Professional Documents
Culture Documents
Full download Challenging Preconceptions of the European Iron Age: Essays in Honour of Professor John Collis Wendy Morrison (Editor) file pdf all chapter on 2024
Full download Challenging Preconceptions of the European Iron Age: Essays in Honour of Professor John Collis Wendy Morrison (Editor) file pdf all chapter on 2024
Full download Challenging Preconceptions of the European Iron Age: Essays in Honour of Professor John Collis Wendy Morrison (Editor) file pdf all chapter on 2024
https://ebookmass.com/product/politics-christianity-and-society-
in-malawi-essays-in-honour-of-john-mccracken-kenneth-r-ross-
editor/
https://ebookmass.com/product/classical-economics-keynes-and-
money-essays-in-honour-of-carlo-panico-john-eatwell/
https://ebookmass.com/product/beyond-the-fascist-century-essays-
in-honour-of-roger-griffin-constantin-iordachi/
https://ebookmass.com/product/bombay-before-mumbai-essays-in-
honour-of-jim-masselos-prashant-kidambi-editor/
Managing Globalization in the Asian Century: Essays in
Honour of Prema-Chandra Athukorala Hal Hill (Editor)
https://ebookmass.com/product/managing-globalization-in-the-
asian-century-essays-in-honour-of-prema-chandra-athukorala-hal-
hill-editor/
https://ebookmass.com/product/landscapes-of-human-evolution-
contributions-in-honour-of-john-gowlett-john-mcnabb-editor/
https://ebookmass.com/product/age-of-iron-on-conservative-
nationalism-colin-dueck/
https://ebookmass.com/product/challenging-economic-journalism-
covering-business-and-politics-in-an-age-of-uncertainty-henrik-
muller/
https://ebookmass.com/product/tenacious-of-life-the-quadruped-
essays-of-john-james-audubon-and-john-bachman-john-james-audubon/
Challenging Preconceptions
of the European Iron Age
Essays in Honour of Professor John Collis
Edited by
Wendy Morrison
Challenging Preconceptions
of the European Iron Age
Essays in Honour of
Professor John Collis
Edited by
Wendy Morrison
Archaeopress Archaeology
Archaeopress Publishing Ltd
Summertown Pavilion
18-24 Middle Way
Summertown
Oxford OX2 7LG
www.archaeopress.com
ISBN 978-1-80327-006-7
ISBN 978-1-80327-007-4 (e-Pdf)
Cover: Extract from Pilbrow’s original illustration (1871, Plate XXIII) of Roman Antiquities from Canterbury.
Hillfort at Berber Hill, Kenn, Devon. Photograph F.M. Griffith, Devon County Council, 29.06.1984.
All rights reserved. No part of this book may be reproduced, or transmitted, in any form or by any means, electronic,
mechanical, photocopying or otherwise, without the prior written permission of the copyright owners.
This book is available direct from Archaeopress or from our website www.archaeopress.com
Contents
Preface������������������������������������������������������������������������������������������������������������������������������������������������������������������ iii
The Atlas of Hillforts of Britain and Ireland: a consideration of the coastal and inland promontory forts and
enclosures of Scotland�������������������������������������������������������������������������������������������������������������������������������������������1
Stratford Halliday and Ian Ralston
The European Iron Age. John Collis (1984). London: Batsford. a late review������������������������������������������������������� 49
Chris Gosden
Since John left Devon: some unanticipated outcomes of aerial reconnaissance in the county����������������������� 70
F.M. Griffith and E.M. Wilkes
Mam Tor, Derbyshire: new plans outlining hill and fort, internal platforms and all��������������������������������������� 79
Graeme Guilbert
Some reflections on phenomenology, structure, agency and actancy in medieval pottery studies�������������� 136
C. G. Cumberpatch
‘Friendly Hills by Nature Guarded Round’: Recent work at Bathampton Down, Bath������������������������������������������ 143
Lisa Brown
i
ii
Preface
Challenging Preconceptions of the European Iron Age is a collection of essays by some of the leading researchers in
the archaeology of the European Iron Age, paying tribute to Professor John Collis. Since the 1960s, John has been
involved in investigating and enriching our understanding of Iron Age society, and crucially, questioning the status
quo of our narratives about the past. He has influenced generations of students and peers alike and has been
one the strongest voices in the demystification of the ‘Celtic’ world. John has never held back on his continual
questioning of the past, and he has been instrumental in helping us unpick the labyrinthine tapestry of myth
and misinterpretation, interrogating the way later prehistory has been traditionally investigated, packaged, and
presented, to reveal a much more interesting past, filled with more nuance, possibility, and humanity than the
uniformly structured Iron Age of the early 20th century narrative.
The idea for this volume was born nearly a decade ago in conversation with my then doctoral supervisor who had
been a student of John’s (and one of the contributors to this volume) Chris Gosden. The list of contributors is not an
exhaustive cohort of those who would wish to honour John, nor of those to whom he has been a profound guide to
their careers and research – indeed that would stretch to many volumes This volume brings together papers from
more than a dozen of Professor Collis’s colleagues and students to mark his 75th birthday. The contributions range
across later prehistory and the European continent, taking in major themes that have been his prime interests -
hillforts, how we use archaeological data, socio-political structures of the past, and of course, the ‘Celts’.
The volume begins with Strat Halliday and Ian Ralston’s treatment of Scottish promontory forts. Drawing on the
wealth of data from the recently published Hillfort Atlas Project,1 they offer a comprehensive treatment of over 500
promontory and coastal sites. They highlight the challenges facing any attempt to classify such a diverse feature
type but the challenges are addressed neatly and convincingly. Through careful and rigorous analysis of their data,
Halliday and Ralston offer a compelling challenge to previous received scholarship on the siting of these features.
Their conclusion invites an exploration of the wider Atlas dataset,2 freely accessible to all, so that multivocality
of interpretations will enrich our growing understanding. In the best tradition of Collis, they are continually
questioning the status quo, backed up by good evidence and a dedication to making that evidence, and the new
interpretations drawn from them, available to wider audiences.
