Age. Yet the artefact assemblage of the early centuries AD, albeit limited, is much more closely datable (IA2) than, say, undecorated metal

work of `local' manufacture, and it should be regarded as essential to our

understanding of the relations between Ireland and her closest neighbour, Britain.

Generalisations and assumptions form the backdrop to even contemporary analysis,

with few scholars pausing to question, how these assumptions were formulated, where and how they originated, or even whether they remain relevant in modern analysis. There is no place in modem archaeological study for the use of old classifications

such as ̀ early Christian' ring-forts, (most recently used by Kerr in his published PhD

235

thesis in 2007) for there are many thousands of ringforts in Ireland whose occupants

may or may not have been Christians. We cannot accept the continued classification

of a site, or monument, based on the presumed faith of its occupants, especially where there is not a single shred of supporting evidence that would indicate their religious

predisposition. This is not a criticism which is made lightly, given that modem

scholarship is founded on a set of principles that rely on typological classifications

against which we are taught to read the archaeology being uncovered first hand, or

reassessed through previously published work.

Our reliance on the existing dating schemes and their persistent use, despite attempts

to challenge them, has resulted in an archaeology that is rigid and inconsistent in

interpretation. Understanding the people of Iron Age Ireland remains problematic, as

outlined in the previous chapters, but this is not just because of the fragmentary

remains left to us; it is due, to a great extent, to how the archaeology has been

interpreted in the past. The Iron Age has been read in relation to its material remains

and found to be a pale reflection of that which came before it in the glorious Bronze

Age and indeed that which followed, heralding the onset of the new religion of Christianity and the emergence of seats of learning and literacy within ecclesiastical

centres. The parallel between the perceived civilizing influence of Christianity in

relation to Irish archaeology and the concept of the civilizing influences of the Roman

Empire in Britain becomes very apparent within a study such as this (Hingley 1997:

85). Only for Ireland authors appear to have forgotten that Christianity in the 4m and Sh centuries AD was a fundamentally Imperial Roman religion. As a consequence of

our reliance on constructed temporal epochs, the importance of the transition taking

place within and between the centuries, and the external cultural influences which helped Ireland become a focal point for Christian learning, has all but been ignored.

Ireland was not isolated in the Bronze Age and nor was it throughout the Iron Age, as

evidenced through the regularity with which material classified as ̀ intrusive' has been

located at sites, as outlined in Chapter 5. Whether material was imported in the 2°d

century AD or 4t° century AD, it can no longer be classfied as intrusive as it is an

essential component of the materiality of Iron Age Ireland and it needs to be

rehabilitated within it, rather than as an adjunct to it. Some of the material is Roman,

or more probably Romano-British, provincial in origin and part of the wide

236

dissemination of products reflecting the cultural hegemony spread throughout those

areas encompassed within the Roman Empire and beyond. The ripples of the Roman

economy spread far beyond its formalized limes which stretched to its greatest extent in the time of Emperor Trajan in the 2d century AD. Imported material came as part

of a package of cultural influence, alongside those whose livelihoods depended on the

trade systems, the merchants and travellers pedalling their wares, and the information

they carried with them about the mighty Roman Empire.

The sticking point in the Warner debate of the 1990s rested on a single question: were

the Romans in Ireland? The answer to that question relies on a further question: what

or who exactly do we mean by `Romans'. To answer this, the critical dialogues that

have taken place within Roman studies in recent years are essential to the

deconstruction of a general misconception within archaeological study in Ireland

about what is meant by the term `Roman' and the concept of `Romanisation'. The

most important result of the application of new post-colonial models and theoretical

frameworks is that historiography has been replaced to include the consideration of

the `others', the vast majority of ordinary people who farmed their land and foddered

their cattle. The resultant collapse of the use of binary oppositions, previously

endemic in Roman studies and the acceptance of variability in response to Roman

influence has undermined the ascription of the generalised term `Romanisation'. This

has been replaced with models of human social interaction that consider localised

responses to new cultural influences and allow for a more holistic approach that

reflects the multicultural nature of communities in the Roman provinces. In trying to

reconstruct a past for local people we need to see transitional change through their

material culture, but it must form part of a broader consideration and be used

alongside equally valid archaeological methods that assess other social and economic indicators, retrieved through burial evidence and landscape studies.

In Chapter 2,1 stated unequivocally that the premise behind this research was to build

upon the existing narrative for Ireland. In so doing, however, it was essential to

highlight what I regard as the inadequacies and inconsistencies endemic within the

existing culture historical approach and its reliance on outdated chronologies and typologies as set out in Chapter 4. I discussed how the historical narrative was created in Ireland against the background of separation from Britain, stengthened by the

237

demarcation of a border around six of its 32 counties, and the social and political

aftermath of rebellion and civil war. Archaeological investigations were directed at finding Ireland's `Celtic' past and in the process, what was essentially a myth-history, became part of a nationalist perception of an indigenous past. As those in government in Ireland learnt to articulate a place as a modern nation state they turned to events

which had defined its more recent troubled past, and looked to the distant past to find

a more heroic age, one which did not include British rule. For deValera and MacNeill

the evidence of the ̀ Celtic' past lay firmly embedded within the previously

unexplored monuments that dotted the landscape throughout Ireland (O'Sullivan

1998: 182).

The need to give a voice to indigenous archaeology has encouraged contributions from all sections of society in the creation of a narrative which more closely reflects the multiple ways in which the archaeological interpretation may be formulated

(Bahn & Paterson 1986: 265; Parker-Pearson 1999). With issues of ownership come issues of how voices should be heard and to what extent different voices should bear

greater weight in the overall consideration (Hodder 2008: 198). I mention this here,

because the pressing need for an indigenous archaeology was created through a

reactive move against (western) imperialist narratives, that have in the past, been used

to tell only one side of the archaeological story (Trigger 2008, Hodder 2008). In this

respect, we might regard the Irish narrative as reflecting an indigenous archaeology bound up as it is with nationalism, colonialism and imperialism. The archaeology

was not, however, allowed to inform the interpretation of prehistory in Ireland; there

was no ground swell of public opinion that insisted on how it was interpreted, instead

it was used to support an already agreed national political ideology. A post-colonial

approach such as that presented here acknowledges that the archaeological interpretation for the Later Irish Iron Age has maintained the continuation of binary

oppositions and collective cultural constructs, which do not reflect the contemporary theoretical approaches of scholarship outside Ireland. The absence of `straight roads, bath houses and mortared buildings' was used by Ryan to completely refute any `Roman' involvement in Ireland (2001: 250) and this reflects how Roman Britain is

characterised or perhaps caricatured in Ireland. This is not only relevant to the Iron Age, however, as there is a marked reluctance of scholars in Ireland to regard the

238

archaeology from the high medieval period to independence, as reflecting anything

other than the ̀ archaeology of the British' (O'Keefe 2004; Horning 2006: 185).

What we have in Ireland is a complex interaction between those regions and people that engaged in different ways with Roman Imperial policy and those who did not. There is no singular explanation for the circumstances of how and why Roman

material was transferred through to those areas outside the frontiers, and the

variability in the evidence from beyond even the Rhine itself calls into question any

attempt to map Roman material beyond the buffer zone of interaction within a

generalised model. The Roman Empire created a political, military and administrative

system that encompassed different peoples from all of its provinces and from areas

well beyond its frontiers. The work done by scholars for Roman Britain, discussed in

Chapter 3, has started to disentangle who these people were, and where they

originated, from generalisations which have obscured them from our view. It seems

clear that a relationship had been negotiated between communities in the eastern

counties of Ireland and the vast might of the Roman military encamped a few short

miles across the Irish Sea. The Annals record the usurption of power within the 4d'

century AD in the Kingdom of Meath by the dynasty known as the Ui Nil. This shift in power in Ireland corresponds with the Classical accounts of the barbarica

conspiratio of AD 367 and the records of the first raiding taking place on Roman

Britain (Rance 2001). Notably Marcellinus records that a pre-existing agreement between the Romano-British administrators and the Scotti had been broken. If we

accept that the powerful Uf NU had an origin in the north of Ireland then this offers

an explanation for the displaced group known as the bitheachtüatha (the Attecotti)

known from the classicial records as conscripted into the Roman army (History

27.8.5; Rance 2001). Furthermore, the silver hoard found at Balline in southern Ireland (see Chapter 5) is more characteristic of an official payment than the rewards

of random plunder (Painter 1971; Raftery 1994; Warner 1995).

I discussed earlier Ptolemy's map of Ireland and his listings for community groups

and their settlements all over Ireland. Many of these names are more familiar to us from discussions of Roman Europe: the Menapii and the Chauci are from Europe and the Brigantes from northern Britain, although all are recorded as settled in Ireland at this time. The etymology of County Fermanagh, fir manch, can be translated to the

239

`men of menapi', and the neighbouring county of Monaghan is simply `mansch'

(remembering that Gaelic at this time did not have a ̀ p', and it would have been

replaced with a ̀ kw' sound, ̀ch'). The controversial site of Drumanagh in County

Dublin comes from the Irish `druim manach'tr. the place (or rise) of the menapi (Warner 1995; Mongan 1995). The most famous individual of Menapian origin was Carausius who usurped power in AD286/7 and declared himself Emperor of Britain

and Gaul (Frere 1987; Mongan 1995). Coins issued under Carausius are rare, but it is

notable that a silver coin of Carausius was found in County Kildare (published in

1898, its whereabouts are now unknown) (Bateson 1973; 52 citing Day 1898; 51).

The Menapii were a maritime trading people within the Roman period and it is

entirely possible that they were engaged in trade with Ireland at this time.

So how then do we operationalise a theoretical framework such as that presented here

in order that we can gain new insights into both the materiality of the Later Iron Age

and the identities forged within and between the communities in Ireland and beyond

its shores. As stated in Chapter 3 we have to start by accepting that the concept of independent ̀Celtic' Ireland is a myth-history, designed to validate the ideology of

politics and nationalism and it does not reflect the diversity of archaeological record in parts of Ireland. We can no longer classify individuals or communities within our

perceived limitations on their cultural homogenity, nor can we create a form of boundedness in terms of the limits of where they may have lived or died: migration happened, people moved throughout prehistory, as today. We cannot map burial

`type' against local or non-local practices to determine whether an individual or a

group are `intrusive' or `Celtic'. Nor can we assume that the presence of so-called `intrusive' material at a site somehow offers an identity for those buried close by.

The new radiocarbon dates and the scientific analysis presented in Chapters 6 and 7

completely undermine the existing relative dating method used for burial in Later Iron

Age Ireland. When we compare the strontium and oxygen isotope results for the faunal samples against those for the human burials, we can find amongst the variation in signatures some individuals who appear to have originated close to where they

were buried. Amongst the isotope evidence from the sites sampled are individuals

who have an origin in the northern and western counties of Ireland, which supports

many of the references in the Annals to a usurption of power by the northern Ui Neil in the 4th century AD. The isotope analysis has highlighted the probable geographic

240

origin for these individuals but we must remember that these people were buried in

Ireland during a period where the rite of formal burial appears to be the exception

rather than the norm. Many of these individuals have an origin outside Ireland, some

are consistent with an origin in Roman Britain, and yet they are morphologically indistinguishable to those classified as of `local' burial type.

What the isotope analysis has demonstrated is level of mobility in this period in

Ireland and this must have created the need for mediation in identity, the need to fit

in, or conversely, the need to remain distinctive within a new area or indeed a new homeland. But it does not follow that these people were regarded as ̀ foreigners' or

excluded from communities within that new cultural setting. Clear examples of such

mediation between identities, in both life and death, can be offered through

contemporary burials in Eboracum, Roman York The work of Eckardt and her

colleagues at Reading, has recently established, through scientific analysis and forensic anthropology, that the lady buried in a fine stone sarcophogus from

Sycamore Drive was a noble lady of African origin (Eckardt 2008; pers. comm). This

lady was buried with all the trappings of an elite Roman lifestyle along with a set of ivory bangles, objects which were originally interpreted as exotic imports. Eckardt's

work has demonstrated that these were not imports, they were aspects of her own

material culture, brought with her and worn as a reminder of her homeland (Eckardt

et al: pers. comm 2008). A further example, the famous Barates and Regina tombstone

from Arbeia, inscribed in Latin and Aramaic, recorded the multiple paths Regina had

taken, from a member of the Catauvallauni tribe, to slave woman, freed woman and finally becoming the wife of (her probable owner) Barates (Alcock 2006). This

commemoration of Regina, reminds us that often the manner of a burial, or objects found within a burial, can obscure an otherwise invisible former cultural affiliation. During the period within which this work is focussed, Ireland's closest neighbour was

under the ideological, economic, political and essentially militarist control of Rome.

Yet even in Roman Britain, in death individuals appear to have expressed, or had

expressed for them, an identity which often transcended their Roman-ness, whether through the choice of grave goods, burial practice or epigraphical commemoration.

241

The importance of Burial BTSXII from Betaghstown is that it links an individual who

originated in Southern Spain/Portugal or north Africa to the growing number of sites in the hinterland of Betaghstown where later imported ceramics have also been found.

Here we have the evidence of trade with the Mediterranean, or at least one individual

who accompanied his cargo to this coastal site in County Meath. The suggested points

of manufacture for both LRA and PRSW are in the eastern Mediterranean, but the

areas with the highest concentration of fords are the coasts of southern Spain/Portugal

and northern Africa from ARSW originated (Campbell 2007). This distribution

reflects the ports where cargo was placed onto ships built to ply the stormier waters

around the Atlantic coasts (Kelly 2010). In the same way as the relative chronology for the Iron Age can be challenged, so too can previous assessments that Ireland

remained outside a direct trading route from the Mediterranean in Late Antiquity. All

of the evidence presented within this research supports regular interaction and trade

between communities on both sides of the Irish Sea before, during and after the period

when Britannia was occupied and ruled by Rome.

From the Late Roman period areas on the western coast of Wales, Cornwall and Scotland were colonised by Irish migrant groups, discussed in Chapter 2, and an

additional form of evidence, the ogham inscriptions, record the individuals

commemorated in both Gaelic and Latin for posterity (McManus 1992; Charles-

Edwards 2000). In Ireland, the fording of a small ogham inscription at Raffin fort

within a dated 3d century AD context, suggests that this form of literacy was being

used much earlier than previous dating methods had indicated and it raised the

possibility that literacy was much more widespread, given that ogham is equally

suitable to notation on leather or wood (Newman 2007; McManus 1991). I mentioned

earlier that there are at least six examples of ogham inscriptions in Ireland that record Roman names and these are clustered at sites on the River Barrow and River

Blackwater in the south-east. Three of these record the Roman names Sagittari,

Mariani and Amadu (Amatus), and two of these are in the same area where the

Stoneyford Roman cremation burial and the collyrium stamp from Golden were found

in County Tipperary. The other three inscriptions record fathers with Roman names

who had given their sons Irish names; Dunaidonas son of Marianus, Coimagnas son

of Vitali(us) and Maqiddeccedas son of Marin(us) (McManus 1992: 113; Thomas

1998: 11). A complete understanding of the Later Irish Iron Age can only be achieved

242

if the evidence of imported material is incorporated alongside ostensibly local

material and viewed as a legitimate indicator of the transcultural influences that helped shape ancient Ireland. In a single generation, a seemingly Roman identity had

been replaced: these individuals were becoming ̀ Irish'.

243

Bibliography

Primary Sources

Annäla Rloghacta Eireann: Annals of the kingdom of Ireland by the Four Masters fr' om the earliest

period to the year 1616, translated by J. O'Donovan, Dublin 1951 (New York 1966)

Annals oflnishfallen, (MS Rawlinson B503), translated by S. Mac Airt (Dublin Institute of Technology, 1951)

Anndla Ülaidh, Annals of Ulster, from earliest times to the year 1541, introduction by N. Ö'Muraile,

2nd ed., (Blackrock, E. Burke, 1998)

Tdin b6 Ciialinge, edited by C. O'Rahilly Dublin 1976 (reprint 1984)

CELT, Corpus of Electronic Texts, Gaeilge and English, University College Cork, on-line data

resources for primary reference to the Irish Annals

Irish Script on Screen (ISOS) database: all manuscripts at Irish institutions including the Royal Irish

Academy, Dublin Institute of Advanced Studies

Ammianus Marcellinus, History Vol II, translated by John C. Rolfe, (London: Loeb classical library,

1937)

Caesar and Hirtius, The Conquest of Gaul, translated by SA Handford, revisions by J Gardner,

(Penguin Classics 1982)

Herodotus, The Histories, translated by A. de S6lincourt, revised by J. Marincola, (Penguin Classics

1996)

Tacitus: The Agricola and The Germania translated by H. Mattingly, revised by SA Handford,

(Penguin Classics 1970)

The Geography of Strabo, translated by H. L. Jones, (Loeb Classical Library 1932)

Geography of Claudius Ptolemy, translated into English, and edited by Edward Luther Stevenson

(New York: New York Public Library 1932)

Ptolemy's Almages* translated and annotated by G. J. Toomer (London: Duckworth 1984)

244

Bibliography

Aitchison, N. B. (1988) Roman wealth, native ritual; coin hoards within and beyond Roman Britain,

World Archaeology, Vol. 20,270-284

Alcock, L. (1972) The Irish Sea Zone in the Pre-Roman Iron Age, in C Thomas (Ed), The Iron Age in

the Irish Sea Province, CBA Report No. 9,99-112

Alcock, J. (2006) Life in Roman Britain (Stroud: Tempus)

Alcock, S. E. & Van Dyke, R. M. (2003) Archaeologies of Memory, (Blackwell Publishers)

Alekshin, V. A. (1983) Burial customs as an archaeological source, Current Anthropology, Vol 24, No.

2,137-149

Allason-Jones, L. (2008) Appendix A: The small finds, in Grogan E, The Rath of the Synods, Tara Co.

