Coins in Churches: Archaeology, Money and Religious Devotion in Medieval Northern Europe 9780367557065, 9780367557072, 9781003094814

This book focuses on the formative period of Church reform in the Middle Ages in Northern Europe, when the Church paved

222 62 49MB

English Pages [483] Year 2021

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Coins in Churches: Archaeology, Money and Religious Devotion in Medieval Northern Europe
 9780367557065, 9780367557072, 9781003094814

Table of contents :
Cover
Half Title
Series Page
Title Page
Copyright Page
Table of Contents
List of figures
List of tables
List of contributors
Preface
List of abbreviations
PART I: Money and religious devotion
1 Money and religious devotion in medieval Northern Europe
2 Pious gifts: Coins and church interior
PART II: Coin finds in Scandinavian churches
3 The archaeology of church floors
4 Digitizing the past: A rewarding challenge for the present. The digitizing process of analogue data from excavation in medieval churches
5 Moving money, ritual money – Studying monetary and ritual space in Bunge church on medieval Gotland
6 Coin finds of Høre stave church, Oppland Norway: Reflections of regulation and conflict in the Middle Ages
7 Eidskog church revisited: Coin finds, architecture and the devotional use of money
8 Coins and monastic liturgy in the Middle Ages: A study of St Mary’s Benedictine nunnery in Bergen, Norway
9 The archaeological landscape under church floors: Coins and contexts in Aggersborg church, Jutland, Denmark
10 Coins in church contexts: Hedensted Church, Jutland, Denmark
11 Building for glory: Coins in the houses of the Lord in Gränna and Arby, Sweden
12 Coin finds and coin use in the medieval round church at Klåstad, Östergötland, Sweden
13 Jomala church, Åland Islands: Coin offerings to the Virgin Mary and the long Reformation
PART III: Coin finds in churches in Central Alpine Europe
14 Inside the church: Coin finds and liturgical topography. A survey of the Central Alpine Europe evidence
Bibliography
Index

Citation preview

Coins in Churches

This book focuses on the formative period of Church reform in the Middle Ages in Northern Europe, when the Church paved the way for the development of money economy on its own doorstep. Church archaeology provides evidence for patterns of monetary use related to liturgy, church architecture and devotional culture through the centuries. This volume encompasses Alpine European evidence, with emphasis on Gotland, Denmark, Norway, Sweden, Finland and Switzerland, which opens up a new field of research on religion and money for an international audience. Based on 100,000 single finds of coins from the eleventh to eighteenth centuries from 650 Scandinavian churches, the volume offers an in-depth discussion of the concepts of ritual, liturgy and devotional uses of money, monetary space and spiritual economy within the framework of Christendom, the medieval church and church architecture. Written by international scholars, Coins in Churches will be a valuable resource for readers interested in the history of religion, money, the economy and church architecture in Northern Europe in the Middle Ages. Svein H. Gullbekk is Professor of numismatics and history of money at the Museum of Cultural History, Oslo University; and is Principal Investigator for the project “Religion and Money: Economy of Salvation in the Middle Ages”. Christoph Kilger is Senior Lecturer in Archaeology at Uppsala University, Campus Gotland. He specializes in Viking and medieval archaeology and numismatics, especially relating to social, economic and monetary history of Northern Europe. Steinar Kristensen is Senior Engineer at the Museum of Cultural History, University of Oslo, with responsibility for digital documentation of major excavation projects. He specializes in GIS technology in archaeological contexts. Håkon Roland is Associate Professor of numismatics and classical archaeology at the Museum of Cultural History, University of Oslo. He has a long record of publication on numismatics and archaeology and within Heritage studies.

Religion and Money in the Middle Ages

Edited by Svein H. Gullbekk, Museum of Cultural History, University of Oslo, Norway

Religion and Money in the Middle Ages explores the connections between two of the most dominant aspects of medieval society and culture: religion and money. Recognizing the importance of both multi- and single-disciplinary perspectives on the issues and questions connected to religion and money, the series accepts joint as well as individual authorship and editorship. All disciplinary perspectives are welcome, particularly from history, archaeology, anthropology and numismatics. The series operates with a broad chronological range: in western European terms from late Antiquity to the Reformation. While the geographical and cultural focus lies in western Christendom, the series will be open to cross-cultural comparative studies, and to treatments of money and religion in all religious communities within the period, within Christendom and without. Of especial interest are studies that explore issues on the theory and practice of money within religious contexts, and those that further reveal the interconnections and contrasts, overlaps and distinctions between these attitudes and practices are particularly encouraged. How differences between theory and practice emerge, how they are reconciled and how they remain unresolved are further questions the series is keen to explore. The range of source material available and the centrality of both subjects to medieval life, culture, belief and activity allow for breadth and depth of investigation and insight into the medieval past at its most intimate and in its largest institutions and social structures. Coins in Churches: Archaeology, Money and Religious Devotion in Medieval Northern Europe Edited by Svein H. Gullbekk, Christoph Kilger, Steinar Kristensen and Håkon Roland Money and the Church in Medieval Europe, 1000–1200: Practice, Morality and Thought Edited by Giles E. M. Gasper and Svein H. Gullbekk Divina Moneta: Coins in Religion and Ritual Edited by Nanouschka Myrberg Burström and Gitte Tarnow Ingvardson

Coins in Churches Archaeology, Money and Religious Devotion in Medieval Northern Europe

Edited by Svein H. Gullbekk, Christoph Kilger, Steinar Kristensen and Håkon Roland

First published 2021 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 605 Third Avenue, New York, NY 10158 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 selection and editorial matter, Svein H. Gullbekk, Christoph Kilger, Steinar Kristensen, and Håkon Roland; individual chapters, the contributors The right of Svein H. Gullbekk, Christoph Kilger, Steinar Kristensen and Håkon Roland to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Names: Gullbekk, Svein H., editor. | Kilger, Christoph, 1967– editor. | Kristensen, Steinar, editor. | Roland, Håkon, editor. Title: Coins in churches : archaeology, money and religious devotion in medieval Northern Europe / edited by Svein H. Gullbekk, Christoph Kilger, Steinar Kristensen and Håkon Roland. Other titles: Archaeology, money and religious devotion in medieval Northern Europe Description: London ; New York, NY : Routledge/Taylor & Francis Group, 2021. | Series: Religion and money in the Middle Ages | Includes bibliographical references and index. Subjects: LCSH: Coins, Medieval—Scandinavia. | Coins, Scandinavian. | Coins—Religious aspects—Christianity. | Coin hoards—Scandinavia. | Archaeology, Medieval—Scandinavia. | Scandinavia—Antiquities. | Excavations (Archaeology)—Scandinavia. | Church buildings— Scandinavia—History—To 1500. Classification: LCC CJ3094 .C65 2021 | DDC 737.49368—dc23 LC record available at https://lccn.loc.gov/2021003637 ISBN: 978-0-367-55706-5 (hbk) ISBN: 978-0-367-55707-2 (pbk) ISBN: 978-1-003-09481-4 (ebk) Typeset in Sabon by codeMantra

Contents

List of figures List of tables List of contributors Preface List of abbreviations

vii xix xxi xxiii xxv

PART I

Money and religious devotion

1

1 Money and religious devotion in medieval Northern Europe

3

S V E I N H . G U L L B E K K , C H R I S T O P H K I L G E R , H Å KO N RO L A N D A N D STEI NAR KR ISTENSEN

2 Pious gifts: Coins and church interior

29

M A RT I N WA N G S G A A R D J Ü RG E N S E N

PART II

Coin finds in Scandinavian churches

57

3 The archaeology of church floors

59

H E N R I E T T E R E N S B RO

4 Digitizing the past: A rewarding challenge for the present. The digitizing process of analogue data from excavations in medieval churches

65

STEI NAR KR ISTENSEN

5 Moving money, ritual money – Studying monetary and ritual space in Bunge church on medieval Gotland C H R ISTOPH K I LGER

76

vi Contents 6 Coin finds of Høre stave church, Oppland Norway: Reflections of regulation and conflict in the Middle Ages

143

J O N A N D E R S R I S VA AG

7 Eidskog church revisited: Coin finds, architecture and the devotional use of money

169

H Å KO N RO L A N D

8 Coins and monastic liturgy in the Middle Ages: A study of St Mary’s Benedictine nunnery in Bergen, Norway

195

A L F T O R E H O M M E DA L

9 The archaeological landscape under church floors: Coins and contexts in Aggersborg church, Jutland, Denmark

237

H E N R I E T T E R E N S B RO A N D J E N S C H R I S T I A N M O E S G A A R D

10 Coins in church contexts: Hedensted Church, Jutland, Denmark

269

G I T T E TA R N OW I N G VA R D S O N

11 Building for glory: Coins in the houses of the Lord in Gränna and Arby, Sweden

292

N A N O U S C H K A M . B U R S T RÖ M

12 Coin finds and coin use in the medieval round church at Klåstad, Östergötland, Sweden

332

H E N R I K K L AC K E N B E RG

13 Jomala church, Åland Islands: Coin offerings to the Virgin Mary and the long Reformation

351

E E VA J O N S S O N

PART III

Coin finds in churches in Central Alpine Europe

377

14 Inside the church: Coin finds and liturgical topography. A survey of the Central Alpine Europe evidence

379

B E N E D I K T Z ÄC H

Bibliography Index

417 447

Figures

1.1 Scandinavian churches and church sites mentioned in this book. Map: Steinar Kristensen, UiO: Museum of Cultural History 4 1.2 A selection of coins from Lom stave church, Oppland County, Norway. Photo: Museum of Cultural History, University of Oslo 6 1.3 Letter of indulgence dedicated to the altar of the Apostles in Våsterås ­cathedral, Sweden. Letter issued on 10th January, 1471 by Swedish archbishop Jakob ­Ulvsson and bishops of church provinces Västerås, Skara, Strängnäs, and V ­ äxsjö. Foto: Christer Åhlin, The Swedish History Museum, Stockholm, SHMM License, CC-BY 7 1.4 Pious laity at the offertory during Mass, offering money at the offering box, the altar, and giving alms. Regensburger Altar, c. 1480. Photo: Bayerisches National ­museum, München, MA 3304 9 1.5 From the excavation of Ringebu stave church in 1981/82. Museum of Cultural History, University of Oslo, unknown photographer 18 1.6 Scandinavian church sites investigated in the project. Map: Steinar Kristensen, UiO: Museum of Cultural History 22 2.1 The Three Magi painted in Fjenneslev church on the Island of Zealand. M ­ ural, c. 1125–1150. Photo: Martin Wangsgaard Jürgensen 30 2.2 The small village church of Tystrup on the island of Zealand. Photo: Martin Wangsgaard Jürgensen 36 2.3 a–e:  Reconstruction of the offering as it could have been performed in the village church of Tystrup on Zealand by the end of the fifteenth century. See text for explanation for the individual stages. Reconstruction by Martin Wangsgaard Jürgensen, based on a plan of the church by Danmarks Kirker 37 2.4 Collection plate from the church of Brændekilde on the island of Funen, D ­ enmark, c. 1500, with the image of an unidentified saintly bishop. Photo: Arnold Mikkelsen 38 2.5 Joachim’s Sacrifice in the temple. Note the coins on the floor by the altar. Etching, Israhel van Meckenem, c. 1490–1500 39 2.6 The Romanesque interior with the altars oriented towards the east. Plan: Martin Wangsgaard Jürgensen 41 2.7 The church of St Olof in Scania, southern Sweden. Photo: Martin Wangsgaard Jürgensen 43 2.8 The “Gothic” interior of St Olof’s church with altars facing in other directions than east. Plan: Martin Wangsgaard Jürgensen 43

viii Figures 2.9a Fight Between Carnival and Lent, oil on canvas. Pieter Brueghel the Younger, 1559. Kunsthistorisches Museum Vienna 47 2.9b Fight Between Carnival and Lent; close-up of beggars in front of church 48 2.9c Fight Between Carnival and Lent; collecting inside the church door 48 3.1 Reconstructed section from the chancel of Malt church. Two coins from the twelfth century were found in layer 3, the subconstruction of a floor surface of stone paving. It was not possible to determine whether the coins were lost, deliberately deposited in the church during construction work or if they were brought into the church in the possibly redeposited soil making up layer 3.  The coin from around 1,400 in layer 4 might also have been brought into the church as part of the possibly redeposited levelling layer (4), which most likely is part of a sub-construction underneath an unknown floor surface. Drawing by Knud Krogh, from the annual publication Fra Ribe Amt (1971), 537 60 4.1 Example of the vectorizing process showing objects on the digitized plan of Bunge church, Gotland, Sweden. Map by the author 68 4.2 Quantitative analysis of recovered coins presented by their find context as color-coded polygons. The color indicates the number of coins found. Bunge church, Gotland, Sweden. Map by the author 70 4.3 Quantitative analysis of recovered coins presented by their find context as center points. The varying size represents the amount of coins found. Bunge church, Gotland, Sweden. Map by the author 71 4.4 Quantitative analysis of the number of recovered coins presented by randomly placed points within their find context. Each point (coin) has unique information according to the finding and cataloguing data. Bunge church, Gotland, Sweden. Map by the author 72 4.5 The complex layout of the excavation of Ringebu stave church, Norway. Map by the author 75 5.1 Distribution map of medieval churches on Gotland. Churches with more than thousand coins: Gothem (4220 sp.); Bunge (4147 sp.); Burs (3103 sp.); St Olof’s chapel (c. 2000 sp.); Silte (1672 sp.); Fardhem (1648 sp.); Hamra (1364 sp.); Bro (1092 sp.); Grötlingbo (1057 sp.). Map: S. Kristensen, UiO: Museum of Cultural History 77 5.2 Weighing scene with Archangel Michael and devils, south-eastern wall of the nave in Mästerby church, Gotland. Photo: C. Kilger 79 5.3 Bunge church, northern Gotland. Photo: C. Kilger 80 5.4 Sketch of the medieval rectory at Bunge church, made by CG Hilfeling in 1800 81 5.5 Wooden sculpture of St Olof from Bunge church, dated to the second ­quarter of the fourteenth century. Photo: K. Nimmervoll, The Swedish History ­Museum, Stockholm 83 5.6 The central area in Bunge parish. Rectified fair copy of the GM 1700 map, Land survey board archive 84 5.7 Frieze on the northern wall of the nave in Bunge church, c. 1400.  Main scenes: A) Crucifixion, B) Gregory’s vision of Arma Christi, C) Gologotha. Photo: C. Kilger/S. Kristensen/Frida Roland 86 5.8 Gregorian mass and the Arma Christi scene. Photo: C. Kilger 87

Figures  ix 5.9 Lafrans’ offering box, second half of the thirteenth century. Photo: C. Kilger 88 5.10 Unornated back of Lafrans’ offering box with iron brackets. Photo: C. Kilger 89 5.11 The excavations of Bunge church in 1971/72 carried out by RAGU 91 5.12 Map showing architecture of standing church in Bunge, extant interiors and the extension of the archaeological investigation in the nave and the tower chamber. Map: Steinar Kristensen, Museum of Cultural History, UiO 92 5.13 The foundations of the earlier Romanesque church in Bunge in 1971/72. Medieval grave slab or platform visible in the northeastern corner of the Gothic nave 93 5.14 Map showing reconstruction of the foundation of the Romanesque church. Map: Steinar Kristensen, Museum of Cultural History, UiO 94 5.15 Medieval burials in the tower chamber 95 5.16 Supporting slab of the wooden floor in the nave 96 5.17 Rectangular stone packing or platform, northern wall of the nave 97 5.18 Coin vault in Lafrans’ offering box. Photo: C. Kilger 104 5.19 Map showing the distribution of all numismatically classified coins. Main scenes paintings northern wall: (A) Crucifixion, (B) Gregory’s vision of Arma Christi, (C) Golgotha. Map: Steinar Kristensen, Museum of Cultural History, UiO 105 5.20 Map showing the distribution of coin fragments. Main scenes paintings northern wall: (A) Crucifixion, (B) Gregory’s vision of Arma Christi, (C) Golgotha. Map: Steinar Kristensen, Museum of Cultural History, UiO 110 5.21 Map showing the distribution of human skulls in relation to the placement of the modern pews. Map: Steinar Kristensen, Museum of Cultural History, UiO 113 5.22 Brass pins from Bunge church, Type A-C. Photo: J. Svensson 114 5.23 Map showing the distribution of brass pins. Map: Steinar Kristensen, M ­ useum of Cultural History, UiO 115 5.24 Beads from square 1, north-western corner of the nave, next to the ­northern portal of the Gothic church, Gotland museum, Bunge inv. 2–14. Photo: C. Kilger 117 5.25 Map showing the distribution of beads in Bunge church. Map: Steinar ­K ristensen, Museum of Cultural History, UiO 118 5.26 Early Gotlandic penny (c. 1150–1220), Myrberg I, Bunge church. Photo: G. Hildebrand, Royal Coin Cabinet, Stockholm 122 5.27 Map showing distribution of early Gotlandic pennies in the oldest stone church. Map: Steinar Kristensen, Museum of Cultural History, UiO 123 5.28 Hatched cross penny, Gotland (c. 1220–1290), Myrberg II, Bunge church. Photo: G. Hildebrand, Royal Coin Cabinet, Stockholm 124 5.29 Map showing distribution of hatched cross pennies in the rebuild ­Romanesque church after c. 1220/1250.  Map: Steinar Kristensen, Museum of Cultural History, UiO 125 5.30 Early W-bracteate with high silver content. Photo: G. Hildebrand, Royal Coin Cabinet, Stockholm 126

x Figures 5.31 Map showing distribution of early W-bracteates with high silver content in the Romanesque church after 1280/90.  Original position of Lafrans’ offering box is indicated. Map: Steinar Kristensen, Museum of Cultural History, UiO 127 6.1 Høre stave church. Between the 1660s and 1680s, the church is believed to have been subjected to extensive repairs. Plans for major repairs were also presented in 1724 and 1740.  The church was given its current shape in 1822, when it was expanded, and the exterior walls of the stave church were replaced. Additional repairs were undertaken in 1857, 1888–1889 and 1952.  Photo: Jiri Havran 144 6.2 Høre church divided into polygons according to find descriptions. ­Map: Steinar Kristensen, UiO: Museum of Cultural History 145 6.3 Map of building stages and burials. (A) Posthole church c. 1100–1180, (B) Stave church c. 1180–1822.  Drawing: Jørgen Jensenius, Riksantikvaren 147 6.4 First drawing of the posthole church. Posthole 15 containing a coin marked with “M” (Jensenius’ excavation cards/log 2 September 1979). Drawing: Jørgen Jensenius, Riksantikvaren 149 6.5 Drawing of Høre stave church prior to the rebuilding in 1822.  Unknown artist 150 6.6 All coin finds. Map: Steinar Kristensen, UiO: Museum of Cultural History 151 6.7 Coin finds in the chancel. Two nearby clusters of 13 and 100 coins respectively were found in the apsis. Posthole 15 with the Danish eleventh-century coin is marked with a number. Map: Steinar Kristensen, UiO: Museum of Cultural History 152 6.8 Stone-lined pit in the apse. Photo: Anne Sidsel Herdlevær, Riksantikvaren 154 6.9 All coin finds in the nave. Map: Steinar Kristensen, UiO: Museum of Cultural History 155 6.10 Norwegian coins c. 1177–1202.  Map: Steinar Kristensen, UiO: Museum of Cultural History 156 6.11 Norwegian coins c. 1217–1263.  Map: Steinar Kristensen, UiO: Museum of Cultural History 157 6.12 Norwegian coins c. 1263–1319.  Map: Steinar Kristensen, UiO: Museum of Cultural History 159 6.13 Coins c. 1355–1513.  Map: Steinar Kristensen, UiO: Museum of Cultural History 161 6.14 German coins c. 1300–1500.  Map: Steinar Kristensen, UiO: Museum of Cultural History 162 7.1 Eidskog church. Plan drawing from the 1965 excavation and suggested plans of the two stave churches, Eidskog I and II. Original plan drawing by Martin Tvengsberg, remapped by Steinar Kristensen, UiO: Museum of Cultural History 171 7.2 Present-day Eidskog church. The log church from 1665 seen towards the north-east. (Photo: H. Roland) 172 7.3 Martin Tvengsberg’s plan drawing from the excavation in 1965. X and y-­coordinates are referred to throughout the text. Original plan drawing by Martin Tvengsberg, remapped by Steinar Kristensen, UiO: Museum of Cultural History 174

Figures  xi 7.4 Fragments of Norwegian coins of Magnus Lawmender (1263–1280). (Photo: H. Roland) 175 7.5 The extension of Layer 1 towards the east. Stratigraphy and drawing by ­Martin Tvengsberg 1965.  (Museum of Cultural History, UiO) 179 7.6 Distribution of coins recorded with exact coordinates and chronologically divided into four groups: 1220–1319, 1320–1536, 1537–1620, 1621–1850. Plan drawing by Martin Tvengsberg. Original plan drawing by Martin Tvengsberg, remapped by Steinar Kristensen, UiO: Museum of Cultural History 180 7.7 The extension of undisturbed stratigraphy towards east. Stratigraphy and drawing by Martin Tvengsberg 1965.  (Museum of Cultural History, UiO) 182 7.8 Cluster of coins c. 1340–1400 at 49,6x and 44y. Plan drawing by Martin Tvengsberg. Original plan drawing by Martin Tvengsberg, remapped by Steinar Kristensen, UiO: Museum of Cultural History 184 8.1 Layout of St Mary’s Abbey church (Nunnusetr). The darkest lines indicate the extant lower part of the tower to the west and the vestry to the east (see ­Figures 8.5  and 8.6). The fundaments of the rest of the church are based on documentation by the architects Peter Blix and Schach Bull (Figure 8.2). Drawing by Per Bækken and Alf Tore Hommedal, remapped by Steinar Kristensen, UiO: Museum of Cultural History 196 8.2 Building fragments and graves documented by Schach Bull in 1891. The ­eastern part of the chancel (presbytery) had been excavated by Peter Andreas Blix in 1872.  After Bendixen 1893 196 8.3 Bergen around 1300 with its monasteries. The main settled area of the town is marked as darker grey. The Augustinian abbey of St. John’s was founded in the mid-twelfth century. As Munkalif and Nunnusetr, St. John’s was located outside the populated area. The Mendicant friaries, the Dominicans and the Franciscans, were both located inside the populated areas at the time of their foundations in the middle of the thirteenth century. Drawing: P. Bækken and A. T. Hommedal 197 8.4 St Mary’s at Nunnusetr. The extant part of the 12th century west tower with its main entrance to the church, seen from the northwest. Most likely this is the oldest building still standing in Bergen. Photos: A. T. Hommedal 199 8.5 St Mary’s at Nunnusetr. The extant vestry or side chapel erected c. 1300, seen from the east. The church’s chancel would be found on the right side of the vestry. Photos: A. T. Hommedal 200 8.6 The three building phases in St Mary’s at Nunnusetr: the original church; the extended church; and the extended church with the added vestry. The crosshatched irregular field marks the documented tile floor. Drawing by Per Bækken and Alf Tore Hommedal, remapped by Steinar Kristensen, UiO: Museum of Cultural History 202 8.7 The extant chancel wall of the extended St Mary’s church (the second building phase), now the north wall of the extant vestry (the third building phase), seen from the south. The wall represents the external south wall of the presbytery with its plinth, ending at the

xii Figures south-east corner of the chancel, the vestry jutting out to the east. The door opening led from the vestry to the presbytery. Photo: A.T. Hommedal 207 8.8 Stavanger Cathedral with its east front built after a fire in 1272.  Photo: A.T. Hommedal 208 8.9 A suggested division of St Mary’s at Nunnusetr. The red lines indicate the screens separating the nuns’ choir (middle part) from the presbytery and the nave. The crosshatched irregular field marks the documented tile floor. From the top: the original, twelfth-­ century church (A); the extended church, imagined with two or four rows of stalls in the nuns’ choir (B and C). Drawing by Per Bækken and Alf Tore Hommedal, remapped by Steinar Kristensen, UiO: Museum of Cultural History 209 8.10 Four of the coins found in the extended part of the church (Bendixen’s “­chancel”). See Appendix, MA 348/20 (1–4). Photo: S. Skare, University Museum of Bergen 210 8.11 Bracteate minted for Magnus the Law Mender, found in the extended part of the church (Bendixen’s “chancel”). See Appendix, MA 348/15.  Photo: S. Skare, University Museum of Bergen 211 8.12 Hohlpfennig minted for Erik of Pomerania. See Appendix, MA 348/16.  Photo: S. Skare, University Museum of Bergen 211 8.13 Two hohlpfennigs minted for Erik of Pomerania and one bracteate minted for Magnus the Law Mender. See Appendix IV, MA 348/17 (1–3). Photo: S. Skare, University Museum of Bergen 212 8.14 Five coins found in the church without detailed context. The upper one is the oldest coin found in Nunnusetr, a sterling minted in Canterbury for King Richard I or King John of England, between 1194 and 1200.  See Appendix, MA 348/18 (1–4). Photo: S. Skare, University Museum of Bergen 213 8.15 A bracteate of Håkon IV Håkonsson. See Appendix, MA 348/19 (1). Photo: S. Skare, University Museum of Bergen 213 8.16 A bronze stylus in the shape of a hand or a glove found in the extended part of the church (Bendixen’s “chancel”), i.e. the nuns’ choir (MA 348/b). This is the only complete stylus preserved from medieval Bergen. During excavations of the main settle area of Bergen (Bryggen, Figure 8.3), a tabula where such a stylus would have been used was found. The tabula mention “a runic letter to the sister of Olaf Hettusvein. She is in the nunnery in Bergen”, i.e. Nunnusetr (Ommundsen, “Å skrive med stil”, 2011, pp. 53–55). Photo: S. Skare, University Museum of Bergen 225 8.17 The Benedictine Abbess Kunigunde and her nuns. Bohemian drawing from the period 1312–1321.  From the illuminated manuscript The P ­ assional of ­Abbess Kunigunde in the National Library of the Czech Republic. ­Photography: WikiCommons, PublicDomain 227 9.1 Coin list from Allested Church 1958.  Plan of the division of the floor into numbered sections and a coin list, which refers to the sections. Antikvarisk- Topografisk Arkiv, The National Museum in Copenhagen 238

Figures  xiii 9.2 Aggersborg church is a large Romanesque building consisting of a chancel and nave of granite ashlar. Probably there was an apse at the east gable of the chancel. In the Late Middle Ages (fifteenth–sixteenth centuries), a tower of field stones and bricks in the west and a porch in front of the south door of the nave were added. The present porch was built in the same place in the nineteenth century. In the north side of the nave, the original small windows are preserved, and the entrance door is bricked up. Photo: Vesthimmerlands Museum, Denmark 243 9.3 Aggersborg is located at the north side of Limfjorden. There has probably been a crossing point here since prehistoric times. In the Viking Age, the inlet Limfjorden was the direct passage from Western Europe to Eastern Denmark and the Baltic. Close to Aggersborg was possibly previously a waterway from Limfjorden to the north, which may have connected Jutland and Norway. Due to rising of land levels, sand blow out, draining of land and deposition of marine sediments, the waterways to the North Sea closed in the early Middle Ages (before twelfth century). Map by Steinar Kristensen, UiO: Museum of Cultural History 244 9.4 Aggersborg Church and the Viking fortress seen from the south. Traces of eleventh- and twelfth-century settlement have been excavated to the west and to the east of the older ring fortress. The present settlement by the name of Aggersborg is post-medieval and is related to the crossing point. Despite of this, the huge amount of metal-detecting finds from the Agger peninsula suggests that the whole area was densely settled in the late Iron Age into the Middle Ages. Photo: Lis Helles Olsen, Holstebro Museum, Denmark, with additions 245 9.5 Undated graves outside the present churchyard to the north discovered in 1949.  The first Christian graveyard was probably located at the northern boundary of the present churchyard, and perhaps there was a wooden church in this area too. Photo: Hans Stiesdal, The National Museum of Denmark 245 9.6 Two plans showing the distribution of coins and non-numismatic finds in Aggersborg Church. In the first plan each dot represents one coin. The colours indicate how precise the location of the single coin is known. Dark green is the most precise and red is very imprecise. The fact that more of the precisely located coins are in the east might be a result of excavation methods more or less consciously adapting to the fundamental principles of the layout of churches, which is designed to emphasise the east. This of course does not suit a search for side alters very well. In the second plan, each dot marks a registration number. Each number comprises either one or several different kinds of objects. Unlike the coins, very few registration numbers are located precisely. Map by Steinar Kristensen, UiO: Museum of Cultural History 246 9.7 Non-numismatic finds. The number of objects in each registration number is indicated by the size of the dots. Five clusters are very clear in this plan. Map by Steinar Kristensen, UiO: Museum of Cultural History 248

xiv Figures 9.8 Medieval features which are registered in plans. The floor Æ/CH (phase C) is blue, the wall ditch of the previous wooden nave (phase E) is red and the abandoned stone foundation is beige. Only coins (red) and non-numismatic finds (purple) from these features are displayed in this plan. Phase C-chancel is only registered in sections and is, therefore, not displayed here. Map by Steinar Kristensen, UiO: Museum of Cultural History 249 9.9 Distribution of coins by periods. a.1100–1234, b.1234–1400, c.1400–1541, d.1541–. Map by Steinar Kristensen, UiO: Museum of Cultural History 253 9.10 Chronological distribution of non-numismatic finds. 10a. medieval (until 1600), 10b. early modern, 10c. modern, 10d. not dated. Map by Steinar Kristensen, UiO: Museum of Cultural History 255 9.11 Composition of 301 non-numismatic registration numbers displayed as percentages 256 9.12 Dating of 301 non-numismatic objects 257 9.13 Three identified coin clusters. Map by Steinar Kristensen, UiO: Museum of Cultural History 259 9.14 Dating of objects from modern and undated graves 261 9.15 Distribution map of coins versus non-numismatic finds. Map by Steinar Kristensen, UiO: Museum of Cultural History 262 10.1 Hedensted Church seen from south-east. Photo E. Nyborg, National Museum of Denmark (CC-BY-SA) 269 10.2 Hedensted Church seen from north-east with the fifteenth-century tower, before it was demolished in 1939.  Photo from 1908 K. Hude, National M ­ useum of Denmark (CC-BY-SA) 270 10.3 The Archangel Michael and a demon are weighing souls. Ashlar tablet in the south-east corner of the nave in Hedensted Church. Photo A. Mikkelsen, National Museum of Denmark (CC-BY-SA) 271 10.4 Romanesque granite font from Hedensted Church. Photo A. Mikkelsen, National Museum of Denmark (CC-BY-SA) 271 10.5 Older structures within the frame of the Romanesque Hedensted Church. The function of the little pit between the two wells is unknown. Map: Steinar Kristensen, UiO: Museum of Cultural History based on data of N.H. Bendtsen, Horsens Museum 273 10.6 The chancel in Hedensted Church is seen from the West. The thickened clay floor is visible in the profile. The white circles mark the position of the postholes in the east end of the wooden church. Photo National Museum of Denmark (CC-BY-SA) 274 10.7 Example of part of wooden plank floor preserved in Hedensted Church. The wood is surrounded by a layer of dirt, which was formed during the use of the plank floor. Photo N. Bendtsen, National Museum of Denmark (CC-BY-SA) 275 10.8a  Hedensted Church just before excavation seen from the West. The modern church floor is removed, and heating pipes are dividing the church room into 13 sections 276 10.8b Hedensted Church during excavation, seen from the East. Photos N. Bendtsen, National Museum of Denmark (CC-BY-SA) 277

Figures  xv 10.9a  T he two best-preserved examples of the new bracteates found in Hedensted Church. Photo J. Lee, National Museum of Denmark 278 10.9b  Church finds are a potential source for new unknown coin types. Two different variants are present among the four new bracteates, as one of the coins is decorated with four as opposed to three leaves. Coin no. 4 was found in the undisturbed layer of chippings in the chancel and is therefore decisive for the dating of the construction of Hedensted Church. Drawings P. Wöhliche, National Museum of Denmark (CC-BY-SA) 278 10.10 Building phases of Hedensted Church. The excavated sections are marked with numbers 1–13.  Map: Steinar Kristensen, UiO: Museum of Cultural History based on data of N.H. Bendtsen, Horsens Museum 279 10.11 The chorological and chronological distribution of coins found in Hedensted Church. Map: Steinar Kristensen, UiO: Museum of Cultural History based on data of N.H. Bendtsen, Horsens Museum 280 10.12 One of four lead casting pits found during excavation. A residue of lead is visible in the centre of the pit. Photo N. Bendtsen, National Museum of Denmark (CC-BY-SA) 284 10.13 These three rare Valdemar I, 1154–1182, coins were found north of the chancel arch and possibly constitute a small offering at a side altar. Photo J. Lee, National Museum of Denmark (CC-BY-SA) 285 10.14 The chronological distribution of all 1461 coins, which were declared treasure trove from May 2012 to May 2013 at the National Museum of Denmark. Based on data facilitated by Line Bjerg, National Museum of Denmark 287 11.1 Exterior of Arby church (southern façade) in 2005. Photograph by the author 293 11.2 Exterior of the present Gränna church. Photo: Gränna parish. Used with permission 294 11.3 Gränna church in 1875. In the foreground, the later farmstead on the medieval Husaby site, on the stretch/plateau between the scarp (in the background) and Lake Vättern. Xylography by Wilhelm F. Meyer. Photo: KMB. Public domain 294 11.4 Selected finds from Arby church (SHM 30040). Left: mounting, date?, find no. 52. Top right: clasp, c. 1300, find no. 36; below: hymn book clasp with cherub’s head, seventeenth–eighteenth centuries, find no. 31. Not to scale. Photo: Ola Myrin, SHM. Used with permission 296 11.5 a) Arby church, excavation plan. Drawn by Steinar Kristensen, UiO: Museum of Cultural History after Andersson & Svennebring (1989, Figure 43); b) Arby, church, coin finds 1220-1584 and 1632-1827. Map by Steinar Kristensen, UiO: Museum of Cultural History after Andersson & Svennebring (1989, Figure 43) 297 11.6 Selected finds from Gränna church (SHM dnr 198/87). Top left: candleholder, date?, find no. 167. Top right: ring clasp, date?, find no. 261. Below: comb, date?, find no. 247. Not to scale. Photo: Ola Myrin, SHM. Used with permission 298

xvi Figures 11.7 Gränna church, fifteenth-century crowbar. JLM dnr 198/1987, find no. 286. Photo: Göran Sandstedt, JLM. Used with permission 298 11.8 Gränna church, foundations of the Romanesque (brown) and present church, and the excavated trenches. Drawn by Steinar Kristensen, UiO: Museum of Cultural History after Varenius (1993, Pl. 2 and 3) 299 11.9 Number of coins per century, Arby church 303 11.10 A selection of coins found in Arby church. From top left to bottom right: 1: Sweden, Avesta mint, Karl XI, 1/6 öre 1676 (Cu, 6.16 g., no. 48); 2: Sweden?, spangle, reused coin? (Cu, 2.77 g., perforated, no. 58); 3: ­Sweden, Avesta, Karl XIV Johan, 1/4 skilling 1825 (Cu, 3.69 g., no. 65); 4: ­Denmark, Malmö mint, Kristian II, klipping 1513–23 (square, Cu, 1.99 g., no. 55); 5: Sweden, Erik Eriksson?, penning 1220–50 (Ag, 0.14 g., perforated, no. 91); 6: Sweden, Gotland?, penning 1220–50 (Ag, 0.08 g., no. 131); 7: S­ weden, Valdemar, penning 1250–75 (Ag, 0.14 g., no. 90); 8: Sweden, Magnus Birgersson Ladulås, halvpenning 1275–90 (Ag, 0.06 g., no. 89); and 9: ­Germany, Mecklenburg, pfennig 1400–50 (Ag, 0.20 g., no. 139). Nos. 91, 90, 89 and 139 are bracteates. Photo: Ola Myrin, KMK 304 11.11 Number of coins per century, Gränna church 305 11.12 A selection of coins found in Gränna church. From top left to bottom right: 1: Sweden, Avesta, Karl XI, 1/6 öre 1666 (Cu, 6.37 g., no. 55); 2: France, Lyon, denier after 1200 (Ag, 1.30 g., no. 21); 3: Germany, jeton, eighteenth century (?) (brass, 0.89 g., perforated, no. 51); 4: Gotland, Visby, gote 1420–50 (Ag, 0.78 g., no. 108); 5: Norway, Olaf Kyrre (?), penning c. 1065–80 (Ag, 0.45 g., no. 1); 6: Sweden, Valdemar, halvpenning 1250–75 (Ag, 0.17 g., no. 12); and 7: Sweden, Uppsala?, penning 1260–75? (Ag, 0.23 g., no. 35). Nos. 12 and 35 are bracteates. Photo: Ola Myrin, KMK 305 11.13 Arby church in 1747 exhibited still with medieval windows, doors towers. Details are not to scale, but measures are indicated in writing on the drawing, showing that the towers are in reality twice as broad in the E–W direction. Drawing by Petrus Frigelius. photograph: RAÄ/KMB. Public domain 310 11.14 Gränna church, plan from report with added structures, burials and coin finds. Drawing and GIS by Steinar Kristensen, UiO: Museum of Cultural History after Varenius (1992) 312 11.15 Gränna church, interior of the chancel and apsis after the fire in 1889. Photo: Carl Curman, RAÄ/KMB. Public domain 313 11.16 Gränna church, Credo paintings in the apsis of the Romanesque church. Drawing by Nils Månsson Mandelgren (published 1862, Pl. XIV Figure 4). Photo: RAÄ/KMB. Public domain 314 11.17 Seal matrices from Gränna church: one from the sixteenth century (cross keys, right), the other from the seventeenth century (church, left). Photo: Torgny Lundqvist. Used with permission 315 12.1 Map of southern Sweden. Map: Steinar Kristensen, UiO: Museum of Cultural History 332 12.2 The seal matrix of the Knight Abjörn Sixtensson, AD c. 1295.  Photo: by permission SHM 333

Figures  xvii 12.3 12.4 12.5 12.6

The clearance cairn at Klosterstad. Photo: Rikard Hedvall 334 The remains of the round church at Klåstad. Photo: Rikard Hedvall 334 The traces of the wooden church at Klåstad. Photo: Rikard Hedvall 335 The church ruin at Klåstad: early and later graves. Dark green signifies burials associated with the older wooden church, while light green signifies burials associated with the round church. Plan by P. Carlsson, Riksantikvarieämbetet, and Steinar Kristensen, UiO: Museum of Cultural History 337 12.7 The church ruin at Klåstad: coin finds from AD 1150−1300. Plan by P. Carlsson, Riksantikvarieämbetet, and Steinar Kristensen, UiO: Museum of Cultural History 338 12.8 The church ruin at Klåstad: coin finds from AD 1300−1400 and fragments of wooden floor. Plan by P. Carlsson, Riksantikvarieämbetet, and Steinar Kristensen, UiO: Museum of Cultural History 339 12.9 The church ruin at Klåstad: coin finds from AD 1400−1520 and fragments of brick floor. Plan by P. Carlsson, Riksantikvarieämbetet, and Steinar K ­ ristensen, UiO: Museum of Cultural History 342 12.10 Portable offertory plate (lat. tabula) from Hemse church, Gotland, fifteenth century. Photo: by permission ATA 343 12.11 Coins found in the church ruin at Klåstad (not to scale). The find numbers refer to Table 12.1.  Photo: Ola Myrin, layout by Frida Roland 345 13.1 The church of Jomala. Exterior from the north. Photo: MrFinland, ­WikiCommons (CC-BY-SA) 351 13.2 Plan of the Romanesque stone church in Jomala. Map: Steinar Kristensen, UiO: Museum of Cultural History after Markus Hiekkanen 2007 354 13.3 Distribution of the pierced coins in Jomala. Map: Steinar Kristensen, UiO: Museum of Cultural History 359 13.4 Distribution of the medieval Swedish coins found in the church of Jomala, Åland. Map by Steinar Kristensen, UiO: Museum of Cultural History 363 13.5 Coin finds in the church of Jomala during the reign of Gustav I (1523−1560). Map by Steinar Kristensen, UiO: Museum of Cultural History 365 13.6 Coin finds in the church of Jomala during the reign of Johan III (1568−1592). Map by Steinar Kristensen, UiO: Museum of Cultural History 366 13.7 Distribution of coins struck in the seventeenth century in the church of Jomala, Åland. Map by Steinar Kristensen, UiO: Museum of Cultural History 367 13.8 Distribution of the one öre copper coins struck in 1719−1726 in the church of Jomala, Åland. Map by Steinar Kristensen, UiO: Museum of Cultural History 368

Background maps in this publication are used under licence from the Norwegian Mapping Authority CC BY 4.0 (@Kartverket)

Tables

1.1 5.1 5.2 5.3 5.4 6.1 7.1 8.1 8.2 8.3 8.4 8.5 9.1 9.2 9.3 10.1 11.1 11.2 12.1 13.1 13.2 13.3 14.1

Approximated number of coins found in Scandinavian churches. In comparison, almost 22,000 coins are found at about 640 ecclesiastical locations in Switzerland Medieval coins minted and used during the period of the Romanesque church building Medieval coinages minted in the period of the Gothic church building Early modern and modern coinages in Bunge church Foreign coins from churches on Gotland up to the seventeenth century, compilation based on Fridh 2014 Coin finds of Høre stave church Chronological distribution of coins in identified stratigraphical layers Proposed chronological phases based on orders resident at the Bergen Nunnusetr St Mary’s. Proposed chronological dating of the church’s building phases, containing both the periods of work and completion The Benedictine period, c. 1120–1455. Coins identified from the excavations of 1891 (42) and 2006 (1) with their context Post-Benedictine periods (c. 1455–1891) and non-identified coins, with their context. All coins identified from the excavation of 1891 List of coins found in Nonneseter monastery The most important excavated layers and constructions (phases 2–7) in the excavation report with datings suggested by the excavators Geographical and chronological distribution of the 341 coins found in Aggersborg church. The numbers refer to the publication of the coins by Jørgen Steen Jensen The chronology of coin finds from graves, pit J and top layer (phase A) Coin finds related to building phases of Hedensted Church Coin finds from Arby church Coin finds from Gränna church Coin finds from the church ruin at Klåstad (KMK 711-609-2005). Find number in bold means that the coin is pictured in Figure 12.11 Medieval coin finds in the church of Jomala Post-Reformation coin finds in the church of Jomala Number of possible coin offerings to the Virgin Mary during the medieval period and the reign of Gustav I (1523−1560). Only the Swedish coins are included The largest church coin finds in Switzerland

21 99 100 101 102 146 180 204 209 214 214 228 250 252 260 282 300 306 346 356 357 364 405

Contributors

Nanouschka Myrberg Burström  is Docent and Senior Lecturer at the Department of Archaeology and Classical Studies, Stockholm University. She is a specialist in the material culture of Viking and Medieval Scandinavia and in numismatics. Her major publications include Ett eget värde. Gotlands tidigaste myntning, ca 1140– 1220 (A Worth of their Own. The earliest coinage of Gotland, c. 1140–1220) (2008) and the edited volume Divina Moneta: Coins in Religion and Ritual (2018). Svein H. Gullbekk is Professor of numismatics and history of money at Department of Numismatics and Classical Archaeology, Museum of Cultural History, Oslo University; and is Principal Investigator for the project “Religion and Money: Economy of Salvation in the Middle Ages”. He is particularly interested in money and its use in Viking and medieval societies. Alf Tore Hommedal is Associate Professor at the University Museum, University of Bergen. He is a specialist in Norwegian archaeology, especially of ecclesiastical sites from the Middle Ages. He has published widely on church archaeology and medieval religious studies. Gitte Tarnow Ingvardson  is Curator at the Historical Museum, Lund University, Sweden, and a specialist in Viking and medieval numismatics, especially coin use, hoards in archaeological contexts and coins in detector archaeology. She has published Møntbrug. Fra vikingetid til vendertogter (2010) and Vikingeskattenes mennersker: Bornholmske sølvskatte som aktører i det økonomiske, sociale, kulturelle og symbolske felt fra ca. 850–ca. 1150 (2020) and is the editor of Divina Moneta: Coins in Religion and Ritual (2018). Eeva Jonsson is Researcher at the University of Stockholm, specializing in medieval church archaeology and numismatic history. She is currently a PhD candidate at Åbo University. Martin Wangsgaard Jürgensen is Editor for the Danish Church record for the National Museum, Denmark. He is a specialist in medieval liturgical practices and church architecture, and he has published Changing interiors: Danish Village Churches c. 1450 to 1600 (2011). Christoph Kilger is Associate Professor in the Department of Archaeology and Ancient History, Uppsala University, Campus Gotland. He specializes in Viking and medieval archaeology and numismatics, especially relating to social, economic and monetary history of Northern Europe. He has a long record of publications

xxii Contributors discussing a wide range of subjects and issues related to archaeology, coinage and money in Viking and medieval societies. Henrik Klackenberg is State Herald, Swedish National Archives, Stockholm; is former director of the Royal Coin Cabinet in Stockholm; and is the author of the standard reference book on coin finds in churches, Moneta nostra: monetarisering i medeltidens Sverige (1992). Steinar Kristensen is Senior Engineer at the Museum of Cultural History, University of Oslo, with responsibility for digital documentation of major excavation projects. He specializes in GIS technology in archaeological contexts. Jens Christian Moesgaard is Professor of numismatics and head of Stockholm Numismatic Institute at Stockholm University. He is an expert on Danish and Norman Viking Age and medieval coinage. He has a special interest in coins and archaeology. He has a long record of publications, including King Harold’s Cross Coinage. Christian Coins for the Merchants of Haithabu and the king’s Soldiers (2015). Henriette Rensbro  is Curator at the Department of the Middle Ages, Renaissance and Numismatics, National Museum of Denmark. She is a specialist in medieval archaeology, especially church archaeology; and is the author of Spor i kirkegulve. De sidste 50 års arkæologiske undersøgelser i kirkegulve som kilde til sognekirkernes indretning og brug i middelalder og renæssance (2007). Jon Anders Risvaag is Associate Professor, Department of Archaeology and Cultural History, NTNU University Museum, Norwegian University of Science and Technology, Trondheim. He specializes in Norwegian numismatic and monetary history of the Middle Ages and has published Mynt og by. Myntens rolle i Trondheim by i perioden ca. 1000–1630, belyst gjennom myntfunn og utmynting (2006) and The Mint in the Nidaros Archbishop’s Palace. Coin production under Archbishop Gaute Ivarsson (1475–1510) (2010), with O. Lohne, P. Ulseth and J. Lohne. Håkon Roland  is Associate Professor of numismatics and classical archaeology at the Museum of Cultural History, University of Oslo. He specializes in ancient and medieval numismatics and archaeology. He has a long record of publication on numismatics and archaeology and within Heritage studies. Benedikt Zäch  is Director of the Münzkabinett und Antikensammlung der Stadt Winterthur. He has a long record and extensive list of publications covering numismatics, archaeology and the history and use of money in Switzerland and European Middle Ages, with a particular interest in coin finds from ecclesiastical sites.

Preface

The primary objective of this edited volume is to examine the use of money in medieval churches, studying its possible relationship with liturgical practices encompassing devotional ceremonies and ritual offerings. Secondarily, we seek to understand the churches as monetary arenas with an emphasis on rural parishes where the majority of the coin finds are recorded. Individual medieval churches in Scandinavia and Finland provide well-furnished settings for testing various hypotheses, focusing on the period between the late twelfth and the early eighteenth century when coins were circulating in great numbers in ecclesiastical religious environments. For the sake of comparison, coin finds from churches in Switzerland and Central Alpine Europe will be addressed and discussed in the final chapter. In total, there are more than 100,000 known coin finds from Scandinavian churches and another 20,000 from Switzerland. Nowhere in the archaeological record of the High Middle Ages is it possible to study a richer body of single finds of coins within religious contexts. Of course, these numbers reflect but a small fraction of the total flow of coins that undoubtedly passed the doorsteps of these churches over the centuries, and certainly a very small proportion of the total numbers of coins entering churches throughout all of European medieval society. However, what they do reflect well is a large number of individual acts of devotion made by churchgoers, and the robust numbers of finds can be seen to amount to telling representative samples. This volume collects research from different scholarly traditions and approaches to devotional practices across diverse parishes, provinces and landscapes, all within the backdrop of Medieval Latin Christendom with its concomitant awareness of money and its uses. These involve the values and economics of the pervasive Church as an institution, as well as within the specific church buildings themselves (e.g. the giving of alms, tithes). The discussions presented here shed light on significant shared notions of practice and principles in the handling of coinage in ecclesiastical environments throughout the medieval and early modern period, as well as quite dynamic differences on local, regional and national scales. This book is more than a collection of articles. It is the result of a research process that has brought together different methodologies and theoretical frameworks from disciplines such as numismatics, medieval history, archaeology, theology and art history. Interdisciplinary in its scope and ambition, the project has revitalized earlier archival and collections research on medieval money and its use in medieval and early modern churches. Some contributions may speak more to the archaeologist than to the historian, or to the architect or the art historian, but each brings a novel transdisciplinary approach to our understandings of how ordinary people in the

xxiv Preface Middle Ages thought about money in the humdrum, quotidian interactions of faith and responsibility. Contributions range from focus on the history of buildings and how they shaped liturgical rituals and the daily religious lives of people, empirical analyses of religious space and its functions, to the topic of monetization in Medieval and post-Reformation societies. Many of the contributors in this book also present and discuss a range of methodological and source-critical issues related to the analysis and interpretation of coins and other small finds from complex and often disturbed archaeological layers in churches. In addition, working with excavated materials poses many challenges and raises many questions. This is not least because the documentation from church excavations and renovations comes in many different forms and levels of detail and accuracy and is only rarely available in published or otherwise accessible reports. Many of the observations and conclusions presented in this volume are the result of painstaking exegeses of archival sources such as unpublished reports, excavation notes, topographical reports, letters, diaries and other personal communications, all supplemented by original photographs where possible; furthermore, the systematic revisitation of archaeological investigations based on documentation of varying quality and difference in excavation methods is a demanding exercise, perhaps characterized at best as an interpretation of manifold interpretations. Inevitably, the practical preconditions of an excavation and the methodological and theoretical–paradigmatic decisions made in its undertaking directly and indirectly affect and bias such subsequent scholarly interpretations. Many of the contributions here provide valuable insights into how best to navigate and ultimately begin to consolidate such rough scholarly terrain. This book is the outcome of the interdisciplinary research project “Religion and Money: Economy of Salvation in the Middle Ages” funded by the Norwegian Research Council. The Museum of Cultural History, University of Oslo, hosted the project, which generated a range of publications, workshops, conferences and lectures and a museum exhibition, during the project period 2013–2017.1 The Editors would like to thank Sven Svensson’s Foundation for Numismatics for generously contributing to the production of this book. We wish to thank all of the authors who have contributed to this volume, as well as the many colleagues and friends who have participated in and made contributions to this field of study as part of the “Religion and Money” project. Finally, there are a few people who have made significant contributions related specifically to this book to whom we are also grateful: Kristin Bornholdt Collins, Astri Karine Lundgren, Kay Celtel, Sean O’Neill and Anette Sættem, and the external reviewers.

Note 1 This is the fifth volume published under the auspice of this project: Money and the Church in Medieval Europe, 1000–1200. Practice, Morality and Thought, edited by G.E.M. Gasper and S.H. Gullbekk (Farnham: Ashgate 2015); Divina Moneta: Coin Finds in Religious Contexts, edited by N. Myrberg Burström and G. Tarnow Ingvardson (Abingdon: Routledge 2017); S.H. Gullbekk and A. Sættem, Norske myntfunn 1050–1319. Penger, kommunikasjon og fromhetskultur (Oslo: Dreyer forlag 2019); Coins in European Churches: Religious Practice and Devotional Use of Money, SNR vol. 97, 2019, special issue (2020), edited by B. Zäch and S.H. Gullbekk; and Peter Bjerregaard (ed.), Transformation: Faith and Sacred Objects in the Middle Ages (Trondheim: Museumsforlaget, 2019), for the exhibition with the same title opened at the Museum of Cultural History in Oslo, January 2019.

Abbreviations

ÅMU

Ålands Medeltidsurkunder, 1−1400, edited by Johannes Sundswall, Helsingfors: Ålands kulturstiftelse, 1954 Askeladden ID The Norwegian National Cultural Heritage Database ID ATA Antikvarisk-Topografiska Arkivet, Stockholm, Sweden. BAR British Archaeological Reports, Oxford. Bnr Bruksnummer/Title no. DIVA Digitala vetenskapliga arkivet, DiVA portal institutional repository for research publications and student theses written at universities in Sweden. DMS Det medeltida Sverige/Medieval Sweden. The historical topographic research project Medieval Sweden is primarily focused on presenting the documentary sources for the settlements in the country in the Middle Ages and the sixteenth century until the 1570s. DN Diplomatarium Norvegicum. Oldbreve til Kundskab om Norges indre og ydre Forhold, Sprog, Slægter, Sæder, Lovgivning og Rettergang i Middelalderen. Vol. I–XXIII. 1847–2011. ChristianiaOslo: Kommisjonen for Diplomatarium Norvegicum. DS Diplomatarium Suecanum, Svenskt Diplomatarium. Printed edition of Swedish medieval charters, published by Riksarkivet (the [Swedish] National Archives) and Kungl. Vitterhets Historie och Antikvitets Akademien (the [Swedish] Royal Academy of Letters) since 1829 and still continued. FMU Finlands Medeltidsurkunder samlade [Finnish medieval charters], vol. I–VIII (Helsingfors, Statsarkivet, 1910, 1915, 1921, 1924, 1928, 1930, 1933, 1935). FNFBÅ Foreningen til Norske Fortidsmindesmærkers Bevaring, Aarsberetning (Bergen) Gnr Gårdsnummer/Cadastral. Hkr Heimskringla, edited by C.R. Unger (Christiania (Oslo) 1868); trans. Anne Holtsmark og Didrik Arup Seip (Oslo: Gyldendal 1970). IFS Inventar der Fundmünzen der Schweiz. Inventory of Coin Finds from Switzerland. JLM Jönköpings läns museum/Jönköping County Museum, Jönköping, Sweden.

xxvi Abbreviations KHLNM KLNM KMB KMK MA MCH NGL NGL NIKU NMD NNÅ NNUM NTNU RA RAÄ RB

RN SAF SD

SDHK SHM SIF SKAS SML SNL

Kulturhistoriskt lexikon för nordisk medeltid från vikingatid till reformationstid, edited by I. Andersson, J. Granlund et al. 22 vols. (Malmö, 1956–78). Kulturmiljöbild, Internet resource made available by the National Heritage Board, Stockholm, Sweden. http://www.raa.se/kmb The Royal Coin Cabinet, Stockholm, Sweden. arkiv Middelalderarkivet, today part of the Topographical Archive in the Museum of Cultural History, University of Oslo. Museum of Cultural History, University of Oslo Norges gamle Love indtil 1387, edited by Rudolph Keyser and Peder Andreas Munch, 5 vols. (Christiania: Chr. Gröndahl, 1849–1895). annen række Norges gamle Love, anden række, 1388–1604, edited A. Taranger (Kristiania: Chr. Gröndahl, 1904–1912). Norsk Institutt for Kulturminneforskning. National Museum of Denmark. Nordisk Numismatisk Årsskrift/Nordic Numismatic Journal. Nordisk Numismatisk Unions Medlemsblad. Norwegian University of Science and Technology. Riksantikvaren. Directorate of Cultural Heritage, Norway. Riksantikvarieämbetet, the [Swedish] National Heritage Board, Stockholm, Sweden. Biskop Eysteins Jordebog (Den Röde Bog). Fortægnelse over det Geistelige Gods i Oslo Bispedömme omkring Aar 1400, edited by Henrik Jörgen Huitfeldt (Christiania: J. Chr. Gundersens Bogtrykkeri, 1880). Regesta Norvegica, edited by Gustav Storm, Narve Bjørgo, Sverre Bagge et al. 10 vols. (Oslo: Norsk Historisk Kjeldeskriftinstitutt, 1898–2017). Schweizerische Arbeitsgemeinschaft für Fundmünzen. The Swiss Association for Coin Finds. Svenskt Diplomatarium. Printed edition of Swedish medieval diplomas, published by Riksarkivet (the [Swedish] National Archives) and Kungl. Vitterhets Historie och Antikvitets Akademien (the [Swedish] Royal Academy of Letters) since 1829 and still continued. Svenskt Diplomatariums huvudkartotek över medeltidsbreven. Online edition of Swedish medieval charters. http://riksarkivet.se/ sdhk Swedish History Museum, Stockholm, Sweden. Inventar der Fundmünzen der Schweiz. Swiss Inventory of Coin Finds. Suomen Keskiajan Arkeologian Seura. Sveriges mynthistoria landskapsinventeringen, Uppland. Stockholm: Kungliga Myntkabinettet, 2004. Store Norske Leksikon/Great Norwegian Encyclopedia.

Abbreviations  xxvii SNR TM

Schweizerische Numismatische Rundschau. The Swiss Numismatic Review. Trésors monétaires, Bibliothèque nationale, Paris.

Part I

Money and religious devotion

1

Money and religious devotion in medieval Northern Europe Svein H. Gullbekk, Christoph Kilger, Håkon Roland and Steinar Kristensen

What is this book about? Why produce a book dedicated to coin finds in medieval churches? One obvious reason is that, in total, the number of coin finds from medieval churches in Northern Europe is greater than from any other defined archaeological context. In sum, more than 120,000 single finds of coins have been recorded from Scandinavian and Swiss churches. Not even urban archaeology has produced greater numbers of medieval single finds of coins than those produced through archaeological work in churches. There is simply no easier way of finding systematic empirical examples of the relative importance of coinage and its distribution in medieval societies than by following the money found inside churches, especially within rural societies. The single framework within which all these coins have been recorded suggests a strong relationship between Christianity and monetisation. Furthermore, studying the devotional use of money in medieval churches provides opportunities to investigate questions of how the medieval Church and church congregations navigated the Economy of Salvation on socio-economic, religious, and everyday terms. This volume examines the subject of medieval economy first and foremost from spiritual and sacred approaches. First, chapters detailing archaeological excavations and investigations of church floors develop a starting point for the analysis of individual churches as case studies. Research is drawn from specific churches within the medieval borders of Denmark, Norway, and Sweden, i.e., roughly within the borders of today’s Scandinavia. Second, chapters take on diverse theoretical perspectives, from exploring theological logics rooted in money to the interplay between epistemology and praxis in terms of monetary exchange within the Church and its manifestations through the variety of parish churches where devotional practices were carried out. This is not meant as a handbook on archaeology or a history of Christendom in Scandinavia during the Middle Ages. Rather, the present contributions aim to form a relatively cohesive collection of articles which use archaeology as a starting point for engaging with coin finds in churches as a material basis for making broader insights into monetary production and consumption at the intersection of secular and religious life. It also includes discussions of a broad range of archaeological concerns around coin finds and documentation of their contexts. As such, this volume builds a framework for comparison between churches across a similar chronological range (c. 1100–1700) and across a broadly heterogeneous geopolitical and cultural landscape – medieval Scandinavia – against the backdrop of coalescing monetary economies in medieval Northern Europe. This enables a fuller and richer assessment

4  Svein H. Gullbekk et al.

Figure 1.1 Scandinavian churches and church sites mentioned in this book. Map: Steinar Kristensen, UiO: Museum of Cultural History.

of the devotional culture and its economy within the churches presented. From these insights, wider questions are drawn. The case studies address surveys of coin finds and their contexts in churches from throughout Scandinavia and offer an overall survey of coin finds and scholarship on the subject of such finds in churches in the Alpine region, with an emphasis on Switzerland. Looking at these two regions within Western Christendom provides a broad base of evidence for the use of money within a religious context while conveniently contextualising the Scandinavian finds with analogous finds from elsewhere in Europe, and vice versa. Presenting the churches and the material evidence of offerings from the medieval and early modern periods in this way makes an extensive body of material both accessible and relevant to a broader scholarly community. The final aim of the volume is to highlight the advantages and necessities of interdisciplinary work. This work represents the first time that the finds presented here have been studied within a wider medieval and early modern, and not merely

Money, religious devotion  5 Scandinavian, context. Questions of a more general character have been generated using the data from the volume’s geographical and chronological framework which takes new and innovative approaches to the archaeology of medieval Christianity to interpret devotional practices. The articles are all offsprings of the project “Religion and Money: Economy of Salvation in the Middle Ages,”, which deliberately encouraged participants to combine the traditional historical emphasis of Christian archaeology with social archaeological questions to enrich the traditional historical emphasis with questions that focus on social and religious practice. The research questions deal not only with the physical environment, and how and where people worshipped, but also with how particular religious contexts and practices influenced changing relationships between individuals and groups, the negotiation of values and power, and the movement of people and objects. Ultimately, it is our hope that this contribution inspires more questions and further discussion.1 As the contributions cover a range of issues relating to money, its use and significance, and the relationship between tenets of the medieval church and the monetised aspects of value and faith, here we offer a brief introduction to some of the major concepts dissected in the volume. These include concepts surrounding medieval Christian notions of salvation and exchange; ways in which coin finds and other material sources reflect changes in liturgical practice; how the introduction of money economy affected change in various aspects of devotional practice and Church interiors; the function of money and coins as metaphors in early Christian thinking; and church archaeology and coin finds.

The Church, salvation, and money In medieval Christian theology, communion reflects the actualisation of mankind’s salvation through Christ’s crucifixion and resurrection. The sacrifice of Christ and the performance of communion belong to the most circumscribed theme of medieval religious epistemology. 2 This theoretical discourse is kaleidoscopic in its breadth and depth, often so much so that observing and understanding it through an archaeological lens becomes near impossible. Christ’s sacrifice on the cross is not repeated but is made alive to believers at Mass, where the priest essentially acts in the place of Christ, in persona Christi, during the Eucharist. One quintessential aspect of Mass during the medieval period became the collection of offerings either in kind or in money. At the heart of religious belief in the Middle Ages was the concern with finding the way from the agony of earthly existence to heavenly salvation, as so eloquently expressed in a Norse homily from the twelfth century: “Since we’re going to be here in the world just a short while, good brothers […]. And the Kingdom of Heaven is so good that no human can imagine it or portray it to anyone else.”3Another driver in the medieval belief system was the concept of purgatory and the quest to shorten one’s stay in this painful intermediate state after death through various acts of penance.4 To succeed in these endeavours, people had to act as good Christians. 5 Through indulgence, the Church offered guidance on how to be a good Christian along with pardons for sins.6 The logic of indulgence required the sinner to perform acts of penance. The making of offerings was part of this performance and fulfilled the Church’s expectations of what being a decent and pious Christian looked like.7 Offering itself was regarded as an individual act of devotion, embodying the Christian belief in resurrection and salvation. ‘To offer’ comes from the Latin transitive verb offerre, meaning ‘to bring before, to present, to bring forward, to dedicate, to

6  Svein H. Gullbekk et al. offer.8 The verbal expression itself signifies the performative act of moving towards the altar and placing something on the shrine itself. In medieval ecclesiastical texts, the accomplished offering and the things given as a result of this performance are referred to as oblationes. One central and communal act of Mass was the offertorium, during which participants carried and presented their offering to the high altar under the eyes of the congregation. As a ritual, it is mentioned in Swedish law texts from the fourteenth century, which also provide one of the few descriptions of monetary offerings in the shape of coins: On these solemn days, the peasants and their wives will offer their sacrifice to the altar, the farmer one penny and the wife another: on Christmas Day, Candel Mass, Easter Day, Pentecost Day, the Last Spring Eve, the Sabbath Mass, and the Church Day. And the wives also offer their altar gifts […].9 As discussed by the authors in this contribution, there existed a wide spectrum of instances and occasions where offerings of coins were performed when people visited churches. Such offerings were, of course, made during the ordinary church services that every peasant was expected to attend, but they were also made during local trade fairs, a common phenomenon in the medieval economy. Trade fairs and church holidays, described extensively in medieval sources, were another key factor in the implementation of coinage and monetisation of medieval society. Importantly, these gatherings were not only nexuses facilitating the exchange of goods, services, and money, but were also presided over by the Church. The purpose of Church holidays was to celebrate patron saints and feast days, and they developed into pilgrimages that attracted people from beyond the parish seeking to attend Mass and provide offerings for the saints. These church holidays were boosted by indulgence letters issued by various Church authorities and by offering penance for sins to those willing to participate.10

Figure 1.2 A selection of coins from Lom stave church, Oppland County, Norway. Photo: Museum of Cultural History, University of Oslo

Money, religious devotion  7 Through this multifaceted relationship between the Church and the people in the medieval belief system, the concepts of salvation and economy became intrinsically entwined. Thus, in medieval society, piety through monetary offerings related directly to the recognition of Christ’s sacrificial exchange and the fear of purgatory. Just as the manifestation of Christ the victim was transformed into salvation, that salvation became currency and could thus find material expression in the form of money, from offerings of humble pennies to the construction of cathedrals. Importantly, in the eyes of believers, offerings were not regarded as payments but seen as an individual act of free will performed in the church as a means of atonement and redemption from sin. During the Reformation, Luther and other reformers condemned such offerings and the excessive abuse of letters of indulgence for providing a shortcut to salvation and a means by which to bargain with God.11 In medieval society, part of the devotional ideology that included money in its different expressions was the imposition of various expectations which could be remunerated by fees or labours. The performances by which individuals could prove their piety in the eyes of the Church were many and varied, as attested by medieval letters of indulgence. They included tithing, penance for sins, pilgrimage, participation in crusades or payment of crusader taxes, making of various offerings such as the giving of St Peter’s Pence, alms, and altar offerings, among other things. During this period, the transfer of wealth from individuals to the Church became massive. In fact, few researchers now question that the Church was the most important economic player during the Middle Ages. By the twelfth century, the Church had become Europe’s largest landowner, placing it in a unique position of power. Revenues generated were huge. At the same time, maintaining the Church system was an enormous expense. Local and regional economies flourished from the construction of cathedrals, monasteries,

Figure 1.3 Letter of indulgence dedicated to the altar of the Apostles in Våsterås cathedral, Sweden. Letter issued on 10th January, 1471 by Swedish archbishop Jakob Ulvsson and bishops of church provinces Västerås, Skara, Strängnäs, and Växsjö. Foto: Christer Åhlin, The Swedish History Museum, Stockholm, SHMM License, CC-BY.

8  Svein H. Gullbekk et al. and thousands upon thousands of church buildings. Add to this the need to support and feed tens of thousands of priests, bishops, monks, and nuns, the scope of the Church as a business entity, as it became during this period, is staggering. In short, by the eleventh century, the Church had undeniably become an economy in its own right. At the heart of this extensive economy were money-offerings, not necessarily in monetary terms but at least in symbolic significance. These could come in many shapes and sizes, from considerable amounts of money for building projects to insignificant donations. But taken together, these offerings were a driving force, at both the local and the central level, of the significant economy that was the Church.

Piety and money The material sources presented in this publication have rarely been utilised in previous research on the relationship between piety and money. The various customs of using coins as a means of expressing piety have been described in biblical literature and discussed in theological scholarship and are also abundantly referenced in legal documents of the time.12 The culture of piety in the Middle Ages has been studied within a variety of disciplines, such as theology, history of religion, art history, architecture, and archaeology. This book focuses on medieval cultures of piety, specifically, as it was transformed through and into money. Church doctrine in the medieval period dictated where, when, how, and for how long certain individuals and classes of individuals within the congregation could occupy the church space itself. The archaeology of women in religious contexts offers many interesting perspectives on the gender and agency of people attending Mass.13 A ­gender-based approach has contributed to a considerable number of discoveries in which, in some cases, finds are so spatially delimited that it has been possible to map the distribution of coins and, in certain cases, archaeological small-finds like beads and pins within the church with a fair degree of accuracy.14 This has made it possible to observe differences in the frequencies of coin finds in the different sections of the church that were occupied by men and women during Mass.15 The archaeological record opens up new possibilities to bring discussions on segregation and gender-based division further. The devotional rituals involving coins have proved to be exceptionally long-lived, like many aspects of Christian tradition; we can observe and analyse coin usage in relation to changes in liturgy, interior church architecture, building structures, and a broad array of social and economic phenomena from local vagaries to events of pivotal importance, such as the Black Death and the Reformation. Observing this through archaeological assemblages of coins evidences the ­universal character of these exchanges throughout the whole of Christendom. At St. ­Peter’s tomb in St. Peter’s Basilica in Rome, there are coin discoveries that show that ­money-offerings were a widespread custom from the time of the first Christians in the first century AD to when the tomb was closed during the rebuilding of the cathedral in the fifteenth century. For those who flocked to Peter’s titular tomb, the use of money was a natural aspect of the pious acts of good Christian pilgrims.16 Hundreds of archaeological excavations in less-legendary churches tell similar stories: coin finds suggest continuity from the time of the Early Christian to post-Reformation times. As historical records, these archaeological sources provide evidence that the income from offerings was as vital for the construction of rural parish churches as cathedrals.17

Money, religious devotion  9

Figure 1.4 Pious laity at the offertory during Mass, offering money at the offering box, the altar, and giving alms. Regensburger Altar, c. 1480. Photo: Bayerisches National museum, München, MA 3304

Below, we offer just two examples of how aspects of the Church (the liturgical process of worship) and the interior of church buildings affected the relationship and distribution of people and, thus, money inside medieval churches.

Reflections of changes in liturgical procedures Oxford architectural historian and theologian Allan Doig makes an obvious but important comment in the introduction to his magisterial study of liturgy and architecture that “every act of worship must have a setting.”18 The relationship between liturgy (the act) and architecture (the setting) is everywhere close. It is reflected in the shape of the church building itself, the locations, positions, and shapes of its entrances

10  Svein H. Gullbekk et al. and windows, its altar(s), in the placement and types of vessels used to store and present the Eucharist. Devotional cultures and acts need administration; in this way, the liturgy and its context shape one another. Thus, changes in the liturgy are often reflected and perpetuated in church architecture and interior design. In memory of the crucified Christ, a crucifix was placed on or above the altar. As the American scholar Jeanne Halgren Kilde, who works on religious architecture, points out: The Christian altar became a privileged setting for gifts to God and to his church, because it was also the setting for the commemoration of God’s gift to mankind: The sacrifice of the body of Christ which he had enjoined all believers to share in and which defined the membership of his church, the Eucharistic experience of the reciprocity of giving, the constant and necessary repetition of giving, and the spiritual dimensions of giving, were fundamental.19 By the thirteenth century, it was prescribed that all churches should have an image of the Virgin Mary on or near the high altar and that the celebrant at Mass should have an image of Christ Crucified before him on the altar. 20 We see the design of the space itself transformed along with the symbolism attributed to its features. The spatial organisation of people in specific ways is a characteristic shared by all sacred and religious spaces. How people organise themselves and behave within specific places imbue those places with significance, sacred and otherwise. Space is made sacred by human action and behaviour, and certain spaces become sacred simply because people treat them differently from ordinary spaces. 21 Thus, a liturgy that was once a communal prayer became, by the thirteenth century, a clerical ritual separated from the congregation by barriers of language and spatial design. For example, participation in the offertory chant, offertorium or offerenda, and oblation prayers that originally accompanied the procession and offering of gifts had been turned into an extended chant through the priest’s actions or replaced entirely by silence by the fourteenth century. 22 On the other hand, we see an increased individualisation of prayer and offering practices emerging during the same period, especially those related to sacrificial acts at the side altars.

Changing church interiors In some places in Scandinavia, archaeologists have discovered several generations of sacred locales under existing medieval churches. Traces of the oldest manifestations often date back to the eleventh century but rarely beyond 1050 AD. In many cases, cultic continuity can be traced back to pagan times and pre-Christian sacred sites. 23 The location of the church must be regarded as a considerable feature of the building itself, as locations of sacred sites have been recycled, reimagined, and reused. However, turning inward to the interior of the church may provide insights into similar processes of envisioning the sacred setting. The majority of church buildings in Europe have constantly changed over the centuries. As Martin Wangsgaard Jürgensen writes in this volume: A church could never be perceived as a finished monument, since the labour of enlarging and embellishing it held a spiritual reward in itself. Such undertakings

Money, religious devotion  11 might very concretely be rewarded with indulgences, which even pertained to small, rural parish churches, as it very often appears from late medieval papal letters, and while patrons and churchwardens possibly envisioned a certain style or type of church at the outset, the building process itself was in principle never-ending. A further step would always be possible and worthwhile. If the nave was extended, a tower could be built. If the church had one porch, another could be added. Chapels might be built or a plan to rebuild the entire church was perhaps launched. Thus we regularly find evidence supporting the alteration of plans for large-scale building projects midway through or immediately after its completion. 24 The church and the nave were collectively a gateway to eternal life in the Kingdom of God. When people went to church, the purpose was to meet the expectations of participation in Christianity: that part about being a ‘good’ Christian. Against this background, the church building itself became central and a focal point in the landscape for people in both small and large societies throughout Christendom. The relationship between people and church was based on reciprocity on several levels. Where the Church was the way to salvation, the medieval parishioners were committed to the preservation of church buildings, maintenance of roads and bridges, and to the support of the priest and his running of the parish, but perhaps most importantly, parishioners also had to meet at the church and attend fairs and church holidays, which in total could constitute a third of the days of the year. While we know a great deal about the building structures of churches, we often know very little about their interiors through the centuries. A good example is the stave church at Høre in Oppland County, Norway.25 There, apart from the richly decorated portals and corner posts, none of the original medieval interior has survived. Besides the possible main altar base, there are no traces of side altars, baptismal font, alms box, or pews. Thus, our knowledge of how the devotional and liturgical practices were organised at Høre is made on general assumptions rather than specific evidence. That said, in extraordinary cases, we have tantalising descriptions of how individual patrons affected change in the interiors of some churches. Church records commonly refer to patrons’ investments in church decoration and décor. Some patrons operated on the principle that investment in church decoration was a sacrifice in the name of God, and the process would lead them to salvation. Abbot Suger, who rebuilt the church at St. Denis outside Paris in the tenth century, argued that the outward splendour of the finished product reflected his own inner purity and that this purity was a result of the labours that he had selflessly endured on behalf of God. Similar views are expressed in the Norwegian collection of homilies from the twelfth century: Just as we see that the church is an image of all Christendom, we can also say that it is an image of each individual Christian, who by living in purity makes himself a pure temple for the Holy Ghost. For every person shall erect a spiritual church within themselves, not of wood or stone, but with good deeds. 26 In some cases, we can infer the locations of interior features from constellations of coin finds. For example, a coin concentration to the right of the inside entrance at the south portal of the Gothic church at Bunge on Gotland probably marks the location of a collection box from the late thirteenth century onwards. The site is, moreover,

12  Svein H. Gullbekk et al. the most cobbled square metre in the entire archaeologically surveyed section of the church. 27 Such coin concentrations at entrances are known from several other churches in the project, e.g., the stave churches at Ringebu and Høre in Norway. 28 The location of collection devices in entrance areas may be connected to the standing church building, the fabrica, and may have been chosen strategically to collect funds for the maintenance of the church. Indeed, episcopal statues from St Olofsholm on Gotland 1360 decrees that one-third of the money collected in wooden trunks should be spent on the upkeep of the church building. 29 Surprisingly, the number of coins that have been positively identified in graves within churches is relatively few but methodologically important. 30 One pertinent exception is the hoard containing some 600 coins from the late twelfth century found within a burial in the south-western part of Nidaros cathedral in Trondheim. The grave was, however, located outside the church at the time of burial, and its archaeological context within the cathedral is, therefore, of secondary importance.31 In some rare cases, coins have been deliberately placed at the bottom of post holes under stave churches, within altar frontals, or even in wooden sculptures, like the fourteenthcentury Norwegian penny that was placed in the hand of St. Olav in Reinli stave church in Buskerud or the fourteenth-century bracteate found in a wooden cross from Borre church in Vestfold.32 The use of money for devotional practice took on a broad range of expressions. Still, ordinary offerings presented during Mass would have been by far the most common and repeatedly made money-offerings a part of Christian devotional practice everywhere in medieval society.

On the function of money The classic definition of money stresses its function as a means of exchange, a standard of value and a means to store wealth. Money, it can be argued at a general level, is any object that is generally accepted as payment by sellers of goods and services and by purchasers.33 The general acceptability of different kinds of money is established when a large proportion of the community accepts its existence in particular forms. Part of the universal quality of money derives from this interactivity between supplier and demander: more than its intrinsic value, it is the notion of general acceptability that makes money a social convention of standardised universal values.34 Among historians studying both early modern and medieval history from economic, political, and social perspectives, this impersonal medium of exchange, which even a stranger is willing to receive in exchange for goods or services they render unto others, is widely acknowledged to constitute the vital fluid of the urban and commercial organism that developed in the eleventh century, and sometimes in the eighth, ninth, and tenth centuries, depending on what region and society one looks at.35 The relationship between religion and money is less obvious but much more intriguing. It has often been put forward in historical works that the quickening circulation of monetary currency not only sharpened the need for money itself amongst nobles, clerks, and clergy but also catalysed the necessity of its use among the common population. Villagers were inducted into the use of money because they needed it to acquire resources like livestock and the materials necessary to improve methods of cultivation. Money also swiftly became necessary to achieve juridical freedom should the need arise and to buy the articles, tools, and clothing manufactured by artisans in the market towns or hamlets. These processes of commercialisation and monetisation are

Money, religious devotion  13 portrayed as fundamental driving forces for the formation of medieval societies and states from 950 to 1300, a period that in Scandinavia also encompasses the transition from the Viking Age to the Middle Ages. In his book Reason and Society in the Middle Ages, which focus primarily on the eleventh century, historian Alexander Murray writes: The entry of money into the European economy was such a radical change in the environment. It brought in new possibilities; and with them new goals, activities, and habits: so many of all of these, in fact that historians cannot be sure of having spotted them all. There were obvious novelties like a growth in trade and towns. But there were many less obvious: wage-labour for instance.36 The monetary historian Peter Spufford has pointed out that “the availability of adequate and regular money incomes allowed for a revolution in the government of states.”37 In a similar way, one could say that, on the one hand, the development of Church reform was dependent on the existence of coinage and money. On the other hand, the adoption of money as a primary vehicle of exchange meant that the Church became a driving force for monetisation in many rural regions within a network of urban societies in Northern Europe. In Alain de Lille’s (c. 1120–1202) work De planctu naturae, the poet’s heavenly visitor laments the state of the world where “cash conquers, cash rules, cash gives orders to all.”38 What did this mean in a society that was fundamentally Christian? Was money more important than religion, more powerful than Christ? After Pope Gregory VII’s Church reforms of the 1070s, the Societas Christiana became increasingly oriented towards financial matters. At the same time, important members of the ecclesiastic elite proclaimed serious concerns about the role of money and economic gain within theology. This duality in norms and practice says much about attitudes to money and the complexity with which the Church balanced the strange relationship between the earthly and the spiritual. Money played a major role in the development of units of value in complex but interchangeable systems and in the deposit of coins to store wealth, e.g., in hoards, whether within buildings in secular and ecclesiastical contexts, within the boundaries of family estates, or outside the home in geographical space controlled by kin, such as hidden in places with a specific meaning or just places noticeable in the landscape. A progressive expansion of monetary affairs occurred with various underlying effects on the growth of the monetary economy in society. As observed by the numismatist Michael Metcalf: “Year by year the changes were often imperceptible, but over two hundred years, their cumulative effect was great.”39 Again, it is important to note that in this process of monetisation, money was certainly not confined to currency: sophisticated monetary systems could exist without coin.40 However, coinage retained a pre-eminent hold on the high medieval – as it does on the modern – conception of money. Even though money could easily be demonised under the founding tenets of Christianity, and often was in, for example, scholastic writing, the Church quickly learned the value of playing along with money. For instance, the making of offerings became fixed upon the parishioner by custom and Christian teaching and was invested in the doctrine and liturgy of the medieval Church. A parishioner’s quest for salvation became a fundamental reason for going to church, attending Mass, receiving

14  Svein H. Gullbekk et al. communion, and making offerings. As coinage became more widely disseminated, such offerings became increasingly monetised. As a result, the Church played an integral part in developing the widespread use of coinage in, for example, medieval Scandinavia through its organisation of taxation and the prolific monetary obligations that forced churchgoers to acquire cash in order to secure their wellbeing in the afterlife. As the Oxford historian Richard Southern points out: To die with a penance incomplete or without having made provisions for its completion was of all things on earth the most to be dreaded. No man could be saved until his debt was paid. A great man could either pay the stipulated sum or engage other men to undertake the penance for him. In the impersonal society of the early Middle Ages one man’s penance was as good as another’s. It was not a question of individual effort, but the payment of a supernatural debt. No matter how large the debt, it would in the end be paid, and the soul of the sinner would be free.41 Money made it possible for the Church to, whether purposely or not, put a price on salvation, and coinage quickly became the preferred means through which to meet that debt.