In a similar vein, the second paper by Henrietta Quinnell also calls for a ‘reworking’ of prior scholarship based
on the potential of not only new excavations but also re-interrogating older assemblages in the study of Iron Age
Devon. Opening with a wonderful reminiscence of her first professional interactions with Collis, Quinnell quickly
sets the stage for the current thinking about ceramics as chronological markers in archaeology, and the thorny
issue of dealing with largely aceramic communities. She makes the excellent point that many groups around the
globe have done rather well without ceramics for large swathes of time, and there is no reason to suppose that a
paucity of a ceramic assemblage will undermine our understanding of peoples who used different materials for food
and drink preparation and consumption.
In an age where we are indeed swamped with new excavations and the ability to revisit archived investigations as
well, it is fitting to ask some questions about the massive datasets – the embarrassment of riches – we have access
to. Martin Kuna explores some of the challenges this presents us with regard to such variables as archaeological
visibility, drawing on recent work he has done on the prodigious evidence from Bohemia. Mindful of the reliability
of some quantification models, he makes a clear case for why this big data is an important resource, but also why
we should be cautious in the conclusions we draw from it. The article is a refreshing antidote to some approaches
to ‘Big Data’, which advocate that if there is enough data, we can assume the robust will outweigh the unreliable.
Riffing on the title of Pierre Clastres’ anthropological work The Society against the State, Sophie Krausz looks at
prehistoric and proto-historic Gaul and the way we have thought about their socio-political systems. Challenging
the still-dominant narratives of a Mediterranean-imposed shift in organisation, Krausz invites us to think about
things from a point of view that returns more agency of choice to the Gauls, rather than seeing the formation of a
Gaulish state as a natural step in the linear progression of societal development. Krausz’s assessment of the stages
and influences of urban development in the Iron Age is particularly significance against this refreshing backdrop.
1
Lock, G. and Ralston, I.B., 2022. Atlas of the Hillforts of Britain and Ireland. Edinburgh University Press.
2
https://hillforts.arch.ox.ac.uk
iii
John’s interest in non-linear forms of complex progression in political and urban development is also referenced
in Chris Gosden’s contribution. In what must be the latest review of a publication to be written, Gosden looks at
John’s unrivalled publication The European Iron Age, giving it not only its due in being the only major comprehensive
attempt at such a geographical and temporal span, but also offering critiques along the way, as any honest review
should. That some of the omissions of John’s book were the product of those themes only recently coming into
scholarly vogue can hardly be seen as a true criticism; Gosden has written a fitting tribute to a book which has
influence generations of archaeologists and pointed out how we can build upon it.
Andrew Fleming provides the volume with a call to consider persistence of a tradition often taken for granted in
some periods and regions of Britain, and totally overlooked in others. The practice of transhumance, or the seasonal
cycle of moving livestock from one grazing region to another, may, as Fleming argues, have much deeper and wider
roots on the island. He shows through a range of evidence sources, both scientific and observational, that this sort
of movement of people and animals in a landscape needs to be factored into our prehistoric narratives and suggests
that some of the patterns of land modification and earthworks building we see in later prehistory may be indicative
of transhumant activity. Fleming’s images of modern transhumance in action are particularly compelling.
The call to look more closely at patterns in the landscape is reiterated in the chapter by Frances Griffiths and
Eileen Wilkes. Acknowledging John’s early interest but subsequent distraction from the prehistory of Devon, the
two authors report on four decades of archaeological prospection that have, by looking at the ‘big picture’ of a
wide landscape, produced a more densely exploited and settled prehistoric landscape. Drawing on reconnaissance
from above and below, both aerial photographic campaigns flown over decades, and geophysical surveys, Wilkes
and Griffiths show that not only is their dedication to the archaeology of Devon aiding our understanding of the
past, but also preserving it for th e future; they outline how relationships with landowners and farmers are key, and
how discovering new archaeological features enables farmers to get support to look after them – a collaboration
between researchers and the public that is to be commended and which should be emulated more widely.
The lengthiest chapter in the book represent the first in depth treatment of the plan of Mam Tor, a nationally
significant hillfort in the Peak District. Graeme Guilbert has been researching Mam Tor for decades, since his first
introduction to the site by John Collis in the late seventies, making this contribution a fitting example of the sort
of inspiration John’s enthusiasm and passion can lead to in those with whom he comes into contact. Guilbert’s plan
is a first, recording comprehensively the myriad features of the site, whilst evaluating what previous attempts
have observed. The plans, particularly Figure 11, in this chapter will be beneficial for future researchers of the site.
Guilbert also draws our attention to the fact that whilst Mam Tor is on the outer limits of just how high up a hillfort
settlement can be placed and thrive, our modern perceptions of what constitutes hardship and comfort (and the
ability/luxury to have such considerations) may bear little relation to the lived experiences of the hardy people who
chose to modify this landmark hilltop.
The macro scale of one of the larger hillforts in Britain if followed by the micro scale of a single assemblage. Tim
Champion presents new thoughts on what may be the earliest of the south-eastern Late iron Age ‘richly furnished’
burials. Discovered in the 19th century, and frustratingly reliant on descriptions of that period due to loss of many
of the artefacts, Champion nonetheless created a vibrant reconstruction and analysis of the collection of grave
goods, and crucially, sets them in their significant context of the development of pre-Roman Canterbury.
Staying with the artefactual focus, Chris Cumberpatch writes an insightful paper on the study of medieval pottery,
a topic he states that admittedly may seem at odds with a volume honouring someone who has been the voice of
the European Iron Age. Inspired by his early research with John’s pottery assemblages from Iron Age Auvergne,
Cumberpatch takes us on an exploration of what use may be made of the theoretical approaches more commonly
used in prehistoric archaeology when applied to a medieval dataset. It is a fine example of how John’s influence is
far ranging across areas of study.
In the penultimate chapter, Lisa Brown provides a closer look at a hillfort that has seen much attention, but little
examination. Made famous by peter Gabriel’s pop song, Little Solsbury Hill outside of Bath has long relied on the
scant evidence from earlier 20th century small investigations and the generation of ceramic and faunal data cast up
from animals. Brown puts these finds in the context of new work conducted by the Bath and Counties Archaeological
Society and offers a clearer understanding the nature of the site in relation to the early days of pre-Roman Bath.
iv
Finally, the volume concludes with Oliver
Buchsenschutz’s critical assessment of archaeological
mapping of ‘The Celts’. Fittingly drawing on John’s
body of work and acknowledging the pitfalls and
potentials we have available with such multiple data
sources as placenames and linguistics, Buchsenschutz
offers a way through the labyrinth using data mapping
to get the most out of overlapping strands of evidence.