Meath, excavations by Sean P O'Rtorddrin, (Bray: Wordwell in association with University College

Dublin), 107-112

Allen, J. R. L. and Fulford, M. G. (1996) The distribution of South-East Dorset Black Burnished

Category 1 pottery in south-west Britain, Britannia, 27,223-281

Appedurai, A. (1986) Introduction: Commodities and the Politics of Value, in Appadurai, A. (Ed), The

Social Life of Things: Commodities in Cultural Perspective (Cambridge University Press), 3-63

Appleton-Fox, N. (1992) Excavations at a Romano-British round: Reawla, Gwinear, Cornwall, Cornish

Archaeology, No 31,69-123

Arnold, B. (1999) The contested past, Anthropology Today, Vol IS, No. 4,1-4

Arnold, B. (2001) The limits of agency in the analysis of elite iron Age Celtic burials, Journal of Social

Anthropology, Vol 1(2), 210-224

Arnold, B. (2001) Power drinking in Iron Age Europe, British Archaeology, Issue 57,30-33

Arnold, C. J. & Davies, E. (2000) Roman and Early Medieval Wales, (Stroud: Thrupp)

Bahn, P. G. & Paterson, R W. K. (1986), The last rights: more on archaeology and the dead, Oxford

Journal ofArchaeology, Vol 5(3), 255-271

245

Baldwin, R. (1995) Intrusive burial groups in the Late Roman cemetery of Lankhills: a reassessment of

the evidence, Oxford Journal of Archaeology, Vol 4 (1), 93-104

Bartel, B. (1980) Colonialism and cultural Responses: problems related to Roman provincial analysis,

World Archaeology, Vol 12, No. 1,11-26

Barrett, J. C. & Ko, I. (2009) A phenomenology of landscape: A crisis in British landscape

archaeology, Journal of Social Archaeology, Vol. 9, No. 3,275-294

Baillie, M. G. L. (1995), A Slice Through Time: Dendrochronology and Precision Dating, (Batsford

Press, London)

Bateson, J. D. (1973) Roman material from Ireland: a re-consideration, Proceedings of the Royal Irish

Academy, Vol 73C, 21-97

Bateson, J. D. (1976), Further finds of Roman material from Ireland, Proceedings of the Royal Irish

Academy, Colloquium on Hiberno-Roman relations and material remains, Vol. 76C, 171-180

Bathurst, W. H. (1972) Roman Antiquities at Lydney Park, (London: Longmans)

Becker, K., O'Neill, J. & Flynn, L. (2008) Iron Age Ireland: Finding an Invisible People (Dublin: an

Chomhairle Oidreachta) Published On-line

Bentley, R. A., Krause, R, Price, T. D. and Kaufmann, B. (2003) Human mobility at the early

Neolithic settlement of Vahingen, Germany: evidence from strontium isotope analysis, Archaeomeby,

Vol 45,471-486

Bentley, R. A. and Knipper, C. (2005) Geographical patterns in biologically available strontium, carbon

and oxygen isotope signatures in prehistoric SW Germany, Archaeomety, Vol 47,629-644

Bhreathnach, E. (2002) Observations on the occurrence of dog and horse bones at Tara, in Roche H,

Excavations at Rdith na Rfg, Tara, Co. Meath, 1997, Discovery Programme Reports, 6, (Dublin: Royal

Irish Academy/Discovery Programme), 177-122

Bhreathnach, E. (2005) The medieval kingdom of Braga, in Bhreathnach E (Ed), The Kingship and Landscape of Tara, (Dublin: Four Courts Press), 410 -421

Birley, A. R. (1992), A case of eye disease on the Roman frontier in Britain, Documenta

Ophthalmologka, Vol 81,111-119

246

Bourke, E. (1989) Stoneyford: a first century Roman burial from Ireland, Archaeology Ireland, Vol. 3,

56-57

Bourke, E. (1994) Glass vessels of the first nine centuries AD in Ireland, Proceedings of the Royal Irish

Academy, Vol. 124,163-209

Boon, G. C. (1983) Potters, occulists and eye-troubles, Britannia, Vol XN, 1-12

Boon, G. C. (1991) The coin evidence, in Holbrook N, Bidwell P T, Allason-Jones L (Eds) Roman

Finds from Exeter, Exeter Archaeological Society

Bowman, A. K. (1994) Life and letters on the Roman frontier: Vindolanda and its people (London:

British Museum Press)

Bowman, A. K., Garnsey, P. and Cameron, A. (Eds) The Cambridge Ancient History: the crisis of Empire AD 193-337, Vol 13, (Cambridge: Cambridge University Press)

Bradley, R. (1987) Time regained: the creation of continuity, Journal of'the British Archaeological

Association, Vol CXL, 1-17

Bradley, R. (2000) An Archaeology of Natural Places, (London & New York: Routledge)

Bradley, R. (2003) A life less ordinary: the ritualization of the domestic sphere in later Prehistoric

Europe, Cambridge Archaeological Journal, Vol 13 (1), 5-23

Braudel, F. (1972) The Mediterranean and the Mediterranean World in the Age of Philip 11, Vol I&2,

(New York: Harper & Row)

Brindley, A. L. & Lanting, J. N. (1990) A Roman boat in Ireland, Archaeology Ireland Vol. 4,10-11

Brogan, 0. (1936) Trade between the Roman Empire and the Free Germans, Journal of Roman Studies,

26,195-222

Brown, S. J. (1927) Poetry oflrish History (Dublin: Talbot Press)

Brubaker, R& Cooper, F. (2000), Beyond ̀ identity', Theory and Society, Vol 29, No. 1,1-47

Brack, J. (2004) Material metaphors: The relational construction of identity in Early Bronze Age

Burials, Journal of Social Archaeology, 4,307 - 333

247

Buckley, L. Unpublished pathology report on Burials at Rossnaree, County Meath

Budd, P., Chenery, C., Montgomery, J. & Evans, J. (2003) You are where you ate: isotopic analysis in

the reconstruction of prehistoric residency, in M Parker Pearson (Ed) Food, Culture and Identity in the Neolithic and Early Bronze Age, BAR International Series 1117,69-78

Burton, J. H., Price, T. D., Cahue, L. & Wright, L. E. (2003) The use of barium and strontium

abundances in human skeletal tissues to determine geographic origins, International Journal of Osteoarchaeology, Vol 13,88-95

Byrne, F. J. (1973) Irish Kings and High Kings, (London: Betsford)

Byrne, F J, Jenkins W and Kenny G (Eds) (2008) Excavations at Knowth: v. 4: Historical Knowth and Its Hinterland, (Royal Irish Academy Monographs in Archaeology)

Campbell, E. (2001) Were the Scots Irish? Antiquity, 75,285-92

Campbell, E. (2007) Continental and Mediterranean Imports to Atlantic Britain and Ireland AD 400-

800, CBA Research Report 157 (London: Council for British Archaeology)

Carew, M. (2003) Tara and the Ark of the Covenant: A search for the Ark of the Covenant by British-

Israelites on the Hill of Tara 1899-1902, (Dublin: The Discovery Programme)

Carson, R. A. G., & O'Kelly, C. (1977) A catalogue of the Roman Finds from Newgrange, County

Meath, Proceedings of the Royal Irish Academy, Vol 77C, 35-55

Caulfield, S. (1994) Some Celtic problems in the Irish Iron Age, in D. 6 Corrhin (Ed), Irish Antiquity,

Essays and Studies Presented to Professor MJ O'Kelly, (Dublin: Four Courts Prow)

Caulfiield, S. (1977) The beehive quern in Ireland, Journal of the Royal Society of Antiquaries of Ireland, No. 107,104-138

Chapman, J. & Dolukhanov, P. M. (1992) The baby and the bathwater: pulling the plug on migrations, American Anthropologist, New Series, Vol 94,169-174

Chapman, R., Kinnes, I. & Randsborg, K. (Eds) (1981) The Archaeology of Death, (Cambridge:

Cambridge University Press)

Chapman, R, (2003) Death, society and archaeology: the social dimensions of mortuary practices, Mortality, Vol 8, No. 3,305-312

248

Charles-Edwards, T. M. (2000) Early Christian Ireland, (Cambridge University Press)

(henry, C., Muldner, G., Evans, J., Eckardt, H., and Lewis, M., (2010) Strontium and stable isotope

evidence for diet and mobility in Roman Gloucester, UK, Journal ofArchaeological Science, Vol 37, 150-163

Cody, E. (2007) A Brief look at aspects of ringfort morphology and distribution, in Manning, C., (Ed)

From Ringforts to Fortified Houses: Essays on castles and other monuments in honour of David

Sweetman, (Dublin: Wordwell), 1-8

Collis, J. (2001) Across the great divide, in Champion, T& Collis, J (Eds), The Iron Age in Britain and Ireland., Recent Trends (Sheffield Academic Press), 1-4

Collis, J. (2001) Hill-forts, enclosures and boundaries, in Champion T& Collis J (Eds), The Iron Age in

Britain and Ireland. " Recent Trends (Sheffield Academic Press), 87-94

Collis, J. (2003) The Celts: origin, myths and inventions, (Tempus Publishing)

Comber, M. (2002) MV Duignan's excavations at the ringfort of Rathgurreen, County Galway, 1948-

9, Proceedings of the Royal Irish Academy, Vol 102 C, 137-197

Cooney, G. (1993) Lambay: an island on the horizon, Archaeology Ireland Vol. 7, No. 4,24-27

Cooney, G. (1996) Ireland, the Romans and all that (Editorial), Archaeology lrelan4 Vol 10,3

Cooney, G. (1999) Landscapes of the Neolithic, (London & New York: Routiedge)

Cotter, C. (1999) Western Stone Forts Project: Cahercummaun Fort, County Clare: a reassessment of its cultural context, Discovery Monograph No. 5 (Dublin: The Discovery Programme Publications)

Condit, T. (2004) Editorial, Archaeology Ireland, Vol 18,4

Conlin Casella, E. & Fowler, C. (2004), Beyond identification, an introduction, in Conlin Casella, E. &

Fowler, C. (Eds), The Archaeology of Plural and Changing Identities, (New York: Springer)

Creighton, J. (2006) Britannia: The Creation of a Roman Province, (London & New York: Routledge)

Crone, B. A. (1993) Crannogs and chronologies, Proceedings of the Society of Antiquaries Scotland, Vol 123,245-254

249

Crummy, N. (1979) A chronology of Romano-British bone pins, Britannia, Vol. 10,157-163

Crummy, P, Benfield, S., Crummy, N. & Rigby, V. (2008) Stanway: an elite cemetery at Camulodunum, Britannia Monographs 24

Curie, J. (1913) Roman and native Remains in Caledonia, The Journal of Rowan Studies, Vo13,98-

115

Curtin, N. J. (1996) "Varieties of h ishness": historical revisionism, Irish style, The Journal of British

Studies, Vol 35, No. 2, Revisionisms, 195-21

Cunliffe, B. (1988) Greeks, Romans and Barbarians: Spheres of Interaction (Batsford: London)

Cunliffe, B. (Ed) (1988a) The Temple of Sulis Minerva at Bath, Vol 2, The finds from the sacred spring, Oxford Monograph 16, (Oxford: Oxford University Committee for Archaeology)

Cunliffe, B. (1997) Iron Age societies in Western Europe and beyond, 800-140 BC, in Cunliffe, B. (Ed)

The Oxford Illustrated History of Prehistoric Europe, (Oxford University Press), 336-372

Cuniit%, B. (1997) The impact of Rome on barbarian society, 140BC to AD300, in

Cunliffe, B. (Ed) The Oxford Illustrated History of Prehistoric Europe, (Oxford University Press), 411-

446

Cunliffe, B. (2001), Facing the Ocean, The Atlantic and its peoples 8000 BC - AD 1500 (Oxford

University Press)

Daffy, S. (2002) A site for sore eyes: A Hiberno-Roman curative cult at Golden, Co. Tipperary,

Archaeology Ireland, Vol. 16,8-9

Dark, K. R. (1996) External Contacts and the Economy of Late Roman and Post Roman Britain,

(Suffolk: Boydell Press)

Dark, K. R. (2002) Britain and the end of the Roman Empire (Stroud: Tempus)

Daux, V., Lbcuyer, C., Heran, M., Amiot, R., Simon, L., Fourel, F., Martineau, F., Lunnerup, N.,

Roychier, H., and Escarguel, G., (2008) Oxygen isotope fractionation between human phosphate and

ari_ter revisited, Journal of Human Evolution, Vol 55,1138-1147

250

Davies, K. M. (1989) A note on the location of the Roman burial site at Bray, Co. Wicklow,

Archaeology Ireland, Vol. 3,108-109

De Marrais, E., Gosden, C. and Renfrew, C (2004) (Eds) Rethinking Materiality the Engagement of Mind with the Material World, (California: McDonald Institute for Archaeological Research)

Derks T and Roymans N (2002) Seal boxes and the spread of Latin literacy in the Rhine delta in

Cooley AE (Ed) Becoming Roman, writing Latin? Literacy and epigraphy in the Roman west, Journal

Roman Archaeology, Supplementary Series 48,87-134

Diefendorf, A. F. and Patterson, W. P. (2005) Survey of stable isotope values in Irish surface waters, Journal of Paleolimnology, Vol 34,257-269

Di Martino, V. (2003) Roman Ireland, (Cork: The Collins Press)

Dincauze, D. F. (2000) Envronmental Archaeology: Principles and Practice (Cambridge: Cambridge

University Press)

Dominguez, A. J. (2002) Greeks in Iberia: colonialism without colonization in, in Lyons CL and

Papadopoulos JK (Eds), The Archaeology ofColoniallsm, (Los Angeles: Getty Research Institute)

Doody, M. (2008) The Ballyhoura Hills Project, Discovery Programme Monograph

Doyle, I. (1998) The Early Medieval activity at Dalkey Island, Co. Dublin: a re-assessment, The

Journal of Irish Archaeology, Vol IX, 89-104

Doyle, I. (2002) Unpublished Excavations in the Republic of Ireland 1930-1997, Heritage Council of Ireland

Driscoll, S. T. & Nicke, M. R. (1988), Power and Politics In Early Medieval Britain and Ireland,

(Edinburgh University Press)

Driscoll, S. T. (1998) Picts and Prehistory: cultural resource management in Early Medieval Scotland,

World Archaeology, 30 (1), 142-158

Duke, S. (1998) The Romano-Irish of Tara, Technology Ireland, No. 3,16-18

Duday, H. (2006) L'arch6othanatologie ou 1'arch6ologie de Is mort, in Gowland, R

and Knusel, C. J. (Eds) The Social Archaeology of Funerary Remains, (Oxford: Oxbow), 32-52

251

Dupras, T. L. (2001) Strangers in a strange land: stable isotope evidence for human migration in the

Dakhleh Oasis, Egypt, Journal of Archaeological Science, Vol 28,1199-1208

Eckardt, H. (2002) Illuminating Roman Britain, (Editions Monique Mergoil)

Eckardt, H. & Williams, H. (2003) Objects without a past, in Williams H (Ed), Archaeologies of

Remembrance, Death and Memory in Past Societies, (New York: Kluwer Academic Press)

Eckardt, H. & Crummy, N. (2004) Regional identities and technologies of the self nail-cleaners in

Roman Britain, Archaeological Journal, Vol 160,44-69

Eckardt, H. (2004) Remembering and forgetting in the Roman provinces in, B. Croxford, H. Eckardt,

J. Mende & J. Wakes (Eds. ) TRAC: Proceedings of the Thirteenth Theoretical Roman Archaeology

Conference, Leicester 2003. (Oxford: Oxbow Books), 36-50

Eckardt, H. & Crummy, N. (2006) Roman' or 'native' bodies in Britain: the evidence of Late Roman

nail-cleaner strap-ends, Oxford Journal ofArcheology, Vol 25, No. 1,83-103

Edmunds, M. (1999) Ancestral Geographies of the Neolithic: Landscapes, Monuments and Memory,

(London & New York: Routledge)

Edwards, N. (1990) The Archaeology of Early Medieval Ireland, (London: Routledge)

Edwards, N. (2005) The archaeology of Early Medieval Ireland, o400-1169: settlement and economy,

in D 6'Cr6inin , (Ed) A New History of Ireland Prehistory and Early Ireland (Oxford University

Press), 235,296

Eggers, H. J. (1951), Der Romasche Import bit Freien Gemahlen, Hamburgisches Museum fur

Velkerlcnnd und Vorgeschichte, (Hamburg)

Eogan, G. (1974) Report on the excavations of some passage graves, unprotected inhumation burials

and a settlement site at Knowth, County Meath, Proceedings of the Royal Irish Academy, 74 C, 11-112

Eogan, G. (1977), The Iron Age and Early Christian settlement at Knowth, in Markotic, V. (Ed),

Ancient Europe and the Mediterranean: Studies presented in honour of Hugh Hewken, (Warminster:

Arris & Phillips Press)

Eogan, G. and Roche, H. (1997) Excavatiow at Knowth (2) (Dublin: Royal Irish Academy)

252

Eogan, G. and Roche, H. (2005) Ceremonial enclosure at Lugg, County Dublin, in Gosden, C.,

Hamerow, H., de Jersey, P. and Lock, G. (Eds) Communities and Connections: Essays in honour of

Barry Cunliffe, (Oxford University Press)

Eogan, J. (forthcoming) Excavation of an unenclosed Early Medieval cemetery at Bettystown, County

Meath

Erdrich, M. (1999) Continuity or discontinuity: native and Roman metal finds, in Besteman, J. C, Bos,

J. M., Genets, D. A., Heidinga, HA., and De Koning J., (Eds), The Excavations at Wynaldum: Reports

on Frisia in Roman and Medieval Times (Rotterdam: Balkema), 171-183.

Evans, J., Stoodley, N. and Chenery, C. (2006) A strontium and oxygen isotope assessment of a

possible fourth century immigrant population in a Hampshire cemetery, Southern England, Journal of Archaeological Science, Vol 33,265-272

Evans, J. (2008) Appendix G: The Roman pottery from the Rath of the Synods, in Grogan, E. The Rath

of the Synods, Tara Co. Meath, excavations by Sean P 6'Rfordäln, (Bray: Wordwell in association

with University College Dublin), 123-126

Featherstone, M. (1990) Global culture: an introduction, Theory, Culture and Society, 7,1-14

Featherstone, M. & Lash, S. (1999) Spaces of Culture: City, Nation, World (London: Sage)

Fisher, J. (Ed) (1994) Global visions: towards a new internationalism in the visual arts, (London: Kala

Press)

Flanagan, D. & Flanagan, and L. (2002) Irish Place Names (Dublin: Gill & McMillan)

Fowler, E. (1960) The origins and development of the penannular brooch in Europe, Proceedings of the Prehistoric Society, 26,14-177

Fowler, E. (1963), ̀ Cultic metalwork of the Fifth and Sixth Centuries AD, Archaeological Journal, Vol

120,98-160

Fredengen, C. (2003) A study of people's interaction with lakes with particular reference to Lough

Gara in the North West, (Bray: Wordwell)

Freeman, P. M. (1995) Greek and Roman views of Ireland: A Checklist, Emania: Bulletin of the Navan

Research Group, No. 13,11-13

253

Freeman, P. M (2001), Ireland and the Classical World, (Austin: University of Texas Press)

Frei K M, Skals I, Gleba M and Lyngstrom H (2009) The Huldremose Iron Age textiles, Denmark: an

attempt to define their provenance applying the strontium isotope system, Journal of Archaeological

Science, Vol 36,9,1965-1971

Frere, S. (1987) Britannia: A history of Roman Britain, 0 Fully Revised Edition, (Suffolk: St

Edmondsbury Press)

Frere, S. S., Roxan, M. & Tomlin, R. S. 0. (1990), in Collingwood, R. G. and Wright, R. P. (Eds) The

Roman Inscriptions of Britain (RIB), Vol 2, Instrumentum domesticum: (personal belongings and the

like), Fasc. 1: The military diplomata, metal ingots, tesserae, dies, labels and lead sealings, (Gloucester: Sutton)

Collingwood, R. G. and Wright, R. P. (1990) The Roman Inscriptions of Britain (RIB), Vol 2,

Instrumentum domesticum : (personal belongings and the like), Fasc. 1: The military diplomata, metal ingots, tesserae, dies, labels and lead sealings, (Gloucester: Sutton)

Fulford, M. G. (1977) Pottery and Britain's foreign trade in the later Roman period in Peacock DPS

(Ed) Pottery and Early Commerce : Characterization and Trade in Roman and Later Ceramics,

(London: Academic Press)

Fulford, M. G. (1985) Roman material in barbarian society c200 BC to AD400, in TC Champion and J

VS Megaw (Eds), Settlement and Society: Aspects of West European Pirehistory in the First

Millennium BC (Leicester University Press) 91-109

Fulford, M. G. (1989) Roman and barbarian: the economy of Roman frontier systems, in john C

Barrett, Andrew P Fitzpatrick and Lesley Macinnes (Eds) Barbarians and Romans in North West

Europe from the later Republic to late Antiquity, BAR International Series 471

Fulford, M. G. (1993) Byzantium and Britain: a Mediterranean perspective on Post Roman

Mediterranean Imports in western Britain and Ireland, Medieval Archaeology, 33,1-6

Fulford, M. G. (2007) Coasting Britannia: Roman trade and traffic around the shores of Britain, in

Gosden C, Hamerow H, de Jersey P and Lock G (Eds) Communities and Connections: Essays in

honour of Barry CunlWie, (Oxford University Press) 54-73

Frankenstein, S. & Rowlands, M. (1978) The internal structure and regional context of early Iron Age

Society in South-Western Germany, Bulletin of the Institute ofArchaeology of London, Vol 15,73-112

254

Galestin, M. C. (2010) Roman artefacts beyond the northern frontier: interpreting the evidence from the

Netherlands, European Journal of Archaeology, Vol 13(1), 64-88

Gardner, A. (1999) Military identities in Late Roman Britain, Oxford Journal ofArchaeology, 18 (4),

403-19

Gardner, A. (2002) Social identity and the duality of structure in late Roman-period Britain, Journal of

Social Archaeology, 2,323-351

Gardner, A. (2007). Fluid frontiers: cultural interaction on the edge of empire. Stanford Journal of

Archaeology, 5,43- 60

Garvin, T. (2004) Preventing the Future: Why was Ireland so poor for so long? (Dublin: Gill and

Macmillan)

Gilmour, G. (2002) Foreign burials in Late Bronze Age Palestine, Near Eastern Archaeology, Vol 65,

No. 2,112-119

Githens-Mazer, J. (2006), Myths and Memories of the Easter Rising: Cultural and Political

Nationalism in Ireland, (Dublin: Irish Academic Press)

Gosden, C. & Marshall, Y. (1999) The cultural biography of objects, World Archaeology, Vol. 31, No.