Coins as metaphors in medieval Christian thinking Since the earliest days of Christianity, coins have played an important role both in practical use for transmitting values, as symbols of morally high deeds as well as greed and other acts of sin, and purely in metaphorically ways for negotiating values between man and the divine. For example, the betrayal of Christ by Judas for 30 pieces of silver was the subject of discourse throughout medieval and early-modern Christendom. In the Gospels, Judas is portrayed as both a traitor and the Apostles’ treasurer; as such, he was always a person related to money. His betrayal is presented as an act of greed for money: Then one of the Twelve who was called Judas Iscariot went to the chief priests and said, ‘What do you wish to give me, if I hand him over to you?’ And they paid him thirty pieces of silver. And from then on, he was seeking an opportunity to hand him over.42 The betrayal of Christ provides a strong case for the evils of money. The story about how Jesus was betrayed stands at the heart of Christian civilisation where questions of loyalty, trust, and betrayal is universally symbolised in this one act. The betrayal of Christ provides a strong case for the evils of money in general and in particular. At the same time, the Gospels provide evidence for perceptions of money in straightforward pragmatic terms. When Jesus was asked whether it was correct to pay tax he answered, “Show me the coin used for paying the tax.” They brought him a denarius, and he asked them, “Whose image is this? And whose inscription?” “Caesar’s,” they replied. Then he said to them, “So give back to Caesar what is Caesar’s, and to God what is God’s.”43 The message provided is that money belongs to this world, it has the image of the emperor,

Money, religious devotion  15 and must follow the rules that exist in the world. The soul, on the other hand, is created in the image of God and belongs to God. Meanwhile, the story of the “widow’s mite” was already of central importance in the Early Church, offering as it did an example of the practice of ritual alms-giving and of the value of offering small change in particular. As the Gospel of Mark made clear, not only was her offering acceptable in the eyes of Jesus Christ but it was also considered far more valuable than the coins offered by the rich.44 And he sat down opposite the treasury and watched the people putting money into the offering box. Many rich people put in large sums. And a poor widow came and put in two small copper coins, which make a penny. And he called his disciples to him and said to them, “Truly, I say to you, this poor widow has put in more than all those who are contributing to the offering box. For they all contributed out of their abundance, but she out of her poverty has put in everything she had, all she had to live on.”45 In medieval interpretations of the parable of the poor widow and her contribution to the temple coffers, small was beautiful in the eyes of Christ. It was a rationale that was universally adopted and reflected in the fact that the coins found under the church floors in churches are not large-value denominations but, in general, the smallest denominations in circulation. When people wanted to make significant contributions to the Church, they made donations in land, rents, husbandry animals, artworks, and larger sums in cash. The earliest author to explore the close connection between the physical aspects of coins and their spiritual meaning was the early ascetic John Cassian. Cassian (c. 360–435) taught that believers must constantly examine their thoughts and feelings in order to ascertain their own true nature. To Cassian, accordingly in a sense, every Christian needed to become a spiritual moneychanger, constantly scrutinising their heart in the same way a secular moneychanger scrutinised the value of coins in the marketplace. As the moneychanger makes sure that a coin (nomisma) is made of purest gold (aurum purissimum) and is not a common brass denarius and verifies the image on the coin and assesses whether its weight is proper, so must the Christian assess and weigh their thoughts and deeds, rejecting those that are not worthy just as the moneychanger rejects those coins that do not weigh.46 Similarly, Pope Gregory the Great (r. 590–604) used the coin as a metaphor in a series of analogies where Christians were urged to weigh the quality of preachers in their hearts, just as money changers test the quality of gold coins. To paraphrase, one should test the intentions of a man like a moneychanger tests the metal content of a coin and observe if a man’s character matches how he presents himself like a moneychanger observes the design of a coin. One should test the quality of a man’s deeds in the same way a moneychanger tests the weight of a coin.47 Five hundred years later, around 1100, Anselm, Archbishop of Canterbury (1093– 1109), used a sermon “On the similarity of the monk and the coin” to discuss the concepts of money at a spiritual and material level. The sermon made a simple analogy between the life of a monk and the qualities of a coin. Through a series of comparisons, it set out examples of moral behaviours: commendable, less commendable, and uncommendable. In a similar frame to Cassian, Anselm’s concern was with the quality of the metal, the image on the coin, and its correct weight as proxies for an

16  Svein H. Gullbekk et al. individual’s qualities. It also mattered whether the coin was struck by the right authority and if it was a forgery. What becomes evident in these examples is that the authors had an in-depth knowledge of coinage and the nature of coins. Indeed, as Archbishop and Abbot of Canterbury, Anselm was responsible for minting operations, at least on an administrative level, and was responsible for large-scale building programmes that involved huge costs. Yet, in spite of Anselm’s involvement with monetary affairs, money did not represent for him a particular category for abstract thought; money does not seem to have been a subject on which he dwelt in any substantive manner, either from a theological or a moral perspective. Anselm’s use of coins as a metaphor in sermons was not only the result of long cogitation on the morality of using money but also a convenient device to his audience, familiar as it was with using coins in day-to-day life. Clearly, this part of Anselm’s sermon repertoire was simply written for a society in which coin was a standard feature.48 Evidently, there was a longstanding Christian tradition which saw the work of money-changers as comparable to the weighing of souls.49 In the eleventh century, Wulfstan, Archbishop of York, wrote: Then let each man consider how much the soul will be tormented on Doomsday when the sins and the soul will be placed on the scales and weighed as gold is weighed against pennies; and if the pennies weigh more than the gold, it soon turns out badly for the man. So will it be with the soul and the sin; if the sin weighs more than the soul, they will both go to destruction.50 The process of separating the good from the bad was at the very core of a moneychanger’s skillset, just as the process of doing good or bad was at the heart of attaining salvation. In this volume, we see how the weighing of souls was a popular theme which reminded the parishioners of judgement day in rural churches on Gotland, and elsewhere.51 While most of these examples pre-date the dissemination of Christianity in Scandinavia, each illustrates the implicit connection in the early Christian mindset between money as a metaphor and the economics of salvation.

Evidence of offerings in early Christian contexts in Scandinavia In Northern Europe, Christianity was already gaining ground within the borders of the Roman Empire by the fourth and fifth centuries AD; Ireland became Christian in the sixth century and Scandinavia in the tenth and eleventh centuries. In Denmark, the conversion was complete by the mid-tenth century, with early Christian communities having formed in the ninth century. In Norway and Sweden, the conversion was fulfilled by the end of the eleventh century. In contrast, the Baltic tribes were not Christianised until the thirteenth and fourteenth centuries. During the Middle Ages, the presence of churches rapidly expanded and forever changed the landscapes of the European continent, both physically and ideologically. While many regions in Scandinavia became Christianised around the year 1000, archaeological remains suggest that church building did not become widespread until the second half of the eleventh century. There was a dramatic increase in building in the first half of the twelfth century, when technological advances provided the basis for the stave churches and influences from Europe opened up the building of stone churches, especially within town centres. Coin finds from churches are, however,

Money, religious devotion  17 relatively scarce until the last quarter of the twelfth century, even though moneyofferings for oblation and atonement had become established practice shortly after the time of the conversion. In Norway and Scandinavia, the first reference to money-offerings is from the 1050s or 1060s, in a letter from the Archbishop of Hamburg, Adalbert, in which he reproached King Harald Hardrade for tyrannical abuse recorded in the Gesta Hammaburgensis Ecclesiae Pontificum. One of the accusations was that the King had acquired the offering gifts that the faithful had presented to the shrine of Saint Olaf in Nidaros and had used the money to pay the hirdmen in his retinue.52 Looking at the archaeological evidence, it would seem that the coins in question were to a large degree the king’s own debased pennies. In the process of establishing a national currency, King Harald replaced the foreign coins in circulation with his own: pennies that contained one-third silver. These debased pennies were, however, invested with Christian iconography: a symbol of the Holy Trinity on the obverse and a cross on the reverse, with small crosses in both the obverse and reverse legends.53 In the same book, Adam of Bremen describes the Norwegians as treating their priests and churches with great respect and observes that those who do not act accordingly, i.e., by presenting offerings at Mass, are seen as bad Christians. Adam’s reflections present his archbishopric as a successful instigator of Christendom in a society still influenced by pre-Christian traditions. Yet, on the other hand, he also complains that the Norwegian clergy are greedy for money because they charge for the performance of the sacraments.54 This suggests that the clergy had to take their income from moneybased services in the reigns of Harald Hardrade (1047–1066) and his successor Olaf the Peaceful (1067–1093).55 The first mention of alms being paid specifically for repentance in a Scandinavian context was in a letter from Gregory VII to the Danish King Harald IX (r. 1074– 1080) dated 1077. In this letter, the Pope praised Harald’s father, Sven Estridssen, for being “second to almost no other King” and claimed that “the grievous sins he had made during his lifetime would have been purged by his own repentance and memorial prayers and alms.”56 Of course, money wasn’t the only commodity being exchanged through the Church. Sources from later medieval Scandinavia describe how people, usually peasants, offered gifts in kind. These would be placed at a side-altar, usually the altar of Mother Mary. These side-altars were often called “cake-altars” or “cheese-altars”, reflecting the types of gifts that were frequently offered.57 The practice of offering foodstuffs at sacred places goes back to the pre-Christian period. Since such offerings were largely perishable, there remains little empirically to suggest that offerings of this kind were less, more, or as common as those of a monetary nature once money entered the scene.

Coins and church archaeology Church archaeology has been an important field within European and Scandinavian archaeology since the eighteenth century, but the focus on coin finds from churches has sharpened since the 1950s as a consequence of numerous excavations. 58 The first to make such finds the primary basis of a grand study on the religious, economic, and social uses of money was the Swedish archaeologist Henrik Klackenberg, in his seminal study Moneta nostra (1992). 59 His work focused on medieval monetisation in rural Sweden c. 1150–1400 AD and has influenced the current generation

18  Svein H. Gullbekk et al. of Scandinavian archaeologists and numismatists, forming the basis for all studies of coin finds in churches, including the works here. Since the 1960s, scholars have debated the reasons behind the rich finds of medieval coins beneath the floors of Scandinavian churches. Noting the many coins that came to light during excavations in Danish churches in the 1950s and 1960s, the renowned Danish historian and archaeologist Olaf Heymann Olsen has suggested that the coins found beneath church floors were the result of casual loss, i.e. the members of the congregation dropped petty change while handling coins for offerings at the altar(s) and offertory trunks.60 Olsen’s view was soon adopted and further elaborated in numerous publications over the next two decades by Danish and Swedish scholars and the leading Norwegian numismatist Kolbjørn Skaare.61 However, a series of articles published by Norwegian archaeologists and historians in the 1980s and 1990s argued against the theory of random coin loss and, to various degrees, in favour of deliberate acts of offering.62 Today, the total body of research and empirical data from excavations of medieval churches throughout Scandinavia clearly shows that the causes of the accumulation of coins under the church floors were complex. There are instances of single coins having been deliberately put in cracks in altars and crucifixes and coins being part of burials and other acts of offering alongside other ritual objects. Still, casual loss during the handling of coins for offering at altars, in offertory trunks, and on collection plates or for the payment of religious services is the reason behind the majority of coins found beneath church floors.63 There seems to be a scholarly consensus that considers the coin finds to be the product of a general phenomenon of coin use in liturgical contexts in all church buildings throughout Scandinavia, which makes comparisons between different regions and countries and across time possible. This is the fundamental

Figure 1.5 From the excavation of Ringebu stave church in 1981/82. Museum of Cultural History, University of Oslo, unknown photographer.

Money, religious devotion  19 principle on which the present work is based, and we leave the loss versus offering debate behind. There have been a prolific number of finds of single coins from medieval Scandinavian churches dating from the twelfth century onwards. In addition to the Scandinavian finds, Swiss churches have produced significant numbers, respectively more than 100,000 and 20,000 (add to this a dozen or so from Iceland, but none from Greenland). As such, Scandinavian and Swiss churches provide a richer body of archaeological remains from the Middle Ages than anywhere else within the Latin West. In the British Isles, France, the Low Countries, Germany, and Italy, finds like these are few and far between in general.64 In Central Europe, coin finds from churches constitute a rich and important body of material, especially in the Alpine regions such as Switzerland and Austria, parts of France, Germany, and Poland, where significant numbers of coin finds from churches have been recorded.65 There are significant differences in finds from churches in different regions. Variations are not only the result of the degree of monetisation, or differences in devotional practice or archaeological methods but arise rather from the building structures adopted in different regions of Europe. For example, as we have noted above, floor types are consequential to the presence or absence of coin finds in churches. In southern Europe, most had stone floors, whereas in northern Europe many medieval churches – both stone and stave churches – had wooden floors. Only the very earliest church buildings had dirt floors, and the most prestigious ones, such as the cathedrals, had stone floors. However, mortar floors were widespread in some periods, and a single church could have several different types of floors, evident through the medieval and early modern periods. As a consequence, the circumstances for archaeological excavations under each church floor are different and potentially quite varied even within a single stratigraphic floor, as is discussed by Henriette Rensbro in this volume.66 Finds from beneath the floors of churches that were furnished with stone are very limited, while the numerous finds from below the floorboard level in churches with wooden floors reflect the activities that took place within the church over an extended period of time. Under these wooden floors, the artefacts number in the hundreds, and sometimes thousands. The radical increase in church excavations has provided an unparalleled possibility to develop church archaeology methods and interpretations of archaeological material from a large sample of churches. The churches with the most profound number of finds are from Gotland, with as many as c. 4,200 coins in Gothem church and c. 3,260 coins in Bunge church.67 In Norway, the highest numbers of coin finds have been produced in stave churches: c. 2,400 in Lom stave church, c. 1,400 in Kaupanger stave church, and c. 1,100 in Ål stave church, all of which are situated in remote areas in the midst of mountains and at the deep end of the Norwegian fjords and must be considered small village churches from a European perspective.68 By comparison, the largest number of coins found in a Danish church is 345 coins from Aggersborg church on Jutland.69 In Sweden, the richest finds come from two churches in the southern province of Småland: Hagby with 940 coins and Åseda with 904 coins. However, despite the seemingly impressive number of coins recovered from church floors, much evidence has been lost due to early church excavation methods. In 1959, Olaf H. Olsen published a famous assessment of Scandinavian church archaeology in which he contended that the careless treatment of Danish church floors by archaeologists had largely destroyed their value as research objects.70 He estimated that prior

20  Svein H. Gullbekk et al. careless excavations had caused a dramatic loss of artefacts from perhaps hundreds of medieval wooden churches, with tens of thousands of old coins and other small items being thrown out with improperly excavated and discarded floor layers.71 On a related note, the publication of the excavation of St. Jørgen’s hospital church in Lund, carried out by Mats P. Petersson [Malmer] in 1948, is widely considered to be a progressive study when it comes to coin finds. In it, Petersson opens up a discussion of coin finds in relation to spatial distribution and its significance for studies of devotional practice.72 However, what remains lacking is an in-depth presentation and discussion of devotional practices and the use of money in religious contexts.73 The methods applied for archaeological finds of artefacts in early church archaeology were in many cases devastating for contextual documentation. In a large number of churches, the soil from beneath the church floors was moved outside the church to be sifted with no concern for its temporal or spatial contexts within the church. For decades, the archaeology of medieval churches was in the care of architects and art historians primarily in search of building structures. Such excavations were concerned with uncovering features in the ground that could reveal older building structures beneath existing ones. They were focused on unravelling the structures of each church from the earliest foundations onwards. In these cases, very few artefacts came with contextual information, and the majority can only be spatially connected to within a specific church building or section. Sadly, a lack of further contextual documentation makes most of the earlier church excavations irrelevant in discussions relating to issues such as liturgy, devotional practice, church interiors, gender, dating, and so on. As a consequence of new awareness of the importance of finds of artefacts beyond the consideration of building structures, archaeology developed better methods. Meticulous investigations by Scandinavian architects and archaeologists from the 1950s onwards have secured important contributions to the history of architecture, ecclesiastical buildings, and floor finds. Within this modern period of Scandinavian church archaeology, methods and theories developed along with archaeology in general and featured a strong focus on diligence and precision in documentation as well as in the overall ways in which archaeological sites and finds are approached from a research perspective. In this volume, scholars revisit these excavations, discussing the methods employed and their qualities, and interpreting their findings in relation to devotional practices in medieval and early modern times.

Parish churches and surveys of coinfinds No institution had as conservative an attitude to change in moral issues as the Church, and none had a greater impact on society. This conservatism was translated into expressions of architecture, liturgy, and continuity of religious practice in rural churches and town churches individually and comparative perspectives. From the cradle to the grave, churches were an important part of individual lives. Generation after generation of parish children attended the same church and followed services which, with certain regional and local variations, were similar in design and liturgy in all parts of Christendom. The Christian message followed broadly the same lines in all parts of the Western world. Thus, the expectations for living a good Christian life changed little between Southern Italy and Northern Norway. This brings us to the churches themselves. At first glance, one may ask why this study does not include churches from England, France, Germany, Italy, or elsewhere

Money, religious devotion  21 in European Christendom. As we have explained above, the answer is simply that churches and cathedrals in these countries were mainly built with stone floors. In Danish churches, stone floors were more widespread than wooden. That is the reason why finds from Danish churches are less prolific than in Norwegian churches. Stone floors only rarely allow archaeological finds below their floor surface. Not much anthropogenic activity goes on beneath a stone-paved flooring besides burials and other specific events during which floors be opened up. However, churches with wooden floors offer completely different archaeological structures to be explored. Spaces between wooden floor planks have, over the centuries, swallowed up an exceptional record of the materials lost between parishioners’ feet. Excavations expose the array of objects that have slipped through the cracks, including everything from burials to pilgrimage marks, building remains, buttons, needles, beads, rope sticks, manuscript fragments, and, of course, coins. Of all of these categories of objects, it is coins that occur most frequently and with the longest continuity and that have the broadest identifiable geographical distribution, by far.74 Being able to observe and identify archaeological finds of single coins within strictly defined ecclesiastical contexts in several different kingdoms, several church provinces, a number of dioceses, and many parishes over a period of seven or eight centuries, sometimes even longer, provides extraordinary comparative evidence. This provides an ample opportunity to study religious practice in a long durée. In the countries under investigation –here – Denmark, Norway, Sweden and Finland, together with Switzerland – we can conclude that the extent of church finds of coins far exceeds that of other complexes with single discoveries of coins from European medieval times. In Scandinavia, more than 100,000 single coin finds have been recorded from some 840 churches and monasteries. Overall c. 11,400 coins, of which c. 5,700 are medieval coins, have been found inside some 370 churches in Denmark.75 In Norway, c. 20,000 coins of which c. 13,000 Table 1.1 Approximated number of coins found in Scandinavian churches. In comparison, almost 22,000 coins are found at about 640 ecclesiastical locations in Switzerland79 Country Denmarka Norwayb Swedenc Finlandd Gotlande Total Scandinaviaf

No. of churches 370 125 200 80 65 840

Medieval coins

Total no. of coins

5,700 13,000 7,650 1,850 31,200 59,300

11,400 20,000 24,500 13,700 31,200 100,800

a This count dates to pre-1994. The number today is higher (Keld Grinder-Hansen, «Mønter som kilde til middelalderens økonomiske historie – en presentasjon av et kildemateriale,» Fortid og nutid 2 (1994): 101–33). b Gullbekk and Sættem, Norske myntfunn, 33. c Klackenberg, Moneta nostra. For an updated survey, see Kenneth Jonsson, “Myntfynden i landsortkyrkor i det medeltida Sverige,” Myntstudier 1 (2011): 1–16; Klackenberg, “Coins in Churches,” 269. d Frida Ehrnsten, Pengar för gemena man ? Det medeltida myntbruket i Finland. Archaeologia Medii Aevi Finlandiae XXVI (Helsingfors, 2019), 296; Klackenberg, “Coins in Churches,” 273. e Susanne Fridh, «Mynt 1150–1699 i lösfynd och hopade fynd på Gotland,” Myntstudier 1 (2014): 1–37. Additional finds of c. 1,200 coins from the 2014 excavations at St Olofsholm church not reported by Fridh are included in our count. f In addition, less than a dozen coins are found in Icelandic churches.

22  Svein H. Gullbekk et al. coins are medieval, from some 125 churches.76 From Sweden mainland, c. 24,500 coins, of which c. 7,650 are medieval coins from c. 200 churches.77 In Finnish areas of Sweden, c. 13,700 coins are found, of which c. 1,850 are medieval coins from some 80 churches.78 On Gotland, the exceptionally high number of c. 31,200 coins have been found in some 65 Gotlandic churches.80 In total, more than c. 100,000 coin finds have been recorded from some 840 churches across Scandinavia. In comparison, Switzerland boasts a total of ecclesiastical excavations around 640 with coin finds which add up to c. 21,600.81

The case studies and analysis Most of the churches discussed here were excavated from the 1950s onwards but before the use of modern spatial techniques such as, for example, total station or GIS was possible. The churches under investigation in this volume were excavated as follows: St. Mary’s Abbey church, Nonneseter monastery, Bergen, Norway (1891), Jomala church, Åland, Finland (1961), Eidskog church, Hedmark, Norway (1965), Bunge church, Gotland (1971–1972), Aggersborg church, Jutland, Denmark (1976), Høre stavechurch, Valdres, Norway (1979), Gränna church, Småland, Sweden (1987), Arby church, Småland, Sweden (1990s), Klåstad church, Östergötland, Sweden (2000), and Hedensted church, Jutland, Denmark (2007–2008).82 The precision with which finds were recorded is, thus, a frequent subject of discussion and interpretation in the

Figure 1.6 Scandinavian church sites investigated in the project. Map: Steinar Kristensen, UiO: Museum of Cultural History.

Money, religious devotion 23 present volume. This noted, in many cases, the archaeologists in charge recorded finds using square metre quadrants (either 1-by-1 m or 2-by-2 m). Thus, soils from each quadrant were sieved and the finds recorded, albeit randomly within each quadrant, allowing for only a general spatial understanding of find locations. Understandably, interpretations are still hampered by such uncertainties.83 The research presented in this volume draws on numerous sources, including, of course, excavated materials from churches in public collections, archival sources, and interviews with the people involved in prior excavations. Thus, the data is diverse in nature, quality, and quantity. As such, churches within the study area were selected on the basis of adequate archaeological documentation, such as interpretable excavation reports, notes, correspondence, etc. In other words, the selected surveys are all based on sufficiently quality of documentation of finds so these records can be analyzed and utilized in geographical and historical contexts. What became clear (and somewhat troubling) from early on in the development of the research agenda was that in many cases excavators did not treat artefact finds as the principal priority. In fairness, one should say that many of these investigations were carried out with rigorous ambitions, but in some cases, the quality of the documentation provided for each individual find is just adequate, at best, by modern standards. Thus, many of the chapters here present an exercise in excavation reconstruction and document interpretation (archival archaeology) and a retranslation of earlier observations based on the record provided by the archaeologists and architects responsible for the original excavations. In spite of the tight definition of the case studies, contributions should not be considered comprehensive. It has also been our goal that each chapter is self-contained yet speaks to other chapters and invites comparisons between sites, specific churches, and their contexts, as well as between the interpretations and insights offered. The interdisciplinary nature of the project which initiated the current volume has allowed for different pathways towards understanding both old and new problems. Overall, several chapters discuss the archaeology of a variety of medieval Scandinavian churches, addressing a wide range of issues related to medieval concepts of devotional practice using the coin finds as starting points and forming more general discussions based on archaeology, history, numismatics, theology, art history, and architecture. Each chapter targets a number of questions with in-depth investigations of specific case studies that have been discussed in plenum with scholars belonging to different research traditions within archaeology, numismatics, history, and theology.

Notes 1 This is the fifth volume to be published under the auspices of this project: Money and the Church in Medieval Europe, 1000–1200. Practice, Morality and Thought, edited by G.E.M. Gasper and S.H. Gullbekk (Farnham: Ashgate, 2015); Divina Moneta: Coins in Religion and Ritual, edited by Nanouschka Myrberg Burström and Gitte Tarnow Ingvardson (New York – Abingdon: Routledge, 2017); Svein H. Gullbekk and Anette Sættem, Norske myntfunn 1050–1319. Penger, kommunikasjon og fromhetskultur (Oslo: Dreyer forlag, 2019); “Coins in European Churches: Religious Practice and Devotional Use of Money,” edited by B. Zäch and S.H. Gullbekk, Schweitzerische Numismatische Rundschau (SNR) 97 (2020): 125–68, special issue 2019. 2 David Grumett, Material Eucharist (Oxford: Oxford University Press, 2016).

24 Svein H. Gullbekk et al. 3 “Nu med tvi goder brødr at vér sculum her vera litla stund i tessum hæimi [---]. En himinriki er sva got at engi madr kan tat hyggja eda odrum sægia,” from the “Sermo necessaria,” cf. Gamal norsk homiliebok, ed. G. Indrebø (Oslo: Universitetsforlaget, 1931), 88. English translation: George T. Flom, Codex AM 619 Quarto: Old Norwegian Book of Homilies Containing the miracles of Saint Olaf, and Alcuin’s De virtutibus et Vitiis (Urbana: University of Illinois, 1929). 4 Jaques Le Goff, The Birth of the Purgatory (Chicago: University of Chicago Press, 1984). 5 Olav Tveito, “Rituell reformasjon. Fra ordo sepulturae til luthersk gravferdsskikk,” Teologisk Tidsskrift 1 (2017): 25–41. 6 Carl-Gustaf Andrén, «De medeltida avlatsbreven – instrument för kyrkans verksamhet, Kungs-Husby i Trögd. Kungsgård, kyrka och socken,» Studier till det medeltida Sverige 6 (1992): 89–104. Myrberg Burström, this volume, ch. 11. 7 See also contributions by Wangsgaard and Myrberg Burström, this volume, chs. 2 and 11. 8 Accessed 2020-09-11 https://www.online-latin-dictionary.com/latin-english-dictionary. php?parola=obferere 9 “§. 1. þæsse höghtiþis dagha aghu bönder oc husfrur sit offer [oc] til altara bæræ. penning bonde annæn husfru. Juladagh. Kyndil messu dagh. Pascha dagh. Pingizdagha dagh. Waræ frua dagh. þen fyrræ. Helghuna messu dagh. Kirkiu messu dagh. oc swa husfrur altara byrþir sinæ [---].” From Codex iuris Sudermannici (c. 1330), the law of the Swedish medieval province Södermanland, H. S. Collin & C. J. Schlyter (eds.), Sammling av Sveriges gamla lagar, vol. 4 (Stockholm and Lund, 1838). 10 Kilger and Myrberg Burström, this volume, chs. 5 and 11. 11 Anders Fröjmark, “De missförstodda avlaten,” in Religionsdidaktiska studier, edited by Torsten Löfstedt (Växsjö: Linneaus University Press, 2015), 75–87. 12 Jürgen von Hagen and Michael Welker (eds.), Money as God? The Monetization of the Market and Its Impact on Religion, Politics, Law, and Ethics (Cambridge: Cambridge University Press, 2014); Wendy Davies and Paul Foracre (eds.), The Languages of Gift in the Early Middle Ages (Cambridge: Cambridge University Press, 2010); Joel Kaye, Economy and Nature in the Fourteenth Century. Money, Market Exchange, and the Emergence of Scientific Thought (Cambridge: Cambridge University Press, 1988). 13 Roberta Gilchrist, Gender and Material Culture: The Archaeology of Religious Women (London and New York: Routledge, 1994). 14 Kilger, this volume, ch. 5. 15 Martin Wangsgaard Jürgensen, “Syddør, norddør òg det kønsopdelte kirkerum,” Kirkearkæologi i Norden. Hikuin 9 (2007): 7–28; Svein H. Gullbekk, “Medieval Scandinavian Women in Search of Salvation,” in Divina Moneta: Coins in Religion and Ritual, edited by Nanouschka Myrberg Burström and Gitte Tarnow Ingvardson (New York – Oxford: Routledge, 2017), 209–27. 16 Pietro Fedele, Il giublieo del 1300, Gli Anni Santi (Torino: Società Editrice Internazionale, 1934), 7–25; Lucia Travaini, “Saints and Sinners: Coins in Medieval Italian Graves,” Numismatic Chronicle 164 (2004): 174–75. 17 Vim Vroom, Financing Cathedral Building in the Middle Ages: The Generosity of the Faithful (Amsterdam: Amsterdam University Press, 2010). For Sweden see Henrik Klackenberg, Moneta nostra. Monetarisering i medeltidens Sverige (Stockholm: Almqvist & Wiksell International, 1992), 36–37. Not surprisingly, the most systematic study of shrines and offerings has been carried out in England, where the medieval records from the Middle Ages often are exceptional, see Ben Nilson, Cathedral Shrines of Medieval England (Woodbridge: Boydell & Brewer, 1998). 18 Allan Doig, Liturgy and Architecture from the Early Church to the Middle Ages (Ashgate: Farnham 2008), xxi. 19 Jeanne Halgren Kilde, Sacred Power, Sacred Space. An Introduction to Christian Architecture and Worship (Oxford: Oxford University Press, 2008), 7. 20 Paul Binski, “The Cruxifiction and the Censorship of Art around 1300,” in The Medieval World, edited by P. Linehan and J. L. Nelson (London – New York: Routledge, 2001), 342–60. 21 Andrew Hughes, Medieval Manuscripts for Mass and Office. A Guide to their Organization and Terminology (Toronto: University of Toronto Press, 1995), 3.

Money, religious devotion 25 22 Hughes, Medieval Manuscripts, 87. 23 John McNicol, Plasseringen av de første kirkene i Norge i forhold til de hedenske kultstedene: en histografisk studie omfattende tiden etter 1830 (Oslo, 1997). For a different opinion, see Olaf Olsen, Hørg, Hov og Kirke. Historiske og arkeologiske vikingtidsstudier (København: G.E.C. Gad, 1966). 24 Wangsgaard Jürgensen, this volume, ch. 2. 25 Risvaag, this volume, ch. 6. 26 “En sva sem vér sægium kirkio merkia allan cristin lyd, sva man hon merkia sérhværn cristin mann, tann er samlega geresc mystere hæilags anda í gódum sidum. Dvi at hvær madr scal smida andlega kirkio í sér, æigi ór triom ne stæinum, hældr or godum vercum,” from “In dedicatione templi sermo,” G. Indrebø (ed.), Gamal norsk homiliebok, 102. English translation: Flom, Codex AM 619 Quarto. 27 Kilger, this volume, ch. 5. 28 Terje Hellan. “A Stave Church Revisited. Ringebu Stave Church, Ringebu Municipality, Oppland County,” Schweizerische Numismatische Rundschau 97, 2019 (2020): 219–43; Risvaag, this volume, ch. 6. 29 “Pecunia vero in truncis proueniens in tres partes equales diuidatur. Quarum unam pro fabrica ipsius cappelle deputamus …” DS 6324. 30 Rensbro and Moesgaard. Myrberg Burström and Jonsson this volume, chs. 9, 11 and 13. 31 Gullbekk and Sættem, “Norske myntfunn,” no. 294. 32 A Norwegian Håkon VI bracteate, found during cleaning of the so-called Borre cross from Borre church in Vestfold county, today in the collection of the Museum of Cultural History, University of Oslo. The coin was put in a wooden crack at the base of the cross. UMK find protocol 1043 (8/12-1966). 33 Fadhel Kaboub, “Money,” in Encyclopedia of World Trade from Ancient Times to the Present, vol. 3, edited by Cynthia Clark Northrup (New York: ABC Clio, 2005), 670. 34 The nature of what becomes designated as money in a given society is as culturally constrained as conventions deciding standards of time. The importance lies in the fact that something is chosen as money, not the particularity of the choice. Within world history, money takes on a wide variety of forms, from the huge stone money from the island of Yap in Oceania, to shells, iron and bronze in other Asian societies, and within Europe at various times tea, pepper, hides, corn, livestock, butter, silver, gold and coins. See, James Tobin, “Money,” in The New Palgrave Dictionary of Money & Finance (London: The Macmillan Press Ltd, 1992), 770–78. 35 Alexander Murray, Reason and Society in the Middle Ages (Oxford: Clarendon Press, 1978); Lester K. Little, Religious Poverty and the Profit Economy (Ithaca: Cornell University Press, 1978); Robert S. Lopez, The Commercial Revolution of the Middle Ages, 950–1350 (Cambridge: Cambridge University Press, 1976, 2 ed.); Peter Spufford, Money and Its Use in the Middle Ages (Cambridge: Cambridge University Press, 1988). 36 Murray, Reason and Society, 59. 37 Spufford, Money and Its Use, 247. 38 Odd Langholm, “A Herald of Scholasticism: Alain of Lille on Economic Virtue,” in Gasper and Gullbekk (eds.), Money and the Church, 93–104, att. 96. 39 Michael Metcalf, Coinage in South-Eastern Europe 820–1396 (London, 1979), 18. 40 Two standard studies of the issue, with an emphasis of commodity-money and natural economy are A. Steinnes, “Mål, vekt og verderekning i Noreg i millomalderen og ei tid etter,” in Mål og Vekt, edited by Sven Aakjær (Oslo-Stockholm-København, 1936), 84–154 and K. Lunden, Korn og kaup. Studiar over prisar og jordbruk på Vestlandet i mellomalderen (Oslo-Bergen-Tromsø, 1978). For a survey of prices in Norwegian medieval sources and further advocacy of natural economy, see Gunnar Pettersen, Priser og verdiforhold i Norge ca. 1280–1500. Riksarkivet Skriftserie (Oslo, 2013). The issue of natural economy or money economy has been the subject of debate for Norway, especially between Kåre Lunden and Svein H. Gullbekk. See contributions in English: S. H. Gullbekk, “Medieval Law and Money in Norway,” Numismatic Chronicle 158 (1998): 173–84; K. Lunden, “Money Economy in Medieval Norway,” Scandinavian Journal of History 24 (1999): 245–65; S. H. Gullbekk, “Natural or Money Economy in Medieval Norway,” Scandinavian Journal of History 30 (2005): 3–19.

26

Svein H. Gullbekk et al.

41 Richard W. Southern, Western Society and the Church in the Middle Ages (London: Penguin, 1990), 227; Lucia Travaini, I Trenta denari di Giuda. Storia di reliquie impreviste nell’Europa medievale e moderna (Rome: Viella 2020). 42 Matthew 26: 14–16. 43 Matt 22: 20–22. Indeed, the disciple Peter admits that Jesus himself paid tax, as recounted in Matt 17: 24–27. 44 Travaini, “Saints and Sinners,” 173; David Ganz, “Giving to God in the Mass,” in The Languages of Gift in the Early Middle Ages, edited by Wendy Davies and Paul Foracre (Cambridge, 2010), 18–32. 45 Mark 12: 41–44, English Standard Version (ESV). Mark 12: 42: Greek two lepta, which make a kodrantes; a kodrantes (Latin quadrans) was a Roman copper coin worth about 1/64 of a denarius (which was a day’s wage for a labourer). A copper lepta in the Jerusalem marketplace would have been considered small change or petty coinage. For a medieval Scandinavian source from the 1430s that reflects on money-offerings and mirrors the Biblical parable of the widow’s mite, see Sven Erik Pernler, “En mässa för folket?” in Mässa i medeltida socken. En studiebok, edited by Sven Helander, Sven-Erik Pernler, Anders Piltz and Bengt Stolt (Skellefteå, 1993), 102–34, att. 125–26; Margareta Kempff Östlind, Kyrkorummet som social arena. Det sociala rummet. Bebyggelsehistorisk tidskrift 46 (2004): 35. 46 Greti Dinkova-Brun, “Nummus falsus: The Perception of Counterfeit Money in the 11th and Early 12th Century,” in Giles E.M. Gasper and Svein H. Gullbekk (eds.), Money and the Church, 77–92; Rory Naismith, “Denarii mixtii: Debasement and Rhetoric in the Early Middle Ages (Fifth–twelfth Centuries),” in Debasements: Manipulation of Coinstandards in Pre-modern Monetary Systems, edited by Kevin Butcher (Oxford: Oxbow, 2020), 195–207. 47 St. Gregory the Great, Morals on the Book of Job, vol. 3, pt. 2 (Oxford, 1850), 611. 48 Giles E. M. Gasper and Svein H. Gullbekk, Money and Its Use in the Thought and Experience of Anselm, Archbishop of Canterbury (1093–1109), “Journal of Medieval History” 38 (2012): 155–82. Thanks to Kay Celtel for her advice. 49 Jane Welch Williams, Bread, Wine and Money. The Windows of the Trades at Chartres Cathedral (Chicago and London: University of Chicago Press), 137. 50 Dorothy Whitelock, “The Numismatic Interest of an Old English Version of the Legend of the Seven Sleepers,” in Anglo-Saxon Coins, edited by Michael Dolley (London, 1961), 193. 51 Kilger, this volume, ch. 5, fig. 2. In contrast the weighing is supervised not by a moneychanger but by the Archangel Michael. The pictures also vividly portray the fight for the soul of the doomed by the devils and demons hanging on the scales. See also Ingvardsson this volume, ch.10, fig.3. 52 The first record of money-offerings in Scandinavia occured in Adam of Bremen’s Gesta Hammaburgensis Ecclesiae Pontificum (Deeds of the Bishops of the Hamburg Church), written down in the years between 1073 and 1076, liber III, ch. 17: “Videbac haec ille derelictus a Deo, nihilque compunctus oblationes quoque ac tesauros, qui summa fidelium devotione collatisunt ad tumulum fratris, ipse Haroldus unca manu corrodens militibus dispersit” (Adam av Bremen, Historia Norvegiae, edited and trans. by Halvdan Koht, Gamalnorske bokverk IV (Kristiania [Oslo]: Samlaget, 1921), 79. 53 For a discussion of the coin design on coinage in the name of Harald Hardrade, see Kolbjørn Skaare, Coins and Coinage in Viking Age Norway (Oslo: Universitetsforlaget, 1976), 68–70. 54 “Preterea sacerdotum et ecclesiarum tantam habent venerationem, ut vix christianus habeatur, qui non cotidie obtuleritad missam, quam audierit. Verum baptismus et confirmatio, dedicationes altarium et sacrorum benedictio ordinum apud illos et Danos care omnium redimuntur. Quod ex avaritia sacerdotum prodisse arbitror; quia barbari decimas adhucdare aut nesciunt aut nolunt, ideo contringuntur in ceteris, qua deberent gratis offeri” (Adam of Bremen, Gesta Hammaburgensis, liber IV, ch. 31). 55 In the early phase of Christianity, the practice of charging for clerical services was first expressed in the Kristendomsbolken in the Gulathing Law, where the normative version states that “um nokon vert sjuk og vil han senda etter presten, då skal han senda etter den presten som han kjøper messer hjå.” The priest could charge 2 øre for holy oil, 1 ½ øre for

Money, religious devotion  27

56 57 58 59 60 61 62

63

64

65 66 67 68 69 70 71 72 73

74

hymns, and an unspecified sum for the Requiem Mass. Priests were forbidden to receive money for services in districts of other priests: “fara i sokni åt ein annan prest for å tena pengar” (Gulatingslovi § 23.15). This suggest that the possibility to make income from services was strictly limited to the parish or parishes that a priest was responsible for. For a discussion of Pope Gregory VII and his relations to Scandinavia, see E. John Cowdrey, Pope Gregory VII, 1073–1085 (Oxford: Clarendon Press 1998), 454–59. Bengt Stolt, “Kyrkorum och kyrkoskrud,” in Mässa i medeltida socken, edited by Sven Helander et al. (Skellefteå, 1993), 161–62. Olaf Olsen, «Kirkegulvet som arkaeologisk arbejdsmark,» Nationalmuseets Arbejdsmark (1958): 17–30. Klackenberg, Moneta nostra presents the discussion on coin-finds and church archaeology in Sweden and Scandinavian context. See also Rensbro and Moeasgaard, this volume, ch. 9. Olaf H. Olsen, “Kirkegulvet …”; “St:Jørgensbjerg kirke.”‘ Aarbøger for Nordisk Oldkyndighed og Historie (1960): 1–71. Kolbjørn Skaare, “Myntfunn fra kirker – hvem er eier?” Museumsnytt 3–4 (1964): 40–42; K. Skaare, “Om myntfunnene i Bø gml. Kirke,” Telemark historie 7 (1986): 28–38; K. Skaare, “Myntene under kirkegolvet,” Uvdal stavkirke forteller (Uvdal, 1992), 62–69. Herleik Baklid, “Hvad der har bragt dem gjennem gulvet er desverre efter al sansynlighet en hemmelig ofring [---],” Heimen, 32 (1995): 181–96. Karin Berg, “Coins in Churches: A Means of Payment? Part 1,” in: Coins and Archaeology. BAR International Series 556, edited by H. Clarke and E. Schia (1989), 77–82. Anne-Marie Mørch von der Fehr, “‘Løsfunn’ fra Høre kirke,” Universitetets Oldsaksamling Årbok 1984/85 (1986): 135–43; “Myntkonsentrasjoner i Ringebu stavkirke,” Universitetets Oldsaksamling Årbok 1986/88 (1989): 161–69. Inger Helene Vibe Müller, “Kirkefunn som arkeologisk og kulturhistorisk kildemateriale,” Universitetets Oldsaksamling Årbok 1982/83 (1984): 183–98; “Coins in churches: A means of payment? Part 2,” in Coins and Archaeology. BAR International Series 556, edited by H. Clarke and E. Schia (1989), 83–89; “From the Battle for Power to the Battle of the Souls: The Basis for Parochial Division and Its Consequences,” Collegium medievale 3 (1992): 137–44. Henrik Klackenberg introduced the term ‘offertory wastage’ in his standard work on Swedish medieval churches (Klackenberg, Moneta nostra, 35–40). A recent study of medieval collection plates from Gotland has made it clear that they were in use everywhere on the island by the thirteenth century, Pia Bengtsson Melin and Kenneth Jonsson, “Medeltida kollekttavlor från Gotland,” Myntstudier 2 (2019): 1–92. “It is true that secular churches often yield little in the way of conventional archaeological data (pottery, coins, small-finds etc.),” Richard Morris, The Church in British archaeology, Research reports from the Council for British Archaeology 47 (London: Council for British Archaeology, 1983), 10. Zäch, this volume, ch.14; Daniel Schmutz and Franz E. Koenig, Gespendet, verloren, wiedergefunden. Die Fundmünzen aus der refromierten Kirche Steffisburg, vol. II (BernStuttgart-Wien: Haupt Verlag AG, 2003), 27. Rensbro, this volume, ch. 3. See Kilger, this volume, ch. 5. Gullbekk and Sættem, Norske myntfunn, nos. 90, 127 and 272. See Rensbro and Moesgaard, this volume, ch. 9. Olsen, “Kirkegulvet.” Jørgen Steen Jensen, “Kirkegulvsmønter,” Hikuin 3 (1977): 296. Olsen, “Kirkegulvet,” 18. Mats Petersson [Mats Malmer], “S:t Jørgen i Åhus,” Medelanden från Lunds Historiska Museum (1948): 191–257. For surveys of research on coin finds in churches, see Klackenberg, Moneta nostra, 34–40; Henrik Klackenberg, “Coins in Churches − 25 Years after Moneta Nostra,” in Coins in European Churches: Religious Practice and Devotional Use of Money, SNR 97, 2019 (2020): 267–81. For finds related to religious sites in Britain, see Martin Allen, “Coins and the Church in Medieval England: Votive and Economic Functions of Money in Religious Contexts,” in Divina Moneta: Coins in Religion and Ritual, edited by N. M. Burström and G. T. Ingvardson (Oxford – New York: Routledge, 2017), 160–73.

28  Svein H. Gullbekk et al. 75 This count dates to pre-1994. The number today is higher (Keld Grinder-Hansen, “Mønter som kilde til middelalderens økonomiske historie – en presentasjon av et kildemateriale,” Fortid og nutid 2 (1994): 101–33). 76 Gullbekk and Sættem, Norske myntfunn, 9. 77 Klackenberg, Moneta nostra. For an updated survey, see Kenneth Jonsson, “Myntfynden i landsortkyrkor i det medeltida Sverige,” Myntstudier 1 (2011): 1–16. 78 Frida Ehrnsten, “Mynten i Finlands kyrkor,” NNÅ New Series 1 (2014): 153–99, att. 157. 79 In 1991, the comprehensive list of church finds published in IFS 1 had 450 entries. Since then almost 200 finds could be added; they are listed in the yearly bibliography in the Bulletin IFS (most recent: 25, 2018, published 2019). Zäch, this volume, ch. 14. 80 Susanne Fridh, “Mynt 1150–1699 i lösfynd och hopade fynd på Gotland,” Myntstudier 1  (2014): 1–37. The finds of c. 1200 coins from the 2014 excavations at St Olofsholm church are not included in this count. 81 In 1991, the comprehensive list of church finds published in IFS 1 had 450 entries. Since then, almost 200 finds could be added; they are listed in the yearly bibliography in the Bulletin IFS (most recent: 25, 2018, published 2019). Zäch, this volume, ch. 14. 82 In addition, excavations of several churches have been revisited with an emphasis on the coin finds: Jens Christian Moesgaard, “Reconstructing the Context of Metal Detector Finds from Top Soil. A Case Study of the Redundant/Abandoned Churches of Oldrup and Uld, Jutland, Denmark,” The Journal of Archæological Numismatics 8 (2018): 175–206; Eeva Jonsson, “Rukoile meidän puolestamme, Pyhä Jumalanäiti. Keskiaikaiset ja reformaation jälkeiset rahauhrit Neitsyt Marialle Ahvenanmaan Jomalan kirkossa,” SKAS (Suomen Keskiajan Arkeologian Seura) 2 (2018): 20–39; Hellan, “A Stave Church,”; Alf Tore Hommedal, “A bell foundry in Sola Church – Bell-casting and Coin-finds in Medieval Norway,” Schweizerische Numismatische Rundschau 97, 2019 (2020): 245–66. 83 Kristensen, this volume, ch. 4.

2

Pious gifts Coins and church interior Martin Wangsgaard Jürgensen

Introduction Medieval Latin Christianity inherited late-Antique practices concerning the giving of gifts to the temple and the placing of ex votos on martyr graves.1 That is, gifts left behind out of thankfulness and in the hope of aid and divine intercession. This chapter attempts to offer explanations on why and how coins were handled in the church and presents a view into the theological and cultural practices where coins came to play an important part as symbolic tokens of faith and devotion. The timeframe is primarily the late Middle Ages, and the sources discussed to a large extent come from Denmark. The specific Danish context gives a certain bias to the material, but it is to be argued that most of the conclusions reached pertain to large parts of the late medieval world of Northern Europe. Over the following pages, the relation between ritual and coin is presented not from the etic approach very commonly employed in research on rituals and archaeological evidence, where the ritual proceedings are discussed from the outsider’s perspective, but from an emic angle in order to present what the ritual handling of money ideally was believed to facilitate or realize for the participant, according to commonplace theological thinking of the medieval period.2 One thus has to keep both the direct and indirect biblical call in mind, beckoning all to leave behind their earthly goods and follow Christ as read in the Gospel of Matthew 19:21: “Jesus answered, ‘If you want to be perfect, go, sell your possessions and give to the poor, and you will have treasure in heaven. Then come, follow me.’”3 This call for apostolic asceticism was already during the Early Church, slowly transforming into a symbolic donation of gifts, demonstrating the willingness of each individual Christian to give up his or her possessions in complete dedication to God.4 Inside the church, this concept was visually grasped through the depictions of The Three Magi (see Figure 2.1) who approached Mary and the Christ-child with their most precious offerings. The Magi, here, fulfilled the role as the congregation, whereby they set the example for everyone to follow, and Mary as the sedes sapientiae (the seat of wisdom) or the Church in whose lap Christ sits enthroned. This image filtered from Byzantine mosaics into the Romanesque art of the West and was repeated again and again throughout the medieval period where it more or less encapsulated the theological backdrop for what it was medieval parishioners in Scandinavia emulated when offering coins, valuables and food in connection with both public and private ceremonies in their local parish church.5 This leads to the second point of this introduction. No matter if a coin was left behind on an altar, hidden somewhere in a crack or stuck between floorboards or handed

30  Martin Wangsgaard Jürgensen

Figure 2.1 T he Three Magi painted in Fjenneslev church on the Island of Zealand. Mural, c. 1125–1150. Photo: Martin Wangsgaard Jürgensen.

to the priest or one of his helpers at the end of a ceremony as an offering, the coin should be perceived as a gift; never as payment. The danger of being accused of simony was always present and a problem the Church, at least in principle, sought to combat with all means.6 While one of course is actuality paid for Masses and the services of the Church, the money given should be regarded as gifts or charity if we want to understand the religious mindset from which the Church and the community interacted. No service was simply bought, although this is exactly how it appears to the outside beholder. Votive Masses, for instance, while financed by a private person were still not bought in the sense that the financer acquired a service as when buying something at the local market. Money paid for a Mass was gifts offered out of thankfulness to be used for the upkeep of the clergy, the upkeep of the church building and the celebration of Christ. The biblical story of Christ driving the money changers from the Temple7 loomed large behind this line of thinking and to a certain extent was the reason why no bartering over religious services could take place. It was thus stressed again and again that no trading or haggling over prices should be done inside the church and that money in the church room was gifts to God and alms for the poor. As the Danish theologian Christiern Pedersen (†1554) notes in his small book on the Mass from 1514: But one who wants to have a Mass spoken for oneself or one’s parents is to say to the priest: Sir, I may not buy the Holy Mass, neither are we allowed to haggle over it, but say so many Masses for me or my parents or this or that person. I will reward you accordingly for your effort. One who lets a Mass be said should give the priest so much that he [the priest] can live on it that day beyond his food and drink. As Saint Poul writes – He who serves the altar should live by the altar.8 As can be gleaned from this, money and wealth existed within the church in a highly ambivalent state, oscillating between being perceived as practical objects of

Pious gifts: coins and church interior  31 exchange and dirty symbols of everything which did not belong in the house of God. In fact, when seeing coins depicted inside churches from the fourteenth century and onwards, it is mostly with negative connotations, as for instance when Judas grasps his purse with the 30 silver pieces. It is also noteworthy how depictions of the Seven Deadly Sins show avarice – Avaritia – through coins, and in scenes depicting Purgatory or Hell, the greedy can be shown as being punished by pouring red-hot coins down their throats.9 In fact, the deeply tainted perception of coins and riches seems substantiated by the fact that the misers depicted on the church-wall often would be shown embracing the nude Luxuria – carnal lust. Together, this pair made up everything earthly and depraved. Coins and genitalia were consequently depicted side by side as symbols of this world.10 However, no matter what stance the Church took towards coins and secular wealth, the institution was embedded and operated in a steadily more monetarized society. Money simply became if not the only, then at least the most convenient, mode of exchange and thus unavoidable and bound to influence ecclesiastical and pastoral thinking. Mundane concerns of everyday life intersected here with theological ideas and biblical thoughts of purity, which had tremendous consequences for the act of offering in the church. It even, as we shall see in the following at times, led to a positive appraisal of coins as a metaphor for Christ. Unsurprisingly when all this dirty money was to be handled in the church, it happened through ritualized actions serving both to avoid any charges of simony and ensuring that the offering was accepted in the right spirit as well as to the right effect. Such ritual moments were, as we shall discuss below, found at certain points in the liturgy of Mass or during private devotional ceremonies, where the act of giving always was accompanied by brief prayers of gratitude and acceptance.11 Some of this, if not all, may be common knowledge to the reader but has been repeated anyway as it constitutes the necessary background for the following conclusions of the chapter, which sheds light on less obvious aspects of the role and use of coins in the church. Three themes are unfolded over the next pages: firstly the theological role of coins in the church interior is to be described; then we are to approach the ritualized habitus which was necessary when money was handled and exchanged between clergy and laity or laity and altar. Especially when and where such exchanges took place is of importance. Finally, some words about the spatial implications of exchanging coins in the church are to be touched upon. All of this aims to help in the interpretation of the patterns that emerge when examining coin finds from church floors, especially small parish churches in rural regions where we often face a dearth of medieval written evidence which otherwise could have explained what altars the church housed and what devotional activities took place.

Coins offered during Mass In order to understand how the offering of coins took place in rural parish churches in medieval Scandinavia and most other parts of Northern Europe, it is useful to examine the historical roots of the practice and consider how the act of offering slowly was transformed from late Antiquity and into the early Middle Ages. This may not tell much about what took place in the Scandinavian churches presented throughout the chapters of this book, but by elucidating this process of change, the cultural implications of offering during the later Middle Ages become easier to understand. Coins were offered as alms and donations during Mass and acts of private devotion. We are

32  Martin Wangsgaard Jürgensen to begin this survey with the Mass, because one could say that the offerings made here establish the framework for all other instances when coins were given in the church. The medieval Roman Mass, and for that matter the modern Catholic Mass as well, is traditionally explained as structured around two parts.12 During the first part, prayer and reading take place, and during the second part, The Offertory, the congregation approaches God, and the Communion is completed. It is this second part of the Mass which is of interest in our context, as it begins when the congregation is carrying forth their offerings to the altar and thereby to God. From the beginning of the Early Church, the Offertory was problematic because it on the one hand was to allow the congregation to present their offering of thanks to God, but on the other hand, this offering could not be allowed to conflate with the pagan or Jewish practices and altar sacrifices.13 It is generally assumed that the offerings on the altars or tables of the first Christian congregations were wine and bread. These were to be used during the Eucharist, but as all members of the congregation could bring something, there would be more than what was needed during this part of the ceremony. What bread and wine were left after the ceremony was then to be shared between the clergy and the poor. Although different, we can here identify the blueprint of the ritual offering as it was performed during the medieval period many centuries later.14 It was not only wine and bread the community offered; other gifts such as jewellery, fruit, cake and oil could also be brought forth; however, from a theological stance, this was criticized as wine and bread was considered the only things proper; something which was already stressed at the Synod of Hippo in 393.15 At some stage, coins must have entered into the proceedings. Moving rapidly from the Early Church of the fourth and fifth centuries and into the eleventh century, a pronounced change has happened. Peter Damian (†1072 or 1073), Benedictine monk and church reformer, later canonized, mentions in one of his writings that two women had offered gold coins at the altar during Mass.16 As with all such theological writing, the content is shaped according to specific aims and very likely not based on any actual situation. One could even ask what gold coins Peter Damian had in mind when writing, but that is of lesser interest here. What we may note is that he expresses his surprise concerning the women’s offering. To Peter Damian, the idea that someone should put coins on the altar of a church was stunning. We may suspect that Peter Damian played the role of the conservative who catered a former “golden age” where such things did not take place; a role which is in tune with his ambitions of church reform.17 It is, accordingly, noteworthy that other sources from the beginning of the eleventh century mention the use of coins during the Offertory.18 It thus seems as if the beginning of the regular, ritual use of coins as an offering can be placed in this century. How fast and in what way coins found their way into private and public ceremonies, undoubtedly, was different from region to region and very likely faced opposition; an opposition which the example above of Peter Damian seems to echo.19 However strong this opposition was, it ultimately failed, in so far as coins were the most common gift by the twelfth century. This we find confirmed in the writings of Honorius of Autun (†1154), whose elucidations on the Mass and Church dogma in the Gemma animae came to serve as commonplaces in his own lifetime and the following centuries. 20 To Honorius, coins are the central gift during Mass, and based on biblical concepts, he categorizes their prestige as follows and while one text by no means confirms a social practice across subsequent decades and places, I would argue that Honorius presented ideas which were mirroring practices in the society around

Pious gifts: coins and church interior  33 him which is why his ideas were so readily copied and disseminated, not in the least by Durandus who we shall touch upon below. 21 The most splendid gift is the gold coin, which he likens to the gifts of the Three Magi; this is followed by the silver coin, which Honorius compares to the poor widow’s offering, 22 and then he goes on to list other kinds of gifts such as jewellery and precious items. Only then, at the bottom of his enumeration, is he referring to wine and bread as gifts ranking over other edible and drinkable donations. In fact, the positive understanding of the coin as a symbol gradually came to reach beyond its mere presence as a gift. Honorius likens the Host with a coin and states: “Which bread [the Host], is formed in the manner of a coin because Christ, the living bread, was sold for the price of coins and he himself is the true coin…”23 Christ as the true coin, the image is evocative, and Honorius continues to explain the coins as offerings in the following words: And because, since people were not taking Communion, it was not necessary that such large bread should be made, it was established that it [the Eucharistic bread] should be made and formed in the form of a coin; and the people offer coins for the offering of flour [to bake Eucharistic bread from], so they recognize that the Lord was betrayed for [coins]…24 The role of coins in the Passion is stressed here and the direct relation between the offered money and Christ confirmed again. Durandus repeated Honorius’ coin metaphor in his incredibly popular and long-lived Rationale divinorum officiorum, finished some time before 1286, by stating: But the bread [the Host] is thus formed in the manner of a coin, first because the bread of life was betrayed for the sake of the coin, also because [the Host] ought to be given as a reward to those laboring in the vineyard [living faithfully].25 Thereby the coin/Host analogy became a popular metaphor of the late medieval period, which tells us how profoundly imbedded coins became in the devotional vocabulary and understanding of religious practices from the twelfth century and onwards. 26 It furthermore goes to show how thoroughly any offering in the church during both Mass and private devotion was entwined with notions of Christ offering Himself on the cross, an offering repeated during every Mass. 27 While gold and silver coins may have been outside the reach of ordinary parishioners in the time of Honorius as well as in later centuries, the hierarchy he presented to his readers does show how much had changed from the time of the Church Fathers of late Antiquity. 28 And one could, without any problems, substitute the silver coin in his listing with the omnipresent penny of the later Middle Ages or simply add the petty cash below the silver. While grand offerings thus remained the most prestigious and beneficial in terms of intercessory power during the late Middle Ages, it is important to remember that the Church stressed the great value of any offering – even humble ones. 29 As Augustine already had stated: “Whatever is given to God becomes sacred, but especially the oblation of the holy altar.”30 Whatever offered on the altar became sacred: this line of thinking shows how even modest pennies gained a spiritual worth as an offering which by far superseded their limited value outside the church walls. Later, these coins would then once again be transformed back to worldly currency

34  Martin Wangsgaard Jürgensen when the clergy put them to use in society for pastoral care, upkeep of buildings, living expenses and so forth. It is also through this that we may understand the phenomenon pointed out in several chapters in this book that old or seemingly obsolete coins appear among the coin finds in churches. Hence, we can ask if they were kept by the parishioners as special “church money” to be offered instead of current coinage, and that may indeed have been the case as such old currencies could be translated into a precious spiritual economy on the altar.31

Donations and church tithe As stated above, it was common practice by the twelfth century in southern and central Europe to donate petty cash during the Offertory of the Mass, and the practice seems to have reached Denmark and Norway at least by the late twelfth century, Sweden and Gotland somewhat later in the late thirteenth century.32 How common or widely disseminated into the different layers of society it was though is a question of regional monetization and outside the scope of this chapter.33 But, at least until the general introduction of coinage of small denominations, it would seem probable that the act of giving of money would have been a symbol of high status that elevated the donor above the majority within a parish or community where food and produce still was the common gift. A social dimension should in other words be kept in mind when discussing the regularity of coin offerings of the twelfth century and beyond, as the act of giving, along with so many other practices in the church interior, was a way to promote and display one’s dedication and religious fervour.34 A host of offerings to the Church were at some point or time either demanded from the laity or given voluntarily, as for instance grants of land and goods handed over to the Church in testaments, all of which was interlinked with notions of piety and devotional practices. However, here we are to take a somewhat closer look at church tithes which at least in Scandinavia came into use during the twelfth century with the division of the land into the parochial system.35 Tithes are, as we know, the taxation of all parishioners where one-tenth of the income is given to the Church who then parts it three ways: one part to the upkeep of the local parish church (the fabrica), one part to the parish priest and the last part to the see as such. One component in this, which has importance for this chapter, is the fact that tithes resulted in a doubling of the gifts to the Church. Tithes were, at least symbolically, offered in the form of produce, wax or coin inside the church during Mass on specific “offering days”, where the priest would accept these donations from his flock, standing in the chancel arch (we are to return to this practice below) and probably have the parish clerk carrying them to the high altar.36 It would, unsurprisingly, be easy for the laity to conflate such presentations of the tithes with the ordinary offerings of any Mass, which indeed seems to have been the case; but is to be remembered that tithes and offerings, while similar in many ways, were two different things. The difference is subtle though. Tithes were understood by the Church as a gift, not a tax, and just like offerings during Mass, tithes were distributed in the same three-way split, with only the see being the benefactors instead of the poor. However different tithes and offerings were, this doubling of tributes to the Church had consequences for the liturgical offerings in so far as the tithes, perhaps unavoidably, took the guise of taxation, while the shift also seems to have strengthened the private or intimate character of offerings during Mass as an individual transaction between the donor and God.37 This differentiation

Pious gifts: coins and church interior  35 is important, and it resulted in a certain unwillingness to make offerings during each Sunday Mass. A homiletic explanation of the Mass in Middle High German (Mittelhochdeutsch) from the mid-thirteenth century by the famous Franciscan Berthold of Regensburg (†1272) addresses exactly this theme: And thus they [the peasants] say: “But the priest is rich enough; why should we make offerings to him?” God does not want to be deprived of that which is his, whence the wise man [the psalmist David] says “offer ye the sacrifices, etc.,” because it is a great thing and holds such good meaning.38 The unwilling laity was in other words encouraged to make offerings, and it was stressed that it was the ideal always to give during all Masses, but in reality, the Church bowed under the pressure and accepted that an offering was not necessary for every Mass, especially as the number of Masses celebrated at the same time was growing during the Middle Ages. At least, the result was that it became uncommon to give offerings during the ordinary Sunday Mass in the late Middle Ages. This can be followed to such an extent that the Offertory slowly shrinks and disappears from late medieval missals, making it the rule rather than the exception, that liturgical books from the late fifteenth century contain any information on offerings during the Sunday Mass.39 Instead, the major celebrations of the church calendar – Christmas, Easter, the inauguration day of the local church building – and the important feast days in the lives of the individual parishioner – baptism, wedding and funeral – were when offerings were performed and which furthermore was the feast days on which the laity since the Fourth Lateran Council in 1215 were required to come to church.40 A woman to be churched, for instance, would even by the end of the Middle Ages offer money or goods, sometimes yarn fabricated during her exile at home, on the northern side altar dedicated to the Virgin Mary.41 A shift was taking place from the twelfth century when tithes were introduced to the fifteenth and early sixteenth centuries where the tithes gradually had come to be perceived as the general gift of thanks to the Church, a role previously held by the offering during Mass; an offering which by then had turned into something intimate and private, closely aligned with other acts of private devotion. In fact, it was only after the Council of Trent in the middle of the sixteenth century that this picture was reversed in the Catholic Church and the offering during the Sunday Mass once again became the primary place for such liturgical gifts.42 When considering how coins were handled and given inside the church during the late medieval period, it is thus an anachronism to focus solely on the High Mass, as this first and foremost would reflect an early medieval practice and, for the Scandinavian countries, the way offerings were performed after the Lutheran Reformation. An anachronism which may seem of little importance, as the offering more or less would be performed the same way during any type of Mass, but if we argue that the spread of coins in the church floors reveal something concerning the use of these buildings and the devotional practices in which they were included, such differences in the understanding of the rituals where the coins were employed gain tremendous importance.43

The act of offering – the Offertory performed We are now to take a closer look at the act of offering, and it should from the outset be stressed that we only to a certain extent know how this proceeded, as local variations, home-grown customs and none the least the specific spatial settings of each

36  Martin Wangsgaard Jürgensen

Figure 2.2  The small village church of Tystrup on the island of Zealand. Photo: Martin Wangsgaard Jürgensen.

individual church shaped the ritual. Regional differences in the understanding of the ceremony, the motions performed and church furnishings, consequently, all contribute to make each church building something unique.44 When we are to understand the act of giving as such, it is, therefore, the broad outline of the ritual which can be described, but as shown in the chapters of this volume, coins found in the church floors at times enable us to close in on the ceremonial activities and substantiate our knowledge of how donations were handled specifically – something written and pictorial sources only rarely, if ever, supply. As an example of an offering, the small rural parish church of Tystrup on the island of Zealand in Denmark will in the following serve as a spatial setting and illustrate how priest and parishioners moved during a late medieval Mass (see Figure 2.2).45 When the first part of the Mass was concluded with the Oremus, where the priest spoke his “let us all pray” and voiced thanks to God on behalf of the congregation, the second part of the Mass, the Offertory, was begun (see Figure 2.3a).46 Once this prayer of thanks was finished, the parish clerk would from his seat initiate a psalm. This singing is important. The words changed according to the church calendar, but the mood expressed was always festive – a point which theologians writing on the nature of the Mass had stressed since Augustine.47 The song was believed to attune the mind of the congregation to the offering and fill them with excitement. The biblical commonplace used to elucidate this was the words of the Second Epistle to the Corinthians:48 “Each one must give as he has decided in his heart, not reluctantly or under compulsion, for God loves a cheerful giver.” The parishioners should, therefore, look forward to the offering, and the parish clerk’s song was both believed to express and further this joy. While the parish clerk began singing, the priest rose from the altar, where he had concluded his prayer kneeling, and moved to the chancel arch, the boundary between choir and nave, and faced the congregation who was gathered there (see ­Figure 2.3b).49 In many churches, the priest would perhaps only step to the side of the high altar, and the only thing important actually was that the priest moved from the

Pious gifts: coins and church interior  37

Figure 2.3 a–e:  R  econstruction of the offering as it could have been performed in the village church of Tystrup on Zealand by the end of the fifteenth century. See text for explanation for the individual stages. Reconstruction by Martin Wangsgaard Jürgensen, based on a plan of the church by Danmarks Kirker.

centre of the altar to the Gospel side (the left side when facing the altar), from which position he would receive his flock. A crucial question is of course whether the laity had access to the chancel or not, as the congregation had to enter this area in order to put their offering on the altar. In the medieval church, one of the most pronounced boundaries was the separation between chancel and nave. The writings of the medieval liturgists state unequivocally that the chancel is not to be barred as such but shielded from the laity in general in order to avoid desecration or what we might call “pollution” of the sacred, but in the late medieval period, this clear distinction no longer was as obvious as before.50 The veneration of the consecrated Host evidently played a vital role in this, since it spurred a wish for proximity and a first-hand impression of the wondrous relic (we are to return to this below). The same was of course the case with the elevation during Mass. People, or at least some, were allowed to get very close to the altar and the Host – resulting in friction and social tension. Two Danish theologians, Christiern Pedersen

38  Martin Wangsgaard Jürgensen who has been mentioned before in this chapter and Poul Helgesen (†1534), discuss this problem. The former writes that: “Lay persons should not stand too close to the high altar or under the nose of the priest while he says Mass, and especially women should not stand or lie too close to the altar.”51 While the latter states: “Some think of themselves as being very pious when the priest makes the body of God appear and then rush so close that they may see that same body of God very clearly…”52 Helgesen rejects such behaviour as an empty, impious practice. However, Christiern Pedersen also declares that if someone in the congregation should possess the ability to understand Latin, he may sit in the chancel during Mass, since he is capable of following and appreciating the liturgical readings.53 This notion seems to echo an older source from Norway from 1320, stating that a layperson had nothing to do in the chancel unless the person was reading or singing with the priest.54 Furthermore, if we go to the synod statutes from the Århus diocese in Jutland, dating from 1443, or the statutes from Roskilde, formulated 1517, it is stated that the consecrated Host could be displayed in the church and visited by the laity, but should be outside their reach – that is, locked in a protective cabinet – especially out of reach for women. 55 But, if we, for instance, note that the English mystic Margery Kempe (after †1438) sat in the choir to “offer her Mass penny after the custom of the place…,”56 it is obvious that there was a difference between learned theological writing as well as between church legislation and the actual conditions in parish churches. The boundaries were fluent and as easily crossed as the right to special privileges was easily obtained. While the most proper place for the priest to receive the offering would be in the chancel arch, this ceremony would probably more often than not take place by the high altar. When collecting the offerings, the priest would very likely bring some sort of container or collection board along to hold the coins (see Figure 2.4).57 A mite box close

Figure 2.4  C  ollection plate from the church of Brændekilde on the island of Funen, ­Denmark, c. 1500, with the image of an unidentified saintly bishop. Photo: Arnold Mikkelsen.

Pious gifts: coins and church interior  39 to the chancel arch could also be the place where the parishioners dropped their donations. It might even be possible that the parishioners in some places put their offering directly on the altar, perhaps on a plate or a corporal. No rule existed concerning how the offerings of the laity should exactly be received, and this part of the ceremony is where we see the flexibility and possible diversity in medieval liturgical practices, which show why coin finds can grant a degree of knowledge concerning the ritual which is otherwise evading out attention. Now in place, the priest was ready to greet those in the congregation who were to make their sacrifice (see Figure 2.3c). These would, in turn, be probably men first according to tradition, who would then move up to the priest and make their offering. 58 The priest would then speak the blessing: “What you give here shall in the name of the Lord be returned to you a hundred-fold” and then let the donor kiss his hand, stole or the place where the coin was given (the box, altar etc.). To see how such a situation could have looked, we might tentatively look to Israhel van Meckenem’s depiction of the apocryphal story Joachim’s Sacrifice; an engraving made c. 1490–1500 (see ­Figure 2.5). Despite the fact that the scene depicts a Jewish sacrifice, the model for the proceedings is certainly taken from van Meckenem’s own time, and it would not seem

Figure 2.5  J oachim’s Sacrifice in the temple. Note the coins on the floor by the altar. Etching, Israhel van Meckenem, c. 1490–1500.

40  Martin Wangsgaard Jürgensen far-fetched to envision the same situation taking place during the previous centuries of the Middle Ages as well. We here see the priest standing next to the altar, receiving the offered coins which are placed directly on the altar cloth.59 Joachim is in front while Anne is standing behind him with a rosary in hand, undoubtedly reflecting the devotional nature of the ceremony. Of particular interest for us are, of course, also the depicted coins on the floor which have rolled from the altar, a telling detail of the everyday atmosphere which contrasts Anne’s prayer beads. But, to return to the offering, as it could happen in a small village church: When the last among the laity had made the offering, the parish clerk could finish his singing while the priest walked back to the altar, if not already standing there, and knelt down (see Figure 2.3d). The parish clerk, moving from his seat, would then take the donations and put them on the altar or take them aside immediately (preferably into the sacristy if the church had such one) while the priest was preparing the Eucharist (see Figure 2.3e). This ritual exchange of coins or gifts between the priest and congregation prior to the Eucharist is undoubtedly the most important place in the Church-controlled use of money in the liturgy. The Offertory was the official place during Mass where, under the auspice of God so to speak, economic transaction could take place; hence when land or property was given to the Church, the transaction was symbolically concluded during the Offertory. This ritualized exchange was consequently the time when the institution freely could accept deals with the congregation and confirm otherwise secular transactions; a situation performed exactly like the above-mentioned act of giving tithes. The ritual context shielded the proceedings from, for instance, accusations of bartering in the church or even worse of simony, because the deals obviously were made and accepted in the name of God and, therefore, above earthly, everyday matters.

The church interior, side altars and private devotion – an excursion When interested in the occurrence of coins found in church floors, an understanding of the above-described proceedings is obviously important, but it is also obvious that the controlled proceedings of the ritual left little room for large quantities of lost coins. It is, therefore, relevant to examine some of the other situations where the laity offered coins. In order to do so, we need to leave the coins aside for a moment and make a small excursion into developments within the private devotion, as this plays crucially into what is at the heart of this chapter. Traditionally, when analyzing medieval church architecture, the church room is treated as one east–west oriented space, usually separated into two major ­compartments – the chancel and the nave, as discussed above. This structure had already been formed and refined in early, pre-Romanesque sacral buildings and was retained throughout the period. Architectonically, the compartmentalization of the interior would often be emphasized through the chancel area being elevated from the nave by stairs or through the entire choir being constructed like a separate architectonical addition to the nave under a roof of its own. A screen – a so-called rood screen – or a lectorium might also often mark the boundary between the two generic segments of the building, limiting the direct view from one part to the other. In this structuring of the church building, the attention was centred on the high altar found in the east end, which can be said to be the natural climax of the interior. The most

Pious gifts: coins and church interior  41 eastern altar was the liturgical or ritual heart of the church and continued to be so throughout the Middle Ages. It was here the “Holy” was addressed at the clearest and the part of the church that received the greatest reverence. But as the spiritual climate of the period changed, side altars gained in popularity as well as importance and thereby contested the exclusive status of the high altar. Two very simplified models of church interiors can, therefore, be presented, in the following called the “­Romanesque” and “Gothic” type. When dealing with Romanesque architecture, a simple east–west oriented model of the parish church interior can be used. The high altar as well as the side altars were all usually placed facing east, allowing for the celebration of the Mass towards the west (see Figure 2.6). In doing so, it leaned on a heavy light-symbolism that paralleled Christ, God, The heavenly Jerusalem and everything else positive with the rising sun. But this strict liturgical east–west orientation of the church interior did not last. As the cult of saints and relics grew rapidly along with the conception of Purgatory, there was soon a demand for more altars which could not all face east. It was necessary to accept a break-away from the east-west-bound orientation that had defined the liturgy of Mass since early Christianity, simply to make room for new altar space.60 In the Late Middle Ages it, as a result, became possible to celebrate the Mass facing north – the direction traditionally interpreted as symbolizing all evil – but for that to happen required some profound changes.61 A large number of scholars have discussed and demonstrated the growing importance fixed to the idea of seeing the Elevation of the Host in late medieval devotional culture – a devotional culture which fetishized the act of perceiving the sacred visually into an almost haptic experience. Especially, Miri Rubin’s work stimulated new research into the subject, and by now, there seems little doubt that the idea of participating in the Eucharist through vision alone created an enormous impact on the way church interiors were structured from the late thirteenth century and onwards.62 At the same time, the research done by, for instance, the late art historian Michael Camille on Gothic art stresses a different development of visuality in the late Middle Ages.63 He, as also Sixten Ringbom and Hans Belting, points towards a rising individualization of the devotional relationship between the beholder and the focal point of the beholders’ gaze – be that an altarpiece, the altar itself or any given image for that matter.64 And the laities’ increasing use of devotional literature during Mass – prayer books and such – seems to be part of the same phenomenon, in that the gaze

Figure 2.6  The Romanesque interior with the altars oriented towards the east. Plan: Martin Wangsgaard Jürgensen.

42  Martin Wangsgaard Jürgensen and focus of the beholder shifted to a self-contained relation between book, reader and altar, that although used in the church – a collective space as such – in principle it disclosed everyone other than the reader. These two trends appear to have existed side by side influencing each other. The steadily growing visual veneration of the Eucharist and the internalization of devotional imagery in the late Middle Ages created a basis for a new layout of the church interior, not based on the old symbolic east–west axis, but a layout that was carried by the idea of a link between altar and beholder. And as the relationship between devotional images, or rather what the images represented, and their audience steadily became more privatized from the twelfth century and onward, the need for private altars also became more urgent and spread to all parts of society. Whereas, in the Early Middle Ages, the ability to finance and uphold an altar in a church had primarily been reserved to the highest-ranking members of society, it now became possible for almost everyone to participate in the establishing of a private altar through joint ventures in the shape of guilds and confraternities. Even peasants could join together and finance their own altar in their local parish church and thus establish a direct link between themselves, as part of a group or as individuals, and their altar.65 When speaking in broad terms, there was no difference between the high altar in the east end of the church and the smaller side altars found elsewhere in the churches.66 The smaller altars had to be consecrated in exactly the same way as the high altar, and the Mass was probably celebrated in more or less the same way in front of the smaller altars when they were used.67 This created the possibility for a series of smaller focal points in the church interior almost interchangeable with the high altar. The situation described above is what contributed to the creation of that we might call the “Gothic type” of church interior. A specific example to illustrate this could be St Olof’s church in eastern Scania, the southern part of Sweden that was part of Denmark in the Middle Ages. Here, a small thirteenth-century village church rapidly became a renowned point of pilgrimage due to a miraculous silver axe – the attribute of St Olof – retained in the church and a sacred spring that suddenly started to flow next to the graveyard. This resulted in a massive rebuilding during the first part of the fifteenth century due to pilgrim donations, and several side altars were erected (see Figure 2.6).68 Seven out of nine known altars still exist in the church today. Although the church of St Olof might be an extreme example due to its great popularity, it still gives us a first-hand impression of how the growing number of altars inside a late medieval church could be placed; a placing that could almost appear random, creating a disorganized overall structure of the liturgical celebration of Mass that had been unthinkable in the early Middle Ages (see Figure 2.8). Side altars were physically, spiritually and financially perceived as small spaces or rooms inside of the church building. Some altars, of course, were placed in separate chapel buildings attached to the church, which in a very tangible way demonstrates the space-generating effects of the side altars.69 But, more often, altars were simple stone or brick constructions, perhaps elaborated by superstructures such as baldachins or flanked by walls or grills, placed where there was space available in the church, and they still functioned as private spaces partitioning the church interior.70 This can be illustrated from three points. Firstly, written sources, secondly the images related to the altars and finally the ritualistic behaviour practised in front of and around the objects of devotion.