His admonition to beware the great swathes of time we
are dealing with, encompassed in a single period is a
relevant one, and he rightfully calls for mapping to be
reflective of this.
v
vi
The Atlas of Hillforts of Britain and Ireland:
a consideration of the coastal and inland promontory forts and
enclosures of Scotland
2
For this study, we have retained the pre-1974 counties as the most
convenient and best-known geographical divisions. The regions we
1
Individual sites in the database can be accessed by their name and have used include Greater Tyne-Forth – by which we designate all of
four-digit numbers, here provided in square brackets. The database south-east Scotland plus Northumberland.
entries include site bibliographies generally not quoted directly in 3
Northumberland includes 12 Confirmed, 4 Unconfirmed and 3
this paper. Irreconciled Issues sites.
Map 1. A maximising view pf the distribution of inland (N= 201) and coastal (N = 301) promontory forts in Scotland drawing on Atlas data (N = 502).
2
The Atlas of Hillforts of Britain and Ireland
Defining the Dataset remains a nagging doubt, which led Alcock to consider
the neighbouring and unlikely promontory of Bowduns
The criteria against which all the entries in the Atlas as a possible alternative. And yet the impregnable
have been judged rest on three key characteristics, character of the conglomerate promontory beneath
namely physical advantages of the topographical Dunnottar Castle, rivalled only on this stretch of coast
position, scale of the enclosing works, and size of the by the tiny sea stack, presumably once a more extensive
interior, for which a lower limit of 0.2ha was set (Lock promontory, at nearby Dunnicaer [3111], renders it
and Ralston 2020). Applied rigorously, however, it was almost inconceivable that it was not the location of the
foreseen that numerous well-known Scottish forts sieges referred to in entries in the Annals of Ulster for the
would be omitted, and the decision was taken to operate years AD 680 and 693 (Alcock 1981, 171-2; Alcock and
these qualifications to include enclosures that met at Alcock 1992, 267). More recently evidence of a timber-
least two out of the three criteria. This had far reaching laced defence dating from the 3rd-4th centuries AD has
consequences, leading to the inclusion of numerous been recovered from Dunnicaer (Past Horizons 2015),
small fortified enclosures which otherwise would not but it would be surprising if the Dunnottar promontory
have qualified. What was not foreseen, however, is that had escaped occupation in the pre-Roman Iron Age;
these same criteria also embrace the morphological this, after all, is the premier location for a promontory
character of a series of other promontory enclosures fort anywhere along this coast. The inclusion of several
that have never been conventionally considered as other castles in the Unconfirmed category is more
forts, be that prehistoric or early medieval. To some speculative, such as at Borve Castle, Sutherland [2809],
extent this may be a reflection of the extensive use of and the Castle of Old Wick, Caithness [2827], where the
Scottish promontories for various purposes at different character of both the promontories and the defences
times, though shades of the same issues are clearly re-creates the form of other promontory sites identified
present in Ireland, but we felt it was better to err on the as forts along the coastline of northern Scotland (cf
side of caution by including this wider range of ancient Lamb 1980, 90-2). Another probable castle included
promontory enclosures, rather than omitting them and under this morphological pretext is the earthwork
denying the choice of further exploring these sites to overlooking the bridge over the River Esk at Gilknockie,
the user of the Atlas data. Dumfriesshire [1122], and, for completeness, what
is probably a late medieval stronghold known as
The overall total of 502 entries thus represents all the Lady Lindsay’s Castle, Perthshire [3044]. Castle Qua,
promontory works in Scotland that appear to be of Lanarkshire [1568], had already been included in the
some antiquity, the principal exceptions being most Atlas database when fieldwork recognised traces of the
medieval stone castles exploiting promontory positions. abutments for a bridge crossing the ditch; its status was
The emphasis here is placed on the word ‘most’, altered to Irreconcilable Issues.
because several castles clearly occupy earlier Iron
Age promontory forts, such as Cullykhan, Banffshire This theme of medieval works occupying promontories
[2982], and Dundarg Castle, Aberdeenshire [2983]. In that may be the sites of earlier promontory enclosures
other cases, notably at Greenan Castle, Ayrshire [1289], also runs through the Irish data in the Atlas, for
and Innerwick Castle, East Lothian [3928], cropmarks example at Ballyoughteragh South, Kerry [1259], or
have revealed ditch-systems that appear eccentric to Downmacpatrick, Cork [0901], this latter where a 15th-
the visible medieval defences and are thus considered century tower and curtain wall cut off a promontory
more likely to relate to a different and earlier phase of identified as an earlier tribal stronghold. Reviewing
activity. This contrasts with several castles where early the descriptions of all these Irish sites against satellite
medieval texts indicate the likely site of a promontory imagery, however, also suggests that the ‘promontory
fort, but no evidence of any earlier defences is visible fort’ classification has been employed generously in
and none has been recovered by small-scale evaluation; both the Archaeological Survey of Ireland Database of the
Dunnottar Castle [3112] on the Kincardineshire coast, National Monuments Service and the Northern Ireland
examined by the Alcocks (1992), is the prime example, Sites and Monuments Record to include a number of
but others might include the rocky outcrop of Dunaverty, sites that are more likely to be post-medieval stock or
Argyll [4309], according to the Annals of Iona besieged agricultural boundaries. A 17ha enclosure at Greenane
in AD 712 (Alcock 1981, 157), and the cliff-girt Castle on Bear Island, Cork [0851], serves as one example,
Rock of Edinburgh Castle [3713], where again a siege while the 58ha enclosure on Innishark, Galway [1104],
is noted for the year AD 638. Excavation on Edinburgh representing about a quarter of the surface area of
Rock has at least revealed evidence of Late Bronze Age this small island, is another. These Irish examples
and early medieval occupation, though any trace of are marked Unconfirmed in the Atlas database, but
earlier defences is entirely lost beneath the castle walls. in Scotland a conscious decision was taken to try to
On balance, Edinburgh Castle has been accepted as a exclude the majority of such works, which are found
Confirmed fort, but in the case of Dunnottar Castle there widely on coastal headlands large and small, notably in
3
Stratford Halliday and Ian Ralston
Shetland and the Outer Hebrides. In some cases they identified as a possible promontory fort by Raymond
have also previously been mistakenly identified as Lamb (1980, 83) but appears to have a row of three
fortifications and have been included in the Atlas only rectangular structures midway along its top, might just
to avoid confusion. Weinnia Ness, Shetland [4176], for as easily be monastic and has been marked Unconfirmed.