2,169-178

Gosden, C. (2004) Archaeology and Colonialism: Cultural Contact from 5000 BC to the present, (Cambridge University Press)

Gowan, M. (2005), The Lisheen Mine Archaeological Project 1996-8, (Bray: Wordwell Press)

Gowland, R. and Knusel, C. (Eds) (2006) The Social Archaeology of Funerary

Rehurins, (Oxford: Oxbow)

Graeme, M (1998) Rome with Romanization: cultural change in the pro-desert of Tipolitania (First - Third Centuries AD), Oxford Journal ofdrehaeolog+, 17(1), 93-111

Griffiths, D., Philpott, R. & Egan, G. (2007) Meols: The Archaeology of the North Wirral Coast

(Oxford University School of Archaeology Monographs)

255

Grogan, E. (1984) Excavations of an Iron Age mound, at Furness, County Kildare, Journal of the

Kildare Archaeological Society, 16,298-316

Grogan, E. (2008) The Rath of the Synods, Tara Co. Meath: excavations by Sean P. O'Rfordbin,

(Dublin: Wordwell)

Gudeman, A. and Haverfield, F. (1900) Agricola's invasion of Ireland once more, The Classical

Review, Vol 14, No. 1,51-96

Guido, M. (1978) The glass beads of the prehistoric and Roman periods in Britain and Ireland, Society

of Antiquaries London, Research Report

Habu J, Fawcett C and Matsunaga JM (2008) Evaluating Multiple Narratives: Beyond Nationalist,

Colonialist, Imperialist Archaeologies, (New York: Springer)

Halpin, A. and Newman, C. (2006) Ireland. - An Oxford Archaeological Guide, (Oxford University

Press)

Hanson, R. P. C. (1983) The life and writings of the historical Saint Patrick (New York: Seabury Press)

Harke, H. (1990), "Warrior Graves"? The background of the Anglo-Saxon weapon burial rite, Past and

Present, No. 126,22-43

Härke, H. (1998) Archaeologists and migrations: a problem of attitude? Current Anthropology, Vol 39,

19-45

Harding, D. W. (2006) Redefining the northern British Iron Age, Oxford Journal ofArchaeology, Vol

25(1), 61-82

Harding, D W. (2004) The Iron Age in Northern Britain: Celts and Romans, Natives and Invaders,

(London & New York: Routledge)

Harding, D. W. (2007) The Archaeology of Celtic Art, (London & New York: Routledge)

Harvey, A. (1985) The significance of Cothraiges, Eriii , 36,1-9

Harvey, A. (1987) Early literacy in Ireland: the evidence from ogham, Cambridge Medieval Celtic

Studies, 14,1-15

256

Harvey, A. (1990) The ogham inscriptions and the Roman alphabet: two traditions or one? Archaeology Ireland, 4 (1), 13-14

Harvey, A. (2001) Problems in dating the origin of the ogham script, in J Higgitt and K Forsyth (Eds)

Roman, Runes and Ogham: Medieval inscriptions in the insular world and on the Continent,

(Donington: Shaun Tyas)

Harvey, D. (1990) The Condition of Postmodernity: An Enquiry into the Origins of Cultural Change,

(Oxford: Blackwell)

Hartley, B. R. & Fitts, R. L (1988), The Brigantes, (Gloucester: Alan Sutton)

Haverfield., F (1913), Ancient Rome and Ireland, The English Historical Review, Vol. 28,1-12

Hawkes, C. F. C . (1981) The wearing of the brooch: Iron Age dress among the Irish, in B Scott (Ed)

Studies on Early Irelang Essays in honour of MV Duignan (Belfast: Association of Young Irish

Archaeologists)

Hedeager, L. (1978), A quantitative analysis of Roman imports in Europe north of the Limes (0-400

AD) and the question of Roman-Germanic exchange, in K Kristiansen and C Paludan-Muller (Eds),

New Directions in Scandinavian Archaeology, Studies in Scandinavian prehistory and Early history,

Vol. 1,191-217

Hedeager, L. (1987), Empire, frontier and the barbarian hinterland: Rome and northern Europe from

AD 1-400, in M Rowlands, M Larsen &K Kristiansen (Eds) Centre and Periphery to the ancient

world, New Directions in Archaeology (Cambridge University Press

Hedeager, L. (1992), Iron Age Societies, (Oxford: Blackwell)

Hodges, R. E. M., Housley, R. A., Ramsey, C. B. and van Klinken, G. J. (1993) Radiocarbon dates

from the AMS system: datelist 17, Archaeomety, 35 (2), 313- 316

Hedges, R. E. M., Housley, R. A., Ramsey, C. B. and van Klinken, G. J. (1995) Radiocarbon dates

from the AMS system: datelist 19, Archaeometry, 37 (1), 195-214

Hedges, R. E. M., Housley, R. A., Ramsey, C. B. and van Klinken, G. J. (1996) Radiocarbon dates

from the AMS system: datelist 22, Archaeomedy, 38 (2), 391-415

Hencken, H. O'Neill (1938) Cahercummaun: A stone fort in County Clare, Journal of the Royal Society of Antiquaries Ireland (Separate Edition), 38,1-82

257

Hencken, H. O'Neill (1938-39), Unpublished excavation notes for Lagore Crannog, Topographical

files, National Museum of Ireland

Hencken, H. O'Neill (1942) Ballinderry Crannog No. 2, Proceedings of the Royal Irish Academy, Vol

XLVII C, 1-76

Hencken, H. O'Neill (1950), Lagore Crannog: an Irish royal residence of the 7'h to 10th centuries AD,

Proceedings of the Royal Irish Academy, Vol 53C, 1-247

Henderson, J. (1987) The Iron Age of Loughey and Meare: Some inferences from glass analysis, Antiquaries Journal, Vol. 67,29-42

Henry, F. (1945) Remains of the Early Christian period on Inishkea North, Journal of the Royal Society

ofAntiquarians of Ireland, 75,127-155

Herity, M (1988) Rathcroghan and Carnfree: Celtic Royal Sites in Roscommon (Dublin: Na Clocha

Breaca)

Hill, J. D. (1995) The Pre-Roman Iron Age in Britain and Ireland (c. 800 BC to AD 100): An

Overview, Journal of World Prehistory, Vol 9 (1), 47 - 98

Highem, N J. & Jones, B. (1985) The Carvetti, (Gloucester: Alan Sutton)

Higham, N. J. (1989), Roman and native north of the Tees: acculturation and its limitations in, Barrett J

C, Fitzpatrick AP& Macinnes L (Eds), Barbarians and Romans in North-West Europe, BAR

International Series 471

Hingley, R (1992) Society in Scotland from 700 BC to AD 200, Proceedings Society Antiquaries

Scotland, Vol 122,7-53

Hingley, R. (1994) Britannia, origin myths and the British Empire, in S. Cottam, D. Dungworfh, S.

Scott and J. Taylor (Eds), TRAC The Proceedings of the Fourth Annual Theoretical Roman

Archaeology Conference, Durham University, (Oxford: Oxbow)

Hingley, R. (1997) Resistance and Domination: social change in Roman Britain, in Mattingly DJ (Ed),

Dialogues in Roman Imperialism: Power. Discowie and Discrepant experience in the Roman Empire,

Journal of Roman Archaeology, Supplementary Series no. 23,

258

Hingley, R (2005) Globalising Roman culture; unity, diversity and empire (London & New York:

Routledge)

Hingley, R (2010) Cultural diversity and unity: Empire and Rome, in Hales S and Hodos T (Eds),

Material culture and social identities in the ancient world, (Cambridge University Press), 54-75

Hodder, I. (1982) Theoretical Archaeology: a reactionary view, in I Hodder (Ed), Symbolic and Structural Archaeology, (Cambridge University Press)

Hodder, I. (1997) Always momentary, fluid and flexible: towards a self-reflexive excavation

methodology, Antiquity, Vol 71,691-700

Hodder, I. (1999) The Archaeological Process, (London: Blackwell)

Hodder, I. (2008) Multivocality and Social Archaeology, in Habu J Fawcett C and Matsunaga J M,

(Eds) Evaluating Multiple Narratives: Beyond Nationalist, Colonialist, Imperial Archaeologies, (New

York: Springer), 196 - 200

Hodder, I. and Hutson, S. (2003) Reading the Past: Current Approaches to Interpretation in

Archaeology, 3' Edition, (Cambridge University Press)

Hodges, R. (1982) Dark Age economics: the origins of towns and trade AD 600-1000, (London:

Duckworth Press)

Hodges, R (1988) Primitive and peasant markets, (Oxford: Blackwell)

Hodos, T. (2006) Local Responses to Colonization in the Iron Age Mediterranean, (London & New

York: Roudedge Press)

Hodos, T. (2010), Local and global perspectives in the study of social and cultural identities, in Hales S

and Hodos T (Eds), Material Culture and Social Identities in the Ancient Works (Cambridge

University Press), 3-31

Hollinrake, N. (2007) Dark Age traffic on the Bristol Channel, UK: a hypothesis, The International

Journal of Nautical Archaeology, 1 -8

Homing, A. J. (2006), Archaeology, conflict and contemporary identity in the North of Ireland:

implications for theory and practice in comparative archasologies of colonialism, Archaeological

Dialogues, 13 (2), 183-200

259

Howe, S. (2000) The politics of a historical ̀ revisionism': comparing Ireland and Israel/Palestine, Past

and Present, No. 168,227-253

Humphries, Mark (2002) In Mommsen's shade: Roman historiography, past and presen4 Classics

Ireland, 9,2845 (http: //eprints. www. nuim. ie/367)

Hunter, F. (1997) Iron Age hoarding in Scotland and northern England, in A Gwilt & Haselgrove C

(Eds), Reconstructing Iron Age Societies, Oxbow Monograph 71,108-133

Hunter, F. (2001) Roman and native in Scotland: new approaches, Journal of Roman Archaeology,

Vol. 14,289-309

Hutchinson, J (1987) Cultural nationalism, elite mobility and nation-building: communitarian politics in modern Ireland, The British Journal of Sociology, Vol 38,482-501

Hutchinson, J. & Aberbach, D. (1999) The artist as nation-builder: William Butler Yeats and Chaim

Nachman Bialik, Nations and Nationalism, Vol. 5,501-521

Hutchinson, J. (2001) Archaeology and the Irish rediscovery of the Celtic past, Nations and Nationalism, Vol. 7,505-519

Iacumen, P., Bocherens, H., Mariotti, A., and Longinelli, A. (1996) Oxygen isotope analyses of co-

existing carbonate and phosphate in biogenic apatite: a way to monitor diagenetic alteratio of bone

phosphate?, Earth and Planetary Science Letters, Vol 142,1-6

Jackson, K. N. (1964) The Oldest Irish Tradition: A Window on the Iron Age (Cambridge University

Press)

Jackson, R. (1996), A new collyrium stamp from Staines and some thoughts on eye medicine in Roman

London and Britannia, in J. Bird, M. Hassall & H. Sheldon (Eds), Interpreting Roman London.

(Oxford: Oxbow Books), 177-181.

James, S. (1999), The Atlantic Ce/tr: Ancient People or Modem Invention (London: British Museum

Press)

Janes, S. and Rigby, V. (1997) &kain and the Celtic Iron Age (London: British Museum Press)

James, S. (2004) Excavations at Dura-Europas, Final Report VII, the Arms and Armour, and other Military Equipment, (London: British Museum Press)

260

Js away, R. C. (1997) The decay of buried human remains and their associated materials, in Hunter J,

Roberts CA and Martin A (Eds), Studies in Crime: Introduction to Forensic Archaeology, (London:

18stsford) 58-85

JduW C. (1996) The jewellery of Roman Britain: Classical and Celtic traditions (London: Routledge)

Johnson R (1999) Ballinderry Crannog No. 1: a reinterpretation, Proceedings of the Royal Irish

Academy, Vol 99 (C), 23-71

Johnston, S. A. and Wailes, B. (2007) Drin Äilinne: Excavations at the Royal Site 1968 - 1975

( Ldelphia. Museum of Philadelphia, Archaeology and Anthropology Monographs)

Jones, B. & Mattingly, D. (1990) An Atlas of Roman Britain, (Oxford: Blackwell)

äoaes, A. (2003), Technologies of Remembrance, in Williams H (Ed), Archaeologies of Remembrance,

Death and Memory in Past Societies, (New York: Kluwer Academic Press)

Jones, M. J. (2004) Cities and Urban Life, in Todd M (Ed) A Companion to Roman Britain (Oxford:

Blaclcwell), 162-192

Jordan, K. and O'Connor, T. (2003) Archaeological sites of interest around the Turoe Stone, Journal of lit Galway Archaeological and Historical Society, Vol. 55,110-116

jwge sen, L. B. (1984) Journal of Danish Archaeology, Vol 3, (Odense: Odense University Prom)

Sensen, L. (2003), The Spoils of Victory, in Jorgensen L, Storpard B and Thomsen LG (Eds) The

dis of Victory. The North in the Shadow of the Roman Empire. (Copenhagen: Nationalmuseet), 12-

is

byes A. and Lopipero, J. (2005) Postscript: Doing Agency in Archaeology, Journal ofArchaeological

mod and Theory, Vol 12, No. 4,365-374

Kamm, A. (2004) The Last Frontier: Romar Invasions of Scotland (London: Tempus)

Kauf, F. (2003) The bog - gateway to another world, in Jorgensen L, Storgaard B and Thomsen LG

) The Spoils of Victory. The North in the Shadow of the Roman Empire, (Copenhagen:

jk*nahnused), 18-44

Kren, R. (2000), Tlk Green Flag: A History of Irish Nationalism, (Fully Rev. Ed), (London: Penguin)

261

Keely, V. (1996) Excavations at Ballydavis, Co. Laois, IAPA Newsletter, 22

Kelly, A. (2007) A Turkish import in County Meath, Seanda: NRA Archaeology Magazine, 16-18

Kelly, A. (in press) The discovery of Phocaean Red Slip Ware (PRSW) form 3 and Bii (LRAI) on site in Ireland: an analysis within a broader framework, Proceedings of the Royal Irish Academy

Kelly, E. P., Burials at Betaghstown,: Unpublished excavation notes, Topographical files National

Musuem of Ireland

Kelly, E. P. (2002) Antiquities from Irish Holy Wells and their wider context, Archaeology Ireland,

Vol. 16, No. 2,24-28

Kelly, E. P. (2006), Secrets of the bog bodies: the enigma of the Iron Age explained, Archaeology

Ireland, Vol 20, No. 75,26-30

Kerr, T. R. (2007) Early Christian Settlement in North West Ulster, BAR Series 430 (Oxford:

Archaeopress)

Kilbride-Jones, H. E. (1980) Zoomorphic Penannular Brooches (Society of Antiquaries London)

Killeen, J. F. (1976) Ireland in the Greek and Roman writers, Proceedings of the Royal Irish Academy,

Vol. 76C, 207-215

Koch, J. T. (2006) Celtic Culture: a historical encyclopedia (Santa Barbara: ABC-CLIO)

Koch, P. L., Tuross, N. & Fogel, M. L. (1997) The effects of sample treatment and diagenesis on the

isotopic integrity of carbonate in biogenic hydroxylapatite, Journal ofArchaeological Science, Vol 24,

417-429

Kunow, J. (1990) Relations between the Roman occupation and the Limesvorsland in the province of Germania Inferior in, Blagg T& Millen M (Eds) The early Roman Empire in the West, (Oxbow Press)

Kurtz, D. C. and Boardman, J. (1971) Greek Burial Customs (London: 'Blames and Hudson)

Kurzmann, R. (forthcoming) Irish sanctuaries and Roman sacrifices: evidence for cult connections between Iron Age Ireland and north-western Roman Provinces, in A. Morino (Ed. ) Limes XX-

Proceedings of the XXth International Congress on Roman Frontier Studies (Leon)

Lane, A. and Campbell, E. (2001) Dunadd. An early Dalriadic capital (Oxford: Oxbow)

262

Lapidge, M., Blair, J., Keynes, S. and Scragg, D. (Eds) (2001) The Blackwell Encyclopaedia ofAnglo-

Saxon England, (Oxford: Blackwell)

Liversage, G. D. (1968) Excavations at Dalkey Island, County Dublin 1956-1959, Proceedings of the

Royal Irish Academy, Vol 66 C, 53- 253

Le Huray, J. D. and Schutkowski, H. (2(05) Diet and social status during the La Tbne period in

Bohemia: carbon and nitrogen stable isotope analysis of bone collagen from Kutna Hora-Karlov and

Radovesice, Journal of Anthropological Archaeology, Vol 24,2,135-147

Lightfoot, K. G. (1995) Culture contact studies: redefining the relationship between Prehistoric and

Historical Archaeology, American Antiquity, Vol 60, No. 2,199-217

Liversage, D. (1983) Comments in response to Alekshin V A, Burial ustoms as an archaeological

sources, Current Anthropology, Vol 24, No. 2,137-149

Longinelli, A. (1984) Oxygen isotopes in mammal bone phosphate: A new tool for paleohydrological

and paleoclimatological research?, Geochimica et Cosmochimica Acta, Vol 48,385-390

Lund Hansen, U. (2003) 150 years of weapon - offering finds, research and interpretation in, Jorgensen

L, Storgaard B and Thomsen LG (Eds) The Spoils of Victory. The North in the Shadow of the Roman

Empire. (Copenhagen: Nationalmuseet), 84-90

Lynn, C. J. (1983) Some ̀early' ring-forts and crannogs, The Journal of Irish Archaeology, Vol 1,47-

58

Lynn, C. J. (1985) Lagore, County Meath and Ballindecry No. I County Westmeath Crannogs: some

possible structural reinterpretations, The Journal oflrish Archaeology, Vol III, 1985/86,69-73

Lynn, C. J. (1992), The Iron Age mound in Navan Ford a physical realization of Celtic religious

beliefs? Emania: Bulletin of the Navan Research Group, No. 10,33-58

Lynn, C. J. (2003), Navan Fort: Archaeology and Myth, (Dublin: WordweU Books)

Lyttleton, J. and Monk, M. (2007) Sites, social change and warfare in the second half of the first

millennium AD: a reappraisal, in Manning C (Ed) From Ring forts to Forti ied Houses: Essays on

castles and other monuments in honour ofDavid Sweetman, (Dublin: Wordwell), 9-20

263

Maas, J. & Byrne, C. (1996) Fort discovery proves Romans invaded Ireland, The Sunday Times, 21 January 1996

Macalistair, R. A. S. & Praeger, R. L. (1928), Report on the Excavation of Uisneach, Proceeding of the Royal Irish Academy, Vol XXXVIII, C, 69-127

Macalistair, R. A. S. (1996) Corpus Inscriptionum Insularurn Celtfcarum (First published 1945:

Reprint) (Dublin: Four Courts Press)

Macalistair, R A. S. (1916-17) A report on some excavations recently conducted in County Galway,

Proceedings of the Royal Irish Academy, Section C, xxxiii, 505-510

McCarthy, D. P. (2008) The Irish Annals: Their Genesis, Evolution and History, (Dublin: Four Courts

Press)

Mac Giolla Easpaig, D. (2005) Significance and etymology of the placename Temair, in Bhreathnach,

E. (Ed) The Kingship and Landscape of Tara, (Dublin: Four Courts Press), 423 - 448

MacNamara, M. (2000) The Psalms in the Early Church (Sheffield: Sheffield Academic Press)

Mallory, J. P. and Ö'Donnabhäin, B. (1998) The origins of the population of Ireland: a survey of

putative immigrations in Irish prehistory and history, Emania, Bulletin of the Navan Research Group,

No. 17,47-71

Mason, D. J. P. (2001) Roman Chester: City of the Eagles, (London: The History Press Ltd)

Mattingly, D. (Ed) (1997) Dialogues in Roman Imperialism: Power, Discourse and Discrepant

experience in the Roman Empire, Journal of Roman Archaeology, Supplementary Series no. 23

Mattingly, D, (2004) Being Roman: Expressing identity in a provincial setting, Journal of Roman

Archaeology, Vol 17,5-25

Mattingly, D. (2006) An Imperial Possession: Britain in the Roman Empire (London: Penguin)

Mattingly, D. (2010) Cultural Crossovers: Global and Local Identities in the Classical World, in Hales S and Hodos T (Eds) Material Culture and Social Identities in the Ancient World (Cambridge: Cambridge University Press), 283-295

264

McArthur, J. M., Howarth, R. J. and Bailey, T. R. (2001) Strontium isotope stratigraphy: LOWESS

Version 3: Best fit to the marine Sr isotope curve for 0.509 Ma and accompanying look-up table for deriving numerical ge, The Journal of Geology, Vol 109,155-170

McCormick, F. (1995) Cows, ringforts and the origins of Early Christian Ireland, Emania: Bulletin of the Navan Research Group, No. 13,33-37

McCormick, F, Cribbin G, Robinson M E, Shimwell DW& Murphy E (1995) A Pagan-Christian

transitional burial at Kiltullagh, Emania: Bulletin of the Navan Research Group, No. 13,89-98

McCormick, F. (2002) The animal bones from Tara, in Roche H (Ed), Excavations at Rdith na Rig,

Tara, Co. Meath, Discovery Monograph no. 6 (Dublin: Royal Irish Academy), 103-114

McCormick, F. and Murray, E. (2007) Excavations at Knowth 3: Knowth and the Zooarchaeology of

Early Christian Ireland, (Dublin: Royal Irish Academy)

McDonald, S. and Fyfe, G. (2003) Theorizing Museums, (London: Blackwell)

McKone, K. (1990) Pagan Past and Christian Present in Early Irish Literature (Maynooth: An Sagart

Press)

McEvoy B, Richards M. Forster P& Bradley DG (2004) The longue durde of genetic ancestry:

multiple genetic markers and Celtic origins on the Atlantic facade of Europe, American Journal of Human Genetics, Vol 75,693-702

McGarry, T. (2005) Identifying burials of the Irish Iron Age and transitional periods c800 BC - AD600, Dolan B and Kenny N (Eds), Trowel, Vol. X (Dublin: University College Dublin), 2-18

McGarry, T. (2007) Some exotic evidence amidst Irish late prehistoric burials, in Davis 0 P, Sharpies

NM and Waddington K (Eds), Changing perspectives on the First Milleniwn, (Oxford: Oxbow) 215-

234

McKay, D. J. (2003) Information Theory, Inference and Learning Algorithms (Cambridge: CUP)

McManus, D. (1983) A Chronology of the Latin loan-words in Early Irish, &J XXXIV, 21-69

McManus, D. (1991) A Guide to Ogham, (Maynooth: an Sagart Press)

Megaw, J. V. S. & Megaw, M. R. (1996), Ancient Celts and modern ethnicity, Antiquity, Vol 70,175- 181

265

r

Megaw, J. V. S. & Megaw, M. R. (1998), The mechanisms of (Celtic) dreams?: a partial response to

our critics, Antiquity, Vol 72,432-435

Meskell, L. (2004) Divine Things, in DeMarrais, E., Gosden, C. and Renfrew, C. (Eds) Rethinking

Materiality: the Engagement of Mind with the Material World, (California: McDonald Institute for

Archaeological Research) 249-259

Millets, M, (1990) Romanization: historical issues and archaeological interpretation, in Blagg T&

Millen M (Eds) The Early Roman Empire in the West, (Oxbow Press)

Millett, M. (1994) Treasure: interpreting Roman hoards, in S. Cotton, D. Dungworth, S. Scott and J.