Pious gifts: coins and church interior  43

Figure 2.7  The church of St Olof in Scania, southern Sweden. Photo: Martin Wangsgaard Jürgensen.

Figure 2.8  T he “Gothic” interior of St Olof’s church with altars facing in other directions than east. Plan: Martin Wangsgaard Jürgensen.

Through late medieval written sources, the private, intimate status of side altars quickly becomes apparent. As an example, we can note that, in connection with the Lutheran Reformation in northern Germany and Scandinavia, conflicts often arose locally between the new Protestant clergy and the founders of altars in the parish. The families and guilds that had paid dearly for the establishing of a shrine, be it in a separate chapel or as a small altar somewhere in the church, were reluctant to remove what they thought of as “their altar”– be it Catholic or not. By this, they in fact

44  Martin Wangsgaard Jürgensen claimed private ownership of a small part of the inside of the church. In the town of Malmö – again in Scania – this resulted in a time-consuming trial between the merchants’ guild and the clergy of the St Peter’s church, based on the question whether the Lutheran priest had the right to use the merchants’ chapel or not.71 A different category of a source material is the contemporary descriptions of churches of which there are not many. But especially in relation to burials in the interior of churches, we get glimpses of the fact that the inside layout was very much perceived as a building housing several smaller units of space – or altars that is. Thus, a testament can point out that the deceased should be buried next to St Lawrence in the nave for instance – meaning in front of the altar dedicated to this saint. In that way, the altars contributed to the definition of the internal geography of the church. This brings us to the images. Of course, all pictures in the churches should, following most of the leading medieval theologians, primarily be seen as a didactic help for illiterate or unlearned, but some images received a special status either through indulgences connected with them or by popular beliefs. This could be close-ups of the face of Christ, as it for instance was known from the Veil of St Veronica. Sculptures in particular seem to have been prone to take on a cultic character and convey an intimate experience to the laity of being close to the given saint, Mary or even Christ himself in flesh and blood so to speak.72 What is meant by this is that certain images had a presence inside of the churches that promoted them as something special. Such devotional images would usually be found on altars or have altars close by. In other words, decorations could – especially through sculptures – reinforce the presence of the altar. This space-generating effect seems even strengthened when looking at ­fi fteenth-century sculpture that almost always was placed inside of small housings indicating a room for the saint or holy person depicted.73 But, besides the liturgy of Mass, several other rituals related especially to venerated depictions of saints or images of Christ created an atmosphere around the devotional object that must have blurred the lines for instance between the veneration of a depiction of the Virgin Mary and Mary herself. What is referred to here is the nursing and handling of pictures and sculptures on the feast days and the processional carrying of the cross into the fields and streets of the towns as part of the communal blessing. All of which must have formed a very strong bond between community, image and the individual viewer when the object was exhibited on the altar.74 It can furthermore be argued that chanting and singing of psalms in front of altars, contributed to the trends discussed here. When chanting and singing at altars or even in front of devotional objects, the idea of the celestial choir praising the given saint or scene was of course present. Along the same line, we very often find angels with musical instruments placed in the wall paintings around altars ultimately evoking the idea of the 150th psalm and its celestial Hallelujah. But in itself, music seems not to establish any sense of space inside of the church. On the surface that is, because a closer look at the liturgy of feast days and the texts sung and chanted will very likely reveal a very distinct sense of spatiality that has to be taken into account along the same lines as the visual arts. Side altars thus became of such importance in the late medieval devotional culture that at times they could almost seem to overshadow the status of the high altar, but we still find plenty of evidence demonstrating their subordinate status. To be able to see the high altar from the side altars appears to have been of some priority among the founders of altars. In the Scandinavian material, we find little or no architectonical

Pious gifts: coins and church interior  45 evidence for this, except for the very regular widening of the arcade opening between chancel and nave. But in English parish churches, the so-called “squints” – peepholes in the walls that create visual contact between chapels, chantries or side altars and the high altar are a relatively common phenomenon. The idea is here to emphasize the necessity of contact between the altars.75 This illustrates very clearly that although there was a wish for personal or more direct contact between beholder and altar, the high altar remained a unifying element that kept the individual units of altars, chapels and so on connected. To return to what has previously been stated, Romanesque church interiors were bound in an east–west axis with the focal point resting in the east-end of the church. The church interior was almost exclusively constructed in a one-way direction pointing towards the main altar, which to a certain level makes it possible to perceive the entire Romanesque interior as one singular space, with no difference between individual and the communal celebration of the Mass since it happened facing the same direction. Although what we might call cult images already were widely spread in early medieval Europe, the idea of private altars and private devotion in itself was only slowly developing on a wider scale. But this it did, and by the end of the Middle Ages, church interiors had changed tremendously. The cohesive Romanesque church room no longer existed. Gothic interiors, no matter how harmonious and unified they may seem to us today when cleansed of all medieval visual signs of piety, have to be regarded as eclectic spaces compiled under a single roof. The example from Scania, the church of St Olof, illustrates this. The unified sense of common direction was now solely reserved for the main altar and the high Mass. Everything else could be placed and directed where there was room for it, although east was still the preferred direction for altars whenever possible.76 The trend for private devotion made this shift possible and very much illustrates the need in the late Middle Ages for the laity to be closer to the sanctity and personalize their devotion – a closeness that the high altar apparently was unable to give. The growing number of private altars can accordingly be interpreted as a turn towards the individualization of devotion and one could point out, that the actual breaking away from the east-western axis of worship linked to the high altar perhaps created an even stronger feeling of personal closeness between devotee and altar. For us today, when looking at medieval churches, these spatial, liturgical or ritual compartmentalizations of church interiors are no longer evident, but by accepting the complicated spatial use of the church, it perhaps also becomes a bit easier to understand the rising demand for processions and rituals embracing the church building in its entirety. Therefore, exactly as civic authorities and citizens in general could use processions through towns and the countryside to demonstrate ownership, authority and attachment or belonging, the Church as an institution did the same inside of its ecclesiastical buildings. The Mass celebrated from the high altar surpassed all other Masses sung in the church, and the processions of clergy visiting the altars of the church during the Sunday procession and on feast days tied a space together that on the inside, exactly as the society that surrounded it, was separated by divisions of status, ownership, gender and profession. What has all of this to do with coins? As the importance of the private devotion grew, it became of equal importance to leave a personal imprint or presence in the church; an individual token showing personal dedication to a saint or ultimately, Christ. The English anchorite Christina of Markyate (c. †1155) was forced to leave the monastery into which she had entered, and according to her biography, she is said

46  Martin Wangsgaard Jürgensen to have etched a cross into the door of the monastery church with her fingernail as a symbol that she left a piece of her heart in the place when leaving: “With one of her nails, she wrote the sign of the cross on the gate at the monastery, as a signifier that she had placed her affection there.”77 And, relevant to our topic, she sacrificed a coin before leaving the monastery completely. Whether or not she actually etched a cross into the door is of little relevance to us. However, what is important is the fact that the very act was feasible to the composer of her biography. By leaving her mark on the church Christina left a bit of herself in the place and thus under the auspice of God. It is also here money enters into the discussion because the act of leaving coins by side altars and images is closely akin to leaving ex votos, lighting candles or etching graffiti in the church. It is an act of self-representation and ultimately the same as financing a votive Mass or even a new altar in a church albeit on a minute scale. One can in other words see the gifts of money, no matter if they ended up in a collection box or were pushed down between the floorboards of the building, as something parallel to the tremendous sums often given to the church by members of the upper strata of society. The coins of small value, the so-called “pledge pennies” thereby came to be a way in which the individual parishioner of modest means could embrace and partake in the culture of offering and giving in order to demonstrate faith and hope for mercy on the Day of Judgement. To conclude one’s prayer by leaving a penny by the altar as a token of devotion was to open a channel to Heaven and the donor could in exchange be understood to hold a spiritual token of divine gratitude, as when an anchorite tells the previously mentioned English mystic, Margery Kempe, that she held an earnest penny of heaven, thanks to her great piety.78 But, as we shall see below, the altars were far from the only place where offerings could be made in the church, a host of opportunities for sacrificing money existed within the late medieval church – to all at once and not all of them in all churches, but the possibilities were many.

The act of offerings – side altars, private devotions and the church interior As was already stated in the first part of this chapter, it slowly was less and less usual for ordinary parishioners to make offerings during every Sunday Mass; instead, important feast days of the church and days of private importance became the time to leave donations. The act of offering thereby gradually became a private concern reflecting the life-situation of the individual members of the congregation. But as it also has been stated, it was not solely during the High Mass or votive Masses or even only at altars that offerings could be performed. We have numerous references to altar images and relics shown in churches with collection boxes or trays placed next to them.79 An alms-box next to the church door, a statue of a saint or some other image is nothing unusual. As Bernard of Clairvaux (†1153) wryly remarks – great images and relics bring even greater income.80 One could say that the late medieval church interior slowly came to be constituted by isolated, non-permanent spaces, created and shaped by the internalization of altars, images and ceremonies of each individual churchgoer. Each altar and each image used in the private devotion of the congregation came to form small spaces under the shared roof of the church. When looking at coin finds from church floors, the painted decorations on the church-walls and the placing of side altars are therefore all-important to consider as

Pious gifts: coins and church interior  47 it is here, at the altars and images, that many of the coins were donated.81 Altars and images, put differently, constitute focal points on which one may concentrate when examining why certain patterns of distribution occur in coin finds from a church, but we should not forget all the temporary places and installations where coins also could be donated. We could here first and foremost think of the collection boxes put up with a special purpose, such as for instance to gather financial support for building projects, crusading and a host of other common purposes.82 Such collections constitute extra components in a church, present for only a certain period, probably leaving only a few traces behind. More relevant than temporary collection boxes is the recurring devotional celebrations or situations occurring throughout the year. The painting by Pieter Breughel the elder, entitled The Fight between Carnival and Lent, from 1559, offers an interesting window into an exact cycle of recurring events (see Figure 2.9a). The famous picture has several insights to offer, contrasting and illustrating the festive carnival which is taking place immediately prior to the Lent. In our context, the interest lays in a church depicted in the upper corner of the picture, clearly shrouded appropriately for the season.83 When approaching the church step by step, we first note the pilgrims begging alms and then stalls put up next to the church door where money also changes hands (see Figure 2.9b). Both situations seem to speak into an urban milieu, as does indeed the whole picture, but similar situations could also take place in the porches and even the west-end of nave in smaller, rural churches as well. Nevertheless, it is more interesting that we see a small, cloth-covered desk put up just inside the church with a crucifix (see Figure 2.9c). This is undoubtedly a very simple version of the so-called Christ’s Tomb sculptures we know from medieval churches all over Europe. That is, sculptural depictions of the entombed Jesus that served both as devotional objects and components during the Easter liturgy.84 In Breughel’s painting, we clearly see how coins are placed next to the crucifix. This exactly seems elucidating

Figure 2.9a  Fight Between Carnival and Lent, oil on canvas. Pieter Brueghel the Younger, 1559. Kunsthistorisches Museum Vienna.

48  Martin Wangsgaard Jürgensen

Figure 2.9b  Fight Between Carnival and Lent; close-up of beggars in front of church.

Figure 2.9c  Fight Between Carnival and Lent; collecting inside the church door.

Pious gifts: coins and church interior  49 because it teaches us how more or less fleeting installations could be installed in the church during the year and takes a place in the gift economy of the place whereby it, at least potentially, could leave an imprint on the way coins could occur afterwards under the church floor. Such recurring feast offered possibilities for the laity to interact with the sacred, to form bonds between themselves, the divine and the counsel they sought in their church. One way to do so was to leave candles, coins and a host of other objects behind as tokens of thanks and reminders of the fact that the parishioner had visited the altar, image or shrine. This was already discussed at length above, but what still needs to be addressed is that up until now only the formal or traditional ways of offering coins has been touched upon. Underneath the kinds of donations so far presented, a sprawling number of practices and customs must have existed; small rituals, probably performed by laity and clergy as well, which drew some of their forms from the generally acknowledged ceremonial and liturgical practices but still existed outside the formal church rituals. As an example, one could mention the custom of leaving a coin overnight beneath the crucifix in the church on Good Friday and collecting it the next day as a blessed talisman; something which was considered an abuse, but still performed.85 Although mentioned often in written sources, such practices are almost impossible to substantiate today.86 The same goes for the coins which unquestionably were intentionally hidden in the churches as more or less secret offerings. Compared to the amount of “normal” offerings, such hidden coins were probably few, but they were undoubtedly performed and should not be disregarded completely, as more recent numismatists have had a tendency to do.87 The reasons for hiding the offering could be many, but as has been argued in this chapter, offerings grew more and more into an expression of personal piety and something between oneself and the divine. To leave something behind as an offering to God – even when hidden – was thus a token of faith, in principle, as permanent as building an altar or something even more elaborate. When envisioning a particularly late medieval church interior, we thus need to think of a space filled with permanent and moveable features which all had relevance as places for sacrificing coins. To us today, this division of the church interior into many islands of devotional opportunities is not obvious any longer, but if we think of the church as a complicated landscape or topography, it becomes easier to understand how the furnishings and images were defining the patterns of use in the church and how coins under ideal conditions in fact seem to enable us to trace these patterns of use. Nevertheless, a certain bias is also present in this in so far as it necessarily would be the permanent fixtures like for instance the high altar, where most coins would be found, while devotional images, side altars and impermanent installations perhaps played a much greater role in the religious lives of the parishioners but only during a few years and decades, whereby the imprint left through coin finds would be much lower. The information we gain of the use of the church interior and the private devotion in the medieval period through coin finds is thus operating at its best in a longue durée perspective. Accordingly, it is also highly interesting when excavated coin finds indeed seem to cluster over brief timespans, as we close in on devotional activities which otherwise could only be suspected.88

Concluding considerations There can be little doubt concerning the impact the introduction of petty cash like pennies had on the way the offerings took place in the church. One could for instance

50  Martin Wangsgaard Jürgensen note how Benedicta Ward remarks, in her seminal study on medieval shrines and miracles, how the poor and lowly almost seem to miss completely from the recorded lists of visiting pilgrims at the shrine of Thomas Beckett during the late twelfth century. But interestingly, she notes that the Church was aware of this and that many of the poor, out of shame, did not announce what gifts they left at the martyr grave because of humility and prudency.89 Exactly, this pattern changed in the twelfth century as the use of coins of small denominations filtered into all levels of society. These small coins in principle enabled everyone to participate in the mode of ritual-giving in the church which had been characterized by social divisions based on those who had access to valuables and those who could only give natural produce: a division which was apparent from the previously quoted Honorius of Autun, who emphasized gold and silver above all else. The medium of the petty coins enabled everybody to take part in these economies of salvation, and one should thus not be blind to the potentially very strong influence that that coins had on the way devotional activities developed in the church, a concern which is rarely touched upon as coins and private devotions traditionally seem to be perceived as belonging to completely separate worlds. Coins did not become the sole type of gift donated in the church, but cash did to a certain degree succeed in pushing foodstuff out of the church and made it easier for the individual to make offerings whenever, no matter the time of year and the situations, as small change always could be brought along. What has been presented in this chapter is a broad view on the use and role of coins in the church interior. One could say that it is the rules by which money was used and exchanged in the church which has been presented here. How these rules were understood and executed locally is a different question which can only be answered through local studies of specific places, such as the chapters in this book. It is of course a shame not to be able to clearly define how and where money was given in a medieval parish church, but the use of the medieval church interior was much too flexible to make such firm conclusions. The medieval, and in particular the late medieval devotional life, was rich and complex and operated under rather unrestrained conditions, much of which the Protestant reformations of the sixteenth century in northern Europe abolished or limited drastically. We may today speculate how a particular coin ended up in a church floor: was it dropped by accident or perhaps hidden by the owner, and the question is of course interesting, but if we lift the gaze a little, the difference between the lost and hidden becomes irrelevant as they both from a devotional perspective are an expression of the same devotional culture. At the one end of this culture, we find donors building whole churches; at the other end, we find the donation of petty cash placed on the altar and stuck between the floorboards – all of it is gifts given in the hope of salvation and in these coins prove a fascinating trail to follow. To this, we can add the number of accidentally lost coins which give further insights into the patterns of use in the church interiors as they with some caution can be perceived by us as further footprints on the path to understanding how and why medieval men and women acted as they did.

Notes 1 The literature on ex votos is comprehensive, but some insightful discussions of this subject can be found in Lenz Kriss-Rettenbeck, Ex Voto: Zeichen Bild und Abbild im christlichen Votivbrauchtum (Zurich: Atlantis, 1972).

Pious gifts: coins and church interior  51 2 See the perspectives of commonplace thinking examined in Eamon Duffy, The Stripping of the Altars: Traditional Religion in England c. 1400–c. 1580 (New Haven: Yale University Press, 1992); Beat A. Kümin, The Shaping of a Community: The Rise and Reformation of the English Parish c. 1400–1560 (Aldershot: Routledge, 1996); Andre Vauchez, The Laity in the Middle Ages: Religious Beliefs and Devotional Practices (Notre Dame: Notre Dame Press, 1993); Susan J. Wright (ed.). Parish, Church and People: Local Studies in Lay Religion 1350–1750 (London: Hutchinson 1988). 3 Matthew 19:21. 4 Peter Brown has in particularly, as in so many other respects, shed light on this in “The Rise and Function of the Holy Man in Late Antiquity,” The Journal of Roman Studies 61 (1971): 80–101; The Body and Society: Men, Women, and Sexual Renunciation in Early Christianity (New York: Columbia University Press, 1988); Poverty and Leadership in the Later Roman Empire (Hanover: Brandeis University Press, 2002). 5 A seminal introduction to the role of The Three Magi in medieval art is to be found in Hugo Kehrer, Die Heiligen Drei Könige in Literatur und Kunst I-II (Leipzig: Verlag von E.A. Seeman, 1908–9). See also Ernst H. Kantorowicz, Laudes Regiae: A Study in Liturgical Acclamations and Mediaeval Ruler Worship, University of California Publications in History, vol. 33 (Berkley: University of California Press, 1946). 6 General discussion of the theme of simony can be found in Sarah Hamilton, Church and People in the Medieval West, 900–1200 (Florence: Taylor and Francis, 2013), 64–105; Joseph H. Lynch, Simoniacal Entry into Religious Life from 1000 to 1260: A Social, Economic, and Legal Study (Columbus: Ohio State University Press, 1976). Further insights into the intricacies of dealing with money within an ecclesiastical context can be found in Rory Naismith, “Turpe lucrum? Wealth, Money and Coinage in the Millennial Church,” in Money and the Church in Medieval Europe, 1000–1200. Practice, Morality and Thought, edited by Giles E. M. Gasper and Svein H. Gullbekk (Farnham: Ashgate, 2015), 17–37. 7 Matthew 21:12. 8 “Men hwo der vil lade sige messe for sig eller for syne foreldre daa skall han sige til presten: Herre ieg maa icke købe den hellige messe icke mwe wii helder køpsla der om Men siger en messe eller saa mange messer for mig eller mine foreldre eller for den eller den Jeg wil løne ether for ethers umage tilbørlig. Hwo der lader sige en messe han bør ath giffue presten saa meget som han kan skellige leffue den dagh udoffuer met til maad oc dricke Fordi Sancte Pouild scriffuer hwo der tien altered han skall och leffue aff altered.”, Christiern Pedersen, “Om Messen,” in Christiern Pedersens Danske Skrifter, vol. 2, edited by Carl Joachim Brandt and Rasmus Theodor Fenger (Copenhagen: Gyldendal, 1851), 467. 9 Piero Camporesi, The Fear of Hell: Images of Damnation and Salvation in Early Modern Europe (Oxford: Polity Press, 1991); Joanna S. Norman, Metamorphoses of an Allegory: The Iconography of the Psychomachia in Medieval Art (New York: P. Lang, 1988), 124. 10 See this theme followed more closely in Nicola F. McDonald, “Lusti Tresor’: Avarice and The Economics of the Erotic in Gower’s Confessio Amantis’,” in Treasure in the Medieval West, edited by Elizabeth M. Tyler (Woodbridge: York Medieval Press, 2000), 135–56. 11 A presentation of the major church ceremonies and the use of offerings can be found in Susan C. Karant-Nunn, The Reformation of Ritual: An Interpretation of Early Modern Germany (London: Routledge, 1997). 12 To understand the historical developments of the Mass from the Early Church and throughout the Middle Ages, the seminal work of Josef Andreas Jungmann is still the starting point, and this chapter also builds on his conclusions. Josef Andreas Jungmann, Missarum sollemnia: eine genetische Erklärung der römischen Messe. 1, Messe im Wandel der Jahrhunderte, Messe und kirchliche Gemeinschaft (Vienna: Herder, 1962). 13 See Peder Borgen, Early Christianity and Hellenistic Judaism (Edingburgh: T&T Clark, 1996). 14 Jungmann, Missarum sollemnia, 3–34. 15 Charles Joseph Hefele, A History of the Councils of the Church: From the Original Documents, to the Close of the Second Council of Nicaea A.D.787, vol. 1, trans. William R. Clark (Eugene: Wipf & Stock, 2007), 394–402. 16 Petrus Damiani, “Ep. V, 13,” in Patrologia cursus completus: Series Latina, vol. 144, edited by Jacques Paul Migne (Paris: Migne, 1844–65), col. 359D.

52  Martin Wangsgaard Jürgensen 17 See further in Patricia Ranft, The Theology of Peter Damian: “Let Your Life Always Serve as a Witness” (Washington: The Catholic University of America Press, 2012). 18 Peter Browe, Die Eucharistischen Wunder des Mittelalters (Breslau: Verlag Müller & Seiffert, 1938); Gerald Ellard, “Bread in the Form of a Penny,” Theological Studies 4 (1943): 319–46. 19 Coins were for instance already playing an important role in the liturgy in seventh-century Spain. See Ellard, “Bread in the Form of a Penny,” 328; Jungmann, Opfermesse, 10–18. See also more specific considerations concerning the changes in the offertory during the late eleventh century in Joseph Martos, Doors to the Sacred. A Historical Introduction to the Sacraments in the Christian Church (London: SCM Press, 1981), 263–64. For an interesting and authoritative discussion of the monetary development in the eleventh century see Alexander Murray, Reason and Society in the Middle Ages (Oxford: Clarendon Press, 1978), 25–58. 20 Concerning the status of the writings of Honorius of Autun see Valerie I. J. Flint, “The Elucidarius of Honorius Augustodunensis and Reform in late eleventh-century England,” Revue bénédictine 85 (1975): 178–89. 21 Damiani, “Honorius Augustodunensis, Gemma animae,” I, 27, Patrologia cursus completus: Series Latina, vol. 172, edited by Jacques Paul Migne (Paris: Migne, 1844–65), col. 553. 22 Mark 12:41–44. 23 “Qui panis ideo modum denarii formatur, quia Christus panis vivus pretio denariorum venditur”, Damiani, “Honorius Augustodunensis, Eucharistion,” 1256C-D. 24 “Et quia, populo non communicante, non erat necesse panem tam magnum fieri, statum est eum in modum denarii formari vel fieri, et ut populus pro oblatione farinae denarios offerent, pro quibus traditum Dominum cognoscerent.” Damiani, “Honorius Augustodunensis, Gemma animae,” col. 1052. 25 “Panis autem hic formatur in modum denarii, tum quia panis uite pro denariis traditus est, tum quia idem denarius in uinea laborantibus in premio dandus est…”, Guillelmi Duranti Rationale divinorum officiorum, Corpus christianorum/Continuatio mediaevalis, vol. IV, edited by Anselme Davril and Timothy. M. Thibodeau (Turnhout: Brepols, 1995–2000), 443. 26 Ellard, “Bread in the Form of a Penny”; Aden Kumler, “The Multiplication of the Species: Eucharistic Morphology in the Middle Ages,” Res: Anthropology and Aesthetics 59/6 (2011): 179–91. See also Greti Dinkova-Bruun, “Nummus falsus: The Perception of Counterfeit Money in the Eleventh and Early Twelfth Century,” in Money and the Church in Medieval Europe, 1000–1200. Practice, Morality and Thought, edited by Giles E. M. Gasper and Svein H. Gullbekk (Farnham: Ashgate, 2015), 77–91. 27 Peter Browe, „Die Elevation in der Messe,“Jahrbuch für Liturgiewissenschaft 9 (1929): 20–66; Die Eucharistischen Wunder des Mittelalters (Breslau: Verlag Müller & Seiffert, 1938); Jungmann, Opfermesse, 331–240**. See also Michal Kobialka, This Is My Body: Representational Practices in the Early Middle Ages (Ann Arbor: The University of Michigan Press, 1999). 28 The social consequences of the introduction of minted gold and silver are elucidated by Rory Naismith, “The Social Significance of Monetization in the Early Middle Ages,” Past & Present 223 (2014): 1, 3–39. 29 The theme of the intercessory powers of offerings can be followed in Jacques Le Goff, The Birth of Purgatory (Chicago: University of Chicago Press, 1984). 30 “Voventur autem omnia, quae offeruntur Deo, maxime sancti altaris oblation,” Augustine, Epistle 149, 16, in Corpus Scriptorum Ecclesiasticorum Latinorum, vol. XLIV, ­edited by Alois Goldbacher (Vindobonae: Tempsky, 1895–1923), 363. 31 Gullbekk et al., Kilger, Klackenberg, this volume, chs. 1, 5 and 11. 32 Svein H. Gullbekk and Anette Sættem, Norske myntfunn 1050–1319. Penger, kommunikasjon og fromhetskultur (Oslo: Dreyer forlag, 2019), 136–69; See the conclusion of Henrik Klackenberg, Moneta nostra: Monetarisering i medeltidens Sverige (Stockholm: Almqvist & Wiksell International, 1992), Kilger and Myrberg Burström, this volume, chs. 5 and 11. 33 A theme which is discussed and analyzed in Klackenberg, Moneta nostra and Gullbekk and Sættem, Norske myntfunn.

Pious gifts: coins and church interior  53 34 A discussion of donations and the use of donations as a way to display piety in the church interior can be found in Martin Wangsgaard Jürgensen, Changing Interiors: Danish village churches c. 1450 to 1600 (Copenhagen: Faculty of Theology, University of Copenhagen, 2011), 494–504. 35 The most recent survey of this in a Danish context has been carried out by Ebbe Nyborg, while Axel Bolvig has discussed the parish church and its founding. See Ebbe Nyborg, “­Enkeltmænd og fællesskaber i organiseringen af det romanske sognekirkebyggeri,” in Strejflys over Danmarks bygningskultur: Festskrift til Harald Langberg, edited by Robert Egevang (Copenhagen: Nationalmuseet, 1979), 37–64; Ebbe Nyborg, “Kirke – ­sognedannelse – bebyggelse,” Hikuin 12 (1986): 17–44; Axel Bolvig, Kirkekunstens storhedstid: Om kirker og kunst i Danmark i romansk tid (Copenhagen: Gyldendal, 1992); Ebbe ­Nyborg, “Kirke, sogn og bebyggelse o. 1000–1300,” in Marsk, land og bebyggelse gennem 10.000 år, vol. 1, edited by Stig Jensen (Aarhus: Aarhus Universitetsforlag, 1998), 191–210. See also Beat A. Kümin, The Shaping of a Community: The Rise and Reformation of the English Parish c. 1400–1560 (Aldershot: Scolar Press, 1996). For Sweden see Burström, this volume, ch. 11. For Gotland see Anders Andrén, “Vem lät bygga kyrkorna på Gotland?” Saga och sed. Kungl. Gustav Adolfs akademiens årsbok (2009): 31–59. 36 See Eamon Duffy, The Stripping of the Altars (New Haven: Yale University Press, 1992), 125; Robert W. Scribner, Popular Culture and Popular Movements in Reformation ­Germany (London: Hambledon Press, 1987). 37 Perspectives on the character of gifts and offerings can be found in Esther Cohen and Mayke De Jong (eds.), Medieval Transformations: Texts, Power, and Gifts in Context (Leiden: Brill, 2001). 38 “Etlîche die sprechent: „der pfaffe ist doch rîche genouc: war zuo solten wir im opfern?“ Got der wil sîn niht entbern, unde dâ von sprichet der wîse man: sacrificate sacrificium, etc.”, “wan ez eht sô grôz dinc und sô guotiu dinc bediutet,” in Berthold von Regensburg, vollständige Ausgabe seiner Predigten‚”Von der Messe,” vol. 1, edited by Franz Pfeiffer (Wien: W. Braumüller, 1862), 499. 39 This trend seems also to become apparent in the churchwardens’ accounts. As Henrik Klackenberg notes the income from offerings at the high altar were very small in late medieval Sweden, Klackenberg, Moneta nostra, 32. 40 See Norman P. Tanner (ed.), Decrees of the Ecumenical Councils, vol. 1 (London: Sheed & Ward, 1990). 41 Adolph Franz, Die kirchlichen Benediktionen im Mittelalter, vol. I (Freiburg: Herder, 1909), 233–34. Particularly from Sweden we know of gifts placed on a side altar, usually the one dedicated to altar of the Virgin Mary. Mentioning of such altars occasionally occur in the sources, particularly from Sweden, as so-called “cake altars” or “cheese altars” reflecting the type of gifts placed on them. See Bengt Stolt, “Kyrkorum och kyrkoskrud”, in Mässa i Medeltida Socken. En studiebok, edited by Sven Helander, Sven-Erik Pernler, Anders Piltz and Bengt Stolt (Skellefteå: Artos, 1993), 161–62. 42 Jungmann, Opfermesse, 3–34. 43 Jonsson, this volume, ch. 13. 4 4 For a discussion of regional differences in Danish rural churches see Martin Wansgaard Jürgensen, Changing Interiors. 45 For more information on this church see “Danmarks Kirker: Sorø amt,” in Danmarks kirker. Bd 5, Sorø Amt, edited by Victor Hermansen, Poul Nørlund and Erik Moltke (­Copenhagen, Nationalmuseet: Gad, 1933–), 962–71. 46 The material concerning the medieval Mass presented here in the text and the plan is mainly based on information drawn from Missale Lundense, paired with the general presentations of the medieval Mass found in Jungmann, Missarum sollemnia while more specific information concerning motions and gestures has been taken from Christiern Pedersen’s book on the Mass: “Om Messen,” 415–76. It needs to be mentioned that Pedersen’s gloss on the liturgy of Mass builds on Johannes Herolt, De eruditione christifidelium (Strassbourg: Georg Husner, 1476) and the collection of homiletic guidelines by Vincentius Ferrarius. A number of other sources have also been employed to a lesser degree, which will appear from the references. See also Duffy, The Stripping, 91–130; Anne Riising, Den middelalderlige prædiken (Copenhagen: Gad, 1969), 237.

54  Martin Wangsgaard Jürgensen 47 Joseph Dyer, “Augustine and the ‘Hymni ante obationem: The Earliest Offertory Chants?” Revue des études augustiniennes 27 (1981), 85–99; Wunibald Roetzer, Des Augustinus Schriften als Liturgie-Geschichtliche Quelle (Munich: M. Hueber, 1930), 114–15. 48 Corinthians (9:7). 49 Note the concentrations of coin finds by the chancel arch in Høre stavechurch, see Risvaag, this volume, ch. 6. 50 This theme can be followed through the study of Dawn Marie Hayes, Body and Sacred Place in Medieval Europe, 1100–1389 (London: Routledge, 2002). On lay people’s access to the choir see Risvaag and Roland this volume, chs. 6 and 7. 51 “ligfolck skulle icke stonde for ner høye altere eller i nesen paa presten men han siger messe Oc besynderlighe skulle icke qwindfolck stonde eller ligge for ner altered”, see Pedersen, “Om Messen,” 462. 52 “Somme synis seg wære megett gudelige, nar presten lader Gudz legomme till syne, oc the tha løbe saa nær, att the samme Gudz legomme kwnne klarlige” see “Een christelig forligelse,” in Skrifter af Paulus Helie, vol. 5, edited by Marius Kristensen (Copenhagen: Gyldendalske, 1935), 285. 53 Pedersen, “Om Messen,” 465. 54 This quote read: “Leikmonnum skal æigi ϸolas at standa i sognhusi hia klerum medan ϸeir flyktia tidir. utan ϸeim einum sem lesa edr syngiaϸ med ϸeim,” in Norges gamle Love indtil 1387 (NGL), vol. III, edited by Rudolf Keyser and Peter Andreas Munch (Christiania: Chr. Gröndahl & Son, 1846), 266–67. See also this quote discussed by Risvaag, this volume, ch. 6. 55 “Decreta qvædam Arhusiensia,” in Samling af Danske Kirke-Love, edited by Grimur Jonsson Thorkelin (Copenhagen: Godiche, 1781), 59; Holger Fr. Rørdam, “Biskop Lage Urnes Synodalstatuter fra Aaret 1517,” in Kirkehistoriske Samlinger II, vol. 3 (1864–1866), 262–91. 56 “to offeryn hir messe peny aftyr the custom of the place…,” in The Book of Margery Kempe: A New Translation, Contexts, Criticism, edited and trans. by Lynn Staley (New York: Norton, 2001), 39–40. 5 7 On medieval collection plates from Gotland, Pia Bengtsson Melin and Kenneth Jonsson, “Medeltida kollekttavlor från Gotland,” Myntstudier 2 (2019): 1–92. 58 To let men offer first would be in tune with most of the church ceremonies where both sexes participated and Honorius of Autun confirms this ideal, when he, concerning the offering in Gemma animae, writes that men should make their sacrifice first, as they are the strongest in faith, then women should sacrifice, as they are weak and finally the clergy should perform their offerings. “Primo viri offerunt, qui fortes in Christo desiguant, quia in primitiva Ecclesia justi (sub persecutoribus dura perpessi) victima Christi occubuerant. Deinde feminæ sacrificant, quæ fragiliores significant; quia tempore pacis ûdeles non eisos sed hostiam laudis Domino immolant. Ad ultimum sacerdotes et ministri offerunt, qui doctores et ductores populi sunt, quia sub Antichristo fideles per diversa tormenta sacrificium vinum se Christo offerunt,” Damian, Honorius Augustodunensis, Gemma animae, I, 28, col. 553. See also Martin Wangsgaard Jürgensen, “Syddør, norddør og det kønsopdelte kirkerum,” Hikuin 36 (2009): 7–28. 59 For coin offerings on altars and altar cloth offerings see Jonsson and Kilger, chs. 13 ch. 5. 60 This process was long and can in this context only be presented in simplified terms. Thus, the radiating chapels added to the choirs of early medieval pilgrim churches and cathedrals as it for example could be seen on Chartres Cathedral from 858 must have added to the acceptance of altars not facing strictly east. See Kenneth John Conant, Carolingian and Romanesque Architecture 800–1200 (New Haven: Yale University Press, 1993), 139–40. 61 Concerning the allegorical interpretation of the corners of the world see Joseph Sauer, Symbolik des Kirchengebäudes und seiner Ausstattung in der Auffassung des Mittelalters, 2nd ed. (Freiburg: Herdersche Verlagshandlung, 1924). See also Wangsgaard Jürgensen, “In the Sphere of Sacrosanctity: Altars as Generators of Space in the Late Middle Ages,” in Ritual Dynamics and the Science of Ritual, vol. 5: Transfer and spaces, edited by Gita Dharampal-Frick, Robert Langer and Nils Holger Petersen (Wiesbaden: Harrassowitz, 2010), 323–38.

Pious gifts: coins and church interior  55 62 Miri Rubin, Corpus Christi: The Eucharist in Late Medieval Culture (Cambridge: Cambridge University Press, 1991). 63 Michael Camille, The Gothic Idol: Ideology and Image-making in Medieval Art (New York: Cambridge University Press, 1989). 64 Hans Belting, Bild und Kult: Eine Geschichte des Bildes vor dem Zeitalter der Kunst (Munich: C.H. Beck, 1990); Sixten Ringbom, Icon to Narrative: The Rise of the Dramatic Close-up in Fifteenth-Century Devotional Painting, 2nd ed. (Doornspijk: Davaco, 1984). 65 In the Eastern Church it had already from the ninth century been common for wealthy families to establish private chapels or shrines housing the graves of deceased family members. Such chapels would almost always be attached as small houses of their own to the church and function as separate ecclesiastical building. Earlier parallels to the developments proposed in this paper can, therefore, be found and should be taken into account when considering altars in the late medieval Western Church. On this see for example ­Jungmann, Opfermesse, 281. Danish and a few Scandinavian examples of this can be found discussed in the important book of Lars Bisgaard, De glemte altre: Gildernes religiøse rolle i senmiddelalderens Danmark (Odense: Odense Universitetsforlag, 2001). 66 Concerning the status and functions of altars see the still crucial work by Joseph Braun, Der Christliche Altar I–II (Munich: Alte Meister Guenther Koch, 1924). 67 Braun, Der Christliche Altar I–II. 68 Nils-Arvid Bringéus, Vallfärder till S:t Olof (Simrishamn: Fören. för fornminnes- och hembygdsvård i sydöstra Skåne, 1997 ). Parallel examples from Denmark of huge rebuilding during the late medieval period due to the income from the donations of pilgrims could be the churches of Boeslunde and Holmstrup, both on the island of Zealand. Find descriptions in “Danmarks Kirker: Sorø amt,” 753-66; “Danmarks Kirker: Holbæk amt,” in Danmarks kirker. 4 Holbæk amt Bind 4, edited by Hugo Johannsen and Marie-Louise Jørgensen (Copenhagen: Nationalmuseets: Gad, 1933–), 1789–856. 69 As a general study on chapels and altars see Antje Grewolls, Die Kapellen der norddeutschen Kirchen im Mittelalter (Kiel: Ludwig, 1999). 70 Note for instance the situation in Bunge Church on Gotland, see Kilger, this volume, ch. 5. 71 Sven Rosborn, “Krämarnas kapell i S:t Petri kyrka,” Elbogen: Medlemsblad för Malmö Fornminnesförening 8 (1978): 21–48. 72 Wangsgaard Jürgensen, “Altering the Sacred Face,” in Resonances: Historical Essays of Continuity and Change, edited by Andreas Bücker, Eyolf Østrem and Nils Holger Petersen (Turnhout: Brepols, 2010), 77–106. 73 As an example the wall paintings in the nave in Bunge church on Gotland, see Kilger, this volume, ch. 5, fig. 7. 74 For rituals that involved both civic and theological elements as well as objects from the church that were carried out into the surroundings see for example Edward Muir, Civic Ritual in Renaissance Venice (Princeton: Princeton University Press, 1981); Ritual in Early Modern Europe (Cambridge: Cambridge University Press, 1997); Scribner, Popular Culture. 75 See for instance the study on this subject by Simon Roffey, Chantry Chapels and Medieval Strategies for the Afterlife (Stroud: Tempus 2008). 76 In Haraldsted on Zealand, for instance, an altar dedicated to St George, facing south in the east end of the longhouse, was dismantled as late as 1816 to make more room in the chancel area, see “Danmarks Kirker: Sorø amt,” 448–50. 77 “Illa signum crucis uno unguium suorum scripsit in porta scilicet quod in illo specialiter monasterio suum recondidisset affectum,” in The Life of Christina of Markyate, a Twelfth Century Recluse, edited and trans. by Charles Hugh Talbot (Toronto: Toronto Press,1998), 38–39. 78 The Book of Margery Kempe, 14. 79 As for instance can be gleaned from the useful collection of references to chancel arch crucifixes from Scandinavia with an attached collection box in Marie Louise Jørgensen, “Pengeblokke,” in Synligt og Usynligt. Studier tilegnede Otto Norn på hans 75 års fødselsdag den 13. dec. 1990, edited by Hugo Johannsen (Herning: Poul Kristensens Forlag, 1990), 71–82.

56  Martin Wangsgaard Jürgensen 80 “Tali quadam arte spargitur aes, ut multiplicetur. Expenditur ut augeatur, et effusion ­copiam parit. Ipso quippe visu sumptuosarum, sed mirandarum vanitatum, accenduntur homines magis ad offerendum quam ad orandum… ostenditur pulcherrima forma Sancti vel Sanctae alicuius, et eo creditor sanctior, quo coloratior,” in Apologia, Sancti Bernardi Opera, III, edited by Jean Leclercq and Henri M. Rochais (Rome: Editiones Cistercienses, 1963), 105. 81 Kilger and Jonsson, this volume, chs. 5 and 13. 82 For several Danish examples see Janus Møller Jensen, Denmark and the Crusades: 1400– 1650 (Leiden: Brill, 2007) and Risvaag, this volume, ch. 6. 83 Renate Kroos already pointed to Pieter Breughel’s painting as a good pictorial source for the concerns of this chapter in an art historical survey of the subject. See Renate Kroos, “Opfer, Spende und Geld im mittelalterlichen Gottesdienst,” Frühmittelalterliche Studien 19 (1985): 502–19. 84 A good introduction to this type of sculptures can be found in Johannes Tripps, Das handelnde Bildwerk in der Gotik: Forschungen zu den Bedeutungsschichten und der Funktion des Kirchengebäudes und seiner Ausstattung in der Hoch- und Spätgotik (Berlin: Gebr. Mann, 1998). 85 Peder Madsen, priest in Ribe in Jutland during the middle of the fifteenth century, mentions this as an abuse in his collection of sermons, Liber Petri Mathie, fol. 56v. See Liber Petri Mathie curati ecclesie Sancti Petri Ripis: The Book of Peder Madsen at St. Peter’s Church in Ribe. 1454–1483, edited by Anne Riising (Copenhagen, n.d.), fol 56v: https://www. sdu.dk/-/media/sidste_chance/files/om_sdu/institutter/ihks/projekter/middelalderstudier/ ny_peder_madsen.pdf (accessed 24 November 2020). For a discussion of coin-offerings as memory, see Lucia Travaini, “Saints and Sinners: Coins in Medieval Italian Graves,” The Numismatic Chronicle 164 (2004): 159–81. 86 See references in Scribner, Popular Culture. 87 This has been a subject of huge discussion, as summarized brilliantly and at the same time continued by Klackenberg in Moneta nostra, 34–40; Gullbekk et al., this volume, ch. 1. 8 8 Kilger and Risvaag, this volume, chs. 5 and 6. 89 Benedicta Ward, Miracles and the Medieval Mind: Theory, Record, and Event, 1000–1215 (Philadelphia: University of Pennsylvania Press, 1982), 96.

Part II

Coin finds in Scandinavian churches

3

The archaeology of church floors Henriette Rensbro

For the antiquarian, the term ‘church floor’ has several meanings. In the broadest sense, it includes everything not visible underneath the present floor surface. This may be described as the space between the outer walls and under the present floor surface. In a narrower sense, the term ‘floor’ describes the surface material, i.e. the flooring. Technically, however, a floor most often consists of both a surface and some kind of sub-construction. If the surface of a floor in a stone church is made of wooden planks, the sub-construction is generally made up of wooden beams placed on padstones. In wooden churches, the beams are often fixed to the wooden sill plates, which are raised above the ground. A floor surface of mortar often covers a stone paving, whereas a surface of floor tiles is always constructed on a levelling layer, which is sometimes a reused older layer from within the church or, occasionally, is a redeposited layer brought into the church.1 When dealing with a floor as an archaeological feature, it is important to use the prepositions ‘in’ and ‘under’ with consideration. Addressing the floor in general, perhaps ‘under the floor’ is a better phrase than ‘in the floor’, as ‘under’ indicates more clearly that this has to do with the dirt, the different layers, constructions and traces of activities under the present floor surface. In some cases, an old floor should be called ‘old floor surface’ to distinguish it from the most recent floor and ‘the floor’ in a broader sense. Following the same reasoning, objects found under a floor surface, e.g. of bricks, are either found ‘in’ or ‘under’ the brick floor, as ‘under the brickfloor-surface’ may mean the same as ‘in the sub-construction of the brick floor’. This distinction is often important for the interpretation of function and dating of a floor layer and of the church in general. Floors in medieval churches vary widely. The surface material may be clay, dirt, wood, fieldstone, mortar, brick or tile – often it is a mixture of different materials. Some floors are thoroughly constructed by levelling layers, sub-construction and surface, while others are just placed on top of an older floor. Many factors influence the choice of floor type including financial capacity, availability of materials and local traditions. Consequently, one must also expect cathedrals to have different flooring than parish churches, and that a twelfth-century floor will be different from subsequent floors from the fourteenth and sixteenth centuries. Furthermore, different regions would have very different preconditions for both construction and material. The following is primarily based on studies in Danish parish churches.2 It is a general assumption that the floor surface material has a crucial impact on the type and number of objects that are to be found under church floors. The difference between an old floor surface of wooden planks with large cracks between the

60  Henriette Rensbro

Figure 3.1 Reconstructed section from the chancel of Malt church. Two coins from the twelfth century were found in layer 3, the sub-construction of a floor surface of stone paving. It was not possible to determine whether the coins were lost, deliberately deposited in the church during construction work or if they were brought into the church in the possibly redeposited soil making up layer 3. The coin from around 1,400 in layer 4 might also have been brought into the church as part of the possibly redeposited levelling layer (4), which most likely is part of a sub-construction underneath an unknown floor surface. Drawing by Knud Krogh, from the annual publication Fra Ribe Amt (1971), 537.

floorboards and a new surface of large stone tiles on mortar is obvious. The latter will allow fewer small items to disappear below the floor surface. In between these two extremes, there is a wide range of variations in the conditions of floors, which influence the chances of the accidentally lost object disappearing. The same conditions would also influence the opportunities for coins and non-numismatic objects to be deliberately placed (offered) under the flooring. The number of finds under church floors is thus dependent on the construction of the church, the material of the floor surface, geographic differences in monetisation and offering,3 not to mention the use

The archaeology of church floors  61 and interior of the individual church. Therefore, information on old floor materials and on the constructions of a floor, as well as differences between concurrent floors in different parts of a church, is of vital importance when analysing distribution patterns and the composition of finds. In Scandinavia, no first-generation wooden churches from the eleventh and twelfth centuries still exist, and excavations reveal very little information on floor constructions and materials. The twelfth- and thirteenth-century stone churches in southern Scandinavia originally had floor surfaces of dirt, clay or mortar. Most of these types of floor surfaces rest on sub-constructions, e.g. a stone padding and/or a levelling layer. In northern Scandinavia, the building of wooden churches continued alongside that of stone churches throughout the Middle Ages. Some of these wooden churches are still in use. They have (and always have had) wooden floors elevated from the ground, and consequently, excavations uncover huge numbers of coins and other objects under the floors.4 In some parts of Scandinavia, many stone churches also have wooden floors, 5 and even in south-west Scandinavia, a few traces of wooden floors are found.6 Most church floors have been intensively used, meaning that the space has been fixed for hundreds of years, as at a tell-site, and hence, most parts have been in use for different actions and functions on multiple occasions. Accordingly, the stratigraphy in some parts of a church can be extremely complex. Furthermore, in some parts of the church, most layers and constructions are disturbed and the preserved parts are scattered over a larger area. The state of conservation of floor layers and other features might vary significantly within a few centimetres, and it is often difficult to identify the relationship between the preserved parts of a layer. Aside from construction activities, the extent to which older contexts under a church floor are disturbed depends primarily on three factors: burial customs, fixed furnishings and the need for open space in the execution of rituals. All three interact, as the location of graves in the church room depends to some degree on the location of fixed furnishings, which thus help to preserve layers in certain areas. At the same time, fixed furnishings prevent traces of new activities from being created. Accordingly, certain burial customs and the need for open space affected the location of fixed furnishings, and thus preservation. The change of floorings may also have had a great impact on larger parts of a church room. A new floor could either be based on top of the older floor, or the older one could be removed. The latter would often produce a lot of disturbance, as well as dirt being removed from the church, causing finds and layers to be absent.7 The removal of floor material can sometimes be detected by analysing the dating of the older and newer contexts and the composition of the finds from the building as a whole. The opposite of removal is the redeposition of layers into churches from the outside. This was sometimes undertaken in order to level or raise a new floor surface in the church. The most obvious indications of redeposited layers brought into the church are all thick levelling layers, which logically must have come from somewhere else (unless, of course, they are made of disturbed older layers). When working with finds under church floors, it is important to identify which of the finds were unintentionally brought into the church along with the dirt to raise the floor surface. In many cases, this is very difficult to detect, but it is crucial to focus on the possible levelling layers and the sub-constructions under the old floor surfaces to assess the risk of ‘contamination’ by finds from a secular context. One method is an evaluation of the find composition in order to distinguish which of the finds are connected to religious

62  Henriette Rensbro activity, which are relevant to congregational visits in church and which are related to secular activities. This distinction is difficult to make as few artefacts are used in churches exclusively. Exceptions include large numbers of fragments of window glass, censers, fragments of altarpieces, crucifixes, all of which were probably only used in churches. Many objects from secular life are found in relatively large numbers in churches. Some (e.g. knives, clothes accessories, jewellery, needles and other sorts of personal equipment) may have been lost or left in the church during mass or other pious activities. Yet other types of finds include slags from iron mining, huge amounts of potsherds or sherds with traces of wear after breaking into pieces and cooking equipment. All of these probably had not been brought into the church on purpose and must therefore indicate dirt from outside being brought into the church. When examining the composition of the finds in a layer or in the church as a whole, contamination from secular contexts is sometimes the most likely explanation. This inevitably influences the evaluation of all finds from this context or the church as a whole. But, then again, dirt brought into the church does not necessarily contain finds at all. In the case of elevated wooden floors in wooden churches, the preconditions are different. Presumably, levelling layers are rare, as is the act of taking dirt out of the church. Such activities are primarily associated with large construction works like grave chambers.8 But, due to space between the wooden floor and the ground when digging new graves, the direction in which the workers chose to throw the dirt, and the extent to which it was returned, has affected the distribution patterns of finds, ultimately whether they were right- or left-handed. Dropped coins disappear much more often into cracks in wooden floors than floor surfaces of mortar, brick and tile. At first glance, this might explain why there are so many more coins under Norwegian stave church floors than under Danish church floors. But wooden floors alone can hardly explain the many coins under the floors in Gotland, where the number of finds makes analysis of distribution patterns relevant regardless of the quality of the information on context. Generally, few coins and other objects are found on or in floor surfaces. Most finds come from layers under or in between floor surfaces. This, for obvious reasons, is particularly the case when dealing with wooden floors. Floor surfaces of clay and dirt can contain coins, while floors of stone, tiles, mortar and so forth are often seen to let small items pass through joints and cracks. In general, small denominations of coins are much more common in finds from churches than coins of larger denominations. The influence of graves on the preservation of church floors and the location of finds is vast, but the number, location and date of these graves vary considerably from one church to another. This is clearly demonstrated in this volume by Hedensted church, where the few graves are situated in the west of the nave.9 In contrast, Aggersborg contains many graves, and some churches house even more graves.10 Unfortunately, there are neither comparative studies of the influence of graves on the location of finds nor studies about the differences in burial practices in stave churches and in stone churches with and without wooden floors, and we do not know how this influences preservation. We can only speculate about whether the construction of graves has not moved coins very far in two dimensions, even though the layers are disturbed, which is good news for the use of distribution plans.11 The key questions concerning coin finds in churches are: where, how and why? Step one is to locate in which parts of the church, and in what contexts, coins are found. Step two is to analyse how they ended up under the floor. From there, it is possible to

The archaeology of church floors

63

investigate both the interior design and the use of the church building. It also allows a discussion on why coins were brought to the church and for what purpose. It is necessary to assess the value of the archaeological finds by analysing the formation of contexts, i.e. whether a layer was grown on the spot, was redeposited from another place in the church or whether it was brought into the church. Naturally, finds from redeposited contexts (often levelling layers) are a different kind of source for the history of the church and for numismatics than finds from accumulated layers, e.g. construction layers and dirt layers on or under floor surfaces. The experience of a large number of excavations provides an empirical basis for concluding that it is often difficult to detect whether layers are redeposited or are created by disturbance of older contexts on the spot (see Figure 3.1). Despite these difficulties, one should always attempt to identify the nature of the layers. Disturbances might also be a source of information on activities like new furnishings and changes in liturgy. For many reasons, the registration methods of past excavations rarely focused on context and formation. This is a challenge that has been handled differently in the various reinvestigations of archaeological excavations presented in this volume. To produce usable excavation results requires both appropriate methods of excavation and appropriate methods of registration. The ideal choice to best serve future research in and under church floors is no doubt the use of a combination of single context and square metre grids along with a clear focus on the formation of contexts. Unfortunately, this is a time-consuming and expensive method. Therefore, in cases when one has to be less ambitious, it is of immense importance to choose the excavation strategy and methods with consideration for what is expected to be found and how it might be used in the future.12 It is even more important to allow for a change in methodology should the excavators realise that the object is different from what was anticipated.

Notes 1 For a discussion of the archaeological methods in particular churches, see Roland, Rensbro, Moesgaard and Ingvardsson, this volume, chs. 7, 9 and 10. 2 Mouritz Mackeprang, Vore Landsbykirker, 2nd ed. (Copenhagen: Andr. Fred. Høst & son, 1944); Olaf Olsen, ‘Rumindretningen i romanske landsbykirker’, in Kirkehistoriske Samlinger, 7 6 (1965): 236–57; Birgit Als Hansen, ‘Gulve – i kirker’, in Bolig og familie i Danmarks Middelalder, edited by Else Roesdahl (Aarhus: Aarhus Universitetsforlag, 2003), 247–53; Birgit Als Hansen and Henriette Rensbro, ‘Kirkearkæologi – fra pionertid til nutid. 50 års arkæologiske undersøgelser af kirkegulve’, Nationalmuseets Arbejdsmark (2007): 165–80 (English summary); and Henriette Rensbro, Spor i kirkegulve – De sidste 50 års arkæologiske undersøgelser i kirkegulve som kilde til sognekirkernes indretning og brug i middelalder og renæssance, PhD thesis, unpublished (University of Aarhus, 2007). 3 ‘Offering’ is here used in a broad sense. This is explained by Klackenberg as offertory wastage, see Henrik Klackenberg, Moneta nostra: Monetarisering i medeltidens Sverige (Stockholm: Almqvist & Wiksell International, 1992). Martin Wangsgaard Jürgensen regards all uses of coins in churches as religious acts. This primarily relates to the Middle Ages, but might also to some extent apply to later periods. Martin Wangsgaard Jürgensen, this volume, ch. 2. 4 See Jon Anders Risvaag, this volume, ch. 6. 5 Christoph Kilger, this volume, ch. 5. 6 Gitte Tarnow Ingvardson, this volume, ch. 10. Svend Bruun Jørgensen, ‘Rapport fra undersøgelsen af nogle gulvlag i Søndersted Kirke’. Fra Holbæk Amt (1977): 71–79; Niels Knud Liebgott, Dansk Middelalderarkæologi (Copenhagen: Gad, 1989), 142. 7 Rensbro and Moesgaard, this volume, ch. 9.

64 Henriette Rensbro 8 For a case where an old wooden church was dismantled and replaced by a stone church at Arby in Småland, Sweden, involved levelling and coin finds, see Myrberg Burström, this volume, ch. 11. 9 Michael Andersen and Hans Mikkelsen (eds.) Hedensted Kirke. Undersøgelser og restaureringer, Nordiske Fortidsminder, serie B, vol. 26 (Odense: Syddansk Universitetsforlag, 2016); Tarnow Ingvardson, this volume, ch. 10. There may have been graves in the centre aisle that were disturbed by a modern heating system. 10 For example, Henriette Rensbro, Nils Engberg and Jakob Kieffer-Olsen, «Kirken som gik i havet- Udgravninger i Mårup Kirke 1998, 2009 og 2015,» Nationalmuseets Arbejdsmark (2016): 130–43. 11 For discussions of coin finds and burials in churches, see this volume Kilger, ch. 5, Risvaag, ch. 6, Rensbro and Moesgaard, ch. 9, Burström, ch. 11 and Jonsson, ch. 13. 12 Kristensen, this volume, ch. 4.

4 Digitizing the past: A rewarding challenge for the present The digitizing process of analogue data from excavations in medieval churches Steinar Kristensen Introduction A major challenge facing interdisciplinary researchers within the Religion and Money project has been the attempt to bring a more useful coherence to the myriad (and often disparate) archaeological and numismatic “legacy data” sources available on Scandinavian medieval churches. Not only did we all want to better understand the changing relationships between architectural phases and furnishings, the evolving liturgy and inferred attendant behavior, the provenience of coin finds and other portable artifacts – but we also wanted to be able to better compare the situation between the churches at different locations, even across different regions more effectively. This required a new congruence between hitherto incomparable widespread data materials and more effective tolerance for the inevitable lacunae in the data. In order to support the effort, we developed a methodology for importing all relevant data from the past 150 years (mostly printed or written on paper) into a fully integrated, multilayered digital environment, with the objective of creating a more comprehensive overall picture of the history of the churches for use by all of our colleagues in related disciplines. Currently, the most standard and useful approach to archaeological mapping is geographic information systems (GIS). These employ computer hardware, software and geographical data for the storing, management and manipulation of spatial information and provide newly systematic analyses of man-made sites.1 Many different forms of data, when available, can be integrated into a GIS map: excavation plans and stratigraphic elevations, photographs, written reports and lists of objects. When the data are available, it can also integrate high-resolution detail on provenience, i.e., the precise (or more generalized) horizontal and vertical position of multiple artifacts, and features all within their contextual matrices, and can better indicate their spatial relations to each other. The following is an account of the process we undertook, its many challenges and opportunities, and specific decisions made along the way as to how to most effectively use newly emerging patterns. We can divide the process of mapping and use of the powerful GIS process into four main stages: firstly, caching, curating and making the paper documentation digital; secondly, the vectorizing of the information presented in the new digital documentation; thirdly, exploring different options for the spatial analyses of the data; and finally, the selection of the most useful analysis, with emerging patterns

66  Steinar Kristensen making most possible the exploration of new questions via an ongoing hypotheticodeductive cycle.

Stage one: converting diverse analogue data into digital A daunting plethora of many different types of historical, architectural, archaeological and numismatic documents exist for the study of medieval Scandinavian churches, archived in many different places. Excavation plans and photographs were chosen and scanned to TIFF raster format and reports and lists to PDF format. The resolution for all was set at a minimum of 300 dpi (to get the thinnest lines on a paper plan visible digitally, minimum highresolution scanning is necessary). This arduous tactical work was done by the various researchers on each church, as they had the most expertise concerning – and control over – the source materials. The project created and distributed a new worksheet template in Microsoft Excel, with default attribute cells where the researchers filled in information about coins from contexts they analyzed, enabling comparison between the numismatic data within and across churches. Excel was chosen, as most of the scholars in the project were familiar with the software. This enabled a more systematic comparison between weight, mint, coin issuer, date and depositional context (where and when known). A key for locating provenience was established: detailing which excavation unit and specific stratigraphic matrix each coin was found within. This way the position of coins could be potentially mapped in relation to each other and also in some instances it was deemed possible to infer if the coins were found in primary or secondary deposits (e.g., the latter as a result of the re-leveling of church floors). The Excel files were later imported into a Microsoft Access database, structured and checked for errors such as misspelling and observing any uncertainty of spatial relations. For conveniently and quickly sharing large datasets and documentation across the team, the project used DropboxTM , a commercial cloud storage service. After the scanning process was completed, the GIS analyst on the team critically assessed the potential (and any possible drawbacks) in the analogue documentation presented by the researchers. First, it was important to revisit the excavation methods and the quality of the documentation and contextual information provided. Critical in the evaluation was the quality (accuracy) of the spatial relations for the position for the coins. The degree of spatial accuracy of coin finds can be divided into seven classes: • • • • • • •

Coins recorded with exact coordinates (in a local coordinate system). Coins recorded as deriving from archaeological structures smaller than 1 m 2 . Coins recorded as deriving from 1 × 1 m 2 (1 m 2 excavation units). Coins recorded as deriving from larger excavation units or archaeological structures larger than 1 m 2 , with variation in shape and size. Coins recorded as deriving from a major part of the church, e.g., the nave. Coins recorded as deriving from the church. Coins with no spatial reference whatsoever but believed to be associated with the church.

Digitizing the past  67 The coins were usually recorded as being found in excavation units, often 1 × 1 m 2 . In some cases, the units were larger and not always in the shape of squares (see below). Other coins were recorded in relationship to archaeological structures such as graves, postholes or ditches. As demonstrated by the seven classes above, there existed a wide variation in the spatial resolution of the recorded find contexts. This had to be taken into account when using, analyzing and interpreting the datasets. An ongoing consultative, positive feedback loop was set up between the GIS analyst and the researchers, because widely varying excavation and documentation practices, often based on very different theoretical assumptions and pragmatic goals through the past century, meant different strategies for digitizing. Evaluating the extent of digitizing was time-consuming when deciding what to proceed with, and what to omit.

Stage two: vectorization of data The latter part of the digitizing process was the vectorization of the datasets. Vectorization requires that features on the digitized raster are redrawn in the GIS software. Vector data in GIS consist of points, lines and polygons, which are mathematically generated and will be infinitely scalable without losing quality. Raster data on the other hand consist of pixels and will lose quality when enlarged. Our GIS vector files were recorded in the ESRI Shapefile format. The main goal for the vectorization was to create a digital-based layout of the church plans to begin with, and as accurate and precise treatments of the archaeological contexts of coin finds and other portable artifacts as possible. The first step in the vectorization process was to georeference the digitized (scanned) plans. Georeferencing means aligning the raster data to location with the real-world coordinate system; in order to do this, we used the WGS84/UTM32n coordinate system. The process connected positions on the raster data with the same position in the GIS map. A minimum of four reference points was used in our project. For each church, we first aligned a vector grid on top of the building layout as presented in public maps. The grid consisted of 1 × 1 m 2 . Then, the raster-based data on the churches and the known excavations plans were carefully aligned to this grid. By doing this, we could scale the plans correctly and place them within the global coordinate system with an acceptable degree of accuracy. The georeferencing process usually ran smoothly and with a fair degree of accuracy, but in several instances, the raster data contained errors and deviations, which made it difficult to achieve an acceptable level of quality (this will be discussed further below). But in the majority of cases where an acceptable level of accuracy was achieved, the raster plans were used as background maps for the actual vectorizing process. After the georeferencing process, the layout of the church and major architectural features like foundations, stone walls and pillars were vectorized, as well as interior furnishings with known positions, such as altars, pews and baptismal fonts, being included (see Figure 4.1). Secondly, archaeological features and structures revealed and documented during the excavations as cultural layers, burials or postholes were vectorized together with the actual measured recording of methodological interventions during the excavations such as trenches, sections and especially the varying reference grids

68  Steinar Kristensen

Figure 4.1 Example of the vectorizing process showing objects on the digitized plan of Bunge church, Gotland, Sweden. Map by the author.

each with datum used for the actual pinpointed spatial location of the coins and other objects found. Some of the excavations had used a standard m 2 grid as reference grid (excavation units) for finds discovered during the survey. However, some of the archaeological investigations operated with non-regular grids, with rectangles of uneven sizes and shapes ranging from 1 to 35 m 2 . Others operated with spatial relations using archaeological structures as the reference object, ranging in size up to 10 m 2 . Some cases even combined two or more of these systems together at the same site. The last step in the digitizing process was to plot the positions of the coins (and other portable artifacts) found in relation to them in situ, and as shown above, there were several categories of accuracy in the spatial resolution.

Digitizing the past  69

Stage three: exploring different options for spatial analyses In order to perform the most effective spatial analysis on the coins based on attributed data for each and every known coin in situ, including consideration of such things as the year of minting or place of origin of the coin, different approaches to spatial analyses were discussed with the entire team, as there are several methods for visualizing spatial patterns based on a quantitative dataset. One method was to show the overall distribution of the coins through colored polygons (see Figure 4.2). The polygons, either squares or archaeological structures, were color-coded according to the number of coins related to (found within) those polygons (e.g., the higher the number of coins, the darker the color of the polygon). The method involves database queries. Database queries can be described as questions asked toward the datasets. Depending on the analysis, different criteria are set to fulfill the selection. A typical query could be that we want the database to return a list of objects that, e.g., were minted in Norway in 1290. In this case, the criterion would be “Country = Norway AND Year = 1290”. Database queries can be simple, or they can be quite complex depending on the research question and the analysis. Using complexed queries with colored polygons is a time-consuming method because only one combination of attributes could be visualized at the same time, and the number of categories to compare in dyadic relationships of this kind is often considerable, sometimes very large. With this in mind, the actual method could be more suitable as a background simply displaying the total amount of coins found within the polygon. A second mode of displaying the amount of coins found within each square was with dots of different sizes in the center of the squares (see Figure 4.3). The size of the dot indicates the number of coins. This mode is not suitable for combining several different patterns, as the dots will overlay. Adding information only increases the numbers of queries and thereby limits the full functionality of the GIS as with the polygons discussed above. Although these two approaches already could provide illuminating general like-forlike comparisons of floor usage between different churches, neither of these methods were deemed adequate for the more complex analyses the project wanted to perform. Therefore, the next step was to give every single coin a unique representation of a pinpointed location in the map. Some of the coins had exact coordinates within the excavation, and for them, it was a straightforward operation to create positional points based on their local coordinates. For the others, a different approach was required. We had to create as many dots within the squares or archaeological structures as there were numbers of coins. In this way, a single coin would be represented by its own point, creating the possibility of examining patterns based on a single coin’s attribute information, to start with, or the same information across many coins. We used a function called “random points” in our GIS software toolkit to generate these points. We summed up the total number of coins related to the square or archaeological structure, and then, the tool created that number of points within the polygon’s boundary. These unique, randomly placed points and the one that had precise known provenience (coordinates) were later joined to each unique coin in the database. After joining, it was possible to do queries directly on each single coin for analysis, based on their attribute values as shown in Figure 4.4, where all coins in Bunge church are randomly placed, and the four coins from Norway are queried out and marked with red points.

70  Steinar Kristensen

Figure 4.2 Quantitative analysis of recovered coins presented by their find context as color-coded polygons. The color indicates the number of coins found. Bunge church, Gotland, Sweden. Map by the author.

Digitizing the past  71

Figure 4.3 Quantitative analysis of recovered coins presented by their find context as center points. The varying size represents the amount of coins found. Bunge church, Gotland, Sweden. Map by the author.

72  Steinar Kristensen

Figure 4.4 Quantitative analysis of the number of recovered coins presented by randomly placed points within their find context. Each point (coin) has unique information according to the finding and cataloguing data. Bunge church, Gotland, Sweden. Map by the author.

Digitizing the past  73

Stage four: choosing our weapons All three methods described above have their pros and cons. The methods with colored polygons or center dots represent abstract visualizations of the find context and do not indicate any precise location of the finds beyond the fact that they were found within the polygon boundaries. These are positive in the sense that they give the reader a general picture of distribution. But both methods are also limited regarding the flexibility of specific queries and visualization for the reconstruction of close-athand human micro-behavior. Only one or two queries can be displayed simultaneously, as the symbols will overlay each other. The queries are also time-consuming. The project chose the third method using the random points as this produced the spatial image of displayed patterns of the accidental losses of coins best. This was especially the case where hundreds of objects were plotted within a limited area inside particular churches. In this way, all the coins could be visible in the map at the same time without overlying each other. This method provides a spatial find distribution that is a product of mathematical algorithms, rather than specific archaeological observations alone. One of the most intense debates within the project as regards GIS analysis was about the use of this random point method. The main objections to the method were that the distribution maps provided merely an impression of the accurate find distribution of each coin. As it can show patterns and clusters that are not proven, it is indeed a qualified objection. But as the benefits of being able to do queries on every single coin and their general related stratigraphic contexts – which would suggest their temporal relationship to each other – were deemed to far outweigh the disadvantages regarding the visualization, the project team chose this method as the preferred one. The challenge is to use the randomly placed coins carefully in interpretation and to be explicit about the method when discussing the results of the analysis. The pursuit of accuracy During the process, two issues provided both challenges and a deeper understanding of the processes of digitizing, and these were enlightening to us: When georeferencing the raster excavation plans, the pursuit of accuracy starts. You want the results to be optimal, but this is not always the outcome. What degree of accuracy can you expect? Sometimes, the raster twists, bends and simply does not to fit the coordinates. This could be caused by measuring errors in the analogue plan, or distortions occurring during the scanning of the paper plans. Another factor that affects the result of the georeferencing is the scale of the original plan. This comes to the surface as you are trying to “stretch” a relatively small-scale map up to “1:1 scale” in the GIS. For example, the original plan of the church of Bunge on the island of Gotland in Sweden was in scale 1:300. The drawn lines on the plan had a 0.3 mm thickness. This is equivalent to 9 cm in “real life”. When digitizing the middle of the drawn lines of the raster, there would be a minimum of 4, 5 cm deviations depending on whether you were interpreting the line in relation to the inside or outside. So, when the accuracy achieved approached 15 cm, this was regarded as satisfactory. A similar example is another church from Gotland, Västergarn (which is not presented in this volume). The church is smaller than Bunge and its plan was at scale 1:100 with a line thickness of approx. 0.5 mm. For this church, we have accurate measurements from the excavation. Comparing those with the one on the digitized plan gave a deviation of 25 cm over a length of 27 meters.

74  Steinar Kristensen 25 cm inaccuracy sounds a lot, but in that scale, this would be considered marginal, as it is less than 1%. Plans with different scales are also hard to merge into the same dataset and will produce inaccuracies, which must be within acceptable tolerance margins. So one challenge in a digitizing process is to have realistic expectations of the original legacy data and all of the possible new applications of it and to decide what is within the set, acceptable margins of error for the project aims. The fact that the data will never get better than the original is important to remember at every step of the process. The pursuit of accuracy can challenge your focus and be time-consuming. The complexity of excavation units Another topic that caused some challenges was the complexity of excavation units. In some of the churches, it was hard to figure out how and why the documentation was done the way it was. With our 2021 knowledge of GIS, it seems odd that the original documentation in some cases is missing aspects that we today consider as necessary and all-important. Although we may allow ourselves to question the methods of our predecessors, it is still important to respect the way our past colleagues conducted their observations and recorded their documentation, in their time. In the stave churches like Høre and Ringebu (latter not in this volume), both from Norway, and also in the stone church of Aggersborg in Denmark (in this volume), the coins were related to squares of different sizes and shapes. In the Ringebu church, excavated in the beginning of the 1980s, more than 800 coins have been found. The accuracy of the coin positions ranges from precise coordinates to excavation units up to 28 m 2 , but the whole church is less than 200 m 2 (see Figure 4.5). The size of the area where the coins were retrieved will of course affect the analytic possibilities and what kind of spatial analysis could be meaningful from the dataset. This frames and puts limitations upon analyses – e.g., distances from certain objects like altars or the baptismal fonts. The varying sizes of the excavation units are not the only obstacle to overcome. To make it even more complex for us today, the documentation shows that coins have been found and related to excavation units that are overlying each other to some extent. This gives a rather complex picture of the condition of the find.

Conclusions In this paper, we have addressed some issues regarding the digitizing process. The quality of the documentation must, as always, be carefully examined and evaluated as suitable for digitizing. Not all older excavation data are suitable for quantitative, nor qualitative analysis in a GIS. Expectations of accuracy must be commensurate with the quality of the available datasets. The analytic results, and how they are visualized, will and must always be discussed at length with the team and be continually reconciled with the overall objectives of the program. The intersections between the disciplines in a multidisciplinary project like Religion and Money are rewarding. Through different approaches and research agendas, the project evolves, and new knowledge is indeed possible to achieve. But there can also be challenges because of the nature of the different professions and vocations involved. The experience and perceptions of a GIS analyst are not necessarily the same as those of a numismatist, and vice versa. The challenges of digitizing old archaeological documentation, with its flaws and historical context, are daunting. But the setbacks are

Digitizing the past 75

Figure 4.5 The complex layout of the excavation of Ringebu stave church, Norway. Map by the author.

nothing compared to the reward gained in bringing new life to rich legacy data in an archive not far from you.

Note 1 For an introduction to GIS in Archaeology, see James Conolly and Mark Lake, Geographical Information Systems in Archaeology. Cambridge Manuals in Archaeology (Cambridge: Cambridge University Press 2006); David Wheatley and Mark Gillings, Spatial Technology and Archaeology: The Archaeological Applications of GIS (New York: Taylor and Francis, 2002).

5 Moving money, ritual money Studying monetary and ritual space in Bunge church on medieval Gotland Christoph Kilger

Money can only perform its role in intensifying economic interaction if the public has faith in it. If faith is shaken the currency is useless. So too with ritual: its symbols can only have effect so long as they command confidence.1

Introduction The island of Gotland is referred to in late medieval world maps as the island of the 90 churches. 2 Between c. 1150 and 1350, Gotlandic society invested enormous resources in the building of stone monuments: churches, stone houses, towers and not least the town wall of Visby. The rural stone churches of the Middle Ages – 96 in all – are visible landmarks to this day and reflect the ambitions and dreams of their time (see Figure 5.1). Most rural churches had Romanesque predecessors, which since the end of the thirteenth century have been replaced with even bigger buildings in the Gothic style. Apart from the actual church buildings, parishioners also invested in lavish furnishings and liturgical equipment such as rood crosses, wooden statues, fonts, frescoes, altarpieces and gravestones.3 Another, less-known feature is the huge amount of coins discovered in ecclesiastical environments on the island. More than 31,000 medieval coins have been found and documented in rural churches.4 Thus, Gotland compared to other regions of Scandinavia has the largest number of such finds, nearly 50% of the total.5 However, the vast number of coins – henceforth “church finds” – has not attracted scholarly interest to the same extent as the well-known Viking-Age hoards or the stone churches. What are the reasons behind this phenomenon? And what functions and meanings did coins have on rural Gotland in the Middle Ages? This study will discuss these issues from an archaeological perspective by examining the distribution of medieval coins and other archaeological finds in Bunge church on the northern tip of the island of Gotland. Covering an area of 321 m 2 , Bunge belongs to the group of large, but not the largest, rural churches on Gotland.6 During renovations in 1971 and 1972, the nave and parts of the tower chamber in Bunge were archaeologically investigated. At this time, the foundations of an older Romanesque church were documented, and extensive archaeological and numismatic material recovered. The rich material included more than 3,200 coins, as well as finds such as beads, fittings, bells, clothes’ fasteners, shards of stained glass window, brass pins. This makes Bunge one of the most archaeologically well-documented churches as well as one of the richest in Gotland in terms of the number of coin finds. The chronological scope of this contribution will encompass the use of coins during the Gotlandic Middle Ages. This period, stretching around 400 years from about 1150 to 1550, corresponds to the period of the domestic Gotlandic coinage. The use

Moving money and ritual money  77

Figure 5.1 Distribution map of medieval churches on Gotland. Churches with more than thousand coins: Gothem (4220 sp.); Bunge (c. 3200 sp.); Burs (3103 sp.); St  Olof’s chapel (c. 2000 sp.); Silte (1672 sp.); Fardhem (1648 sp.); Hamra (1364 sp.); Bro (1092 sp.); Grötlingbo (1057 sp.). Map: Steinar Kristensen, UiO: Museum of Cultural History.

78  Christoph Kilger of early modern coinages commencing after the Reformation from the seventeenth century onwards will be outside the scope of this article. Summing up, as a case study Bunge offers good opportunities for investigating devotional practice and coin use in medieval Bunge parish and when and how an economy based on coins unfolded in a Gotlandic church room in the Middle Ages.

Perspectives and aims Church finds have been studied by Scandinavian medievalists since the 1990s to explore the monetisation of medieval society, a process during which coins became accepted as a common means of payment. Considered as both historical and archaeological sources, coin finds in churches have been mainly discussed within the context of the formation of medieval states and the growing importance of a state-controlled currency, the influence of administered trade and taxes.7 A shift from the central power perspective towards a perspective more focused on coin-using communities and their mentalities appeared in later studies.8 Other potential approaches which have been recently articulated present themselves through the perception of monetisation as a result of ecclesiastical processes, such as parish formation, the building of churches and the implementation of church doctrine on a parish level. In medieval Sweden, the earliest minting coincides with the emergence of a more centralised church organisation.9 Influenced by Gregorian reforms and the papacy in Rome, the Scandinavian church, gradually building up its own organisation, became more independent in relation to royal power. The archdiocese of Uppsala was founded in 1164, and Gotland was included in the diocese of Linköping in the second half of the twelfth century.10 Svein H. Gullbekk has argued from a Norwegian perspective that coins and the coinbased economy evolved into one of the cornerstones of the ecclesiastical economy, acting as a guarantee for its independence and control over its assets. The Peter’s pence, introduced with the foundation of the archdiocese of Nidaros in 1153, is one example of how church decrees and levies initiated the use of coins within an ever-growing part of the population.11 However, the introduction of coinage as a financial means represents only one side of the coin. There is also the question of what ordinary parishioners believed in, under what circumstances they developed a relationship to coined money and monetary value. To underline this approach, I want to stress two theoretical dimensions in the analysis and interpretation of medieval coin use in churches. First, it is important to consider the influence of church doctrine for the monetisation and the developing economy of medieval Europe. Common concerns in Christian beliefs such as the dogmas of eternal sin, hell and purgatory underwent significant changes in the reformed church between the eleventh and thirteenth centuries. According to the English historian Giles Gasper, the doctrine of an economy of salvation, although never explicitly normalised by the church, evolved in this period. In the learned literature, atonement of eternal sin and the collective sinfulness of mankind was conceived in economic terms such as debt and payment. The consequences of sin were expressed in more monetary terms. Eternal damnation was balanced with the prospect of salvation.12 When church organisation was finally established, in practice, each parish and each congregation in Christian Europe was interconnected within the Catholic world system and eventually became accustomed to these ideas.13 An example of how the concept of salvation was communicated to parishioners can be seen in the wall paintings in Gotlandic churches. A popular motif often depicted in the nave shows how the archangel Michael weighs the soul of the German emperor

Moving money and ritual money  79

Figure 5.2 Weighing scene with Archangel Michael and devils, south-eastern wall of the nave in Mästerby church, Gotland. Photo: C. Kilger

Henrik II (1014–1024) on judgement day (see Figure 5.2).14 The paintings convey an impression of how money was conceived in a Christian world view, and that the possession of worldly riches was regarded as potentially sinful. They also support the suggestion that contemporary theological views were probably more influential than worldly powers in the monetisation of rural communities. A second important aspect of coin use in a religious sphere stresses the importance of the church monuments themselves. Parish churches, cloisters and cathedrals acted as dynamic interfaces and nodal points within the Catholic belief system. The church building itself was a focal point in the medieval landscape, and the rituals performed in the church were – as argued by Roberta Gilchrist – important for the “habitual learning” of medieval people since they shaped their habits, disciplined the human body and created meaning in life.15 The church building not only conveyed ecclesiastical doctrine, but also created, through its architecture, interior and not least the liturgy, a physical space and spiritual environment that governed actions and meetings, and – most interestingly for our concerns – enabled exchanges on various levels between people and God. Built in stone, replicas of the heavenly Jerusalem, churches were mediators facilitating, tuning, redistributing and materialising concepts of money. Taking both these aspects together, it can be argued that coin finds in churches were not only media for making payments in cash, but were also considered as tokens of faith for people practising their religion. The questions which will be investigated in this contribution are not only why, but also how the use of money was implemented and unfolded in a religious and spiritual environment at the parish level in Bunge church, as one of many cases within Northern European Christendom. One strand followed in this contribution is how archaeological context in Bunge can provide clues on the development of a monetary economy within a sacred milieu by examining the distribution of the coins in relation to lost furnishings, frescoes and graves. Due to the relatively high quality of archaeological documentation in Bunge, it is possible to analyse and compare the use of coins in the earlier Romanesque church with the later Gothic church. A second strand followed is whether the coin finds can provide information regarding the monetary and liturgical topography in Bunge church. It is important to consider here the original placement and later relocation of the famous

80  Christoph Kilger Lafrans’ offering box, a unique money box from the High Middle Ages that is still preserved. In this respect, the contextual analysis will also investigate the material and spatial relationship between coins and the church’s floors, building structures and so on, which influenced the distribution of artefacts within the church. Methodologically, it will also be necessary to analyse and discuss if and how the numismatic evidence corresponds with the archaeological data. This strand will therefore touch upon source-critical aspects which influence the archaeological record, such as the redeposition of cultural layers through burials and the significance of wooden floors for the deposition of coins and other materials below. And the final strand will address the agency of medieval coin users, the people behind the church building and the coin offering.