example, was first reported as a fort by Raymond Lamb The Landberg promontory fort on Fair Isle [2861] raises
but later discounted from his synthesis (1980), and yet many of these issues, for the rectangular building visible
it still appears in Canmore – the online National Record within it (Hunter 1996, 89-93) was demonstrated by
of the Historic Environment of what is now Historic excavation to be a chapel and there was evidence of
Environment Scotland – as a fort. Another ‘fort’ an earlier phase of occupation beneath it from which
recorded at Lambigart, Shetland [4172], was included pottery comparable to that from brochs was recovered,
on a recent distribution map (Halliday and Ralston along with moulds for copper alloy artefacts. An
2009, 466, fig 5), but a field visit revealed that the thick evaluation carried out on Brei Holm, Papa Stour [4197],
banks cutting off the promontory are no more than recovered evidence of a complex sequence of occupation
old dykes built of turf stripped from the neck. Likewise and two radiocarbon assays returned dates in the 5th-
visits to a series of large promontory enclosures noted 7th centuries AD, rather earlier than the date suggested
by coastal surveys in the Outer Hebrides showed that by the rectangular buildings on its summit.
the majority were almost certainly agricultural, in
some cases with a 19th-century drystone dyke roughly This problem of disentangling secular and monastic
replicating the line of an earlier bank flanked by turf- enclosures is by no means unique to Shetland. The
stripping scars (Halliday and Ralston 2013, 224-5). In Brough of Deerness on Mainland, Orkney [2840], for
most cases, the relationship of these boundary banks example, with its chapel and rectangular and bow-
to the topography at the neck of a promontory is quite sided structures scattered across its interior, has long
unlike that of defensive works, which almost invariably been regarded as a monastic site, but the supposed
adopt the cliffs and scarps on the seaward or outer vallum monasterii is constructed as a major stone-
side of the neck to enhance the barrier. This sort of faced rampart defending the now collapsed neck of
promontory enclosure is not entirely restricted to the the promontory, with its outer face set on the steep
Atlantic coast, and those at Strath Howe, Aberdeenshire landward slope. Its status as a fort has therefore been
[2984], and Elliot Water, Angus [3080], are possibly no marked as Confirmed. The Brough of Deerness is a
more than post-medieval field boundaries. relatively large enclosure of almost 1ha, but at the
opposite end of the scale there are several smaller
Another group of sites that are more conventionally and heavily eroded fortified promontories on Orkney
considered to be early medieval is a series of what are with traces of rectangular structures on them, such
thought to be undocumented monastic sites, both large as The Brough, also on Mainland [2841], and Castle
and small, mainly situated on coastal promontories and of Burwick on South Ronaldsay [2813]. The offshore
isolated sea stacks. In Shetland these include relatively stack known as The Brough of Burgh Head on Stronsay
large enclosures, notably at Blue Mull, Unst [3723; 8.2ha], [2844] is another with a stout wall along its landward
but also including Outer [4195; 0.74ha] and Inner [4196, flank but where there may be some doubt about its
4.4ha] Brough, Fetlar, and Brei Holm, Papa Stour [4197; original function. The northern coast of Caithness
0.6ha], and the smaller stacks of Clett [4175], Burri and Sutherland includes further examples, on the one
Stacks, Culswick [4177], Kame of Isbister [4182], Aastack, hand with the enclosure of 2.25ha on St John’s Point,
Yell [4198], and Birrier of West Sandwick, Yell [4189]. Caithness [2833], traditionally associated with the
The presence of small clustered rectangular structures remains of a burial-ground and a chapel dedicated to
on many such sites, best illustrated by Kame of Isbister St John, and on the other two minor promontories
and Birrier of West Sandwick, forms the basis for the characterised by spectacular cliffs and the narrowest of
monastic interpretation, but it is difficult to demonstrate razor-backed necks connecting them to the mainland:
conclusively that any of them is not an early medieval the first, Aodann Mhor, Sutherland [2782], is crowded
secular settlement, nor, if they are indeed monastic, that with small rectangular structures, and less-certain
their origins did not lie in secular defended enclosures. traces of similar remains are also visible on the second,
All other forts in the Northern Isles exploit precipitous An Tornaidh Bhuidhe, Sutherland [2790]. Though
promontories and stacks, so it is not unreasonable unusual in this part of Scotland, St John’s Point meets
to suggest that some of these enclosures that have all the criteria for a major fortification with a massive
been claimed as monastic were initially secular. The rampart and ditch, so much so that it is impossible to
inaccessible Birrier of West Sandwick, for example, has a deny its Confirmed status. The other two are much
wall overlooking the razorback neck and though usually more problematic, and though An Tornaidh Bhuidhe
identified as monastic has been marked Confirmed in has a bank facing onto the only access and has been
the database. Conversely the promontory enclosure on accepted also as Confirmed, in truth there is no way of
Burrier Head [4174], also on Shetland, which was first knowing whether either is secular or monastic.
4
The Atlas of Hillforts of Britain and Ireland
Aspects of the same problem re-surface in eastern and prehistoric fort at least for the present configuration
south-eastern Scotland, though these examples have of the site. On Geirum Mor [2484], a cliff-bound islet
all been accepted as forts, such as the Kirk Hill on St of 1ha in the sound between Mingulay and Berneray
Abb’s Head, Berwickshire [4150]. Whether or not this at the extreme southern end of the chain of islands,
is the site of the documented monastery of St Aebbe, a wall blocks the only access at the north-east end
the place-name Colodaesburg implies a fortified and there is a series of rectangular buildings on its
place and Alcock’s excavations uncovered a complex summit. Morphologically this is a fort, but whether
early medieval rampart sequence (Alcock et al. 1986). prehistoric, early medieval or post-medieval cannot be
Elsewhere on this coast at Auldhame, East Lothian determined without excavation, and it could yet prove
[3921], excavations on a promontory enclosure first to be monastic. Biruaslum [2483], a tidal islet off the
identified from cropmarks have uncovered a chapel and west coast of Vatersay, which extends to 9.8ha – the
a long cist cemetery. This has been interpreted as the second largest such site in the Outer Hebrides – poses
remains of another Anglo-Saxon monastery, but there a similar concern. It is defended by a thick wall facing
is nothing inherently in the character of the ditch, the main island; its position recalls the monastic site
which was not bottomed during the recent excavations on the Brough of Birsay in Orkney, though at 18ha the
and is not closely dated, to demonstrate that its original latter island is rather larger and there is no evidence of
construction was as a monastic enclosure (Crone et al. a perimeter wall overlooking the tidal isthmus linking
2016, 129). Nor is it alone in the association of long cist it to the mainland.