Taylor (Eds), TRAC The Proceedings of the Fourth Annual Theoretical Roman Archaeology

Conference, Durham University, (Oxford: Oxbow)

Millett, M. (2000) The Romanization of Britain: A lesson in archaeological interpretation, (London:

Tempus)

Minh Ha, T. (1991) When the moon waxes red., representation, gender and cultural politics, (London

and New York: Routledge)

Mitchell, F., & Ryan, M (2003) Reading the Irish Landscape (Dublin: Townhouse Press)

Mongan, N. (1995) The Menapia Quest. Two thousandXars of the Menapli seafaring Gauls in

Ireland Scotland and Wales and isle of Man, 216 BC - 1990 AD, (Dublin: Herodotus Press)

Monk, M. and Sheehan, J. (1998) Research and Early Medieval Munster: agenda or vacuum? in, M.

Monk and J. Sheehan (Eds), Early Medieval Mrmster. Archaeology. History and Society, (Cork

University Press), 1-8

Montgomery J, Evans J A, Powlesland D& Roberts CA (2005) Continuity or Colonization in Anglo-

Saxon England? Isotope Evidence or Mobility, Subsistence Practice and Status at West Heslerton,

American Journal of Physical Anthropology, Vol 126,123-138

Montgomery J, Evans J& Chenery C (2006) Report of the lead strontium and oxygen Isotope analysis

of the Iron Age burial from Rath, Ireland (unpublished)

Montgomery J, Grimes V and Evans J (2006) Report of the lead, strontium and oxygen isotope analysis

of the Iron Age burial fr' om Ratoatk Co. Meath. Ireland (unpublished)

266

Moore. F, (1998) Munster ogham stones: siting, context and function, in Monk M. and Sheehan J (Eds)

Early Medieval Munster, Archaeolog , History and Society, (Cork: Cork University Press)

Moore, M. (1987) Archaeological Inventory of County Meath, (Dublin: Stationery Office)

Morris, J. H. (1987) The northern belt of the Longford-Down Inlier, Ireland and Southern Uplands,

Scotland: an Ordovician back-arc basin, Journal of the Geological Society, 144 (5), 773-786

Mytum, H. (1981), Ireland and Rome: The maritime frontier, in King A., and Honig M., (Eds), The

Roman West in the Third Century: Contributions from Archaeology and History, BAR International

Series 109i, 445-449

Mytum, H. (1995), Across the Irish Sea: Romano British Irish settlements in Wales, Emania: Bulletin

of the Navan Research Group, No. 13,15-22

Mytum, H. C. (1992) The Origins of Early Christian Ireland (London & New York: Routledge)

Nash-Williams, V. E. (1950), The Early Christian Monuments of Wales (Cardiff: University of Wales

Press)

Newman C., (1989), Fowler's Type F3 Early Medieval penannular brooches, Medieval Archaeology

33,7-20.

Newman, C. (1997) Tara: An Archaeological Swwy, Discovery Programme Monographs 2 (Dublin:

Royal Irish Academy)

Newman, C. (1998), Reflections on the making of a'Royal Site' in early Ireland, World Archaeology,

Vol 30,127-141

Newman, C. (2005) Recomposing the landscape of Tara in, Bhreathnach E (Ed) The Kingship and

Landscape of Tara, (Dublin: Four Courts Press), 361-382

Newman, C., O'Connell, M., Dillon, M., and Molloy M (2007) Interpretation of charcoal and pollen

data relating to a late Iron Age ritual site in eastern Ireland: a holistic approach, Vegetation, History

and Archaeobotony, 16,349-365

Neuman de Vegvar, C. (1997) The value of recycling: conversion and early Anglo-Saxon use of Roman materials, in The Haskins Society Journal, Studies in Medieval History, Vol 9,123-136

267

O'Brien, C. C. (1981) Revolution and the shaping of modem Ireland, in R O'Driscoll (Ed), The Celtic

Consciousness (Canada: The Dolmen Press), 427-435

O'Brien, E. (1990) Iron Age burial practices in Leinster: continuity and change, Emania: Bulletin of the

Navan Research Group, No. 7,37-42

O'Brien, E. (1999a), Post-Roman Britain and Anglo-Saxon England, Burial Practices Reviewed, BAR

Report No. 289

O'Brien, E. (1999b) Excavations of a multi-period burial site at Ballymacaward, Ballyshannon, County

Donegal, Journal of the Donegal Historical Society, No. 51 (Donegal Democrat Ltd Press)

O'Brien, E. (2003) Burial Practices in Ireland: first to seventh centuries AD, in Downes J and Ritchie

A (Eds) Sea Change: Orkney and Northern Europe in the later Iron Age AD 300-800 (Balgavies,

Angus) 62-72

O'Brien E (in press) Pagan or Christian? Burial in Ireland during the 5th to 8* centuries AD in N

Edwards (Ed) The Archaeology of the Early Medieval Christian Church, (Association for Church

Archaeology)

O'Brien E (forthcoming) Knowth, Monograph no. 5, (Royal Irish Academy)

O'Connor, K. D. (1998) The Archaeology of Medieval Rural Settlement in Ireland, Discovery

Monograph No. 3, (Dublin: Discovery Programme)

b'Crbinin, D (1995) Early Medieval Ireland 400-1200 (London: Longman Press)

b'Cröinin, D (2005) Ireland 400-800 AD, in D 6'Cr6inin , (Ed) Anew history of Ireland, prehistory

and early Ireland (Oxford University Press), 182-233

O'Drisceoil, C. (2007) Life and death in the Iron Age at Carrickmines Great, County Dublin, Journal

Royal society ofAntiquaries Ireland, Vol 137,5-28

b'Floinn, R. (1999), The date of some metalwork from Cahercommaun reassessed, in Cotter C,

Cahercommaun Fort Co. Clare: a reassessment of its cultural context, Western Stone Forts Project,

Discovery Monograph No. 5, (Wordwell Press), 73 - 79

O'Floinn, R. (2000) Freestone Hill, Co. Kilkenny: a reassessment, in A Smyth (Ed) Seanchas Studies

in Early and Medieval Irish Archaeology, History and Literatome in Honour of Francis J Byrne

(Dublin: Four Courts Press)

268

O'Kelly, M. J. (1989) Early Ireland. - An Introduction to Irish prehistory, (Cambridge: CUP)

b'hbgain, D. (1988) Fionn Mac Cumhaill: Images of the Gaelic hero (Dublin: Gill and Macmillan)

b'hOgäin, D. (2002) The Celts, A History (Suffolk: Boydell Press)

6 h6gain, D. (2001) The Sacred Isle: Belief and Religion in pre-Christian Ireland (Suffolk: Boydell

Press)

Ö'hUiginn, R. (1990), The Literature of the Laigini, Emania: Bulletin of the Navan Research Group,

No. 7,5-9

O'Kelly, M. J. (1962) Two Ringforts at Garryduff, County Cork, Proceedings of the Royal Irish

Academy, Vol 63C, 17-125

O'Kelly, M. J. (1970), Problems of Irish Ring-forts, in D Moore (Ed), The Irish Sea Province in

Archaeology and History, (Cardiff: Cambrian Archaeological Association),

O'Kelly, M. J. (1982) Newgrange: Archaeology, Art and Legend (Thames & Hudson)

O'Kelly, M. J. (2005) Bronze Age Ireland, in D 6'Cr6inin (Ed) A New History oflreland (Oxford

University Press), 98 -133

O'Rahilly, T. F. (1946) Early Irish History and Mythology, (Dublin Institute for Advanced Studies)

Ö'Riordäin, S. P. (1940) Excavations at Cush, County Limerick, Proceedings of the Royal Irish

Academy, 45C, 83-181

Ö'Rlordiin, S. P. (1942) The excavation of a large earthen ringfort at Garranes County Cork,

Proceedings of the Royal Irish Academy, Vol XLVII, 77-151

6'Riordain, S. P. (1947) Roman Material in Ireland, Proceedings of the Royal Irish Academy, Vol L1,

35-82

Ortner, D. J. (2003) Identification of Pathological Condition in Human Skeletons (Elsevier press)

O'Sullivan, A. (1998) The Archaeology of Lake Settlements oflreland, Discovery Monograph No. 4

(Dublin: Discovery Programme)

269

O'Sullivan, A. & Harney, L. (2008) Early Medieval Archaeology Project: Investigating the character

of Early Medieval archaeological excavations, 1970 - 2002, Report for the Heritage Council (An

Comhairle Oidreachta and University College Dublin), published on-line

O'Sullivan, J. (1998) Nationalists, Archaeologists and the Myth of the Golden Age, in Monk, M. and Sheehan, J., (Eds), Early Medieval Munster: Archaeology. History and Society, (University College

Cork), 178 -18

O'Sullivan, M. (2004) Duma na nGiall: The Mound of the Hostages, Tara (Bray: Wordwell and UCD

School of Archaeology)

Owen-Crocker, G. R. (2004) Dress in Anglo-Saxon England (Suffolk: Boydell Press)

Painter, K. S. (1972) A late Roman silver ingot from Kent, The Antiquaries Journal, Vol. 52,84-92

Parker-Pearson, M. (1982) Mortuary practices, society and ideology: an ethnoarchaeological study, in

Hodder, I., (Ed), Symbolic and Structural Archaeology, (Cambridge University Press)

Packer-Pearson, M. (1989) Beyond the Pale: barbarian social dynamics in western Europe, in Barrett,

J. C., Fitzpatrick, A. P. & Macinnes, L., (Eds) Barbarians and Romans in North-West Europe fron the later Republic to late Antiquity, BAR International Series 471,198 - 221

Parker-Pearson, M. (2003) The Archaeology of Death and Burial, (Sutton Publishing)

Pearson, M. N. (2003) The Indian Ocean (London and New York: Routledge)

Pike, A. W. G. and Richards, M. P., (2002), Diagenetic arsenic uptake in archaeological bone. Can we really identify copper smelters?, Journal ofArchaeological Science, 29,607-611

Philpott, R (1991) Burial Practices in Roman Britain: A Survey of Grave Treatment and Furnishings

AD 43-410, BAR Series 219

Polanyi, K. (1957) The Great Transformation (Boston, Mass: Beacon Press)

Power, C. (1990) Unpublished pathology report on Betaghstown Burials, Topographical Files National Museum of Ireland

Quinnell, H. (1986) Cornwall during the Iron Age and Roman Period, Cornish Archaeology, No. 25, 111-134

? 7n

Quinnell, H. (1987) Cornish gabbroic pottery: the development of a hypothesis, Cornish Archaeology, No. 26,7-12

Quinnell, H. (1993) A sense of identity: distinctive Cornish stone artefacts in the Roman and post Roman periods, Cornish Archaeology, No. 32,29-45

Radford, C. A. R (1956) Imported pottery found at Tintagel, Cornwall, in Harden (Ed) Dark Age Britain, (London: Methuen), 59-70

Raftery, B. (1969) Freestone Hill, Co. Kilkenny: An Iron Age hillfort and Bronze Age cairn, Proceedings of the Royal Irish Academy, Vol 68C, 1-108

Raftftery, B. (1970) A decorated strap-end from Rathgall, Journal of the Royal Society of Antiquaries Ireland, Vol 100,200-211

Raftery, B. (1972) Irish hillforts, in Thomas, C., (Ed), The Iron Age in the Irish Sea Province, CBA Report No. 9,37-59

Raftery, B. (1981) Iron Age burials in Ireland in, b Corräin, D., (Ed), Irish Antiquity, Essays and Studies Presented to Professor MJO'Kelly, (Dublin: Four Courts Press),

Raftery, B. (1989) Barbarians to the West, in Barren J C, Fitzpatrick AP& Macinnes L (Eds),

Barbarians and Romans in North-West Europe, BAR International Series 471,117 -147

Raftery, B. (1994) Pagan Celtic Ireland (London: Thames & Hudson)

Raftery, B. (1996a) Drumanagh and Roman Ireland, Archaeology Ireland Vol 10,17-19

Rafteiy, B. (1996b) Iron Age Studies in Ireland: some recent developments, in Champion T. C., & Collis, J. R, (Eds), The Iron Age in Britain and Ireland. " recent trends, (Sheffield University press),

155-161

Raftery, B. (1996c) Trackway Excavations in the Moundillon Bogs, Co. Longford 1985-1991, (Dublin: Crannög Publications University College Dublin)

Raftery, B. (2000) A bit too far. Ireland's Transylvanian link in the Later Iron Age, in Smyth, A., (Ed) Seanchas Studies in Early and Medieval Irish Archaeology, History and Literature in Honour of Francis J Byrne (Dublin: Four Courts Press)

271

Raftery, B. (2005) The Iron Age in Ireland in, b'Cröinin, D., (Ed) A New History oflreland (Oxford University Press), 134 - 181

Raftery, B. (20(6) Celtic Ireland: problems of language, history and archaeology, Acta Archaelogica, Vol 57,273-279

Raftery, B. (forthcoming) Late Iron Age Burials at Knowth, in Knowth, Monograph no. 5, (Dublin: The Royal Irish Academy)

Raftery, J. (1940) A suggested chronology for the Irish Iron Age in Ryan, J., (Ed), Feilscribhinn Eoin

Mhic Neill (Dublin: Four Courts Press)

Raftery, J. (1972) Iron Age and the Irish Sea, in Thomas, C., (Ed) The Iron Age in the Irish Sea

Province, CBA Report No. 9 (Council for British Archaeology)

Raftery J, (1980) Mainland Europe and the Offshore Islands, Archaeologla Atlantica, Vol 3,1-7

Raftery, J. (1981) Concerning Chronologies, in b Corriin, D., (Ed), Irish Antiquity, Essays and Studies

Presented to Professor MJ O'Kelly, (Dublin: Four Courts Press)

Rance, P. (2(01), Attecotti, D6isi and Magnus Maximus: The case for Irish federates in Late Roman

Britain, Britannia, Vol 32,243-270

Rankin, D. (1999) Celts and the Classical World, (London: Routledge)

Reece, R. (1990) Romanization: a point of view in Blagg T& Millen M (Eds) The early Roman

Empire in the West, (Oxbow Press)

Reed, N. (1971) The fifth year of Agricola's campaigns, Britannia, Vol. 2,143-8

Renfrew, C. (2004) Towards a theory of material engagement, in DeMarrais, E., Gosden, C., and Renfrew, C., (Eds) Rethinking Materiality: the Engagement of Mini! with the Material World,

(California McDonald Institute for Archaeological Research) 23-33

Richards, M. P. and Hedges, R. E. M. (1998) Stable isotope analysis reveals variations in human diet

at Poundbury Camp cemetery site, Journal of Archaeological Science, Vol 25,1247-1252

Rigoir, J., Rigoir, Y. and Mehra, J. F. (1973) Los deriv6e palbochrd ienaes du groupe atlantique, Gallia, 31,364-409

272

Rivet, A. L. F. (1980) Celtic names and Roman places, Britannia, Vol. 11,1-19

Robertson, A. S. (1970) Roman finds from non-Roman sites in Scotland: More Roman ̀ drift' in Caledonia, Britannia Vol 1,198-226

Robertson, A. S. (1975) Birrens (Blatobulgium), (Dumfriesshire & Galloway Natural History Society)

Robinson, D. J. (2000) The Romans and Ireland again: some thoughts on Tacitus' Agricola Chapter 24,

Journal of Chester Archaeological Society, 75,19-31

Roche, H. (2002) Excavations at Räith na Rig, Tara, Co. Meath, 1997, Discovery Programme Reports,

6, (Dublin: Royal Irish Academy/Discovery Programme), 19-82

Rowlands, M. (1993) The role of memory in the transmission of culture, World Archaeology, Vol. 25,

No. 2,141-151

Roymans, N. (1983) The north Belgic tribes in the l' century BC: a historical-anthropological

perspective in Slofstra, J., and Brandt, R., (Eds) Roman and Native in the Low Countries Spheres of

Interaction, BAR International Series 184

Roymans, N. (1990) Tribal Societies in Northern Gaul: An Anthropological Perspective, (Universiteit

van Amsterdam)

Russell, L. (2004) ̀ Either, or, neither nor': resisting the production of gender, race and class dichotomies in the pre-Colonial period, in Conlin Casella E& Fowler C., (Eds), The Archaeology of Plural and Changing Identities, (New York: Springer)

Rynne, E. (1976) The La T6ne and Roman Finds from Lambay County Dublin: A re-assessment, Proceedings of the Royal Irish Academy, Colloquium on Hibe no-Roman Relations and material

remains, Vol 76C, 231- 243

Rynne, E. (1972) Celtic stone idols in Ireland, in Thomas, C., (Ed) The Iron Agee in the Irish Sea

Province, CBA Report No. 9 (Council for British Archaeology)

Said, E. W. (1990) Yeats and decolonization, in Eagleton, T., Jameson, F. and Said, E. W., (Eds),

Nationalism, Colonialism and Literature, (Minneapolis: University of Minnesota Press)

Sargent, A. (2002) The north-south divide revisited: thoughts on the character of Roman Britain,

Britannia, Vol 33,219-226

273

Selway, P. (1993) A History of Roman Britain (Oxford: Oxford University Press)

Sherman, S. (2000) Population, culture history and the dynamics of culture change, Current

Anthropology, Vol 41, No. 5,811-835

Sikora, M. and Cahill, M. (forthcoming) Burial in Ireland from Ist to 8th centuries AD: recent

research in the archives of the National Museum of Ireland (Dublin: National Museum of Ireland)

Slofstra, J. (1983) An anthropological approach to the study of Romanization processes, in Slofstra J

and Brandt R (Eds) Roman and Native in the Low Countries Spheres oflnteraction, BAR International

Series 184,

Shotter, D. (2004) Roman Britain, Second Edition, (London: Routledge)

Smith A G, Pearson GW and Pitcher JR (1971) Belfast Radiocarbon Dates IV, Radiocarbon, 13 (2),

450-467

Smith, J. (2005) Europe after Rome: A New Cultural History 500-1000, (Oxford: Oxford University

Prosa)

Smyth, A. P. (1982) Celtic Leinster, Towards a Historical Geography of Early Irish Civilization AD

500 - 1600 (Dublin: Irish Academic Press)

Stapleton, C. P., Freestone, I. C. & Bowman, S. G. E. (1999) Composition and origin of Early

Medieval opaque red enamel from Britain and Ireland, Journal of Archaeological Science, Vol 26,913-

921

Stein, G. (2002) Colonies without Colonialism: A Trade Diaspora Model of Fourth Millennium BC

Mesopotamian enclaves in Anatolia, in Lyons, C. L. and Papadopoulos, J. K., (Eds) The Archaeology

of Colonialism, (Los Angeles: Getty Research Institute), 27-64

Stevenson, J. (1989) The beginnings of literacy in Ireland, Proceedings of the Royal fish Academy,

Vol 89 C, 127-165

Stormed, B. (2003) Cosmopolitan aristocrats, in Jergensen L, Storgaard B and Thomsen LG (Eds)

The Spoils of VictoryThe North in the Shadow of the Roman Empire. (Copenhagen: Nationalmuseet),

106-126

Stoat, M. (1997) The Irish Ringfort (Dublin: Four Courts Press)

274

Strang, A. (1998) An analysis of Ptolemy's Geography, The Cartographic Journal, Vol 35, No. 1,27-

47

Stout, G. (2002) Newgrange and the Bend of the Boyne, (Cork University Press)

Sweissing, MM and Grupe, G. (2003) Stable isotopes in human teeth and bone: a key to migration

events of the late Roman period in Bavaria, Journal ofArchaeological Science, Vol 30.1373-1383

Swift, C. (1997) Ogam Stones and the Earliest Irish Christians, Maynooth Monographs, Series Minor

2, (Maynooth, Ireland: The Cardinal Press)

Swift, C. (2008) Commentary: The Knowth Oghams in Context in, Byrne F J, Jenkins W and Kenny G

(Eds), Excavations at Knowth: v. 4: Historical Knowth and Its Hinterland (Royal Irish Academy

Monographs in Archaeology), 120-132

Swift, E. (2000) The End of the Western Roman Empire: An Archaeological Investigation (Stroud:

Tempus)

Sweetman, D. (1982) Reconstruction and partial excavation of an Iron Age burial mound at Ninch

County Meath, Mock na Midhe, No. 7,42-57

Tait, J. and Jobey, G. (1971) Romano-British burials at Beadnell, Northumberland, Archaeologia

Aeliana 49,53-69

The Täin (1989) translated by Liam Mac Uistin (Dublin: The O'Brien Press)

Tarlow, S (2000) Emotion in archaeology, Current Anthropology, Vol 41(5), 713-746

Taylor, J. (2001) The Isle of Portland: An Iron Age Port of Trade, Oxford Journal of Archaeology, Vol

20,2,187-205

Thomas, C. A. (1973) Irish Colonists in south west Britain, WorldArchaeology, Vol 5, No. 1,5-13

Thomas, C. A. (1998) Early Medieval Munster: thoughts upon its primary Christian phase, in M.