Setting the stage – church and parish in Bunge Bunge is situated on the coast of north-eastern Gotland, by Fårösund (see Figure 5.3).16 During the Middle Ages, the church and the parish belonged to “Rute setting” (“the sixth of Rute”) and “Nordertredingen” (“Northern third”). The present church in Bunge consists of a High Gothic church with a double-aisled nave, a rectangular choir without an apse and a vestry on its northern side. Apart from the rafters, which were lowered during repairs in 1857, the building as a whole has been untouched since the Middle Ages. The western end of the building is flanked by a tower, which is the oldest remaining part and originally belonged to the Romanesque church. The crenellated wall and the massive tower enhance the fortified character of the building; this also applies to the cemetery wall which originally was 2.5 m high and built like a fortress. There were loopholes in several places and on the top of the wall were saddle roof-shaped stones set at an angle, just as on the Visby town wall. In the year 1800, the artist Carl Gustaf Gottfried Hilfeling visited Bunge and documented the church, parts of the interior and the rectory.17 The remains of the medieval rectory are found 40 m north-east of the chancel. The rectory, which was built in the fourteenth century, probably at the same time as the Gothic church, was a part

Figure 5.3 Bunge church, northern Gotland. Photo: C. Kilger

Moving money and ritual money  81 of a multi-house complex in which the so-called monks’ house was the main building. In Hilfeling’s drawing, the monks’ house is depicted as a three-storey stone building with a highly embellished Gothic portal and traced stained glass windows. Most of this building was demolished in 1807 (see Figure 5.4). In terms of communications, the proximity to Fårösund harbour must have been of major importance to Bunge parish. During the Early Middle Ages, Fårösund had a key position in the shipping routes connecting Gotland with the Baltic regions and is mentioned as a crusader port in 1210 and 1215 in the Chronicles of Henry of Livonia.18 Fårösund also emerged as a rural port for Baltic traders in early modern times.19 Another important nearby ecclesiastical site was the Chapel of St Olof in Akergarn (see Figure 5.1). This church was situated in the neighbouring parish of Hellvi and became a place of pilgrimage for the whole of Gotland in the thirteenth century.20 We will return to St Olofsholm as written sources provide several detailed accounts of the offering activities at this important religious site within the vicinity of Bunge parish. An archaeological investigation of the Chapel’s choir and the eastern part of the nave was conducted in 2013 and 2014.21 More than 2,000 coins were found, and their contexts described in detail. Since St Olofsholm is close to Bunge, in terms of both dating and location, this site will serve as an analogy for interpreting the Bunge coin finds. According to the seventeenth-century chronicle Chronica Guthilandorum written by the priest in Vall, Hans Nielssøn Strelow, Bunge church was consecrated in 1169. 22 The question is whether Strelow referred to the Romanesque church, the foundations of which were exposed and documented during excavations in the 1970s. However, the absence of finds dating from the late Viking Age suggests that Bunge was established as a church site relatively late in comparison with other Gotlandic churches. 23 The excavations did not reveal any traces of a wooden church or earlier activities. Moreover, no Viking-Age coins are present in the ample coin material from Bunge. Further support for a later date is the large collection of beads recovered during the excavations. More than 200 beads were found, which is exceptional even by Gotlandic

Figure 5.4 Sketch of the medieval rectory at Bunge church, made by CG Hilfeling in 1800. Source: The Swedish National Heritage Board.

82  Christoph Kilger standards. The beads were predominantly of medieval types and there were no obvious Viking-Age examples.24 It therefore seems unlikely that an older Christian cultbuilding stood here, at least not at the site where the stone church was erected. The earliest reference to Bunge church is found in the “taxus list.”25 The medieval list, which is preserved in two sixteenth-century manuscripts, is an account of how much the priests from each parish on Gotland should pay the bishop in Linköping in compensation in the years when he did not visit their churches. Visitations were meant to take place following a rota, according to which half the churches on Gotland were visited every three years, which meant that compensation was paid every six years. 26 Bunge church paid 1marks, while other parishes in Nordertredingen, e.g. Fole and Follingbo, paid 4 marks each, and Lärbro paid the highest compensation of 6 marks. The sum is thus very low in relation to the size of the Gothic church and rectory. However, the evidence of the coin finds from the Gothic church indicates that the parish had a substantial economic capacity. The question is, therefore, does the relatively low taxation in the taxus list in the case of Bunge reflect the circumstances of the fourteenth and fifteenth centuries? What seems more plausible is that the size of the compensation refers to an earlier phase in the history of the parish, probably before the Gothic church was built. In the so-called Visby version of the taxus list, specific years are mentioned in relation to around half the churches, seemingly referring to conversions and extensions of the churches. In the case of Bunge, the year 1221 can be associated with the first phase of conversion, probably when the tower was constructed to the west of the Romanesque nave. The tower constitutes the oldest preserved part of the church and remained after the older Romanesque church was demolished and was integrated into the new Gothic building. However, the Gotlandic churches were constantly rebuilt and there are obvious source-critical problems in matching up the years mentioned in the documents, such as the taxus list and Strelow’s chronicle, with actual building phases.27 The construction of the Gothic church began at the end of the thirteenth century, and this phase seems to have taken gigantic proportions. The building itself shows that the parish had considerable economic assets to start the building process. The Romanesque church with its nave and choir had a base footprint of c. 75 m 2 , while the corresponding footprint of the Gothic church, including the addition of the vestry, was 285 m 2 . This means that the interior footprint was extended almost fourfold (see Figure 5.14).28 Another historical building feature is the consistent and coherent base profile which surrounds the Gothic nave, chancel and vestry, suggesting that all three components were part of a continuous building phase and that there must have been sufficient funds to finish the building. Even the church fixtures show that the local community had significant economic resources (see Figure 5.5).29 All preserved wooden sculptures, woodwork and liturgical items from Bunge belong to the Gothic church and were acquired during the fourteenth and fifteenth centuries. One such example is the well-known Bunge Olof, a wooden figurine portraying the Norwegian patron saint, which probably stood on the southern side altar. Originally, there were more wooden figurines in the church, and according to Hilfeling, there were also larger depictions of saints: St Mary, St Peter and St Paul, and others that were placed in groups in the “nooks and crannies” of the church. Based on the work of human geographer Per-Göran Ersson,30 the archaeologist Anders Andrén has proposed a new interpretation for how ownership patterns around the Gotlandic churches can provide insights into the earlier proprietary conditions, which

Moving money and ritual money  83

Figure 5.5 Wooden sculpture of St Olof from Bunge church, dated to the second quarter of the fourteenth century. Photo: K. Nimmervoll, The Swedish History M useum, Stockholm

in turn can be used to determine who ordered the churches to be built.31 The property situation in Bunge, which might reflect the social geography of the parish in the Middle Ages, is therefore important to consider here. Andrén suggests that there were large farms in almost every parish on Gotland that initiated the building of churches.32 He argues that some of the parish names may have originated from these farms. Some churches were probably also erected in the vicinity of large farms. Another argument relates to property fragmentation, where the land of one farm was intermixed with the land of other farms, which appear quite often on Gotland. Mixed parcels of land in the vicinity of churches are interpreted by Andrén as the traces of large farms.33 However, a re-examination of the property conditions in Bunge points to a different scenario. In the cadastral maps of 1694, most of the farms in Bunge are situated in the southern part of the parish. Parcels of land are very mixed for the farms in the south-eastern corner of the parish along the coast, but fragmentation is noticeably less developed around Bunge church. The church and the church property, its annex, are surrounded by the Ducker and Hägur farms only. In addition, the farm of Utbunge, which may have given the parish its name, is situated more than 2.5 km

84  Christoph Kilger from the church (see Figure 5.6).34 This suggests that the land on which the church was built originally might have belonged to the farm of Ducker or another big farm in the vicinity, and may have already been partitioned off during the Middle Ages. In this particular instance, the property conditions in Bunge parish provide evidence to support Andrén’s suggestion of a large farm that could have initiated the building of the church, but the evidence is not necessarily in line with his arguments. The property surrounding the church was kept together and not mixed up. This suggests a close relationship between the church and the nearby Ducker farm in the Middle Ages. In addition to the early modern cadastral maps, the medieval grave slabs in the church also indicate such a connection. Grave slabs from ecclesiastical contexts are common on Gotland and they provide important information on the history of each church, the congregation and individuals buried inside the church building.35 During the first known inventory, in 1830, there were ten grave slabs in the choir, of which four bore inscriptions in Gothic script. The inscriptions marked the graves of the clerical family who held office in Bunge church during the fourteenth century. All of them were originally placed in front of the high altar in the choir and all persons buried there were in some way related to the priest Botulf Ducker. The inscriptions suggest that there was a close relationship between the church, the priesthood and the Ducker farm.36 The oldest preserved gravestone is dedicated to Johannes, Botulf’s father, who according to the inscription died in 1294.37 That the layman Johannes of Ducker was buried in front of the high altar suggests that he was a prominent figure in the parish and that he as the head of a leading peasant family might have been a church benefactor. If there was a transfer of property in the parish between the Ducker farm and

Figure 5.6 T he central area in Bunge parish. Rectified fair copy of the GM 1700 map, Land survey board archive. Source: Svedjemo 2014, Landscape dynamics, Figure 5.21.

Moving money and ritual money  85 the church in the form of one or several donations, this may have taken place at the beginning of the fourteenth century when the construction of the Gothic church and rectory commenced. A change in ownership would mean that the parishioners were more closely engaged and took more economic responsibility. According to the Gotlandic tax book from 1653, Bunge parish had 12 farms which paid taxes to the crown.38 These farms were also represented in the church, where every farm had its own pew.39 Stux, Ducker and Änge were among the largest farms in the seventeenth century. It is, however, somewhat uncertain whether the farms, their number and taxation in more recent times are an accurate reflection of the situation during the Middle Ages. An analysis of maps and the landscape of the farm at Hultungs in Bunge parish suggests that there were more farms in the parish, which may have been abandoned in the Late Middle Ages as a result of the Black Death.40 However, it is reasonable to suggest that the majority of the coins offered in the church passed through the hands of the individuals living on the farms of the parish. At the end of the fourteenth century, the newly built Gothic church and its interior were reshaped once more. The walls were covered in pictures and friezes connecting the choir and the nave. The myriad figures and representations form one of Gotland’s largest collections of late medieval frescoes (see Figure 5.7).41 The paintings on the northern wall also create a spatial effect, dividing the northern half of the nave into separate zones or compartments. This compartmentalisation also becomes apparent in the distribution of finds under the church floor, where concentrations of coins, brass pins and beads correspond with the paintings (see Figures 5.23 and 5.25). The friezes in the nave and choir depict various saints’ legends, and the scenography is dominated by saints in the foreground and a depiction of God’s city, or castle in the sky, in the background. The systematic and even placement of the figures seems to follow a well-planned programme. The depictions of the saints and apostles in larger-than-life-size mimic statues placed in front of altar screens or retables. The regular placement of these enshrined saints in the nave and the choir amplifies the three-dimensional effect of a painted ritual scenography. The paintings appear to have been created by a master with a background in the eastern Baltic style and who may have been active when the Teutonic Order occupied Gotland.42 A frieze on the northern wall of the nave below the saints depicts a knightly battle, which according to current interpretations depicts the martyrdom of Achasius in the fourth century, and the last battle of the Theban legion.43 Three large scenes along the northern wall are what catches one’s eye when entering the church: the Crucifixion over the northern portal in the west, the Gregorian Mass depicting the Pope’s vision of the Man of Sorrows in the middle and a scene where Christ is taken down from the cross at Golgotha in the east. The middle scene is dominated by the Arma Christi motif showing the suffering Christ standing in a field, coloured red by blood, behind an altar-shaped grave surrounded by instruments of torture. Gregory’s vision of the Man of Sorrows was not a common feature in the Nordic countries until the fifteenth century, which makes Bunge one of the earliest churches in Scandinavia with this popular mystical motif of the Passion of Christ (see Figure 5.8). On the way out of the church, the visitor passed beneath the Last Judgement depicted on the western front of the nave, with Christ portrayed as the world’s judge in the centre, with an illustration of Hell with devils hanging in the balance weight, on the right, reminding the church visitor of death, judgement and purgatory. On the way out, the visitor would also enter the tower and the ringing chamber and finally leaving the church through the western porch. The walls of the ringing chamber in the tower are covered in graffiti and runic inscriptions with, e.g., personal names,

Photo: C. Kilger/S. Kristensen/Frida Roland

Figure 5.7 Frieze on the northern wall of the nave in Bunge church, c. 1400. Main scenes: A) Crucifixion, B) Gregory’s vision of Arma Christi, C) Gologotha.

86  Christoph Kilger

Moving money and ritual money  87

Figure 5.8 Gregorian mass and the Arma Christi scene. Photo: C. Kilger

magical incantations, writing exercises, dates and saints’ names.44 Inscriptions of these types are commonly found in the ringing chambers of the Gotlandic churches. They are popular manifestations in a liminal zone of the church, which in Bunge constitutes a balancing element to the frescoes of the nave and choir. However, the archaeological finds from Bunge suggest that the ringing chamber also had significance as a zone where people carried out devotional exercises such as money offerings. The runic inscription which reads Den helga kyrkan (“The holy church”) on the southern wall of the ringing chamber appears in the Creed, and in other confessions of faith from the fifteenth and sixteenth centuries, may cast some light on the significance of the church in the popular mind in the Late Middle Ages. The inscription is found to the right of the stairs leading up into the tower and on the left-hand side where a collecting box may have been placed.

Lafrans’ offering box A well-known monument of Bunge church is a collecting box made of grey limestone with runic inscriptions. The damaged and incomplete inscription on the front of the box reads: “Lafrans made this stone.”45 Known as Lafrans’ offering box, the object probably served as a base or pedestal for a crucifix, which is now lost. The construction and ornamentation of the box suggest both a monetary and ritual function. Its use as a container for coins is indicated by a crude iron plate with a coin slit set into the top of the box, and a cube-shaped space in the capital that could be opened and closed with an iron hatch (see Figure 5.9). Tracing the placement and use of the box, through the coin finds, provides an opportunity to study the development of coin offerings in the church. It will be therefore necessary to present a more detailed account of the collecting box and its appearance, functions and interpretations. The pedestal is 112 cm high and consists of a base, a shaft and a die-shaped capital. Two sides of the capital are decorated; on the front is the main scene the Majestas Domini motif, with Christ standing in a mandorla. On the other ornate side, to the

88  Christoph Kilger

Figure 5.9 Lafrans’ offering box, second half of the thirteenth century. Photo: C. Kilger

right of Christ, there is a lion with a crown on its head placed within a medallion. To the left is the iron hatch, while the back of the capital is crude and without decoration. Two iron brackets, the lower one still intact, are set into the rough surface. Three sides of the base are decorated with leaves, while the fourth side has an animal mask. When the box was moved outside the church in modern times, a sundial was mounted on its top. The art historian Johnny Roosval drew attention to the collecting boxes, which are a common feature in Gotlandic churches and are popularly referred to as “mite” or “alms” boxes. According to Roosval, the collecting boxes were originally used as supports for rood crosses, and he therefore introduced the term “rood-cross pedestal” and examined their place in the churches.46 Since this time, the placement of the rood cross and its use as a collecting box has been debated by art historians. Torsten Svensson argued that Roosval’s terminology is misleading since it primarily emphasises their function as supports for the crosses, but not necessarily their ritual importance. Through references to written sources, the choices of motifs and decoration, Svensson emphasised the use and importance of the boxes in the Christian cult of offering. The use of Lafrans’ offering box as a liturgical object has also been discussed by the art historian Anna Nilsén. She adhered to the prevalent opinion that the offering function of the box is secondary and was not introduced until the Late Middle Ages. She argued that the

Moving money and ritual money  89

Figure 5.10 Unornated back of Lafrans’ offering box with iron brackets. Photo: C. Kilger

coin hatch on the side was not properly mounted, indicating that it is a late addition. Also, the crude construction of the hatch does not correspond to the artistic decorations of the box. The iron brackets on the back are placed behind the stone’s mid-point, which is where one would expect the cross to be placed, and does not fit the symmetry of the stone (see Figure 5.10).47 Another suggestion is that the iron brackets may have held an offering plate. Indeed, an old sculpted plate disappeared during the renovations of 1857, and this may have been mounted onto the box.48 Re-examination of the stone’s design, however, makes Nilsén’s argument that the box was initially not used for offerings rather implausible. It seems more likely that the compartment in the capital is an original feature created at the same time as the box. The crucifix may have been mounted on a broad wooden base, which in turn rested on top of the capital.49 The wooden base or the rood cross itself may have had a coin slit connected to the coin compartment. The iron brackets did not necessarily support the cross itself, but may instead have been mounts for a wooden support, to which the crucifix was attached. On the whole, it seems likely that the box had been made and put into use in the second half of the thirteenth century, with an earliest possible date around 1250.50 It is not clear when the box fell out of use. The Catholic offering traditions remained strong in many of Gotland’s rural churches as late as the seventeenth and the beginning of the eighteenth centuries.51 Interestingly, there is a lacuna in coin finds from Bunge church in

90  Christoph Kilger the period between the 1560s and the 1620s, which coincides with an edict issued by the Protestant church authority of Gotland. The Visby general chapter had decided that the offering boxes were to be removed and replaced by mite boxes.52 It is possible that the Lafrans’ box was transformed from a Catholic offering box into a Protestant poor box in this period. In any case, the box is mentioned again for the first time in the eighteenth century, in Bishop Jöran Wallin’s book, Analecta Gothlandensia Walliniana. When Wallin visited Bunge, the box had been placed outside the church in the cemetery.53 The rood cross was a focal point for devotional practice.54 As such, the cross constituted one important element in the ritual landscape of the nave and was probably on equal terms with the side altars. There is still some uncertainty where these crosses were originally placed. This concerns especially the characteristic Gotlandic ring crosses which are believed to have been produced during the thirteenth century.55 In this respect, it is important to stress that the box served as a pedestal supporting and emphasising the rood cross, and as a collection and storage box for coins, thus linking liturgy to coin usage. The pattern caused by the spillage of offerings encircling the box will be analysed in the coin find analysis. Here, the possibility of pinpointing the original placement of the box in the Romanesque church is discussed, as well as tracking its subsequent movements later in the Gothic church.

Renovations and archaeological investigations The Reformation and renovations in modern times changed the spatial layout of the church room and radically altered the original medieval appearance of the church. The wall paintings in Bunge church not only tell stories of saints and apostles, but they also skilfully depict towns, towers, shrines, wall carpets and figures in three dimensions, and through this convey an impression of the original medieval interior now lost. A closer look at the renovations provides the reader with an idea of the source-critical challenges that arise when analysing the material culture below the floorboards. The earliest documented renovations of the church in the modern period took place in the second half of the seventeenth and the first half of the eighteenth centuries.56 The most profound changes were made during the renovations of 1856 and 1857.57 There is no record of what happened to the medieval grave slabs in the church. A few were broken up and used as fill when niches in the choir were blocked off. A row of pews in the middle of the church was removed and joined to the present-day pews. This division and moving of the pews to the walls was probably the most significant change in the nave. In order to make room for the pews in the north-eastern corner of the nave, the northern side altar, which was situated to the left of the chancel arch, was demolished; the altar base was truncated and fitted into the choir floor.58 At the end of the nineteenth century, iron stoves were installed and flues were drilled through the walls. These stoves were placed in the corners of the nave close to the earlier side altars, thus possibly disturbing earlier archaeological remains (see Figure 5.20). During restorations in 1917/8, attempts were made to reverse some of the changes made in the nineteenth century and a new restoration plan was developed with the Swedish National Heritage Board.59 In 1937, central heating was installed, when, e.g., a flue was installed between the vestry and the nave, and this connected to a boiler which had been constructed on the outside of the northern wall of the nave. An archaeological survey of the partly exposed foundations of the eastern part of the Romanesque church was carried out at this time.60 Plumbing connected to the central heating system was also installed, which means that floor layers are likely to have been disturbed in the northern part of the aisle and behind the western pillar.

Moving money and ritual money  91 During the restorations of 1971/72, an archaeological survey of the entire nave and parts of the tower was carried out. The church was investigated by RAGU61 in the late autumn of 1971 and early spring of 1972, under the direction of archaeologists Waldemar Falck and Ragnar Engeström (see Figure 5.11). The excavations at Bunge are characterised by the detailed documentation of all the finds pinpointed within a square metre grid, which provides a good opportunity to study the distribution of the coins and other artefacts. The investigations also revealed that the archaeological remains have been affected by the various building activities and burials, in both the nave and the tower. In total, an area of 207 m 2 was investigated, comprising the entire nave (192 m 2) and the eastern part of the tower chamber (15 m 2) (see Figure 5.12). There is no final excavation report, only a draft version and notes made during the work.62 The finds are not registered in relation to features or structures, which imposes some restrictions on the assessment of individual find contexts. There is no information available on how the cultural layers were excavated and how the finds were recorded. The available data allow the assumption that each square was excavated, emptied and sieved separately. Small beads with a diameter down to 0.2 mm were retrieved, and coin fragments were also recorded. Recording finds according to layers was not applied in the nave. However, more detailed stratigraphic information is available for the burials in the tower chamber. Since all finds were registered in relation to their grid square, a distribution analysis is, however, still possible. Most likely, not all of the archaeological remains in the nave were excavated. The excavation of the tower was hurried. However, compared to other church excavations on Gotland and in Scandinavia as a whole, Bunge is a relatively well-documented church from an archaeological point of view.

Figure 5.11 T he excavations of Bunge church in 1971/72 carried out by RAGU. Source: The Swedish National Heritage Board.

92  Christoph Kilger

Figure 5.12 Map showing architecture of standing church in Bunge, extant interiors and the extension of the archaeological investigation in the nave and the tower chamber. Map: Steinar Kristensen, Museum of Cultural History, UiO.

Moving money and ritual money  93

The church under the church During the renovations of 1916/17, the foundations of an earlier Romanesque church were found. In 1971, there was an opportunity to survey the entire foundation (see Figure 5.13). The outer wall of the church was originally 120 cm wide and additional 20 cm with the wall base. No wall remains were preserved above ground, and the excavators noted that the northern walls in particular had been severely damaged by burials. The eastern part could not be examined since it extended underneath the current choir floor. The remains of the wall in the choir suggest that an apse once existed. This idea is supported by a slab found in the Romanesque choir, interpreted as the foundation of the altar. This was 160 cm wide and aligned with the mid-axis of the church. It cannot be ruled out, however, that the slab may also have formed the base of a mid-altar, which was placed under the chancel arch in the later Gothic church (see Figure 5.19). It should also be noted that no certain remains of side altars in the Romanesque church have been found. The Romanesque church had four cross vaults which were supported by a single pillar. The pillar base of cut limestone with painted plaster was found under the western pillar in the current church. In addition, two portals, formerly part of the Romanesque church, were reused in the Gothic church: the northern portal, which was originally the southern portal of the nave, and the vestry portal which had served as the southern portal of the choir (see Figure 5.14). The excavation records do not indicate the exact position of the side portals, but two worn slabs inside the entrance of the Gothic church give an indication of the original position of the southern portal. The only remaining interior features of the Romanesque church are the font and Lafrans’ offering box. The baptismal font was created c. 1250 and consists of polished limestone with a simple arc-shaped ornamentation which adorns the outside of the

Figure 5.13 T he foundations of the earlier Romanesque church in Bunge in 1971/72. Medieval grave slab or platform visible in the north-eastern corner of the Gothic nave. Source: The Swedish National Heritage Board.

94  Christoph Kilger

Figure 5.14 Map showing reconstruction of the foundation of the Romanesque church. Map: Steinar Kristensen, Museum of Cultural History, UiO.

Moving money and ritual money  95 63

bowl. In 1916/17, when the entire wooden floor of the church was replaced, the large stone base of the font was found underneath the floor. The stone base was placed close to the western pillar of the nave, in both the Romanesque church and the Gothic church.

The tower: layers and graves in the ringing chamber During the excavations of 1971/72, the eastern half of the ringing chamber was also exposed, an area of 3 × 5 m, which was covered by a wooden floor.64 According to the excavation report, four undisturbed medieval graves were identified (see Figure 5.15).65 Underneath the floor, a dirt layer of deposits, 30 cm thick, was found, containing most of the coins and the other finds from the tower chamber. No layering within the dirt layer was observed, and the distribution of the finds was perceived as even. The same observations were made in the nave. Stratigraphically, all of the burials lay underneath the dirt layer. The grave fill primarily consisted of lime and contained no traces of soil from above the layer, and the graves were thus deemed to be medieval, no later than the fourteenth century. Immediately adjacent to grave no. 1, by the northern wall of the tower, five thirteenth-century coins were found, all in grid square 195. They were all Gotlandic pennies, with hatched crosses, which were minted between the 1220s and 1280s. The coins were found close together, but in different layers.66 Unfortunately, their mutual contextual relationship and their placement in relation to the burials was not further commented on in the report. In the adjacent square 198, a copper thread with 20 beads (small white glass beads and two large jet beads) was found, together with a woven textile band, the mouth of a glass beaker with pieces of yellow prunted glass and blue glass. The consistent composition of the coins and the find contexts may suggest that the coins, the copper thread and beads belong together and were deposited at the same time. Similar fragment of oil bottles was found in Hedensted church, which probably relate to ritual practices.67

Figure 5.15 Medieval burials in the tower chamber. Source: The Swedish National Heritage Board.

96  Christoph Kilger

The Gothic nave: wooden floors, cultural layers and a “platform” The Gothic nave was the largest contiguous area excavated in 1971/72. The wooden floor was removed and the entire cultural layer examined. At the time of the renovations of 1916/17, the whole wooden floor had already been removed. At this time in parts of the nave, three different wooden floors, one on top of the other, were found.68 Under the wooden floor, a layer of decomposed spruce needles, covering most of the nave, was found. The spruce needles were added to insulate from rising damp when the floor was replaced in 1916/17.69 The groundwater level is only one metre below floor level, and the wooden floor and needle layer were separated by a 15 cm gap. The wooden floor was supported by slabs, of which some were in their original position, such as a pillar plinth from the older Romanesque church, which was reused as a support (see Figure 5.16). It is evident that the nave was excavated by removing the culture layers in slices of equal depth or spits and the finds were then documented within these. In the finds register, almost all of the finds from the nave are said to be from layer 1. Some finds were also registered in layers 2 and 3 from a trench along a north–south axis in the middle of the nave (see Figure 5.12). Layer 1 was c. 30 cm thick and consisted of a mixture of soil and moraine. Stratification is visible within the layer, with a spruce needle horizon on top, c. 4–5 cm deep. An incomplete section drawing suggests that there were more coins near the surface of layer 1, i.e. between 0 and 10 cm, rather than further down. In addition, there is no evidence of any stratification separating the earlier medieval and the more recent coins, indicating that the wooden floor was sturdy and that the coins and other finds underneath accumulated over a long period of time. This phenomenon has also been observed in the latest excavations of the ruined church of St Olofsholm, where the exact location of the coins has been documented in relation to the original wooden floor.70

Figure 5.16 Supporting slab of the wooden floor in the nave. Source: The Swedish National Heritage Board.

Moving money and ritual money  97 During the excavations of 1971/72, one single grave slab was found in the nave. The slab, which had grooved edges underneath, was found in the north-eastern corner of the nave with its long side against the northern wall (see Figure 5.13). There were no inscriptions on this trapezoid-shaped slab, and Falck interpreted it as a medieval gravestone. Based on the photographic documentation of the excavations, it seems clear that the slab was still visible and on the same level as the wooden floor from 1916/17. One coin, a half shilling minted in Avesta in 1802, was found underneath the slab and thus could have been moved at the earliest in the beginning of the nineteenth century.71 However, there is also the possibility that the grave slab was found in its original position. The slab was placed on top of a rectangular stone packing, which may have been a foundation for some kind of structure (see Figure 5.17). In the photograph, bits of plaster are visible on the stones adjoining the wall. It seems unlikely that the stone packing is the remains of a stone altar. It is more likely that the packing was the base to a kind of platform that was raised a few centimetres above the wooden floor. The platform was located directly underneath the eastern scene of the passion suite, depicting the removal of Christ from the cross at Golgotha and his burial (see Figure 5.7). Based on the contextual analysis of brass pins, which will be referred to later, it is possible that the area in front of the platform was used for burials from the fifteenth century onwards. The fact that the packing was not removed during the renovations in the nineteenth century indicates that the construction with the grave

Figure 5.17 Rectangular stone packing or platform, northern wall of the nave. Source: The Swedish National Heritage Board.

98  Christoph Kilger slab was still visible at this time. Why it was spared when the northern side altar was demolished is not clear. Similar constructions have been documented in porches of churches in Scania in southern Sweden, where they are associated with paintings showing the tomb of Christ.72 We are to return to the platform and the tomb of Christ below. In conclusion, Bunge church provides a good example of the importance of wooden floor layers for the preservation of the culture layers below. The archaeological context indicates that the floor layers in Bunge probably were not removed. Rather, when altered or renovated, new wooden floors were placed atop the earlier ones. In comparison with church excavations recorded in Denmark, the preservation conditions for cultural layers are rather different. Massive reconstruction work was conducted in Danish churches throughout the medieval and early modern period. The redeposition of layers probably also affected the presence of coins and other small objects to a considerable degree.73 These observations provide a touchstone for assessing the coin finds from Bunge from a source-critical point of view. How representative are the coin finds from Bunge church and to what extent do they reflect the coin circulation in Gotlandic churches during the Middle Ages?

The coins from Bunge church – a numismatic survey In 1972, the first survey and analysis of the coin material from the excavations of the nave and part of the ringing chamber of Bunge church was published by Ragnar Engeström in Gotländskt Arkiv and states that 4,147 coins were found.74 The coins and all coin-like objects were transferred from Bunge church to the Royal Coin Cabinet in Stockholm in 1976 and were classified after conservation. There are 3,260 intact coins in the final finds list, of which 3,236 are numismatically definable.75 The pronounced difference in number probably relates to the fact that fragments were not included in the revised list. These fragments were primarily of the very common W-bracteates. In total, the fragments weighed c. 70 gram. The average weight of a W-bracteate from Bunge church is c. 0.08 gram, and the weight of all the fragments would thus correspond to 875 intact bracteates. This estimate naturally represents the minimum number of bracteates.76 Unfortunately, no weight for the entire coin material is available and it is therefore impossible to estimate the weight of the intact coins compared to the fragmented ones. The calculations presented here suggest that the coin fragments make up nearly a quarter of the entire coin assemblage. The primary research publications dealing with the medieval coin finds of Gotland which have been consulted for the numismatic analysis that follows are outlined below. An overview of the history of Gotlandic coinage by Jens-Christian Moesgaard discusses questions on coin circulation, trade contacts and political relations between Sweden and the Baltic area.77 His research is based on a comparison and analysis of the coin material from 43 churches. Susanne Fridh’s MA dissertation in archaeology from Stockholm University expands upon Moesgaard’s research and deals with all known stray finds of coins and accumulated coin finds from Gotland, dating from the Middle Ages onwards.78 The standard work on the earliest Gotlandic minting is Nanouschka Myrberg Burström’s PhD thesis.79 In later articles, Myrberg Burström also interpreted the thirteenth-century Gotlandic coinages, and this will form the basis for the spatial analysis of coins in the Romanesque church presented in the final section of this paper.80

Moving money and ritual money  99 The coins from Bunge will be presented in three groups. Group I consists of the coins minted between 1150 and 1300, which were in circulation in Gotland during the time of the Romanesque church (Table 5.1). The coins of group II correspond to the minting period from c. 1300 to 1560, representing the coins used in the Gothic church until the time when Gotlandic minting ceased (Table 5.2). Group III contains coins from the seventeenth century to modern times (Table 5.3). In order to discuss how representative the finds from groups I and II are, they have been compared with the numbers presented for greater Gotland. In this context, the figures presented by Fridh are primarily based on church finds but also include stray and accumulated finds from other locations. Despite this, the comparison provides an idea of the proportions of the different coinages, i.e. how common or rare they are on Gotland. Together, the coins from group I only make up 2% of all the coins found in Bunge church. This represents approximately one lost or deposited coin every other year, if we estimate that the Romanesque church existed c. 100–120 years, i.e. from the late twelfth century to the beginning of the fourteenth century. This does not include the period of rebuilding, which may have taken place in the first decades of the fourteenth century. In this group, the “older W-bracteates” are included. W-bracteates were minted in Visby from the c. 1280 to the 1450s but are difficult to date accurately. The older W-bracteates, with a strikingly high silver content, form part of group I and are separated from the younger lower silver content W-bracteates. The distribution pattern shows that the earliest W-bracteates were already in circulation in the Romanesque stone church (see Figure 5.31). About 93% of the coins come from group II (Table 5.2), which corresponds to a substantial rise of c. 12 coins per year in the period from c. 1300 to the 1560s. The absolute majority, over 2,800 coins, represent the late low silver grade W-bracteates. Comparison with other accumulated finds from Gotland shows corresponding proportions, which suggests that the coin finds of Bunge reflect more general tendencies of coin circulation.81 The comparison also provides a good case for arguing that the late medieval floor layers in the Gothic church were largely undisturbed. The coins from group III make up 5% of the total and thus date from the seventeenth to the twentieth centuries. There is a chronological gap; the sixteenthcentury coins stop with a half-shilling struck in 1554 and one German sechsling minted by Duke Johan Albert in 1557. The seventeenth-century coins start with the shillings minted by King Christian IV in 1620 and 1621. This lacuna may be due to the minting of speciedaler in Denmark, which started in 1560. These coins Table 5.1 Medieval coins minted and used during the period of the Romanesque church building c. 1150–1300 (66 specimens = 2%) Gotlandic coinages Gotland, Myrberg I, LL XX (c. 1150–1220) Gotland, Myrberg II, LL XXII (c. 1220–1290) Gotland, early W-bracteates, LL XXXIV:1 (c. 1290–1360) Foreign coinages Pomerania (c. 1250–1325)

Bunge 4 sp. 30 sp. 30 sp.

Gotland (Fridh 2014) 48 sp. 636 sp. ?

2 sp.

?

100  Christoph Kilger Table 5.2  Medieval coinages minted in the period of the Gothic church building c. 1300–1560 (3018 specimens = 93%) Gotlandic coinages Gotland, later W-bracteates, LL XXXIV:2 (c 1360–1450) Gotland, Örtugs (c 1340–1380) Gotland, Örtugs (c. 1380–1420) Gotland, Örtugs (c. 1420–1450) Gotland–Denmark, Hvid (c. 1450–1480) Gotland–Denmark, King Hans, Hvid (c. 1480–1500) Gotland–Denmark, Hvid (c. 1500–1520) Gotland–Denmark, Hvid? (c. 1450–1520) Gotland–Denmark, King Kristian III (1534–1559), Schilling Gotland–Denmark, –”–, ½ Schilling Total

Bunge 2823 sp. (95%)

Gotland (Fridh 2014) 27815 sp. (95%)

   3 sp. (0.1%)    9 sp. (0.3%)   69 sp. (2.3%)    5 sp. (0.17%)    3 sp. (0.1%)

   28 sp. (0.1%)    52 sp. (0.18%)   581 sp. (2%)    52 sp. (0.18%)    15 sp. (0.05%)

  20 sp. (0.67%)    3 sp. (0.1%)   20 sp. (0.67%)

  131 sp. (0.45%) –   343 sp. (1.18%)

  3 sp. (0.1%) 2958 sp. (100%)

  136 sp. (0.47%) 29153 sp. (100%)

Foreign coinages Sweden, King Magnus Eriksson, Penny Lion/Crown LLXXVI (c. 1319–1340) Sweden, – “ –, Penny Lion/Three crowns LLXXVII (c.1340–1354) Sweden, – “ –, Penny Crown/Strålring, LLXXVIII (c. 1354–1363) Sweden, King Albrekt of Mecklenburg, Bracteate Crowned head (c. 1370–1380) Sweden. Sten Sture the older, Bracteate Crowned head (c. 1470–1480) Sweden, King Magnus Eriksson, Bracteate Opposite crowns, LLXXIX (c. 1360–1370) Norway, Kg Magnus Eriksson, Opposite crowns, LL XXIX (c. 1360–1370) Norway, King Hans (1483–1513), Hvid Denmark, Kg Kristian II (1513–1523), Klipping Denmark, King Fredrik I (1523–1533), Penny Baltic states, Livonian knights, Reval (c. 1350–1390) Poland, Kg Kasimir Jagiello (1447–1492) Teutonic knights (c. 1390–1410) Germany, anonymous, bracteate with eagle Mecklenburg, Oxhead bracteate (c 1450–1500) Mecklenburg, Towns Grevesmühlen, Rostock, Wismar, Sechsling (c. 1540s–1550s) Pomerania, Vinkenauge (c. 1350–1450) Total

Bunge    3 sp.   19 sp.    1 sp.    1 sp.    2 sp.

   1 sp.    2 sp.    1 sp.    1 sp.    7 sp.    3 sp.    1 sp.    1 sp.    1 sp.    4 sp.   12 sp.   60 sp.

a Fridh, Mynten på ön, 19, fig. 14. This number includes all known single finds of W-bracteates from Gotland registered by Fridh. b Örtugs minted c. 1340–1360, 10 specimens; örtugs minted c 1360–1380, 18 specimens, Fridh, Mynten på ön, 1.

Moving money and ritual money  101 Table 5.3  Early modern and modern coinages in Bunge church c. 1600–1900 (152 specimens = 5%) Seventeenth century Eighteenth century Nineteenth century Twentieth century Total

  41 sp.   43 sp.   63 sp.   5 sp. 152 sp.

comprise c. 25 gram of pure silver and were thus unlikely to be lost or involved in any actions of devotional character in the church room. It therefore seems unlikely that the gap is an indication that people stopped using coins or that there were fewer church visitors in Bunge. However, the significance of early modern coins as offerings after the Reformation is outside the scope of this contribution and remains to be analysed and discussed in more detail in future. There are no reports of coin finds in the choir with the exception of one. Approximately 40 silver coins were found under some loose planks in front of the main altar during the renovations of 1917.82 A short article in the newspaper Stockholmstidningen on 8 February 1917 provides more detailed information: 42 silver coins were found when a few wooden steps in front of the altar in the choir were removed. These coins were primarily Gotlandic, dating between the fourteenth and seventeenth centuries.83 It is not known where these coins are today. No further reports of coin finds in the choir are known. This numismatic overview shows that the use of coins increased considerably in the Gothic church. The W-bracteates make up the majority of finds. Since the ­W-­bracteates were minted in the same style for c. 160 years, they are difficult to date with any degree of accuracy. The figures presented above give an indication that coin use in Bunge started in the thirteenth century in the Romanesque church. However, the real take-off in terms of a more frequent use of coins in Bunge happened in the late fourteenth to fifteenth centuries. A similar development can also be suggested for other churches on Gotland based on the evidence of the W-bracteates, which totally dominate coin circulation in all excavations of ecclesiastical contexts. In the late thirteenth and early fourteenth centuries, we see the commencement of new building projects taking place at almost every rural church on Gotland: the rebuilding of the Romanesque churches in Gothic style. This may be a coincidence, but if the frequent use of coins can be understood as a sign of prosperity and monetary proficiency, it is reasonable to suggest that there is a connection. Even though the quantification of coins in the church room alone cannot explain the rebuilding frenzy of the churches in Gothic style – the absolute majority of the coins consist of low-valued ­bracteates – there is an indication of a shift in coin use that engaged Gotlandic society on a broader scale and at different levels.

The evidence of foreign coinages The publications of Moesgaard and Fridh clearly show that Gotlandic coinage – mainly based on the evidence of church finds – dominates the coin circulation on

102  Christoph Kilger medieval Gotland from the thirteenth century onwards. Fridh has registered 28,822 coins from the rural churches of Gotland.84 The domestic Gotlandic coins, which were produced until the mid-sixteenth century, make up 99% of the coin circulation in the rural churches, of which the W-bracteates make up 93%. Foreign coinages on the other hand only constitute 1% of the finds. Both studies confirm that coin circulation was highly regulated on Gotland in the Middle Ages, i.e. that foreign coins were not legal tender. Bunge has a relatively large proportion of foreign coins compared to other churches on Gotland (Table 5.4). This overview is based on Fridh’s work and shows which Gotlandic churches have the highest number of foreign coins. Swedish and Danish coins are also listed here, although they would have been legal tender on Gotland in certain time periods. Churches with large numbers of foreign coins, excluded from Fridh’s work, are the known pilgrim destinations of Bro and St Olofsholm. The latest excavations at St Olofsholm, in 2014, resulted in several foreign coinages. This overview is therefore incomplete and needs to be complemented with the recently excavated materials, but despite this, certain tendencies can be seen. According to Moesgaard, the relatively high proportion of foreign coins, especially Polish, Baltic and German ones from the second half of the fifteenth century and the sixteenth century, found in churches on Gotland’s east coast, can be attributed to the growing importance of local trade and rural ports in the Late Middle Ages.85 Several sources report that skippers and merchants from Danzig visited Hamra on the southern tip of Gotland in the sixteenth century. Nearby, Fårösund also emerges as a trading place for grain and beer in the seventeenth and eighteenth centuries. The sources describe how the people of Fårö and Bunge partied in the harbour with visitors from the Baltic area, referred to as “people from Ösel” and “Estonians.”86 There is also evidence of visiting skippers from Denmark in the early eighteenth century who contributed donations to the neighbouring rural churches in Hellvi and Rute.87 In other words, Bunge was probably no ordinary rural church during the Late Middle Ages, but was instead a place which attracted international contacts and agents. Both Bro and St Olofsholm were well-known places of pilgrimage in the Middle Ages.88 Hilfeling, who visited Bunge in the beginning of the nineteenth century, once again provides valuable information, stating that in the Catholic period, the church was frequently visited by pilgrims wishing to see its many shrines.89 There are, however, no contemporary medieval written sources, such as letters of indulgence, Table 5.4 Foreign coins from churches on Gotland up to the seventeenth century, compilation based on Fridh 2014 Baltic region

Denmark

Sweden

Germany

Norway

St Olofsholm (13 sp.– 8%)

Burs (27 sp.–1.08%)

Stenkyrka (54 sp.–12%)

Bunge (19 sp.–10%)

Bunge (3 sp.–30%)

St Göran Visby (8 sp.–11%)

Bunge (17 sp.–0.68%)

Bunge (51 sp.–11%)

St Hans/St Per Visby (17 sp.–9%)

Hogrän (2 sp.–20%)

Bunge (7 sp.–10%)

Väskinde (14 sp.–0.56%)

Burs (?)

Hogrän (14 sp.–7%)

Gammelgarn (2 sp.–20%)

Total 72 sp. (100%)

250 sp. (100%)

465 sp. (100%)

191 sp. (100%)

10 sp. (100%)

Moving money and ritual money  103 to corroborate this. Based on the numismatic overview and comparisons with other churches, it can be argued that Bunge may have been a place of pilgrimage which was visited by foreign traders, although it was not of the same importance as St Olofsholm and Bro. However, the proliferation of coins in Bunge might also be taken as evidence of the importance of northern Gotland as a place where foreign merchandise and trade was available. In that case, the rich coin finds rather indicate the presence of a seasonal market in the vicinity of Bunge. From that point of view, the attraction of Bunge might have been both the patron saints of the local church and commercial activities.

The use of coins in the Gothic church This chapter will examine the use of coins in the Gothic church in their archaeological contexts and establish how distribution patterns of coinage might provide clues on devotional practices. What conclusions can be drawn regarding the importance of the altars situated in the nave for monetary offerings? And what kind of other activities took place within the church? In order to answer the last question, the general distribution pattern of the coins in Bunge will be compared to the many finds of beads and shroud pins. The contextual analysis also deals with the materiality of both church space and coins, along with how architecture and interiors, such as floors and standing structures in the church, which have disappeared, influenced the distribution patterns. From a source-critical point of view, it is also important to consider the reasons why and extent to which the coin material was disturbed and redeposited. A question which sparked great discussion in earlier research was whether the coins under the church floors represent accidental losses or deliberate deposits.90 This debate had reached an impasse until Henrik Klackenberg introduced the idea of “offering spillage,” which encompasses both the intentional and the more random dimension of the church finds. Klackenberg argued that coins in churches are primarily a result of random spillage, from monetary offerings at mass, in solitary worship, during pilgrimage or other transactions. It is on these occasions that the coins fell through the cracks in the floor.91 The large coin vault in Lafrans’ offering box can be used as an example. The box may on many occasions have contained hundreds of tiny bracteates. When the hatch was opened and the box was emptied, many of the coins may have ended up out of reach under the box or the floor (see Figure 5.18). Extant accounts from the Danish governor Ivar Axelsson Tott bear witness to the fact that large sums of coins could be offered on a single occasion on Gotland in the latter half of the fifteenth century. The sources state that the governor himself travelled to Bro church on a pilgrimage to make money offerings.92 On one occasion, in 1485, an offering of 2 marks was recorded, and this was repeated in 1487.93 This example illustrates that during certain holidays, there could be a spiritual exchange of hundreds, if not thousands of coins at known places of pilgrimage, such as Bro church. In his doctoral thesis, Klackenberg discusses this issue with reference to the church accounts of Kumla church in the Swedish province of Närke. The accounts have been preserved in the period between 1421 and 1590. Klackenberg estimates that three marks were offered in this pilgrim church on a yearly basis. Calculated in coins, this corresponds to about 576 pennies. In 1971, 270 coins were found during excavations of Kumla church in the western half of the building. Based on the comparison of preserved accounts and the coins found in churches in archaeological

104  Christoph Kilger

Figure 5.18 Coin vault in Lafrans’ offering box. Photo: C. Kilger

excavations, he argues that the preserved coin material only presents a small fraction of the sums once handled and exchanged in the church room.94 Therefore, the preserved coins in Bunge church can, in line with Klackenberg’s arguments, be seen as “offering spillage” from this, at times extensive, activity.

Coin clusters and offering boxes A distribution analysis of the coin material from the excavations was published in 1972 by Ragnar Engeström.95 In this early publication, he made some interesting observations regarding the find distribution. He also compared the time of minting with the distribution pattern and, in this way, distinguished between the coins used in the Romanesque church from the ones that circulated in the Gothic church. Figure 5.19 shows the distribution of 3,238 coins which have been numismatically classified and attributed to a grid square. The coins are predominantly W-bracteates. A total of 198 m 2 was examined, of which the tower measured 15 m 2 and the nave 183 m 2 .96 The average number of coins per square metre is 21, if the coin fragments are also included. Comparable coin distribution analyses have been undertaken for Silte (a small parish church in south-western Gotland), the Romanesque church ruin in Västergarn and St Olofsholm. The nave in Silte church is 96 m 2 , i.e. half the size of Bunge, and was surveyed in 1972.97 A total of 1,672 coins were found, which gives us an average of

Moving money and ritual money  105

Figure 5.19 Map showing the distribution of all numismatically classified coins. Main scenes paintings northern wall: (A) Crucifixion, (B) Gregory’s vision of Arma Christi, (C) Golgotha. Map: Steinar Kristensen, Museum of Cultural History, UiO.

17 coins per square metre. The remains of the Romanesque church by the medieval harbour Västergarn were fully excavated in 1973/4.98 Despite a total floor area of 185 m 2 for the nave and choir, only 147 coins were found, which gives us an average of 0.8 coins per square metre. For St Olofsholm, where the whole choir and the

106  Christoph Kilger south-eastern part of the nave were excavated, no such detailed figures are available. There are, however, detailed records of the coin concentration in the nave, which can be compared to Bunge.99 In comparative perspective, in the churches at Silte and St Olofsholm, the coin density is significantly higher in Bunge. The grid square with the highest coin density in St Olofsholm was found adjacent to a mid-altar and contained 48 W-bracteates. A grid square which contained more than 60 coins was located in the middle of the nave at Silte.100 The densest grid squares in Bunge contained 78 to 94 coins. These concentrations are divided between three zones which indicate where the offering boxes were placed, in front of the southern side altar devoted to St Olof (94 coins), in the middle of the nave along the northern wall (79 coins). In this particular area, in front of Gregory’s vision of the Passion of Christ, there is a cluster of coins which provide more accurate dating than the W-bracteates. They indicate that a collection box was placed here from the late fourteenth to the early sixteenth centuries.101 Finally, in the half-metre grid square immediately to the right of the entrance inside the southern portal and close to the wall, 78 coins and many coin fragments were found, which in relation to the small area represent the densest concentration of coins. Ragnar Engeström, one of the excavators of Bunge church, suggested that this spot marks the place of Lafrans’ offering box.102 However, if this had been the case, the offering spillage would have created a different, even larger distribution pattern, not least due to the opening radius of the substantial money hatch. There are simpler collection boxes on Gotland that are made out of hollowed tree trunks with iron lids with coin slits on the top and a small coin compartment at the front. Such boxes are still standing inside the southern portal of churches, such as Bro, Hamra and Rone.103 It is more reasonable to suggest that the limited concentration was caused when the coins were offered, dropped behind the collecting device and disappeared in the gap between the floorboard and the church wall. It is possible that such a tree trunk, often referred to as a truncus concavus in the written sources, may have been placed at the entrance in Bunge.104 The isolated cluster of coins found by the stairs, on the southern side of the ringing chamber, indicates the existence of yet another offering box, which the parishioners passed on their way out through the western portal, this time on the left-hand side. Also, this concentration is entirely made up of W-bracteates. A Polish denier minted during the reign of King Kasimir Jagiello (1447–1492) was found in the neighbouring grid square which indicates that a truncus concavus may have been in use until the late fifteenth century. A cluster of more than 60 coins from the same area in Silte church suggests that a collection box was in use for similar purposes in the Middle Ages.105 Comparing the coin clusters by the entrance and the exit at Bunge church, it seems likely that the spillage resulted from the same collection box. Such an interpretation is supported by the chronological composition of the coins present in the two concentrations. The concentration by the entrance includes several early W-bracteates with a high silver content and a Swedish penny of King Magnus Eriksson from the midfourteenth century, but no later coins. The collection box may have been placed by the entrance portal to collect money for the building of the new church. Once complete, the box may have been moved to the exit to provide for the maintenance of the church. The interpretation of coin clusters indicating the placement of offering boxes is also consistent with preserved contemporary written sources. A decree issued by Bishop

Moving money and ritual money  107 Karl of Linköping in the year 1329 explicitly states that anywhere on Gotland where processions are conducted, one-third of the money from all collection boxes offered to either a saint or relics should be used to finance the maintenance of the church.106 A similar decree regulating the income from offering trunks is mentioned in the extant episcopal statutes from St Olofsholm dated to 1360.107 Putting the evidence together, it can be established that the coin density is more pronounced by the southern side altar and especially along the northern wall of the nave. The clusters, which will be discussed more in the following sections, suggest that there were at least five possible locations for offering boxes: (1) in front of the southern altar, (2) by the southern entrance, (3) by the western exit, (4) in the middle of the northern wall and (5) adjacent to the platform in the north-eastern part of the nave. Another cluster is recorded in front of the northern side altar, which could indicate a sixth location. Interestingly, the western part of the nave, i.e. the rectangular area between the north and south portals and the western wall, contained relatively few coins. What we probably see here based on the archaeological evidence is the spatial extent of a late medieval offering zone and offering spillage accumulated over 150 to 200 years. What we can also imagine from the distribution is that this is the material outcome of hundreds or even thousands of acts of individual devotion.

The northern half of the nave – offering spots and empty spots Most remarkable for Bunge are the widely scattered coin densities along the northern wall of the nave. It is reasonable to assume that these are associated with offerings performed in front of the frescoes and that there could have existed installations such as altars and offering devices now gone (see Figure 5.19). There is one offering zone in the mid-section with a concentration of 79 coins in the densest square, which probably relates to the central scene depicting Gregory’s vision of the Passion of Christ. Another concentration with 66 coins is visible beneath the Golgotha scene and to the left of the rectangular stone base with the medieval grave slab. A more widely scattered coin assemblage, covering four grid squares, is situated three metres further to the south in the north-eastern section of the nave. Remarkably, between these two concentrations – if they are interrelated – there is a zone almost devoid of coins. Yet another dense coin area with 56 specimens appears in the north-eastern corner of the nave. However, it is not clear if this concentration relates to the northern side altar or the platform with the grave slab and the rectangular stone base. There are some source-critical factors to consider. The northern half of the nave, especially the area in front of the stone base, was used for burials between the fifteenth and seventeenth centuries (see Figure 5.23). Thus, the cultural layers could have been redeposited to a greater or lesser extent. However, the grid squares adjacent to the left corner of the stone base which are virtually devoid of coins, but are surrounded by squares with high coin densities, are remarkable. They probably cannot be explained as simply the result of dugout holes left by burials. There are several possibilities to explain their absence. One possibility is the placement of a broad-based offering box here, such as Lafrans’ offering box, or the presence of a small kneeler or hassock. Offering spillage ended up underneath the floorboards, not directly below but encircling the base of the box, and thus creating a negative footprint of the box’s or kneeler’s location. Another possibility is that this area was paved and regularly cleaned. The pavement could have consisted of grave slabs which were later removed. As a result

108  Christoph Kilger of sweeping, coin material was successively deposited in the cracks and edges of the stones, thus leaving this particular distribution pattern. However, due to the fact that the finds were recorded in square metres, the resolution is too low to provide any additional clues and more certain answers on this question. The existence of at least two or three coin-offering spots, presumably situated along the northern wall of the nave, is a phenomenon which to my knowledge has not so far been recorded and discussed in earlier numismatic or archaeological studies of other medieval rural churches in Scandinavia. The observation of an extensive activity zone in the northern half of the Gothic nave, which is based on the proper archaeological recording of coins, adds more clues to the suggestion already made in the numismatic survey. The status of Bunge was presumably a place of worship known beyond the borders of the parish. Other archaeological features such as the distribution of coin fragments, pins and beads are considered below to substantiate this observation.

The chancel arch – mid-altar, wooden post and stone cist Few coin finds have been made in the eastern part of the nave, and in addition, the almost empty grid squares are all situated in the transition between nave and choir, i.e. under the chancel arch (see Figure 5.19). It is possible that there was a physical barrier which also left a negative footprint here, perhaps a threshold or a rood screen. Compared with St Olofsholm, the find situation is very different. In the pilgrim’s chapel, the concentration of coins is rather high in the transition between nave and choir.108 The concentration in the zone of the chancel arch in St Olofsholm clearly relates to a rectangular stone base, which on the basis of its proportions (1 × 1.8 m), and its position on the church’s mid-axis, is interpreted as a mid-altar.109 A cluster of 48 W-bracteates and remains of wood were found in a small rectangular stone cist, consisting of vertical slabs, immediately behind this altar. The excavators view this cist as having marked the place of an offering box.110 At Bunge, a rectangular slab was also found in the eastern part of the nave and has been interpreted as the base of the main altar from the older Romanesque church (see Figure 5.14). Using the evidence from St Olofsholm as an analogy, the slab in Bunge – besides indicating the main altar of the Romanesque church – may also be interpreted as the foundation of a mid-altar in the later Gothic church (see Figure 5.19). Reconsidering the evidence in St Olofsholm from a liturgical point of view, it seems unlikely that the stone cist behind the altar marks the location of a collecting box. In that instance, visitors would have had to reach over the altar to make an offering. It is more plausible that a collecting box stood in front of, and not behind, the altar. The stone cist is more reminiscent of a stone-lined posthole, which was also the excavator’s interpretation. Was the wood from the cist at St Olofsholm really the remains of an offering box? It seems more likely to be the remains from a wooden post, which stood behind the altar and was wedged into the chancel arch, maybe as a support for a crucifix which was mounted on the post. The many W-bracteates found in the posthole can also be explained as having resulted from lavish offerings given to and placed on the altar itself. Some coins tipping down from the altar cloth were caught in the cist behind the altar.111 That a wooden pillar may also have existed in Bunge church in the transition between nave and choir is indicated by several sources. According to the antiquarian Pehr Arvid Säve’s travel log from 1864, a wooden post is said to have stood in the

Moving money and ritual money  109 chancel arch. This post was removed in 1807.112 A similar observation was made by Hilfeling after his visit a few years earlier. He wrote: “This church is a joyful and beautiful building. [It] has three pillars from the choir, along and after the church, and two aisles on either side.” A foundation for such a post was found during the renovations in 1916/17 when the wooden floor of the choir was removed. Theodor Erlandsson, the local teacher of Bunge who participated and reported about the renovations, mentions a square stone base placed in the mortar of the choir floor which was positioned directly below the tip of the chancel arch. In his report, Erlandsson further mentions that the slab was probably the bottom of a square stone cist referred to in local traditions, where it is called an “offering fountain.” It seems that in 1807, the stone cist was removed and placed in the south-eastern corner of the vestry, from where it later disappeared.113 The evidence for the existence of a stone cist, popularly imagined as an “offering fountain,” is circumstantial and has to be handled carefully from a source-critical point of view, but despite the uncertainties, it is nonetheless important to mention here. The comparison between the coin finds from the areas surrounding the chancel arches in Olofsholm and Bunge reignites the question regarding the relationship between the altar, the coin finds and the cult of the saints. Extant written sources dated to 1246, 1248, 1360 and 1376 specifically mention offerings at the altar of St Olof in Akergarn.114 Taking the evidence together, the coin finds and the postconstruction indicate that the rectangular stone base in Olofsholm was possibly the site of the altar dedicated to Saint Olof. Since the chapel in Akergarn was not an ordinary parish church, the offerings to the saint may have been concentrated along the mid-axis of the church and not, as customary, by the side altars. Also, no side altars have been documented in the excavated areas of St Olofsholm. If there was a mid-altar at Bunge, it was probably not intended for coin offerings. However, the Gothic chancel in Bunge was never excavated properly, so this remains inconclusive.

Coin fragments and fragmentation Coin fragments make up a quarter of all the coin finds from Bunge. In this section, I will analyse this particular phenomenon and how the archaeological record of Bunge provides an opportunity to study the contextual relationship between fragmented coins and the physical environment in Bunge. The construction of the floors, the building materials used such as stone or wood, the length and width of planks and the orientation and angle of the beams, all influence the way in which the coins are deposited.115 Paper-thin bracteates are hard to keep hold of in a dark and cold church. They are easier to drop and have also a stronger tendency to disappear through the gaps in the floorboards than larger and heavier coins. When lying on the floor, coins are also affected by the hardness or smoothness of the surface, e.g. stone or more flexible materials such as wood. The question remains as to whether the fragmentation is a consequence of fragmented bracteates being used as monetary offerings, or due to recurring physical movements subjecting them to extreme weights and forces, or even to the existence of different floorings. Or should it be regarded as the result of different activities performed by the churchgoers? The map shows the spatial relationship between the coin fragments, primarily W-bracteates, and all other coins (see Figure 5.20). In order to compare both categories, the complete coins have been mapped according to quantity as point clouds with

110  Christoph Kilger

Figure 5.20 Map showing the distribution of coin fragments. Main scenes paintings northern wall: (A) Crucifixion, (B) Gregory’s vision of Arma Christi, (C) Golgotha. Map: Steinar Kristensen, Museum of Cultural History, UiO.

Moving money and ritual money  111 a random distribution within each square, and the coin fragment according to weight. The weight is illustrated grid by grid through a shifting colour scale. The distribution map indicates that coin fragments are not present in the whole nave but are confined to concentrations in the northern half of the nave, in a rectangular area extending from the northern portal, c. 7 m to the east and c. 5 m to the south. The largest numbers of coin fragments are found in the grid squares along the northern wall of the nave, probably due to coins ending up in the gap between the wall and the wooden floor. The rectangular area covers an area of 30 m2 and corresponds roughly to the offering area described above, the northern half of the nave where many coins were recovered. In contrast to the area covered by coin fragments on the northern side, the distribution on the southern side of the nave is limited to rather small areas. A distinct area of fragments stretches diagonally from the western pillar, past the baptismal font, towards the western wall and continues to the exit and the western porch. Could this pattern be related to the movement around the font and the busy area leading out of the church? Also, a small cluster of fragments is found by the southern side altar. Here, the contrast between complete and coin fragments is particularly evident since the area close to the southern altar marks the zone with the highest density of coins in Bunge. The grid square with the fragments is where the iron stove was installed in the late nineteenth century. The base of the stove was found during the excavation. In this instance, the fragmentation of the coins could be the result of modern disturbance to the archaeological remains. Yet another small concentration is found by the entrance in the same grid square where collecting box no. 1 was situated. The contrasts between grid squares with few fragments and surrounding areas with heavily fragmented material are also noteworthy. Two virtually empty grid squares are seen in the rectangular area, otherwise rich in coins and fragments, in the northern half of the nave. The westernmost of these is located next to the central motif of the Gregorian mass, c. 1 m from the wall, while the eastern grid square was next to the stone base with the medieval grave slab found during the excavations. Interestingly, both grid squares are located to the left of their respective scenes at the same distance from the wall. Fragments generate another pattern which strengthens the already proposed placement for collecting boxes nos. 4 and 5. Thus, a large collecting box with a broad foot generating this negative footprint was probably placed on one or both of the spots. There is the possibility that Lafrans’ offering box could have been positioned here in the Later Middle Ages, from the late fourteenth century onwards.116 Through the clear spatial contrast between intact and fragmentary coins, it is possible to interpret the rectangular area as evidence of extensive activities that took place in the northern aisle of the nave, and which was focused around the frescoes and possibly constructions other than the collecting boxes. On the basis of her study of the frescoes, art historian Mereth Lindgren has already argued that there may have been an altar dedicated to the Body of Christ underneath the Arma Christi scene.117 According to the excavation report, no base or other construction that may have supported an altar was found. It is therefore possible that a portable altar on a wooden table may have stood here. There is also a “corridor” in which no fragments have been found, stretching from the northern portal to the baptismal font. It is located near the area rich in fragments, which suggests that there was a barrier, possibly a screen, preventing coins from moving further west. The distribution pattern reinforces the earlier observation that the western part of the nave, dominated by the crucifixion scene above the portal, was not part of the zone where coins were actively used.

112  Christoph Kilger Another logical option might be that the coin fragmentation may also have been caused by the burials in the nave. Many of the grid squares where human skulls were found correspond with areas with many coin fragments, which could indicate where graves and grave slabs were originally located. The areas without fragments could indicate “soft surfaces,” i.e. areas covered by wooden floors. If this suggestion is accepted, the question is: To what extent did the burials interfere with cultural layers, and thus also the coin distribution? Is there a way of pinpointing this activity in time and space?

Pins and skulls – the burials in the nave During the restorations of 1916/17, a large number of skeletons were found. The soil was loose and filled with human bones. During the excavations of 1971/72, human bones were also found, embedded in the layer of spruce needles underneath the wooden floor. Falck’s report stated that only a few graves were undisturbed, and unfortunately, the skeletal remains had not been documented with the same care as the other finds. It is not known whether the bone material was collected or redeposited. One of Engeström’s plans shows the distribution of skulls, which numbered 34 in total (see Figure 5.21). These skulls were primarily found in the northern half of the nave and close to the baptismal font. It is noteworthy that a concentration of 19 skulls was found in six connected grid squares running in an east–west direction in the north-eastern part of the nave.118 Perhaps this distribution pattern in some way reflects the original location of the skeletons? This idea is supported if we consider the placement of the modern pews along the northern side of the church. The specific location of the skulls in rows may be explained by the moving of the pews and may indicate modern reburials. In 1856, the church and its interior were refashioned. The pews, which previously were placed in the middle of the nave, were removed and the wood was reused as building material for the current pews, which were then placed along the northern and southern walls of the nave. The skulls are physically located underneath the outer wall facing the aisle. It therefore seems likely that the redeposition of the skulls was carried out before the installation of the pews. The distribution of the skulls also corresponds to the distribution of the brass pins in the nave (see Figure 5.23). A total of 697 pins have been classified and interpreted as burial or shroud pins.119 The pins were not taken care of to the same extent as the bodies when they were moved. Thus, the distribution of the pins can be interpreted as a “shadow” of the original distribution of the graves. As a specific theme, brass pins from Scandinavian churches are a relatively unexplored topic, and it is therefore difficult to assess these finds within a larger church archaeological context. If the pins mark an area of burials, this also has consequences for the interpretation of the coins. I will return to this issue below. The chronology and typology of the pins can provide approximate points of reference in terms of where, when and for how long burials took place in the nave. Pins made from drawn wire with rectangular heads (henceforth type A) probably belong to the earliest known burials.120 Type A is attested in fifteenth-century contexts from England and probably was used there until c. 1550. Type A pins make up 30% of the pin finds and are primarily located in the north-eastern part of the nave. The next generation of pins, type B, is also made from drawn wire but has a distinctively more

Moving money and ritual money  113

Figure 5.21 Map showing the distribution of human skulls in relation to the placement of the modern pews. Map: Steinar Kristensen, Museum of Cultural History, UiO.

114  Christoph Kilger

Figure 5.22 Brass pins from Bunge church, Type A-C. Source: Svensson 2015, Knappnålar som gravmarkörer. Photo: J. Svensson

rounded head. There are 512 pins of type B, manufactured between c. 1550 and 1700, making up 73% of the finds. This could suggest an increase in the number of burials, but also a change in burial practices in which more pins were used for attaching the shroud to the dead body. The distribution of these pins shows that an area further west in the nave came into use. Only 41 pins of the subsequent type C, with a soldered spherical head, have been found, making up 6%, which indicates that the burial activity ceased at the beginning of the eighteenth century (see Figure 5.22). The chronology of the pins suggests that the burials started in the later Middle Ages, probably in the fifteenth century, and lasted until the late seventeenth century. There were several grave slabs from the sixteenth and seventeenth centuries (see Figure 5.23). One of the latest slabs is dedicated to Margareta Jakobsdotter who, according to the inscription, died in 1690. Extant documentation shows that many grave slabs were probably removed from the church in the 1850s and some were taken back again during the renovations in 1916/17.121 Large areas of the choir were then used for redepositing the slabs. Probably none is in its original position.122 A reasonable guess is that some may originally have been located on the northern side of the nave, in the same area where a large number of pins were found. The distribution of the pins shows a similar tendency to that of the coin fragments and reinforces the impression that late medieval and early modern burials were consigned to the northern half of the nave. The study of the pins also provides clear evidence that the cultural layers of the northern half of the nave have been disturbed by the burials. On the other hand, the distribution pattern does suggest that the redepositing of the cultural layers did not move sideways or across great distance. Another important observation is that the concentration of skulls coincides with the area with the highest density of pins, as, e.g., grid square 59, which contained five skulls as well as the largest cluster of pins. It is therefore reasonable to assume that skulls were reburied close to the original location of the bodies. The pins may have also had other functions, e.g. for hair arrangements or clothing on living persons. However,

Moving money and ritual money  115

Figure 5.23 Map showing the distribution of brass pins. Map: Steinar Kristensen, Museum of Cultural History, UiO.

116  Christoph Kilger the quantity of pins discovered, almost all in the northern half of the nave, and the spatial association with human remains, rather indicates that the pins reflect burial practice and not necessarily the gender-based division of church space as evidenced by many different sources.123 But who were the people that were put to rest in the nave at Bunge? Were the individuals buried there related to the church building and the parish congregation in some way? Late sources tell us that the church pews were reserved for the farms of Bunge parish. Each farm had had its own row. The fixed order dictated that churchwardens should occupy the first row, the inhabitants of the church land, the so-called annexe, in the second, the Ducker farm on the third row, etc.124 This social and possibly economic ordering of the parish in early modern sources could probably also apply to the medieval situation. The concentration of burials in a limited and confined zone of the nave is reminiscent of a necropolis where people from the parish were buried together. A burial context which has clear resemblances with the necropolis in Bunge has been recorded during the excavation of Silte church on south-western Gotland. Fourteen graves and remains of further 25 individuals had been buried in the western part of the nave of the stone church.125 A source which further illuminates the situation in both Bunge and Silte is the famous runic inscription on the northern wall of Anga church in eastern Gotland. The inscription dates to the late thirteenth century and tells the deeds and the economic contributions of the farmers who built the church.126 Taking the evidence together, it is reasonable to assume that the individuals buried in the northern half of the nave in Bunge church belonged to the farms and families that had contributed to the building and upkeep of the nave. The obviously intentional placement of the burial site in the northern hemisphere of the nave may have been due to the location of altars along the wall. The concentration of early shroud needles of type A indicates the location of the earliest burials in front of the platform and the Golgotha crucifixion scene. The third and final find category to be considered here is the many beads recovered, which provides further information on the spiritual environment in the Gothic church.

The beads In addition to the pins and coins, a large number of beads, 219 in total, were recorded. Most of the beads are medieval single-coloured glass beads, while a small number are made of carnelian, amber, bone and jet (see Figure 5.24).127 Several concentrations of beads in the nave and the tower chamber have been recorded (see Figure 5.25). The beads from the find concentrations can be categorised according to size, colour and shape, which suggest that some of them were part of one or several chains. The find context, shape and colour suggest that at least 130 beads may have been part of strings of beads and deposited at selected places in the church.128 Originally threaded on a cord, beads made of jet, amber or rock crystal may have been part of paternosters or rosaries. Used in private meditations, paternosters were associated with the religious practices of the laity.129 However, beads may also have been sewn on textiles that were used for decorating the church room, attached to the clothes on wooden sculptures, or used as embroideries on altar cloths. Small glass beads are also documented on liturgical equipment such as reliquaries, portable altars and chasubles.130

Moving money and ritual money  117

Figure 5.24 Beads from square 1, north-western corner of the nave, next to the northern portal of the Gothic church, Gotland museum, Bunge inv. 2–14. Photo: C. Kilger

Many of the beads found in the northern aisle of the nave seem to relate to the frescoes on the northern wall. A cluster of 71 beads consisting of several types was found in square 1 in the north-western corner, where hardly any coins have been found. This concentration is situated next to the northern portal of the church. The question is whether the cluster also relates to the depictions of saints on the walls – one female saint on the northern wall and a male saint on the western wall. Given that this large concentration can be interpreted as one or several intentional depositions, it might indicate devotional practices performed in a secluded and intimate corner of the church. A particular use of rosary beads which could explain the concentration is mentioned in late medieval English sources dating to the fifteenth century. According to those sources, rosary beads were provided for public use as communal prayer beads in churches. Rosaries would have been used in prayers to the Virgin Mary. When leaving the building, the beads were to be left where they were found.131 Another possibility is that this concentration could indicate the placement of a statue in this

118  Christoph Kilger

Figure 5.25 Map showing the distribution of beads in Bunge church. Map: Steinar Kristensen, Museum of Cultural History, UiO.

Moving money and ritual money  119 corner of the nave. In medieval parish churches in England, statues of female saints were adorned with rosaries given by local women, especially women concerned with successful childbirth, who used rosaries during pregnancy by presenting them to statues of patron saints, such as St. Margaret, in their parish church.132 A less obvious concentration of 20 beads, dispersed over three square metres in the eastern section of the northern wall, may indicate another activity zone. The beads were found in front of the fresco of another female saint and the scene from Golgotha. A more in-depth analysis of the beads from both contexts in terms of their composition and patterns of wear might provide additional information on their significance and use in the church room at Bunge and determine whether they relate to furniture and church fixtures such as saint sculptures and side altars placed along the northern wall. Yet another cluster, consisting of 44 beads, was found in grid squares 132 and 133 in the southern half of the nave. There is no information regarding stratigraphic context. Most beads were made of translucent yellow or opaque black glass. Some show clear signs of wear and may have been mounted on a string and possibly formed part of one or more rosaries. This means that they would have formed part of the same stratigraphic layer. Since the grid squares are situated inside the foundation of the Romanesque church, immediately adjacent to the presumed position of the southern portal, these strings of beads may have been deposited during the time of the Romanesque church. Finally, 20 beads (18 small white glass beads and two large jet beads) on a copper wire were found in the ringing chamber. The context suggests that the beads were deposited at the foot of grave 1 (see Figure 5.25). Like the brass pins, the beads provide valuable information regarding taphonomic and contextual conditions. The distribution pattern of the beads suggests that the cultural layers in certain areas of the nave have not really been redeposited or moved any great distance. As symbolic tokens of faith and devotion, beads were also handled in the church space. In the same sense as coins, beads can, in a church context, suggest underlying intentions, hopes and expectations. The archaeological context indicates that beads were probably regarded as personal objects which in some cases have been intentionally deposited. Comparisons between the distribution patterns of beads and coins can therefore provide clues on different devotional practices performed in a spiritual environment. Finally, the combined distributional analysis of the coin fragments, burial pins and beads corroborates observations made earlier that there was significant activity going on in the northern aisle of the nave.

Glimpses of a late medieval devotional milieu There is now room to argue that there exists a mutual relationship and spatial dynamics between the hidden archaeological materials deposited beneath the wooden floor and the visual material culture and architecture of the church room above the wooden floor. The scenography which appears so clearly on the northern wall of the nave has its counterpart in the cultural layers beneath the floor. The intention of the frescoes was not only to convey the well-known stories of the saints, but also to spatially divide and organise the room. By depicting people and buildings in perspective, a sense of scenography could be created. Knowledge of this technique existed outside Gotland and the frescoes may initially have been intended for a larger audience, not only for the parishioners of Bunge and neighbouring areas. There may be a connection between

120  Christoph Kilger the spectacular frescoes and the influence of the Teutonic Order during its military dominance of Gotland between 1397 and 1408.133 One of the female saints on the northern wall, found between the Gregorian mass and the scene from Golgotha, has been interpreted by the historian Birgitta Eimer as the Order’s patron saint, Dorothy of Montau.134 Anders Andrén in turn suggests that the paintings with knightly motifs, the defensive character of the church, may imply that Bunge was of military importance as a stronghold for the knights of the order.135 This idea seems plausible from a numismatic perspective as well. There are seven coins from the Livonian Order, minted in Reval between the 1350s and the 1390s. This small group forms almost half of all of the fourteenth-century Baltic coins found on Gotland.136 There is actually also one bracteate from the Teutonic Order, minted between c. 1390 and 1400.137 The archaeological contexts suggest that the entire northern wall of the nave formed one coherent stage in front of which multi-faceted devotional practices may have been played out. The contrasts between the zones with few and many coins may reflect the placement of late medieval furnishings, such as screens and offering boxes. Remains, such as the platform in the north-eastern corner of the nave, spatially correspond to the iconic portrayals of the Passion mystique: the Gregorian mass, also known as Imago pietatis, and the removal of Christ from the cross at Golgotha and the mourning women. Archaeologist Mattias Karlsson suggests that such platforms or niches could have been used during ceremonies as an elevated final stage for performing the burial of Christ. Besides the niches representing the “tomb of Christ,” there existed other items which were used during these ceremonies, such as coffins and dolls representing the Body of Christ. The frescoes, the platform with the grave slab, the many beads and of course the coins may provide glimpses of a late medieval devotional milieu dedicated to cult of the Body of Christ, a popular mystery play which developed its own symbolism and liturgy centred on the Passion of Christ.138 The cult surrounding the Body of Christ was celebrated during both mass and outdoor processions where the hostia, representing the Body of Christ, was carried. In Bunge church, the scenographic portrayal of the various stages of the Passion of Christ may also be seen as stations that the individual pious church visitor passed in both body and spirit, either in private devotion or as part of a procession. The peculiar distribution pattern of the coin fragments shows the spatial extent of actions performed in front of the wall paintings. The fragmentation of the coins may hint at the movement of many people in this area reflecting the favoured routes of pilgrims. The question is whether the fragmentation is conditioned by the environment itself and the liturgy. For instance, acts of devotion such as kneeling may have contributed to the unintentional crushing of discarded coin material still lying on the floor. Other practices affecting the coin material were the ceremonies conducted when burying the dead in the cemetery, which meant that the church floor was opened on several occasions. If there existed burial slabs in this area, fragmentation could be caused by moving the stones back and forth. Another conceivable interpretation could be that the coins were damaged or broken on purpose. The custom of bending coins has been observed in church contexts from medieval England, but less so in Scandinavia.139 A possible interpretation for the offering of so many coins in the northern nave is because it was in this zone that the Body of Christ rituals took place. The Catholic pilgrimage site that Hilfeling received information about in the early 1800s may refer to the rites of passion performed in Bunge church during the

Moving money and ritual money  121 fifteenth and early sixteenth centuries. The suggestion put forward here is that Bunge was a pilgrim site of regional importance devoted to the cult of Corpus Christi, which also attracted an international audience. This could explain the extreme proliferation of coin finds which, even by Gotlandic standards, is exceptional. When and why this cult started are further questions to be investigated. A detailed study of the distribution of late medieval coin types, which is not part of this contribution, might possibly illuminate these questions.