burials with a promontory enclosure. In 1831 a long cist
cemetery was discovered not far away at Castle Dykes, Another problem found along the Atlantic coasts relates
Dunglass Dean, Berwickshire [0486], and also occupies to the interpretation of the outworks of brochs and
the site of a multivallate promontory fort, while duns, or Atlantic Roundhouses as they are commonly
other cists were said to have been found before 1853 called, themselves excluded from consideration in the
on the neighbouring promontory fort at Castle Dykes, Atlas database on size grounds. The very term outwork
Bilsdean, East Lothian [0487]. Yet another promontory assigns primacy to the Atlantic Roundhouse, but in
enclosure where cists occur is at Whiting Ness, Angus some cases the supposed outworks may have been the
[3100], which was traditionally the site of a burial- primary fortifications, with the Atlantic Roundhouse
ground and a chapel dedicated to St Ninian and marked subsequently set within them. This is a sequence
as such by the Ordnance Survey. familiar in southern Scotland at Edinshall [4069] and
Torwoodlee [3542], and the excavator of the broch at
Another group of Unconfirmed entries in the Atlas Crosskirk, Caithness [4348], Horace Fairhurst, postulated
database relating to promontories and sea stacks worth that the broch there succeeded an earlier promontory
drawing attention to is in the Outer Hebrides. In their fort (1984). No stratigraphic evidence was advanced
case, the scope for confusion is not related to the use in support of his case, which largely rests on two
of such locations by early monastic communities, but radiocarbon dates obtained in the 1970s that purport to
rather as post-medieval strongholds. The small and predate the assumed chronology of the broch supplied
now inaccessible Stac Dhomnuill Chaim at Mangursta, by sherds of Samian ware and late Roman pottery
Lewis [2759], is thought to be the remains of a refuge (see discussion by MacKie 2007, 407-26). More recent
constructed in the early 17th century by the Uig excavations at Nybster, Caithness [2820], were unable
warrior, Donald Cam Macaulay, while Dun Eistean to establish the stratigraphic relationship between the
[2772] on the east coast of Lewis is traditionally a wall across the neck of the promontory and the broch,
stronghold associated with the Morrisons since the and radiocarbon dates for samples underlying the wall
16th century (Burgess 2008, 60-2). The excavations indicate a phase of Late Bronze Age activity rather
by the Barrowmans at Dun Eistean found no evidence than the date of the outwork itself. Four out of nine
of occupation before the medieval period, but the Unconfirmed promontory works in Argyll fall into this
possibility that this is also the site of an earlier fort category,4 while for Sutherland and Caithness the figure
cannot be discounted. is 6 out of 16,5 in Orkney 6 out of 10,6 and in Shetland 3
out of 17.7 This uncertainty of the relationship between
All the same issues arise at Dun Eorradail [2773], another Atlantic Roundhouses and their supposed outworks is
large and inaccessible stack north of Dun Eistean, where
there are traces of at least ten rectangular buildings
and the possibilities for its use range from fort to early
monastic community or post-medieval stronghold.
4
Dun Bhronaig [2444], Dun Haunn [2503], Dun Aorain [2546] and Dun
Chruban [2550]
On reflection, the same could be said of several other 5
An Dun, Clachtoll [2793], Altanduin [2806], Poll Gorm [2810],
examples that have been accepted as Confirmed forts. Scarfskerry [2816], Nybster [2820] and Crosskirk [4348]
Rudha Shilldinish near Stornoway on Lewis [2765], for
6
Weems Castle [2811], Yesnaby, Broch of Borwick [2845], Midhowe
[2846], Riggin of Kami [2847], Lamb Head [2848] and Broch of Burrian
example, carries a suite of large rectangular buildings [2849].
which suggest a medieval or later date rather than a 7
[Broch of Aithsetter [4187], Noss Sound [4190] and Sna Broch [4260].
5
Stratford Halliday and Ian Ralston
not limited to those on coastal promontories and recurs in the top of the headland. The character of several
amongst those in other locations inland. others could be resolved likewise, probably including
four relegated to the status of Irreconciled Issues.
The rest of the sites that make up the Unconfirmed Three of these, Court Hill, High Skeog, Wigtownshire
promontory enclosures break down into several types [0227], and An Fang, Craignish Point [2449] and Creag
of record. Twelve are long-recorded sites that have a’ Chaisteal, Stillaig [2472], Argyll, are where the
been so heavily degraded that there is insufficient observations recorded by the archaeologists who first
information to judge their true character with any visited the sites have been disputed by subsequent
confidence.8 Twelve others turn on the interpretation OS surveyors, while the fourth is a site reported in
of cropmarks, either because a ditch is relatively 1993 on the shores of Loch an Iasgaich, Skye [2743],
narrow for a defensive work or the definition of the which appears an unlikely candidate to be a fort on
cropmarks is too diffuse;9 and finally some are simply the grounds of either its topographical position or the
miscellaneous earthworks on promontories, most of slightness of the supposed defences. In other cases, the
them of uncertain date or purpose. Most spectacular only resolution is by invasive evaluation, though as the
of these is the multivallate earthwork isolating 54ha on experience of the Mull of Galloway proved, there is no
the Mull of Galloway [0201], which despite excavation guarantee of success. Nevertheless, clarifying the date
(Strachan 2000) remains undated; it is either the of the Mull of Galloway is evidently vitally important
largest fort in any setting in the whole of Scotland or for the interpretation of the Iron Age landscape of
an extraordinary enclosure with some other function south-western Scotland, and the same might be said
that finds its only morphological parallels with locally of Dun Evan in Morayshire if this proved to be a
several equally large promontory works in southern fortification of Iron Age date.