Monk &J Sheehan (Eds), Early Medieval Munster, Archaeology, History and Society (Cork University

Press), 9-16

Thomas, C. A. (1959) Imported pottery in Dark Age western Britain, Medieval Archaeology, 3,89-113

275 ß.

Y4!.:

Thomas, C. A. (1994) And Shall these Mute Stones Speak? Post Roman inscriptions in Western Britain

(Cardiff University Press)

Thomas, J. (1999) Understanding the Neolithic, Revised Ed. (London & New York: Routledge)

Thorpe, C. (2007) The earthwork at Restormal Farm, Lostwithiel, Cornwall: archaeological site and

finds evaluation, (Cornwall County Council, Historic Environment Service)

Tierney, M. (1998) Theory and politics in Early Medieval Irish archaeology, in M. Monk and J.

Sheehan (Eds), Early Medieval Munster: Archaeology, History and Society, (University College Cork),

190-199

Tilley, C. J. (1994) A Phenomenology of Landscape: Places, Paths and Monuments (Oxford: Berg)

Todd, M. (1977) Germanic burials in the Roman Iron Age, in R Reece (Ed), Burial in the Roman

World, CBA Research Report 22,39-43

Todd, M. (1987) The Northern Barbarians IOOBC to AD 300 (Blackwell: Oxford)

Todd, M. (1998) The Germanic peoples and Germanic society, in Bowman A K, Garnsey P and

Cameron A (Eds) The Cambridge Ancient History: the crisis of Empire AD 193-337, Vol 13,

(Cambridge: Cambridge University Press), 440 - 460

Townshend, C. (2006) Easter 1916: The Irish Rebellion (Chicago: Ivan van Dee Publishers)

Trigger, B. G. (1989) A History ofArchaeological Thought, (Cambridge University Press)

Trigger, B. G. (2008), ̀ Alternative Archaeologies' in historical perspective, in Habu J Fawcett C and

Matsunaga J M, (Eds) Evaluating Multiple Narratives: Beyond Nationalist, Colonialist, Imperial

Archaeologies, (New York: Springer), 187 -195

Toynbee, J. M. C. (1971) Death and Burial in the Roman World (London: Thames & Hudson)

Tyers, P. A. (1996) Pottery in Roman Britain, (London: Batsford)

Ucko, P. J. (1969) Ethnography and archaeological interpretation of funerary remains, world

Archaeology, Vol. 1 (2), 262-280

Van Dommelan, P. (1997) Colonial constructs: colonialism and archaeology in the Mediterranean,

World Archaeology, Vol. 28, No. 3,305-323

276

Van Dommelan, P. (2002) Ambiguous matters: colonialism and local identities in Punic Sardinia, in

Lyons CL and Papadopoulos )K (Eds), The Archaeology of Colonialism, (Los Angeles: Getty

Research Institute), 121-147

Waddell, J. (1990) The Bronze Age Burials of Ireland (Galway University Press)

Waddell, J. (2005) Foundation Myths: The Beginnings of Irish Archaeology, (Dublin: Wordwell)

Wailes, B. (1990) Dein Ailinne: A summary excavation report, Emania: Bulletin of the Navan Research

Group, No. 7,10-21

Wallace, P. F. (2008) Irish archaeology and the recognition of ethnic difference in Viking Dublin, in

Habu J Fawcett C and Matsunaga J M, (Eds) Evaluating Multiple Narratives: Beyond Nationalist,

Colonialist, Imperial Archaeologies, (New York: Springer), 166 -183

Wallerstein, I. M. (1980) The Modern World System (London: Academic Press)

Warner, R. B. (1976) Some observations on the context and importation of exotic material in Ireland,

from the first century BC to the second century AD, Proceedings of the Royal Irish Academy,

Colloquium on Hibemo-Roman relations and material remains, Vol 76C, 267-289

Warner, R. B. (1982) The Broighter hoard: a reappraisal, and the iconography of the collar in, BG

Scott (Ed) Studies on Early Ireland: Essays in honour of M VDuignan, (Dublin: Assocation of Young

Archaeologists ) 29-38

Warner, R. B. (1983), Ireland, Ulster and Scotland in the earlier Iron Age, in A. O'Connor & D. V.

Clarke (Eds), From the Stone Age to the 'Forty-Five: studies presented to RB. K Stevenson, (Donald:

Edinburgh), 161-187

Warner, R. B. (1985) The date of the start of Lagore, The Journal of Irish Archaeology, Vol 1I1,

1985/86,75-77

Warner, R. B., Mallory, J. P. and Bailie, M. G. L. (1990) Irish Early Iron Age sites: a provisional map

of absolute dated sites, Emania, Bulletin of the Navan Research Group, 7,46-50

Warner, R. B. (1995) Tuathal Techtmar: A myth or ancient literary evidence for a Roman Invasion? Emania: Bulletin of the Navan Research Group, No. 13,23-32

277

Warner, R. B. (1996a) De Bello Hibemico: A less than edifying debate, Archaeology Ireland, Vol 10, 38-40

Warner, R. B (1996b) Yes, the Romans did invade Ireland, British Archaeology, No. 14,16-17

Warner, R. B. (1996c), Navan and Apollo, Emania: The Bulletin of the Navan Research Group, 79

Warner, R. B. (2000), Clogher: an archeological window on early medieval Tyrone and Mid-Ulster, in

Dillon C& Jefferies H (Eds), Tyrone History and Society, (Dublin), 39-54

Weale M E, Weiss D A, Jager R F, Bradman N& Thomas MG (2002) Y Chromosome evidence for

Anglo-Saxon migrations, Molecular Biological Evolution, Vol 19, No 7,1008-1021

Wedlake, W. J. (1982) The excavation of the shrine of Apollo at Nettleton, Wiltshire, 1956-1971,

(London: Society of Antiquaries)

Wells, PS (1996) Production within and beyond Imperial boundaries: goods, exchange and power in

Roman Europe, Journal of World Systems Research, 2(13), 1-26

Welshe, W. (1999) Transculturality: the puzzling form of cultures today, in Featherstone M., and Lash

(Eds) Spaces of Culture: City, Nation, World (London: Sage) 194-213

Wheeler, R. E. M. (1954), Rome Beyond the Imperial Frontiers, (London: Bell & Sons)

Whimster, R. (1981) Burial Practices in Iron Age Britain: A Discussion and Gazetteer of the Evidence

c. 700BC-AD42, BAR Reports 90

White, R. H. (1988) Roman and Celtic Objects from Anglo-Saxon graves: a catalogue and interpretation of their use, BAR Report no. 191

Whittaker, C. R (1994), Frontiers of the Roman Empire. A social and Economic Study (Baltimore:

John Hopkins University Press)

William, H. (2003), Tales from the Dead: Remembering the Bog Bodies in the Iron Age of North

Western Europe, in Williams H (Ed), Archaeologies of Remembrance, Death and Memory in Past Societies, (New York: Kluwer Academic Press)

Wilmot, G. F. (1938), Three burial sites at Carbury, Co. Kildare, Journal of the Royal Society of Antiquiries oflreland 68,130-42.

278

Wood-Martin, W. G. (1886) The Lake Dwellings of Ireland or Ancient Lacustrine Habitations of Erin,

facsimile reprint 2003, (Lampeter: Llanerch Press)

Woolf, G. (1992) Imperialism, empire and the integration of the Roman economy, World Archaeology,

Vol 23 No. 3,283-293

Woolf, G. (1997), Beyond Romans and natives, WorldArchaeology, Vol 28, No. 3,339-350

Woolf, G. (1998) Becoming Roman: The Origins of Provincial Civilization in Gaul, (Cambridge

University Press)

Wylie, A. (1992) The interplay of evidential constraints and political interests: recent archaeological

research on gender, American Antiquity, Vol 57,15-35

Wylie, A. (1993) A proliferation of new archaeologies: beyond objectivism and relativism, in Yoffee N

and Sherratt A (Eds) Archaeological Theory: Who sets the Agenda? (Cambridge University Press), 20-

26

Wylie, A. (2008) The integrity of narrative: deliberate practice, pluralism, and multivocality, in Habu J

Fawcett C and Matsunaga J M, (Eds) Evaluating Multiple Narratives: Beyond Nationalist, Colonialist,

Imperial Archaeologies, (New York: Springer), 201- 212

Young, R. (1995) Colonial Desire: Hybridity in Theory, Culture and Race (London and New York:

Routledge)

279

Appendix A

0 >o 0 ID CD M CL

W ývýv c% ece iyi eE dIß E oCO 00% 0 o+ 0 c+ N

it. ä Z'dCZ Z'= aim Z'äiaö L'c3 ° c3 °C L' üiaý °1aö vº °1öý 2 «vg -mg «a e_ CO 0% 0% «o

X7199 3

9u 3m3m E3ß 3rý, 't f ", - "-i

011004.30 c': c O': c d oý oý ö cý C7 cý -ý ý .,., 3.03 ruU3 °

99 OW Ws 00

CC oC CC cC v4 00 0t e W. d %0 rý Ln r*. Ln In %D CO P.

1O t%0 0NMM In U .

OO

CO CO m Ch m ca CO CO CO 0 >

In 0

OOD 000 90 Ln

M

'_ ODD U oUG c12

9 oUG COQ 9 QQQ o0

n ýQ ýQ ýQ ýä üQ ciü ýQ c CD

üä -Ip J O- O- O- _O 00 1Q

000 00 CD

OODU NLANtOMU

(ý d0U dV MV cU MU MM

Ul W !i

CO m CO CO m CO CO CO m CO 0 ý iN

Zmö 0NN8m

r' Ö Z LA In iö00 In 00

cý ö in m

Is o+ 0o

ýQ 0 CO $0 vl

Z v4 9-4 o cN7 . -1 9.1 N ONNNa

t cm Ai r_ äät

to CL

°. tu . äg

0EEEE EöE Et ui

12 19 - %f- c

11 «3 11 tý CO e to (7 v üt oäüüEU ný us U"-ýuýv5

Z t2 O

t t,, QO 'CI '' Oº

ms D 6-4

b. 4 ffl

tLD

a- ==

je

ö E

cl az oB o8°° F'- 3 LL 1 CO ZEbbF,

bbBFbT ki- Uäädääääää aý

L0 d) 0 (L)

Co CR$

oL WLn 3v 3ý >o cccc X00,3 3m 2 in crt 3: 3:

OV 00 200 o0 6 'I Q°

;c 10 '

,0 ;cc `v =mm Co 0 OD 00 Co m Co Co pp Co r0 Co m

VA 1--4 -1 0 (L sir

Ny r_ C2E W_ýO_2O c% dQ1 C Cm r- 2O C 00

W i0 ýo ý0 tC j, iÖ C Oýº to

ý0 0% ö Omi c am, zu ý, TO ýioe3

cýxcýxý m. vý. Am. ým. ým. 4ý3. o23

QW p0 Cº Oý C1

-H -H -H

00 0% Cm9 Ln le I

W cc oc oc r; a.. o000 °C oc oc

of cm Ln CO 00N CO Co in

N CO Co NNrO CO

M >>>> ääää ä> oo CO ao

0000 0' GUUU QQ aQQQC Co mUmU

99 mUCý

j QQ aß Öd ja %0 0 00

Vm iOA NUMm . O-1

Cý ^UmMM

Vi W

I- Caaa F'" CO Co Co Co Co m Co Co a in o0 in in in Ln

in Z00MM01! 1 V1 0N OO tD ,1NN CO , -r

e-t "4 VA 4)

CID -4

. -4 NNNNNO

r

EppC3 Vl

V tCj iw 4-' -w 4. ' yE

Ö

C) ;aa,

_E 18 m2 *0 «o ß_ w- °- EEEE '° E E E

9 U)

1.9

W ýýºuýa ý5cßC= ESE EJt! fE

t7 v ¬ii-'oü o2ü2äu m 9r- cc i= cj cQ ö t5 Co 92

Z °C

r- EC .: IW. 1 W. JW JWJ W. JWWx ýW

a Co Co 1222

CL tf t? t! t "p qyq Uli-

c r. tOB cý OOä (5 ßß

LL ä cý cý ziz

Ti

w -i

i Co Z du 9999919 ir ir 9

'A

a, QI NN QI Cý C Oi Di Ci wd NN L

g "' m 3ý 3LC"0 'LLC '"

Co eý ýLGC (4 L

C"2

3ý X rj

ý

Z ýc

c ýo c c

ýv c r c ýo

c w (30 m

ý' 00 m Co to 00 49

z 00 m

-ýL 00 10

74 00 m

, ter 00 to

4 oc ý tu r0 ¬_°ý r-_° ß ¬ °ý re ýý cý r ¬ý 0tJZZ JZZ JZ rZ i

Jf ßi C% Jf "4

C% JZ 0 JZ "l

J Z"4

Z I- wI

In

IL CC

o rn m g Co M 0

U, Co 16

Q N V

0 r*ý tu

(%i W

Qmm o

Z Ln O° cN

N Q

_Ö

O ~Ö

Fw

Wx

VÜO

+r ý E

I

W

wß ' 0~C cý (% tiw

m

Co N

Z t! ?5WÄ

JI>

5 (~Ä Z ! tom

4

JZ CD 2 '3 E 0

0

M

c ce ce 0 cC ce oC

c% CIN CIN "l le le qr Co Co Co r4 c13 Co c12

O N O IA O N

ää ää m m m m ÜV Ü _ UQ

ÜQ _ Q _ 00 M t0

O0

-ý M

O-1Q MV

O-fQ M

O_1Q MV

O 1p

. MU

Co m m m N

_H

Co

-H

00

41

Co

-H

O

-H

tý

-H 0 M

. ^-4 ý 0

N N N Ö1

ß

U Q 0' c

= go Ä c

ü R E E

H ý £c E

U -5

u oz 2 N C

m U

z to ä

d U

d o«

Z c ý'

p U U C o 1 Co

c a!

c W Y -

E ate.

E E E .JW :W O C)

2 ff!

2 'd II)

2 ? (11

2 o fq

to m m to m m Co

N N N N 2 N

Jul N :@ N

t'

tý

tp'

1L

t'

ýL.

t'

1ý

t'

IL

tp

IL

t'

LL

z z z z z z z

L L L L t L L

cp

(A

ä ä ¬ E E 3

V) W

ä 0 z 0 m ýc 0 0

W

a z 0 cc 0 W Z

m

a J J

LL

W J m Q

EEE k

-YYöc^ oi` 90 0.. 3° 3c

c2 äö cä M i^ i'^ nNN ö' Co

nNi %o iö iö c ýo c .. c a+ in vW Co i0 90 r0 90 t0 Co i0 OM f0 :Mý Co 0

CÖOC2OCÖOCÖON iD

ÖoýC

ui crt c> (D CD >. C% '

Vl JJF ri J . -I . -ý E .1fpNNJOO +ý

Cl tu W

kn

91 Q CA

oC oC a p ap Q ' ce ce ce ce

^^^^0MML

mN Iý N 01 NMMN Co to Co Co N Co Co mm Co c12

.. 1 >>>> O' >>>>>> 0? 000ä

cý 12 c mäh _-m - _m

m ääm ^ää _i _ä 44 cM (30 32 W. 4

4 00 0_1p o_1Q o_1Q ooo_ýqet %D 00p00 V Mý . -I U

MV NM

CO (4 r, MVIA 10NM

G

CL Co im

Co Co Co to °ä a i

mo MN cm IOA

N. e ILA "f

_H _H _H _H fi _H fi u) 0OO U) 0 CO U)

Ö in tt 1N N %0 N Co N CO N CO N

N0N1.4 "4 "4 M Ln 41 N +O-I N i-1 .

i-1 r0.4 O

ýQj yý

ý1p d

m in 37 Ö

Ec Ec E2 EE E Ec

22 JI ti 18 2E

ýc ü

(5 t; i0 u«a üä c) u cýÜ Eü EI6

(4

ae ý

c W .0

i t;

O

Et li E Et; '° -ö 2 CD

ch 2c -& (A

ý °v ä °v tt m U) 2. z Co Co Co ao ao

! ±t 2 Jul ll, N (A

U) (A

t t' e ti UO- li ti ti QQ M

r_ r_ ÄÄÄ

ýiy

m CM > iob G WÄ

ZZZZZ 1) Cý VV

C9

_c

'A

D1 C b ih1

WWZ C% 8V-4

10 ý= LN

6g w Slibo"ö°+fI UO

0 19 law Wt

GO

n 0

ý ý N

n 0 N

n 0 N o o o o o p 1

t .p N N N N N

Y Y NY Ö Ö Ö Ö

r_ L- 00t

ät ä p p O O Ö

LU öý

(n N co (A 1-4 N co Go co

ý ä ä ä

ää c _ cl ü 74 V-4

Ln V-0 J. 00 .

N A (A N N

m LA N %D N

w Q 0m

co N

1-4 N

N -H -H 41

M Cs O C9D cn

ko m

N

Li

j t pý

37 1" ÖL

.v 0

-

ui Ö

Öt5 w ao

(, ý uu c3

Z LD oc

L ä t t cl,

S °' ° C w oc '

co a r J WL 2

_P 2

U-

1; 4

W CO c 3

c 3Q ý 3

c 3

c 3

ý.. {ýi O O O O D

N to N `- rv

(A (A 94

9 c2

9 oo

UÜ 0ä Co

_ää o oCý o o_1qoo Mý MmMývM

Co CL CL Co CL CL Co Co M 0 ii

m0 _H _H

M

_H Ö

N vi "4

0 Kn in . -1 O

N

%D N N NN +-t

a ý ý ac e!