The introduction of coins in the Romanesque church A coin-based economy with its own monetary standard was introduced in Gotland in the twelfth century, when local minting modelled on continental standards also began.140 In this context, it is interesting to note that the Gotlandic minting and use of coins coincides with the introduction of stone architecture and stone churches.141 The following questions will therefore be addressed here. At what point in time can we trace the introduction of coins in the archaeological context in Bunge church? What conclusions can be drawn about the use and importance of coins in the early Gotlandic stone churches based on the spatial relation between the architecture and coin distribution? In this context, we will also examine in greater detail the original location of Lafrans’ offering box in the older stone church which according to the written sources was consecrated in 1169. The entire Romanesque church at Bunge has been investigated archaeologically, which makes it possible to examine the coin distribution across the whole building. In the following section, the three phases of the early coinage of Gotland, which form the basis for the distribution analysis, will be sketched out. One methodological problem is the synchronisation of the time of minting and usage with the phases of rebuilding in the church. Changes to the church building do not necessarily coincide with changes in coin circulation. Since the coins theoretically may have been in circulation after the estimated time of minting, and thus may have been deposited at a later stage than the actual building phase, the coin distribution of the individual groups of coins will be shown cumulatively. The area with the brass pins in the Gothic building will be highlighted to show where walls and cultural layers may have been disturbed through burials. The placements of the offering boxes in the Gothic church will also be addressed in order to determine if the thirteenth-century coin material may have been in circulation and offered at a later date. The ground plan of the Romanesque church, which has been recreated through ArcGIS, shows the wall, the entrances and the chancel arch (the passage between nave and choir). The earliest Gotlandic pennies c. 1150–1220 The building phase of the oldest stone church, presumably between c. 1170 and 1220, corresponds to the time of the oldest Gotlandic pennies, minted c. 1150–1220 (see Figure 5.26). These coins represent the first controlled monetary system on Gotland, making foreign silver coins obsolete.142 The dating of the earliest building is tentative, however, and the lack of “churchyard finds,” i.e. graves of Viking-Age character and Viking-Age beads, suggests that the building process did not start until the later twelfth century. There are no records of an older baptismal font at Bunge, and the extant one was acquired after c. 1250. An older font, if there ever was one, should

122  Christoph Kilger

Figure 5.26 Early Gotlandic penny (c. 1150–1220), Myrberg I, Bunge church. Photo: G. Hildebrand, Royal Coin Cabinet, Stockholm

have been located by the entrance, to the west of the pillar. This potential spot has been tentatively marked on the plan. The plan also shows the location of the tower, although this was added after the construction of the nave (see Figure 5.27). Five early Gotlandic coins have been found, four of these inside the Romanesque building: one inside the original entrance of the western portal in the older Romanesque church, one in the eastern part of the nave, two by the northern chancel arch and the fifth one in the southern part of the ringing chamber. For the latter, there is stratigraphic information. This coin was found in the fill underneath the floor, but above the later lime layer, which suggests it was deposited after the period of minting had ceased. It may have ended up here after c. 1220, i.e. after the construction of the tower had begun. Overall, the use of coins was very limited in the oldest stone church. The hatched cross pennies c. 1220–1290 The tower is the oldest extant part of the church. It was kept when the older Romanesque church was demolished and integrated into the new Gothic church. The historic building analysis shows that the tower was adjoined to the western gable of the nave of the older church.143 Around the same time, a new phase of coinages was introduced on Gotland. The minting of hatched cross coins starts c. 1220 and continued until approximately 1290, and these coins turn up in greater numbers in Gotlandic churches. Here, the iconography is characterised by a more pronounced Christian symbolism with a cross as the main element. Some emissions are only struck on one side (see Figure 5.28).144 In total, 30 Gotlandic hatched cross pennies have been found

Moving money and ritual money  123

Figure 5.27 Map showing distribution of early Gotlandic pennies in the oldest stone church. Map: Steinar Kristensen, Museum of Cultural History, UiO.

124  Christoph Kilger

Figure 5.28 Hatched cross penny, Gotland (c. 1220–1290), Myrberg II, Bunge church. Photo: G. Hildebrand, Royal Coin Cabinet, Stockholm.

in Bunge church, which represents a pronounced increase in the use of coins compared to the previous period. After c. 1250, the western gable of the church was demolished, the western portal moved to the tower and a vault was incorporated into the ringing chamber. It is assumed that the reason for this was that the old church had been damaged.145 Another more likely explanation is that the wall between the tower and the old west façade was demolished in order to integrate the ringing chamber with the Romanesque nave, resulting in a larger church. During the same period, the baptismal font and Lafrans’ offering box were also acquired, presumably in connection with the extension of the church space. In this phase, graves also appeared in the ringing chamber. The distribution map shows the situation after the ringing chamber had been integrated into the nave (see Figure 5.29). A total of 25, or 83%, of the cross coins were found within the Romanesque church and the tower, indicating that the majority of these coins were roughly contemporary with the Romanesque church. Five coins come from outside the foundations, of which two were adjacent to the old chancel portal. Alternatively, these two coins may be later “offering spillage” from the St Olof’s altar in the Gothic church. If we examine the distribution pattern in relation to the different areas of the church, it becomes clear that 5 coins were found in the choir (1 of which was in front of the high altar), 13 in the nave and 6 in the ringing chamber. Two concentrations of coin finds appear, one containing seven coins in the zone to the south of the pillar and baptismal font, and one consisting of five coins, in the ringing chamber. As previously argued, the cluster in the ringing chamber could be part of grave 1. The documents from the excavation contain no information regarding the location of the pennies in relation to the skeleton. Finds from the adjacent grid square may strengthen the argument that this was an intentional coin deposit; here, 21 beads, small white ones and three large jet beads on a copper thread were recovered.146

Moving money and ritual money  125

Figure 5.29 Map showing distribution of hatched cross pennies in the rebuild Romanesque church after c. 1220/1250. Map: Steinar Kristensen, Museum of Cultural History, UiO.

126  Christoph Kilger The seven coins from the entrance area were distributed across four grid squares, but may all belong to the same context.147 A cluster of 44 loose beads, probably belonging to a paternoster, was found by the southern portal, inside the southern entrance to the old Romanesque church.148 In the same area as the coin and bead cluster, i.e. between the font, the pillar and the entrance area, three skulls were documented (see Figure 5.21).149 This means that the coin cluster south of the pillar displays several similarities to the finds from the ringing chamber, which could suggest that the coins and beads may also come from graves. Yet another cross penny was found inside the original northern portal. In the same area, an undisturbed burial was documented,150 which could indicate that a large portion of the cross pennies in Bunge church could possibly be linked to burial contexts. The early Wisby bracteates c. 1280–1361 The earliest emission of the Wisby-, or W-bracteates, can be connected to the late Romanesque church, i.e. before the new Gothic church was built. This issue has a high silver content and is thus easy to distinguish from the later bracteates with lower silver content (see Figure 5.30).151 There is no absolute chronology for this group of coins and the dating that does exist is based on historic events. The minting began presumably around the time of the Gotlandic civil war in 1288, which is the starting point for the political rise and independence of Visby. The emission with the high silver content is considered to end in 1361, the year of the Danish invasion. A total of 30 such coins have been recorded in Bunge.

Figure 5.30 Early W-bracteate with high silver content. Photo: G. Hildebrand, Royal Coin Cabinet, Stockholm

Moving money and ritual money  127

Figure 5.31 Map showing distribution of early W-bracteates with high silver content in the Romanesque church after 1280/90. Original position of Lafrans’ offering box is indicated. Map: Steinar Kristensen, Museum of Cultural History, UiO.

128  Christoph Kilger The distribution pattern shows all the coin groups that may have circulated in the Romanesque church in the thirteenth century (see Figure 5.31). Like the cross pennies, the early W-bracteates show a clear spatial relationship to the building. The cluster in front of the chancel arch is striking, with 11 W-bracteates in grid square 105, 2 and 3 in grid squares 106 and 102, respectively. One or two cross pennies and an older Gotlandic penny of Myrberg type I may also belong to the cluster, which probably indicates that a collection box was located here. It is likely that this concentration marks the original position of Lafrans’ offering box and the rood cross. It is notable that the cluster is situated to the south of the church’s mid-axis, since the offering spillage should have ended up north of the mid-axis had the box been placed directly in front of the chancel arch. The box’s coin hatch opens to the north, provided that the depiction of Christ on the front of the capital is facing west. Thus, the coin distribution also provides indirect evidence for the original placement of the rood cross. If the rood cross was mounted on the box lid, it was not positioned on the mid-axis of the nave and the same line of sight as the main altar, but instead put up slightly to the south. The crude back of the box provides another argument for an off-centre position on the right-hand side of the chancel arch, and the back would not be visible from the choir. In spite of this skewed placement, a clear intention to create symmetry exists with regard to the mid-axis of the church. Three corners of the box are decorated with palmettes and the fourth with an animal head. The western pillar base in the current church shows a similar composition with three corners having ornate leaves and one with an animal mask. The decorations on the pillar base and the box base obviously correspond to each other, which further strengthens the argument of the original placing of the box and rood cross by the chancel arch. The presumed opening of the ringing chamber towards the nave after the mid-thirteenth century may also imply that construction works of the nave and the western pillar of the older church had taken place. In the light of this, the rebuilding process, from the Romanesque to the Gothic church, seems to have been more complex and prolonged. The existence of an offering box in Bunge at this time also signals a change in the use of coins within the church. A cluster of early W-bracteates with high silver content is also noted by the offering boxes of the Gothic church.152 These concentrations may mark the transition phase between the older church and the newer church and could indicate the start of coin offerings in the new church. Since the dating span of the early W-bracteates is so vague, as is our knowledge of their period of circulation, it is not possible to accurately date the building process and activities within the church during the transitional phase.

Coins forging the congregation The Viking-Age coins that made Gotland rich in silver between the ninth and the twelfth centuries have left few traces in the island’s churches. Viking-Age coins from Gotlandic church contexts are almost exclusively found as coin pendants in graves, in accordance with the Viking-Age tradition of burying the dead in their clothes. Also, very few specimens of the earliest Gotlandic coinage have been found in churches. In a Gotlandic context, these coins are almost entirely known from hoards.153 The composition of the hoards suggests that a regulated monetary system was introduced on Gotland around the middle of the twelfth century, when foreign silver such as German and English coins quickly disappeared from circulation. These processes, such as the abolishment of the Viking-Age silver economy and the establishment of a new monetary economy based on coins, have left virtually few traces in the churches.

Moving money and ritual money  129 Thus, it seems reasonable to suggest that the beginnings of the monetary economy on Gotland and a system of payment based on coins are not directly linked to the earliest wave of church buildings. Thus, the offering of coins as a religious and Christian act seems subordinate when the first stone churches in Romanesque style were built. With the introduction of the next generation of Gotlandic coinage, the native cross pennies in the 1220s, the presence of coins rises noticeably in Bunge and in other churches on Gotland. Cross pennies are found in the choir, in the nave and in the tower chamber. These coins are also found in larger numbers in various contexts of the urban settlement of Västergarn on the west coast of Gotland, such as in the Romanesque church, graves and several house foundations.154 The evidence of coins from Västergarn appearing in both secular and ecclesiastical contexts seems to suggest that the use of coins in wider segments of the Gotlandic population was rising during the thirteenth century. The hatched cross itself represents a peculiar symbolic innovation. Based on a comparison with heraldic symbolism, Nanouschka Myrberg Burström has suggested that the cross pennies may be associated with the crusader ideology, which was starting to gain a foothold in the Baltic region in the early thirteenth century.155 Myrberg Burström also puts forward another alternative interpretation. The inscriptions and the choices of motifs above all allude to Christian expressions and concepts, which seemingly express a growing need to connect with the networks of the western church, which were also gaining a foothold on the island.156 During the first half of the thirteenth century, the bishops of Linköping manifested their presence on Gotland through statutes and decrees.157 The evidence from Bunge also makes it possible to interpret the cross coins as personal objects possibly associated with burials rather than as coin offerings as part of the liturgy. In order to support this further, more research on the use of coins in Gotlandic churches is needed, and the results compared to those of Bunge. A clear shift in the circulation pattern appears with the earliest high silver content W-bracteates. The cluster found near the chancel arch of the Romanesque church most likely shows where Lafrans’ offering box was located. This placement adheres to a symmetrical line of thought accentuating the mid-axis of the church. The placement of the box in front of the chancel arch, together with the cluster of coins, suggests a more controlled and directed use of coins within the church, compared to the earlier distribution pattern of the cross pennies. It is possible that this represents the first traces of coin offerings in a liturgical context on Gotland. Equally remarkable is that it is with Lafrans’ offering box that the first W-bracteates make their first appearance in a Gotlandic church. From a liturgical viewpoint, it is interesting that the first phase of coin offerings at Bunge was concentrated in front of the chancel arch along the mid-axis of the church, and not in the outer aisles of the nave. The archaeological investigation did not detect any side altars from the old Romanesque church, but that does not exclude the possibility that these existed. The role of the side altars for coin offerings is not clearly apparent in Bunge until the Gothic church was built in the fourteenth century. The Lafrans’ offering box in Bunge provides both a liturgical and a monetary dimension to the church, where the placing of coins in the box, and under the cross, became an act of offering motivated and sanctioned by the church. It thus seems likely that the rood cross with its prominent place on the offering box close to the chancel arch represents an initial phase of coin offerings in the liturgical context of the Gotlandic churches.158 One example of how the act of offering is explicitly referred to in the ecclesiastical regulations is the older Law of Västgötergötland: in manuscript B59, it is stated that the Good Friday offering is to be placed at the foot of the cross.159

130  Christoph Kilger Another development in this context should be mentioned here. At the end of the thirteenth century when Lafrans’ offering box was placed in Bunge, changes were made to the ecclesiastical visitations and the procuration rights. An agreed taxation determined how much the bishop should receive from the parish when he and his entourage visited the church. Since this decree caused regular conflicts between bishops and parishioners, the Pope tried to influence this via canon law. The right of procuration was an important and essential source of income for the bishops. In the late thirteenth century, Pope Boniface VIII lifted the ban on cash payments as means of procuration to the bishops as evidenced by the Liber sextus. In this way, parishioners and priests were able to fulfil their obligations to the bishop through hard cash.160 If the papal decrees can be applied to a Swedish and Gotlandic context making coins, an accepted and preferred form of payment for the spiritual care of the church has to be analysed in more detail in further studies. In any case, the proliferation of coin finds in churches in medieval Sweden in the late thirteenth century could indicate a shift in attitude towards coined money as a means of payment in spiritual environments on the lowest level of the church hierarchy, the parish congregation.161 The observation of a more frequent use of coins is also discernible in other written sources. A survey of monetary terms in the Diplomatarium Suecanum such as moneta, denarius, nummus and pecunia undertaken by Henrik Klackenberg shows a consistent pattern in terms of chronology. Whereas there are only a few mentioned in the diplomas before c. 1250 – the earliest one is from 1244 – there is a considerable rise of a coin-related terminology in the last three decades of the thirteenth century. According to Klackenberg, more than 80% of all known documents belong to this period.162 Based on his analysis of church finds from parish churches, Klackenberg concludes that around 1300, the kingdom of Sweden was monetised to such an extent that coins were regarded as legal tender by the peasants.163 Based on the evidence of Bunge, we can also suggest that the island of Gotland was part of the same process. In the year 1285, the island was politically tied more closely to the Swedish kingdom by accepting the treaty ledungslamen with the King Magnus Ladulås (1275–1290).164 The last decades of the thirteenth century on Gotland were also a period of urban consolidation and civil war. After having defeated the rural Gotlanders, the town of Visby won its independence in 1288. The distinguished appearance of Visby bracteates in the chancel arch of the Romanesque church in Bunge provides an indication that the town’s own coinage dominated coin circulation in the rural districts of the island. However, it is not clear whether this monetary development can be placed before or after the civil war; either the town’s monetary policy was one of the reasons for the outbreak of the hostilities, or it was the result of its victory. The grave slabs in front of the high altar in Bunge allow various interpretations regarding the church economy and the relationship between church and parish. The fact that the layman Johannes, father of the priest Botulf Ducker, is given a final resting place in front of the high altar suggests that he was an influential person, possibly a representative of the family that owned the land and which may have financed the building of the Romanesque church. The slabs in front of the church’s most sacred spot may represent as an area reserved for the members of the Ducker farm, who once owned the land. It is possible that when the new Gothic choir was completed, the remains of Johannes of Ducker were brought from the still extant Romanesque church to their final resting place in front of the high altar. The placement of the slabs also marks a symbolic shift in the parish history from the perspective of ecclesiastical law and political power: from the family members of the large farm who owned the land to the clerical family who managed the church and led the parish.

Moving money and ritual money  131 The ritual coin economy which developed from the end of the thirteenth century within a liturgical framework seems connected to the phase of expansion, when stone churches were first rebuilt in Gothic style. The rebuilding of the church in Gothic style signifies a change in the use and meaning of coins in Gotlandic society. Whereas coins in the Romanesque church were handled as personal tokens, coin use in the Gothic building indicates a different attitude to coins and a change in monetary behaviour. Combining the archaeological and numismatic evidence from Bunge, it can be argued that it is not until this time that the parish emerges as a communal organisation, which was given a much larger responsibility for the church building.165

Final conclusions This contribution is a study of medieval coins discovered in the nave and the tower of the present church of Bunge in north-eastern Gotland. With more than 3,200 coins, recovered in excavations conducted in 1971 and 1972, Bunge represents one of the most prolific and productive Christian ritual sites on Gotland and in Scandinavia. The church was rebuilt in the early fourteenth century with a building in Gothic style replacing an earlier Romanesque church. Starting from a local context, Bunge church provides an opportunity for studying the establishment and significance of the coinbased economy on Gotland. During the Middle Ages, the church building was the hub of every parish and the centre for each individual. Its monumentality conveyed the Christian hierarchical world view;166 it shaped the knowledge of the surrounding world, in terms of both heaven and the earth. The church also represented a threshold to the life cycle of individuals, and sanctioned through the sacraments the various stages of human life, from cradle to grave.167 The material culture of the church, both above and underneath the floors, reflects a dynamic tension between law and practice, between the economic policies of the church and the local parish community, i.e. the people who built the churches and were responsible for their upkeep, and not least those who attended the services. Coin finds can therefore provide knowledge of how people of the Middle Ages practised their religion within the church, individual attitudes to and interpretations of church doctrine and the message of salvation. The relationship between the physical and the liturgical space, material culture and the many different material footprints left by medieval parishioners provides an extraordinary possibility to study various behavioural patterns, intentions and conventions. The aim of this study was to analyse and compare the movement of coins in the Romanesque church and the Gothic church of Bunge and the issue of how coins, as a means of payment, relate and interact with the liturgical and ritual space. Another strand discussed in this contribution is the issue of the monetary and liturgical topography in a religious environment and its development in the long run. Crucial for understanding the movement of coins and the interconnectedness between liturgical and monetary space in churches are mechanisms to handle coins on a large scale such as collecting boxes which would have been placed strategically within the church interior. Comparing the distributional pattern of coins in the Romanesque church and Gothic church, there are several implications on the liturgy involving the offering of coins. The study shows that coins were used on a limited scale in the Romanesque church, with only 2% of the excavated coins. At the end of the thirteenth century, the placement of the offering box of Lafrans in front of the chancel arch signals a deliberate change in coin use. The archaeological evidence shows that the circulation of coins

132  Christoph Kilger in the church space from that point onwards was more controlled and confined to liturgical fix points indicating the growing significance of coin offerings. In an initial stage, coins seem to have been offered in front of the chancel arch where the rood cross was positioned, highlighting the mid-axis in the Romanesque church. Although the archaeological evidence is ambiguous, there were no clear signs of side altars in the earlier church. Also, the distribution of coins seems to indicate an absence of offering activities in the corners of the nave in the earliest phase. The pattern markedly changes in the Gothic church with offering activities moving sideways towards the side altars and leaving the area in front of the chancel arch almost devoid of coins. This change seems to be in accordance with the development of the veneration of the local patron saints in the parishes on Gotland and concomitantly the increasing significance of the church doctrine of indulgence. As the church historian Lars-Erik Pernler points out, every parish church could, in principle, become a local site for pilgrimage.168 This development took off during the second half of the thirteenth century. The prerequisite was that a parish church had to present an attractive saint, one who people were willing to go on a pilgrimage for and make offerings to. Such undertakings were encouraged by the highest church authority and were promoted by the aid of letters of indulgence. One such example is seen in the time of Pope Boniface VIII, who in 1296 issued a letter of indulgence to Näs church and Grötlingbo church on southern Gotland. The letter stated that the people who, ready to repent their sins, visited the church of St John the Baptist on the Feast of St John, on the feasts of St Nicholas, St Blasius, the Virgin Mary or on Ascension Day, would be given penance of 1 year and 40 days.169 Pilgrimages within the local area evolved into important celebrations and significant sources of income for the parish churches, who were in need of money, not least because of the colossal tasks of transforming their churches to Gothic temples. Extensive coin offerings in front of side altars, which were often consecrated to patron saints, should be seen in this context. In the late fourteenth to fifteenth centuries, the use of coins increased with more than 93% of the coins, predominantly Gotlandic W-bracteates, found within the Gothic church. From this point of view, it can be argued that the large handling of coins is an indication of a deeply monetised society and prospering communities that became increasingly linked to the church building and the cult activities performed there. However, a more precise chronology of how offering activities developed could also turn this argument on its head. If most of the coins were offered in the second half of the fourteenth century and fifteenth century when disasters struck Gotland, such as the Black Death in 1351, the Danish invasion in 1361, the incursions of the Vitalian pirates, Teutonic knights and foreign mercenaries in the early fifteenth centuries devastating the countryside and finally the severe debasement of Gotlandic coinages under Eric of Pomerania after 1420, the picture might radically change. According to Jes Wienberg, the tipping point was the many Gothic building projects that were started and which exhausted the economy of the congregations and led finally to the demise of Gotlandic society.170 An abundance of coins in this context could be interpreted as a sign of a deep crisis rather than monetary wealth. Another distribution pattern which seems like a more peculiar feature of Bunge church is the coin concentrations along the northern wall. Based on the spatial analysis of coins, beads and pins and the comparison with wall paintings, it is argued that the church in the early fifteenth century probably changed to a pilgrim site for the cult of Corpus Christi, highlighted by new wall paintings and a changed liturgical topography in the northern half of the nave. However, there is no written evidence to

Moving money and ritual money

133

substantiate this interpretation. The towering Gothic church building, the furnishing of space with wall paintings, sculptures, screens, roods and altars, above all the offering of thousands of coins are material evidence of the spiritual and monetary engagement of the parish congregation preparing for the afterlife. The necropolis on the northern side of the Gothic nave in Bunge could be an expression of the parishioners taking up more space in the church.171 The coins, which were used in the liturgical context, also gave the priest and the parishioners an opportunity to gather capital and develop an economic framework with a basis in the coins. The title of this paper encompasses three different dimensions of a monetary economy. As moving money, coins represent a moveable resource and economic potentials enabling and allocating knowledge and services in the church building. The churches were long-term investments that required the organisation of craftsmen from outside the parishes: master builders, stonemasons, carpenters, smiths, all of whom required materials and transport, along with technical and logistical know-how. In addition to all this, the buildings programmes projects required movable resources, money. Secondly, the well-documented and richly furnished church, the ubiquity of coins, almost littering the space, and the presence of other rich archaeological material such as beads and burial pins make it possible to study the coins as an integral part of material assemblages deposited, offered, discarded and lost in the church. And finally, as ritual money, coins are tokens of value and faith, devotional devices creating links between God and people. Church finds also reflect monetary practice, ideologies and provide clues on how people in Bunge perceived and valued money in a religious environment. The stone churches of Gotland, as described in this contribution, emerged as a result of the interaction between coin-based economic ideas and the spiritual economy which was introduced into Gotlandic churches in the thirteenth century and lasted until the early sixteenth century. In this context, money, as a portable and driving resource (“moving money”), made it possible to realise various potentials, knowledge and resources. Together, money and faith created a landscape of churches. Coins represent a materialisation of value, a presence of economic potential and constellations.

Acknowledgements This article is a product of collaborations and discussions with many colleagues. Nils Blomqvist, Henrik Klackenberg, Austin Main, Victor Melander, Sven-Erik Pernler, Tryggve Siltberg, Gustaf Svedjemo and Joerg Widmeier gave me constructive comments on the manuscript. Especially, I want to thank my students Marlene Andersson, Adam Hultberg and Jennilie Svensson whose diligent and knowledgeable seminar papers have aided my interpretation of the rich material culture of the churches in Bunge and Silte. Monica Golabiewski-Lannby at the Royal Coin Cabinet, Stockholm, has given me access to the coin database of Bunge, which constitutes the empirical basis for all coin analyses conducted in this contribution. Thanks to Fredrik Sundman who translated this text from Swedish to English, and to Kristin Bornholdt Collins who patiently revised and refined the final version of the manuscript. I also wish to express my gratitude to the Berit Wallenberg foundation for providing a generous grant (BWS 2014-0061) for the documentation and analysis of the coin finds from Västergarn, Gotland.

Notes 1 Mary Douglas, Purity and Danger: An Analysis of the Concepts of Pollution and Taboo (London: Ark Paperbacks, 1984), 70.

134

Christoph Kilger

2 Jörg Widmaier, “Insel der Kirchen, Die Kartographie der Gotländischen Sakrallandschaft vom 14. bis 18. Jahrhundert,” in La cartographie de l’espace. Les voyageurs de la Mer baltique et de la Méditerranée du Moyen Âge au XIXe siècle, edited by Burghart Schmidt (Montpellier: in press). 3 An informative introduction to the medieval interior of many Gotlandic churches is provided by Justin Kroesen and Regnerus Steensma, The Interior of the Medieval Village Church, Second revised and expanded ed. (Leuven: Peeters Publishers, 2012). 4 A numismatic survey and analysis of the Gotlandic church finds has recently been presented by Susanne Fridh Mynten på ön. En studie av de mynt från 1150–1699 som hittats i lösfynd och hopade fynd på Gotland (Stockholm: unpublished MA thesis, Department of Archaeology and Classical Studies, Stockholm University, 2014). Fridh has published her results in “Mynt 1150–1699 i lösfynd och hopade fynd på Gotland,” Myntstudier 1 (2014): 1–37. Altogether, Fridh lists here 30.197 coins from 64 churches. Additional finds of c. 1,200 coins from the 2014 excavations at St Olofsholm church have not been listed by her but are included in my count. 5 An overview of all church finds in Scandinavia is provided in the introduction to this book, see Gullbekk, Kilger, Roland and Kristensen, ch. 1. 6 A comparison of church sizes, architecture and taxation on Gotland has been presented by Isabelle Wårfors, “Kyrkolandskapet i Västergarn. En studie av faktorerna bakom nedgången i det medeltida kyrkobyggandet på Gotland” (Visby: unpublished BA thesis in Archaeology, Gotland University, 2006). 7 The influence of the medieval state in the monetisation of Sweden has been stressed in Henrik Klackenberg, Moneta nostra: Monetariseringen i medeltidens Sverige (Stockholm: Almqvist & Wiksell International, 1992). Another example is found in the work of Professor Thomas Wallerström, Norrbotten, Sverige och Medeltiden: problem kring makt och bosättning i en europeisk periferi (Stockholm: Almqvist & Wiksell, 1995). Wallerström applied Michel Foucault’s discourse of power in his analysis how coins and coinage were of paramount importance in the implementation of royal power in northern Sweden. The most thorough study of the dynamic relationship between coinages, royal powers, lawgiving and coin use in the kingdom of Norway has been presented by Professor Svein H. Gullbekk, Pengevesenets fremvekst og fall i Norge i middelalderen (Copenhagen: Museum Tusculanums Forlag, 2009). 8 For instance, Dr. Nanouschka Myrberg Burström has emphasised the significance of coinage in the twelfth and thirteenth centuries and the active role of coins on Gotland in the formation of regional, urban and ecclesiastical identities in the Baltic Sea zone. See Nanouschka Myrberg [Burström], Ett eget värde. Gotlands tidigaste myntning, ca 1140–1220 (Stockholm: Institutionen för arkeologi och antikens kultur, Stockholms universitet 2008) and “The Colour of Money: Crusaders and Coins in the 13th Century Baltic Sea,” in Making Sense of Things, Archaeologies of Sensory Perception, edited by Fredrik Fahlander and Anna Kjellström (Stockholm: Department of Archaeological and Classical History, Stockholm University, 2010), 83–102. 9 Kenneth Jonsson, “Från utländsk metall till inhemskt mynt,” in Myntningen i Sverige. 995–1995, edited by Kenneth Jonsson et al. (Stockholm: Svenska numismatiska föreningen, 1995), 43–61. 10 Sven-Erik Pernler, Gotlands medeltida kyrkoliv: biskop och prostar: en kyrkorättslig studie (Visby: Barry Press, 1977): 60–65. 11 Svein H. Gullbekk, “The Church and Money in Norway c. 1050–1250: Salvation and Monetisation,” in Money and the Church in Medieval Europe, 1000–1200, Practice, Morality and Thought, edited by Giles E. M. Gasper and Svein H. Gullbekk (Burlington: Ashgate, 2015), 223–44. 12 Giles E. M. Gasper, “Anselm of Canterbury,” in Christian Theologies of Salvation. A comparative Introduction, edited by Justin S. Holcomb (New York: New York University Press, 2017), 124–42. 13 How liturgy was practised within the Swedish parish church in the Middle Ages and how the cathedral chapter within each diocese influenced the liturgical profile of each parish have been discussed by Sven Helander, “Sockenkyrkans liturgiska profil,” in Kyrka och socken i Medeltidens Sverige, edited by Olle Ferm (Stockholm: Riksantikvarieämbetet, 1991), 189–230.

Moving money and ritual money

135

14 Bengt Söderberg, “Kejsar Henrik i vågskålen och Nicolaus på skeppet: två intressanta motiv i Gotlands medeltida muralmåleri,” Gotländskt arkiv 35 (1963): 19–36. 15 Roberta Gilchrist, Medieval Life. Archeology and the Life Course (Woodbridge: Boydell Press, 2012), 169–81. 16 A description of the historic building, its architecture and the cemetery wall is provided by Sven Hedlund, “Bunge kyrka,” in Sveriges kyrkor häfte 42, edited by Sigurd Curman and Johnny Roosval (Stockholm: Riksantikvarieämbetet, 1935), 17–39. Valuable information on the history of Bunge church, the parish and renovations has been published by the local teacher and founder of Bunge museum, Theodor Erlandsson, “En gotländsk landskyrka,” Gotländskt arkiv 18 (1946), 77–86; Bunge socken genom tiderna (Visby: Bunge kommun, 1951), 23–42. An overview on all medieval churches of Gotland is provided by Gunnar Svahnström and Erland Lagerlöf, Die Kirchen Gotlands (Kiel: Conrad Stein Verlag, 1991). 17 Carl Gustaf Gottfried Hilfeling, C G G Hilfelings gotländska resor 2,1800 och 1801, edited by Torsten Gislestam (Visby: Gotlands fornsal, 1995): 47–49. 18 Nils Blomkvist, The Discovery of the Baltic: The Reception of a Catholic World-System in the European North (AD 1075–1225) (Leiden: Brill, 2005), 172–72, 290–95. 19 Tryggve Siltberg, “Lanthamnar på Gotland och hamnordningar i Sverige, Danmark och på Gotland,” Kust och kyrka på Gotland: Historiska uppsatser, edited by Per Stobaeus (Visby: Landsarkivet i Visby, 2010), 305. 20 Sven-Erik Pernler, “St Olavs kapell i Akergarn,” De hundra kyrkornas ö (1978), 24–35. 21 Dan Carlsson, Niklas Björk and Julia Hillberg, Rapport från den arkeologiska undersökningen på St Olofsholm 2013, Archaeological report, no. 7 (Visby: Arendus, 2014) and Hellvi, S:t Olofsholm. Arkeologisk undersökning, Archaeological report, no. 39 (Visby: Arendus, 2014). 22 Hans Nielssøn Strelow, Cronica Guthilandorum: Den guthilandiske cronica, huor udi beskrifuis, huorledis Guthiland er opsøgt oc paafundet, 1633 (Visby: Barry Press, 1978). 23 Viking-Age finds in churchyards form part of a burial tradition that marked the transition to Christianity on Gotland, where the deceased were still buried with ornaments and dress fittings. Not all church finds of Viking-Age character are unearthed near older wooden churches, though this is often the case, Lena Thunmark-Nylén, “Churchyard Finds from Gotland (11th–12th Centuries),” in Archaeology East and West of the Baltic, edited by Ingmar Jansson (Stockholm: Department of Archaeology, Stockholm University, 1995), 161–93. 24 A typological overview, chronology and distribution analysis of the medieval beads from Bunge has been presented by Marlene Andersson, “Medeltida pärlor. En studie av pärlor påträffade in Bunge kyrka år 1971–72,” (Visby: unpublished BA thesis in Archaeology, Department of Archeology and Ancient History, University of Uppsala, 2014). There is no comprehensive and absolute chronology for the medieval beads from Gotland. For the purpose of dating, reference needs to be made to finds from settlements and churches. 25 There are two versions of the taxus list preserved, the so-called Visbylist and the Linköpingslist. An overview and discussion on the taxus list and other bishop tithes has been provided by Tryggve Siltberg, “Biskopsinkomster och tionden på Gotland fram till 1600–talet,” Kust och kyrka på Gotland: Historiska uppsatser (Visby: Landsarkivet i Visby, 2010), 214–53. 26 Pernler, Gotlands medeltida kyrkoliv, 89–107. 27 A source-critical evaluation of the dates provided in the copies of the taxus list has been presented by Efraim Lundmark, “Bilefeld, Strelow och de gotländska kyrkornas kronologi,” Fornvännen (1925): 162–80. A more recent interpretation of the list and the chronology of the Gotlandic churches has been discussed by Lena Thunmark-Nylén, “Om de gotländska kyrkornas ålder,” Gotländskt arkiv 52 (1980): 17–34; Pernler, Gotlands medeltida, 108–10. 28 The areas of the Romanesque and the Gothic churches have been reconstructed with the help of ArcGIS. The tower 36 m 2 , the nave 56 m 2 and 176 m 2 , respectively, the choir 19 m 2 and 87 m 2 , respectively, the vestry 22 m 2 . 29 The fixtures have been listed and described on several occasions. The first inventory was carried out in 1830, Antikvarisk Topografisk Arkiv (ATA), Stockholm. I have also had access to a copy of P.A. Säves travel journal from 1864 (ATA).

136

Christoph Kilger

30 Per-Göran Ersson, Kolonisation och ödeläggelse på Gotland, Studier av den agrara bebyggelseutvecklingen från tidig medeltid till 1600–talet (Stockholm: Kultur- Geografiska Institutionen vid Stockholms Universitet, 1974). 31 Anders Andrén, “Vem lät bygga kyrkorna på Gotland?” Saga och sed (2009): 31–59. Andrén’s paper has had considerable impact in recent debates in Gotlandic historiography discussing the social environment and the initiatives behind the many church building projects on the island. 32 Andrén carefully underlines that these large farms were not magnate farms or estates owned by a landed gentry as known from other parts of Scandinavia such as the central regions of medieval Sweden and Denmark, Andrén, “Vem lät bygga,” 53. 33 Objections to Andrén’s model have been recently raised by Gustaf Svedjemo, “Landscape Dynamics, Spatial Analyses of Villages and Farms on Gotland AD 200–170,” Occasional Papers in Archaeology 58 (2014): 194–97. Mixed parcels of land which Svedjemo interprets as traces of village structures rather than large farms are found at any location of Gotlandic parishes. In addition, several clusters can be found in one parish and often they are far from the church than close to it. 34 The property pattern in Bunge has also been thoroughly analysed by Svedjemo, “Landscape Dynamics,” 160–61. 35 Anders Andrén, “Herr Peter Bingel i Vallstena,” in Skandinavische Schriftenlandschaften: Vänbok till Jürg Glauser, edited by Klaus Müller-Wille et al. (Tübingen: Narr Francke Attempto Verlag, 2017), 102–107; Jörn Staecker, “Stellvertreter auf Erden. Studien zur Ikonographie der mittelalterlichen Grabplatten Schwedens,” The European Frontier. Clashes and Compromises in the Middle Ages: International Symposium of the Culture Clash or Compromise (CCC) Project and the Department of Archaeology, Lund University, held in Lund October 13–15 2000, edited by Jörn Staecker (Lund: Almqvist & Wiksell International, 2004), 177–208. 36 Today, the four slabs are placed in pairs on either side of the high altar. They were originally located below the altar steps in front of the high altar but were moved during renovations in 1916/17. 37 The inscription in Latin says “In the year of the Lord 1294, three days before the first of April, Johannes, father of Botulf Ducker died. His soul may rest in peace, Amen.” 38 Eva Östlund-Stjärnegårdh (ed.), Revisionsbok för Gotland 1653: (“Jordeboken 1653”), Generalundervisningsbok om alla hemmans beskaffenhet i Gotlands län, Nordertredingen, vol. 3 (Visby: Barry Press, 1979). 39 Erlandsson, Bunge socken genom tiderna, 41. 40 An archaeological survey and landscape study of the farm of Hultungs in Bunge parish is provided by Dan Carlsson, Arkeologisk förundersökning och särskild arkeologisk undersökning rörande fynd av vikingatida mynt, Ducker 1:63 och Hultungs 1:6, Bunge socken, Gotland (ArkeoDok: Rapport 7, 2008). Traces of at least four more deserted medieval farms in the parish have been presented and discussed by Svedjemo, “Landscape Dynamics,” 161. 41 A thorough description and interpretation has been provided by Bengt Söderberg, Gotländska kalkmålningar 1200–1400 (Visby: Föreningen Gotlands Fornvänner, 1971), 177–84. The murals were covered with whitewashed in 1816 by the priest and Doctor of Theology D. Daniel Söderberg, despite parishioners’ protests. The reason was that churchgoers preferred looking at the paintings than listening to his sermons. 42 According to Bengt Söderberg, certain features in the paintings, such as the use of perspective in the depiction of architecture, point to influences from Italian and Bohemian art styles, which had a strong impact on northern Europe at the end of the fourteenth century, Söderberg, Gotländska kalkmålningar, 182. 43 The historian Birgitta Eimer has interpreted the battle scenes as a detailed account of a contemporary event on Gotland, the bloody battle between Danish troops and the Teutonic Order in 1404, see Birgitta Eimer, Gotland under dem deutschen Orden und die Komturei in Schweden zu Årsta (Innsbruk: Universitätsverlag Wagner, 1966), 219–21. 44 Helmer Gustavson, “Bunge kyrka. Bunge socken G 327–333,” Gotlands runinskrifter 3 (Stockholm: Riksantikvarieämbetet), 1–46. Online publication without year, https:// www.raa.se/app/uploads/2013/09/23_bunge.pdf, accessed 2020-02-26.

Moving money and ritual money

137

45 Gustavson, “Bunge kyrka,” 4–8. 46 An overview of research on the history, use and meaning of the “rood-cross pedestals” and medieval collecting boxes from Gotland is provided by Torsten Svensson, “Gotlands medeltida offerstockar och krucifixpiedestaler,” Gotländskt arkiv 64 (1992): 69–105. For a survey of medieval collection tables in Gotlandic churches, see Pia Bengtsson Melin and Kenneth Jonsson, “Medeltida kollekttavlor från Gotland,” Myntstudier 2 (2019): 1–92 [Bunge]. Tables [Latin tabula], used for collecting coins offered to the shrine of St. Olav at St. Olofsholm, are explicitly mentioned in extant episcopal statutes from Gotland, see footnote 107. 47 Anna Nilsén, Kyrkorummets brännpunkt. Gränsen mellan kor och långhus i den svenska landskyrkan. Från romantik till nygotik (Stockholm: Kungliga Vitterhetsakademien, 1991), 168–69. 48 Erlandsson’s letter to B. Schnittger 31 March 1921 (ATA). 49 Crosses with stair-shaped, probably wooden, bases are a common motif on medieval Gotlandic grave slabs, Staecker, “Stellvertreter auf Erden,” 190–92. 50 Sven Hedlund suggests a dating between c. 1235 and 1270, Hedlund, “Bunge kyrka,” 38. The art historian Carl Ramsell af Ugglas estimates a later dating between 1280 and 1300, see Carl Ramsell af Ugglas, Gotlands medeltida träskulptur till och med höggotikens inbrott: bidrag till kännedomen om stilströmningarna i Norden under den äldre medeltiden (Stockholm: Bonnier, 1915), 303. Based on the syllabic evidence of the runes and comparison with a second runic inscription referring to Lafrans, Helmer Gustavson suggests a dating before c. 1250, Gustavson, “Bunge kyrka,” 12. 51 Bengt Stolt, Kyrkligt – Kulturellt – Gotländskt (Visby: Odins Förlag AB, 2007), 9–24. Parish accounts of the early eighteenth century from Ethelhem and När mention that the sick “gave money to the cross to ensure good health”; see Svensson, “Gotlands medeltida Offerstockar och krucifixpiedestaler,” 77. 52 This decision is mentioned in a manuscript from 1611 in the Visby cathedral archive, but probably dates back to an edict by the Danish Governor Christoffer Valkendorff, who governed Gotland between 1570 and 1573, Svensson, “Gotlands medeltida Offerstockar och krucifixpiedestaler,” 78. 53 Gustavson, “Bunge kyrka,” 9. 54 Johnny Roosval, “Triumfkrucifixet i Stånga. Ett meddelande om dess konservering och nyuppsättning,” Gotländskt arkiv 2 (1939): 3–21; Lennart Karlsson, Romansk träornamentik i Sverige. Decorative Romanesque woodcarvimg in Sweden (Stockholm: Almqvist & Wiksell International, 1976), 32. 55 Nilsén, Kyrkorummets brännpunkt, 151–53. 56 In 1689, an altarpiece of sandstone was placed on the main altar. In 1702, the pulpit was built and placed above the southern side altar. The frescoes in the choir and nave were whitewashed in 1816. 57 The medieval stained glass was removed from all windows, and a loft for the organ was created on the western wall of the nave. The baptismal font was moved from the entrance and instead placed in front of the chancel arch. During the Middle Ages, the font was placed behind the western pillar in the nave. The changes have been thoroughly accounted for by Theodor Erlandsson in the newspaper Gotlänningen, 15 and 17 July 1918. Notes regarding the renovations in 1856/57 and later in 1916/17 are found in a document by Erlandsson, Något om kyrkans restaurering in 1916 and 1917, dated 24 December 1917 (ATA). 58 Today, the truncated altar slab is found in the vestry. 59 The whitewash was removed from the arches and walls. The northern portal, which had been bricked up, was reopened, and attempts were made to recreate the niches in the choir. In order to fill the voids of the medieval inscribed slabs in front of the main altar, gravestones were brought in. These post-Reformation gravestones had most likely been kept inside the church and had been moved outside during the restoration of 1856. 60 The results have been presented by Erik Bohrn (ATA). The plan and six photographs taken during the survey were not, however, available at ATA. 61 Riksantikvarieämbetets Gotlandsundersökningara, Gotland branch of the Swedish National Heritage Board.

138

Christoph Kilger

62 A compilation of all documents from the excavation has been established by Malin Eriksson at the Museum of Gotland (Visby: unpublished report, 2012). It must be pointed out that not all the material that was transferred from the Museum of Gotland to ATA in Stockholm was accessible for the assessment of the excavations and the various contexts excavated in 1971/72. More in-depth information about the excavations can be found on photographs, each of which has been given a typed description. Of the 18 posters transferred from the Museum of Gotland to ATA, the author had access to 8, 5 describing the excavations in the tower and 3 regarding the nave. 63 Hedlund, “Bunge kyrka,” 38, fig. 23. 64 The descriptions of the stratigraphy in the tower are mentioned in an incomplete draft report Sammanfattning av vetenskapliga rön vunna vid en arkeologisk undersökning av Bunge kyrka (ATA). Additional stratigraphical information is found in Waldemar Falck (ed.), Fyndförteckning till arkeologiska undersökning i Bunges långhus och tornkammare (ATA). The artefacts are recorded in relation to their respective layer. 65 Only three burials are visible on photographs and plans, graves 1–3, and the location of grave 4 is not divulged. Therefore, only three graves are included in the distribution analysis. 66 According to Falck, these were the construction layers belonging to the Romanesque nave and the tower. 67 See Ingvardson, this volume, ch. 10. 68 Account of the renovations by Theodor Erlandsson, 24 December 1917 (ATA). 69 Falck presumed that the spruce pine layer that covered the area was added during the first renovations of 1856/57. It is not clear, however, what his reasoning was based on. 70 Carlsson, Bjørk and Hilleberg Helliv, St Olofsholm, 27–28, 69–71, fig. 65. 71 Today, the slab is located in the modern stone floor, in the nave, under the chancel arch. In 1807, the rectory was demolished and a wooden pillar and a stone cist placed in the chancel arch were removed. 72 Mattias Karlsson, Konstruktionen av det heliga. Altarna i det medeltida Lunds stift. Skånsk senmedeltid och renässans 23 (Lund: Humanistiska Fakultet, Lunds University, 2015), 222, with further references. On depictions of the entombed Christ and Christ tomb sculptures as devotional objects, see Jürgensen, this volume, ch. 2. 73 See Rensbro, this volume, ch. 3; Rensbro and Moesgaard, ch. 9; Ingvardson, ch. 10. 74 Ragnar Engeström, “De medeltida myntfynden från det sista årets utgrävningar i gotländska kyrkor,” Gotländskt arkiv 44 (1972): 89–94, fig. 1–4. This number is also used by Engeström in unpublished summaries and preliminary statistical analyses in the preserved documentation (ATA). 75 Monica Golabiewski-Lannby, Royal Coin Cabinet Stockholm, has classified and listed the coins (unpublished coin list Bunge, Royal Coin Cabinet, inventory no. 5635/76). A comparison between the revised finds list of Falck and Golabiewski-Lannby’s data list differs on a few occasions regarding the number of coins and the original classifications. I have not had an opportunity to examine the whole coin material again. For future studies of Bunge, it would be important to compare these classifications to the results of this study. Monetary artefacts such as modern tokens, counting coins and other materials such as fittings and ornaments have been excluded. 76 Golabiewski-Lannby estimated that there were at least 564 fragments, letter to Gotlands Museum 2001 (ATA). 77 Jens-Christian Moesgaard, “Kirkegulvsmønter fra Gotland,” Nordisk Numismatisk Unions Medlemsblad 6 (1987): 134–41. Moesgaard’s article is based on his MA dissertation, “Myntomløbet på Gotland i middelalderen, numismatisk, økonomisk og politisk set” (Aarhus: unpublished MA thesis in History, Department of History, Aarhus University, 1986). 78 Fridh, Mynten på ön; “Mynt 1150–1699,” 1–37. 79 Myrberg [Burström], Ett eget värde. 80 Myrberg [Burström], “The Colour of Money,” 83–102. See also Myrberg [Burström], “The Hatched Cross. Gotlandic Coins of the Thirteenth Century in the Baltic Sea,” in Monetary history of the Baltics in the Middle-Ages (12–16th c.). International Symposium in Tallinn, 9–10 Dec. 2010, edited by Ivar Leimus (Tallinn: Eesti Ajaloomuuseum, 2012), 180–96.

Moving money and ritual money

139

81 On the basis of this, the suspicion expressed by Carl Johan Gardell and Stefan Simander, that some of the coin material from Bunge may have disappeared before it was registered by the Royal Coin Cabinet in Stockholm, therefore appears rather implausible, Carl Johan Gardell and Stefan Simander, Fallet Engeström, mannen som plundrade det svenska kulturarvet (Uppsala: Romlid Förlag, 2015), 90. 82 Report from the ongoing restoration made by the local church council to the Swedish National Heritage Board, 1 May 1917, item 6 (ATA). 83 In another account provided by Theodor Erlandsson to the Swedish History Museum, approximately 100 coins are mentioned, most of the very thin Gotlandic coins were supposedly found in front of the altar, Erlandsson, letter dated 10 April 1923 (ATA). 84 Fridh, Mynten på ön, 44. This count relates to the numbers presented in her master thesis. In her later publication, the total number for the church finds is higher, altogether 30197 specimens, Fridh, “Mynt 1150–1699,” 18–22, fig. 38. 85 Moesgaard, Myntomløbet på Gotland, 11–12. 86 Siltberg, “Lanthamnar på Gotland,” 305. 87 Jürgen Beyer, “Gaver fra Sønderborg-skippere till kirkene i Hellvi og Rute,” Kust och kyrka på Gotland: Historiska uppsatser, edited by Per Stobaeus (Visby: Landsarkivet i Visby, 2010), 366–77. 88 Pernler, “St Olavs kapell,” 24–35. 89 Hilfeling, C G G Hilfelings, 47–49. 90 This issue has been extensively debated in Norwegian research, Karin Berg, “Myntfunn fra Høre kirke i Valdres,” Årbok: Foreningen til norske fortidsminnesmerkers bevaring 135 (1981): 69–84; Anne-Marie Mørch von der Fehr, “Løsfunn fra Høre kirke,” Universitetets Oldsakssamlings Årbok 1984–1985 (1986): 135–43. The offering theory has been vividly argued for in Herleik Baklid’s Nær folkje kallar på Gud. Myntfunn under kirkegolv i sosialhistorisk perspektiv (Oslo: unpublished MA thesis in History, Department of History, University of Oslo, 1993). 91 Henrik Klackenberg, “Kyrkfynden och medeltidsforskningen,” in Festskrift till Lars O. Lagerqvist, red. Ulla Ehrensvärd. Numismatiska meddelanden 37 (1989): 209–19. 92 Bengt Stolt, Kyrkligt – Kulturellt – Gotländskt (Visby: Odins Förlag AB, 2007), 14: “han for till broo til offerpenninge.” 93 According to the Gotlandic currency system, one mark consisted of 288 pennies. Yrjö Hyötyniemi, “Om gutnisk mynträkning på 1400–talet,” Nordisk Numismatisk Unions Medlemsblad 3 (2000): 45–53. 94 Klackenberg, Moneta nostra, 37–38. For quantifications of money offerings in medieval society, see discussions based on Norwegian churches (S.H. Gullbekk and A. Sættem, Norske myntfunn 1050–1319. Penger, kommunikasjon og fromhetskultur (Oslo: Dreyer forlag, 2019). 95 Engeström, “De medeltida myntfynden,” fig. 1–4. 96 In the nave, there were 192 grid squares, 1 × 1 m. Grid squares 177, 178, 181,182, 186– 88, which were located along the southern wall of the nave, are all half squares and are therefore counted as seven full squares. Grid squares 159, 160, 175, 176, 191 and 192 correspond to the area of the southern side altar and the pulpit and are thus not included. 97 Mats Bergman, Silte kyrka: Hablinge ting, Gotland band VIII:3 (Stockholm Almqvist & Wiksell International: 1992), 15–19. The stratigraphy of Silte church, based on the coin finds and graves, has been analysed by Adam Hultberg, Kyrkorna i Silte. Kyrkobyggnader och kyrkfynd i det tidig medeltida Gotland (Visby: unpublished BA thesis, Department of Archaeology and Ancient History, University of Uppsala, 2015). 98 The coin finds from the Romanesque church ruin in Västergarn have not been published. 99 Carlsson, Björk and Hillberg, Hellvi, St. Olofsholm, 73f, fig. 69. 100 Hultberg, Kyrkorna i Silte, 24, fig. k2. 101 This area of 16 m 2 corresponds with grid squares 5–8, 21–24, 37–40 and 53–56. The cluster consists of foreign coins: 7 Swedish pennies of King Magnus Eriksson (c. 1340– 1353), 1 Swedish penny of King Albrecht of Mecklenburg (c. 1370–1380), 1 penny of the Teutonic Knights (c. 1390–1410), 1 German Oxhead bracteate (c. 1400–1450), 1 Norwegian hvid of King Hans (c. 1480–1500), 1 Polish penny of King Kasimir Jagiello (1447– 1492) and 1 Swedish penny of Sten Sture the Elder (c. 1470–1500), further domestic

140 Christoph Kilger

102 103 104 105 106 107

108 109 110 111 112 113 114

115 116 117 118 119

120

121

122 123

Gotlandic coins, 2 örtugs (c. 1380–1420), 14 black örtugs (c. 1420–1450) and 5 hvids (c. 1450–1520). Engeström, “De medeltida myntfynden,” 91. These churches have been visited by the author. In Silte, on the other hand, there is no cluster of coins by the southern portal. One reason could be that the entrance area was not fully excavated. Hultberg, Kyrkorna i Silte, 24, fig. k2. Sven-Erik Pernler and Anders Piltz, “1329 års förordning för Gotland – ett 650 års minne,” De hundra kyrkornas ö (1979): 23–27. DS 6324 (Diplomatarium Suecanum) “Ex pecunia] vero, que in truncis provenui[t et ex ta]bulis [tam] sancti Olaui quam monialium ex indiusio una marcha argenti dabitur curato prenominato et due marche denariorum dabuntur eidem – de pecunia sancti Ola[ui vel cappelle racion]e seruicij et laborum quos quos impendit pro eadem [et h]ijs omnino sit contentus/ Pecunia vero in truncis proueniens in tres partes equales diuidatur. Quarum unam pro fabrica ipsius cappelle deputamus; reliqu[as vero duas par]tes monasterio et monialibus prefatis mandamus pro suis usibus assignari et ídem fieri precipimus de pecunia que de tabulis sancti Olaui et monialium proueniat annuatim.” Carlsson, Björk and Hillberg, Hellvi, St. Olofsholm, 73f, fig. 69. Carlsson, Björk and Hillberg, Hellvi, St. Olofsholm, 67, fig. 62. Carlsson, Björk and Hillberg, Hellvi, St. Olofsholm, 66, fig. 51. On the topic of altar cloth offerings, see Jonsson, this volume, ch. 13. One hundred coins were found in a stone-lined pit in the apse in Høre church, probably as a result of coins being swept on the floor behind the main altar, see Risvaag, this volume, ch. 6. Säve, copy of travel journal of 1864 (ATA). Erlandsson Något om kyrkans restaurering in 1916 and 1917, dated 24 December 1917 (ATA), 9–10. DS 336 (1246), “Hinc est quod communi consensu & consilio nostro & vestro, cum deliberatione divina, Altare S. Olavi in Ackergarn cum toto proventu sacrificii, quod ibidem ab advenientibus Christi fidelibus offertur, perpetuo possidendum eisdem assignamus …” DS 362 (1248), 6324 (1360) and 9274 (1376). See Rensbro, this volume, ch. 3. See also footnote 101. Mereth Lindgren, “Kalkmålningarna,” in Signums Svenska Konsthistoria. Bd 4: Den gotiska konsten, edited by Jan-Erik Augustsson et al. (Lund: Bokförlaget Signum, 1996), 309–412, 372–73. Grid square 57: two skulls; 58: four; 59: five; 60: four; 61: two; 62: two. A chronological, typological and spatial analysis of the pins from Bunge is provided by Jennilie Svensson, Knappnålar som gravmarkörer. En studie av knappnålar påträffade i Bunge kyrka år 1971–1972 (Visby: unpublished BA thesis, Department of Archaeology and Ancient History, University of Uppsala, 2015). The manufacture of pins can be traced through written sources back to the mid-fourteenth century, when the Pinners’ Guild of York was founded. Barclay, Catherine and Martin Biddle, Artefacts from Medieval Winchester: Object and Economy in Medieval Winchester (Oxford: Clarendon Press, 1990), 560. Theodor Erlandsson was given first-hand information about the renovations in 1858 from the former school teacher, sexton and juror Gustaf Källström who participated in the work as a 20-year-old, as well as from the landowner Lars Petter Dahlström of Ducker, unpublished account of renovations in Bunge church in 1917 (ATA). Report on the ongoing restoration, from the church council to the Swedish National Heritage Board, 1 May 1917, item 5 (ATA). Different aspects of this topic in a Scandinavian context have been presented and discussed by Martin Wangsgaard Jürgensen, “Syddør, norddør òg det kønsopdelte kirkerum,” Hikuin 36 (2009): 7–28. Svein. H. Gullbekk interprets beads and pins as ornaments and therefore as evidence of female activity in the northern half of the nave. Svein H. Gullbekk, “Scandinavian Women in Search of Salvation. Women’s Use of Money in Devotional Practice,” in Divina Moneta. Coins in Religion and Ritual, edited by Nanouschka Myrberg Burström and Gitte Tarnow Ingvardson (London: Routledge, 2017), 207–25, att. 212.

Moving money and ritual money

141

124 Erlandsson, «Bunge kyrka,» 41–42. 125 Nora Liljeholm, “Gravfält kontra kyrkogård. Bysans kontra Rom? Diskussion kring det senvikingatida begravningsskicket på Gotland utifrån gravfältet Stora Hallvards och Slite kyrkas kyrkogård,” Fornvännen 89 (1994): 152–56. However, it is not entirely clear if these individuals belong to the stone church or if they relate to the earlier wooden stave church which was discovered below the church floor. A contextual analysis of both the burials and the coin finds has recently been presented by Adam Hultberg, “Kyrkorna i Silte,” 16–20. His stratigraphic analysis of the coin finds in the burial area suggests that the individuals probably belong to both churches. Hultberg has further developed this interpretation and published his results in, Adam Hultberg and Victor N. L. Melander, “Begravningar, bebyggelse och begravd bebyggelse: Långtidsperspektiv på Silte kyrka och socken, ca 1000–1400,” in Arkeologi på Gotland 2. Tillbakablickar och nya forskningsrön, edited by Paul Wallin and Helene Martinsson-Wallin (Uppsala/Visby, Department of Archaeology and Ancient History, Uppsala University and Gotlands Museum, 2017), 258–62. 126 Fourteen names are listed explicitly mentioning that they made this church, which probably represents 14 units or farms in the parish. According to Nils Blomkvist, the inscription is a statement of those worth mentioning because they belonged to the circle of elite farmers in the parish, but probably not all individuals contributed in the building process. Nils Blomkvist, “Folk och gårdar på medeltidens Gotland: En nyckelfråga för Östersjöforskningen,” Från Gutabygd (2010), 89, 88–93. 127 Andersson, Medeltida pärlor. The beads are listed in Appendix 1; photographs of all beads are presented in Appendix 3. 128 Andersson, Medeltida pärlor, 25. 129 Gilchrist, Medieval Life, 157–58. 130 Edith Holm, Glasperlen, Mythos, Schmuck und Spielereien aus fünf Jahrtausenden (München: Callwey, 1984), 15–16. 131 Roberta Gilchrist, “The Materiality of Medieval Heirlooms. From biographical to sacred Objects,” in Mobility, Meaning & Transformation of Things. Shifting Contexts of Material Culture through Time and Space, edited by Hans Peter Hahn and Hadas Weiss (Oxford: Oxbow, 2013), 170–82, 179. 132 Gilchrist, “The Materiality of Medieval Heirlooms,” 178; Katherine French, “The Material Culture of Childbirth in Late Medieval London and its Suburbs,” Journal of Women’s History 28/2 (2016): 129. 133 Gun Westholm, Visborgs slott. Danskt herresäte och piratnäste. Gotland i den politiska hetluften 1407–1679 (Klintehamn: Gotlandica förlaget, 2015), 39–70. 134 Eimer, Gotland unter dem deutschen Orden, 219–21. Eimer’s interpretaion has been supported by the art historian Bengt Söderberg, Gotländska kalkmålningar, 181. 135 Anders Andrén, Det medeltida Gotland. En arkeologisk guidebok (Lund: Historiska media, 2011), 183–84. 136 Fridh, Mynten på ön, 31, 38. 137 See also footnote 101. 138 Miri Rubin, Corpus Christi. The Eucharist in Late Medieval Culture (Cambridge: Cambridge University Press, 1991). 139 Ronald C. Finucane, Miracles and Pilgrims. Popular Beliefs in Medieval England, 2nd ed. (New York: Palgrave Macmillan, 1995), 94–96. 140 Myrberg [Burström], “A Worth of their Own,” 161–65. 141 Evald Gustafsson, “Vänge, Väte, Västergarn. Anteckningar om Gotlands romanska kvaderstenskyrkor,” in Gotlandia Irredenta. Festschrift für Gunnar Svahnström zu seinem 75. Geburtstag, edited by Robert Bohn (Sigmaringen: Thorbecke Verlag, 1990), 103–15. 142 Myrberg [Burström], “A Worth of their Own,” 161–63. 143 Hedlund, “Bunge kyrka,” 28. 144 Myrberg [Burström], “The Hatched Cross,” 182–84, fig. 2. 145 Hedlund, “Bunge kyrka,” 23–28. 146 All five cross pennies were found in grid square 195. The string of beads came from the adjacent square in the south, 198. 147 The coins were found in the following grid squares: two in grid square 101, one in 102, two in 117, two in 118.

142

Christoph Kilger

148 The beads are from grid squares 132–33. 149 Grid squares 100, 101 and 118. 150 The grave is marked on the plans but not mentioned in the report. There is no stratigraphical or contextual information. It is unclear if the grave belongs to the burial area of the Gothic or the older church. 151 The early W-bracteates are classed as LL XXXIV:1, Lagerqvist, Svenska mynt under vikingatid och medeltid. 152 Three bracteates by box no. 2, by the southern portal in the nave, and three by box no. 4 along the northern wall of the nave. 153 Myrberg [Burström], “A Worth of their Own,” 162, 175. 154 Christoph Kilger, Frédéric Elfver and Gustaf Svedjemo, “Bebyggelseutvecklingen inom Västergarnsvallen ur ett numismatiskt perspektiv,” in Myntstudier. Festschrift to Kenneth Jonsson, edited by Tuukka Talvio and Magnus Wijk (Stockholm: Svenska numismatiska föreningen, 2015), 74–89. 155 Myrberg [Burström], “The Colour of Money,” 98. 156 Myrberg [Burström], “The Hatched Cross,” 194–95. 157 Pernler, Gotlands medeltida kyrkoliv, 66–68. 158 Clusters of coins in the same location have also been recorded in the Høre and Jomala churches in Norway, see Risvaag, this volume, ch. 6, and Jonsson, this volume, ch. 13. In Høre, the cluster consists predominantly of coins minted under the Norwegian king Magnus the Law-mender (1263–1280). Risvaag argues that an offering box was placed in the chancel arch for a very limited time, probably for collecting money for the six years’ tithe for financing a crusade to the Holy Land conducted in 1276. This interpretation could also explain the coin cluster of early W-bracteates in Bunge. 159 Per Axel Wiktorsson (ed.), Äldre Västgötalagen och dess bilagor i Cod. Hol. B 59, vol. 2 (Skara: Föreningen för Västgötalitteratur, 2011), 21. B59 is a complete and late version of the law, edited around 1290. 160 Pernler, Gotlands medeltida kyrkoliv, 75–77, 86–89. However, as Sven Erik Pernler stresses, written sources such as Liber sextus are ambiguous in this respect. 161 For a similar conclusion, see Nanouschka Myrberg Burström, this volume, ch. 11. 162 Klackenberg, Moneta nostra, 38–39. 163 Klackenberg, Moneta nostra, 179–81. 164 Hugo Yrwing, Gotlands medeltid (Visby: Gotlandskonst AB, 1978), 23–25; Per Stobaeus, “Gutasagan, några tankar om dess uppkomst och ålder,” Kust och kyrka på Gotland: Historiska uppsatser (Visby: Landsarkivet i Visby, 2010): 79–83. 165 For a similar conclusion, Mats Anglert, “Från gårdskyrka till sockenkyrka – Bjäresjö i Skåne,” in Sveriges kyrkohistoria, missionstid och tidig medeltid, edited by Bertil Nilsson (Arlöv: Verbum, 1998), 208–15. 166 Leif Gren, “Platon mitt i byn. Kyrksocknens uppkomst från monumentologisk synpunkt,” in Mänsklighet genom millennier: en vänbok till Åke Hyenstrand: sammanställd med anledning av Åke Hyenstrands 50-årsdag den 13. juli 1989, edited by Mats Burström (Stockholm: Riksantikvarieämbetet, 1989), 63–70. 167 Gilchrist, Medieval Life, 169–70. 168 Pernler, “St Olavs kapell,” 30–33. 169 Stolt, Kyrkligt, 11. 170 Jes Wienberg, “Medieval Gotland – Churches, Chronologies and Crusades,” in The European Frontier. Clashes and compromises in the Middle Ages: International Symposium of the Culture Clash or Compromise (CCC) Project and the Department of Archaeology, Lund University, Held in Lund October 13–15 2000, edited by Jörn Staecker (Lund: Almqvist & Wiksell International, 2004), 285–98. 171 On the topic of privatisation of church space, see Jürgensen, this volume, ch. 2.

6

Coin finds of Høre stave church, Oppland Norway Reflections of regulation and conflict in the Middle Ages Jon Anders Risvaag

Introduction Scholars have long been drawn to the Valdres Valley and to the Høre stave church1 (Vang, Innlandet, Norway), which sits on a plateau on its northern slope. The valley is the westernmost of the large valleys of south-east Norway and ends at the Filefjell mountain pass in the north-west, which leads to the inner parts of the Sognefjord. Filefjell is the lowermost and least harsh of the mountain passes, and the Valdres– Filefjell route was the shortest and most widely used passage between the eastern and western parts of Norway during the Iron Age and the Middle Ages. 2 Yet, Høre was still physically far from the most important centres of Norway, even though the members of the upper echelons of the area maintained close relations with the royal elite of Norway. It is this remoteness, combined with the fact that the excavations at the site are amongst the best documented in Norway, 3 and the number of coins found there form a solid statistical basis for analysis that makes Høre such an interesting site. Both the building history and the interior of Høre have been published and discussed by a number of scholars since the nineteenth century.4 In the 1980s and 1990s, the archaeologists Berg and Fehr and historian Baklid took important steps in discussing the coin finds with a focus on using coins in churches as a means of dating and determining whether coins were accidentally lost or purposely deposited. 5 Fehr also discussed the coins as a tool for identifying baptismal fonts or offering boxes. While highly indebted to their findings, this study seeks to go further. In the first part, the general outline of the building history of Høre stave church will be presented according to the finds of archaeological excavations. Subsequently, coins will be mapped within the church in order to discuss the possible relationship between the coins and religious and liturgical practice – Offertory and votive offerings – preferably in Pre-Reformation times. 6 Parallel to the work of Fehr, distribution patterns will be used to discuss church interiors, i.e., possible locations of side altars, pews and alms boxes. On a macro level, some of the coin finds will also be used in a discussion on national and international affairs, i.e., crusades and the collection of tithes in relation to these. As this study will demonstrate, even though shifts can be detected in the way coins were used in some parts of the church, and in one case specifically reflects the Church–State conflicts of the thirteenth century, most coins of Høre fall neatly into the general and regulated pattern of the Offertory.

144  Jon Anders Risvaag

Figure 6.1 Høre stave church. Between the 1660s and 1680s, the church is believed to have been subjected to extensive repairs. Plans for major repairs were also presented in 1724 and 1740. The church was given its current shape in 1822, when it was expanded, and the exterior walls of the stave church were replaced. Additional repairs were undertaken in 1857, 1888–1889 and 1952. Photo: Jiri Havran

The history of Høre and the data from the excavations The church of Høre is mentioned eight times in written sources between 1327 and 1355,7 and over its life was subject to several changes of status, use and jurisdiction. As a parish, Høre was originally subject to the diocese of Stavanger, but it was transferred to the diocese of Oslo sometime before 1400.8 In 1586, the cathedral of Stavanger demanded, to no avail, that the income from Valdres and Hallingdal be restored to Stavanger.9 Sometime after the Reformation, Høre was reduced to being an annex-church subject to the Vang church and parish. In 1720, Høre was sold to local farmers, but in 1790, it was again mentioned as “the eastern annex” of Vang church.10 Høre has belonged to the parish of Vang and has been subject to the diocese of Hamar, since 1864. The church has undergone several repairs and modifications. The majority of the data used in this article stems from the archaeological excavations undertaken by the architects Håkon Christie and Jørgen Jensenius, working under the auspices of Riksantikvaren,11 in May–September 1979.12 The excavation covered all of the floors inside the existing exterior walls since the beam layer in the chancel seemed to be failing. In addition, the northern corner post had sagged by 15 cm, suggesting the possibility that rot had set in. Severe cracks were also uncovered in the southern masonry and parts were being pushed out. It was thus deemed necessary to remove the soil and stones putting pressure on the walls, stabilise the posts and establish sufficient ventilation beneath the floors. Thus, the interior was removed, and all floorboards were taken out.13

Coin finds of Høre stave church  145 Until the late 1970s, excavators of churches focused mainly on the history of the building being investigated, providing only limited detailed contextual documentation of stray finds. However, the excavators of Høre stave church employed a more refined approach to archaeological finds, albeit one that still raises some challenges. Finds of coins and other artefacts were documented in different ways, from exact XY-positioning (e.g., 2.5X/11.5Y) via wider XY-positioning (e.g., 5–8X/14–17Y) to a simple record of the part of the building structure and type of layer. The varying quality of documentation has made it necessary to plot the finds in GIS into polygons of varying sizes.14 Apart from coins with exact positioning, the coins form clusters

Figure 6.2 Høre church divided into polygons according to find descriptions. Map: Steinar Kristensen, UiO: Museum of Cultural History.

146  Jon Anders Risvaag Table 6.1 Coin finds of Høre stave church Origin

First

Last

Total

%

Medieval

Norway Denmark Sweden Mecklenburg Parchim Hamburg Rostock Uncertain Coin missing

1177–1202 1042–1047 1360–1450 14th Century After 1389 15th Century 14th Century Medieval

1876 1779 1881 14th Century After 1389 15th Century 1424/25 Post Medieval

253 36 3 14 1 1 3 46 2

70.47 237 10.02 24 0.84 2 3.90 14 0.27 1 0.27 1 0.84 3 12.81 44 0.54

%

Postmedieval

66.02 15 6.68 12 0.54 1 3.90 0 0.27 0 0.27 0 0.84 0 12.26 2

% 4.18 3.34 0.27 0 0 0 0 0.54

within certain areas of the church. Unless otherwise noted, distribution maps are thus to be understood as representing the general distribution of coins within certain areas and not as providing the exact positioning of each specimen.15 The excavation produced 359 coins: the oldest is a cut fragment of a Danish penny issued by Magnus the Good (1042–1047) in Viborg; the most recent is a Swedish 25 öre from 1881. Two of the coins found are at present unaccounted for. Only 30 of the total number of coins are post-medieval (8.35%). Of the remaining coins that can be dated with certainty, 232 date from before 1320 (64.6%) and 50 from 1320 to 1510 (13.9%). Of the remaining 44 (12.26%) indeterminate medieval coins, the lion’s share could reasonably be deemed to be thirteenth-century bracteates, thus boosting the numbers for the earliest period. The physical framework for archaeological finds at Høre changed several times between c. 1100 and 1822, and Christie and Jensenius recorded the changing sequence of buildings with extraordinary care. Their work provides important information on the structural remains and boundaries for the artefactual finds from different periods. Based on the 1979 excavations, the history of the site of the church can be divided into four chronological phases,16 which will form the basis for the discussions of coin finds below: 1

2

3

Eleventh-century Christian burial site Consisting of a total of 22 burials, nine of which can be attributed with certainty to the phase. The remaining 13 all have the same orientation, suggesting they belong to the same phase.17 No building or structure has been identified in the vicinity. c. 1100 Posthole church Consisting of 18 postholes that outline two rooms, one larger than the other, oriented east–west, measuring 3.6 x 3.6 m (nave) and 1.9 x 2.1 m (chancel). The building was set on top of the existing Christian burial site. c. 1180 Stave church Consisting of a rectangular nave and a chancel. Both nave and chancel had raised roofs. The church was divided into nave, chancel with apsis and porch, and pentices on three sides: north, south and west. The interior was richly decorated in what is often referred to as “Sogn-Valdres style.”

Coin finds of Høre stave church  147

Figure 6.3 Map of building stages and burials. (A) Posthole church c. 1100–1180, (B) Stave church c. 1180–1822. Drawing: Jørgen Jensenius, Riksantikvaren.

4

1822 demolition of stave church walls and expansion to the current shape Maintained the inner structure but expanded the length of the church by c. 4.5 m. The north and south walls were positioned along the outer perimeters of the former pentice, making the church more than two metres wider. Only one coin has been found in connection with the phase, and the phase itself is not included in the general discussion.

The Christian burial site Only one coin find can be related to this earliest phase: an eleventh-century fragment of a Danish penny found in a grave. The archaeological evidence from this first phase consists solely of graves, a common feature in Norwegian stave churches. The burial site was situated on the Kvie farm. Based on the evidence of place names, several of the farms of Høre can be dated from the fifth century to the seventh century, and the vin farm Kvie was apparently the most prominent.18 However, so far, no building or structure in the vicinity has been identified as belonging to the burial site. The site was, as Jensenius puts it, “with or without a building.”19 Nine graves most certainly belong to the phase, while another 13 graves have the same orientation, and are thus considered as possibly belonging to the same phase.20 The phase should be viewed in the context of the forced Christianisation of Valdres by St Olav in 1023, 21 when the area became, apart from the northernmost parts of Norway, the last area of Norway to be Christianised. 22 Kvie’s apparent prominence seems to have made it the chosen place for a burial site, and maybe a church, for the new religion, although we do not know whether the burial site was in use only for the farm or for a wider populace. The absence of a church may be explained by one of two factors: either a church was built near the cemetery and has yet to be found and archaeologically examined, or the cemetery was in place well in advance of the church. 23 In the south side of what is today called the chancel, a coin was found in or underneath a cranium in grave 2: a cut fragment of a Danish penny issued by Magnus the Good (1042–1047) in Viborg. 24 Taking a broader view of the Norwegian material, there is significant evidence showing that a custom of placing coins in graves existed

148  Jon Anders Risvaag both in Christian and non-Christian graves until the end of the eleventh century.25 This may well be interpreted as a form of cult continuation in a time of religious transition. The fragment is in a poor state of conservation, and this may even have been the case before being put in the ground. It is reasonable to assume that it has been in circulation for a considerable time, presumably several decades, as similar coins that have been found in Norwegian hoards from the period c. 1065–1080. 26 The fact that the coin has been cut certainly suggests that it arrived in Norway at a time when there was a predominant silver economy, before or during the establishment and consolidation of the Norwegian coinage by Harald Hardrada (1047–1066) and Olav Kyrre (1067–1093).

The posthole church Apart from the building structure, little is known of the earliest church building at Høre. Eighteen postholes were uncovered in the course of the 1979 excavations, constituting a small wooden building, possibly constructed at the end of the eleventh century, with two rooms, one larger than the other. The structure was oriented east– west and was built on top of the burial place. 27 The nave measures 3.6 x 3.6 m, and in both the north and south walls of the nave, there were traces of two middle posts. The chancel measures 1.9 x 2.1 m. The posts were circular with a diameter of 30–40 cm, and these were burned at the end and placed in holes at a depth of 1 m and with a diameter of up to 1.5 m. A flat rock was placed at the bottom of each hole, and the posts were surrounded by tilted stones. 28 As for the burial site, only one coin find can be related to the posthole church: a fragment of a Danish penny from the second half of the eleventh century found at the bottom of a posthole.29 The fragment was found in a sand layer underneath the stone at the bottom of posthole no. 15, which is presumed to belong to the lower part of the roof construction.30 Two possibilities arise from this find: (i) the coin was deliberately placed in the posthole during the erection of the church, 31 or (ii) the coin was originally lost or placed in a grave at the burial site predating the church and accidentally fell into the posthole during construction works. The absence of any grave in close proximity could point towards the coin having been deliberately placed in the posthole. This is strengthened by the fact that the coin was recovered from underneath the bottom stone and not from the sides or filling. A parallel find has been registered in the Uvdal stave church in Buskerud, where two coins c. 1100–1130 were found under two stones in the stabilising filling for a posthole belonging to an older posthole church. The coins could not have fallen into the fill and must have been placed there prior to the erection of the post. That particular posthole was considered to be one of the best preserved and has been seen by several as providing a reliable date for the construction of the posthole church of Uvdal.32 Despite its poor conservation, the attribution to Denmark and minting in the second half of the eleventh century of the coin from the Høre posthole, together with considerations on the history of circulation, provides an indication that the coin could hardly have been in use later than the mid-twelfth century and more likely ended up in the posthole shortly after c. 1100. However, the coin’s contextual relationship to the building structure tells us that a clear purpose was at play when it was deposited. Like the previously mentioned coin from the grave, this may be perceived as a form of cult continuation or a reminiscence of earlier belief systems. Still, it may also be regarded in a Christian context as an offering to God or a patron saint made at the foundation of the building.

Coin finds of Høre stave church  149

Figure 6.4 First drawing of the posthole church. Posthole 15 containing a coin marked with “M” (Jensenius’ excavation cards/log 2 September 1979). Drawing: Jørgen Jensenius, Riksantikvaren.

The stave church The stave church at Høre is a long building divided into a rectangular nave, chancel with apsis and porch, with pentices to the north, south and west. Above the central nave is a raised roof supported by four corner posts, and the chancel also had a raised roof. The current structure has only a west entrance; however, “two doors in the south wall” were mentioned in 1740.33 The church is closely related to the stave churches of Lomen, Hegge and Hemsedal.34 The original ridge turret, now serving as part of the entrance to the cemetery, is similar to the turret of Borgund, Sogn, c. 80 km west of Høre, at the end of the Valdres–Filefjell route. Apart from the richly decorated portals and corner posts, none of the original medieval interior has survived. 35 The stave church is dated to 1179 both by dendrochronology and by a runic inscription on the pulpit, which reads: “It was that summer the brothers Ellingr and Auðun had (trees) cut for this church, when Jarl Erlingr fell in Nidaros/Niðarós.”36

150  Jon Anders Risvaag

Figure 6.5 Drawing of Høre stave church prior to the rebuilding in 1822. Unknown artist.

Elling was known as lendmann (lendr maðr) at the time and had his seat at the Kvie farm.37 Jarl Erling, who was King Sverre’s main opponent and father to the king sanctioned by the Church, Magnus (1161–1184), was killed at the battle of Kalvskinnet in Nidaros (Trondheim) on 19 June 1179.38 The inscription is corroborated by dendrochronology, which shows that the trees for the church were actually cut in 1178/79 and that the church must have been built that year or shortly after.39

The chancel One issue raised by the numerous coins found in the chancel area in Høre stave church, and in many other medieval churches, is that of lay people’s access to the choir. The altar, the holiest space in the church, both at Høre and in other churches, was situated in the chancel. How this affected the way the room itself was regarded is not clear, although there are clues in some provisions by Norwegian archbishops. In the 1320s, Archbishop Eiliv Arnesson Korte (1309–1331) issued a statute forbidding laymen, except for those participating in the reading or song, from entering the choir during Mass.40 We do not know whether this was a confirmation of a general practice or an attempt to reinstate the general rule of barring the laity from the choir. Later in the statute, Eiliv makes a very distinct exception to the rule, allowing access to the choir on Easter Sunday and on other days when the priest should offer the sacrament of the Body of God (guds likama). All receiving the sacrament should fall to their knees at the altar.41 This would be in line with, and was clearly inspired by, the Fourth Lateran Council of 1215, which highlighted Easter Sunday as a day for confessions and receiving the sacrament of the Eucharist.42 Later, Archbishop Aslak Bolt (1428–1450) stated that offerings at the main altar should be undertaken on 17 Sundays per year.43 Laypeople were given access to the choir on these particular occasions, but all these provisions suggest that laypeople were restricted, if not forbidden, from partaking in devotional practice in the chancel at most other times. Yet, the fact

Coin finds of Høre stave church  151

Figure 6.6 All coin finds. Map: Steinar Kristensen, UiO: Museum of Cultural History.

remains that the majority of the coins discovered at Høre were found in the chancel (approximately 66%, 238 coins). At first glance, this might appear to contradict the injunctions of the archbishops prohibiting the laity access to the choir. But, the appearance of so many coins in the chancel could, in fact, prove to be perfectly by the book considering the number of occasions offerings at the altar were to be performed. It cannot be dismissed, however, that offerings may have been given to the priest who subsequently brought the offerings to the altar, but still, the provisions of the archbishops, Eiliv and Aslak, that specifically mention offerings at the altar point in another direction. In the 1946 excavations of S:t Jörgens, Åhus, Scania, an interesting

152  Jon Anders Risvaag observation was made concerning the role of the altar in connection with coin offerings. Of the 176 coins found inside the church, 69 were found in the chancel, which makes a total of 36.9%. Interestingly, the number of coins in the chancel drops radically c. 1360, from 65 coins before to four after. Admittedly, there is a significant drop in the finds from the nave, but it is not as marked as in the chancel. The explanation seems to be closely connected to the whereabouts of the altar. In 1361, the original altar was removed from the chancel, and a new altar was erected at the centre of the nave. Concurrently, there is an increase in the number of coins in the area of the new altar.44 It is not unusual to find high proportions of coins in chancels in Norway, Sweden and Denmark; the churches of Bø, Lom, Eidskog, Bjärka, Dverstorps, Nedre Ullersrud, Hamneda and Aggersborg are some examples.45 At Høre, the proportion of coins in the chancel was already high shortly after the church was built after c. 1180 –

Figure 6.7 Coin finds in the chancel. Two nearby clusters of 13 and 100 coins respectively were found in the apsis. Posthole 15 with the Danish eleventh-century coin is marked with a number. Map: Steinar Kristensen, UiO: Museum of Cultural History.