Ireland, such as the 83.5ha enclosure on the headland
at Ballynacarriga, Cork [1970]. Others falling into this Confirmed Promontory Forts
category in southern Scotland are inconsequential by
comparison. Haly Jo, Lumsdaine, Berwickshire [4092], for This rehearsal of the range of promontory works
example, a slight enclosure on a coastal cliff is probably found in Scotland and the problems of applying
no more than a small settlement, though others, such morphological classifications to identify those that
as Drummoral, Wigtownshire [0223], appear defensive, are at least potentially prehistoric and early medieval
here with rock-cut ditches but no evidence of an fortifications serves to clarify the definitions that
accompanying rampart. Elsewhere there are: the tiny lie behind those that are attributed the status of
stone-walled enclosure (claimed as unfinished: MacKie Confirmed forts. In short, they comprise fortified
2016) overlying the broch at Leckie, Stirlingshire [1471]; enclosures where thick walls or earthen ramparts and
the enclosure beneath a later medieval burial-ground ditches bar access on one side, usually the easiest line
on Innis Bhuidhe, an island set in the river at Killin, of approach, and the rest of the perimeter is apparently
Perthshire [2609]; the slight boundary, undated despite defined by little more than cliff-edges or steep scarps.
limited excavation, enclosing two rectangular buildings In this definition the character of most of these forts
on a low promontory projecting into Loch Kinord, is synonymous with the description of the topography
Aberdeenshire [3075]; the precipitous promontory on which they stand. Originating from its Latin root
known as Tronach Castle, Banffshire [2944], where to describe a raised headland jutting out into the sea,
there are no visible defences; and Dun Earn, Morayshire it has been adapted more generally to describe other
[2918], where a ditch 4m-5m in breadth but with little projecting landforms and raised spits of ground, and
trace of a rampart cuts off about 2.5ha on an inland in archaeological terminology further extended to
promontory. embrace inland enclosures set on interfluves and thus
often exploiting angles formed in escarpments along
In some cases a visit would solve issues raised by the streams and rivers, usually where a tributary has cut
existing records, as for example at Lambigart discussed down at its confluence with the main flow. They are
above, or Hynish, Tiree, Argyll [2486], where fieldwork thus largely defined by natural declivities on at least
(by SH) since the completion of the database clearly two sides, often creating a roughly triangular plan in
demonstrates that this is not a promontory fort as such, which the artificial defences form the third side.
though there are traces of a fortified enclosure taking
This basic format, however, has also led to the
8
Grennan, Grennan Point [0180], Killantrae Bridge [0217], Port o’ term being applied to the plans of forts that are in
Warren [0311], Gunnerton Crag Camps [0520], Ebb’s Nook [0920], hilltop positions, or on the ends of ridges, where
Salter’s Nick [1977], The Heugh [2038], Machrihanish [2222], Keir,
Easter Tarr [2617], Firbush Point [2619], Coldstream [4079] and Siccar the defences were apparently constructed only on
Point [4115] one side, complemented by abrupt or at least steep
9
Clanyard Bay [0196], Leffnoll [0342], Loch Quien [1152], Rousland descents elsewhere. Scottish examples of these tend
[1838], Wester Tullynedie [3046], West Lindsaylands [3230], Milton
Mill [3782], Bara [3859], Nether Hailes [3883], Lumsdaine Dean [4098], to be located in prominent elevated positions, such as
Coveyheugh [4101] and Ayton [4142]. Dumglow, Kinross [3203], Ben Effrey, Perthshire [2648],
6
The Atlas of Hillforts of Britain and Ireland
Craik Moor, Roxburghshire [3453], An Sgurr on the in the Inner Hebrides. This is perhaps a reflection of
island of Eigg [2527] or Sithean Buidhe, Argyll [2292]. the longer hard coastlines in relation to the surface
All told, there are only thirteen in this type of setting areas of the islands and probably aspects of their
in Scotland and they are better analysed with other geology too. Islay (about 600 sq km and 165 km of
hilltop forts.10 Much more problematical in this sort of coast), for example, has 38 forts all told, of which no
morphological classification driven by topographical fewer than 26 are promontory works (68%), and 24
descriptors is the exclusion of scarp-edge forts where of them coastal, whereas Mull (about 880 sq km and
the circuit was evidently left incomplete along one side with 250 km of coast) has equivalent figures of 22, 10
and the interior simply backs onto the lip of a cliff or (45%) and 10. Skye (1650 sq km and 650 km of coast),
escarpment, and where the distinction between these the northernmost of the Inner Hebrides and formerly
and some promontory forts is one of degree. This part of Inverness-shire, has figures of 31, 15 (48%) and
issue is discussed further below under the heading of 12 respectively. It is worth noting in this context that
Scarp-edge Forts. At least 97 other Confirmed forts the 9 Confirmed forts on Orkney, and 15 on Shetland,
share this feature, eight of them in coastal locations, are all promontory forts and are all coastal.
and their entries in the Atlas database are variously
labelled Contour Fort (40), Hillslope Fort (12), and Other regions show similar patterns of variation within
Level Terrain Fort (45). While these terms serve as them. Thus, while promontory forts make up 34 % of all
topographic descriptors, it is unwise to apply any of forts in the South-West, in Wigtownshire, again with its
them too prescriptively in terms of their archaeological relatively long coastline, the figure rises to 65%. The lowest
significance. regional percentage is in the Greater Tyne-Forth region,
where it is no more than 11%, and here exceptionally 87
Whereas Lamb’s schematic map suggested that of the 98 promontory forts occur inland. In the hillier
Scottish promontory forts formed discrete inland landscapes of Peeblesshire, where forts typically
concentrations in Shetland, Orkney, Caithness, Angus stand on spurs along the sides of valleys, the figure is
and Galloway (1980, 5, fig 1), the Atlas data reveal only 1%, representing a single fort at Castlecraig [3636].
that they are much more widespread and occur In Roxburghshire and neighbouring Northumberland
along virtually every coastline where there is some only 7% of Confirmed forts are on promontories, but
form of normally rocky escarpment or cliff-line. for Berwickshire, with its long predominantly rocky
Furthermore, this same defensive format is equally coastline, it rises to 25%. Without the 8 coastal examples
widespread in inland locations. Nevertheless, the known there, however, the percentage would be no more
new map hides some general trends. Reference to than 18%, a figure more akin to the 17% in Dumfriesshire,
the regional and county table (Fig. 1) shows that the or 15% in Lanarkshire.