ö ö ö ö V

E am am in E 3

A CpCOCOCOCp 3Z3 .C7r3t7t

CCCC

docod(Daudo

CCCCC

333 0ÖÖCÖ

0

N

Ln d Ob

C

3 U a n

öö0° eý ° eg Z'ý äö Z'O Z'$, rtr ýý rC

41 p CA

V ß' cr, ßi-4 in - CD Ch Co ty% ßo 01 i0'q^TO"ý r ß'4 n ýuVA

"t pl WZdfd-, f CL) Nto., f 3? c% fdc o+ LL) c o+ fdc

16- N0

CZ C Cl C7 CZ W -c Au cßC tz

NCß

NdßyN! Cg

_y N .2E pWp =c ý`ocö 2yß ýv'ý°o'Ä iaö'. ömto

t7iX 3 IS Z r- Co ßf 3ßu cjýcý ß33 ß3N33 ýo3N33 ßN 0 me CW Wi

OG

N OC OC OC WMN

. -ý OC OC C% Ln a

.i f4 MM '"i Nn

0ý0 N fV N N0 cm

M 90 0MM°

111MQ st

ääm ää 93

Z to mä m

ücýcýý cici ýä ýO ä ccý üä

M Jx c, "o VN in re ONQUMV ý- INMV

(%i m W Qm

Co mmIm to Co

12 M le le 0m %D "l PS 00 Co

Z ci LA 0 LA u) Un 0 In In 0 Un %D

iN Cýp1

. '-I N

. 9-4

. 1-1

. '-I ýN

pCYI

13 ß'a Q L. 0 (A ill

ößN aß aE týii IS IS

O E, Ea ýß 0s ää

rOa°N°Nß 9c

Lý Z8mozL t2

W 890 C 00 : 3.

E 'ma c Vv üý U3 üC a. C. 0 ü-2 OLr;

OIn 8

Z

& c1 Eö v' °' =

fA Z U, -acä=ii

t

cccc

ßE C»

E Z)

1a (3 X aö Q cl I- oC Yc Ym

ä1 J°

`c 332äÄ3

Ö Ö Ö Ö Ö

2t et LE

O

0r -0% Cp A1A 0 pm .o

O1'- N fNý. 1. -1 N

M Im Z Ot ý 47 C p-, M 90 01 fdC Nß0

01 ZdC -4 N to ß

01 Z dl r- I"ý 90 Cº 0C

0 ON L- r-1 r1 lu

WM . -r r0

WNCd ,1 ýC Ww 12

t ý

. -r tC .. 4 lo

N, u t2 d

Iý c` 2N

.1 ýo dCN

fl

LM

C

W i0 CÖ0 _ 1 i0 e0 CC0 ÄmCoom i6 Cm

. I, CmmCC

10 Co pp

cNCF. 0 ýac33. v3N 33 r-3N 3. o3N 33. v3 N33 X3 ". 4 -. 4cn

oft C iu W< 00

116 re cC a aC cC cC ei N

00º

IA

001

O

0º

CO ^ lqr

OOº

r1-8 4 in GO

NNN

i i i i üý n c s n n aa O N

i

ä ä ää äiä mý ämä2 ä ä ö ö 0ö 0 öo

ON MM

0 ON N In

OM .ýM

0 %0Ln *'I N

00 O1 . -+ M . -I

o OOOON0 ß M r% v-i m (yb

W

Q 0. 0

to m CL. to 0. Co CL CL c12 Co O Co O O In O,

In Co

In 00

c> tn

NO . -4

Z 41 In ff

* N Ö 0 -H -H in

Ö N

M In m O

O N

9X 0 -4

o YýYE CCCC

L M t2 0ý

V-Q Of

Yl N Yl N C V& _ ý° ý° ä ä°

1 ýp ý1 ý0 C ýC Ü 'c oÜ Q

ý d y ttt

0 W 42 wo wa cm im cm

= Ul 2 Z Z S 'C 'C

Co

ä _C

c_ c_ c c

Lo

00

h 7

W

7 ý

> =E t 12121 cm

ý J O O ö ýC. G

Z Y Y Y Y Y ZZ

. Xt 0 to ß m0 CO Q 22 2

IS ß 2

to 2

-a 32 E 3 x

Fi- V2 2 22Z

O

2' O Z' OO it o C ýp c ip YYYN 99 W . -ý

0 O% - 0% Oº 1. M - Oº NN up (U

OVY (0 tl Qº rp "i ° i0 '`m -C) ` Gº

m 0% 0%

CD E Z7 N

wt0uf2a, fdEd0 c6 1° r- r_ 0ö o+ o+ o+ °1 °c cl m

2W g0

Ä «o

EE «a

°EE °0°

0E In O °O'

Q ^° 2N2N0t! 0.0+

X3f3 e w3 mt73 m(93 m5. °'-ýf. rn-ý rn o, ßo Sao1Q o c+ ýo+ 3r1 *ýUNNVNm. -ýX. 1

d

o

CC GC CC oC CC le» U. ce fl, - M rZ, v4 0% 1; %D Ln le 0 Ln Co

M Nj M2

Qg mm Co mý m< mý mý uü u1 %D o- O00 00 Cº O In GO 1ý O 1Q O? O0 O0 CO

0O CON M "1 N

M ri NUNNM eý M eý NMN

Uý W ui

mm Co m Q0 Co Co 0

OM O% Oº 0NVMM

Z9 r4 CD CD CD 0 Ln r. N IV-4 0 90

%0 rQZN0 01 0^M ý1

in In 0 en ZNN V-1 NNN -4 -I . -4 Nm

>

Eßdo, ,ý CL ßa - O3 1C LLL

03EEEE 9o2 oo ö Win'

_ýy ýý> ro

v+ 00 tu E

b. du 0) (P e8 -a 8 C7 8äE je 3: üý üý o; 0FS Q

O d

E t! t2 t (

W� LqC p1

p1 4.0

_<Vf äý aC o7

'v myccc

90 90 tu tu tu

cm cm cn «c aö t7 t7 (7 ff<<<Ü

U-

w r» 12 e 19 .8 10 mv1 ßv 19 13

1

vý '6º 'ý, 3 03 333

47 4i Qi 41 47 O r'

ä e -2 G L:, a O N l 89m d

0% 0EC ,D 4J

EC 0 , 0EC

t0 u OEC

t N C C

LO d C01 EC

G p =

W yý 3 .o 2 '

1 0

3 .o '

0

., * do Z'Z' 9-6 m e e

"

"i to e 2 1 'i ? ýº 10 e e vý s X eZ Z

.. Z - 7 .. 7 .. c

zuW md 00 ý

dl Op 2 ý0+ ädýO8 ß 2W3i

00 e

Oý sO mo m

ý00 ýOý io C N üocm

. 43: 4 m 0%

1.4ZC. º oý. ýZ. -1 1- 4Z 4 m c; Z'4 CT -43VI

%D 0 O Oýº

r1.

do Ö

t. ov, W tco cý ý-ýaOoýCý I

cý I ý; ýaü

DOLn<

3 ý N rlý le Co

CO Co . "i ý-1 ý-1 r1

inn r1 .y

, -1

Cý C7 (7 C7 Cý C7 W

5 Qt0 LIMA

fýtr c> Co CA NUlb %-4 t0

Vi W

W I-

to Co % a CO

a a c12

a CO 0 CD o in p CD

Z ýý

N O M

1A O

to

ZZ in r1

it, .f

in N N N

I N

A im

a 25 go R ß go g 9

ö ö ö ö ö c _

C W (, 7

E VF

E F

E F

E F

E F

E Fcä

_ cý -ý

0 e to to m de 4 l G L

LLI ýý r . a+

Ö

O O ý Cý J Wý

ý 0

Ü 8 O O O O i LL

w ö E E E E E ýi

O

ap M de NNN\

OV- -c+ - cyb c CD WZ '1 Oº 0 rp NN

Ec

cr, WC pp ýC ýp Q

t710 O '. 4O to t9

aC OW Wä CO

e ce te N. CO O . -+ NM WM 1O In NNM em mMmm %D %D N in IA in IA IA IA IA IA

NNN le O 00 OD 00 () Co Co (3D

c12 m Co c12 >>>>jZZMZMZ

5>

CD C) OpQ QQQP'Q

0m tVD

Q CUD ýýUUý (j ý

cin 1Nn

u< U(3 ÜQ32 Sir, -I i Zr, -Qý O 00 O_ _NNN Lh (3

Co i NV NN N0N0

NNN Mý {J L0 3N

W

W CL CL Co a. CL

in

CO Co Co Ln fu Co m Co Co m Co Co

D %O -4 0% .400 .4MM0OO M Co v 1-1 MM et MMNMN

Z 44 Zx +1 41 +1 -H 44 -H 41 41 -H '0O0M0 Co dNN,. ý ON 0% In NM9.4 O% 1- 01 LC 0%

0 Cl 0 00 00 O% CO tt 0% 0 O tD d N ý4 N rl r4 ". 4 1-1 N r1 N r. 1 0%

E

Co M V .v`tc ýn ä. E '' '4 ä EE Er EE6 EJ- E ýo 1E

QOL A> CQOOuOQc> yO C E+ý E+r o oýF

ýß '3 äL m; =m,

ý

'3'3'3Eý8.5 o8ý- vº°c °-° 0LLLL '0 ' lC - ""

ui E AE toEmotv_oc(E( Boa foo ý. > Toc Orzort mt M LM "cs"i. e z 06 ý. c e . ý°

QV Uý Usk UeU. _U ao.. +U au0U o u.. tn caU aaiiº °J

Z 0 tu _0o° alai ý, ßäi

IL t-

aci ößßßöäßß

wm5O3C Ö1

'5 Cf

-5 CI

pl ' pº

-3 ch

ý' ph

7777

= tf- t-v m' E"c E"c ErEc E-c EcEcEc U, 1 -6 vvv «ul -O

Z o, cm t cii cif & üý x jý cif chi cccccc

üttrttt je Lo 2 l! t! Lo l! Lo Lo

JJ2 LYFAOC IAfýýý1ýLXFý ¢iý ýFý

CCý

Lii CO O9a=ö3 Qux F- 7. x7xxex F-

n

w

Z 0 Co CC U O

w C7

O oG 0 w

Co

LL

w J m

O

0 Op MMpOppoM VOO 01 jý 0%jC

ri =0 ON

C) 76 yZj CCCCO

aVCW EEEEcc E ýý zzzz 33Ä3v 3ö

W Z I- m 09 in 9

n MMMMQO .10 Z in In in IA IA

A

OD GO Co Co GD GO 11 NG 12 0NN

92

CD Nýý cý

ýýýý äßäm m_ý Cl% fQ4 iQiQR}cý

u cý od v1 . O4

UQN 1ý VVV

'°

pý ps NN p1 '0- CD 00000-. 00 in in ttl NN to O (0 01 ONON0 00 NNMMNMNNNM %0 "i .IMM .i

G

9aaaamam aaaa 00 Co ac Co ra o Co 2m Co Co M Co

NNNO 1-4 LA in in j00N

r% Öº 000 0NGOONNP.

ý Cs 0

ZN t0 .-N CO .iO to 0% -4 .1 A) +1 . -+ NN *i NNNN Al

ö to 4c ä

ýi Y IO in 1' Nt ä äý z

if Z%cE3 E' ý

EEEE 'Q EEücEEEö C 't a+

d fA CmNC

Or `° ö2c Et ör c Er's öcä .o fit- t- E

uu c6z u 8u. cu aeuf tom 3 mcn uu uäý;

.WW -w NO Z

: n. 2,2 Ule

75 rar-3rrcZZcEcE EEcEcEc E"c E to faZ cEi .w ]w u

cccccc ät äÄ

Lo e t! l! t! Lo aý `F-° oC `!

-° oC oca aý zZ oC oýC m

tTf m

ttt

O a10i 10 a`0i ar äi d 41 1) ar öcöw

9fffxffxfxff2fA

0

ÖO Öý Oý 00 001 0 01 000º 000 000

000 00

ÖO ÖO M 19 WO-.

ý-. 1- 0-ý Go -'. 4

ý0- -, 4.. 4 M

q0% at r

ý0- . 0-i0 -E 4N pu ýo ýo ý ýo ß c+ ýo ý ýo pý to ýa ýo ev ýo ýo ZýorfdfdfdiS? XI? XLy Xdle f. Ly =fdEd=fdýfd

OL W aýý °rý daB °'cB 4)a'$ý °'a`$0 °1ýv °'aÖ c ýo$ c pol ua c a'$ V W'E°c'E°c' E°

c' E°cßcmßc

m 'E°caMc=Mc=Mc=mc Äc me I

oc3 ova X 3: rv3 X03 ßv33 X033 ßv33 m m3 mco 3m m3 (0cD m co 3: ß 0 GO ow 19 ooi mM OW

ac cc cc

0 0% 0 Cie cc cc 116

w "q go Ln NN

C4 tN0 CV)W

T LA N

N W-4

Q (7 W Cri JJJJJ

O N

i aU QE OD000'0U

mE_ -a

<< as c_'< as 0< mä_ aCaLl 45 ad ach ada add 33 ücd 0000 0_ oo O 00 00 oO 00 0 r10M 1A 1Q OM ON MN 4D T-1 MO inN 3 NtF01Go mU N (A (n NLfl r-4 -4 N( rýrl

W 'CIO

CL CL CL CL M Q0O Co Co m [D 0 Co Co Co

0NNM in

ýN LM in

+1 41 -H -H -N -44 ZOOO in OOO0 N .4NN$ %D M ý"f 01 Qt "4 GO t0 co N Op d :O 01 O

co 2 . -4 N ri .. i .4 .4 74 . -4 1.4 I

Q ýö .ö 1v Ä

. `°'v ä

ccocE 22 L° t°. LEcEcE .8°EEE

La. S EEEEEccE

W scctc110c EE to c `° t0c 0u33333u 10. u to .00. muuCu

p9 aw x

QW333 CL CL 06

pt

>% >% : 0% >. >. EEgEEE

WecC

ca -6 "a co aa iL 0L i= EEE3

tm cm im fo La

C20 a ü cý cý

WEE bi Jk . 2t. z 2: 1 21

0 -9 .9 A

O

WÖ N 00º ýCOºnr4 m r14 99r-9 T^0001n 001n00r, C2, g 9ý "M LA rýJ de 4 1-4 "j "4 ', 1-4 .l 1-4 4 Irl 4 1-4

W Z` 2" Co Co Co ? .n Co in m . -ý C, 2 CO 75

W 10 Cd 4J dO Q7 Q1 d 4I IU 47 07 O7O 'C

o°c 0W W

G0 IA .+ Co In

OC r4 IV-4 Z Co &e OC aC cC ad cC 1.1

5mJ0 12 000 12 0m

0O

Cm mä _

mm _ý<

<< _m< _mä N c3ä as aLi cßä a(3 3 ae cic'3 üü

Co V. 4 (. J et oMNV

of N1 . -4 . -r %0 it v-i N1 Li M

W

Q0 LO O0 Co

O Co Co c% 0O0 OO

Ln "4 "4 9.4 1.1 "1 r4 Im 1-1 Z*000 -fi 0C000

C)

0NNNNNNN +-4 N wý

dJ Q t V 3ýZ m9> `g'- Ng1Zä Vn 0 (P `; i0 Co

- LL

29 E! 3 16 =o 2! tdg

EEE EQ röc02 0 ýy

V $+' y 117 týliýj

ý {/º

y_ O< O ýy

WaßEm W> c ('> uu cgs' U A'ä 92u p'2 ö3 L' 3ü9

Z WÖ

r_ Z r_ NCGcGcGG 12 w um _ÖÜuuÜVÜÜ

If) LLt. Ltt

.s>>Z R '. yCC AC GcGcC

Lo Lo l! Lo : go» LG LO D

U) EJ 'agt9 0ýcý mc3 o&cý of3 cl t7 t23

LL

39 F. - Nf MM Nf N to NyNFx

0pý 0 öpf ýöf

öpý ö 629 O0º ir1 41 If1 M Q1 iÄ It1 iÄ iÄ Oº iÄ tý1 iÄ N j0 e-q -; a V-1 Iý -A "-1 N ýp ri C1 1- j0 rl 41 1ý i0 ý"4 ^ jp

Wýf arE aý o°1-ýf ý; °1+ý °ýfg . o-

9.4 'o

Z 41 41 CI NWN 41 ddNddNd! dmNddmdN

slb3i X03 m 33.03 3t 3: go 300 ° 33 ß3 33 10 333 10 3

O OC

wig cl a

0

N

cC oC ne oC 116 M %0 r" cx

r4 (41 00 e0i

NNMNN le N

in Co Co Co Co Co Co

i ýý° qßä Qä °ä_ä äýý Qä Qä

üü ýü ýcý cýü cýcý ýü cýü QOºý} Om Oý Hemd 00% OM In MN' U9 0Q4 1A M ri %D N 1fl

W I-

Q0 0o ö Co m Co Co Co 0 r% le

vii *cýö0 er

O N. r+ wNN

3

(5 m9ccca0r. r. r. E

0. Erv', E_ EEE

ge2g9

E

(ý ü üý ü8 üc f. 4 ü 0-

Zp O

@6 v

= cý zzzzi

i c E Je

a ILLtLC= J

ýWy Y4ý Opl N Vtfffl U0

U-

-i LO CCCCC

kz l= F- m» igt x

w Qö===

otwN 4-4

^ ý+iri aN EV

° Ei, i. ° E0

WZ x

I -1 . 2gºov+fyä+of v a+AC o+AC v+ßc ...... _. r.. .r3 ýo ýº ß . -ý 3 iv

ý-moo C1 'VN al 01 OI NE cl y ..

ÖdÖodÖ

E 0% cl, ac33 ß3 33 A3 A If o", I. +o f"ý

oc oc oc ac pN

fß+1 N 41 en W4

NNN co co P4 co co

gg ýg <Q 0< N3 ýcý 015 45 19 a 43 so co $ 0 8e

W CL A. Ul to co

in in T-4 Ln

r4 0 .1N. 41 -H . -H Z in vii

* in 0 in

ONM 1D N%oo =n GD co ý-1 W4 1 ri e4 .4

15 Qä

O2c p

E°k v ,

£°k ä

A ßg1 ß ßý _ '_ L9 6 69 x

aC z O cr- G 19t ttt p0 1ý 402 W==__ __22_

to I M q 'COX

LL

e I'' 333

0

In as of R

C

A n

W J U Z D

H x W F- Z 0 U W w W I

W F-

M 0 cc U- J

W

(V W J

Fa-

L (II ý- 'L

-O L C Y

L O

m Ü N

OCmV -r- Co

0= M6-

, 33 0 -O

C J

Co - ýON ý 0 d) Q) L OZ ui E NL O> N- O p

.ý N

Z m äm° om e r- 3 L > Z (0 )

0ö ºý, M Co

ý M M - r

MC 0O

V0 CCN

0) C_ `- C0 N

Ca O HTL m

(D cu m cq r, - 00

3 mý m w&- Mt Ný

HTL

a'- >, Z Co

1- 0 CD

C

O

C5

E

N

Cr c0

NM C .. 0

oc I-

m Ea

cava) o Z:, 3ý 'p c c c

U) .! IV C N L EC C_

CmLp .

- E Jv) CO U

D c_ I-

Cym (1) S 0-0 ý rnrne> Ooýý MME awoc_Zc 4-U)

>U)mow. m ýE o c. ý (Sc`ý'ý-'ycýcmE0

0- ýOýN Oo 0i (0 IX cýa `v c cýa ma ch ö Ocooa"c -o

o- ýº- o t' NC N+S.