Coin finds of Høre stave church  153 over 50% at the time of King Sverre Sigurdsson (1184–1202) – and rose significantly in the Late Middle Ages. Coins are found spread from the middle to the eastern part of the chancel. Two clusters of 13 and a remarkable 100 coins were retrieved close to each other in the apsis. Nearly 90% of the coins from the chancel were found in the eastern two-thirds of the structure. The areas on both sides of the chancel arch, both in the nave and the chancel, were heavily disturbed, and notably, a large section of the western part of the chancel, bordering up to the screen, produced very few finds.46 The disturbances of the archaeological context in this area indicate that soil and coins had been moved or removed. Another remarkable feature must be noted: apart from two coins from King Håkon Håkonsson (1217–1263), all medieval coins found in the western side of the chancel arch area – in the nave – stem from the time of Magnus the Law-mender (1263–1280). The concentration of coins from Magnus the Law-mender will be discussed below when dealing with the nave. The most extraordinary feature of the chancel coin finds is the group of 100 coins found in a stone-lined pit in the apsis (coordinates 5x/17y). The pit is described as a semicircle, resembling some sort of “mini tomb” that sliced through parts of the older postholes nos. 1 and 14.47 The coins found there represent the whole Medieval period from King Sverre Sigurdsson (1184–1202) to Archbishop Gaute Ivarsson (1474–1510), and no post-Reformation coins were discovered. The coins were found in brownish soil and seemingly deposited “in a way that they are presumed swept down over time.”48 Alongside the coins in the pit were a wide variety of fragmentary artefacts, such as personal belongings, remains of foodstuff and building material (C 35005: 264–203).49 No matter what the intended use of the pit may have been, it seems to have developed quite rapidly after the construction of the church into a place into which dust and debris were swept. This may also account for the nearby cluster of 13 coins in the fillings of the apsis (4–5.5x/16–17y), which, apart from one Norwegian skilling from 1694, display the same pattern as the ones from the pit. During the 1979 excavations, a large stone, measuring c. 60 cm at its widest point, was found in the centre of the chancel between postholes 2 and 13. The top of the stone was flat and level, and the base ended in a prong. All signs indicate the stone having been deliberately placed in this position, leading Jensenius to suggest that it was the base of an altar. The possible identification of the foundations of the altar may prove to be the very key to explaining the high proportion of coins in the chancel. The coins from the chancel could conceivably have spilled from the communion offering on the altar and subsequently been swept with other debris into pits, fillings and crevices. This interpretation would fit nicely with the archbishops’ provisions for offerings being left on the altar up to 17 times per year, which would on average allow laypeople into the chancel nearly every third week throughout the year. These offerings are thus to be regarded as a standard element of the liturgy proper, as part of the Offertory.50 The coins found in the chancel, therefore, reflect the laity’s part in this official liturgical practice, sanctioned by Mother Church and authorised by Canon Law. Antedating the Fourth Lateran Council, the coins clearly show the longstanding importance of the high altar as a place for offering and how a universal practice even

154  Jon Anders Risvaag

Figure 6.8 Stone-lined pit in the apse. Photo: Anne Sidsel Herdlevær, Riksantikvaren

extended into a remote Norwegian valley as early as the late twelfth century.51 Furthermore, the absence of post-Reformation coins may indicate both a shift in offering practices and the fact that the Høre church seems to have more or less fallen out of use for a time in the sixteenth century.

The nave The fact that only roughly 44% of the coins at Høre were found in the much larger nave, where the laity was free to enter, would suggest that the majority of coin offerings took place in the chancel on the designated occasions, namely the Offertory. However, clearly, coins were used and lost in the nave. In the following, the distribution and chronological development of the coin finds will be discussed in the light of possible side altars, alms boxes or pews, even though no traces of these, aside from the possible main altar base, were found during the excavations.52 In total, 28 coins were found in the central nave. Apart from two seventeenthcentury coins found in the eastern part near the chancel arch, the coins were evenly distributed throughout the area: 31 coins were found in the west aisle, 37 in the south aisle and seven in the north aisle. An analysis of the coins and disturbances of the soil in the church reached the conclusion that the coins in the south aisle should be regarded as found in situ.53 In addition, three coins were found in the porch. Ten of the coins found in the south and west aisles can with some certainty be placed in the south-west corner where the aisles meet and seven in the south-east corner. Ideally, one would have hoped for a clearer pattern for the coins in the aisles. As noted, the choice was made when constructing the maps, however problematic, to allocate all coins found in the existing aisles to the pre-1822 aisles, thus leaving the medieval pentice virtually bereft of coins.54 There is a noticeable difference in the frequency of finds between the northern and southern parts of the nave. This raises the question of gender separation in the church. As a general rule, the nave would be divided into male and female sides, south and north respectively.55 However, this explanation for the distribution of coins in the nave does not explain why the finds were distinctly concentrated in the aisle and not

Coin finds of Høre stave church  155

Figure 6.9 All coin finds in the nave. Map: Steinar Kristensen, UiO: Museum of Cultural History.

more evenly spread across the whole southern part of the church. Another factor was clearly at play in the south aisle of Høre church.56 A distinct pattern can be seen among the coins of Sverre Sigurdsson (1177–1202) and Håkon Håkonsson (1217–1263), which are concentrated in the south and west aisles. The south aisle might be related to a side altar and the concentrations in the west aisle to a baptismal font or an alms box.57 Concerning a possible side altar, it is presumed that most parish churches had at least two side altars, preferably one on each

156  Jon Anders Risvaag

Figure 6.10 Norwegian coins c. 1177–1202. Map: Steinar Kristensen, UiO: Museum of Cultural History.

side of the chancel arch.58 The Virgin facing the congregation on the Gospel side, i.e., the north side, and the patron saint on the south side (Epistle side). Still, there are no indications to the identity of the patron saint of Høre, neither in the written evidence nor in the physical remains from the church. In contrast to the offerings made at the main altar, which are to be regarded as part of the liturgy proper, and the official act of receiving the sacrament, the offerings made at the side altar are better seen as votive offerings to the saint, thus reflecting the personal beliefs or relationship to the divine of the individual.

Coin finds of Høre stave church  157

Figure 6.11 Norwegian coins c. 1217–1263. Map: Steinar Kristensen, UiO: Museum of Cultural History.

The possibility that a baptismal font may once have been placed in the western part cannot be completely ruled out, but its likelihood seems remote, and evidence for it is hard to find. As the floor has been replaced, there are no traces of water spilling during baptisms over time, nor have traces of water deposits or rot underneath the floor been registered. The latter might be due to the simple fact that this was not investigated during the excavation. Rot identified in 1979 was concentrated in the northern part of the nave; specifically, in the raft beam and both of the two northcorner posts.59 Relating a baptismal font to the rot in the north-west post appears to

158  Jon Anders Risvaag be a long-shot as the rot was a problem for the north side in general, and the issue appears to have had structural causes. Yet, there are other possibilities left to explore. Given the existence of a south entrance to the church in Medieval times, one may have to view as a whole the coins in the west and south aisles. Even though the ten coins placed with certainty in the south-west corner, where the aisles meet, are by rights too few to be considered a cluster or concentration, one might still cautiously suggest that they could point to something having been placed in this corner, just between the west and south entrances. This “something” might have been an alms box. Whether the alms box was primarily used as the congregation entered through the south door or left the church through the west door (or both) is not clear. The intended purpose for the income from the possible alms box is not clear either. One suggestion may be a connection with the fabrica ecclesiae.60 Thus, in summary, I would suggest that the explanation for the concentration of coins in the south and west aisles was that there was an alms box to the left (the southwest corner) of the south entrance and a side altar in the south-east corner of the nave. The coins dating from the period 1263–1319 display tendencies that are similar to those of the coins of Sverre and Håkon. The 44 indeterminate medieval coins show some of the same tendencies, with 24 coins in the chancel and the rest in the nave: seven spread in the middle of the space, five in the west aisle, four in the south aisle and two in the north aisle. Although not shown on the maps, most of these can reasonably be regarded as thirteenth-century Norwegian bracteates, thus enhancing the picture already seen. In the complex spatial and chronological distribution of coin finds within the nave and the chancel, there is a notable difference in the distribution presented by a number of coins from the reign of King Magnus the Law-mender (1263–1280) in the chancel arch area (i.e., on both sides bordering the chancel arch).61 Being limited to the coins of one king and a time span of less than 20 years, this phenomenon stands out as an oddity. One possible answer might be that an alms box was temporarily located by the crucifix, as has been identified in Svinnegarns church, Uppland, Sweden.62 Some rudimentary evidence suggests the introduction of some sort of layman’s altar in Scandinavia during the High Middle Ages, which might also provide an explanation.63 However, as the offering and spillage in the chancel arch area seem to have occurred over just a few years, another explanation presents itself; importantly, this explanation echoes events in the Mother Church. Upon becoming pope in 1271, Gregory X conceived an ambitious plan, which he hoped would bring political and religious stability to Europe, unite the Eastern and Western Churches by a treaty with Michael VIII Palaeologus (Nicaean emperor, 1259–1261; Byzantine emperor, 1261–1282) and resolve the problem of the Holy Land. Three themes were outlined for discussion at the Church council in Lyons in 1274: union with the Greeks, the reform of the Church and the crusade.64 It was this last theme that is likely to have been reflected in the Høre church and which could explain the concentration of Magnus’ coins. As a direct consequence of the first constitution of the Lyons council, a six-year tithe for a crusade to the Holy Land was instigated in 1274.65 The pope allowed the Church in Norway to delay the start of the collection, not only in part because of Archbishop Jon Raude’s shipwreck on his return from Lyons but also because of the number of officials needed to carry out the collection in each vast Norwegian diocese.66 In addition, instructions were issued for

Coin finds of Høre stave church  159

Figure 6.12 Norwegian coins c. 1263–1319. Map: Steinar Kristensen, UiO: Museum of Cultural History.

tithes to be given in milk produce, fish or poor Norwegian coins to be changed into silver or gold.67 The accounts of the papal nuncios show that a total of 9,900 marks of Norwegian coins were collected in Norway in 1285.68 The concentration of coins solely from the reign of Magnus the Law Mender found at Høre suggests that, for a limited time, an offering box dedicated to the collection for the crusade was placed in the vicinity of the chancel arch.69 As far as we can establish from coin finds in churches all over Scandinavia, concentrations of coins typically are not limited to a short period like this but rather encompass finds of coins from considerably longer

160  Jon Anders Risvaag stretches. When a concentration of coins of one specific coin type from a very limited period of time presents itself within a cluster, it strongly suggests special measures and considerations. Assuming this is correct, it would be the only evidence left in a Norwegian church of a large-scale operation running in the Archdiocese of Nidaros.70 Considering the fact that the collection period ran from 1276 to 1282, one would expect some coins belonging to Eirik Magnusson (1280–1299) to appear in the finds. Admittedly, we do not know with certainty when Magnus’ bracteates disappeared from circulation. A large part of the coins in circulation in the earliest years of Eirik’s reign would almost certainly have been issued by his predecessor, and it is not unreasonable to argue that bracteates would continue to be used for taxes and offerings some years into the new reign.71 Three pennies found by a metal detector at Overvik, Trondheim, in 2013 may serve as an illustration for the period of transition: two of the coins were issued by Magnus and one by Eirik. The coins were corroded together and remain inseparable.72 However, it is not clear that the collection was carried out in full during the reign of Eirik. Beginning just a couple of months into the new reign, as a consequence of a bitter conflict between Archbishop Jon Raude and some of the most influential members of the regency government of the 12-year-old Eirik, the Church suffered several political and economic blows. In 1280, the regency issued an amendment to the law (rettarbot), rejecting, among other things, the clerics’ claims to tithes from trade and income and forbidding the sale of silver to the Church.73 The latter was a direct blow to the Norwegian church’s ability to export the six-year tithe in accordance with the papal provisions mentioned above. The conflict culminated in the revocation of the Norwegian concordat of 1277 Sættargjerden and Archbishop Jon’s escape to Skara, Sweden in 1281, where he died the following year.74 Concerns for the collection must have reached the Holy See as, in a letter dated 4 March 1282, Pope Martin IV (1281–1285) reminded the king of the sad situation in the Holy Land and urged him not to hinder the export of the tithe.75 There are indications of local resistance to the collection of the six-year tithe during the reign of Magnus, and the Church seems to have been dependent on pressure from the king on his local subjects in order to carry out the collection.76 During the conflict between the Archbishop and the regency government of Eirik, the focus was on creating obstacles for the Church, not helping it. However farfetched it may seem, it is not unlikely that the explanation for the absence of Erik’s coins at Høre may be a reflection of the ongoing conflict between the Church and State in which the local powers at Kvie apparently sided with their close relative, the king, and abandoned the collection altogether at an early stage. Looking at the Scandinavian coins from the period 1355–1513 and the German coins from c. 1300–1500, one particular tendency stands out: apart from one coin struck for Erik of Pomerania (King of Norway 1389–1442, Denmark and Sweden 1396–1439) sometime between 1400 and 1420, all coins disappear from the south and west aisles. Of a total of 47 coins, only seven were found in the nave, placing 85% of the coins of this period in the chancel. In fact, this tendency seems to have started with the coins of King Håkon V Magnusson (1299–1319), with only one coin in the south aisle and none in the west aisle. This clearly indicates some changes in the religious practice and the way money was being used. The most obvious conclusion seems to be that the original offering box had gone out of use by the early 1300s. The overall number of coins found also dropped significantly in the period. This could be explained by a recession in the fourteenth century, which reduced the output of

Coin finds of Høre stave church  161 Norwegian coins and limited the amount of coin in circulation.77 Valdres was also hit by the Black Death in October 1349, a time recorded in written evidence from the area as “the autumn of the (great) mortality,”78 and in general, it took both the population and the economy of Norway several generations to recover. Of the 27 post-Reformation coins identified, 13 coins were found in the nave and 13 in the chancel, while one was found in the porch. Six of the coins from the nave were spread evenly across the mid-room, and five across the south aisle. Still, the number of coins is so low that they are best left out of the analyses on distribution.

Figure 6.13 Coins c. 1355–1513. Map: Steinar Kristensen, UiO: Museum of Cultural History.

162  Jon Anders Risvaag

Figure 6.14 G erman coins c. 1300–1500. Map: Steinar Kristensen, UiO: Museum of Cultural History.

The significant drop in the number of coins in Høre sometime after the Reformation can be explained by two factors. Firstly, there was the altered status of the church as it went from being a parish church to being merely an annexe of the new parish church at Vang.79 As there is no written evidence for a priest assigned to Høre in the Post-Reformation period, the church probably went without one and was served by the priest of the main church. In contrast, the names and vitae of all priests since 1537 in the parish of Vang are known and have been published.80 Secondly, this

Coin finds of Høre stave church  163 coincides with the introduction or intensification of offering plates carried around the congregation for the collection, which helped reduce the number of coins lost through spillage.81

Conclusions Despite its position in a rather remote Norwegian valley far from the centres of the Mother Church, the coin finds of Høre distinctly reflect some of the central policies and decisions around liturgy and the collection of tithes throughout the Middle Ages. Ultimately, the changing practices brought about by the Reformation are also mirrored in the shifting patterns of the coin finds. There were three major concentrations of medieval coins in the stave church – in the chancel, the south and west aisles of the nave and in the chancel arch area – with the chancel being the only area in use throughout the time from the foundation of the stave church to the early sixteenth century. The fact that the largest body of coins was found in the chancel suggests that the majority of offerings throughout the Middle Ages took place on the main altar on Easter Sunday and other days intended for offerings, which allowed the laity to enter the otherwise off-limits chancel on these occasions. The two other concentrations have more limited time frames. The finds from the south and west aisles of the nave suggest that an offering box was placed in the south-west corner by the south entrance to the church, and a side altar was positioned in the south-east corner. These seem to have already been in place from the late twelfth century but must have fallen out of use sometime at the beginning of the fourteenth century. The reason for this abandonment is difficult to assess but stands out in sharp contrast to the continuous offerings made at the altar in the chancel throughout Medieval times. The last concentration of coins, at the chancel arch, seems to reflect an offering box put up to collect the six-year tithe for the coming crusade to the Holy Land during the reign of Magnus the Law-mender (1263–1280). This offering box seems to have fallen out of use at the very beginning of Eirik Magnusson’s reign due to the bitter conflict between the king and the Norwegian Church. Furthermore, the lack of post-Reformation coins reflects the reduced status of the church as it shifted from being a parish church to an annexe under the parish of Vang, apparently without a priest assigned to the church. Offerings may also have shifted from fixed locations to offering plates carried around the congregation. The eleventh-century coin found in the grave of the early Christian burial site may be regarded as a transitional element or cult continuation at a time when the new religion was in its initial stages. Whether this was also the case for the coin in the posthole of the first church is more open to question.

Acknowledgments I am grateful to Steinar Kristensen for creating the distribution maps for this article and forpatient guidance and inspiring discussions during their creation. Thanks also go to Harald Bentz Høgseth and Terje Thun for their assessment on the time of construction of the stave church; to James E. Knirk for advice on runes and the English translation of the inscription on the pulpit; and to Kay Celtel for copy-editing. I also

164 Jon Anders Risvaag wish to thank Axel Christophersen, Svein H. Gullbekk and fellow participants in the Religion and Money project for commenting on this article at various stages. All errors and misunderstandings remain the fault of the author. Ch06

Notes 1 Høre was referred to as Hurum prior to the 1970s. 2 Yngvar Nielsen, Det norske vejvæsens udvikling før 1814 (Christiania: Det Mallingske Bogtrykkeri, 1876), 3–4; Sverre Steen, Ferd og fest. Reiseliv i norsk sagatid og middelalder (Oslo: Frydenlunds bryggeri, 1929), 211–16. 3 All excavated objects were handled by Universitetets Oldsaksamling (now the Museum of Cultural History, University of Oslo). In addition to coins, other kinds of finds were recorded, e.g., textiles, rope, nails, wooden sticks, window glass, nuts, shells and animal bones (MA-arkiv). All non-numismatic objects are given the inventory number C 35005 and additional sub-numbers 1–1408. Each sub-number may include one or several fragments (MA-arkiv). 4 Some of the most significant studies being: Lorenz Dietrichson, De norske stavkirker: Studier over deres system, oprindelse og historiske udvikling: et bidrag til Norges middelalderske bygningskunsts historie (Kristiania: A.B. Calmeyers Forlag, 1892); Anders Bugge, “Kirkene i Valdres,” in Valdres. 900-årsskrift 1923, edited by Gudbrand O. Hovi et al. (Gjøvik: M.O. Mariendals boktykkeri, 1923), 34–110; Aslak Liestøl, “564. Hurum kirke VI,” in Norges innskrifter med de yngre runer, vol. 5, edited by Magnus Olsen (Oslo: Kjeldskriftfondet, 1960), 169–75; Roar Hauglid, Norske stavkirker: dekor og utstyr (Oslo: Dreyer, 1973); Norske stavkirker. Bygningshistorisk bakgrunn og utvikling (Oslo: Dreyer, 1976); Jørgen Jensenius, Nils Marstein and Eivind Bratlie, “Sikring av en stavkirke. Arbeidene i Høre i 1979.” Vern og virke. Årsberening fra Riksantikvaren (1979): 1–10; James E. Knirk, The Runic Inscriptions in Høre Stave Church. Høre, Vang (Valdres) Oppland (Archaeological Report, unpublished, Oslo, 1990); Erla Bergendahl Hohler, Norwegian Stave Church Sculpture, vol. I–II (Oslo: Scandinavian University Press, 1999); Jørgen Jensenius, “Trekirkene før stavkirkene. En undersøkelse av planlegging og design av kirker før ca. år 1100” (PhD diss., The Oslo School of Architecture and Design, 2001); Claus Ahrens, Die frühen Holzkirchen Europas. Schriften des Archäologischen Landesmuseums, vol. 7 (Stuttgart: Theiss, 2001); Leif Anker, De norske stavkirkene (Oslo: Arfo, 2005); Sigrid Christie, Ola Storsletten and Anne Marta Hoff, “Høre kirke,” Norges Kirker, accessed 28 November 2014, http://www.norgeskirker.no/wiki/H%C3%B8re_kirke. 5 Karin Berg, “Myntfunn fra Høre kirke i Valdres,” Fortidsminneforeningen Årbok 135 (1981): 69–84; Anne-Marie von der Fehr, “Løsfunn fra Høre kirke,” Universitetets Oldsakssamlings Årbok 1984–1985 (1985): 135–43. Herleik Balid, “Hvad der har bragt dem gjennem gulvet er desverre efter al sansynlighet en hemmelig ofring [---],” Heimen 32 (1995): 181–96. 6 On the Offertory and votiv offerings, see Gullbekk et al. and Wangsgaard Jürgensen, this volume chs. 1 and 2. 7 1327: ecclesia de Ordun (Pavelige Nuntiers Regnskabs- og dagböger, førte under Tiende-opkrævningen i Norden 1282–1334. Med et anhang af diplomer, edited by Peder Andreas Munch. Christiania: Brögger og Christies Bogtrykkeri, 1864, 24); 1343: a Horðini (DN II, no. 257); 1355: j Haurdins kirkiu sonk a Valdrese (DN XVI, no. 16); 1389: a Hordini j Hordins kirkiiu sokn (DN IX, no. 182); c. 1400: Horðini sokn, á Valdresi (RB, p. 235); 1424: Hordins kirkiu sokn italics (DN II, no. 679); 1477: Hordenne sokn (DN II, no. 902); 1538: i hardh[en] kirke sokn DN XXI, no. 835). The only priest mentioned is sirra Alfver Þronderson in 1355 (DN XVI, no. 16). 8 Mentioned for the first time as part of the diocese of Oslo in Biskop Eysteins Jordebog (RB), written sometime around AD 1400. 9 Ivar Aars, “Ut av 1500-talet. Valdresdiplom 32. 1598–1600,” Årbok for Valdres 89 (2009): 197–99. 10 Christie, Storsletten and Hoff, “Høre kirke.”

Coin finds of Høre stave church

165

11 The Norwegian Directorate for Cultural Heritage Management. 12 All excavated objects were handled by Universitetets Oldsaksamling (now the Museum of Cultural History, University of Oslo). Other participating institutions were Treteknisk institutt (Norwegian Institute of Wood Technology), Norges Tekniske Høyskole (now the Norwegian University of Science and Technology [NTNU] and the Municipality of Vang, Jørgen Jensenius, Innberetning om Hurum stavkirke etter opphold sept. – okt. 1975 (Archaeological Report, unpublished, Oslo, 1975). Prior to the excavation, Christie and Jensenius had also made preliminary surveys of the church; Håkon Christie, Hurum stavkirke, 3.1.1977 (Archaeological Report, unpublished, Oslo, 1977). Jensenius, Innberetning om de bygningsarkeologiske undersøkelsene i Høre Stavkirke 1979. 1.4.84 (Archaeological Report, unpublished, Nesbru, 1984), 7–9. Additional reports on the coin finds were made by Berg and Skaare, Karin Berg, Funnliste mynter – Høre stavkirke (Archaeological Report, unpublished, Oslo, 1979); Høre stavkirke, Hurum sogn, Vang prestegjeld, Oppland fylke (Archaeological Report, unpublished, Oslo, 1980). Karin Berg, Mynter fra Høre stavkirke. 21. november 1980 (Archaeological Report, unpublished, Oslo, 1980). 13 Jensenius, Innberetning bygningsarkeologiske, 1–2. All joints were drawn in 1:5. The four phases of the church were drawn in 1:20 and some details in 1:1. The main phases and details were photographed during the ongoing excavations, Jensenius, Innberetning bygningsarkeologiske, 14. 14 Se discussion Kristensen, this volume, ch. 4. 15 The rebuilding and enhancing of the church and the removal of floors on several occasions were important factors that needed to be taken into consideration in constructing the distribution maps. Apart from organising the coins into polygons, a decision had to be made whether the coins in the nave’s aisle were to be presented as belonging to the pre- or post-1822 aisle, i.e., all placed inside the medieval exterior walls, omitting the medieval pentice, or inside the present exterior walls, including the pentice. The documentation from the excavations did not suggest one or the other. Hence, based upon the assumption that the coins were offered or lost inside the church, the coins have been placed inside the medieval walls, omitting the pentice. As a consequence, the maps presented here will differ somewhat from the ones presented in Fehr, “Løsfunn,” 135–43, and the tendencies in the material will not be as pronounced as in Fehr. 16 Jensenius, Marstein and Bratlie, “Sikring,” 2–3. 17 Jensenius, Innberetning bygningsarkeologiske, 74. 18 Vin: natural pasture. For an introduction to Norwegian placenames see Berit Sandnes, “Linguistic patterns in the place-names of Norway and the Northern Isles,” in Northern Lights, Northern Words. Selected Papers from the FRLSU Conference, Kirkwall 2009, edited by Robert McColl Millar (Aberdeen: Forum for Research on the Languages of Scotland and Ireland, 2010), 3–14. 19 Jensenius, Trekirkene, 153. 20 Jensenius, Innberetning bygningsarkeologiske, 74–80. 21 Hkr. ch. 121; Oluf Kolsrud, “Då Valdres vart kristna,” in Valdres. 900-årsskrift 1923, edited by Gudbrand O. Hovi et al. (Gjøvik: M.O. Mariendal, 1923), 223. 22 Kolsrud, “Då Valdres,” 209. 23 For discussions of coin finds and burials in churches, see this volume Rensbro, ch. 3, Kilger, ch. 5, Rensbro and Moesgaard, ch. 9, Burström, ch. 11 and Jonsson, ch. 13. 24 Berg, “Myntfunn,” 77; Jensenius, Innberetning bygningsarkeologiske, 48; Jensenius, Trekirkene, 153. 25 Kolbjørn Skaare, Coins and Coinage in Viking-Age Norway: The Establishment of a National Coinage in Norway in the XI Century, with a Survey of the Preceding Currency History (Oslo: Universitetsforlaget, 1976), 171; K. Skaare, “Myntene fra Lom kirke,” Foreningen til norske Fortidsminnesmerkers bevaring Aarsberetning (1978): 120–24; Karin Berg, “Myntene fra Ringebu stavkirke,” Heimgrenda (1983): 51; Inger Helene Vibe Müller, “Coins in Churches: A Means of Payment? Part Two,” in Coins and Archaeology: Proceedings of the First Meeting at Isegran, Norway 1988, edited by Helen Clarke and Erik Schia (Oxford: BAR, 1989), 85. 26 Stange Churchyard, Hedmark (Skaare, Coins and Coinage, no. 19); Stavenesodden, Krødsherad, Buskerud (Skaare, Coins and Coinage, no. 31); Nordrum, Hedrum, Buskerud

166 Jon Anders Risvaag

27

28 29

30 31 32

33

34 35

36

37

38 39

40 41

(Skaare, Coins and Coinage, no. 49); Nomeland [II], Valle. Aust-Agder (Skaare, Coins and Coinage, no. 70); Måge, Ullensvang, Hordaland (Skaare, Coins and Coinage, no. 102); Gresli, Tydal, Sør-Trøndelag (Skaare, Coins and Coinage, no. 143). Jensenius, Innberetning bygningsarkeologiske, 48–50; Trekirkene, 153. A fourteenthcentury sample was taken from one of the postholes, which was given a calibrated date AD 935 ± 85 (Sample ref: T-3775). However, it was not specified which posthole the sample was taken from, nor which part of the posthole. Jensenius, Marstein and Bratlie, “Sikring,” 2. The fragment is in a poor state of preservation and cannot be identified more precisely. Berg, “Myntfunn,” 76–77; Karin Berg, “Rapport om myntfunnet,” in Jensenius, Innberetning bygningsarkeologiske, 19–20, attachment 3; Berg, 1981; e-mail from Frédéric Elfver, 6 May 2015. Jensenius, Innberetning bygningsarkeologiske, 68; Trekirkene, 151–53. Berg, “Rapport.” Herleik Baklid, “… nær folkje kallar på Gud…” Myntfunn under kirkegolv i sosialhistorisk perspektiv (MA-thesis, unpublished, University of Oslo, 1993), 58–59; Sigrid Christie and Håkon Christie, Norges kirker, Buskerud, vol. III (Oslo: Gyldendal, 1993), 22; Jensenius, Trekirkene, 147. Knut Hermundstad, Høre stavkjyrkje (Fagernes: Valdres Trykkeri, 1968), 34. After the rebuilding, the two south portals were included in the building but were not in use as doors. One portal (Hurum II) is positioned on the south wall of the porch, the other (Hurum III) in the chancel screen (Hohler, Norwegian, 174–77; cat. no. 107 and no. 108). Roar Hauglid, Norske stavkirker. Bygningshistorisk bakgrunn og utvikling (Oslo: Dreyer, 1976), 328; Hohler, Norwegian, 15; Ahrens, Die frühen Holzkirchen, 283. Stylistically, the decorations belong to the so-called “Sogn-Valdres style,” a term first coined by Dietrichson, De norske stavkirker, 62; Roar Hauglid, Norske stavkirker. Dekor og utstyr (Oslo: Dreyer, 1973); Hohler, Norwegian, Ahrens, Die frühen Holzkirchen, and Anker, De norske stavkirkene have written extensively on the topic more recently. “Þá, um þat su[mar létu] þeir brœðr Erlingr ok Auðun hǫggva till kirkju þessar, er Erlingr j[arl fell] í Niðarósi”; Liestøl, “564. Hurum kirke,” 174; James E. Knirk, Befaringsrapport: Høre stavkirke. 12.11.1989 (Archaeological Report, unpublished, Oslo, 1989b), 2. According to Knirk, a more literal English translation would be: “Then, that summer, the brothers Elling and Audun had (trees) cut for this church, when Jarl Erling fell in Nidaros” (email from James E. Knirk, 4 May 2015). Lendmann (pl. lendmenn) refers to men receiving income from royal lands in order to uphold peace and conduct certain administrative tasks. The title lendmann was replaced by baron in 1277. In 1308, the title ceased to exist (SNL, accessed 4 May 2015, https://snl. no/lendmann). The title jarl is the highest-ranking chieftain next to the king, equivalent to the Anglo-Saxon eorl. Jarl is found translated as comes in Latin inscriptions on early thirteenth-century Norwegian coins. Ola Storsletten, Høre kirke. Vang kommune. Dendrokronologiske prøver, rapport. 28.7.2000 (Archaeological Report, unpublished, 2000); Terje Thun, Dendrokronologi Høre stavkirke, Vang kommune, Oppland, notat 29.8.2000 (Archaeological Report, unpublished, Trondheim, 2000). Terje Thun, the foremost Norwegian expert on dendrochronology, is of the opinion that the timber was in use the very same year as the trees were cut. The trees were cut during the winter and spring to be used the following summer. To his knowledge, this was the case in churches where the year of construction is certain. Major constructions like stave churches would have taken approximately two years to build and the timber was cut over two winters, as is the case at Urnes. Drying of the wood would make it more impractical and harder to use as a building material (email from Terje Thun, 13 March 2015). “Leikmonnum skal æigi ϸolas at standa i sognhusi hia klerum medan ϸeir flyktia tidir. utan ϸeim einum sem lesa edr syngiaϸ med ϸeim,” NGL III, 266–67. On access of the laity to the chancel, see also Wangsgaard Jürgensen and Roland, this volume, chs. 2 and 7. “Prestar skolo telia fyrir soknarfolki sinnu á pascha dag edr ϸa daga adra. er ϸeir skolo ϸeim guds likama gefa. at menn gangi med allri spekt ok virdingu till ϸess heilags embættis.

Coin finds of Høre stave church

42 43 44 45

46 47 48 49

50 51 52 53 54 55

56

57 58 59

167

ok falli huerr a kne fyrir alltari [sa er kal skal. ok standi eigi upp fyrr. edr gangi i brott en hann hafi alϸingis neytt guds likama ok skolad uel munn finn i uini edr uatni,” NGL III, 267. Concilium Lateranense IV, Constitutiones 1.21, Norman P. Tanner, Decrees of the Ecumenical Councils, vol. 1 Nicaea I to Lateran V (London: Sheed and Ward, 1990), 245. Olav Tveito, “Mynter i messen – Kirkefunnene som bidrag til kunnskap om offerpraksis og kirkeskikker (11.–17. årh.),” Historisk Tidsskrift 3 (2015): 395; NGL 2. r. I, 552. See also Wangsgaard Jürgensen, this volume, ch. 2. Mats Petersson [Malmer, Mats], “S:t Jörgen i Åhus,” Meddelanden från Lunds Universitets Historiska Museum (1948): 238–40. Henrik Klackenberg, Moneta nostra: Monetarisering i medeltidens Sverige (Stockholm: Almqvist & Wiksell International, 1992), 202, 204, 229–30 and 241–42; Roland, this volume, ch. 7; Svein H. Gullbekk and Anette Sættem, Norske myntfunn 1050–1319. Penger kommunikasjon og fromhetskultur (Oslo: Dreyer, 2019), 269. Jensenius, Innberetning bygningsarkeologiske, 42–45. Jensenius, Innberetning bygningsarkeologiske, 47–48. Jensenius, Innberetning bygningsarkeologiske, 54. The fragments are as follows: pin (1); stud (2); rivet (1); copper (2); leather (2); textile (7); paper (17); thread (15); mussels (17); nuts (1); bone (2); mortar/shell filling (1); iron (7); wooden stick (1); unidentified wooden object (1); window shard (46); unidentified glass (2); knife? (1); bottle? (1) (MA-Arkiv). On how the Offertory was performed, see Wangsgaard Jürgensen, this volume, ch. 2. Tveito, “Mynter i messen,” 391–96. According to the excavation report, there was generally much disturbance of the soil due to rebuilding, repairs and graves. Jensenius, Innberetning bygningsarkeologiske, 29–42. Baklid, “…nær folkje kallar på Gud,” 208–17. Three coins with an accurate positioning in the south pentice are visible in the maps. Alternate divisions have been discussed by Margaret Aston, “Segregation in the Church,” in Women in the Church, edited by William J. Sheils and Diana Wood (London: Basil Blackwell,1999); Katherine L. French, “The Seat under Our Lady: Gender and Seating in the Late Medieval English Parish Churches,” in Women’s Space: Patronage, Place and Gender in the Medieval Church, edited by Virginia Chieffo Raguin and Sarah Stanbury (Albany: State University of New York Press, 2005); Corine Schleif, “Men on the RightWomen on the Left. (A)Symmetrical Spaces and Gendered Places,” in Women’s Space: Patronage, Place and Gender in the Medieval Church, edited by Virginia Chieffo Raguin and Sarah Stanbury (Albany: State University of New York Press, 2005), 207–49 and Martin Wangsgaard Jürgensen, “Syddør, norddør og det kønsopdelte kirkerum,” Hikuin 36 (2009), 7–28; Svein H. Gullbekk, “Medieval Scandinavian women in search of Salvation,” in Divina Moneta: Coins in Religion and Ritual, edited by Nanouschka Myrberg Burström and Gitte Tarnow Ingvardson (New York – Oxford; Routledge, 2017), 209–27. East–west divisions are known, but divisions other than north–south seem to be exceptional. In the following, the medieval coin finds of Høre have been divided into four maps according to chronology and one map displaying late medieval German coins. The division is based on the fact that all coins found, apart from the German coins, are regarded as domestic. This is valid for all Swedish coins 1319–1355 due to the union between Sweden and Norway, and Danish coins post-1380 due to the union between Denmark and Norway (1380–1814). As the bulk of the coins were found in the chancel and have been dealt with earlier, the following focuses mainly on the results from the nave. Fehr, «Løsfunn.» Bernt C. Lange, «Altare, Norge,» in Kulturhistorisk leksikon for nordisk middelalder, vol. I, Abbed-Bilde, edited by Finn Hødnebø (Oslo: Gyldendal,1956), cols. 113–14. Nils Marstein, “Rapport om reparasjon av stavene. 2. oktober 1979,” in Jensenius, Innberetning bygningsarkeologiske, attachment 1 (Archaeological Report, unpublished Oslo, 1979); Eivind Bratlie, “Sikring av råtesvekket tre med plastmaterialer i Høre kirke sommeren 1979,” in Jensenius, Innberetning bygningsarkeologiske, attachment 2 (Archaeological Report, unpublished, Oslo, 1979).

168 Jon Anders Risvaag 60 Martin Wangsgaard Jürgensen, “When the Coin Left the Hand: Devotional Use of Money for Offerings in (Late) Medieval Northern European Societies,” in Coins in European Churches – Religious Practice and Devotional Use of Money, edited by B. Zäch and S.H. Gullbekk. Schweizerische Numismatische Rundschau 97 (2020): 87–88; Kilger and Myrberg g, this volume, chs. 5 and 11. 61 Eighteen coins were spread in the easternmost part of the nave and westernmost part of the chancel. No other coin types have been found in relation to these. 62 Klackenberg, Moneta nostra, 293. 63 Tveito, “Mynter i messen,” 397–98. 64 Tanner, Decrees, 303–307. 65 Concilium Lugdunense II, Constitutiones (1b) (Tanner, Decrees, 310–11). 66 DN VI, no. 37; DN VI, no. 38; DN VI, no. 40; RN II, nos. 154–60. 67 DN VI, no. 39. 68 Pavelige Nuntiers Regnskabs- og dagböger, 12–14; Svein H. Gullbekk, Pengevesenets fremvekst og fall i Norge i middelalderen (Copenhagen: Museum Tusculanum, 2009), 79. 69 För a similiar cluster of coins near the chancel arch from the same time period, see Kilger, this volume, ch. 5. 70 I wish to express my gratitude to Svein H. Gullbekk for pointing this out. 71 The example of Peter’s Pence being collected from Bergen diocese in the years 1294–1300 is telling: a sum of five marks six ores subtilis monete was part of this, Pavelige Nuntiers Regnskabs- og dagböger, 12–14; Gullbekk, Pengevesenets, 79–83. 72 Gullbekk and Sættem, no. 304. 73 NGL III, 1–12. 74 Jon Anders Risvaag, “Ikke-kongelig utmynting i Norge frem til reformasjonen,” Nordisk Numismatisk Årsskrift 1994–96 (2000): 143–44. The only written evidence of the revocation is Archbishop Jørund’s (1287–1309) complaint, dated 9 March 1291, concerning the unjust treatment of the Church by the regency government (DN III, no. 30). 75 DN I, no. 71. 76 Pål Berg Svenungsen, Norge og korstogene: en studie av forbindelsene mellom det norske riket og den europeiske korstogsbevegelsen, ca. 1050–1380 (Bergen: University of Bergen 2016), 236–37. 77 Jon Anders Risvaag, Mynt og by: myntens rolle i Trondheim by i perioden ca. 1000–1630, belyst gjennom myntfunn og utmynting (Trondheim: Norwegian University of Science and Technology, 2006), 308–11; Gullbekk, Pengevesenets, 162–85. 78 Ole Jørgen Benedictow, Plague in the Late Medieval Nordic Countries. Epidemiological Studies (Oslo: Middelalderforlaget, 1992), 94–97; Ole Jørgen Benedictow, The Black Death, 1346–1353: The Complete History (Woodbridge: Boydell Press, 2006), 157. 79 The last time Høre is mentioned as a parish is 15 May 1538 (DN XXI, no. 835). 80 Kjell T. Lund, Han far: prestehistorie og bygdeliv, Vang i Valdres (Vang: Valdres historielag, 1997). 81 Tveito, “Mynter i messen,” 413–14. The Scandinavian term is offertavle or kollekttavle. It needs to be noted that the use of offering plates in Scandinavia can be traced back to at least the thirteenth century. From Gotland, 55 medieval plates have survived, the oldest from the churches of Bro and Norrlanda, both dated c. 1220–1260, and Endre Church dated to the thirteenth century, Pia Bengtsson Melin and Kenneth Jonsson: “Medetida kollekttavlor från Gotland,” Myntstudier 2 (2019): 1–92. The spread of the practice in Norway during the Middle Ages is difficult to assess; however, the Reformation Bishop, Peder Palladius, writes polemically in favour of the offering plate in his Visitation book (written c. 1540 or 1541), comparing the lack of willingness of offering on the (offering) “plate of the poor” to the willingness of offering on the “monk’s altar,” Peder Palladius visitasbok, published by Kaare Støylen (Oslo: Johan Grundt Tanum, 1945). If this signals the abandonment of a predominant practice (offering on the altar) and enhancing another hitherto parallel (the offering plate), or exchanging one for another, is difficult to establish.

7

Eidskog church revisited Coin finds, architecture and the devotional use of money Håkon Roland

Eidskog church This article presents a re-examination of the remains of two stave churches found beneath the present-day Eidskog church and documented by excavations in 1965. A thorough re-examination of the stratigraphy based on the available documentation is presented. Furthermore, the coins, coin distribution and coin clusters are analysed in detail. The material is then presented and contextualized with the available written sources. The main questions follow two main directions. Firstly, can the excavation coins provide new knowledge on the construction and architecture of the earliest stave church in Eidskog? And, secondly, can an analysis of the coin material provide new insights into offering practices at the site? The investigation attempts to provide clues on the practice of coin offering at Eidskog by comparing the coin material with the well-attested pilgrimage activity at the site. Eidskog church in eastern Norway is a relatively small parish church in Skotterud, in the municipality of Eidskog in Hedmark County.1 Eidskog is the southernmost municipality in Hedmark County and borders the county of Värmland in Sweden to the east and southeast. Eidskog, or Eidaskog, is a newer name on the church site.2 The original name of the rectory, and most probably also the name occasionally used for the church, was Matrand (NG 224).3 The oldest name, perhaps, is Midtskogen (Mesku), the farm that Matrand was separated from in the twelfth or thirteenth century. The first account of a church in Eidskog is from Håkon Håkonsson’s saga (r. 1217–1263), where we are told that King Håkon passed by Eidskog church on his way to Värmland in 1225.4 The church holdings were listed as Matrand in 1394 (RB 461).5 Many farms were established in the Eidskog area in the eleventh to thirteenth centuries. The farm names tell us that 69 out of the 73 principal farms existed before the Black Death (i.e., 1349).6 There are also indications of increased activity in the area in the eleventh to thirteenth centuries. Recent archaeological investigations have shown extensive iron production in the south-eastern parts of Hedmark county. Iron production peaked in the period 1150–1250 and coincided with a significant increase in farms in the area. The volume of production was significantly above self-supply, and a significant part of it must have been transported to other areas continuously in the period c. 1100–1300. The extensive production was of great importance for communication and traffic throughout the area. Eidskog was located at an entrance to Hedmark County in the far south and must have been increasingly affected by the huge activity in the period after 1150.7

170  Håkon Roland The church was possibly under the bishopric of Hamar from its earliest days, but written sources from the middle of the fourteenth century firmly put the church under the control of the bishopric of Oslo. Eidskog church was centrally located on the main route from Nidaros, in the north, through Solør in Østerdalen, to the Swedish border in the south.8 The early medieval trade routes were based mostly on the sailing routes along the west coast. However, during the winter, heavy weather conditions often prevented sailing. The well-established western road through Gudbrandsdalen, which passed through the mountainous areas of Dovrefjell, was also closed during the winter, leaving the eastern route through Eidskog virtually the only option for transport and travel. Most travellers heading northwards from Sweden to Norway to reach Nidaros passed Eidskog on the way. The pilgrim trails to St Olav’s shrine in Nidaros that evolved during the eleventh and twelfth centuries followed the already existing road network.9 In short, Eidskog church was centrally located on an important road. It was perhaps the main road from Sweden to Norway in the wintertime, and a road that developed into an important pilgrim route in the twelfth and thirteenth centuries.10 Today it is c. 15 km south from Eidskog to the current Swedish border. The location of Eidskog church might well have influenced the composition of finds from the medieval church. In the following, a review of the previous excavations is followed by a re-examination of the archaeology and an analysis of the composition of the coin finds in light of the fact that the stave churches at Eidskog were pilgrim churches in close proximity to the Swedish border. Eidskog churches Three churches can be traced at the same location in Eidskog: two stave churches and one, the latest, a log church. In the following discussion, the oldest and smallest stave church is Eidskog I, the large stave church is Eidskog II and the present-day log church is Eidskog III. The excavations carried out by the architect Per Martin Tvengsberg in connection with the restoration of the current church in 1965 revealed the remains of two former stave churches at the site.11 The most comprehensive study of the stave churches and coin finds from Eidskog is still archaeologist Anne-Marie Mørch von der Fehr’s work from 1984.12 A church in Eidskog is first mentioned in 1225.13 This probably refers to the earliest stave church on the site, Eidskog I, which was identified by a rectangular area of approximately 5 × 7 m beneath the better-preserved remains of a larger stave church (Eidskog II), of which the full ground plan can be reconstructed. The first church on the site was built on stone foundations and, according to the state of research, was without a chancel. There are few clues to dating the earliest stave church, although dates as early as the end of the eleventh or early twelfth century have been suggested.14 Finds of two Viking-Age beads, one fibula with Urnes-style ornamentation, and a few diagnostic objects are held as proof of an early date.15 It has also been suggested that parts, or perhaps the whole aisle, of the first church were reused as the chancel in the new stave church (Eidskog II), since the width of the chancel in Eidskog II is identical to that of the aisle of the oldest church (Eidskog I).16 A detailed analysis of the coin finds makes it clear that both the chronology and reconstruction of the church buildings is in need of re-evaluation (see discussion below).

Eidskog church revisited  171

Figure 7.1 Eidskog church. Plan drawing from the 1965 excavation and suggested plans of the two stave churches, Eidskog I and II. Original plan drawing by Martin Tvengsberg, remapped by Steinar Kristensen, UiO: Museum of Cultural History

A large stave church replaced Eidskog I in the second part of the fourteenth century. Eidskog II is the stave church that is mentioned in written sources on several occasions during the fourteenth to seventeenth centuries. The church is addressed briefly for the first time in written accounts during the fourteenth century, most importantly in Bishop Eystein’s Register in 1394, when the bishop gave orders in a letter for the repair of the St. Olav’s cross, which had been located near the church for centuries, and for raising a cross-house (chapel) over it.17 He also gave orders for an altar to be built in the chapel to provide a place for the pilgrims for prayers and offerings. The income thereof was to be divided into two parts: one for the priest and church and one for the cost and maintenance of the cross and chapel.18 According to Tvengsberg’s reconstruction, Eidskog II was a fairly large stave church measuring 10.5 × 13.5 m. A gallery surrounding the church was a later addition. Two sleepers of the old stave church still remain in use as horizontal supports in the present-day tower. Even though they have been partly cut, their dimensions and shape indicate that they originally functioned as the southern and northern sleepers of the stave church.19

172  Håkon Roland Also, three wall planks from Eidskog II that had been reused as flooring for a period are now stored beneath the floor of the present tower. Vague traces of original paint are still visible. The dimensions and decoration show that the stave church had inner cladding in parts that were at least 3.5 m in height and painted and decorated with a red and white winding pattern. Hardly anything is known about the interior or decoration of the first stave church. A medieval baptismal font is today located in the chancel to the north. The font is dated c. 1175–1250 with a main production period in the first half of the thirteenth century.20 The font is not in its original position, and we do not know whether or not it belonged to the preceding stave church(es). One of two medieval church bells is still located in the church;21 it probably dates back to the fourteenth century, and thus belonged to Eidskog II. It was suggested that the odd crucifix currently mounted in the altarpiece may be of medieval origin, but this theory has now been rejected. 22 The present-day church at Eidskog is cruciform in shape and seats 350 people. It was built in 1665 and constructed by the architect Knut Mortensen. The church was modelled after Kvikne church in the municipality of Tynset, Hedmark County, which Mortensen had constructed a few years earlier, in 1652–54. The modern church is c. 19 m long with a width of c. 15 m. The nave, chancel and transepts are all c. 7.5 m wide. The church has a central, pointed tower over the crossing, and the timber walls are covered in white, painted cladding.

Figure 7.2 Present-day Eidskog church. The log church from 1665 seen towards the north-east. (Photo: H. Roland)

Eidskog church revisited  173 The largest stave church (Eidskog II) is mentioned for the last time in written sources in 1656, and the modern church (Eidskog III) for the first time in August 1665. However, the church office’s accounts make no mention of the removal of the stave church or the rebuilding of a new, cogged timber church.23 Several phases of repairs and rebuilding are, however, recorded for the new church. A new loft was built, a new roof covering laid and 19 windows replaced in 1694, and works on the foundations of the tower were carried out during the following year. The church and most of the surrounding properties were sold in 1723, and the church declined. Local farmers bought the church in 1758, and several periods of repair and rebuilding and alteration of the interior followed in the decades thereafter. The tar-burnt outer walls were covered with plank cladding in 1785, and heavy restorations and repairs followed in the period from 1850, when the church got new flooring, interior, seats and enlarged windows, and the walls and carrying constructions were heavily reinforced. This period of repairs and rebuilding might have had an impact on the situation in the ground level inside the church. Archaeological excavation In the following will be presented a new interpretation of the archaeological features and stratigraphy of Eidskog that partly diverges from the works of Tvengsberg and von der Fehr. The arguments are based on a comparative and contextual analysis of the coin finds and the archaeological layers. A comprehensive analysis in which the coin material is more closely integrated will provide new insights into the stave churches’ construction phases and architecture. The following accounts utilize the coordinates of Tvengsberg’s original plan as the points of reference for the discussion and analysis. Eidskog church was excavated between 11 January and 15 May 1965. The excavation and architectural survey of the present church were undertaken under the auspices of the Directorate for Cultural Heritage (RA) and conducted by the architect Per Martin Tvengsberg with local assistance from Kjell Magne Bjørnebye. Tvengsberg submitted a preliminary report on 29 May 1965, and the undated final report later the same year.24 The church is built on a moraine plain, consisting of rubble and gravel, which is flooded regularly by the nearby river Vragnselva. These floods wash much of the soil away from the moraine plain. The excavation reports hardly give any information on the methods used. The excavator was a trained architect, and the reports reveal no critical considerations regarding the choice of methodological approach. However, the excavation was well laid out, with 1 m intervals measured from 44 to 58x (S–N), and 37–57y (W–E) on the plan drawing. Heights are measured in relative 1 m intervals, +1 to +3 (upper). Based on the accuracy of measurement, one can create a Cartesian three-dimensional coordinate system for the objects with precise information in the archaeological record. However, not all objects are recorded with exact coordinates. Only one-fifth of the coins and other objects have accurate coordinates, the remaining have only a description of finds’ location related to the architectural remains (e.g., “found close to the inner wall of the SW corner in N transept”). The reports say nothing about sieving. However, the recorded coin finds and finds of many small objects such as needles and fragments of coins and so forth suggest that the excavation was conducted very carefully. It is difficult to see that the very small, heavily damaged fragments of bracteates would have been found at all without sieving. The exact recording of coordinates further tells us that the soil must have been processed inside the church,

174  Håkon Roland during excavation and not, as is often the case with Norwegian church excavations, after the soil had been moved outside the excavated area. 25 The excavation beneath the floor of the present church identified three or four (see below) distinct layers, post holes, graves and grave fill. The most distinct layer is a fire layer covering much of the excavated area with the exception of a 5 × 7 m rectangle.26 This layer consists of 1–3 mm of compressed ash, with an inclination of charcoal, and with a white and reddish layer of gravel beneath (due to heating). Several graves and other finds are undoubtedly located beneath the fire layer, and the amount of charcoal in the underlying layer point to a fire in a building preceding the one that caused the distinct fire layer. Tvengsberg concluded that at least one major building burnt and that older remains of activity existed, prior to the building that was destroyed.27 The 5 × 7 m rectangular area that breaks the fire layer has traces of human activity, most notably post

Figure 7.3 Martin Tvengsberg’s plan drawing from the excavation in 1965. X and y-coordinates are referred to throughout the text. Original plan drawing by Martin Tvengsberg, remapped by Steinar Kristensen, UiO: Museum of Cultural History

Eidskog church revisited  175

Figure 7.4  F  ragments of Norwegian coins of Magnus Lawmender (1263–1280). (Photo: H. Roland)

holes. A fire-fill layer covers most of the rectangular area (the fire fill must be interpreted in connection with the surrounding fire layer). Contours of a post setting as well as removal can be seen, and traces of coal and ashes point to a conclusion that this post was removed after a fire (different from and earlier than the one that caused the fire layer and thus attesting activity and possibly a significant fire prior to the construction of the first stave church). The pits inside the rectangle belong to a different situation and seem to have been difficult to interpret during the excavation. Tvengsberg probably mentioned the situation to his superior at the RA, since in April 1965 architect Håkon Christie asked the archaeologist Arne Emil Christensen and geologist Inge Bryhni specifically to examine the area in question. In his report, Christensen describes “older Christian graves” that cut through a fire layer beneath the largest stave church.28 These graves are, however, outside the rectangle in question.29 Some of the pits are older than the post holes; they are mostly c. 50 cm deep and circular in shape, although oval-shaped ones are also found. They have been dug in sterile gravel and filled with gravel, humus and charcoal. There are no traces of bones, so the possibility of earlier cremation graves is ruled out. Christensen hesitated to identify them as pre-medieval cooking pits since the characteristic reddish, burnt soil and stones with burn marks are missing. A row of larger stones indicates the presence of the wooden sleepers of the largest stave church (Eidskog II). Large lumps of solidified tar were found immediately outside the outer row of stones, most likely a reminiscence of spilt tar from the outer walls of the older church, although it could also be the result of a fire. Tvengsberg identified the outer gallery, reaching all around the stave church, and the inner main wall. The gallery is c. 1.5 m wide, with wooden floor. Later rebuilding and the construction of a grave tomb in the seventeenth century destroyed much of the feature, which has been interpreted as the chancel of Eidskog II (see below). Several graves were found, including children’s graves. However, none of the graves were examined in any detail.30

176  Håkon Roland Approximately 3,500 objects, of which 485 are coins and coin-like objects dating from c. 1208–1216 to 1905, were recovered during the excavation. Other finds include parts of books, fragments of metal objects in copper, silver and iron, bolts and different tools, lead bullets, amulets, a medieval shield, pencils, horseshoes, knives, combs, fragments of textiles and leather, an embroidered cap, keys, clay pipes, broken bottles, fragments of jewellery, small pins, scales, fragments of window glass etc.31 As mentioned, the area is regularly flooded in the spring, and this has created an O2-rich environment in the ground beneath the church that is unfavourable to the preservation of objects. Much of the coin material is in a very bad state of preservation, and it must be assumed that smaller objects and organic material have been lost over the years. One of the most notable finds is a medieval shield. It was found in the aisle with a coin beneath it. The shield has been dated to c. 1200–1300 typologically and for its late-Romanesque and early-Gothic ornamentation.32 The coin beneath it is significantly later (1450s). The shield and coin are most probably part of a secondary burial (after 1665) of items that had already been in the church for several centuries. The display of family shields in churches is a tradition attested from c. 1200. 33 The fifteenth-century coin seems to have been deliberately and accurately positioned beneath the con of the shield. Coin finds Bracteates of King Sverre (1177–1202) are absent in Eidskog even though they are well represented in other Norwegian stave churches. A general survey of coins from Norwegian churches gives the following numbers: Sverre, 1020 coins; Håkon Håkonsson, 3260; Magnus Lawmender, 2125; Eirik Magnusson/Duke Håkon Magnusson, 1409.34 The oldest coin in Eidskog is either a Swedish bracteate of Erik Knutsson (r. 1208–1216) or a Norwegian bracteate of Håkon Håkonsson (r. 1217–1263).35 It is tempting to interpret the absence of coins earlier than c. 1200 as an indication of the earliest date of Eidskog I. However, a survey of the church finds reveals a similar pattern in other churches in the eastern parts of Hedmark, Akershus and Østfold (parenthetical numbers refer to no. of coins dated before c. 1200): Rødenes church (0), Eidsberg church (0), Våler church (1), Hvaler church (3), Eidsvoll church (2), Nes church (0) and Eidskog church (0).36 The number of pre-1200 coins is much higher in the neighbouring counties of Oppland, Buskerud and Vestfold: Lom stave church (242), Ringebu stave church (33), Høre stave church (27), Torpo stave church (63), Uvdal stave church (35) and Sandar church (19).37 The significant differences in finds of pre-1200 coins in churches from the eastern part of Norway in comparison with churches on the western side of the Oslofjord and in the inland regions pose questions about ritual and devotional practices, the availability of coins and monetization, and church architecture. It seems clear that archaeologists have been able to harvest large numbers of twelfth-century coins from the stave churches inland. The availability of coins was hardly less in the more eastern and southern parts of Norway, rather the opposite in comparison with the regions in the valleys and mountains in the inland. Thus, the absence of coins prior to 1200 in this region must find another explanation. Similar observations have been made recently and discussed in further detail, and it has become clear that there is strong evidence for monetization developing at a different pace in the inland versus coastal regions.38 These patterns of coin finds provide significant empirical material for further discussion. In this, the finds from

Eidskog church revisited  177 Eidskog church provide important insights, as they do not fit the generally accepted notion of the Norwegian economy in the late twelfth and early thirteenth century. A particular feature at Eidskog is the high portion of foreign early medieval coins. Only one other Norwegian church matches the percentage of foreign coins: Stange church, also in Hedmark County, with 48 early medieval coins, of which 22 are foreign. The Swedish coins from Stange and Eidskog churches make up 45% of the total number of foreign coins older than 1319 in Norwegian churches.39 In addition, a number of bracteates from c. 1350 to 1400 are difficult to attribute to either Sweden or Norway. From Eidskog, there are 23 Swedish and 48 Norwegian coins older than 1319, and 23 Swedish and 4 Norwegian coins from 1319 to 1360, giving a total of 46 Swedish and 52 Norwegian coins older than 1360. The share of Swedish coins earlier than 1319 is especially noticeable. Swedish coins are mostly absent from church finds and hoards alike in Norway from this period.40 On the other hand, the percentage of Norwegian pre-1319 coins is not particularly high. After the 1360s, foreign coins provide an increasingly large share of the finds: 12 Swedish bracteates dating c. 1360–1520, and 24 bracteates from fifteenth-century Mecklenburg, 5 coins and 1 jeton from German states, and 23 uncertain fifteenthcentury bracteates. This is a trend that is observable in the finds from many Norwegian churches and a reflection of the increasing availability of foreign coins in circulation in the late medieval Norway. Re-examination of coin finds related to stratigraphy and church plans We shall now turn to a re-examination of the stratigraphy, coin finds and archaeological findings related to the stave churches, Eidskog I and II. Close examinations of the excavation reports and plan drawings reveal several uncertainties regarding the stave churches and the stratigraphy of the site. After looking closely at the stratigraphy, we will then analyse the distribution of coin finds within the present-day church. The results will then be used to examine the possible building phases and ground plans of the earlier stave churches. The documentation of the excavation has only partially survived. Tvengsberg’s original drawings of the church plan and profiles are detailed and the best source for the stratigraphy of the site (cf. Figure 7.1). Unfortunately, all photographs from the excavation have been lost,41 although there are a few, very bad photocopies in the archives of the Museum of Cultural History. About one-fifth of the coins are plotted within a 10 × 10 cm position. The rest are documented with various degrees of accuracy.42 Most can, however, be positioned within 1–2 × 1–2 m coordinates. This is based on the archival notes and listings made by von der Fehr when she prepared her major archaeological study of the church in 1980–1982.43 Four stratigraphic layers are defined, and coins with exact coordinates are analysed independently of the others. Coins from graves and grave fill are separated. Special attention is given to the earliest coins, coins with burn marks and clusters of coins that can be defined with certainty. The stratigraphy and coin distribution will be analysed before considerations on the various phases and plans of the stave churches are discussed. Only one distinct stratigraphic layer is mentioned in Tvengsberg’s excavation report: the “thick and distinct fire layer” covering the excavated area except within a clearly defined rectangle.44 The fire layer covers several graves.45 The 5 × 7 m area where the

178  Håkon Roland fire layer is absent, and which is filled with a fire-fill layer, corresponds by and large with stone rows that supposedly supported the earliest stave church (Eidskog I). One of the graves dated by Tvengsberg as “younger than 1250” is located to the west of this rectangle46 and is important to the chronology of the earliest church(es) on the site. Two Viking-age glass beads and one fibula with Urnes-style ornamentation, but no coins, were found in this layer. The fibula is dated to the twelfth century, and a similar type has been found in an early medieval grave in Leksand church in Sweden.47 Von der Fehr summarizes the identified layers of the Eidskog excavation as: (1) the fire layer; (2) a mixed layer of reddish woodchips, sand and gravel; and (3) a dust and woodchips-layer immediately beneath the floor of Eidskog III (i.e., after 1665).48 Von der Fehr’s simplified stratigraphic division was needed in order to study the find distribution of almost 500 coins and more than 3,000 additional objects from what was a complex excavation. The coin material was analysed through chronological groups and three distribution maps of each stratigraphic layer.49 According to Von der Fehr, all objects other than coins from the fire layer and fire-fill layer (with the exception of a fragment of a clay pipe and a button) belong to the period before c. 1355. The objects from the mixed layer all belong to after c. 1360. However, to a certain extent, this layer also contained later objects like clay pipes. The dust layer contained only objects from 1600 onwards, with the exception of a few stray coins. Von der Fehr makes the observation that the earliest and latest layers were mostly undisturbed, but disturbances were occasionally seen in the mixed layer. The previously published plans of Eidskog I and II are very approximate. In my analysis, I want to go a step further and use the coin finds as a means to achieve a higher resolution of the layers and their relationship to other archaeological features. This will contribute to a broader understanding of the architecture and the layout of the medieval church buildings at Eidskog. In the initial phase of the documentation work from the excavation, Von der Fehr progressed significantly in defining four or five main characteristics that separated layers of different phases. But, as mentioned above, she ended up with a simplified model with three layers in her analysis of the find distribution. Tvengsberg’s profile drawings show the situation north–south, cut through the nave and chancel of Eidskog I,50 and east–west, cut through the centre of Eidskog I.51 If we look strictly at the profile drawings, it becomes clear that Von der Fehr’s so-called “mixed layer” can be divided into two separate layers. Immediately above the fire layer is a layer consisting of gravel and red woodchips. Above is a layer of sand and woodchips with no red colour. Thus, in this analysis, a division in four layers can be introduced: (1) the fire layer; (2) red woodchips and gravel layer; (3) woodchips and sand layer and (4) dust and sand level. As mentioned above, the fire layer covers the whole excavated area except for a rectangular part with fire fill (layer 1) that is supposedly identical to the first stave church.52 However, it becomes clear that layer 1 is less rectangular than it would be if it corresponded to the aisle of Eidskog 1. At the profile cut at 48,7y at the western end, it matches the width of the aisle of Eidskog I, 53 but it narrows towards the east end and is narrowest at the profile cut at 49,5y. 54 The east–west profile cut at 51x clearly shows that the fire-fill layer ends at 50,2 y to the east. The red woodchips and sand layer (layer 2) follows an almost identical pattern. Towards the west, at the profile cut 48,7y, it matches the aisle of Eidskog I.55 But at the profile cut at 49, 5y, it narrows in the same manner as the preceding layer.56 Layer 2 extends further than layer 1 to the east and ends at 53,5y (it is limited by the grave crypt from Eidskog III).

Eidskog church revisited  179

Figure 7.5 The extension of Layer 1 towards the east. Stratigraphy and drawing by Martin Tvengsberg 1965. (Museum of Cultural History, UiO).

The woodchips and sand layer (layer 3) covers a smaller, more rectangular area.57 Like layer 2, the grave crypt of Eidskog III marks the extent of layer 3 to the east. The dust and sand layer (layer 4) covers the area beneath the floor of Eidskog III. The stratigraphy beneath the transepts of Eidskog III differs, as expected, from the stratigraphy within the area of the earlier stave churches. Layer 4 contains modern coins (i.e., after c. 1620), with the exception of one Magnus Eriksson coin58 from a child’s grave to the north59 and a Magnus Lawmender coin and unidentified bracteate (c. 1200–1400) from grave fill to the south.60 The combination of coin finds and the identification and extent of these four layers opens the way to a new interpretation of the different generations of stave churches. In the following, the distribution and chronology of coin finds and their relationship to the identified layers will be discussed. Coins from graves and grave fill will be added to the coins from the four layers. A general survey of coins attributed to layers shows the following pattern: Layer 1 has a majority of early coins, with only four extending into the period immediately following 1319. The coin distributions in layers 2 and 3 seem almost similar at a first glance. However, layer 2 contains 14 of the total of 15 early medieval Swedish coins. There is also a tendency that coins from 1220 to 1319 dominate in layer 2, and coins from the following period dominate in layer 3. When it comes to the latest coins, about half the numbers in parenthesis in Table 7.1 are coins from the transepts of Eidskog III (the post-1665 church). The coins from this part have been assigned to layer 2 since the soil contains reddish woodchips. However, Tvengsberg’s description clearly states that the woodchips there are mixed with brown soil and thus differ strongly from the composition of layer 2 as it is defined above. The other half of the coins come from parts where layer 2 and layer 4 are compressed and partly mixed.61 Only 7 of 178 coins from layer 4 are from the medieval period. Most of these medieval coins are found in secondary sand fill above graves at the west end of the aisle.62 The

180  Håkon Roland

Figure 7.6 Distribution of coins recorded with exact coordinates and chronologically divided into four groups: 1220–1319, 1320–1536, 1537–1620, 1621–1850. Plan drawing by Martin Tvengsberg. Original plan drawing by Martin Tvengsberg, remapped by Steinar Kristensen, UiO: Museum of Cultural History

Table 7.1 Chronological distribution of coins in identified stratigraphical layers

Layer 1 Layer 2 Layer 3 Layer 4 Graves/fill TOTAL

1220–1319

1319–1536

16 35 25 1 25 102

4 32 42 6 18 102

1537–1620

1620–1844

2 2 1

(29) (33) 170 7 239

5

Eidskog church revisited  181 coin distribution tells us that the layers are mostly intact, except in a few areas where they have been disturbed by loadbearing supports for the floor of Eidskog III, burials or other local disturbances. During the excavation in 1965, Tvengsberg noticed that a substantial part of the coins had burn marks. They were later examined, in 1983, by Per Mathiesen at the conservation department at the University Museum of Antiquities.63 Fourteen of the coins with burn marks came from layer 1, 19 from layer 2, 12 from layer 3 and 10 from a grave fill. All except four were found within the area of Eidskog II, as defined by Tvengsberg and von der Fehr. The majority of the earliest coins (before c. 1319) and a substantial part of the unidentified bracteates from 1200 to 1400 contain burn marks. A few later coins also bear burn marks: Swedish coins of Thordeman XVI, XVII, XX, XXII64 and Mecklenburg bracteates of type Oertzen 183.65 An early medieval bead and fibula (see above) were also burnt.66 It must be noted that far from all the medieval coins show burn marks. Among the coins from before c. 1360, a total of 55 have burn marks and 71 do not. The burned coins appear to be evenly distributed between the aisle and chancel of Eidskog II. The coins with burn marks probably belong to the building or part of the building that burned down. Thus, the burned material provides us with a terminus post quem after which the fire took place. It is reasonable to suggest that the church was destroyed around the middle of the thirteenth century. We find, then, a discrepancy between the coin finds in layer 1 (fire layer) and the coins with burn marks. As a consequence, the absence of coins from the fire-fill layer within the aisle must be considered. As we have seen above, recorded coordinates can position about one-fifth of the coins accurately. Plotting them on the excavation plan and to the stratigraphic layers reveals certain clusters that provide clues to the original extent of the stave churches. Coins numbered 714 and 71567 are from grave fill that must have been outside the western wall of the stave church in the mid-fourteenth century. This is significant since it makes it clear that the west wall at this time ended east of 43y. Several burials68 and the re-burial of a thirteenth-century shield with a c. 1450-coin beneath (no. 679) provide the southern limit of the fourteenth-century church and prove that the woodfloored gallery was added significantly later than the nave and chancel.69 The northern limit of the church in the same period is also attested by burials.70 The construction of a grave crypt in 1665 to the east makes it more difficult to estimate the eastern limits of the largest stave church (Eidskog II). It certainly continued east of 53y, but whether or not it extended beyond the grave crypt wall at 54,5 y is difficult to establish. The distribution of pre-reformation coins is limited to the area of the fourteenthcentury church, with a few exceptions (coin nos. 708, 714–15, 717–21, 722, 724). Coin no. 708 (Gotland, Visby, c. 1225–1288) belongs to a mixed layer and was found together with a Frederik III coin (no. 709). Nos. 714 and 715 (Swedish bracteate c. 1240–1245 and Magnus Eriksson c. 1319–1340) are from grave fill. No. 717 (Gotland, Visby, c. 1225–1288) was found close to a Christian V (r. 1670–1699)-coin (no. 622), near a supporting stone from the foundations of floor III related to the latest church from 1665. Coin no. 718 (Håkon Håkonsson r. 1217–1263) was found in a brown, mixed layer close to a grave. Coins nos. 719–22 (Albrecht of Mecklenburg c. 1370–1390, Magnus Eriksson c. 1319–1340, Hans r. 1483–1513 and Gotland, Visby c. 1225–1288) and no. 724 (bracteate c. 1200–1400) are also from a mixed, brown layer, probably a later fill. Tvengsberg’s suggestion that the rectangular fire-fill layer (layer 1) corresponded to the area of the first stave church at the site has since been repeated and accepted. As mentioned, he also claimed that the chancel of Eidskog II was damaged due to the

182  Håkon Roland

Figure 7.7 The extension of undisturbed stratigraphy towards east. Stratigraphy and drawing by Martin Tvengsberg 1965. (Museum of Cultural History, UiO).

construction of the grave crypt after 1665. However, a detailed analysis of the coin material proves otherwise. As we have seen above, layer 1 narrows towards the east and ends up with the same width as the chancel of Eidskog II. It also extends east of the aisle, and at least 0.5 m into the chancel.71 The area covered by layer 2 seems to have covered the complete area of the chancel and most of the aisle. The rectangular layer 3 covers almost exactly what appears to be the chancel. Layer 1 contained five coins later than 1319. Coin no. 144 (Frederik IV, r. 1699–1730) was found close to a stone that was used to support the church floor from 1665, coins nos. 145–47 (Hans and Christian IV r. 1588–1648) are listed without any coordinates or information about find spot. That leaves coin no. 207 (Magnus Eriksson c. 1319–1340) as the only coin later than 1319 that certainly belonged to layer 1. Some modern coins entered layers 2 and 3 at certain places where the stratigraphy is locally mixed and/or compressed (see above). It remains that pre-reformation coins make up the vast majority of coins in these layers. Plotting the coins with accurate coordinates reveals a very interesting pattern: they were almost exclusively located in the chancel.72 This begs a question about whether the situation in the medieval chancel was destroyed by the construction of the grave crypt from 1665 onwards. The construction of the grave crypt certainly affected the stratigraphy of the chancel. However, if we look closer to the profile cut at 51x, it becomes clear that the stratigraphy was unaffected significantly further to the east73 and that the fill from the construction work was used as backfill for the crypt’s wall at 54y. Only one or two coins were found in the area affected by the backfill. The early medieval coins in this part must have come from the

Eidskog church revisited  183 chancel, since the stratigraphy (layer 2 and 3) clearly shows that soil was moved from east to west.74 The majority of coins from the chancel and the chronological distribution of coins in this specific area clearly indicate the existence of a chancel in Eidskog I that was in use at least from c. 1220 and that coins were brought to the main altar for devotional purposes from the very beginning. Coins are found in all parts of the preserved part of the chancel of Eidskog II, except in the north-eastern corner. Clusters are found in the north-western corner, in the middle (two clusters), and at the south-eastern corner. There is no visible cluster of coins at the opening between the chancel and aisle. A small but interesting cluster of coins is found in the aisle.75 Coins nos. 664, 665, 666, 671 date from c. 1340 to 1400 and are recorded with exact coordinates. According to the fourteenth-century church plan, these coins were found in the southwestern corner of the aisle. This would be right inside the entrance whether it was on the western or southern wall at the time and thus where a truncus concavus or money box would have been positioned.76 Phases and chronology Remains of charcoal, pits and post holes clearly show that the Eidskog church is located on a site with traces of human activity extending back to the pre-Christian period. However, the post holes have proved difficult to interpret, and the excavators were not able to document any regularity or pattern that might bear witness to a post-hole church at the site. The amount of charcoal is, however, evidence of earlier fire(s) on the site. Viking-age beads, arrowheads and an ornamented fibula give some chronological indications of the pre-Christian activity but cannot be considered decisive evidence. Layer 1, covering most of the aisle of the earliest stave church, was most probably made to level out the area before the sill stones of the first church were laid out. The area of the chancel was not in need of levelling, since the older buildings or other remains of activity at the site had not reached this area prior to the construction of the first stave church. Based on this survey, I would argue that Eidskog I was planned and built with aisle and chancel but without gallery. The coin chronology points towards a date of c. 1200–1220 as the most probable starting point for the first church. Burials, coin distribution and stratigraphy show that the simple church remained unchanged, and without a gallery, until a large fire caused severe damage to the church in the mid-fourteenth century. The survival of a baptismal font and the fact that the majority of the medieval coins are without burn marks suggest that complete destruction was probably avoided. The situation after the big fire is uncertain, but we must assume that the rebuilding of a significantly larger church began shortly after the old one was damaged. When the church was rebuilt, it extended further to the west and the width was increased.77 An east–west78 and north–south79 row of stones as well as sill stones supporting the gallery80 are clearly visible on Tvengsberg’s plan. The 1.5-m-wide gallery clearly belonged to the rebuilt and enlarged church. The burials close to the walls of the earlier church, as well as the coin and shield burial in the early- or mid-fifteenth century, also bear witness to the chronology of the enlarged church and gallery. The most uncertain part of the enlarged church is the northern wall of the nave. This wall was left out of earlier reconstructions. The area northwest of the present-day church was less extensively excavated

184  Håkon Roland

Figure 7.8 Cluster of coins c. 1340–1400 at 49,6x and 44y. Plan drawing by Martin Tvengsberg. Original plan drawing by Martin Tvengsberg, remapped by Steinar Kristensen, UiO: Museum of Cultural History.

than other parts of the church. We know little of the situation in this part, where the western end of the north wall would have been. However, the profiles and plan clearly show a row of large stones81 laid out at the same distance from the foundations of the older church, as the very visible east–west wall  on  the south side.82 The large stones on the north side83 are at the same distance from the gallery walls as the thoroughly documented row of stones on the south side of the church.84 The chronology of the enlarged church remains uncertain. The question remains whether the population in the area would have demanded, only a few decades after the Black Death, that such a big church be rebuilt after a fire? Eidskog church is mentioned in written sources in 1367, 1369, 1394 and 1400.85 A large fire in the late fourteenth century is attested in local folk tales.86 In 1367, we are told that Vinger church authorized a change of land to the mensa and fabrica of

Eidskog church revisited  185 87

Eidskog church. Eidskog was apparently an annexe of Vinger church at the time, but the source also testifies that this was a temporary arrangement.88 Fabrica given directly to Eidskog church is attested by Bishop Hallvard in Oslo in 1369.89 Bishop Eystein visited Eidskog in 1394, when – as already mentioned – he ordered the restoration of the St. Olav’s cross at the site and the building of a small house for prayer nearby. The bishop was concerned about the state of the cross and the lack of facilities for visiting pilgrims, but the record does not mention any fire or problems with the church itself.90 If a devastating fire had recently taken place, one would expect this to be mentioned by the bishop during his stay. The fire must, then, have taken place several decades before the bishop’s visit in 1394. Contextualizing coin finds in Eidskog church In the following, we will look into the connection between pilgrimage, communication routes and coin finds in Eidskog church in an attempt to present a framework of interpretation new to this material. Eidskog is one of the easternmost stave churches in Norway and, as we have seen, was on the main road from Sweden to Norway during the winter season.91 That road was a well-established communication route from the eleventh century and a pilgrim route of importance from the twelfth century on.92 Värmland, the area on the Swedish side of the present-day border, has always had strong contacts with Norway, and the road was often used during conflicts between royals and among the shifting alliances of the eleventh to fourteenth centuries. The sagas contain several incidents where Norwegian kings and members of the elites passed Eidskog. King Sverre fled over Eidskog to Värmland in 1177–1178, and the usurpers against Håkon Håkonsson in the 1220s had their camps in Värmland. The Swedish king, Erik Eriksson, refused to take action against the usurpers on the Swedish side, prompting Håkon to gather a large army and march into Värmland in 1225. On the king’s command, Arnbjørn Jonsson, the local governor of Borgarsyssel (a municipality almost identical to present day Østfold), gathered fellow local governors from the eastern part of Norway at Eidskog early in 1225. The campaign was only partly successful, even though a small group of usurpers was defeated at the battle of Holmedal in 1225, and the situation in the area was not stabilized until the peace of Lödöse in 1249. The earliest mention of Eidskog church was written in connection with these events. In the contemporary saga of Håkon Håkonsson, the route passing is described as “easily passable during wintertime.”93 It was certainly less difficult to travel than the western route through Gudbrandsdalen, and even travellers from Oslo sometimes chose the eastern route through Eidskog.94 The detailed route passing through Eidskog, often called Eskoleia in medieval times, is still used today.95 The road through Eidskog created a strong connection between the areas on each side of the border. It was a connection that was strengthened by Eidskog’s position as an important site of pilgrimage and the St. Olav’s cult. Eidskog became connected to St. Olav when, according to Snorri (St. Olav’s saga), King Olav Haraldsson rested there on his way to Värmland in 1028.96 The legend tells that when the king rested at a field near the location of the later churches, a spring suddenly rose beside him. A cross in his honour was later raised and worshipped at the location, and when the first church was built at Eidskog, it was devoted to St. Olav. According to local tradition, parts of the road that passes by the church was called the “pilgrims’ road”97 and several sources confirm Eidskog’s importance to pilgrimage.