proportions of promontory forts to other forts alter
from region to region. The largest single regional In passing, it is worth taking brief note of the
group of promontory forts is in the western and distribution of Unconfirmed promontory works, which
northern Highlands, which includes the whole of the are also recorded in Fig. 1. At a regional scale, these
mainland Atlantic coast from Argyll northwards, and form between 17% of all promontory works in Greater
the Inner Hebrides and Outer Hebrides. Across this Tyne-Forth, but only 2% in comparison to the total
region as a whole 159 of the 370 Confirmed forts are number of forts in the region. The equivalent figures
promontory forts, representing 43%, and 32 of them for the South-West are 13% and 5% respectively, in
occur inland. The pattern within this region, however, Central Scotland 17% and 4%, and in the North-East
varies widely. In the three northernmost counties of 15% and 7%. In the northern and western Highlands,
Ross-shire, Sutherland and Caithness, where there these figures appear to rise overall to 20% and 10%
are relatively few forts, the percentages rise to 59%, respectively, but this masks wide regional differences
64% and 69% respectively. Locally within them, some between Argyll at 14% and 6%, and Caithness at 53%
of the figures are even higher. In the Outer Hebrides, and 76%. The Northern Isles calculations are even more
for example, which were formerly split between Ross- extreme, with 54% and 116%. A significant proportion
shire and Inverness-shire, of the 26 Confirmed forts, of these are the supposed outworks of brochs (supra). In
all bar 3 are coastal promontories or tidal islands themselves, these figures are of little significance, but
(88%); the exceptions are all islands in inland lochs. In they confirm the general pattern that the proportions
Argyll, too, where the overall number of promontory of promontory forts to other forts not only increases
forts forms 38% of all Confirmed forts there, no fewer on the islands of the Atlantic seaboard, but also more
than 50 of the 68 promontory forts occur on islands generally northwards through the mainland, and
that whereas the alteration of the status of a few of
10
The full list comprises Dumyat [1593], Sithean Buidhe [2292], An the Unconfirmed category in the south makes little
Sgurr [2527], Skirley Craig [2633], Ben Effrey [2648], Dun Vallerain difference to the overall proportions of promontory
[2715], Phoineas Hill [2887], Dumglow [3203], Little Trowpenny [3378],
Craik Moor [3453], Earn’s Heugh NW [4094], St Abb’s Head [4150] and forts to other types, in the far north it exaggerates the
An Dun, Cornhill Wood [4381]. contrast already observed still further.
7
Stratford Halliday and Ian Ralston
Figure 1. Table of promontory enclosures by region and county. Percentages of Confirmed Promontory Forts are calculated as a ratio of All
Confirmed Forts regionally and locally. Sizes column totals include multiple measurements from individual sites; for six others there is no
data.
Region % Conf
All Prom Unconf Irrec Conf All Conf
Prom Coastal Inland Sizes
Historic County Encls Proms Issues Proms Forts
Forts
Greater Tyne-Forth 100 12 1 87 667 13% 11 76 103
Berwickshire 36 7 - 29 117 25% 8 21 35
Dumfriesshire 22 1 - 21 116 18% - 21 24
East Lothian 17 3 - 14 81 17% 3 11 19
Lanarkshire 10 1 1 8 54 15% - 8 8
Midlothian 4 - - 4 41 10% - 4 4
Peeblesshire 1 - - 1 91 1% - 1 1
Roxburghshire 10 - - 10 148 7% - 10 12
Selkirkshire - - - - 19 - - - -
SW Scotland 68 8 1 59 172 34% 49 10 61
Ayrshire 9 - - 9 32 28% 4 5 9
Kirkcudbrightshire 12 1 - 11 80 14% 8 3 11
Wigtownshire 47 7 1 39 60 65% 37 2 41
Central 42 7 - 35 164 21% 5 30 42
Clackmannanshire 1 - - 1 1 100% - 1 1
Dunbartonshire - - - - 4 - - - -
Fife 10 - - 10 39 26% 4 6 13
Kinross-shire 1 - - 1 3 33% - 1 1
Perthshire 21 5 - 16 75 21% - 16 20
Renfrewshire - - - - 8 - - - -
Stirlingshire 7 1 - 6 25 24% - 6 6
West Lothian 2 1 - 1 9 11% 1 - 1
NE Scotland 40 6 - 34 83 41% 20 14 42
Aberdeenshire 7 2 - 5 23 22% 3 2 5
Angus 15 1 - 14 36 39% 8 6 15
Banffshire 7 1 - 6 8 75% 5 1 9
Kincardineshire 8 1 - 7 9 78% 3 4 7
Morayshire 3 1 - 2 7 29% 1 1 6
N & W Highlands 200 37 3 160 371 43% 127 33 159
Argyllshire 79 9 2 68 179 38% 58 10 69
Buteshire (Arran & Bute) 11 2 - 9 19 47% 3 6 9
Caithness 19 10 - 9 13 69% 8 1 9
Inverness-shire 50 7 1 42 102 41% 31 11 40
Ross-shire 24 3 - 21 35 60% 18 3 22
Nairnshire - - - - 6 - - - -
Sutherland 17 6 - 11 17 65% 9 2 10
Northern Isles 52 27 1 24 24 100% 24 - 23
Orkney 19 10 - 9 9 100% 9 - 10
Shetland 33 17 1 15 15 100% 15 - 13
Totals 502 97 6 399 1481 27% 236 163 430
8
The Atlas of Hillforts of Britain and Ireland
The 399 that are Confirmed include 12 coastal and 17 and 2ha, and 13 are over 2ha, the largest being two
inland forts with eccentric or wide-spaced lines of coastal forts on Vatersay and Mingulay in the Outer
ramparts, taken here to represent separate structural Hebrides enclosing 9.8ha and 10.4ha respectively [2483,
arrangements probably belonging to successive phases 2480].
of construction. In another six cases no size data are
available. This provides a total of 430 measured internal By breaking this graph (Fig. 2) down into the two
sizes, which are plotted on a single graph (Fig. 2) in datasets, however, different patterns can be detected
increments of 0.1ha up to 2ha, and 2ha increments between those in coastal and inland positions. The
thereafter, the latter in orange. This shows that almost profile of the coastal graph of 247 sizes from 236 forts
half (48%) fall below 0.2ha. The profile of the graph falls (Fig. 3) falls away much more steeply to about 0.4ha,
away steeply from 119 (28%) examples below 0.1ha to and there are almost twice as many sizes below 0.1ha as
about 0.7ha, sites below this threshold representing there are in the 0.1ha-0.2ha increment. A total of 185 fall
87% of all promontory forts. In all, 24 lie between 1ha in the four increment classes below 0.4ha, representing
Figure 2. The Confirmed promontory forts of Scotland by enclosed area. Steps are 0.1 ha to 2ha, thereafter by 2ha divisions. N=430.