S. i§ EVO um

(6 r_ O s», to `OV to O ±+ _ý

(O N! "

-CO

Cl O M0)

O O- C tf) O rNONC Co N

C cß

ý2MIO 3-c020

9- 0 Cl) v cý

V cý Co Q U) >. w 0) r wE

0 F- ý

cN E_

pE. ý mo EE 00

N3( 0cO (oc o OO cn -zu 0

ýým(Dýýncaa

0- m

O N to

V) v (0 0) N O-ý Z Z

N

0 m

L L m fß

(LD 4) C

cm CM CIL) :3 :3 O

0 0)

-j U- cc (a

CC Q N w cu

A

013 E -1

a) 2 Ö

0 c cE ý° - v L

vag - t5 s

o cc _

C p

N 0 ä>

ä-1 c (a Z co > >

t oö N=O C Cc (D 4)

ca rn o

`o Cl) .p vý ýý ýn ov >. o pNCOMÜpO

O C

16- a>"- fld

EC

O i- uE MRI a)Y

C N co C'mOON =>, m3

ui (1) 4- 0O0VNÜ UCj

.a00 dN C cýA OOVO0 in NY

O- LX... ONC U ßwv cn wd Ü-o o

ti 00 ti-T r- rlý NN

C 1ý 1ý Iý 1ý (1 C» ÜQ) f O)Q) 0) 0) (»

CCCC C` C `) O. OOO ý` pO

4) mcy mm ofßo >

rý

CE CM ((0 '`ZC

NL a)

(0 r. 0V

C" OO CL V3VC ý'

OOO Ma ÜpOC0CN N C0

(PR CL CL -Z 3 cu m

p 00 Eý aý L 0.2 00 ýUý 4) x a: c >0 vZ

a G) aY

C Lv, o c

>' U_

p

ýEL Nß

O C_ EC CpCm0

i- O CI)

ti 0ý0 00 Co Q)

oý a; cj Co

MNN

02 2ZZ

L C)

p) r OVL

C _C

N CL, 0

cni3 Nm5 O-L

E c-u m (L) m

CVC Co t/1 0 G) 4) 0 Ü0W

c F-T 3: ýe

Appendix B

bi

in ZZ IZ jZ -Z

ZZ}}Z

ccEo

pcL9E

Ie Iö oC "ý Ny2

ei

2eCdpaCVNCO; Z

Co C, ý1 7 f0 ,m '9 yaNSN L"' pC lC C '. Y E ýp GC ip c

OL NöL>ý. y pC Lý3 IC :7 61 LPL, EJ17

-6.0 22 2? 235 a ýo > 3ivIN

ýyE pN

ý' Co m Sccä >Qa Ci7 NY ,bL..

C _O

O ýQ

> 2S acc mä Em L _'ý

cý Lcc

f'n NLa 23 QorE vi ;c Y5ý aö rn cg go

co

o"Y

cg °° $ätg ;QýycI ifib

lutH

m° ý' TyNNlm 25 N'-arä2 ;Jý,

ýji mmaomsay d_ NOLp0cäpät CpIJ L Ul c= 7cö2c

aU3 dcE=mc c o.

,a°° >Qý1Npcu üi EQ r äý im yc il1I lilI!

wC N_ NNZ

UQd

i1n a

AIM

AY I(; 7N

,NrmY Mrr

ym rrn W Irn rn rn rn a) w vcc

ýrn Irn c `c Ic rý cccc iý Iý

24) g m ea eo mm Co m

° mO 0° JE 9 m Öm m

+5N 5__ m

acep__ °o °o YCo Co ým

y °° °2 IST ä_ cp

8 °ý

Z 12-- Wan`, g

c aý°i 10 t0LC

'0 ? 23ý ',,

c0',; ccM F- jc E°

10

L LY

io mc

ýj ÖEU Nc oßui

CÖ LL o 7a y

o? 0 CNUv (") 7Ov

LL adON IUrI dCsc 47 iiý 4)

5 aý äý fib.

>Ey vý W 'c

1N

aaEN Niäo NýO r;

cm, u0 SC Näa -6 m3Cc ýC Nj

_5_ °ý z v

Ü cý UUU to ý cn U -ý a LLfý !cop2° ý'Er. 8dseäöm

.

Ilv

rn Co

w : Co

Co

arl< ao co rCo

ri rc c, cy, C» CD

40

;II ýö

ýpý,

ýYIj ICNJ IS '2 c 35

iE LL

is NN -ýq O_ IN

m_YUUo --

1 ILL

Li r- -

cý II EEEE jE

IE 'E .EEEE tco

OJE CC IC c is SIC 'CSCm

U'a asaaaaadaaU

04

zzzzz zlz z

iz

2zz_Z

c O G: 'y

II' CÖ t0 m vi F,,

'.. ' c l! 20

c Co 2ä_g ö

ýÜ g_ cc0

äm 0

E°n«_m ä yfO (@p m a. 5 %'t:

° Ig2 o

cm dVE y4C N> !ßy0Ny7O :'S.

"- Zaä `° U(°

sm äU° n `ýý

uii ce 4Lg3Z4d aý väoN im mS dcc 25 c Y' ao °Q i3 m °c mýUJ

LL U ltE

Y8 I- 2? Com N 01 18.2 J Q= NNJOÖg ;-I. O lp

iý vmo, g c "ý

i f0 G c. ý aci

cc3: yt

ýsNt; mýý ýýE

N_ 0-mODc 0 cr yc

_

gý p '7' 5ia 3 Co

N 'F� C lV aEjN -y

N f0 yý JT

o fJ yNyo MIA C

CN7 fDn

O Cmn

jOßOäm' . L-. mCCä

Old äCC III

'$- Oc ýwg

°`ý y MC0 '. C C0 . 5. m00 yý a0Q twcNäaj «ä

ci p 2` Q. °° p yý arcýmö ai m>ö -e ý mo

ZCOC °)m Q' Ur.. J 'C UR f4 H f0 OU 11 U .,

J

CEL

M OIrO

SN ý'

zCO I"" p n ý. j 1ý f0

'ý,,

Y N- ÜMMY to 1ý '.,

fh M f") f") f+) NÖMnNMMMW. M li M

_ý _p T On? N Off IT9If0Q! (DO 07 'ý 01 01 6f Ol

Co r r_ , r-

m$2öV 12 m ýi °i 66

'0 Öý 23 mom mm om -o' oN 1o m co oým

_Lm 102 m

41 NNC alE N

-S xp O Im ýtNdCJ

Zm HT ýp

_výp

Ve CO

A ._W

co tC w7 tf ON yý U Vl O/ ýj än eý d'cýo'3ýýt E3

NNCO ýN,

NN dý J ýf .ý

of mSm co N m- 04NC C'a W

(, Co

OO

(0 NO g1

4-LO !

qýý

Oý

4) "ES m öE

r to I. LD6

Q Qc_ Q

LL rU 12 co U -Co !n (%f

, Go Go Wm iN

0 I0 I0

11 LO

CIC

2 ÜIÜ 0

i ', ý. '- 0

cc ý

Li 0

7 f0

2M c cý y-ý 2$ C}

ýýNIU ýýN

O

E

CCý 20 3: 27G CEN

d My ß noy

V c ýE

IQ o

c G n ää

ý

. 111

LL t . ALL .

r_ o(D , r- M

't 90 0) ONE 00 U ;m

yL LLJ

c L Co ,m ,C

ý Iw cä 'ý Co ! ýI- cu ,, m la m

M

zzz ,}zZzZzZzzzz ýz

W Gr Q. 05

r- 15 pÖ

9D

AD _1ýI §Mv ; -0 RW>Y

r. t EE Al EEH

E. or . E@

12 is

LmF y? 2N

lilI

ZJ F- dDxQ Ci

I

'C 7JUmW!!

Ii1

JJ

$5 t0

p-, .6 C2 M -a C4 Cld en co

n co vl n , nn n co ro co W)

C7 O'f i`ý i+i l+) c') f7 tD fM t1 Cl) t0 f7 co OD C) C, C,

annnnnnnnnnrnnnnnnn mo, oe oý cw w rn _w

oý _rn crn _rn

rn_rn rn_ c _rn

rn c 10 c _ý

ccWccccm 01

o CDo m co co möm CO mm co mm M0 cr ö°5

° Iga E, EWn= 'n Nc 16 0 fro - a= La

c8 o. 3 ä_ 8 it 1ö -2s 60 1cý°

äýö9

_8.2Ü W CD r- s JbH N_ HI1I1!

f JIilIlluIH!:! i

1n ? >ý ýn~ N

JO

i'o

. U(3 - 4) 9äM.

c_ is i6 K$öy r- °'

46 ENE -5

°

äL GW g$3WN

I$

cccäQc m2 EE EoEcc

ýn n CD co

co Cl? P- pl: aD

A

fn V) ýlj r

eh Y C)

Cl) f l'ý

N

9 cmýä

YEWE

Lo

t, c

rn U' U t! ý QUOO

15 z

It

h"

Fq

mZZZZTZZYZZzzZ

yp .yWUpC .pO

kC

IT

ýc ýýßä 'E ö'yJ

Hg qTqOn

-ý k

, 16c ° ý0 na

rä ° m' Mds

? mýjoA cycý4

_8 JE 1t

Im 3c .2 r3-.

yp

CyL fU/ý1 OCwCC

p>WCmcOO

LP 61 C +ý- qh m ä z Co

Sý+mv c näh OZAHcmH O_ 2yC aý

E 2 cj5

U fiH c Eqnv

ö

ý° m2ýEocý

mEýý ý°

ZJCpW Ui

l0 C

Z

'n 0 Co ö --(D MQý MMMM NN NNMMMMM fG N C+j r. - rrrrrrrrrrrrv

mW a aý rn 21 rn ý$iýr äiýrn rn rn rn rn rn ýrn

VqýN 21 CIJ

CD .E ei a) irro0 m co

A m°reC)N xSmr' mmmm Co wm

iSýN dý 1-9 gß f0$

II

>

iH iii

o Co °2

-5 w q2 ä MLý_ Ly ZEA cod' 3cn Q _p ýp ýoL L

Cm C7

OC t7) pL ýto

NN CC8 °pI! ýCp

OL il H;

M ýCY+

ý

C» D Vi Z üJ CQJ C7 M CO UO fOý yCm

X« Eyd

$'ý 8

aýiý gý0 `niv LQ°y

ögQ# le m y$ Zö cc0 A-

3C S7 yL

Ö'

. -(p ÖZ

Co lpO J CC -T C Co N (0 l0 HN£23C co

qq .ý .ý iý e0 Vý m q> 2CW

_n qCCCC Co

ii tg F v`i -n ra EEcmUUUU C7 U-

.e o 4) °

ul m <o yfYMNNr

mý of ro^ Ln

N fA NNNN OO OO 0.0 0OO2mZ

$tm icr pD

cu Z qCc to Tp7 q

.L lp $OT

OO I'ý

ýý_--- -- --

ä -a cm Ute'

tr Co

,Zä2ää T5 25 äa

ý :3Zýo o' 0ö0 :5A JE

ö u¬

cD I cy

I')

mzz tz zYzzzz

m=m

32 Z 3- .22 oäc 10 r Co -0 z gC

g Hiz

uä3.30: E2_

UJ1I (V c7,

a i' v tYR}pý c

.s m Cp c a2 o 'a ö mJ O mZ

ÖZ%7Z. mäL

'CO k

J' ÖCö

yý$ g AIL- mWp ei,

.2 T, j7

Of OOCCO (/l 61 LNd

ýO n

Z LL R d' J 9ý

J7N -Q

Sý CL Uý rwo UUG> J, *i Co M ! (4 ICD

AYN CD N

NNM N f: jN NO

cii N

QCOY 1ý

C9 ýj

Nr OBI Oý ý Of rý aiS Q1 C=rOý

§z §z `X r- iögr 3i

m0

MC)

ým cr mO2- GTOý m. ým

Co ? N;

W 22 Z (mJ

y ýp aZG Co N

. t' a0 -@ --

(3 0 j2 . 2?

3E `° 3_' a

sdO .`NO1Nä aý V. ýý

ýLL°ý Iý Eä < (0U; 8cD Um m ýz¬0

=i1 7 ýs 92e1.

E j{I m l0 CCOC

c0 tIý

aa iuI1 C, Ea ý

!

lh m2 31 e00Ü

Ný p ßf7 NOr

N Co x U') c3 : Co

en 2 ;Z2 40 Co 0

piOZZ U) AwO

ys G=L =2 a0 v A

in ci Co

c a' usi um

021 ý2 22ZzzJ

(0

N

Aj

mI} Z I}

Z} cl _

cy 2ZZZ2Z

Nk m /C G) '0 O7t!

EN1

C /0 p .QCO7N 'm ý,

awOc4 °ýi Cmcz

.c0 aý jcy7 "-" cr OI

uni '

,Cä

°D E

0. - 10 eEc ýE

C (D y °? E jedto c

Iflit äd'

O rý.

ONL C- 'co

; fC (Up CmA

02,2 0 E cm L,

im ECC LZ "c

ý1b

ý+ CZ...

31s

ýY

fC dnRyC

_yO l0 yyyO>NxEC GI L

OOACN CV ýV

18 ON0322 80.8ý. _ýJl c

L° EO LeN¬ 25 ýC ° t9 ý~ N4mL ý' _c 2S Z u) aý mN U« « LL a1Z3 > a9 .e

e> rNQ ._d ýý! is ý

4N Q

Z` ä

fýj NQNM Oi l_ O^ CO N CO cýi (O JZU (P 8

up ^Nr NMENM'.

N N C7 m r-4 N0aOY fV rE1 C0

MMMMM Qý M O) MMOM <p M Of nOQO

in n,

Ö rRM fý O 'h 1ý hN Ifs M f, f- ÖOCC Oý

C Oý

N (7) ÖC Of (D c. 2 rZO0

(» W»W Oý

0 rý ö^

2cýý C 'f NN C "fÖ

CMhC fp I^ C lÖ °' fX ýý°° IZ lee

12 r- 12 -51 ä> 2 aý

= ýýÖ mÖ mO: 9 w0

2 �o E

M. 2 d ý'_ -

c' 0°mmTO

OD a `; ý44 ä mýo Dee

% mm mg to 0°c° mm

°t

O `týý 2 m2« ým m Yr'

c£cc yggö°3v

Vm"Vc q"

we Fy

Co w Co

2 -5 yS7 lp C0yCC liqý

"OTÖ

I

-2 «s , 10

amN

Co OD röCCLOmK y-0 LL

NCY_ Wwt a- -

10 m cvd cm c ÜCNCay Ný O-ENCL 'y ýN

f0 f0 ý Z'

CL- O°cmE8 Ev Euýa>a E E°,

i, }NyI, NOm

089 <n c.

Co .,

weS f- e° Co

`0, r- ''i Co

i

12 2 >;, äööc

-9 ýc oci 'c

16 mm dN c2°E ° «" U) A-- Q L' °S3A -t Q

. r3 QyOCCy 1/1 l0 N ý+ 7OC" .

°. d

22 -2 a Im 12 3 ONUENCC f/1 c

fýA cý Ca8 ý} CyW

ýN

c°

(71 `°`0 iyý

eo Hufihil yýýfc_ cy°' `0yýIA v E

Üm `ý ä in E ä$ýi'in'ä3ý E

'Iý cco m t/ý to Q

E$ý öo

W

IC,

4 ilý lad pp

ICS

C» OD gP

18 OJ OO0 Z Z. Z0ZU !-z ýz ftn

lä ! I'ý ý Ig? ýN

rl 1°

Zc2

tn Fý C! J mmUY

15

ZmmY -c0

y3 0° i äi m1 0) °m' T 12 mb

I, -

Z zz

Iga, öi

Co

t E. cw Ü n, ý+ ö

2 y p SIE d

Co $ö

E

$o

Co J _ .. I.. _Q...

Z

le CI) Vp LC) Nr

CýÖ w

to Co m2 Co

Co r ce 2d0

k. 0 ( ßäa (0 lp

ÜN (n er-

n

cr' c

Om

dU

_ e3 LO

ýc3 öm CW

fN0 ý O d'ý

ýi ý+cQ

Uö ývUi

o 8ä

U) LL

Co r

00 Uý

CL E Si

C

y

jC to

1

f0 L

Q7 CO 'V)

oso E 3

C OL C

8 c0 z ö

E=S

o SIN ^

I7

m Ito

ý

.N t7ý

C ý

i p f/1 0

iC C

0.22. - iOO NmE

Cp i1ý I_

C y

«8 l9 ONO ýoý IoD

C 0

C O c

NC dO '

jC u

C Nom ': c

C7 .ý

E - 1121 ý 1.0 Nm QÖ u

ON 7

.; ü c G

Ö od

O- i Ir NO

M r- W

Ig E

äQ

Co C7

N

, 02 2iß

Ic U

ÜÜ !v

M - ýp IN ýp

to cý 2

C7 J D

ZZ IZ ZZ tý"

CCcCC M£ OdEOyCJCd2Ad

C Cl) NC ýC jýO7

_yä ,

24 gJO al e3 I

mý OnE

2 f' 2 1'9- 1 8C 2t2,1 c)

u. -m -12 m o, ý

qb _ý_ s

C (p It C

Co Co E O. IIý

ZOm J(O} ym

ý1U YvZ, -i eE': > -cri :, Ü

aýi a iW äY

c C

N

U) IOS lOm

Cl) jO yc)NCO

f /1

_. i m

E ý

Omi Chi Ol

Cl) th

Ü Iý F lE ýp

I I

N I ' c)

CO `-

w Iö

v oil NNCNN

ýN IF- C)F- fhb mm CO m

1% 1m

Im CmO ja'$. C1 cm

SfOýI Qi

CD c

aVy IZ

iÖ to IZ

Iii i S_ ý Ieýa

cc E co CA yN UY

N iGp 'r IcD a0 m IM

is

c 0 m`

Co

i c, CO IC v

Ito CCM

IN ýI) IN I IZ Z !Z

x

Co

Z: 1

}ZZ E

ä

X _y

Co

7

yrý'

ýI

Cy7CC

J fý ZJ02

Nhýý

Ifs Ifs ý

Of 4iýOf O! c_ W

Co i 9qcC a j2 -C

PD d OrmmO

Co

TöL 8md f0

c5^^m

cä ° E

IHI

HJ 7ý Ö= UJ; y Ný U ýN ý

C f0 E 0

iA ý Ot V f0

N .9

OM ýý pc b

`mom (r1ýý CO Vn r4 N 010.. Q

N_

d=' f0

Lo Lo m

Appendix C

W

N

NýC

,0 j3 Z: 90 Ai

v L

W

D1 V 1gd 9D 3:

j2 do v ß- 16.

w B U)

N R('19Q 4900 W

LU W

--I FWww Ui WW

N! A

N iZ V-1 r-1 P% 'M'I N

V- V "-, 9 "-' "-' 8TTON

WÖWWQWWWW

't (f) co Ln(n0 00 Nýy1ý

G0 ai

Gº 0" c, o) cC

ÖÖ

W W "' W W W

'1V Ö NO f-N

V-1 ri

ö öö S'9? öö ýLný g 8-4Lb Lb N ýý 0 ON

lttý MtGCMui

01 m

ö 9o o9 900 WWWW W Wp WWWW

N to

un yul Ln 009 .y9 00

W WWILJUi WWWWý N °ý 28Ö en en 0NOV cri cri 0 It 00

GC CM C "-1 ri . -I t0 N

t W B U)

N >- C

ö öä999

öö WWWWWWWWWW

tC 01 Ci cri cV M^

. -i

WWWWWWWWWW 0 (0

et m ui ýi cri aci ch d' 90

ý ml OI WI

ýI

YI - _I to 0

CCCCCCCC ý1 NI

999999999 öö WWWWWWWWW bib N rrm0'0

v U. ) M-: oir,: N LA 't

N- 1- 88 1- Ir- tn 1- v- C) 0 WWjWfWWW Wcý W 0+ 0+ ýtiýýÖNtiC7M

in Cri aD1ýýZ0 W rai0 -, 0

44

ö to WWWWWWWWW '' 0T-Ninco ýp

8O's Nv

LA

? e? e44444 öö WWWWWWWW Wld' NNýN0X0r-M >ýC. >aD W >ýUlcD CO et rrc7Nrc4 Ir-

O 01 M 01

e? eQ1QeQQöö WWWWWWWWW '' $A?

00OD Cý NOr

1nw-) 8WW

OýMN rýrCi-: Wr WOi

00

Wýp WN W Wo W Wýp WW Wýp ''

aACOIcýýQ 10923%0

v- In Me- Mv- le- ýq-N f7

l 1-4

WW W8 WW WM w w? W m8089 inm Ci (ye c4 tVM9-ei ui efr%-r

01 ßl'1 NN

NNN NN NNN 9 44 c2, c3 WWWWWWWWW '' 2000e080N WW mMMOD-: N. c4 M

Ll N

Q_1N

ýi Ni N `c iý Ln %D XýXXX MINI co

Ft& H!; co mmN

mm mmmmmm

MMNcnNmmfhN 0000099(? (?