186  Håkon Roland There are, in general, few medieval sources that document the pilgrimage tradition. There are some mentions of pilgrims in the Saga literature, and pilgrim badges, holy water jars and relic pendants are among the archaeological sources. Small crosses were also often placed along pilgrim routes and also cut into rocks. Indeed, this tradition became so widespread that it was prohibited by Archbishop Eilif in Nidaros in 1327.98 Several large crosses were raised as well. We know of crosses in Brandbu, not far from Eidskog, and on Stiklestad and of the cross raised by Archbishop Jørund in Nidaros in the late thirteenth century.99 Indeed, the pilgrim’s route from south to north to Nidaros has been described as a “journey from St. Olav’s memorial to St. Olav’s memorial, from St. Olav’s church to St. Olav’s church, from St. Olav’s spring to St. Olav’s spring.”100 The large cross devoted to St. Olav in Eidskog is also well attested and served as one of those important markers on the pilgrim way to Nidaros.101 An interesting account from Bishop Eystein’s visitation in 1394 provides details on the Eidskog cross and the income from pilgrimage: During his visitation to Vinger on Wednesday after Sunday septuagesima, he took the testimonies of four men, most in their 70s or 80s. They gave a testimony that, when they were young, there was a big cross on the church common west of Eidskog church. Sacrifice was given for its maintenance, and pilgrims were allowed to hold Mass there. Their ancestors had told them that the cross was first raised in honour of St Olav, for he had rested there. Therefore, the bishop now allows the cross to be maintained, and that a small prayer house/chapel is being raised for St. Olav’s glory with an altar and doors without lock so that it is always open, and so that the pilgrims, with the bishop’s permission, can enter and hold Mass when they do not get into the church. Whatever offering locals give goes to the church itself, but offerings dropped in the chest at the cross or in the prayer chapel, shall be divided between the church, and the cross and the prayer house, and each shall receive a half for maintenance.102 It is clear that the bishop requested testimonials from the elders on the existence of a St. Olav’s cross at the site and the older traditions around it. This indirectly implies that the St. Olav’s cross was in a poor state or even destroyed at the time of the visitation. After hearing the testimonies, the bishop allowed the St. Olav cross to be re-erected or properly maintained and gave orders for the building of a small house for prayer “so the pilgrims can keep Mass when the church is closed.” The income from offerings at the cross and chapel was to be divided equally between the cross/chapel and the church. There is no additional evidence to link the poor state of the cross in 1394 with the destruction of the stave church in the big fire. However, one can speculate on the reasons why the cross was in such a bad shape at the time: If the church was destroyed in the mid-fourteenth century, it would naturally have affected the income from pilgrimage and worship at the cross in the decades that followed. Thus, the written account from 1394 would seem to indicate a mid-fourteenth-century date for the destruction of Eidskog I by fire. As we have seen above, and as is supported by numerous sources in general, pilgrimage, coin use and offerings are closely linked. The importance of the eastern pilgrim route is, for example, indirectly confirmed by Papal accounts for quadrennial tithe in Nidaros. The archbishop and cathedral committed 973 Mark in 1358103; almost one-third was paid in Swedish money.104 A monetary union between Sweden and Norway probably existed at the time (after 1319), and Swedish coins circulated widely in Norway in the mid-fourteenth century.105 There is, however, reason to believe that a

Eidskog church revisited  187 substantial part of the Swedish coins that accumulated at the end point of the pilgrimage to St. Olav’s shrine in Nidaros came from pilgrims travelling along the eastern route through Eidskog. We have already seen that early medieval Swedish coins (i.e., before 1319) are numerous in Eidskog church. A survey of church finds in the eastern part of Akershus, Hedmark and Østfold confirms that the situation in Eidskog is special.106 The coin distribution in east Norwegian churches along with Eidskog’s geographical location on a major communication and pilgrimage route make it safe to interpret the early Swedish coins in Eidskog to be mainly the result of pilgrimage and the money offerings made by people travelling from Värmland (and even further south) to Nidaros. Eidskog was on a major communication route from Sweden, and possibly further south, to Nidaros, a pilgrimage site of importance in Northern Europe. The existence of the cross and a holy well and the need for building a prayer house with permanent access bear witness to the importance of Eidskog to pilgrims. While the church and Saint Olav’s cross were, of course, important, the importance of the holy well should not be underestimated. Health and healing were an important part of pilgrimage, and localities such as holy wells associated with healing were clearly an important part of the spiritual magnetism that drew pilgrims to a site. In this perspective, the holy well was a central part of what constituted a therapeutic landscape and the production of place.107 The holy well at Eidskog formed part of the pilgrimage site and thus became a natural liminal space framed by narratives and texts connecting it to the power of the church and authorities in general.108 It is generally accepted that ritual acts of offering involve not only human–divine relations, but also a complex set of social processes by which a community of participants, like pilgrims, understands itself.109 The church, holy well and cross all made up a space where the movement of people/travellers and objects became essential parts of a complex network of people–object relationships. Different identities and cultural practices came together due to the pull of miraculous cures and sacred geography. Acts of devotion as well as performances of healing took place, and leaving objects, or offerings, was one of the most important actions performed at the Eidskog cross, as is attested in literary and historical sources as well as by the many coin finds from the church. A further question is whether Eidskog, with its sacred cross, holy well and pilgrimage facilities and unusual portion of Swedish coins, was a destination in its own right for Swedish pilgrims. Was all the traffic to Eidskog heading to Olav’s shrine in Nidaros or was Eidskog, in fact, the site of a flourishing local cult? The question is worth asking, even though the sources do not allow anything other than speculation. Pilgrims were obviously motivated by complex and many-faceted factors, and the popularity of pilgrim sites could change over time. The Norwegian material attests to the fact that healing was the single most important motivation for pilgrimage until the thirteenth century, when it was gradually supplanted by other motivations. But the final destination of the pilgrim also played a part: longer journeys to sacred shrines in Rome, Aachen, Santiago de Compostela, Canterbury and Nidaros were driven by one set of motives, while shorter, local pilgrimages had other aims. In the Norwegian material, it is clear that the healing aspect continued as the main motivation for shorter journeys ravels also in the late medieval period and even after the Reformation.110

Conclusion Three successive churches can be traced at the same spot in Eidskog. Two of them stave churches (Eidskog I and II) and the latest one a log church from c. 1665 (Eidskog III). This investigation of the coin finds and archaeology of the two stave churches has

188  Håkon Roland demonstrated that offerings and liturgy were performed in a small church from the beginning of the thirteenth century. The existence of a chancel, and thus the location of the earliest main altar of the church, is based on evidence provided by analysis of the archaeological record, stratigraphy and the concentration and chronological distribution of coin finds. The detailed analysis of coin finds and their chronological and spatial distribution in relation to the archaeological context has given us clear evidence that the church Eidskog I was planned and built with aisle and chancel but without a gallery. The simple church remained unchanged until the church was severely damaged by fire in the middle of the fourteenth century. Complete destruction was probably avoided since a baptismal font survived and the majority of medieval coins found are without burn marks. The church was significantly enlarged after the fire. Sill stones supporting the wooden frame for the aisle, chancel and gallery have been identified. The extent to the east is uncertain due to the construction of a grave crypt in 1665. The written sources combined with the coin material and stratigraphy of the site suggests that Eidskog I was partly destroyed early in the 1360s; it was rebuilt (and annexed) in the mid 1360s, and the large stave church, Eidskog II, was in function before 1369. The contextual analysis proves beyond doubt that coins were used extensively in the chancel in the medieval period. This result corresponds with recent findings in other Norwegian medieval churches (e.g., Høre church in this volume).111 The coin use and laymen’s access to the chancel, whether for liturgical or profane reasons, in the pre-Reformation period deserves a closer examination. The coins found at Eidskog church would have been offered there for different reasons, some by people celebrating the institutionalized and important holy sacrifice for the Mass of souls, some by individuals offering an intimate votum seeking redemption and miraculous healing. The coin finds from Eidskog church, situated on an important pilgrimage route and known as a place of sacred springs and healing, articulates these different motivations and mentalities more clearly than most medieval churches. Eidskog church exerted a strong spiritualized magnetism that caused people to give offerings for intimate and individual reasons. The composition of the coin finds makes it reasonable to ask whether the church was more than just one of several stops for pilgrims on their way to Nidaros. Eidskog was perhaps the final destination for Swedish pilgrims and the natural place to worship St. Olav for people travelling the important communication routes from the eastern part of medieval Norway and south-eastwards to Sweden. In this respect, the coins from Eidskog church contribute to our understanding of the medieval mentality behind offerings as well as to the architecture and construction of one of the early stave churches in Norway.

Acknowledgements The archaeologist Anne-Marie Von der Fehr examined the stratigraphy and coin finds in the early 1980s. She has generously made all her documentation and preparatory notes on Eidskog available to this study. Without her cooperation and preliminary studies, this article would not have been possible. The author wishes to extend profound thanks for this generosity. Thanks also to the referees, to Steinar Kristensen for GIS-plans, and to my colleagues Svein Harald Gullbekk, Alf Tore Hommedal and Christoph Kilger for reading early drafts and giving valuable advice. And many thanks to Kay Celtel for significantly improving the language.

Eidskog church revisited  189 Survey of coin finds from Eidskog church Norway Håkon Håkonsson 1217–1263 c. 1263–1280 Eirik Magnusson 1280–1299 Duke Håkon Magnusson 1284–1299 Håkon V 1299–1319

Eirik or Håkon Magnusson 1280–1319 Magnus Eriksson 1319–1355 Archbishops: Jon Raude 1268–1282 Gaute Ivarsson 1474–1510 Erik Valkendorf 1510–1522

1 bracteate 28 bracteates 1 penny 3 quarter pennies 7 quarter pennies 1 penny 1 halfpenny 1 quarter penny 2 half pennies 3 quarter pennies 5 two-sided coins 7 bracteates 1 bracteate 1 bracteate 2 bracteates

Denmark-Norway Eric of Pomerania 1396–1439/41 Christian I 1449–1481 Hans 1481/83–1513 Christian or Hans Christian II 1513–1523 Frederik II 1559–1588 Christian IV 1588–1648

Frederik III 1648–1670 Christian V 1670–1699

Frederik IV 1699–1730

Christian VI 1730–1746 Frederik V 1746–1766 Christian VII 1766–1808 Frederik VI 1808–1814/39

Unknown

3 sterlings (Danish) 2 bracteates (Danish) 5 hvid (Danish) 5 hvid (Danish) 3 hvid (Danish) 5 klipping (Danish) 2 klipping (2 sk, Danish) 12 two-skilling (2 Norwegian) 12 skilling (Danish) 3 søsling (Danish) 41 two-skilling (16 Norwegian) 3 skilling (Norwegian) 1 eighth-skilling (Danish) 45 two-skilling (15 Norwegian) 2 skilling (Norwegian) 2 twelve-skilling (1 Norwegian) 5 eighth-skilling (3 Norwegian) 1 four-skilling (Danish) 43 two-skilling (33 Norwegian) 1 eighth-skilling (Norwegian) 4 skilling (Danish) 5 two-skilling (3 Norwegian) 15 skilling (6 Norwegian) 8 two-skilling (4 Norwegian) 13 skilling (1 Norwegian) 1 four-skilling (Norwegian) 4 two-skilling (Norwegian) 1 skilling 3 fragments

190  Håkon Roland Carl XIII 1814–1818 Carl Johan 1818–1844

Oscar II 1872–1905

2 skilling 10 skilling 3 half-skilling 1 three-skilling/10 øre

Sweden Erik Knutsson, Lödöse 1208–1216 Erik Eriksson 1222–1250 c. 1225–1288, Gotland, Visby c. 1240–1245, Götaland Valdemar, Lödöse and Götaland, 1250–1265 Magnus Ladulås, 1275–1290 Magnus Eriksson 1319–1364

Albrecht of Mecklenburg or later Eric of Pomerania, Västerås, 1430–1480 Karl Knutsson Bonde, Stockholm, 1450–1470 Sten Sture d. E. Gustav Vasa 1523–1560 Johan III 1568–1592

1 bracteate 1 bracteate; 1 penny 9 two-sided coins 1 bracteate 1 bracteate 8 bracteates 1 bracteate 11 two-sided coins (1319–1340 (8); 1340– 1354 (2); 1319–1363 (1)) 5 bracteates (Kalmar, 1354–1363) 7 bracteates (Stockholm, 1363–1370 (1); Västerås, 1363–1380 (1); Stockholm, 1370–1390 (2); Västerås, 1363–1520 (3)) 3 bracteates 1 bracteate 1 bracteate (Stockholm, 1470–1500) 1 half-örtug (Västerås, 1480–1503) 1 fyrk (Stockholm, 1530) 1 half-öre (Stockholm, 1592)

Germany Güstrow, before 1379 Wismar, 1594 Stralsund, sixteenth century Lübeck, fifteenth century Eisleben, second half of fifteenth century Mecklenburg-Schwerin. Friedrich 1756–1785 Mecklenburg, fifteenth century Preussen. Friedrich Wilhelm I 1713–1740 German Order, Livland. Reval, c. 1400

1 quarter-witten 1 sechsling 1 schilling 1 scherf 1 bracteate 1 sechspfenning 24 bracteates 1 schilling 1 bracteate

Uncertain country Fourteenth to sixteenth century Probably medieval

3 two-sided coins 23 bracteates 23 two-sided coins or bracteates

Jetons: Germany, sixteenth century

1 jeton

Eidskog church revisited

191

Notes 1 Gnr. 31/bnr. 1; Askeladden ID, 84073. 2 Norse eið literally means “road” or “path” but is used for roads crossing rivers or alongside falls, for the road itself or a ford. Vrangselva, which flows near Eidskog, had many fords, and the “Eidskog” name means that in summer you could get from there via the waterway to large parts of the stretch through the valley. Norsk stadnamnleksikon, “Eid”. http://www.norskstadnamnleksikon.no/grunnord.aspx?grunnordCode=eid (accessed 26 November 2020). 3 Oluf Rygh, Norske Gaardnavne, vol. 3, Hedemarkens Amt (Kristiania (Oslo), 1900), p. 224, no. 31. 4 Jorunn Engen, “Eidskog kjerke. Prestegarden,” in Jubileumsskrift, 1990. Eidskog kirkes 325-årsjubileum, edited by Arne Optjernsberget (Eidskog: Eidskog menighetsråd, 1990), 8; further ref. Lorenz Dietrichson. De norske stavkirker: Studier over deres system, oprindelse og historiske udvikling: et bidrag til Norges middelalderske bygningskunsts historie (Kristiania: A.B. Calmeyers Forlag, 1892), 246–47. 5 Kulturminnesok.no, “Eidskog kirkested” with refs, https://kulturminnesok.no/ minne/?queryString=https%3A%2F%2Fdata.kulturminne.no%2Faskeladden%2 Flokalitet%2F84073 (accessed 11 November 2020). 6 Engen, “Eidskog kjerke,” 12. 7 Bernt Rundberget, Østnorsk jernutvinning i sen vikingtid og middelalder – særegen metode og kontrollert overskudd (København 2012), fig. 1; Kathrine Stene, Gråfjellprosjektet, bind 4. I randen av Taigaen – bosetning og ressursutnyttelse i jernalder og middelalder i Østerdalen (Oslo: Portal, Kulturhistorisk museum), 54, fig. 4.2.9; 155–56. 8 Sverre Steen, Ferd og Fest. Reiseliv i norsk sagatid og middelalder (Oslo: Aschehoug 1942), 266–67. 9 Inger Smestad, “Middelalderens veier og pilegrimenes vandringer,” in Natur, kultur og tro i middelalderen. En artikkelsamling, edited by Aud Beverfjord (Oslo: Riksantikvaren. Direktoratet for naturforvaltning, 1996), 23–34. 10 Kulturminneplan. Kommunedelplan, Kulturminner og kulturmiljøer 2018–2025. Eidskog kommune, 51–52. https://www.eidskog.kommune.no/_f/p1/i6aa59d19-8d0e-4a74914b-4f1f8ba04788/forslag-kulturminneplan-for-eidskog-til-horing-66883.pdf (accessed 27 November 2020). 11 Per Martin Tvengsberg, Eidskog kirke. Foreløpig orientering om utgravning under gulvet og bygningsarkeologiske undersøkelser. Archaeological Report, unpublished (Oslo: Directorate for Cultural Heritage, 1965a); Per Martin Tvengsberg, Eidskog kirke. Archaeological Report, unpublished (Oslo: Directorate of Cultural Heritage, 1965b). 12 Anne-Marie von der Fehr, «Gjenstandsfunn under kirkegulv. En metode for undersøkelser belyst ved materialet fra Eidskog kirke,» Universitetets Oldsaksamling Årbok 1982/1983 (1984): 199–211. 13 Kulturminnesok.no, “Eidskog kirkested” with refs. 14 Engen, “Eidskog kjerke,” 8. 15 von der Fehr, «Gjenstandsfunn under kirkegulv,» 205–206. 16 Tvengsberg, Foreløpig orientering, 4. 17 DN I 396. 18 Martin Blindheim, “The Cult of Medieval Wooden Sculptures in Post-Reformation Norway,” in Images of Cult and Devotion: Function and Reception of Images in Medieval and Post-medieval Europe, edited by Søren Kaspersen and Ulla Hastrup (Copenhagen: Museum Tusculanum Press, 2004), 47–59; Engen, “Eidskog kjerke,” 11. 19 Tvengsberg, Eidskog kirke, 3. 20 Mona Bramer Solhaug, Middelalderens døpefonter i Norge. Acta Humaniora 89, vol. II. (Phd Dissertation, University of Oslo, 2001), no. 28 with references. 21 Engen, “Eidskog kjerke,” 11. 22 Oddbjørn Sørmoen, “En 1600-talls kirke i 1990,” in Jubileumsskrift, 1990. Eidskog kirkes 325-årsjubileum, edited by Arne Optjernsberget (Eidskog: Eidskog menighetsråd, 1990a), 27–31, att. 30. 23 Tvengsberg, Foreløpig orientering, 4.

192 Håkon Roland 24 Tvengsberg, Eidskog kirke. Foreløpig orientering and Eidskog kirke. 25 A similar method of detecting and documenting small finds was probably practiced in the excavations of Bunge church, Gotland, see Kilger, this volume, ch. 5. 26 48,6–53,5x and 43,6–51y. 27 Tvengsberg, Eidskog kirke, 1. 28 Arne Emil Christensen, Innberetning om befaring til Eidskog kirke, Hedmark, 24/4– 1965. Archival report (Oslo: Directorate of Cultural Heritage, 1965). 29 cf. Tvengsberg, Eidskog kirke, 2. 30 von der Fehr, «Gjenstandsfunn under kirkegulv,» 200. 31 von der Fehr, «Gjenstandsfunn under kirkegulv,» 206. 32 Petter Molaug, “Middelalderskjold,” in Fortiden forteller. Universitetets Oldsaksamling 1829–1979, edited by Sverre Marstrander and Thorlief Sjøvold (Oslo: Dreyer forlag, 1979), 14. 33 Sigurd Grieg, “Norske gravkårder og litt om eldre militær gravskikk,” Vaabenhistoriske Aarbøger VII (Copenhagen, 1954), 77–106. 34 Svein H. Gullbekk, Pengevesenets fremvekst og fall i Norge i middelalderen (Copenhagen: Museum Tusculanum Press, 2009), 87. 35 Kolbjørn Skaare, «Norsk utmyntning på Håkon Håkonssons tid,» NNÅ 1970 (1971): nos. 18–20. 36 Svein H. Gullbekk and Anette Sættem, Norske myntfunn 1050–1319. Penger, kommunikasjon og fromhetskultur (Oslo: Dreyer forlag, 2019), nos. 3, 4, 5, 7, 25, 26, 28 and 73. 37 Gullbekk and Sættem, Norske myntfunn, nos. 90, 93, 95, 126, 130 and 151. 38 Gullbekk and Sættem, Norske myntfunn, 136–61. 39 Gullbekk, Pengevesenets fremvekst, 54–55. 40 Kolbjørn Skaare, «Norske funn av svenske mynter preget før 1319,» Hikuin II (1985): 237–48. Gullbekk, Pengevesenets fremvekst, 163. 41 Directorate for Cultural Heritage, pers. comm. 42 For a general discussion on accuracy of plotting coin-finds in churches, see Kristensen, this volume, ch. 4. 43 Von der Fehr, «Gjenstandsfunn under kirkegulv.» 44 Tvengsberg, Eidskog kirke, 1. 43,6y/51y, and 53,5x/48,6x. 45 At 50,4x/43y and 47x/49y. 46 51x/43y. 47 Inga Serning, Fyndgravar och gravfynd. Tusen år på kyrkudden. Leksands kyrka, arkeologi och bygnadshistoria (Falun: Dalarnas fornminnes och hembygdsförbunds skrifter 25, 1982), 116. 48 von der Fehr, «Gjenstandsfunn under kirkegulv,» 200. 49 von der Fehr, «Gjenstandsfunn under kirkegulv,» 201–202, 204. 50 48,7y and 49,5y. 51 51x. 52 Tvengsberg, Eidskog kirke; von der Fehr, «Gjenstandsfunn under kirkegulv,» 199. 53 48,2x–53,5x at 48,7y. 54 49,8x–52,8x. 55 49,8x–52,8x. 56 From 49,8x–52,8x. 57 From 49,8x–52,8x. 58 r. 1319–1355/63, coin no. 458. 59 55,2x/49,6y. 60 53x/49,5y. 61 c.50x–52x, and 49y–53y. 62 c. 49–50x and 41,8y, and c. 51–52x and 42,5y. 63 Now part of the Museum of Cultural History, University of Oslo. 64 Bengt Thordeman, “Sveriges medeltidsmynt,” in Mønt, edited by Svend Aakjær Nordisk Kultur XIX (1936): 1–92. 65 Otto Oertzen, Die mecklenburgischen Münzen des Großherzoglichen Münzkabinetts: Die Brakteaten und Denare, Vol. 1 (Bärensprung, 1900). 66 von der Fehr, «Gjenstandsfunn under kirkegulv,» 203.

Eidskog church revisited 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102

193

c.51x/42,5y. At c.44x–49x and c.48y–50,5y. Cf. plan at 55x–56x and 51y–52y; Tvengsberg, Eidskog kirke, 2. c. 54x and c.48,5y–51y. At 50,2y. 49x–53,2x and 51y–53,5y. Until 53,2y. From east of 54y and filled in between 53,2y and 54y. At 49,6x–44y. A similar interpretation has been put forward by Risvaag for Høre church, this volume, ch. 6. A cluster of coins suggests the placement of an alms box in the south-western corner of the nave beside the southern entrance. To 47y–55y. At 47x. At 40y. East–west at 45,5x and 56,5 x. At 55 x and 51y, 47y and 44 y. Running east–west at 47x. At 45,5x. At 56,5x. DN XXI, no. 124 (1367) and 130 (1369); DN I, no. 545 (1394); DN V, 505 (1414). Oddbjørn Sørmoen, “Mer enn syv hundre års kirkehistorie under same gulv,” in Jubileumsskrift, 1990. Eidskog kirkes 325-årsjubileum, edited by Arne Optjernsberget (Eidskog: Eidskog menighetsråd, 1990b), 32. Fabrica (lat.), church building, i.e., the funds necessary for maintenance of the church building. Mensa (lat.), altar, i.e., funding for the priest. DN XXI, no. 124. DN XXI, no. 130. DN I, no. 545. Arvid Ernvik, Olof den helige och Eskoleia (Karlstad: Skrifter utgivna av Värmlands museum 2, 1955), 18. Ragnhild Bjelland, Bot og Bedring. Den norske pilegrimstradisjonen i middelalderen. Cand. Philol. Thesis, unpubl. (Oslo: University of Oslo, 2000), 120–21. Håkon Håkonssons saga, trans. Anne Holtsmark (Oslo: Aschehoug, 1964), ch. 112. Steen, Ferd og Fest, 246. Magne Skrede, “Gamle veier i Glåmdalen,” Norske Bygder, Glåmdalen. Norske Bygder, vol. V (Bergen: Johan Griegs Forlag, 1943), 113. Olaf the Saint’s saga, Heimskringla, edited by C.R. Unger (Christiania 1868), trans. Anne Holtsmark and Didrik Arup Seip (Oslo: Gyldendal, 1970), ch. 91. Skrede, “Gamle veier,” 113. Grethe Authen Blom, «Nidaros – vandringens mål in Natur, kultur og tro i middelalderen. En artikkelsamling», edited by Aud Beverfjord (Oslo: Riksantikvaren, Direktoratet for naturforvaltning, 1996), 72–79. Eivind Luthen, “Til Nidaros – på pilegrimsveien fra sør,” in Pilegrimen. Valfartsmotiv og valfatsmål, edited by Gustav Erik Gullikstad Karlsaunet (Trondheim: Tapir forlag, 1996), 107. Luthen, “Til Nidaros,” 109. DN I, no. 545. “Fundas for korset og bønnehuset for pilegrimer ved Eidskog kirke utst av biskop Øystein i Oslo: Under sin visitas på Vinger onsdag etter søndag septuagesima tok han vitnesbyrd av 4 menn, de fleste i 70–80 årene. De avgav beediget vitneprov om at da de var unge, stod det et stort kors på kirkeallmenningen vest for Eidskog kirke. Det ble gitt offer til dets vedlikehold, og pilegrimer lot si messe der. Deres forfedre hadde berettet at korset først var satt til ære for St Olav, for han hadde rastet der. Derfor tillater nå biskopen at korset holdes ved like, og at det reises et lite bønnehus til St Olavs ære med alter og dører uten lås slik at det stadig står åpent, og slik at pilegrimene der etter biskopens tillatelse kan gå inn og holde messe når de ikke kommer inn i kirken. Hele det offer sognemennene

194

103 104 105 106 107 108 109 110 111

Håkon Roland gir, tilfaller kirken selv, men av det offer som faller i stokken på korset eller i bønnehuset, skal kirkens bygning og korset og bønnehusets vedlikehold ha hver sin halvpart.” DN I, no. 545; NGL 2rk I no. 158. English transl. by the author. DN VIII, no. 173. Audun Dybdahl, “Nidaros erkesetes økonomi,” in Ecclesia Nidrosiensis 1153–1537. Søkelys på Nidaroskirkens og Nidarosprovinsens historie, edited by Steinar Imsen (Trondheim: Tapir, 2003), 303. Gullbekk, Pengevesenets fremvekst, 163–67. Gullbekk, Pengevesenets fremvekst, 54. Ronan Foley, “Small Health Pilgrimages: Place and Practice at the Holy Well,” Culture and Religion 14, 1 (2013): 44–62 with refs. Foley, “Small Health,” 49. Catherine Bell, Ritual. Perspectives and Dimensions (Oxford: Oxford University Press, 2009), 109. Bjelland, Bot og Bedring, 206 and 319–20. Risvaag, this volume, ch. 6.

8 Coins and monastic liturgy in the Middle Ages A study of St Mary’s Benedictine nunnery in Bergen, Norway Alf Tore Hommedal Introduction The interpretation of coins excavated from medieval monasteries in general, and monastic churches specifically, has rarely been a subject of study in Norwegian church archaeology. More work has been done examining the dynamics of lay donations of coinage made by people attending services in medieval parish churches; excavations in Norwegian medieval churches generally reveal an abundance of coins. However, we know of relatively few coins from Norwegian monastic sites. This may, on the one hand, reflect the nature of sheltered monastic life, which was radically different from those of lay congregations and the parish churches that participated in daily interactions with the wider world. On the other hand, we can also hypothesise that the dearth of evidence of any possible “sacred economy” within Norway’s monasteries partly reflects the fact that the monastic sites in Norway were mostly excavated in the nineteenth and early twentieth centuries, when there was a bias towards recovering the building structures themselves rather than the finds of portable artefacts such as coins and other paraphernalia. The archaeological coin material excavated at St Mary’s Abbey church in Bergen in 1891 is an exception in this respect.1 Around 60 coins were found, the majority of which were minted in the centuries when St Mary’s was a Benedictine nunnery. This, in combination with the fact that more than three-quarters of the coins were found within the area of the nuns’ choir, begs two questions: (1) are the coins in the nuns’ choir related to the behaviour of the nuns? And (2) if they are indeed related, what use for coins did contemplative, Benedictine nuns have when seated in their prayer stalls? This chapter will present and discuss the coin material from St Mary’s and its interpretation in a liturgical context. To be able to throw light on these questions, the fabric and function of St Mary’s in different periods must also be clarified. St Mary’s Abbey was in Old Norse often called Nunnusetr, meaning “where the nuns live.”2 In the following, Nunnusetr will be used most frequently. The archaeological coin finds from St Mary’s and their contextual interpretation open up new perspectives on life both inside and outside the monastic walls in a ­Norwegian town.

Figure 8.1 Layout of St Mary’s Abbey church (Nunnusetr). The darkest lines indicate the extant lower part of the tower to the west and the vestry to the east (see Figures 8.5 and 8.6). The fundaments of the rest of the church are based on documentation by the architects Peter Blix and Schach Bull (Figure 8.2). Drawing by Per Bækken and Alf Tore Hommedal, remapped by Steinar Kristensen, UiO: Museum of Cultural History.

Figure 8.2 Building fragments and graves documented by Schach Bull in 1891. The eastern part of the chancel (presbytery) had been excavated by Peter Andreas Blix in 1872. After Bendixen 1893.

Coins and monastic liturgy – middle ages  197

Figure 8.3 Bergen around 1300 with its monasteries. The main settled area of the town is marked as darker grey. The Augustinian abbey of St. John’s was founded in the mid-twelfth century. As Munkalif and Nunnusetr, St. John’s was located outside the populated area. The Mendicant friaries, the Dominicans and the Franciscans, were both located inside the populated areas at the time of their foundations in the middle of the thirteenth century. Drawing: P. Bækken and A. T. Hommedal.

The historical context of the coin finds. Problems and main results Three main questions or themes related to the period of the Benedictine nuns are discussed in the following: The first is the Benedictine nunnery’s period of existence. The Bergen Nunnusetr had a rather complicated functional history with at least two, and perhaps up to four, different monastic orders occupying the site on a consecutive basis within the four centuries St Mary’s existed as a monastic site (c. 1120–1528). To interpret the coins found there, it has therefore been essential to establish a historical framework for the different monastic orders in residence, with clear chronological phases that make it possible to correlate specific assemblages of coins that are coeval with these phases and may enable us to distinguish sub-phases. It will be argued that the original Benedictine nunnery existed from the 1120s to about 1455. This was followed by a Bridgettine period until 1507, possibly divided by a short Cistercian period.

198  Alf Tore Hommedal The Bridgettine period was followed by a convent period for canons of the order of St Anthony’s (1507–1528). Thus, for the longest part of its history, c. 335 years, Nunnusetr housed Benedictine nuns; this was followed by a shorter period of c. 73 years of more varied monastic life. After Nunnusetr’s secularisation in 1528, the former monastery was rebuilt as a manor house called Lungegården. This period lasted until 1891, when the wooden manor built on the site in 1725 was destroyed by fire. This enabled the archaeological excavation of the church/manor house area, which yielded the coin material discussed in this chapter. The second main question to address is how the coins from the Benedictine period related to the nuns’ life in the church. Interpretation of the coin material in relation to the nuns’ activities, both devotional and leisure, is complicated by other possible routine activities at the site, such as the work of craftsmen building or repairing the church. With one possible exception,3 none of the coins in St Mary’s are older than the time of King Håkon IV Håkonsson (reigned 1217–1263), which means that there are no coin finds from the first 100 years of Nunnusetr’s existence. The presence of coins from St Mary’s thus seems not to have started before some decades into the thirteenth century. The majority of the 29 reliable coins found from the thirteenth century relate to an eastern extension of the church: the nuns’ new choir. Earlier research has argued that this extension was built during the years c. 1280–1300. However, on this point, the architectural and numismatic evidence is not entirely aligned. Many of the coins found in the extended church and minted with a terminus ante quem of 1263 (the death of King Håkon IV) might have been valid for only a short period after 1263 before a recoinage was undertaken. Accordingly, would these coins have still been in circulation if the nuns began using the new choir in the 1290s or shortly thereafter? Four possibilities thus arise: (1) that the theory that the coins were only valid for a short period before recoinage made them obsolete is false and the coins were in fact still valid around 1300; (2) that the church extension took place some decades earlier than c. 1280–1300, meaning that the nuns’ new choir may have been erected in the 1260s or 1270s, or perhaps even started in the 1250s; (3) that the nuns used old, nonvalid coins for their own purposes in the church; and (4) that the coins are not related at all to the Benedictine nuns’ activities, but to the builders of the nuns’ new choir, and were either deliberately left behind in the soon-to-be-consecrated space by masons and carpenters or dropped accidently over years of construction work. This suggests a new hypothesis: that all coins found at the Nunnusetr may present possible evidence for the phased construction and repair of the church over centuries and may help us to fix the actual construction dates for these architectural accretions more precisely. The third main question to consider relates to the Divine Office of the Benedictine nuns, which is discussed in relation to the coin finds within the liturgical layout of the church building and the nun’s activities, both devotional and personal, in the Benedictine period. The furnishing of the nuns’ choir is thus discussed, with an emphasis on the area of the stalls used by the nuns in the service. The devotional practices surrounding burials have proved decisive for the archaeological remains under church floors, and this is also the case at St Mary’s. So, the relation between burials and coin finds will also be discussed. We must in the following discussion hold several possibilities in mind simultaneously: that the coins were deposited by the nuns or through liturgical practices, like the Divine Office and funerals, in the church, or by the workmen, or by both.

Coins and monastic liturgy – middle ages  199

St Mary’s church and convent Two architectural remnants of the church at Nunnusetr still exist: the lower part of the west tower from c. 1130 and a vestry from the late thirteenth century. Archaeological investigations from the end of the nineteenth century to the beginning of the twenty-first century revealed the main layout of the church and a part of the churchyard, but all excavated monastic fragments have now been removed from the site. The church site between the extant tower part and the vestry is now occupied by two late nineteenth-century buildings erected after the excavation in 1891. Thanks to the excavations, we can state that the original twelfth-century Benedictine church was built with a rectangular layout, externally measuring 11.2 × 33 m (the first building phase, cf. Figures 8.1 and 8.9A). After it was extended with a new nuns’ choir in the thirteenth century, the church was c. 48 m long, still with a rectangular layout (the second building phase, cf. Figures 8.1 and 8.9B).4 This extended part of the church and its dating are of crucial interest in our discussion, since this is where the majority of coins were found. The vestry was not planned as part of the extension, but was added probably in the last decades of the thirteenth century, c. 1300 (the third building phase, cf. Figures 8.1 and 8.9C). The vestry had a special

Figure 8.4 St Mary’s at Nunnusetr. The extant part of the 12th century west tower with its main entrance to the church, seen from the north-west. Most likely this is the oldest building still standing in Bergen. Photo: A. T. Hommedal

200  Alf Tore Hommedal

Figure 8.5 St Mary’s at Nunnusetr. The extant vestry or side chapel erected c. 1300, seen from the east. The church’s chancel would be found on the right side of the vestry. Photo: A. T. Hommedal

architectural position, jutting out from the end of the chancel to the east by nearly 2 m. Both the original and the extended Benedictine church thus seem to have had a rectangular, narrow and aisleless layout connected to a west tower. The main part of the churchyard was situated to the north and north-east of the convent church, with the Benedictine nuns’ own burial area probably to the east of the chancel. We know nearly nothing about the ranges connected to the church. The monastic site to the south is at present occupied by a modern shopping centre. However, some building fragments uncovered by scattered excavations indicate that the nunnery had a conventional quadrangle with a central cloister to the south of the church. There are also a few archaeological fragments interpreted as remnants of the convent’s thirteenth- and fourteenth-century outbuildings.5

The archaeology of the coins: excavation, methods and interpreted results The two very first archaeological surveys at Nunnusetr revealed most of our archaeological knowledge of St Mary’s church and its churchyard (cf. Figures 8.1 and 8.2).

Coins and monastic liturgy – middle ages  201 The first excavation, in 1872, was led by the architect Peter Blix; the second, in 1891, by the architect Schak Bull. Both excavations were published by a third person, the teacher and trained archaeologist Bendix E.R. Bendixen, in 1893. The coin finds relate to the 1891 excavation.6 This excavation lasted from 2 June to 25 July, i.e. eight weeks in the summer, and was conducted “with such speed, as one in consideration of the works nature might, as the owners claimed the site cleared up as soon as possible.” There were found “a lot of burials,” and “spread out in the gravel lay quite a few medieval coins and a number of small objects: needles and mountings in bronze, small pieces of cloth etc.”7 Within the church, the excavated areas covered the entire floor of the nave and most of the chancel, except the very eastern part (the presbytery), excavated in 1872 and in 1891 already covered by new buildings. Neither the extant west tower8 nor the extant vestry was a part of the 1891 excavation area, and thus, the coins discussed do not relate to these parts of the church.9 Unfortunately, no systematic documentation in the two excavators’ own hand seems to have survived aside from the ground plans and drawings published by Bendixen in 1893, which must be regarded as the report for both excavations. However, Bendixen’s information on the 1891 excavation, given both in his publication and in connection to the coins registered in the University Museum of Bergen’s collections, seems highly reliable. In 1872, Bendixen was not involved in the excavation. In 1891, he was the central person in the cultural heritage work in Bergen and at Nunnusetr. Bendixen was then chairman of the Bergen section of the Society for the Preservation of Ancient Norwegian Monuments,10 which was conducting the excavation. His role was pivotal in the Society’s rescuing the extant tower part and vestry of St Mary’s, and at the excavation in 1891 before the church site was reused as building ground. Bendixen was also a member of the board, and soon to be chairman, at Bergens Museum (i.e. the present University Museum of B ergen), which received and catalogued the small finds from the excavation; Bendixen was himself engaged in the daily business of the Museum.11 He was not only an observer of the excavation in 1891, but was also in many ways in control of the entire situation, and can therefore be relied upon as the best authority to consult for further research.12 All that said, Bendixen’s Nonneseter klosterruiner gives hardly any information on the archaeological methods employed in 1872 and 1891. According to contemporary common standards of documentation and field methodology, the documentation of the 1891 excavation seems fairly good, although it lacks a systematic, contextual approach. For late nineteenth-century excavations, we would expect that the soil from the site would have been shovelled and the excavation carried out without a stratigraphic approach to the cultural layers. Also, there would have been a recovery bias focused more on the relationship of objects to the constructed parts of the revealed building, and less on the relationship between the assemblages of the portable artefacts, including coinage, themselves. In some cases, Bendixen provides information for the coins, e.g. “found in the chancel.”13 Probably the soil was sieved, even though this would have been an uncommon practice at the time and would have been time-consuming.14 Still, sieving seems one of very few possible explanations for the collection of the relatively large amount of tiny bracteates and bracteate fragments – several weighing less than 1 gram and some even less than 0.10 gram – found among the coins.15

202  Alf Tore Hommedal

Figure 8.6  T he three building phases in St Mary’s at Nunnusetr: the original church; the extended church; and the extended church with the added vestry. The crosshatched irregular field marks the documented tile floor. Drawing by Per Bækken and Alf Tore Hommedal, remapped by Steinar Kristensen, UiO: Museum of Cultural History.

Another explanation could be that the excavations took place inside still standing remnants of the burnt wooden main building at Lungegården. According to Bendixen,16 the wooden building covering parts of the area of the former church stood on a layer of gravel that covered fragments of the chancel and nave. This indicates that the cultural layers from the church were preserved when the building burned down on 12 February 1891. If some non-burnt parts of the building17 were still standing or were newly torn down when the archaeological excavation started in June,18 this may also explain the relatively large number of bracteates: the dust or soil underneath

Coins and monastic liturgy – middle ages  203 these building parts would still be dry, making the bracteates easier to see; wet, lumpy soil would have made the bracteates especially difficult to observe. However, such an explanation relies on remnants of the main building still standing, since it was definitely raining during the excavation (21.2 mm of rain fell on 3 June, the second day of the work).19 We must then also be aware of the possibility that parts of the church area had lumpy soil and other parts dusty soil during excavation, making the conditions for coin finds unequal. However, we must relate to the coins found, even though we also must be awake to the possibility of coins not sampled, as would always have been the case, to a greater or lesser extent, for all excavations at that time. The fact that small coin fragments were recovered and accounted for suggests that the excavators were carefully documenting even the smallest objects. Therefore, the coins found in the archaeological record can be considered a representative sample of the presumed larger body of coins brought into the church during the whole span of the Middle Ages. The total number of coins from St Mary’s is few considering the church’s 400-year-long existence. However, this may reflect reality, and it is the material that left us. We will willingly go where the data are, and thus propose what the material indicates to us. The conclusions may be vague but may shed some light on small details of monastic life in a nunnery in medieval Norway, at the northern periphery of Western Christendom. The contextual terminology of the church’s areas All the known coins from St Mary’s, except one (Appendix, BRM 650/14), were found during the excavation in 1891. According to Bendixen, 20 the majority of the coins were found in the part of the church area he defines as the chancel (“koret”), i.e. the thirteenth-century extended eastern part of the church, which is interpreted as the “nun’s new choir” in the following (cf. Figures 8.2 and 8.9). Bendixen’s nave (“skipet”) will be used as synonymous with “the original church room,” meaning the original church excluding the tower area, which was not included in the 1891 excavation. “The original church” includes the tower area. “The original church room” then includes the original presbytery, the nuns’ choir and the nave, while “the extended church” means the total church area included the new presbytery, the new nuns’ choir, the nave and the tower area but not necessarily the vestry (cf. Figure 8.9). Nunnusetr: some premises for the coin interpretation Three premises are fundamental to the interpretation of the coin material from Nunnusetr. The first one deals with the delimitation of the Benedictine nunnery phase. When the Benedictines abandon the nunnery and what happened on the church site immediately afterwards are questions whose answers bear heavily on the archaeological interpretation of the Benedictine nunnery phase. The second premise deals with the location of the Benedictine nuns’ choir within the church area already indicated. The third one deals with the dating of the church’s eastern extensions. The historical periods. Nunnusetr as a Benedictine nunnery The Benedictine nunnery seems to have been created as a royal foundation sometime in the 1120s. 21 Nunnusetr (St Mary’s) in Bergen was thus established just a decade or two after Munkalif (St Michael’s), the royal foundation for Benedictine monks dating

204  Alf Tore Hommedal to c. 1110 (cf. Figure 8.3). 22 The two abbeys may reflect a royal idea of establishing parallel Benedictine institutions for monks and nuns, most probably inspired by England. 23 An establishment date for the Benedictine nunnery in the decade before 1130 seems also to concur with the High Romanesque architectural elements of the extant part of the church’s tower, probably built around 1130. 24 In the period before the crisis following the Black Death in 1349–50, St Mary’s seems to have been the largest and wealthiest of Norway’s five nunneries in this period. In 1320, there were 35 qualified nuns residing there, and the nunnery was thus a rather large convent, comparable to nunneries throughout Europe.25 However, the number of nuns declined later in the Benedictine period. There were supposedly just a few nuns left when the nunnery was closed in the reign of King Christern I (1450–1481). The nunnery’s buildings and estates were then subordinated to Munkalif, by then a Bridgettine abbey.26 The event that most likely directly led to the subordination was a fire at Munkalif in 1455. This disaster seems to have been so all-­consuming that it took over 20 years to rebuild the abbey. It is inferred that the Bridgettines – seeking a place of refuge – were taken in by their Benedictine sisters and then took over Nunnusetr.27 However, after some years, a rather unique situation came about: the Bridgettines left Bergen for the Cistercian Abbey at Hovedøya, close to Oslo, while their hosts, the Cistercians, went to Bergen, probably to take on the task of rebuilding the burned-down abbey at Munkalif.28 The Cistercians may also have lived at Nunnusetr for a period, until the rebuilding of Munkalif provided them with suitable accommodation.29 This possible period of temporary use of Nunnusetr, first by the Bridgettines and then by the Cistercians, would, at the latest, have ended in 1478, when the Cistercians were back in Oslo and the Bridgettines in Bergen, and definitely installed in the restored buildings at Munkalif. We do not know how the buildings at Nunnusetr were used for the rest of the period of its subordination to Munkalif, i.e. c. 1478–1507, but the church seems to have been in constant use.30 In 1507, the former nunnery was transferred to the canons of St Anthony’s Order, known as Antonites, on the condition that the canons keep the daily church services and restore the buildings.31 The Antonites were mendicants organised as a hospital order specialising in nursing those with shingles, skin diseases and venereal diseases. We must therefore assume that they ran a hospital at Nunnusetr, as the Benedictine nuns perhaps did. Often, the canons also ran hostels for pilgrims and other travellers. The canons, however, were not allowed to practise their healing arts for more than about 20 years. In 1528, Nunnusetr was confiscated by the King, given to the nobleman Vincent Lunge and rebuilt as a private residence, the manor later known as Lungegården. We can then sum up: St Mary’s existed as a monastic site for about 400 years, ­divided into three consecutive monastic phases; a fourth post-monastic period ­followed, as shown here: Table 8.1  P  roposed chronological phases based on orders resident at the Bergen Nunnusetr Benedictine

1120s–c. 1455

Bridgettine (including possible short Cistercian period) Canons of St Anthony’s Order (Antonites) Post-monastic (ended with fire)

c. 1455–1507 1507–1528 1528–1891

Coins and monastic liturgy – middle ages  205 For the interpretation of the archaeological material in general, and for the coin material in particular, it is important to stress that the church continued to have a monastic function after the Benedictine nuns left around 1455. When it comes to the discussion of the coin finds, there are no indications that St Mary’s had an additional function as a parish church in Bergen in the Benedictine phase as, by comparison, most of the Danish nunnery churches seem to have had.32 This situation remained unchanged in the post-Benedictine phases of Nunnusetr. The Benedictine nuns’ choir If we now look at the position of nuns’ choirs in a general European paradigm, two options seem plausible for the location of the choir in St Mary’s. First, the nuns could have been seated in a gallery or rood loft, preferably in the western part of the nave, as was the case in Danish and partly in tenth-century German nunneries.33 In English nunnery churches, however, western galleries were relatively rare. The second possible location of the nuns’ choir was on the ground level in the eastern part, i.e. the chancel part of the church, next to the presbytery. Such a location seems to have been quite common in German nunneries in the eleventh-to-thirteenth centuries.34 In England, after the Norman Conquest in 1066, even large nunnery churches appear to have operated only at ground level, and there is no evidence of access to galleries. 35 In both architectural conventions for the location of the nuns’ choir, gallery or chancel ground level, the laity – if present – would have been located in the western part, i.e. the nave, of the church. A third option, with the nuns’ choir located at ground level in the west end of the church,36 seems unlikely in St Mary’s because the church has a west entrance. Alas, no remnants of the other Norwegian Benedictine nunnery churches have been preserved to indicate an answer to the question. Both Bendixen and the architectural historian Hans-Emil Lidén place the nuns’ choir on the ground level in the eastern part of St Mary’s, even though they do not unpack the question explicitly.37 In a more thorough discussion of the Benedictine nunnery church in Bergen, it is deduced that St Mary’s followed the English pattern for locating the nuns’ choir.38 This inference rests on two arguments. Firstly, the ground plan of St Mary’s seems to fit into the pattern of most English nunnery churches, i.e. a parallelogram ground plan with frequently narrow, aisleless rectangles, in appearance unusually tall and thin, sometimes very elongated and often with west towers. 39 With the connection that western Norway appears to have had to England regarding monastic foundation in general in the twelfth century, an English influence on St Mary’s church is plausible.40 Secondly, this location for the nuns’ choir is indicated by a large fragment of tile floor found intact in the extended part of the church in 1891 (cf. Figures 8.2 and 8.9. The square floor tiles (14 × 14 cm) were arranged in a diagonal pattern of reddish yellow and green towards a straight, orthogonal edging band, ending in a line c. 1.25 m from the south wall.41 The floor seems then to have been a walkway fitting longitudinal rows of prayer stalls with stools and kneelers running to the north and south.42 There was no need for a beautiful and valuable tile floor to continue underneath the stalls. There, one would rather expect a wooden floor, and, if so, it might well have been elevated above the level of the tile floor to help insulate the feet in the cold winters. The wooden floorboards could have had gaps between them where coins and other small items would fall and disappear beyond reach. Of course, the tile walkway would, in general, produce less coin finds than wooden flooring.43

206  Alf Tore Hommedal However, the tile floor does not necessarily reflect the interior of St Mary’s in the Benedictine period. It could also relate to the choir’s functions in the monastic periods after c. 1455. Lidén dates the floor to the second half of the thirteenth century, but this seems more due to his dating of the extension of the church than to an examination of the floor itself.44 Tiles from the floor are kept in the University Museum of Bergen’s collections, and these do seem more likely to be from the thirteenth or fourteenth century rather than the late fifteenth or sixteenth century.45 Therefore, it is reasonable to associate these tiles also with the Benedictine nuns and not only with their successors. Such traces of tile floor have also been used to identify the location of German nuns’ choirs.46 Thus, moving forward, the interpretation of the coin finds will be based on the solid premise that the location of the nuns’ choir in St Mary’s was on the ground floor in the eastern part of the church, not only in the extended church but also in the original church (cf. Figure 8.9). Dating the church’s extensions to the east Architectural fragments from St Mary’s, combined with written sources, indicate that the extension of the church took place within the second half of the thirteenth century. According to Lidén, moulded building stones from plinths and windows, and perhaps also tracery windows, relate the extension of St Mary’s to the general church architecture in late thirteenth-century Bergen.47 This architecture is characterised by English Gothic, in particular the Early English and Decorated styles, as found in the Franciscan friary church of St Olaf’s, the present Lutheran Bergen Cathedral.48 Lidén also suggests, as Bendixen did, that the extension of St Mary’s was connected to King Magnus the Law Mender’s generous testamentary gift of 40 marks sterling to the nuns in Nunnusetr in 1277.49 Lidén therefore suggests that the extension of the church at Nunnusetr took place after from the King’s death in 1280.50 Both Bendixen and Lidén, in their consideration of the dating of the extension, also discuss the coins uncovered in Bendixen’s chancel. However, both of course relate to Bendixen’s original identification of the bracteates as coming mainly from 1263 to 1355. This, according to Lidén, “confirms the presumption that the chancel extension may be from the last quarter of the 1200s.”51 With most of these bracteates re-dated to the period c. 1220–1263 (see Appendix), the age of St Mary’s extension is open for reconsideration.52 None of the building fragments from the chancel extension indicate a dating exclusively in the last quarter of the thirteenth century. The present north wall in the extant vestry is the only standing masonry part of the extension. It has a moulded plinth (cornice, cyma recta), but this may even be from earlier in the 1200s. Some, now lost, fragments of a blank arch, 53 most likely from the extension, have the same character as a blank arch on the western part of the Franciscan church at St Olaf’s. This western part has been dated to after a fire in 1270; however, it may seem more probable that it was initiated after a fire in 1248.54 A fragment of a tracery window from the last decades of the thirteenth century has been identified to St Mary’s. However, this may be from the third and last building phase in the church, or it may have been re-deposited from another church. 55 The building fragments that have been conserved do not exclude a date for the extension before the last quarter of the thirteenth century, and thus, a slightly earlier dating might be possible, perhaps the third quarter of the 1200s.56

Coins and monastic liturgy – middle ages  207

Figure 8.7  T  he extant chancel wall of the extended St Mary’s church (the second building phase), now the north wall of the extant vestry (the third building phase), seen from the south. The wall thus represents the external south wall of the presbytery with its plinth, ending at the south-east corner of the chancel, the vestry jutting out farther to the east. The door opening led from the vestry to the presbytery. Photo: A.T. Hommedal

Nor does Magnus the Law Mender’s testamentary gift of 1277 exclude an earlier dating of the new choir. Perhaps the idea that Magnus’s testament provided the means to extend the church has been overemphasised. This lump sum of money may instead be linked to the third and last building phase in St Mary’s: the addition of the extant vestry to the south of the chancel (cf. Figure 8.9C), probably with a parallel to the north.57 The two vestries or choir chapels would have thus framed the east wall of St Mary’s, in an architectural parallel, on a smaller scale, of the present east front of Stavanger Cathedral. In Stavanger, this part of the cathedral, where the framing building parts jut out like the extant vestry in Bergen, is interpreted as having been built after a fire in 1272 and in the decades up to 1300.58 Finally, the architectural parallels between St Mary’s vestry and the Franciscan St Olaf’s church and vestry also do not preclude a possible dating for the last building phase in St Mary’s after 1277. This final architectural addition to St Mary’s may then have taken place in the 1280s or early 1290s. The consecration of the extended St Olaf’s, which was also extended due to the testamentary gift of 1277, took place in 1301. The testamentary donation made to the Franciscans and St Olaf’s was more generous than that given to St Mary’s, presumably because the King chose to be buried in St Olaf’s.59 Even though it is undocumented in the narratives, it is possible that Magnus the Law Mender gave another donation to St Mary’s and the Benedictine nuns before 1277. In the autumn of 1262, King Magnus and Queen Ingeborg left their infant firstborn son, Olaf, in the care of the nuns at Nunnusetr, while they continued on to Nidaros (Trondheim) for the winter.60 The boy died some years later, at only five years of age,61 and he may have been disabled.62 If so, and the Benedictine nuns nursed and

208  Alf Tore Hommedal

Figure 8.8  Stavanger Cathedral with its east front built after a fire in 1272.  Photo: A.T. Hommedal

took care of the child, the royal household surely would have supported the nuns, who presumably buried him in St Mary’s. As late as 1275, we here about “the grief, as all men in Norway carried, when Prince (junker) Olaf, son to King Magnus, died.”63 The situation with Olaf may also have been the main reason for King Magnus’s testamentary gift to Nunnusetr in 1277, given ten years after Olaf died and only two years after we hear of the great grief over his death. However, a significant donation from the King in 1262 may have provided the Benedictine nuns caring for his son at the time with funding for an extension to their church as early as the 1260s. The architectural and archaeological remains from the extended church, i.e. the nuns’ new choir, do not contradict such a dating. We must therefore be open to the idea that the new nuns’ choir in St Mary’s was completed and consecrated around 1270 or even earlier, and certainly in the 1270s at the latest. 64 However, the architecture of St Mary’s seems not to be older than c. 1250 at its oldest. The still standing vestry on the south of St Mary’s, in all likelihood with a parallel to the north, was not a part of this first extension. The vestry/vestries were part of the second and third (and final) architectural additions to St Mary’s. 65 This extension may have been erected as a consequence of the testamentary gift in 1277 and then completed in the 1280s or 1290s. The completed St Mary’s, as we know the total church, must then have been finished before 1300.

The coins, their identity and their archaeological context According to Bendixen,66 at least 60 coins were found during the excavation of St  Mary’s in 1891. Approximately 40 of the coins were bracteates “largely fragmented, which makes it hard to give a precise number.”67 In 2014, a new survey of the coins in the University Museum of Bergen’s collections was carried out by Professor

Coins and monastic liturgy – middle ages  209

Figure 8.9  A  suggested division of St Mary’s at Nunnusetr. The red lines indicate the screens separating the nuns’ choir (middle part) from the presbytery and the nave. The crosshatched irregular field marks the documented tile floor. From the top: the original, twelfth-­century church (A); the extended church, imagined with two or four rows of stalls in the nuns’ choir (B and C). Drawing by Per Bækken and Alf Tore Hommedal, remapped by Steinar Kristensen, UiO: Museum of Cultural History. Table 8.2  St Mary’s. Proposed chronological dating of the church’s building phases, containing both the periods of work and completion The original church (building phase 1)

1120s/1130s

The extended church, the nuns’ new choir (building phase 2)

1260s/1270s

The vestry (vestries?) (building phase 3)

1280s/1290s

210  Alf Tore Hommedal Svein H. Gullbekk, and 55 coins can still be identified.68 Many of Bendixen’s interpretations were also confirmed. However, a great part of the type of bracteates dated by Bendixen to c. 1263–1355, according to the nineteenth-century numismatist Claudius I. Schive’s catalogue of Norwegian medieval coins, are now attributed to the reign of Håkon IV Håkonsson (1217–63) according to twentieth- century numismatist Kolbjørn Skaare.69 The 42 coins from the Benedictine period, their periodical spread and their context are listed in Table 8.3. Table 8.3 also includes the German viertelwitten found in Nunnusetr in 2006, making the total number of 43 coins from the Benedictine period of St Mary’s. Table 8.4 contains a corresponding set-up for the coins from the post-Benedictine period, not discussed here. The basis of both tables is listed in the coin catalogue (see Appendix Table 8.5). In the following discussion, when the coins’ museum numbers and dating are given, these relate to the coin catalogue.

Figure 8.10 Four of the coins found in the extended part of the church (Bendixen’s chancel, i.e. the nuns’ choir). See Appendix, MA 348/20 (1–4). Photo: S. Skare, University Museum of Bergen

Coins and monastic liturgy – middle ages  211

Figure 8.11 Bracteate minted for Magnus the Law Mender, found in the extended part of the church (Bendixen’s chancel, i.e. the nuns’ choir). See Appendix, MA 348/15. Photo: S. Skare, University Museum of Bergen

Figure 8.12 Hohlpfennig minted for Erik of Pomerania. See Appendix, MA 348/16. Photo: S. Skare, University Museum of Bergen

None of the coins in Nunnusetr relate to the twelfth century. An English sterling (cf. Figure 8.14) issued between c. 1194 and 1200 (Appendix, MA 348/18 (3)) could theoretically have reached St Mary’s in the last decade of the twelfth century. However, the wear and tear of the sterling indicates that it was in circulation for decades. We must therefore assume that this foreign coin ended up in St Mary’s towards the middle or second half of the thirteenth century. Except for this English sterling, all 28 coins from the thirteenth century are Norwegian, as is common for coin finds in churches within the Norwegian realm. All of them are bracteates, as this was the only denomination minted in the thirteenth century before c. 1263 except for a limited

212  Alf Tore Hommedal

Figure 8.13 Two hohlpfennigs minted for Erik of Pomerania and one bracteate minted for Magnus the Law Mender. See Appendix, MA 348/17 (1–3). Photo: S. Skare, University Museum of Bergen

coinage of half-bracteates c. 1217–1220; there are fewer Norwegian coins from the late thirteenth century, but they are of higher value, such as pennies, halfpennies and farthings.70 Another English sterling (Appendix, MA 348/39 (1)), minted 1272–1377, most probably ended up in Nunnusetr in the fourteenth century. Of the other six coins that found their way to St Mary’s in the fourteenth century, there is one presumed Swedish coin and one without provenance. The three others are probably Norwegian: one bracteate, one hohlpfennig and one halfpenny, the last minted before 1319. Of the fifteenth-century coins, the majority (eight out of nine) were minted in Denmark within the first two decades of the century, but one coin, a Danish Groschen (Appendix, MA 348/10 (2)), dates from the 1430s and could theoretically have ended up in St Mary’s after the Benedictine nuns left around 1455. Norwegian coins were not minted between 1387 and the 1470s or 1480s. We can then conclude that the majority of the coins from the thirteenth and fourteenth centuries are Norwegian. In the Benedictine decades of the fifteenth century, Danish coins dominate.71 In the first periods of coins found in Nunnusetr, the lowvalued bracteates dominate, while coins with a higher value are represented from the last decades of the thirteenth century onwards. The trend in the pattern of coin finds is similar to what we see in parish churches in Norway in general and should be seen as a reflection of the availability of the monetary pool in general.72

Coins and monastic liturgy – middle ages  213

Figure 8.14 Five coins found in the church without detailed context. The upper one is the oldest coin found in Nunnusetr, a sterling minted in Canterbury for King Richard I or King John of England, between 1194 and 1200. See Appendix, MA 348/18 (1–4). Photo: S. Skare, University Museum of Bergen

Figure 8.15 A bracteate of Håkon IV Håkonsson. See Appendix, MA 348/19 (1). Photo: S. Skare, University Museum of Bergen

214  Alf Tore Hommedal Table 8.3 T he Benedictine period, c. 1120–1455. Coins identified from the excavations of 1891 (42) and 2006 (1) with their context Order and period Benedictine, c. 1120–1200 Benedictine, c. 1200–1300 Benedictine, c. 1300–1400 Benedictine, c. 1400–1455 Benedictine, total c. 230 years (c. 1120–1455)

Coins, chancel

Coins, nave

Coins, churchyard and cloister

Coins, no context

Coins, total

12

1

1

1 13

1 2

1

14 6 7 27

0 28 6 9a 43

a Eight of them were found in 1891 and one in 2006.

Table 8.4 Post-Benedictine periods (c. 1455–1891) and non-identified coins, with their context. All coins identified from the excavation of 1891 Order and period

Coins, chancel

Coins, nave

Bridgettine (and Cistercian?), c. 1455–1507 Antonites, 1507–1528 Non-identified hohlpfennig Post-monastic manor, 1528–1891 Non-identified coins Total

1

1

Coins, churchyard and cloister

1

Coins, total

3

5

1

1 1 2

2 1 3

2

3 10

5 16

1

2 3

Coins, no context

Sixteen of the coins from the Benedictine period can be traced back to locations inside and outside the church.73 Thirteen of them can be related to the area of Bendixen’s chancel, i.e. the nuns’ new choir; two can be related to Bendixen’s nave, i.e. the original church room, and one can be related to the churchyard (Table 8.3). The real number of coins found in the nuns’ new choir in 1891 must, however, be higher than the thirteen coins contextually identified to Bendixen’s chancel: according to Bendixen, more than three-quarters of the total number of coins74 were found in the “chancel”, i.e. the new nuns’ choir. We must assume that this division represented the coins from the Benedictine period, indicating that c. 31 of the 42 coins from the Benedictine period and found in 1891 were found in the nuns’ new choir. More than half of the 26 coins found in 1891 and registered with no context in Table 8.3, and probably c. 18 of them, must then have originated from the nuns’ new choir. The 27th of the coins without context in Table 8.3 is the German viertelwitten found in 2006, and thus not a part of the coin finds from 1891. This coin was found in secondary layers deposited in the manor’s post-Reformation garden to the east of the church together with other material, for example tiles of the same type as found in situ in the new nuns’ choir. It is thus a possibility that the German viertelwitten also originated from the new nuns’ choir. However, it is listed in Table 8.3 as a coin without context.

Coins and monastic liturgy – middle ages  215 Even though the post-Benedictine coins are not discussed, they are listed in Table 8.4 to give an impression of their number and extensiveness compared to the coins from the Benedictine period. The coins are few, but this could in part be due to the shortness of each period: the Bridgettines lasted only c. 50 years; the Antonites scarcely 20 years. It is noteworthy, however, that the longer post-monastic and secular manor period, spanning more than 350 years, also shows a lack of coins. The non-identified hohlpfennig from the Middle Ages and the other non-identified coins may be from the Benedictine period, but they have been left out of the discussion. Comparative material from some other nunneries If we compare the coins material from Nunnusetr with coin finds from other nunneries, a difference emerges: it seems unusual to find coins in nunnery churches. At the Cistercian Alvastera Abbey in Sweden, most of the approximately 1,450 coins recovered were found in the monastic buildings, with only a few from within the church itself.75 And the same pattern occurs at other monastic sites. Since the areas that would have contained the cloister and the monastic buildings at Nunnusetr have been lost, we cannot know if this would have been the case in the Bergen nunnery as well. However, any coins from these areas would all the same have come in addition to the coins recovered from the church. We will never know! In the Benedictine nunnery of St Johann in the valley of Müstair, Switzerland, 895 coins were found, but none in the nunnery church (although the church seems not to have been part of the excavation).76 In the Swedish Cistercian nunnery of Gudhem, on the other hand, the situation seems to have been the same as in St Mary’s, with the majority of the coins found in the nuns’ choir, near the nuns’ stalls. However, it seems that nobody who has discussed Gudhem has questioned why the majority of the coins were found in the nuns’ choir. We will now turn to the question as to whether the coins in Nunnusetr can be reliably related to the nuns’ choir in St Mary’s.

Furnishing St Mary’s: the Benedictine’s new nuns’ choir Tile floor and prayer stalls The extended Benedictine church seems to have been finished and consecrated in the 1260s or the 1270s at its latest. A large, intact part of the beautiful tile floor filling the middle of the room and ending in a line c. 1.25 m from the church wall has already been interpreted as a walkway fitting a longitudinal row of prayer stalls with stools and kneelers. The tile floor has been interpreted as dating from the Benedictine nuns’ period in the church. We can thus conclude that the majority of the coins found in the extended part of the church can be related to the Benedictine nuns’ choir c. 1260s/1270s–c. 1455 (cf. Figure 8.9B). The old nuns’ choir in the original church would have been abandoned at the same time, i.e. presumably during the 1260s or the 1270s (Figure 8.9A).77 The extent of the nuns’ choir and presbytery We must assume that rood screens, in stone or most probably wood, separated the nuns’ choir from the presbytery to the east and the nave to the west in St Mary’s.78 No traces of such dividing screens have been documented. The east wall foundation

216  Alf Tore Hommedal of the original church (cf. Figures 8.1–8.2 and 8.9A) has been interpreted as a foundation for the screen separating the nave from the nuns’ choir in the extended church.79 However, the foundation does not seem to demarcate a western limit for the area of the nuns’ choir.80 It is possible that the new nuns’ choir stretched to the west of the foundation and into the area of the original church (cf. Figure 8.9).81 In all cases, the documented tile floor indicating stalls shows that the area directly to the east of the foundation was utilised as a choir, at least in the last part of St Mary’s existence as a church and probably already from the thirteenth century onwards. We would expect that foundations for wooden cross-screens would have been very simple, either rooted in postholes or most probably on sill stones, or simply on the floor. This may partly be the reason why no such foundations were documented during the excavations in 1872 and 1891. Neither have any foundations for lay altar(s) been observed, which were likely to have been located on the nave side of any separating screen.82 Where then should we look for the locations of the screens? If we compare the outstretched, rectangular layout of St Mary’s to English nunnery churches,83 we would expect St Mary’s to have been bisected or divided to give the nuns’ choir space a larger area. But with a rood screen on the original east wall foundation, the nave would cover the largest liturgical area of the extended church. This seems implausible, both compared to English models and taking into account the fact that Nunnusetr in Bergen was at times a relatively crowded nunnery. The extension of the nuns’ choir. Alternative I: two rows of stalls In 1320, and again in 1326, the nuns’ choir in St Mary is indirectly mentioned in the narratives. On both occasions a new abbess had been elected and was to be enthroned in the church. In 1320 we also learn that there were 35 qualified nuns84 at St Mary’s, and with the novices included there were probably 40 to 50 regular women in total. If we consider that the 40–50 nuns and novices in 1320 filled two rows of choir stalls as indicated by the documented tile floor, one to the north and one to the south, this gives 20–25 women seated in each row. The width of each of the stools, including the armrests dividing them, must have been around 75 to 80 cm. The rows would then have been between 16 and 20 m long, indicating that the nuns’ choir had a total length of at least 22 m, allowing for passages and a doorway (cf. Figure 8.9B).85 The middle part separating the two rows of stalls would have been approximately 5.9 m wide. This spacious walkway, with its beautiful tile floor, would have been a worthy setting for the ceremonial enthroning (intronizare) of the new abbess, Jorunn Constantia, in 1320, when she was led to the abbess’s seat by two secular canons, sire Salve and sire Simon, on behalf of Bishop Audfinn of Bergen. In 1326, the ritual was repeated by the same two canons, then enthroning Abbess Ingeborg Cecilia.86 The extension of the nuns’ choir. Alternative II: four rows of stalls What then if the documented tile floor is from a later and less crowded period of the nunnery, or from post-Benedictine times, and the nuns’ choir in the 1320s was

Coins and monastic liturgy – middle ages  217 divided into four rows of stalls, two on each side and with no traces of the existing floor of the walkway preserved? If we also calculate that each of these hypothetical rows jutted 1.25 m into the room, the width of the four rows would cover 5 m altogether.87 With the church’s internal width measuring c. 8.4 m, this would allow for a walkway measuring c. 3.4 m wide in the middle of the room, maybe also covered with a tile floor. Even this narrower mid-section would have provided a worthy passage for the ceremonial enthroning of the two abbesses. The nuns’ choir would then at a minimum have been 11 m long and most probably at least 1 or 2 m longer, giving added space for the doorway from the cloister and side passages to the stalls (cf. Figure 8.9C).88 The extended St Mary’s had a total internal length of c. 38 m, excluding the tower area. If we consider the presbytery, with the high altar, in the 1320s to have been c. 6 m long, and the nuns’ choir with two rows of stalls to have been 22 m long, this would indicate a nave of only 10 m (c. 17 m when the tower area is included) (cf. Figure 8.9B). If we consider the nuns’ choir to have had four rows of stalls, the choir would have been c. 13 m long. Such an extension of the choir would indicate a 19-m-long nave, or c. 26 m long when the tower area is included. This would indicate an exactly bisected division between the presbytery/nuns’ choir and the nave without the tower area (cf. Figure 8.9C). It seems then most likely that the extended church originally had a nuns’ choir with four rows of stalls. When the number of sisters was reduced later on, the choir may have been refurnished with two rows, creating the spacious and luxurious tiled walkway documented in 1891.89 The original choir: the number of nuns and the extension of their choir. Four rows of stalls What then was the division of the original twelfth-century church? Before the extension of St Mary’s, the total internal length of the church, without the west tower area, would have been close to 22 m. If we consider the presbytery, with the high altar, to have been c. 5 m long and the rest of the church to have been bisected, the nuns’ choir and the nave both would have been c. 8.5 m long, and the nave c. 15.5 m with the tower included (cf. Figure 8.9A). If we furnish this nuns’ choir with four rows of stalls and deduct space for the doorway from the cloister and for side passages to the stalls, the rows of stalls may have been c. 6.5 m long, indicating up to 32 pairs of stools and kneelers. Deducting the novices, the number of qualified nuns may have been around 20, which seems a reliable figure. If the nuns’ choir was furnished with only two rows of stalls, the number of stools and kneelers cannot have exceeded 16. The number of qualified nuns must then have been 12. We must assume that both the original and the extended church at Nunnusetr were divided in such a way that the presbytery and nuns’ choir comprised at least half of the total length of the church (without the tower area). It therefore seems likely that both the original and the extended nunnery church at least were bisected. Regardless of the division, we must conclude that the extended church’s nuns’ choir stretched to the west over the foundations of the original church’s east wall and into the area of the original church. As a result, some of the coins found in the easternmost part of Bendixen’s nave, if these date from the period after the church’s extension, may have been left in a part of the original church then in function as the westernmost part of the new nuns’ choir. Alas, it is impossible to identify and pinpoint these coins, which anyway seem to have

218  Alf Tore Hommedal been few. We must also be aware that the original nuns’ choir was situated in the area by Bendixen defined as “the nave,” i.e. the original church. It is then interesting to note that none of the oldest coins was contextualised by Bendixen to this original church. Of the bracteates from the time of Håkon IV, seven are without any context; thus, between none and seven of them may have been found in Bendixen’s nave. If so, however, they relate to a period when a large part of the original church functioned as either the original nuns’ choir (cf. Figure 8.9A) or the western part of the nuns’ new choir (cf. Figure 8.9B). Even if found in the nave in 1891, there is then the potential that these bracteates would have been functionally related to the nuns’ choir. It would seem then that none of the oldest coins from St Mary’s can be contextualised with certainty to the nave’s function in the church. The oldest coin we can contextualise to the nave is an Eirik Magnusson penny (Appendix, MA 348/12) minted c. 1280–1285, from a period when the nuns’ new choir probably was in use. The penny was found in a grave (stone grave III) and was thus probably not related to the ordinary function of the nave. Only two coins from St Mary’s seem to have been found in a nave context, and both are from the fifteenth century, thus from the last years of the Benedictine period or perhaps even the post-Benedictine period (Appendix, MA 348/26 (1) and MA 348/26 (2)). It is then possible that none of the coins from St Mary’s Benedictine period can be related to the functional nave part in the church.

The liturgy in Nunnusetr. The Benedictine nuns’ Divine Office The Benedictine nuns, living according to the Benedictine Rule, ordered their day along the same lines as the Benedictine monks, namely around the liturgical hours of the Divine Office as described in St Benedict’s Rule § 8–20.90 Given the “English” localisation of the nuns’ choir in the Benedictine nunnery in Bergen, we can assume that the liturgical practice in Nunnusetr was also English in its form and rituals. At least, this must have been the situation in the first period of the nunnery’s existence in the twelfth century, and probably also later on.91 The liturgical practice in Nunnusetr then, in all probability, differed very little from what is known of liturgical practice in Benedictine nunneries in England. If the nuns in Nunnusetr followed their English sisters’ horarium, their day began with Matins at about 2 a.m. and ended about 7 or 8 p.m. with Compline. In between, Lauds, Prime, Terce, Sext, None, Vespers and two Masses (the Chapter Mass, followed by the chapter, and the High Mass) provided the framework around which other monastic activities were set.92 In Norway, the short length of the night office on great feasts in the summer may have caused a problem. As David Knowles points out, a solution on those nights may have been to move Matins immediately after Compline the previous evening, thus creating a single period of rest after the office. For the nuns in Nunnusetr, the day-and-night cycle within the walls of the nunnery would then have been divided into three parts. The first third was reserved for the Divine Office with liturgical prayer and mass preceding all other activities. The second third was reserved for reading and studies, and for manual work. The last third was reserved for rest and sleep. The three parts were subdivided into intervals, so that, for example, the nuns congregated in the church to pray eight times during a 24-h cycle, seven times during the day and once during the night. To take part in the full Divine Office, the nuns in St Mary’s were thus required to spend a great part of their day – and life – in their choir and in their stalls.

Coins and monastic liturgy – middle ages  219 In the discussion of the liturgy in Nunnusetr, we must also be aware of two facts: firstly, the nuns themselves could not celebrate mass and administer the sacraments; and secondly, the nuns were not allowed to leave the clausur, and nobody from outside – not even a priest – was allowed to enter it. As a consequence, as already discussed, there must have been a screen totally separating the presbytery, where the priest celebrated mass, from the nuns in their choir in St Mary’s. This separation even had to include the sacrament of communion, so contact between the celebrant and the nuns must have been restricted to an opening in the screen. The sacrament of confession and absolution was also most probably performed at the same screen opening.93 The Rector role in Nunnusetr must have been filled by a secular priest or – more probably – by Benedictine monks from Munkalif. The nunneries were often related to, or even subordinated to, a nearby monastery,94 and a connection between the two Benedictine monasteries in Bergen seems likely. After the Bridgettines took over Munkalif in the 1420s, a Bridgettine priest may have taken over the role of Rector. A part of the liturgy connected to the mass in Nunnusetr may also have been the daily recitation of the Office of All Saints. At Barking, a Benedictine nunnery in Essex, England, this remained part of the liturgy until at least the early fifteenth century even though the Benedictine monks had repealed its use.95 We must also assume that parts of the liturgy that did not involve a priest could be performed outside the church but within the nunnery. For example, among other processions, nuns were expected to hold a Palm Sunday procession and venerate the cross in their own cloisters.96 On some very special occasions, the nuns also seem to have been allowed to leave the clausur and their nunneries. In Bergen, the Bishop’s general ordinance on processions and masses in times of plague, supposedly given in the fourteenth century, perhaps outlines just such an occasion: to prevent God’s wrath, masses were to be celebrated on five consecutive days with the penitents walking in procession around Bergen to the monasteries surrounding the town and to the cathedral. Having fasted for four days, each penitent would walk barefooted in solemn parade, carrying a burning candle, and giving alms to the poor. The participation of the contemplative Benedictine nuns in such an event would of course have lent it an extra dignity and power. If the nuns did not take part themselves, they at least would have hosted the penitents on the third day, when the procession left for Nunnusetr.97 In connection with such events specifically, and with the Black Death and other plagues in general, we must assume that coin alms were given in St Mary’s.

Loss or sacrifice. What do the coins represent? We must bear in mind the fact that more than three-quarters of the coins found in St Mary’s were found in the extended part of the church. We must also bear in mind that the majority of these coins were from the Benedictine period in Nunnusetr and, thus, from the nuns’ new choir. Even though some of the coins can be related to graves, the majority of the coins are not related to graves and, thus, must relate more to the nuns’ general life in their choir. We can then return to the question of what use the contemplative Benedictine nuns had for money in their liturgical life. Why would contemplative nuns carry coins when they were praying? Or, indeed, why would nuns carry coins at all when they were not allowed to enter the secular society beyond the walls of the nunnery? And how did the nuns acquire the coins? In principle, no one but the nuns themselves had access to the nuns’ choir.

220  Alf Tore Hommedal The nuns’ obligations and monasterial rules and regulations would have involved a life in solitude separated from everyday concerns. Yet, even Jesus Christ used coins and was familiar with monetary values, especially when used for good deeds and devotional purposes: while Christ was scathing of the offerings of the rich and the money changers at the temple in Jerusalem, he was full of admiration for the poor old widow who offered her last copper prutahs (the lowest denomination coins at the time).98 The subject of coins and the Bible has produced a rich academic literature. A recent contribution – a monograph published by Lucia Travaini – deals with coins as relics, exploring the coins that allegedly formed part of the “thirty silver pieces” paid to Judas for betraying Christ in Gethsemane that are celebrated as relics as far north as Uppsala in Sweden.99 Studies of perceptions of money in religious societies have provided stories of complexity and pragmatism that do not suggest that coins would have been prohibited items of value and devotion for any of the nuns at Nunnusetr. On the other hand, over the long history of the nunnery, repair work would at times have been necessary. Could the coins have been lost by builders? Or could they have left them there intentionally, knowing that they would remain in a holy room that would be filled with prayer every day? This would seem not to be the case, even though the high number of bracteates from the time of Håkon IV’s Håkonsson seems to have been in circulation in the period when St Mary’s was being extended. If builders accounted for the coins, the coins would probably date only from more limited periods in the church’s history, when building work was being carried out. The coins are few and are distributed over a period of several hundred years, so they would seem to reflect occasional rituals involving money being carried out in the chancel, rather than the general, daily activities of the nuns. We might imagine that the coin material reflects the liturgy for specific dates in the ecclesiastical year. An example could be All Souls’ Day, the prayer day for the departed, on 2 November. The first reading of one of the Masses for this day tells the story of the money collected as a gift for the temple in Jerusalem, as an intercession for the dead, and the already mentioned poor old widow who offers her last copper prutahs. This kind of reading from the Holy Scripture could easily have formed the basis for rituals. Perhaps this was one of the few days of the year when the Bergen nuns brought with them coins to offer the church in imitation of the poor widow, and perhaps – on rare occasions – one of the coins was left behind, intentionally or unintentionally. In the Benedictine period of c. 330 years, this would have happened only once every decade according to the number of archaeologically documented coins; in the second half of the thirteenth century, it would have happened more often, but hardly more than once every three years or so. The number of coin finds from the nuns’ choir is thus not large, considering the long period of the Benedictine nunnery’s tenure. A comparison between the numbers of coin finds in St Mary’s and in parish churches in Norway suggests that monetary activity within the monastery church was limited. However, it is a fact that urban churches produce much fewer coin finds than the rural ones. The explanation is, in general, maybe to be found in the nature of the flooring, that there more often were stone floors in urban churches than in rural churches? In general, stone floors produce fewer finds than wooden floors. Excavations in churches in areas connected to Bergen, such as the stone church in Kinsarvik in Hardanger and the stave churches at Hopperstad and Kaupanger in Sogn, all with wooden floor, have revealed respectively 425, 597 and 1,456 coins.100 In a comparative perspective, some 95% of the coin finds from churches derive from rural societies, so the large number of coins found in urban churches, also in Bergen, is rare.