9
Stratford Halliday and Ian Ralston
Figure 3. The Confirmed coastal promontory forts of Scotland by enclosed area. Steps are 0.1 ha to 2ha, thereafter by 2ha divisions.
N=247 from 236 sites.
Figure 4. The Confirmed inland promontory forts of Scotland by enclosed area. Steps are 0.1 ha to 2ha, thereafter by 2ha divisions.
N=183 from 163 sites.
10
Another random document with
no related content on Scribd:
– Nos és mi az? – kérdezte Eriksen – halljuk! Én mint férj
jogosan érdeklődhetem.
– Hogy lehet ilyet még kérdezni is, hát nem elég biztositék, a
nagyságos asszony meglepően gyönyörü megjelenése.
– Ezt már igazán nem tűrném – kiabált a szőke nagy énekesnő, –
hogy így udvaroljanak jelenlétemben a feleségemnek.
– Én nyugodt vagyok a Lolómat illetőleg – ez volt a háziúr egyik
legkedvesebb, sokat hangoztatott kijelentése.
Lolo megsimogatta az ura rövid ujjú, kövér, fehér kezét. – Igazad
van édes piczikém – hizelgett kedvesen.
– Isten tartsa meg a nyugalmát – mondta gúnyosan a szőke.
Oszkár ránézett az asszonyra, aztán az urára, aztán a szép
Mencsikofra és saját tapasztalatai jutottak eszébe. Vajjon ez a kis
ember csakugyan bízik a feleségében, vagy neki csupán
dísztárgynak kell, a ki bármi áron, de biztosítsa neki, az ambicziózus
kis zsidónak, hogy művészek, írók, szóval nevezetességek
forduljanak meg a házánál?
Nagyon furcsa volt mindez az Oszkár szemében és egyszerre az
asszonynyal eltöltött tavaszi idilljét is egészen más színben látta.
A háziúr felemelte a poharát. – Engedjék meg, hogy ezt az
aranyszínű rajnai bort a Mary barátnőnk egészségére üritsem.
Koczintott a leánnyal és hogy jobban megnézte őt, rajta is
felejtette a szemét. Nagyon szépnek találta a sötétzöld ruhában, a
mely kicsit halaványnyá tette, de annál finomabban tüntette fel a
vonásait.
– Milyen gyönyörű virágai vannak – jegyezte meg aztán, mintha
azokban gyönyörködött volna.
Marynak egy csomó lila-rózsaszin orchidea volt az övébe tűzbe.
– Ilyet még csak kirakatban láttam, – mondta az amerikai
asszony, – soha sem volt még olyan heves bámulóm, a ki ezekkel a
drága virágokkal ajándékozott volna meg.
– Valld be Mary, kitől kaptad – és Lolo jelentőségteljesen nézett
Oszkárra.
– Magam vettem – felelte Mary egyszerüen.
– Mutasson egyet – kérte Mencsikof.
Mary kihúzott egy szálat a csokorból és átnyújtotta Mencsikofnak.
A művész egy ideig a lámpa felé emelve nézte, a mint a villany
fénye átvilágitott a finoman erezett szirmokon.
– Ez a legszebb virág – mondta – titokzatos, csudálatos, szelid
és buja egyszerre. Olyan mintha egy szőke, hideg leány valahol az
indiai templomokban, a tarka pompa között, a hol fényes szemű,
sötét bőrű, ékkövekkel telerakott nők csábítják a férfiakat, vad, keleti
tánczba kezdene a forró éjszakában, disztelenül, karcsún, fehéren,
egy nagy, kétségbeejtő érzelem hatása alatt.
Sápadt arcza kipirul, tiszta bőre ég, szőke haja leomlik a
tánczközben és szemérmesen eltakart válláról lecsúszik szűzi
ruhája.
Nem látják mindezt a rózsaszínű, halaványeres, különösre kinyilt
szirmokon? Milyen üde most még és milyen hirtelenül fog
elhervadni, a mint az én szőke nőmnek is hirtelen kell meghalnia a
táncz után, megöli a szégyen, a mint abbahagyja és tudatára ébred
mezitelen vállainak, égő arczának. Szegény, szép fehér leány
mereven, hidegen alszik az ezer csillagos, indiai éjben. Ezt a virágot
szeretném elmuzsikálni és azt a lányt. Majd egyszer megcsinálom.
– Kegyedhez nagyon jól illenek az orchideák – tette hozzá
Maryhez fordulva.
– Egy dalt írjon róla és én majd eléneklem – mondta Lolo és
ajkához emelte a virágot.
– Oh kedves barátnőm, nem talál magához ez a furcsa virág,
sokkal egyszerűbb, illatosabb diszt és hangulatot kiván a maga
lénye.
Az asszony engedelmesen tette le az orchideát, a melyet a férje
vett a kezébe, egy ideig nézte, de minthogy semmi találó sem jutott
eszébe, hármat köhintett, visszaadta Marynak és a boráról kezdett
mesélni, hogy hol, mikor és mennyiért vette.
Mary hosszú, fehér kezében tartotta az orchideát. Mencsikof
áthajolt az asztalon és halkan mondta. – Tűzze a szép hajába.
– Sohasem tűzök virágot a hajamba – felelte Mary és elengedve
az orchideát, önkénytelen mozdulattal a kontyához nyúlt.
Oszkárt boszantotta a Mencsikof bizalmaskodása, felvette a
virágot és mikor a leány beleegyezőleg bólintott, a gomblyukába
tűzte.
VI.