0) 9 MNM MMMMV)

OOQ4Q000Q WWWWWWWWW W WWWWWWWWW

NO0nNNNC%N O _ %D -I 'N

tDý ýn N N d ID CM

IR ri06

cJNO M4 C6

0000 tO r' 'OO 000009900 rl V-1 V-I "rl r"1 OOQ Q QOOO Q ffff'WWff + iii W W W w Mn 9-4Ln

tr ww 00 (n

W LAN 1 lfN t pO o'

Ö M ' 4 :

Cifn gf N . r-oiOLfl N ci ad

It ( ))

ý4 14 ý4 14 V-1 14 1.4 %6 t6 wN ON

10 1o %D %D %a 9 %0 %0 opOooQOO ÖÖ ýýýj ÖÖÖý W

C%%DNtt Ksoa; N M 000ºO , OMGtO O .4OaN N .! N

1-4 IM V-1 V-4 rr

. -i r r1

%0 %0 %0 %0 %D O 92 %o %D oooooQ oo %D %D OO Ln %0 to UA 00 990009 WW Üb th WWW

tit W W WW Lu WWWWW

Q% CI NM NONM "I 81 , 0) PS fn v-4 C- NN v-4 (D 04 0%NC101NtZcorgM M00 N V OIll. -i

rý Pý L6 LA %6 M ( to -4 . -1 %0

ÖÖ

th Lb W Lb LbWWN I WWW ý o tb .

; rl

-HOC 0O N

000 M O r tp O Ql d' ýp t 0)

ºOO , O -: " d C

. -- MOO

.4 r4 4

rl r4 ri .ý 010% ri ,A .-

oo%D Ln %D ooo9 öö W

UJ WWWWWW M. -mN%Dco N in- r4 . -4 O 4.0 "4 CO 9-1 Imo' ;N 06 MIS

0%0 %D 0999000 000 WW_WW_Wli1W I

t, CMNrýLLAÖOODD d'N 'i N%6

0000099009 co Qv 00 WWWWLLWWdJý W WWWWWLWWW Mmin0%(7,

Ný j ýd *3co ýGtý

01t11ri .. 1,0uiCV

Q10CºLQ 0º ý .-

fl ON O' ý 1ý dN . -1

NNNN%CNNNN N NNNN NNNN 00000 00 Q 0099990009 WWWWW Wiy WWWW Lb

WWWWWWW Wýp

ODNOiMNM. ý-11ý N

'V. ýtG2mORI"-I NM

,4 wi uir<c6rýZri t000 co

MMNU)eTýv"IO

Go 0%0r1N tn %o

1-4 V-1 V-4 VA 1 v-4 16.1

,. ýI NI ý, rýl I Ný of

I .ýN MI

UUUUUUUUUU

1G 1lf M Ce) Q1 1--1 -1 N

=Z=ZZ=i== m le le 2 je

0 00 OO O QQ? ß W

19 O0% . m--

0740 -8 000 +W SWW

LU V- 0in%o ýcDÖ

rr . -i . -1

WWW U" LU Ö

T- V-

%D %D fl% 000

ee4

www le 00 C 0 '-

° o t%0 C N('1N

tD NN 000 Lb Lb öOD N iöö

of ri c+ ci

IIIn W)

UiW Lb Lb Lb op in N Oet%G 1ý2N

00 999

'lot ý

000 ý_4 `'Ö et CV-00

NNM

99 NNN9 C*4 C14 (14 OOO

Lb II ui O^1

N %0

Mq

to ui 6 Cd u

V-1 1.4

Go 0 C#) v0 1,4

00 N-r . Y e-ý

e ly

2II f2P

N M O

m W CO CO

1 W

M

U)

(Ä+'' +

tp OD

N

N C GO

60 2 PD 4-1

eöö9eeeoo

0 Co 22 to Ln N Cl

e= 00 NM Co N

80888800 +wwww ++

Cd MMad ad ad ad Mf'7 OD 00 CD CO

0 ý 0 0 00 m W WWWWWWWW W 3

L rrrr W

0

c» r ýZi

N 0 W

%D

Ln

NNNNNNNNN 99999 WWOO

LD ui uito%0u2viui uitg10

Ul Ln LA LA

_ L

W

B

to O ao y 90 -41.20 f

N O W O

rn

0 0 w 0 N

NmC14MMC4,4N 99999200

00CY00(i ý(4MN

It R T- 11: -! . -, to 1-4

8ööög8 8 ggöö ++ + WWUJW Lb tLONCOCCDNtNNGo 01

1-ý Q1 n OD 00 f-ý L6 "I LA n +-I

e4e4e99 00 WWWWWWWWWWW COýýýrýt)NN 1ý

n NMmltMr'Ne- e=

. -i N

00 WWWWWWWWW+} AÖÖ

C0 OD co CD CD 416 ~ Cd cd MM

co 00

wwwýpw_Wý{WWwwww mýNV ýf-

-C14U! P: " NN

r- rV-ýV-V-rV, T- V-

14 1;

NNNNNNZ; NNN ý? QQRQý? Q OO WWWWWWWWWWW uý lý 1n lC1 !ý 1n ýA %[) cpcpmcpcCCO9cq, 4, %00 0 0 ºnU)t[)I)1[) I0 0

LA L-

g44ý2ge? ý öö Wýp WWWW Wýp Wýp WWWW 9-1-192

. 4i COOmCý Md

c+» cV ,- od c. - r1

888888e8e öö wwwww www ++ rP. tAr 1ý O -ww

ce �, n Co le c4 Co 0 rl% 1

c6 rd cyi L 00

- M 0

MME' c") MMM QOOQc9QOO

Mcr1cýM MMR1M gqQQ449QQOO

Lb Lb W Ä co ONmOf OOOýNO G oN ODa-V

N(bvýN

0 mN MruicMýfU" 0 NC) lVýcd%6 nj

W 0 N

0 W m Oi

10

90 4.1 Z

0

oC

V. - rrrON1.4 V. 4 2999 WWW

Wf 0

in 40 Co :3w ui :3 'le ce ci

tp G r- (0 CO CD 9

"; in

M0ýCDae

mew=CI I _Itn%o .L V-4 1-4 Ch j) th Ch CO COW I CCCCCCC4N

OCaCie W. M0: W. nn

T- 0ö WWWWWWWWWWW

to L6 r,: t, 6cd vi 6011%

WN 00 nm T- r

=1>>x1Xxx5<5 1�11N F-F-1-'1-F-F-F-F-F"

of to

mmmmmmmmm

000009900 WWWWW Lu WWW OD. -iNNNM Q1 O +1 O t0 N . -1 Ld N, -4 , -401tC et CO

0ö 449000

iýMtDtAýtoIR 11 1146 %0%D

00001-40900 00 000 oooooa oo 8 0000000 0 WWWW WWWWW W ++WWW... W+ f "q MMMMMMMMM Co MMM

I CO 00 c16 00 CO CO 00

Co Co ý GO pp

Cý OD Cý CO 00 co

%0 %0 %0 %D 00 Oooo09999 %c %0 %D o0000 LiwwLbWwpw. ý} Ö MÖ

w wwwwwüýwWw M O^ Ný MNN cl ..

'. l ''i '1 4 Cy 1"4 ̀ -4 NNN 1-1 . -i

NNNNO (N NNN 000009999

N 9 NNNNNNNNN

009999999

w ý ý m m ý1wAýý m W WWWW WWW

Uh ºo to in Ln ý i ý W %C tD tO m %D U)L i 0 to W %D 'R 'R q Ct Ct p it i LA Ili LA 6 Ln to U)W u

1! i Lei l! 1 ý to

NNV-IN%0 NNr1 OOooOQQ OO

N 9 fVr"1NMNN N

OO 9 WWWWWWWWW OD 00 . -1 M ri

ýMO tý w IV

QQ 0 W WW

ýp Uj Lb WW

LA LA 2 cl M OD %Q qtqtMOOýNMv N 00

0 t`7NC%coN .M r1 ri NN 'O -I (ý NN

0 -l, --4N09Oý 000000OO OOgg 0008 OO

+' +'}. .. ++ Lli

OOO ++LUWW .. .w

WWW LU00 ýp WW N 9T at 00 N aD N 00 N

WWOO ýrý WWW SON in C4

9O , -i . -i Oq c6 N. Oº N Ci q d. oD Iý It . -I I:

MMNMscMMN OOOOOQQOO

Cl) 9

MMM ý")lý'1mmC 999 9 99Y99

WWWWWWWWW

MO to N Nýý (Fin W Q

WWWWWWWWö

N fýmýý ýa P-W NMCA00Nrf"tN N OGýý Ln ': %6+1LA0; Nr1 %6 GO %6NN

1-1 r4 . -4 . -1 to . -1 O OOOOO QQOO

' 9

1 1-ir- V- . -4 r1 r1 V- O OQQ OOOQ

WWWWWWWWf w wWWWWWWWW 1-+ODQ1V "--IN N M%00%00 N

NOD r0)0(1w 111

ui tC0o t- M-cMIýcC-1

1lý Gº Gº 06 r1 N ý4 it It 4 NY %6

0OO1O, -1N Ln % ýf

v rf ý1 ýI aIs

(n, 1 ., 1 NI chi %0 MM

cý coM U a% V ý V-4 .iN V11u? CP fä :rI =2=i II I= UUUUUÜÜUUÜYYYV. Y

000 ee

9 WWW cl 1

to !? 01 ýp 0p ýýý .1 GO Op

000 $00 000 WWWW Wa W MMMMM

cd cd C6 OD 00 OD

WWWwww M

V! N NMI

rr rl . -i . -1

NNN9 QN

9 C) 0 CD ib IL

wwWWW uJ %D%0 0 Co

Ui LA ui ui ui

NNN NNN

WLb IL map N cM c

M rr C

000 880 0 +++ LLJ in co

NN

O 00 N NC of Cd co 6na,

MMM V) Co) V) 000 W

Ill WWWW

ev- Nr PS vi "4 r*4 0! CV MMm Ir: Ir:

000 999 WWWWWW in 0

tim ti r; %6 06

a to '"1ý co' ý V- VV U- ut

S== t- _ i_

0 OD

U) n OD iO 41

4) 0

N v L

N

90 L U) Co

0

. -. v U) 0

L

U)

a) Q

C m d f

v

L w M

OO 40099O

fWWWW

OWW ýWp

WW W ýC tD O njo

ý 4: 0

Qý OR 1

ö9 9999 _ WM

WE ; Ei ONOOO

CO 19-

0 1*k*k0ý0ý#ýýk

O O W O 0

ßd

gn a

W

cn

G 1C

N

A v

B

n in

i ý U) Go

C

II ai

90

L L

U)

4.,

2V) I_ýo

vý 00 LA ýc aiOi aZE

0 0 w 0 0 ö

0 0 w 0 0 ö

0 0

00 in V-1

W Wff WWW Wp WWÖÖ pM ÖÖ

Cl OO N coOOCOcDcO OO

66 00

III0 I

0o W`LLW

m 8ýOWO ýCiti00

MMMMMCr1ttm; 7

Ir- "4 V-4 Irl, I- Irl V- V-4 v-4

o 900999900 Lb WMMWWW0

iB -0 O

Od0'00 00100+ I.: Qg-: P-Nz 1ý0. O

NNN

888888888 0888800800800 wwwwwwwww 'f' + 888888ööö äö cc oodoodd00

00

888888888 0Q WWWWWWLUWW +'}

888888888 ö LOU ööööööööö00 00

888 8888 öö Wp Wp W Wp Wp Wp W Wp Wp WW OOÖOOOÖ00 OO ÖCÖÖÖÖÖÖÖOO

ÖÖ

88 88888 Wp Wo Wp Wp WW Wp Wp WO++

000888888ÖÖ 000000ÖÖO OO

00

3Sg888S8g WWWWWW WW '+' +

888888 8800 ä CCGCCCCCC9C

00

u9 W_ ýp

WQtý WQW ýWp

WN WW T- cyl ui ui

rrýODrt=Kjr r O'Qý Qº Oº

00 WWWWWWW WW '' 8ýp$14IN CS aSin M Nr- vi --lýjNlýrIN MO

01 et

WWWWWW'} WW '' (A

14D

chýritaichchr. cýcM

.- r- .r r- .- v- 9- 11 14 494999 Q9 oo www, WwilwcLl LL LL 2r0 U) CD

rN 1- e- N

O1ýO

0 W$ O O $ o0000o000 OO o O ff tttW jf t W ff W W W t 't' ý' W 00000OÖOO p

p p p 00ppp000

000006000 0 OO ci 6 ci 000p OÖOOO 00 00 000

OOOO Opp OO 00

p O

008880008 ' 00000 00

fff t. WW ff j o0 -000 ++ WWW fff Lu WWWWW pp OW WO

OOOOOCO U 00 WW0 ppp WWWg

pp 000 00000000 O OOä cs 6000d OOOOÖ 00 00 000

000009900 ..

0 $ 00200000 p

..... jf WWWWW W f tf ÜJ W }WW

ppÖÖ OOOOOOC 8 p p ÖOgpOÖÖÖg

OOOOO0QOO 0 OO6 ci0 000 C5 ÖÖÖÖO 00 66 coo

000000000 0 Ö0

08880008 000 WWWWWWWWW

OOOOO g$ OO

W g

WWWWWWWWW

OOg g$ OOOg 000000000 6 OO C5 p0pOO ci OOOOO 00 00 000

OOOOOOgÖÖ 000000 O OÖoOÖOo

WW p jjWWWW WW WWWWW WW 0000 0$$ 00 O

ý W 00 88800

08 OOOOO ci 0OO 0 OO ci O6Opp0 OOOOOOÖ co 000

NNNNOONN OOOOOQ OO

º- 9 t0O1OO tD0t0D 1G0 9

Wýp WW WWWW 0 0

ý WWWWW Lb WWW V) ONw%DtCm S( W

3 14 N OMMGºC4 N0% OO OO OOO ,i"0 0) 1D 14 N T- V- 4 ý-1 V- Ob N 00 O' Cý N 1 . -i -ý . -1

ÖÖÖÖO Qý ÖO ' ÖÖQQÖÖÖ

WWWWWWWW C1N ONO%O5ON %

WWWW LLJ W NJW. U1N FO;;

Ci (p 0% O +1 r1 d; CD M 11 cfl

0% 00 V- O Gq W CD tO M 00 NTM LA

t0%D%D%DOO %D%0 OOOOOQsOO

? t0tC tGI0ýC OOOOO

W WWWWWWWW M fn N ý (11 V- ED Z'

11 LncicoÖMý 1 M G1NýjMC)1nNM

' MNMM 1 MM 'ýýM d

W-4 W. 4"1- ýx- -, - Q OOOOOQ900 WWWWWWWwww ON Cri 0% (X) et ý Co C1 Oº Co OOOOOD 000 M

V-4 ö99 Ir- V-4 T-4 0 900 000§ WWWWWWW& LW

it -IT V-00%mms

nntýlýtýnnn~

000 888

ooh 888 666

000

WW± 888

ýýý 888 dd6 000

000 $gý 000

ööö 889 dcö 000

0 000 088 +++ wUi w 00 000 0000

ööö 000

000 ÖOÖ coo +.. Wp Wp W

000 ÖÖO odä 000

WWWW "' "' qp MR OD

ýrý

000 Q9 di Lb Lb Lu w

t" C) at C) 7 :7

0 0 0 Lb Lb LL www

Cr) M CV) MM f'7

000 T- 999

1t Lb ui

OÖn g

nn

n

+ W+00 +WW Wp W+00 + Öe0ÖöOSÖÖ

Ö99ÖÖö000 00 Ö0

+800880000 OOOO WWWW gOOggp 800

OORÖÖÖOOO 00 00

8008888800 +++++++ 00

00 cc

00CD.. 00 +++++++ 00

800088800 ö0 o000Ro 00 00

800800000 ++ ++ $öö$8 ++ 00

ö°°ööööo0 00 00

SoýSSSSoý ++ ++ ööööööS00

ö°°ööööo0 00 00

Rt L.

tn 0 co

i ce

00

L U)

r-. a v

L W B N

c a) f

L W

0 00 ýý 00 WWWWWWWWWW

OOOO8 88800 O600660000 O 00 00

8 cc wwWWWW

88888888 öö 00

göög. gSöö $g$g $g. co

öS LOU UOJ öö 888888888 ö LOU 0 00066d00.0 6oo6666oooo 0 00 00 00

0QÖÖ Q'

j9ÖÖ

uj CD w P- ö

W%ov114: - N4Lb

"C1rrt[jtfý

8888888800 00 WWpWpWpWWWWWp 'ý + p 000088888 eo ÖÖÖÖÖÖÖÖÖ O

ÖC

000000$OO OOOOOO 00 WWWWWWpWW

000008000 OOOOO cjcj OO Ö66O6 00

000000000 000000000 ++ WWWWWWWWW

000008800 OOOOO06OO CÖÖ66 cc

ýo g ooo + Wp O

+'ýWpLUýj {' 00OSOÖÖo

d OOÖÖÖOOO0 00 000

O Ö ÖOO800

0

8 0Ö$8 8ÖÖÖg

p 00000000 OÖO 00

888 OOOO Oý g00p00 8000 eoeo

f 00 fWWWfffW fffWWo++

LIJ ÖÖÖÖÖ ci 800 ÖÖ6ÖÖCpÖÖÖö ÖO6OÖ

00 00 000

000 WWW

Lu

W+

LL, 000 ÖÖÖ

6OO

000 + W WW

W"'

000 888 coo OÖO OOÖ

000 00 ýO

jO W T t

000 888 000 000 6ÖÖ

Ci mý0 ýn ýýppppO..

ýrvpny

M NOnp

PabD P

OcYppbý0 mp_ ýPýp

Pý OVýVJa

OP1 °wýgPýO

ýD aäm ,OnP

ýiNNOO S n° rN

o

nOýýPmHýO PNNnmG Nm NÖnP1ýýIýNryÖnnÖmnn Na m°ýmn ^°. ný nubiN un°ý

°«ýiN a mNN SNP ýl m0nOn aý

mONýbnOO

rvý nbT

Nö e ýýN ý mNmx°m° N ýp N Vý ýp mN Tý N Ný mNbb Yf N ýD

N ip 1ý b ýD elf b ýD N ýO NN YIN MAD ýO nNPN tv

Ö. nMIPtiaTNnmpfmýmONNINV M ýn. bi wýmp

gyp,

NPNpppQýýýN^ýl ýmÖ tinNM Nlýp. miN

ýýn$'ýnrvuRNOPwbNNýý PýQ I!

hýý�ýN ÖnýrvýÖNiPD

ýrv'INýMnPGD ýÖ

,? 1'a ,, b L^Onrn$öno

i i`ýnnöönn äön°mm. --! °nöý p un

o ýäý vi viY aýýnýv . i.. 'i c mýýýy, eui d in viT uinn

ýö ýýO ýoýäý': ýöurvi äiöömnrvanöuN+n. äýD ßäö ýn ä`V ýn nwýirvmaäýoTg mä°p

. °. n$uni $$ °n ýö u'"i 'ýaävmi^ä.. m, °ý ýnmý"n.. °n ään°äenöNnmýeN. ýä ýý äN. öö i ý$ýomnöänýoäN °ýin äb'aN q a. +nmmv nnbü

ýe f Tý^i vif ýYi Yi<PYi Iý ýG Yi Yiývi C mbb V výcý ý v+rý! +.: ýr: mýýi viö

býýýýopfýýýpppýýý°ýý°'Xt"pgmýpýýaýamp B8ý NNQÖýpäpmýýMýý, öayp C 

Pýp yfnlýp

äNHh n^GDwýw fýNý.

pýIOýpN

wýNýOppppý PPNn

, ý, p(pmm ýO

E

ýýbýOmNýOý

ryNÖPý wGONOmýP

OýOýýmýgNNOP

N ööPNN

3f OE CD $ýOmn

. °i

mýn nZ ýn° mrym N. ý

N Týv riör:.: qýJ 1i 1vmm v

nnT ti eyf Tri . {iw Yn `^? ui uinYwriý

a nýmNýTPýN^ýnrvýÖf

N"'rvn öÖ fýO1ýMp.

OSTa Mrv

PN

Ný0

Nm OHOýppp SON ýpD fO Iý mNýýOpOyýfOýý ýO

p °nýD nýOm ýNn wýG N aýDPnÖnýÖm On SON mýrvNýO

J nöö u'nnnnfi $ýöm '°N`Yim

, m. nm ̀Y^'ý'ý�^'oý Nýý ýNnNa bný, ý ao ýo öý a, o ýeN n er: ýee f eNV+eeui oýý ý eýýem, ýý: ýeýnýi: eeýýi Pnnaö uien

gNPTNälmm$ rv

nb rv bYbbnMmmP rv mfy

.6Z

I IM

IHGS : ist - ýE _r

gnnýqqýTq ýq

1UNIVERSUM'

OF BRISTOL LIBARY

IS