Coins and monastic liturgy – middle ages  221 101

According to Bendixen, many of the coins found in St Mary’s were worn out and the bracteates were often bent, which may indicate that the coins had been in use over some time and were already old when left or lost in the nunnery church. These qualities can be observed in the material today. The wear and tear could indicate that the coins were used for a long period, as they would have been if, for example, the Håkon IV bracteates were in use until the 1260s/1270s when the church’s second building phase was completed. It seems that at least some of the nuns did at certain times carry money in St Mary’s when partaking in the Divine Office. The archaeological evidence suggests that this was not the situation in the first century of the Benedictine nunnery’s existence. However, a change seems to have happened in St Mary’s in the mid-thirteenth century, after which the nuns occasionally carried money during the Divine Office. This may reflect changes in the liturgy or in the nuns’ contact with the outside world. It may also reflect changes in the fabric of the building, for example the installation of a different type of floor in the new nuns’ choir (perhaps partly wooden, at least under the stools) compared to the original nuns’ choir (perhaps made of slabs?).

The original and the new nuns’ choirs and the coins To summarise, we must assume that the extension of St Mary’s started during the 1260s or even in the 1250s and was completed c. 1270 or in the 1270s. Then, 1280, or most probably some years before, is the upper dating of the functional period of the original nuns’ choir (cf. Figure 8.9A) and the lower dating of the functional period of the new nuns’ choir (cf. Figure 8.9B or C). The original church and the coins These functional periods are important in the discussion of the coins.102 When it comes to the original church, we generally know that less than a quarter of the coin material originated from this part of the excavated area, i.e. less than 13 coins.103 The functional periods indicate that some of the coins found within the original church may have in fact originated from the old nuns’ choir (cf. Figure 8.9A) if they are from before the 1270s. If they are from the fourteenth century, or from before c. 1455, they still may be from the nuns’ choir, but from the western part of the new nuns’ choir, which jutted into the original church area (cf. Figure 8.9B). However, since we do not know the context of the c. 13 coins within the original church area, they may even have originated from the functional nave of the church, i.e. to the west of the original and extended nuns’ choir (cf. Figure 8.9A–C). It is not possible to tell because of limited archaeological documentation. Only three coins in the University Museum of Bergen’s collection can be identified as having been found within the area of the original church. The first, an Eirik Magnusson penny104 minted c. 1280–1285, was found in a stone grave close to the south wall (cf. Figures 8.1 and 8.2). The grave seems then to be from after the extension of St Mary’s and to have been located in a part of the church then most likely functioning as the nave. The two other coins were found somewhere within the area of the original church. Both are from the fifteenth century, and only an Erik of Pomerania105 hohlpfennig (1400–1420) relates to the Benedictine period. It is then interesting to see that none of the thirteenth-century coins from Nunnusetr can be documented to the area of the original church. However, we must assume

222  Alf Tore Hommedal that the only twelfth-century coin from Nunnusetr, the English sterling minted c. 1194–1200 AD, was located there (Appendix).106 Even though the wear and tear of the sterling makes it likely that the coin had been in circulation for a long period, we must assume that it was left in St Mary’s before the extended church was in use (c. 1260s/1270s). In addition, at least 12 coins without any context were minted in “the time of King Håkon IV Håkonsson” or for King Magnus the Law Mender (see Appendix), i.e. in the period when the extension seems to have been carried out. As a result of this lack of archaeological records, with the exception of the English sterling MA 348/18 (3), we are not able to decide where these coins were found within St Mary’s church. The nuns’ new choir and a hoard of thirteenth-century coins We know that more than three-quarters of the total coin finds originated from the new nuns’ choir in the extended part of St Mary’s. Within these coin finds, a cluster of 11 coins have been recorded. The chronological distribution provides evidence for eight bracteates issued in the time of King Håkon IV Håkonsson (reigned 1217–63) – five reliable and three probable – one farthing from Magnus the Law Mender (reigned 1263–80) and two unidentified bracteates.107 That this group of coins reflects a small hoard has never been suggested, and with a terminus post quem in the late 1270s, it would represent the only known hoard recorded from the reign of Magnus the Law Mender.108 That the new nuns’ choir may have been consecrated and fully in use from the 1270s or even before is interesting when it comes to delimiting the time of circulation of these Håkon IV bracteates. According to Gullbekk, the Håkon IV coins are assumed to have been valid for only a limited period after the King’s death in 1263.109 Three other coins from the time of Håkon IV are contextualised to the new nuns’ choir. Altogether 20 of the total of 43 identified coins found in 1891 may be from King Håkon IV’s time (see Appendix), suggesting that some of the non-contextualised coins must also have originated from the nuns’ new choir. There seems then to have been a rather intense period of deposition of bracteates in St Mary’s, either in the period when the extension was being built (building phase 2) or in the first functional decade of the nuns’ new choir. Two of the contextual coins from “the time of King Håkon IV Håkonsson” are related to graves.110 It is then a possibility that the graves originated from before the extension of the church, supposedly begun in the 1260s, and thus relate to a part of the churchyard to the east of the original church that was incorporated into the new nuns’ choir during the extension. However, this seems unlikely since at least one of the graves with a coin, according to Bendixen’s description,111 seems to have been centralised and adjusted to the structure of the extended church. The cluster MA 348/37 (Appendix) is described as having been found “by a grave,” indicating that the hoard was a part of the grave, even though this was not clearly documented. Even this grave may originate from the churchyard before the church extension. Again, this might seem improbable, but it is not impossible. A similar case is known from Nidaros Cathedral, where a hoard of c. 600 bracteates issued in the time of King Sverre Sigurdsson (reigned 1177–1202) was found together with a gold finger ring in a grave in the south-western corner of the church. The burial has been interpreted as a Bishop’s grave, but the grave’s supposed whereabouts inside the

Coins and monastic liturgy – middle ages  223 cathedral is false, because the coins must have been placed in the ground well before that part of the cathedral was built.112 If the grave at St Mary’s reflects a similar situation, the church’s second building phase must have started after the coins were deposited, meaning a numismatic terminus post quem in the late 1270s before the building work on the extension of St Mary’s commenced. Such a late dating for the start on the church’s extension seems, however, unlikely. A numismatic terminus post quem of c. 1275 for the hoard also seems to exclude the possibility that the coin cluster (MA 348/37) could relate to the building site. A purse holding the coins might well have been in the ownership of a mason or another person related to the building process.113 A purse could also have been lost by the abbess or other nuns overseeing the building process and handling cash for a wide range of purposes, including paying wages. However, a purse from the late 1270s or shortly after, containing a fresh quarter penny minted 1275–80, seems to be too late in relation to the dating of the extension of the church. The new nuns’ choir was then most likely in use. The hoard was then most probably contextually located inside the nuns’ new choir and was probably not lost accidentally. It was then either hidden, with full consciousness of the location, as a foundation offering or sacrifice in more general terms, by either the nuns or the builders. This leads us to the thought that the purse or cluster of coins, and even other thirteenth-century single coins, may have been left intentionally in the closing hours of the building work on the nuns’ choir, a room which people knew was intended for everlasting prayer. The builders or others visiting the nearly finished church extension could have left the cluster of coins as a sacrificial offering, perhaps even a foundation offering.114 This is a possibility if the extension of the church (the second building phase) was finished in the late 1270s, but as to its probability, we will never know the answer with certainty. We must, however, be aware of the most probable explanation: that a cluster of coins was left in a new grave or by the nuns themselves in the first decades of use of the new nuns’ choir. If we consider the numismatic evidence, a fair number of coins were found in this new and extended section of the church. Many of these were originally minted during the reign of Håkon IV Håkonsson, who died in 1263. The numismatic and archaeological evidence from the period suggests that a recoinage was carried out in the Norwegian realm shortly after King Magnus was established on the throne; in other words, the bracteates in circulation were replaced by new bracteates issued by the new king, a prudent symbol of a new reign and a new king.115 In all cases, the cluster of coins documents the fact that the bracteates from Håkon IV’s time were still in use after 1275. If we are to consider the numismatic evidence seriously, the question is: how long a period could the bracteates minted before 1263 have been in use? We have several hoards from the 1280s that all provide the same pattern: only pennies and a few halfpennies and farthings minted at the earliest at the very end of the 1270s and the 1280s. The older coins seem to have been replaced. This would, then, support the theory that the new nuns’ choir was in use already by 1270 or at least in the 1270s, even though a pretty high number, 19 or more, of the bracteates from the time of Håkon IV Håkonsson must then have been deposited in the new nuns’ choir within a decade. Maybe this high number could even indicate that the extension of St Mary’s in fact took place even earlier, in the 1260s or 1250s, giving a longer period in which the Håkon IV Håkonsson bracteates could have been deposited. The architectural traces exclude a building period earlier than c. 1250.116

224  Alf Tore Hommedal However, it is also commonly known that coin hoards usually contain the best coins available at the time, and those would not have been bracteates. Then again, accounts for collections of St Peters Pence from the Bishopric of Bergen in the years 1294–1300 include a sum of bracteates even though they would not have been valid currency at the time. We know from other sources that people sometimes used obsolete coins when paying dues,117 although it is difficult to know how long coins were considered valid in the second half of the thirteenth century. The use of old obsolete currencies for church offerings was recorded, for example at Visby on Gotland in 1424, where the town council outlawed the use of poor coinage for offerings: “Many women make offerings using Danish copper pennies that are no longer valid currency anywhere, and this seems like an unfair custom, therefore all women should be restrained from offering such coins.”118 We do not know whether this was a common practice among the nuns or others at the Nunnusetr, but it was certainly a possibility, especially if the coins, as discussed, had only a ceremonial use as part of the liturgy.

Coins and graves The coins from the nuns’ new choir documented to graves are absolutely related to the nuns’ everyday Divine Office, but not directly to the nuns’ everyday lives. In addition to the hoard, coins were found in three graves in the nuns’ new choir,119 some of them already mentioned. The only grave with a coin in the nave documented from the Benedictine period has also already been mentioned.120 Can the coins found in the graves give any indication that they were placed there intentionally? The Håkon IV bracteates MA 348/19 (1) found in a stone grave (X) could again indicate that the grave is older than the church extension. However, it seems plausible, at least in this concrete case, that the coin ended up here by secondary digging into the grave since it was found together with a younger jeton (MA 348/19 (2)). Of the c. 100 graves excavated within the church of St Mary’s in 1872 and 1891, only 17 graves were located to the new nuns’ choir. Only four of the graves have documented coins, three in the new nuns’ choir and one in the nave or original part of the church.121 The nuns’ new choir thus contained only around 17% of the total graves, but more than 75% of the coins from Nunnusetr. Even though these distributions may partly reflect nineteenth-century archaeological excavation methods, the available data probably reflect a reality: that most of the graves at St Mary’s were without coins. This, again, was a common feature of burials in medieval churches and medieval churchyards. The custom of placing one or several coins in burials with the deceased was originally a pre-Christian custom, but one that survived and was adopted occasionally in the Middle Ages.122 The most famous case is the case of St Francis, who has a prominent place in the saga of medieval monetary history for detesting money as an evil medium. When his grave was opened and surveyed in 1818, 12 coins were discovered next to the skeleton alongside a silver ring with an ancient gemstone.123 The number of coins points in the direction of the number of disciples, and we could, of course, speculate regarding the group of 11 coins found in relation to the grave discussed above (Appendix, MA 348/37 (1–11)). If the original number was supposed to have been 12, as in the grave of St Francis, and one was missed by the excavators, who might the buried person have been?

Coins and monastic liturgy – middle ages  225

The nuns’ choir: only for pious prayer or even for sewing? Among the archaeological small finds from 1891 were a bronze stylus indicating writing,124 some metal needles, thimbles and a spindle whorl and even fine gold thread indicating textile work, and a few gaming pieces indicating pastime activities.125 These artefacts beg the question: was the nuns’ choir sometimes used as a place for other activities, for example sewing during the Divine Office? This chapter will not discuss the question in detail. However, there are some indications of the answer. The archaeological context of the Nunnusetr items, combined with a morphological comparison to contextualised archaeological small finds from the Bergen town area (Bryggen) and from other Scandinavian Benedictine nunneries, provides an opportunity to date many of the Nunnusetr items. We are able to indicate that some of the Benedictine nuns in Bergen – to a certain degree, at least during the period of the fourteenth or fifteenth century – also performed textile work (sewing and spinning) when seated in their prayer stalls. Both the stylus and the gaming pieces seem also to be from the Benedictine period, indicating that the nuns on some occasions even played games in their stalls.126

Figure 8.16 A bronze stylus in the shape of a hand or a glove found in the extended part of the church (Bendixen’s “chancel”), i.e. the nuns’ choir (MA 348/b). This is the only complete stylus preserved from medieval Bergen. During excavations of the main settle area of Bergen (Bryggen, Figure 8.3), a tabula where such a stylus would have been used was found. The tabula mention “a runic letter to the sister of Olaf Hettusvein. She is in the nunnery in Bergen”, i.e. Nunnusetr (Ommundsen, “Å skrive med stil”, 2011, pp. 53–55). Photo: S. Skare, University Museum of Bergen

226  Alf Tore Hommedal However, we must be aware that sewing activities may also have been performed in the nave, i.e. the laypeople’s part of St Mary’s. The archaeological context of, for example, the thimbles is not documented, and they may have been related to the functional nave. Sewing activity could have been performed, for instance, by familiares living – not always of their own free will – in the nunnery, but not as nuns. We find an example of such a case in Lady Ragna Martinsdatter of Jåstad, the divorced wife of Lord Torolv of Idse, who had become monk in the Cistercian abbey of Lyse, south of Bergen. In October 1309, Lady Ragna was instructed by the Bishop of Bergen to make a lifelong vow of chastity, which she did before February 1310. She then also vowed to take residence in Nunnusetr in Bergen and live there at her own cost. In return, the Bishop promised not to force Lady Ragna to become a nun.127 In Nunnusetr, Ragna would have been exempted from the Divine Office in the nuns’ choir, and probably sat in the lay part of the church when following the church services. Given her situation, she may well have been more comfortable sewing than praying during mass. Living in Nunnusetr on her own economy, Lady Ragna may also have kept a purse about her, which she may occasionally have used for devotional purposes during her many hours in St Mary’s.

Some conclusions Nunnusetr in Bergen was the largest and wealthiest of the Norwegian nunneries until the fifteenth century, with 35 sworn nuns in 1320. At that time, altogether between 40 and 50 women lived in the nunnery, including the novices and prospective lay sisters. In 1320, and again in 1326, the newly elected abbesses were ceremoniously led from their stools in the nuns’ choir by two canons, acting on behalf of the Bishop, to be enthroned at the abbess’s seat. This chapter has among other things tried to provide an insight into that nuns’ choir. The number of coins found at the site of St Mary’s is not large considering that it functioned as a Benedictine nunnery for c. 330 years. Conclusions about the material are necessarily relatively vague. However, we must assume that the excavated and sampled coin material is representative and that the coins reflect the total existing coin material at St Mary’s. Even though it cannot be guaranteed, it seems a distinct possibility that none of the coins from the Benedictine period can be related to the functional nave part in the church and that the majority of the coin finds from this period can indeed be related to the nuns’ choir. The coins are few in number and are distributed over a period of several hundred years; however, they seem to indicate occasional rituals involving money being carried out in the chancel, rather than the general, daily activities of the nuns. We may imagine that the coin material reflects the liturgy for specific dates in the ecclesiastical year. We can therefore infer that at least some of the nuns, though they belonged to a contemplative order, carried money, even when taking part in the Divine Office. The material also indicates that this was not the situation in the first century of the Benedictine nunnery’s existence and that the carrying of money was a new custom in St Mary’s from the mid-thirteenth century onwards. Whether the nuns’ coins were lost, or whether they were liturgical offerings, is more difficult to determine. If we compare the coin material from St Mary’s with that from other contemporary nunneries, a difference emerges: it seems rather unusual to find coins in nunnery churches. Either that or the question has not been addressed and

Coins and monastic liturgy – middle ages  227

Figure 8.17 The Benedictine Abbess Kunigunde and her nuns. Bohemian drawing from the period 1312–1321. From the illuminated manuscript The Passional of Abbess Kunigunde in the National Library of the Czech Republic. Photography: WikiCommons, PublicDomain

discussed in particular. We have therefore tried to examine the situation as regards St Mary’s Benedictine nunnery in Bergen where some of the other archaeological small finds mentioned above may reflect other activities taking place in the nuns’ choir, such as textile work, writing and game playing, even when the nuns were seated in the stalls intended for praying. And it may be that even some of the coins were lost in connection with these more worldly activities. The archaeological small finds from some hectic weeks of excavation undertaken in the summer of 1891 thus give us a narrow but fascinating glimpse into the lives of contemplative women behind the closed doors of a medieval nunnery in Bergen.

Appendix Nunnusetr. The 44 coins (43 coins found 1891 and the coin found 2006) identified to be from the Benedictine period (c. 1120–1455); the 6 coins identified from the post-Benedictine monastic period (c. 1455–1528); 1 hohlpfennig with no identification; the 3 coins identified from the Manor period (1528–1891); and 5 unidentified, possible coins.

228  Alf Tore Hommedal Table 8.5 List of coins found in Nonneseter monastery Nationality and coin/minter

Dating

Number

Contexta

Museums’ number

England: Richard I or John Norway: time of Håkon IV Håkonsson, bracteates, reliable

1194–c. 1200 c. 1220–1263

1 9

Chancel Chancel, grave Churchyard?b Chancel, by a grave Chancel, by a grave Chancel, by a grave Chancel, by a grave Chancel, by a grave

MA 348/18 (3) MA 348/11 MA 348/19 (1) MA 348/22 MA 348/37 (3) MA 348/37 (4) MA 348/37 (5) MA 348/37 (8) MA 348/37 (10) MA 348/45 (1)

Norway: time of Håkon IV Håkonsson, bracteates, probable

10 (or more, jfr, MA 348/28 (4))

Chancel, grave Chancel, by a grave Chancel, by a grave Chancel, by a grave

MA 348/18 (1) MA 348/18 (2) MA 348/24 MA 348/25 (1) MA 348/25 (2) MA 348/28 (2) MA 348/28 (4) MA 348/37 (6) MA 348/37 (9) MA 348/37 (11)

Norway: Håkon IV or Magnus the Law Mender, bracteate

1

MA 348/39 (3)

Norway: Magnus the Law Mender, bracteates

1263–80

3

MA 348/15 MA 348/17 (3) MA 348/25 (3)

Norway: Magnus the Law Mender, penny

c. 1275–1280

1

MA 348/20 (2)

Norway: Magnus the Law Mender, farthing

c. 1275–1280

1

England: Edward I/II/III, sterling

1272–1377

1

Norway, Tunsberg: Eirik Magnusson, penny

c. 1280–1285

1

Norway: Eirik Magnusson, halfpenny

c. 1285–1290

1

MA 348/23

Norway: Eirik Magnusson, penny

c. 1295–1299

1

MA 348/20 (3)

Norway: Håkon V, halfpenny

1299–1319

1

MA 348/20 (4)

Norway (?): bracteate

c. 1350–1400

1

MA 348/39 (2)

Norway: time of Håkon VI, hohlpfennig

1355–1380

1

MA 348/28 (5)

Sweden (?): Albrecht of Mecklenburg (?)

1363–1389

1

MA 348/18 (4)

Chancel, by a grave

MA 348/37 (7) MA 348/39 (1)

Nave, grave (III)c

MA 348/12

Coins and monastic liturgy – middle ages  229 Hohlpfennig (?), fragment

Late fourteenth 1 century or early fifteenth century

MA 348/18 (5)

Denmark: Erik of Pomerania, hohlpfennigs

c. 1400–1420

7

Denmark, Lund: Erik of Pomerania, gross

1430s

1

Germany, Lübeck: viertelwitten

After the period  1403–1406

1

East of church, secondary context, the manors garden, found in 2006

BRM 650/14 Reinsnos, Nonneseter kloster, 33.

Mecklenburg: hohlpfennig

1450–1500

2

Nave Chancel, grave

MA 348/26 (2) MA 348/42

Denmark, Copenhagen (?): Hans

1481–1513

1

MA 348/8

Norway: Hans, hvid

1483–1513

2

MA 348/10 (1) MA 348/20 (1)

Norway: Erik Valkendorf, hohlpfennig

1510–1522

1

MA 348/21 (1) Bendixen, Nonneseter, 8.

Norway, Oslo: Christian II, shilling

1513–1523

1

Hohlpfennig (?)

?

1

Denmark: Christian V, shilling

1670–1699 1676

2

Sweden: Ulrika Eleonora, copper coin, lostd

1719–1720

1

Unidentified, coins?

?

5

Nave Chancel, grave

MA 348/13 MA 348/16 MA 348/17 (1) MA 348/17 (2) MA 348/26 (1) MA 348/45 (2) MA 348/52 MA 348/10 (2)

Coins identified as being from after the Benedictine period ending c. 1455

The cloister (south of the church)

MA 348/14 MA 348/28 (1)

“Outside the church”

MA 348/9 MA 348/27 Bendixen, “Nonneseter,” 7.

Chancel, by a grave Chancel, by a grave

MA 348/21 (2) MA 348/28 (3) MA 348/31 MA 348/37 (1) MA 348/37 (2)

a Context according to Bendixen, Nonneseter, or information on paper bags in the museum’s collection, MA 348. b According to information on the paper bags in the Museum’s collection, the coin was found “outside the church” (“utenfor kirken”), probably in the churchyard. c According to information on the paper bag in the Museum’s collection, the coin was “found in a stone grave no. IV” (“stensatt Grav IV”). Most likely this is a lapse, and the correct number is III, since there is only one coin with this identification (Eirik Magnusson, c. 1280–1285, Tunsberg), and this coin, according to Bendixen, Nonneseter, 8, was found in stone grave III in the nave. d Bendixen, Nonneseter, 7, mentions a Swedish copper coin from Ulrika Eleonora (1719–1720), now missing.

230 Alf Tore Hommedal

Notes 1 Bendix E. Bendixen, Nonneseter klosterruiner (Bergen: Foreningen til norske Fortidsmindesmærkers abevaring, 1893). 2 Nunnusetr is also used for the Benedictine nunnery in Oslo (Hilde Inntjore, Nonneseter – et middelalderkloster i Oslo). Master’s thesis (hovedfagsoppgave) in history, the University of Oslo (Oslo, 1998). The name must therefore be used in connection with the name of the town to identify which of the nunneries one has in mind. 3 Table I, appendix, Museum accession no. (hereafter MA) 348/18 (3): English sterling from 1194 to c. 1200. 4 There have been discussions of whether the extended church had an apse (Hans-Emil Lidén and Ellen Marie Magerøy, Norges kirker: Bergen, vol. I [Oslo, 1980], 166). However, the extended church also seems to have had a rectangular shape (Alf Tore Hommedal, “Bergens andre Mariakyrkje. Om utforminga av og livet i Nonneseter klosterkyrkje,” Årbok for Foreningen til norske fortidsminnesmerkers bevaring (2020), Forthcoming). 5 Lidén and Magerøy, Norges kirker, vol. I, 166–67; H.-E. Lidén and E.M. Magerøy, Norges kirker: Bergen, vol. III (Oslo, 1990), 91; A.R. Dunlop, Arkeologiske undersøkelser ved Vincens Lunges gate 19/20, Nonneseterkvartalet i Bergen 1997, Norsk institutt for kulturminneforskning no. 57 (Oslo, 1998); Alexander Rory Dunlop, Arkeologiske undersøkelser ved Vincens Lunges gate 19/21, Nonneseterkvartalet i Bergen 1998, Norsk institutt for kulturminneforskning no. 84 (Oslo, 1999), 26; Ambjørg Reinsnos, Nonneseter kloster, Bergen. Utgraving framfor Klosterkapellet, 2006, Norsk institutt for kulturminneforskning (NIKU) oppdragsrapport, no. 29 (Oslo, 2009); Alf Tore Hommedal, “Monks, Nuns, Canons and Friars in Medieval Bergen,” in Lübecker Kolloquium zur Stadtarchäologie im Hanseraum IX: Die Klöster, edited by Manfred Gläser, Manfred Schneider, Claudia Kimminus-Schneider (Lübeck: Schmidt-Römhild, 2014), 621–23. 6 No small finds seem to have been kept from the 1872 excavation and even the architectural stone fragments found were re-deposited and not re-found later (Bendixen, Nonneseter, 5). In an excavation in 2006 (Reinsnos, Nonneseter kloster, 33–4), one coin was found just to the east of the church (see Appendix, BRM 650/14). This was the only coin added to the 1891 finds. 7 The excavation was conducted “med saa meget hurtighed, som man af hensyn til arbeidets art turde, da eierne fordrede grunden ryddiggjort snarest mulig. Man udgrov hele klosterkirkens skib og en større del af koret, samt strækningen af den gamle kirkegaard paa nordøstre side.” One found “en mængde begravelse” and “Spredt omkring i gruset laa en del middelalderske mynter og en del smaating; naale og beslagstykker af bronse, tøistumper m.m.” (Foreningen til norske Fortidsmindesmærkers bevaring, Aarsberetning 1891, signed by the board member of the society B.E. Bendixen, I. Ross, F. von der Lippe, A. Fischer og W.M. Schjelderup (Kristiania, 1892), 165–66). 8 According to Bendixen, Nonneseter, 3–4, the area of the west tower had been dug out in the seventeenth century. For a period in the nineteenth century, coins were stored in this lower part of the tower, but these coins were post-medieval and were also separated from the former nave area by masonry filling the former opening between the tower and nave. The tower coins were thus clearly separated from the coin finds in 1891. 9 Aarsberetning 1891, 165–66. 10 Foreningen til Norske Fortidsminnesmerkers bevaring (Fortidsminneforeningen). 11 Håkon Schetelig and B.E. Bendixen, Foreningen til norske Fortidsmindesmærkers bevaring, Aarsberetning 1917 (1918a): V–IX; H. Shetelig and B.E. Bendixen, “Bergens Museum,” Aarsberetning 1917–1918 (1918b): 6–8. 12 Unfortunately, the survey in Bendixen, Nonneseter, is not related to the present Museum accession number for the small finds (MA 348). This means that any identification and correlation between the description in Bendixen, Nonneseter, and the objects under MA 348 in the Museum collection had to be re-constructed as a part of the present work. For most of the objects, the correlation is absolutely reliable, e.g. where the small finds are drawn as figures in Bendixen, Nonneseter, and therefore identifiable. In other cases, the correlation can be constructively argued, as will be pointed out in the actual cases in this chapter. Originally, the catalogue number MA 348 only included some of the

Coins and monastic liturgy – middle ages

13 14 15 16 17 18 19 20 21 22 23

24

25 26

27 28

29 30 31 32 33

34 35

231

non-numismatic objects catalogued MA 348/a–x. Later on, the coins and more of the other small finds were added to the catalogue, but then as MA 348/1–73, explaining why the sub-numbers of MA 348 consist of both letters and numbers. “fundet i koret.” Aarsberetning 1891, 165. See Appendix, MA 348/11 and MA 348/17(3). Bendixen, Nonneseter, 3. At present, we do not know which specific parts of the building burned down, but this should be possible to research further. Aarsberetning 1891, 165. The measurement was done at “Lungegaardshospitalets målestasjon” by Norwegian Meteorological Institute and then quite close to the excavation site. Bendixen, Nonneseter, 7. For the discussion on the foundation of Nunnusetr, see Åslaug Ommundsen, “To kongar, to dronningar og eit nonnekloster i Bergen,” (Norsk) Historisk Tidsskrift 95 (2016): 7–33. St Michael’s Abbey at Nordnes in Bergen was in Old Norse normally called Munkalif, meaning “the monks’ way of life” (Hommedal, “Monk, nuns, canons,” 616). Å. Ommundsen, “Nonneseter i Bergen – eit benediktinarkloster,” (Norsk) Historisk tidsskrift 89 (2010): 547–71; Å. Ommundsen, “Benediktinsk klosterliv i Bergen,” in: Geir Atle Ersland and Øystein Hellesøe Brekke (eds.), Fragment frå fortida (Oslo, 2013a), 45–9; Å. Ommundsen, “To kongar, to dronningar.” H.-E. Lidén, “Nonneseter klosterruin i Bergen,” in Fremtid for fortiden. Fortidsvern i 125 år, edited by Bernt Heiberg, Arne Berg and Egil Sinding-Larsen (Oslo: Foreningen til norske fortidsminnesmerkers bevaring (1969), 94–99; Lidén and Magerøy, Norges kirker: Bergen, vol. III, 60, 68, 99; H.-E. Lidén, “From Lund or Directly from Speyer?) Import of Rhenish-Lombardic Impulses as Manifested in Some 12th Century Churches in Bergen,” in Romanesque Art in Scandinavia, edited by Ebbe Nyborg, Hannemarie Ravn Jensen and Søren Kaspersen. Copenhagen Papers in the History of Art, no. 12 (Hafnia, 2013), 23. The author wishes to thank the architectural historian Ole Egil Eide for sharing his knowledge on the dating of the extant tower part. Halvard Bjørkvik, “Klostergods og klosterdrift i Noreg i mellomalderen,” Collegium Medievale 8 (1995): 148, 155–56; Ommundsen, “Nonneseter i Bergen,” 565–66. Munkalif was probably functioning as a Bridgettine monastery from AD 1425 (Erik Gunnes, “The Foundation of the Brigittine Monastery of Munkeliv, and Its Struggle for Existence,” Collegium Medievale 3 (1990): 111–22; Hommedal, “Monk, nuns, canons,” 618). Hommedal, “Bergens andre Mariakyrkje.” Christian C.A. Lange, De norske Klostres Historie i Middelalderen (Christiania [Oslo]: Chr. Tønsbergs Forlag, 1856), 301–304 and 415; Knut Helle, Bergen bys historie, vol. I, Kongssete og kjøpstad. Fra opphavet til 1536 (Bergen-Oslo-Tromsø: Universitetsforlaget, 1983), 870. Hommedal, “Bergens andre Mariakyrkje.” Hommedal, “Bergens andre Mariakyrkje.” Ommundsen, “Nonneseter i Bergen,” 567f. Hans Krongaard Kristensen, Klostre i det middelalderlige Danmark. Jysk Arkæologiske Selskabs Skrifter 79 (Århus, 2013), 125. Kristensen, Klostre, 144–45; U. Müller, “Klöster in Schleswig,” in Gläser et al. Kolloquium zur Stadtarchäologie, 363–64; Matthias Untermann, “The Place of the Choir in Churches of Female Convents in the Medieval German Kingdom,” in Women in the Medieval Monastic World. Medieval Monastic Studies, vol. 1, edited by Janet Burton and Karen Stöber (Turnhout: Brepols, 2015), 329, 348. Untermann, “The Place of the Choir,” 329, 348. Roberta Gilchrist, Gender and Material Culture. The Archaeology of Religious Women (London and New York: Routledge, 1994), 97; Edgar Lein, Mittelalterliche Klöster in Deutschland, Österreich und der Schweiz. IMHOF-Kulturgeschichte (Petersberg, 2009), 39–40; Krongaard Kristensen, Klostre, 145.

232 Alf Tore Hommedal 36 37 38 39 40

41 42 43 44 45 46 47 48 49 50 51 52 53 54

55

56

Untermann, “The Place of the Choir,” 334. Bendixen, Nonneseter, 5; Lidén and Magerøy, Norges kirker: Bergen, vol. I, 164–67. Hommedal, “Bergens andre Mariakyrkje.” Gilchrist, Gender and Material Culture, 97–99. A.T. Hommedal, “Kva fortel bygningsrestane av dei norske klostra om kontinental norm og norsk praksis innan ordensliva?” in Norm og praksis i middelaldersamfunnet. Kulturtekster 14, edited by Else Mundal and Ingvild Øye (Bergen: Senter for europeiske kulturstudier, 1999), 172. Bendixen, Nonneseter, 5. The corresponding part of the tile floor’s edging band to the north was not preserved when excavated in 1891 (Figure 8.2), but we must envision a row of stalls along both sides of the floor. See Gullbekk et al. and Rensbro, this volume, chs. 1 and 3. Lidén and Magerøy, Norges kirker: Bergen, vol. III, 141–42. Thor Kielland, “Middelalderlige flisegulver i Norge,” in Foreningen til norske fortidsminnesmerkers bevaring, årbok 77 (1921), 144–56, discusses Norwegian tile floors including St Mary’s, but he does not give an exact dating for the tiles. Untermann, “The Place of the Choir,” 334. Lidén and Magerøy, Norges kirker: Bergen, vol. I, 166–67; Lidén and Magerøy, Norges kirker: Bergen, vol. III, 91. Lidén and Magerøy, Norges kirker: Bergen, vol. III, 91. DN IV, no. 3; Bendixen, Nonneseter, 5. Lidén and Magerøy, Norges kirker: Bergen, vol. I, 166. Lidén and Magerøy, Norges kirker: Bergen, vol. I, 166: “[Dette] bekrefter formodningen om at kortilbygget kan være fra siste fjerdedel av 1200-årene.” A supplementary discussion of the topic has been done in another article (Hommedal, “Bergens andre Mariakyrkje”), and only the main argumentation will be referred to here. Bendixen, Nonneseter, pl. IV; Lidén and Magerøy, Norges kirker: Bergen, vol. I, 166. Lidén and Magerøy, Norges kirker: Bergen, vol. III, 106. The author wishes to thank the architectural historian Ole Egil Eide for the discussion on this western part of the Franciscan church. The parish church as the Franciscans took over in the 1240s and adapted to their friary church also had a west tower, an alien substance to the Franciscans. It seems then for me only natural that the brethren, after the fire in 1248, intended to replace the tower with a western extension of the nave. Hommedal, “Bergens andre Mariakyrkje.” The author wishes to thank the architectural historian Ole Egil Eide for the information and discussion on the tracery fragment: the tracery window fragment may be from the third Church of the Apostles, consecrated in 1302. Lidén (Lidén and Magerøy, Norges kirker: Bergen, vol. I, 166) supports this dating: Etter arkitekturfragmentenes stilpreg å dømme, må forlengelsen av kirken ha skjedd en gang i 2. halvdel av 1200-årene/ Judged from the shape of the architectural fragments, the extension of the church have taken part during the 2. Half of the 1200s.

57 Hommedal, “Bergens andre Mariakyrkje.” 58 Gerhard Fischer, Domkirken i Stavanger. Kirkebygget i middelalderen (Oslo: Dreyer, 1964); Morten Stige, Stavangerkorets utvidelse og innflytelse. Unpublished Master’s thesis (hovedfagsoppgave) in art history, University of Oslo (Oslo, 1997). 59 H.-E. Lidén and E.M. Magerøy, Norges kirker: Bergen, vol. II (Oslo, 1983), 10. 60 Soga om Håkon Håkonsson. Norrøne Bokverk no. 22 (Oslo: Det norske samlaget, 1963), 338. 61 Helle, Bergen bys historie, 613. 62 According to the English–Scottish Chronicle of Lanercost, King Magnus’s firstborn son “resembled more the offspring of a bear than a man, as it were a formless lump of flesh.” See The Chronicle of Lanercost 1272–1346, tr. Sir Herbert Maxwell (Glasgow: J. Maclehose & Sons, 1913), 21–22. For a discussion of this information, see Hommedal, “Bergens andre Mariakyrkje.”

Coins and monastic liturgy – middle ages

233

63 “Magnus Haakonsønns saga,” Norges Kongesagaer 1914–utgaven (Kristiania [Oslo]: I.M. Stenersens Forlag, 1914), 301: “den sorg, som alle mænd i Norge bar, da junker Olav, søn til kong Magnus, døde.” 64 The high number of bracteates from the time of Håkon Håkonsson could theoretically indicate that the church was extended even earlier in the King’s reign (1217–1263). However, the architecture of St Mary’s seems not to be older than c. 1250 at its oldest. 65 One suggestion is that a new and larger treasury window in the east front, as seen in Stavanger Cathedral, was also a part of this last building phase in St Mary’s and that the fragment from such a window identified by Lidén at St Mary’s came from this phase; see note 55. 66 Bendixen, Nonneseter, 7. 67 “… for en stor del i fragmenter, hvad der gjør det vanskeligt at opgive antallet nøiaktig.” 68 MA 348 in the collection at the University Museum of Bergen records 66 coins or conceivable coins in all. According to Svein H. Gullbekk, six of the older registrations are not coins; one is token money (MA 348/29) and four may be coins, even though this cannot be fully clarified (MA 348/28(3); 31; 37(1); 37(2)). We then have 55 identifiable coins. 69 Claudius Jacob Schive, Norges Mynter i Middelalderen (Christiania: H. Tønsberg, 1865); Kolbjørn Skaare, “Norsk utmyntning på Håkon Håkonssons tid,” Nordisk Numismatisk Årsskrift (NNÅ) 1970: 5–36. 70 Svein H. Gullbekk and Anette Sættem, Norske myntfunn 1050–1319. Penger, kommunikasjon og fromhetskultur (Oslo: Dreyer forlag, 2019), 5–68. 71 The Norwegian coins return in the post-Benedictine period of Nunnusetr, however then also with German and Danish coins present. The three post-monastic coins are all foreign; see Table 4 and Appendix. 72 Gullbekk and Sættem, Norske myntfunn; Terje Masterud Hellan, Utenlandsk mynt i Norge, ca. 1350–1483. Unpublished Master’s thesis in archaeology, NTNU (Trondheim, 2012). 73 See, for example, Bendixen, Nonneseter, 7–8; Reinsnos, Nonneseter kloster, 33–34. 74 Bendixen, Nonneseter, 7. 75 Elisabet Regner, Den reformerade världen. Monastisk och materiell kultur i Alvastra kloster från medeltid till modern tid. Stockholm Studies in Archaeology 35 (Stockholm, 2005), 125. 76 Neither do any coins seem to have been found in the Heiligenkreuz chapel (in two stores). The coins at St Johann were excavated 1971–1995 and seem mostly to have appeared in the “Sudhof,” the south cloister and in the churchyard to the east of the cloister. The majority of the coins seem to have been from the eleventh century to the fifteenth century. The church seems not to have been excavated (José Diaz Tabernero, “Die Fundmünzen aus dem Kloster St. Johann,” in Müstair Kloster St. Johann 2. Münzen und Medaillen, edited by José Diaz Tabernero and Christian Hesse (Zürich: VDF Hochschulverlag, 2004), 16 and 30–35. 77 The original church probably had the same layout as the extended church, with the nuns’ choir stationed to the west of the presbytery (Hommedal, “Bergens andre Mariakyrkje”). 78 Gilchrist, Gender and Material Culture, 97; Untermann, “The Place of the Choir,” 345. 79 Lidén and Magerøy, Norges kirker: Bergen, vol. I, p. 166. 80 Hommedal, “Bergens andre Mariakyrkje.” 81 This, however, does not rule out the possibility that the existing foundations were reused for screens in some periods. If this was the situation, it was probably in the postBenedictine periods or in the last period of the Benedictine nunnery’s existence when there seems to have been few nuns at Nunnusetr. 82 Cross-screens have not been discussed thorough for Norwegian monastic churches, but some studies have been done on Norwegian parish churches; see Anne Marta Hoff, “The Area between Chancel and Nave in Norway’s Medieval Parish Churches,” Jaarboek voor liturgie-onderzoek 19 (2003): 147–74. 83 Gilchrist, Gender and Material Culture, 97 and 100. 84 The number of nuns may have been 36 including Brynhild Katharina if she, as the outgoing abbess, was not allowed to vote (Lange, De norske Klostres Historie, 117)

234

Alf Tore Hommedal

85 There is also a possibility that four or six nuns were seated along the east side of the choir screen, with their backs to the screen and facing the presbytery, creating a U-shaped row of stalls. Six stalls here, three on each side of a doorway in the screen, would fill c. 4.8 m of the total width of 8.4 m in the room. This would give a width of c. 3.6 m in the doorway for processions to pass through. Such a solution would make it possible to shorten the length of the choir area by c. 2.4 m. 86 DN I, 162; DN I, 185; DN III, 119; Lange, De norske Klostres Historie, 317–18; Helle, Bergen bys historie, 614. 87 1.25 m × 2 rows on each side × 2 sides. 88 If we include a U-shaped arrangement (see footnote 87) with some nuns sitting in two rows with their back to the nave and facing the presbytery, this would shorten the length of the choir area with c. 2.4 m. The width of the doorway in the choir screen could still be c. 3.6 m, allowing processions to pass by. 89 The re-furnishing may also have been done in post-Benedictine times. However, this does not seem likely due to the supposed age of the excavated tile floor. 90 Bernice Kerr, Religious Life for Women, c. 1100–c. 1350 (Oxford: Oxford University Press, 1999), 122; Lein, Mittelalterliche Klöster, 38; Benedict of Nursia, 2009, 79–88. 91 The English Latinist and researcher Michael Gullick (M. Gullick, “A Preliminary Account of the English Element in Book Acquisition and Production in Norway before 1225,” in Latin Manuscripts of Medieval Norway. Studies in Memory of Lilli Gjerløw, edited by Espen Karlsen [Oslo: The National Library, 2013], 103–21) has discussed and given some preliminary accounts of the English element in book acquisition and production in Norway according to manuscripts and documents of English origin or the work of English scribes in Norway datable to before 1225. None of these can be related to the Benedictine nuns in Nunnusetr. However, Gullick (“A Preliminary Account,” 106, 111) has established that a gospel book written in England in the second quarter of the twelfth century was at Munkalif by the end of the century, and thus in a Benedictine liturgical setting in Bergen. A homily book written in or near Bergen by an English trained scribe c. 1200 (Gullick, “A Preliminary Account,” 112–13) suggests that the scribe was attached either to the Christ Church cathedral or to the St John’s Augustinian Abbey in Bergen. We see, then, glimpses of the English influence on the monastic liturgy in Bergen. Gullick (“A Preliminary Account,” 112–13) can also see a certain English influence in an Old Norse translation of the Rule of St Benedict probably written c. 1200 in Nidaros (Trondheim), whose originator was connected to the Benedictines. For a general account of English influence on the Norwegian Church, see Absalon Taranger, Den Angelsaksiske Kirkes Indflytelse paa den Norske (Kristiania [Oslo]: Den Norske Historiske Forening, 1890). The Norwegian Latinist and researcher Åslaug Ommundsen (Å. Ommundsen, “Psalms Interrupted. The Psalters Fragments in the NRA in Oslo,” in Latin Manuscripts of Medieval Norway. Studies in Memory of Lilli Gjerløw, edited by Espen Karlsen [Oslo, 2013b], 279–305) can also see an English influence in some fragments of psalters either written in England or by a scribe writing in an English style in Norway. 92 David Knowles, The Religious Orders in England Vol. I. The End of the Middle Ages (Cambridge: Cambridge University Press, 1962 [repr.]), 280; Kerr, Religious Life for Women, 122. 93 Lein, Mittelalterliche Klöster, 38–39. 94 Lein, Mittelalterliche Klöster, 36. 95 Kerr, Religious Life for Women, 123. 96 Kerr, Religious Life for Women. 97 Lange, De norske Klostres Historie, 318–19; Helle, Bergen bys historie, 674. 98 Mark 12:41–44. See Gullbekk et al., this volume, ch. 1. 99 Lucia Travaini, I Trenta denari di Giuda. Storia di reliquie impreviste nell’Europa medievale e moderna (Rome: Viella, 2020). 100 Gullbekk and Sættem, Norske myntfunn, nos. 258, 274 and 272. As mentioned, we must here make a certain reservation due to the differences in excavation methods. Hopperstad, Kinsarvik and Kaupanger were excavated in the late 1880s, 1960 and 1964, respectively. 101 Bendixen, Nonneseter, 8.

Coins and monastic liturgy – middle ages

235

102 The use of coins for dating church buildings and ecclesiastical building structures has so far mainly been related to stave churches, e.g. Peter Anker, “Dateringsproblemet i stavkirkeforskningen,” Historisk Tidsskrift (1977): 103–42; Roar Hauglid, “Om dateringsproblemet i stavkirkeforskningen,” Historisk Tidsskrift (1977): 433–42; and Roar Hauglid, “Om stavkirkers datering. Myntfunnenes betydning,” Viking (1983): 118–33, even though both Bendixen, Nonneseter, and Lidén (Lidén and Magerøy, Norges kirker: Bergen, vol. I, 166) discuss the coins in their dating of the extension of St Mary’s. 103 Less than a quarter of the coin material is less than 15 coins. However, since we know that at least two coins were found at the churchyard and cloister, less than 13 must have been found in the original church building. 104 Appendix, MA 348/12. 105 Appendix, MA 348/26 (1). The other one is a hohlpfennig from Mecklenburg, c. 1450– 1500 (Appendix, MA 348/26 (2)), and thus probably from the Bridgettine period in Nunnusetr. 106 Appendix, MA 348/18 (3). 107 Appendix, MA 348/37 (1–11). Besides the five reliable coins ( MA 348/37 (3–5, 8, 10)) and three probable coins (MA 348/37 (6, 9, 11)) from “the time of king Håkon IV Håkonsson,” the hoard also contained one coin from King Magnus the Law Mender minted 1275–1280 (MA 348/37 (7)) and two unidentified bracteates (MA 348/37 (2, 6)). 108 Gullbekk and Sættem, Norske myntfunn, no. 255, list this as a “Church find” but notify that 11 coins were recorded as a group in relation to a burial. The recorded hoards from this period are from Søndeled, Risør, Aust-Agder County, t.p.q. 1280, and Hornö, Svenneby, Bohuslän, t.p.q. 1280 (Gullbekk and Sættem, Norske myntfunn, nos 2 and 374). 109 Svein H. Gullbekk, “Renovatio monetae i Norge i middelalderen,” NNÅ 1992–93 (1994): 52–88. 110 Appendix, MA 348/19 (1); MA 348/24. 111 Bendixen, Nonneseter. 112 Risvaag, this volume, ch. 6; Linn Eikje Ramberg, Mynt er hva mynt gjør. En analyse av norske mynter fra 1100-tallet: produksjon, sirkulasjon og bruk (Stockholm: Stockholm Studies in Archaeology 73, 2017), 290, no. 10; Gullbekk and Sættem, Norske myntfunn, no. 294. One of the rare medieval coin hoards recorded from Bergen is 94 bracteates from the late twelfth century next to the skeleton of a woman in a mass grave, presumably a purse (Eikje Ramberg, Mynt er, 288–89, no. 4; Gullbekk and Sættem, Norske myntfunn, no. 249). 113 Thibault Cardon, “Les usages des monnaies (mi XIIe-début XVIe s.). Pour une approche archéologique, antropologique et historique des monnaies médiévale,” (Unpublished PhD diss., Paris: Ecole des Hautes Etudes en Sciences Sociales, 2016). 114 Jonsson, this volume, ch. 13. 115 Gullbekk, “Renovatio”; S.H. Gullbekk, Pengevesenets fremvekst og fall i Norge i middelalderen (København: Museum Tusculanums Forlag, 2009). 116 Hommedal, “Bergens andre Mariakyrkje.” 117 DN IV, 60; Gullbekk, Pengevesenets, 258. 118 Georg Galster, Unionstidens udmøntninger. Danmark og Norge 1397–1540. Sverige 1363–152 (Copenhagen: Dansk Numismatisk Forening, 1972), 20, no. 70a; Jørgen Steen Jensen, De skriftlige kilder til Danmarks middelalderlige møntvæsen, et udvalg 1085– 1500 (Copenhagen: National Museum, 1989), no. 670. 119 Appendix, MA 348/19 (1); MA 348/24; MA 348/52. A fourth grave, MA 348/42, seems to be from the post-Benedictine period. 120 Appendix, MA 348/12. 121 Bendixen, Nonneseter, 6 and 8. 122 For a discussion of coins in funerary contexts, see Lucia Travaini, “Saints and Sinners and  … a Cow: Offerings, Alms and Tokens of Money,” in Money and the Church in Medieval Europe, 1000–1200. Practice, Morality and Thought, edited by Giles E.M. Gasper and Svein H. Gullbekk (Farnham: Ashgate, 2015), 210–15. 123 Travaini, “Saints and Sinners,” 210–11, pl. 17. 124 Åslaug Ommundsen, “Å skrive med stil – stylus frå Nonneseter gjenoppdaga i museets magasin,” in Årbok for Universitetsmuseet i Bergen (Bergen, 2011), 51–55.

236 Alf Tore Hommedal 125 The small finds mentioned are all catalogued as MA 348 in the collections of the University Museum of Bergen. For analysis of similar finds, see Kilger, this volume, ch. 5. 126 In Lund cathedral in present-day Sweden, a lot of small finds have been found under the flooring of the secular canons’ choir stools, and this supplements our picture of the situation in the nuns’ choir in Bergen; see Mattias Karlsson, Konstruktionen av det helliga. Altarna i det medeltida Lunds stift, Skånsk senmedeltid och renässans 23, (Lund, 2015), 167–68. 127 DN, I, no. 123 and no. 126.

9 The archaeological landscape under church floors Coins and contexts in Aggersborg church, Jutland, Denmark Henriette Rensbro and Jens Christian Moesgaard Introduction This study provides an in-depth numismatic and archaeological study of archaeological finds, with an emphasis on coins, from the excavations of Aggersborg church in northern Jutland, Denmark. It aims to present the unique character of church floors as archaeological objects and discusses the value of coin finds in churches. As in all archaeology, the value of finds is determined by their preservation, context, methods of excavation and documentation. In churches, the historical redeposition of cultural layers during construction and rebuilding works is taken into account by archaeologists and numismatists but rarely in a fully systematic way. In a significant number of cases, the levelling of church floors was carried out using material from outside the building or from another place within the building; in some cases, earth was taken out of the church and dispersed elsewhere. In the Aggersborg church, the distinction between original and subsequent building events is especially important. The complexity of the archaeological situation and the challenge of interpreting coin finds from disturbed cultural layers are examined here. The distribution of the coin finds at Aggersborg will be discussed in light of inferred disturbances in the stratigraphy, and this underlines key methodological challenges faced in the study of church floors generally. The nature of church archaeology is complex, partly not only because of the longue durée and the re-levelling and extension of building structures but also because of the changes in devotional practice that attended an evolving liturgy and, last but not least, because of significant lacunae in the available information due to a lack of high-resolution data capture during excavation. We recommend an increased focus on precision in the archaeological execution and documentation of churches as a consequence of the material from Aggersborg church and experiences of Danish church archaeology in general.

Church archaeology in Denmark In present-day Denmark, 1,700 churches have been in continuous use since the ­Middle Ages, most of them since the twelfth century. The main reason for the good condition of these old buildings is their uninterrupted use as the primary location for Christian practices, coupled with the fact that churches in Denmark are owned and funded by the state. They provide excellent opportunities for the analysis of structural changes related to layout and the destruction of older traces of use and design, resulting from changes in religious practices and liturgy, creating new demands for space to accommodate shifting demographic, psychographic and behavioural needs.1 Lists of coins from all finds in churches from the last 70 years are available. These finds came from an estimated 400 plus churches, of which only some have been fully excavated.2 Distribution

238  Henriette Rensbro and Jens Christian Moesgaard

Figure 9.1 Coin list from Allested Church 1958. Plan of the division of the floor into numbered sections and a coin list, which refers to the sections. AntikvariskTopografisk Arkiv, The National Museum in Copenhagen.

The archaeological landscape under church floors  239

Figure 9.1  (Continued)

240  Henriette Rensbro and Jens Christian Moesgaard plans of coin finds have been made for fewer than 20 churches, and very few have been published. Most of these plans are divided into three or four parts, relating either to each window or vault module or to an even less accurately defined module.3 It is important to note that archaeologists have primarily used coin finds from churches for the dating of layers,4 while numismatists and historians have used them as sources for shedding light on the monetisation of the economy.5 This is perhaps the most important reason why many excavations have been executed without the use of grid-systems, and only a smaller number have used square metre grids. When coins from church floors are presented, they feature in most cases in lists that enumerate their quantity, dating, provenance and other numismatic information, and these studies are occasionally accompanied by floor-plan distribution maps. In general, lists are best suited for statistical analysis and comparative studies between different sites in a region, whereas distribution plans are best suited for spatial analysis, especially where large numbers of coins have been found and when there is adequate information on wider find contexts. However, it is a fact that lists are actually rarely used for statistics, whereas distribution plans for some reason have been used for the neat presentation of finds with few coins (approximately less than 40) and little information on find context. This is because the methods of documentation have often been highly selective and inadequate. Finds of non-numismatic portable artefacts are, in principle, registered in the same way as coins, but in many cases, they are registered with less precision.6 Distribution plans from the 1950s and 1960s were mainly made from excavations in churches prompted by construction work and were supervised by architects, albeit under the supervision of representatives from the National Museum of Denmark in Copenhagen (NMD), who introduced systematic sectioning and sieving methods. Unfortunately, the coin lists and plans of finds from this period provide little or no information regarding differences in excavation depth or data on stratigraphic levels from which Harris matrices could be constructed and which would provide a basic starting point for understanding and interpreting what happened through the history of the church. Most of the reported data is missing this dimension, and this leaves us with only the dates of the coins themselves as a rough guide to the chronology of the deposition history. The main result of these old distribution plans and coin lists is that fourteenth- and fifteenth-century coins are in the majority.7 Since 1970, there have been very few large-scale excavations in Danish churches, and distribution plans have gone out of use. There is no equivalent for Danish churches to Klackenberg’s study providing surveys of coin finds in Swedish churches.8 Perspectives and methods have changed considerably within church archaeology in Denmark in recent decades. In the 1950s and 1960s, scholars were engaged in developing the oldest structural layouts of churches9 and particularly identifying and rendering the possible wooden predecessors of stone churches.10 Excavations of the full extent of churches were often prompted by the installation of new heating systems and floor insulation.11 The most often registered features under the floors were the solid stone foundations for wall benches and altar tables.12 In the 1980s, interest developed in church interior furnishings, especially objects from the fourteenth century onwards. Furthermore, special attention was paid to finds of small fragments of baptismal fonts’ changing placements,13 the changes of floor-surface materials over time14 and experiments with scientific analysis of mortar and soil.15 In the past decade, the focus in Danish archaeology has increasingly turned towards appreciating churches as sacral places and understanding their role in medieval mentality and society, as well as, on a practical level, prising further micro-information out of the individual layers of flooring. The latter is probably an attempt to alleviate

The archaeological landscape under church floors  241 frustrations over the myriad coins and non-numismatic finds in the National Museum’s stores that are hardly usable due to the lack of contextual information and the limitations caused by the small excavation areas.16 It is to be expected that new questions will emerge in time and that contemporary archaeological methods will soon be considered insufficient.17 It is, however, noteworthy that the old idea of using church buildings as sources for economic and cultural changes, which flourished in Sweden in the 1980s and 1990s, is still highly relevant in Danish research today.18 The practice of church archaeology has developed from the use of contractors before the 1960s, through the introduction of archaeological authorities, to today, where professional archaeologists are in charge of the whole process from excavations to documentation and collection management.19 This is both due to an increasing focus on stratigraphy, layers, formation of the layers and find contexts over the last 30 years and the fact that the era of extensive church excavations has ended. 20 Today, excavations are mostly limited to small excavation areas of limited depth. There are rarely opportunities for new excavations encompassing overall distribution plans or data collection for statistical analyses/ modelling of new finds. Contemporary registration methods are, for this reason, designed to deal with trenches of limited size and they fall short in the very few cases where entire church interiors are excavated. Therefore, it is still common that finds are exclusively attributed to layer and not location (x, y). On this basis, distribution plans unfortunately cannot be made. Coins in churches For decades in Scandinavia, it has been debated whether coins found in church floors primarily represented random accidental losses or were votive offerings deliberately deposited in tamped-down earth on the floor of the church.21 In 1992, Henrik Klackenberg significantly furthered research on this dichotomy by introducing the term “offertory wastage” to indicate coins dropped accidentally in connection with their votive use in the church. Since this term has been introduced, it has become easier to use coin finds from churches as sources for hypotheses concerning a church’s interior and use. The question is then, how widely the term “offering” can be used. Coins might be lost in connection with devotional, sacrificial gifts to God and saints, regulated contributions for the maintenance of an altar, charity contributions for the poor, payment (fees) for church services (masses, baptisms, funerals etc.) or one-time contributions for crusades, among other things.22 Wangsgaard Jürgensen argues that these coins are offerings as they are tangible manifestations of faith and thankfulness.23 In medieval society, the ritual use of coins could serve many different purposes, from promoting religious ecstasy to protecting oneself against accidents, to the more secular payment of dues and tithes and, of course, “costly status signalling” to peers in the church and a wider community.24 Different behavioural strategies could distinguish between deliberate offerings and accidental “offertory wastage”, but distinguishing between these strategies by way of the archaeological evidence is of course very difficult to do with certainty. Before using coin finds from churches as sources for offerings, the development of religious practice, the church’s interior and coin circulation, we should both improve our contemporary archaeological methods and use old excavation results in the best possible way. It is necessary to analyse the information from old excavations and evaluate which questions they can or cannot answer. Furthermore, we must clarify what uncertainties are inherent in the registration methods used in older excavations. It is essential to identify which coins found in church floors were deposited there (knowingly

242  Henriette Rensbro and Jens Christian Moesgaard or unknowingly) and which were brought into the church with redeposited soil from outside (or from another place in the church), and from which time periods coin finds are missing because of the removal of dirt from the churches. Of course, this can best be done in future excavations, where the methods aim to answer contemporary questions about old churches and church coins. But such identifications can also be made (to a certain degree) for older excavations if there was a focus on stratigraphy and registering the context of finds. It is especially important that observations on the formation history of the layers, meaning whether they were formed there or brought there (as often happened with layers for levelling or raising the height of a floor), were registered. However, it must be emphasised that concerns about the redeposition and removal of dirt are mainly relevant to stone churches and churches that do not have wooden floors. For instance, in wooden churches from the High Middle Ages and onwards in Norway, the floor is often linked to the rest of the construction and is raised above the ground, which probably impeded burials and other disturbances beneath the floor to some extent. At the same time, this type of floor construction complicates the archaeologists’ chances of recognising redeposits of soil and finds that happened during burials or other construction work. There is no doubt that there is a strong correlation between the material of the floor surface, i.e., stone paving or wooden planks etc., and the number of coins that end up under the floor. But this has to be considered in relation to geographic differences in monetisation and offering as well as the construction, interior design and use of the individual church. We need to keep track of which coin finds represent coins that ended up in churches as a consequence of monetary use related to devotional practices or ecclesiastical economy on the one hand, and what coins may be missing as a consequence of earth being moved within or from outside the church – and thus were taken out of their original context – on the other hand. The most valuable contextual finds are those related to the interior and use of the church: those that have not been brought in by redeposited layers or taken out of the church, causing gaps in distribution plans. Only when we have identified those coins which were related to the interior and original use of the church can we successfully use the coin lists to distinguish original in situ depositions and geographic variations, and to differentiate between church coins and such outliers as detector-found coins in the near vicinity. Furthermore, by applying these methods, distribution plans become more reliable sources for inferring the possible positions of side altars and other aspects of the church’s use at different times. With all of the above in mind, it is vital to examine to what degree the state of preservation of the floor layers, and the possible redeposition and removal of dirt on the floors, affects the distribution patterns in each church. The overall picture of church coins should rightly be examined with comparative analyses of finds from many churches, but since a research programme on such a scale is beyond the scope of this paper, the best method of approach is to examine the specific case of the situation at a single church. Introducing a case study: Aggersborg church The excavation in Aggersborg church in 1976 provides us with a variety of reasons for concentrations of coins under church floors. The excavation, covering most of the interior of the building, yielded 341 coins, the largest archaeological corpus of coin finds in any Danish church to date. There are also a further 301 non-numismatic registration numbers covering an additional 699 artefacts or fragments found in the

The archaeological landscape under church floors  243

Figure 9.2 Aggersborg church is a large Romanesque building consisting of a chancel and nave of granite ashlar. Probably there was an apse at the east gable of the chancel. In the Late Middle Ages (fifteenth–sixteenth centuries), a tower of field stones and bricks in the west and a porch in front of the south door of the nave were added. The present porch was built in the same place in the nineteenth century. In the north side of the nave, the original small windows are preserved, and the entrance door is bricked up. Photo: Vesthimmerlands Museum, Denmark

church (registration numbers can cover one or more objects in a batch). The published data vary a great deal in quality: some reflect a very modern standard. Aggersborg church and the excavation in 1976 are introduced below. The results of the re-investigation performed as part of the Religion and Money project are presented in the next section, and we discuss and present alternative interpretations of three areas with clusters of coins in the church in the section entitled “Case study: Three micro-concentrations in Aggersborg church”.

Aggersborg and the 1976 excavation The site of Aggersborg is located at the north side of the Limfjorden waterway, just west of its narrowest part. 25 The economic importance of this particular site over the centuries should not be underestimated. It was perched on a vital crossroads for both land and sea routes. 26 It has been suggested that the king owned land on both sides of Limfjorden at this place in the Viking Age and early Middle Ages (eighth–twelfth centuries). 27 Around AD 970–80, a vast ring fortress, the largest of its kind, was placed at the north side of the fjord. It was abandoned some 20 years later. Despite considerable research efforts, burial sites associated with the ring fortress have never been found. 28 We know of the existence of a royal estate in the eleventh century, which may have been located on the south side of the water at a natural harbour near the village of Ørbæk, where coins were minted at that time.29 In the thirteenth century, a manor was built on the north side of Limfjorden, between the shore and the abandoned ring fortress.30 By the fourteenth century, this had evolved into a proper

244  Henriette Rensbro and Jens Christian Moesgaard

Figure 9.3 Aggersborg is located at the north side of Limfjorden. There has probably been a crossing point here since prehistoric times. In the Viking Age, the inlet Limfjorden was the direct passage from Western Europe to Eastern Denmark and the Baltic. Close to Aggersborg was possibly previously a waterway from Limfjorden to the north, which may have connected Jutland and Norway. Due to rising of land levels, sand blow out, draining of land and deposition of marine sediments, the waterways to the North Sea closed in the early Middle Ages (before twelfth century). Map by Steinar Kristensen, UiO: Museum of Cultural History.

castle with a defended stronghold, a bailey and a manor.31 Little is known about the location or type of the royal estate at Aggersborg at the time of the construction of Aggersborg church in the eleventh–thirteenth centuries, and there is certainly no sense of whether there was a castle chapel. However, the dating of the chancel of the current church to the mid-twelfth century, the existence of eleventh-century graves and the graves outside the churchyard indicate the existence of a (probably wooden) predecessor nearby.32 In 1976, the church of Aggersborg was renovated and the floors renewed. Funding was granted for a six-week excavation of the nave and chance. One of the key objectives

Figure 9.4 Aggersborg Church and the Viking fortress seen from the south. Traces of eleventh- and twelfth-century settlement have been excavated to the west and to the east of the older ring fortress. The present settlement by the name of Aggersborg is post-medieval and is related to the crossing point. Despite of this, the huge amount of metal-detecting finds from the Agger peninsula suggests that the whole area was densely settled in the late Iron Age into the Middle Ages. Photo: Lis Helles Olsen, Holstebro Museum, Denmark, with additions

Figure 9.5 Undated graves outside the present churchyard to the north discovered in 1949. The first Christian graveyard was probably located at the northern boundary of the present churchyard, and perhaps there was a wooden church in this area too. Photo: Hans Stiesdal, The National Museum of Denmark

246  Henriette Rensbro and Jens Christian Moesgaard

Figure 9.6 Two plans showing the distribution of coins a) and non-numismatic finds b) finds in Aggersborg Church. In the first plan each dot represents one coin. The colours indicate how precise the location of the single coin is known. Blue is the most precise and red is very imprecise. The fact that more of the precisely located coins are in the east might be a result of excavation methods more or less consciously adapting to the fundamental principles of the layout of churches, which is designed to emphasise the east. This of course does not suit a search for side alters very well. In the second plan, each dot marks a registration number. Each number comprises either one or several different kinds of objects. Unlike the coins, very few registration numbers are located precisely. Map by Steinar Kristensen, UiO: Museum of Cultural History.

The archaeological landscape under church floors  247

Figure 9.6  (Continued)

of this project was to look for a Viking graveyard under the church, a search prompted by the church’s proximity to the Viking fortress and the hypothesised continuity of notions of sacred ground between pre-Christian and Christian communities. This excavation was carried out by the NMD in collaboration with Aarhus University.33 Layers and features were excavated stratigraphically, meaning the most recent layer/feature was excavated and documented before moving on to the next layer, and all soil was sieved. The nave and chancel were divided into square-metre grids. The best-documented finds were, therefore, registered either by centimetres in three dimensions or by layer/feature and square metre. However, the accuracy of registration

248  Henriette Rensbro and Jens Christian Moesgaard was not upheld in all parts of the excavation, and it proved very difficult to separate layers in some parts of the church, particularly in the middle of the nave. A less detailed registration system based on parts of the building was employed, and consequently, some areas were excavated without the use of the square metre grid. Non-numismatic finds were registered less accurately than coins, and all of this had an adverse effect on the distribution plans Generally, the registration methods placed little emphasis on the exact position of finds or the extent or formation of layers. Considerable work was done both during the excavation and in the subsequent report to revise descriptions and interpretations of the function and dating of layers, but far from all uncertainties were solved.34 The use of coin finds was restricted to the dating of layers, as was the use of the 301 non-numismatic

Figure 9.7 Non-numismatic finds. The number of objects in each registration number is indicated by the size of the dots. Clusters are very clear in this plan. Map by Steinar Kristensen, UiO: Museum of Cultural History.

The archaeological landscape under church floors  249 records. As mentioned, many of these comprise more than one object; for example, 19 registration numbers of pins account for 52 objects. So, there are a total of 699 objects including fragments available for this survey. These were all finds that passed the assessment of relevance at the excavation and were later transferred to NMD.35 The most important excavated layers and constructions were divided into phases 2–7 in the excavation report, as shown in Table 9.1. They are presented here along with the chronology suggested by the excavators. There was no Viking burial ground beneath the present church, but there was a Christian graveyard from just before 1100;

Figure 9.8 Medieval features which are registered in plans. The floor Æ/CH (phase C) is blue, the wall ditch of the previous wooden nave (phase E) is red and the abandoned stone foundation is beige. Only coins (red) and non-numismatic finds (purple) from these features are displayed in this plan. Phase C-chancel is only registered in sections and is, therefore, not displayed here. Map by Steinar Kristensen, UiO: Museum of Cultural History.

250  Henriette Rensbro and Jens Christian Moesgaard Table 9.1 T he most important excavated layers and constructions (phases 2–7) in the excavation report with datings suggested by the excavators Phases 2015

Dating 2015

A

Disturbed top layers 18th–19th Modern floors C 16th–18th Early modern C floors 14th–15th Medieval floor C in the nave 12th–? Medieval floor in the chancel 1300 Stone nave early (construction) 14th C and demolition of wooden nave mid-12th– Wooden nave 13th C (construction) incl. wall trench mid-12th Stone chancel C (construction) 1100 Abandoned stone foundation (construction) and levelling layer late 11th– Graveyard and early traces of 13th C ploughing divided by e ditch

B1 B2 C-nave C-chancel D

E-nave

E-chancel F

G

H I

Total

Description

Layers and constructions

Coins Non-num. Dating registr.no. 1976

Phases 1976

105 137 BØ/AU, CD/BA, AF/BA U/Ø/BB/DE Æ (CH) DD/DC, EÅ

18 + A-B 54 50 17 2

AL, (AW/DO/ DN),

9 9 + A-D 13

AW/DO/DN, EF, EC, CR, CI, CK

D-E 4 4

7 Modern times

6

13th–15th C 12th–15th C 13th C

5 5 4

mid-12th C

4

BL, BQ

0

DH/AY, BF, R, BD, BM, EN

0 13

1100

3

BI, AØ/BE, DA/ DL, EI, EA, EY, FC, DØ, DY, DZ, EL, EM, CP/CQ, DR, DS, CY BP, BG, BH

3

late 11th C

2

Underground, clay and sand Graves. Early BV etc. modern, modern and undated

0 0 70 50

328 230 + 13a + 71b = 341 = 301

a Without information on context b Not connected to a phase

the graveyard was separated from an area with ploughing tracks by a boundary ditch (phase 2). The first building phase was a stone foundation from around 1100, with no traces of a building on top of it (phase 3). For some reason, the construction of a building was abandoned, and some 50 years later, the present chancel was built along with a contemporary wooden nave, which was replaced by a stone nave by the middle of the thirteenth century (phase 4). The medieval floor in the nave was made of tramped-down

1

The archaeological landscape under church floors  251 earth – not clay or mortar (this comprised phase 5). A few square metres of the floor were unearthed and registered in plans, parts of it were instead registered in several sections. In the chancel, the medieval floor (phase 5) was only registered in two sections. The excavators could not identify the late medieval/early modern floor layers,36 which perhaps were dug up and removed from the church at an earlier date.37 This explains the apparent absence of late medieval and early post-medieval layers (phase B2) which are mentioned several times in the discussions below. The earliest modern levelling layer of sand (phase 6) was observed directly on top of the medieval floor in several areas. In other areas, fragments of lime or mortar floors/layers U/Ø/BB and DE were recorded between these two layers. In conclusion, this short summary of the results from 1976 provides a picture of the complexity of the stratigraphy and the archaeological context in Aggersborg church and the potential difficulties for interpreting the coin finds.

Aggersborg revisited – the 2015 survey To investigate how and why coins ended up under the floor of Aggersborg church, it has been necessary to re-evaluate the excavation results by asking new questions about the function and formation of the layers and by analysing the composition and distribution of coins and non-numismatic finds. Non-numismatic finds are mainly used as an attempt to explore the function and formation of layers, and their distribution patterns have been compared with those of coins. Where possible, all layers have been re-evaluated according to time and building phases, function and formation processes. This has led to adjustments of phases and dating based on a combined assessment of both coins and stratigraphy. The changes are displayed and very briefly explained in Table 9.1 and should be compared to the 1976-phases and 1976-dating. Generally speaking, most layers and features have not been re-interpreted with regard to function, as this proved very difficult to do on the basis of the records available. However, in a few (but very important) cases, the layers have been re-categorised as “construction” or “levelling” layers. In this work, construction layers describe different kinds of layers created during a building or reconstruction process, including wastage of materials, walking surfaces and infill for foundation trenches. Levelling layers, on the other hand, for the purposes of levelling out or heightening floor surfaces were either created by the demolition of older floors or by redeposition of dirt from the outside into the church. It is crucial to identify and not to confuse the two kinds of layers, as finds from construction layers should not be interpreted in the same way as finds connected to the church’s use as the house of God. Furthermore, finds from redeposited levelling layers presumably ended up in a church floor quite accidentally. Both types of layers can be difficult to identify, but one almost certain way to detect a redeposited levelling layer is by the absence of finds. Potentially, such layers can also be identified by the presence of unexpected objects or materials, for example, organic materials that have been subject to taphonomic processes. The following is a brief presentation of the overall results of the 2015 survey.

Assessment of the composition of the coin assemblage With its 341 coins, the find assemblage from the excavation of Aggersborg is the largest church floor assemblage in Denmark.38 The geographical and chronological distribution is shown in Table 9.2. With a few exceptions, the composition is what one would expect

252  Henriette Rensbro and Jens Christian Moesgaard Table 9.2 Geographical and chronological distribution of the 341 coins found in Aggersborg church. The numbers refer to the publication of the coins by Jørgen Steen Jensen. Century Numismatic period 19th C 18th C 17th C 16th C

Denmark, incl. Norway Schleswig-Holstein

Early modern 19 coins: no. 252–70 and modern, 9 coins: no. 243–51 1541 onwards. Late medieval 25 coins: no. 219–42, 285 1400–1541 9 coins: no. 210–18

1 coin: no. 284 5 coins no. 279–83 5 coins: no. 274–78 0

Germany

Others

Total

0

0

20

0

0

14

0

0

30

0

17

0

30

8 coins: no. 307, 322–24, 326, 329–31 15th C 14 coins: no. 0 16 coins: no. 196–209 292–306,319 14th C Medieval 112 coins: no. 0 19 coins: no. 1234–1400 60–88, 77a, 308–18, 320, Early medieval 90–91, 104–06, 321?, 325, up to 1234 117, 120–95 327?, 328, 332–34 13th C 79 coins: no. 6–59, 2 cons: no. 2 coins: no. 89, 92–103, 272–73 290–91 107–16, 118–19 12th C 5 coins -1–5 1 coin: no. 0 271 Undetermined Total 272 14 45

3 coins: 134 no. 286, 288–89 2 coins: 85 no. 287, 339 0 6 5

5 341

from a Danish church floor.39 Strongly debased pennies from the late thirteenth and fourteenth centuries are, as always, the most numerous. Furthermore, as usual, from the time of the introduction of multi-denominational coinage in the late Middle Ages (fifteenth– sixteenth centuries), small change completely dominates here as in other churches. The relatively strong component of German small denomination coins from the last twothirds of the sixteenth century is a well-known feature of the Danish currency.40 Bracteates (thin one-sided silver or base-silver coins) are also common among Danish church floor finds. In Aggersborg, they account for 18% of the medieval coins (47 out of 262). Three bracteates are early Norwegian and two are early German dated to the thirteenth century. Fifteen are fourteenth-century German. Six are Danish from the early fifteenth century, and sixteen are fifteenth-century German. Four are undetermined. Moreover, there are two sixteenth-century blafferts (bracteates worth two pence). The twelfth century is relatively well represented, among others, by two specimens of a rare coin of King Canute (1146–57).41 The presence of two German bracteates of the thirteenth century is unusual,42 as are the three Norwegian bracteates. The latter are explained by the position of Aggersborg as an advantageous geographical “crossroads,” straddling key land and sea routes in northern Jutland. A south Italian coin is also not common and is a notable outlier. Overall, the proportion of postReformation coins is low compared to many other Danish churches, where they often account for half the amount of coins. There are no twentieth-century coins. The distribution pattern changes over time. There is a limited number of coins from the chancel due to massive incursions caused by the installation of large modern

The archaeological landscape under church floors  253 graves. Additionally, due to the fact that almost half of the area has not actually been excavated, it is not known if and how many times the floor has been heightened and lowered in the chancel. Consequently, it is difficult to deduce patterns of spatial relevance for the discussion of devotional practices throughout the Middle Ages from the

Figure 9.9 Distribution of coins by periods. a.1100–1234, b.1234–1400, c.1400–1541, d.1541–. Map by Steinar Kristensen, UiO: Museum of Cultural History.

254  Henriette Rensbro and Jens Christian Moesgaard small amount of coins recorded in this section of the church. However, some chronological developments of coin use in the chancel can be observed, as both the early medieval and the late medieval coins are centred in front of the main altar, while the thirteenth- and fourteenth-century coins are equally spread over the excavated area. The distribution pattern for the coin finds in the nave is greatly skewed: those dated up until 1541 are predominantly found in the eastern half of the nave, towards the chancel. The exceptions are mainly late medieval coins, but these are very few, and one should perhaps not put too much emphasis on the equal distribution in the nave. This small number of late medieval coins reflects the partial lack of late medieval and early modern floor layers (phase B2). Early modern and modern coins are almost completely absent in the middle sections of the nave by the south and north wall: the exceptions are coins found in early modern graves (phase I). So, it is safe to infer a correlation between the absence of coins and the absence of graves. This is explained by the removal of the modern floors. Only the coins redeposited in graves survived. Nine of the coins found in layers from phase D are dated around 1300. It is worth hypothesising that some of these were lost by the workers constructing the stone nave. If so, their presence on the site would suggest that builders’ salaries were paid at least partially in coinage rather than with bartered goods.43 Non-numismatic finds The finds from medieval contexts (phases C–G resp, phase 5–2) are by far the fewest; they account for 40 of 121 medieval registration numbers. The modern layers (phases A-B1) are the most prolific, followed by the post-medieval graves in phase I. Altogether, there are 301 registration numbers for non-numismatic finds, which can be divided into three categories: 1 2 3

artefacts most likely to have been used as parts of accoutrements and furnishings in the church, e.g., book claps, window glass and floor tiles44 etc., representing 22% of the registration numbers. artefacts most likely to have been in use largely outside the church, such as large numbers of ceramics, representing 26% of the registration numbers.45 artefacts that might have been in use in the church intermittently but that were left there, including tools and casting waste, as well as personal equipment, clay pipes, pins, brackets and jewellery, including beads, representing 52% of registration numbers.46 This category is the most comprehensive in the sample.

The presence of personal equipment (category 3) indicates that the parts of the church where they were found were accessible to individuals from the congregation, who occasionally accidentally dropped and lost personal equipment – pipes, pins and more. Some of these finds of course might derive from disturbed graves. The large amounts of ceramic fragments (category 2) on the other hand indicate the redeposition of dirt from outside the church. Potsherds and window glass are the most numerous find categories after coins, and one must consider how and why ceramics ended up in the floor. Some of them at least were possibly the remains of builders’ meals, taken during construction work, and some might have actually been used in the working process. But these conjectures cannot explain the large numbers found. Of the 78 ceramic-related registration numbers (comprising 272 objects), 55 are not connected to phases, whereas 11 belong to phase I.47 Therefore, a plausible explanation for a fair share of these potsherds must be that they were redeposited in the church within dirt used as levelling layers.

The archaeological landscape under church floors  255

Figure 9.10  C  hronological distribution of non-numismatic finds. 10a. medieval (until 1600), 10b. early modern, 10c. modern, 10d. not dated. Map by Steinar Kristensen, UiO: Museum of Cultural History.

256  Henriette Rensbro and Jens Christian Moesgaard

Figure 9.11 Composition of 301 non-numismatic registration numbers displayed as percentages.

For the purposes of distribution plans, it has proved feasible to split dating of non-numismatic finds into three types: • • •

not dated, representing 159 registration numbers (47% of the total) medieval (eleventh–sixteenth centuries), 126 registration numbers (42%) post-medieval (sixteenth–twentieth centuries), 35 registration numbers (11%)

Among the non-numismatic registration numbers, 159 are not dated in spite of the excavators’ dedicated efforts to date all objects if possible. Furthermore, only 11% of the registration numbers have been dated post-medieval, compared to 19% of the coins (Table 9.2: seventeenth–nineteenth centuries). This is probably partly because many of the finds, which are tools, pins, everyday ceramics and casting waste, are difficult to date precisely. And it might also be connected to the principles governing the keeping and discarding of finds (see note 35). The distribution of the 44 non-numismatic finds in the chancel is obscure (see Figure 9.13). Seemingly, there is a cluster in the north-east corner and a less substantial one in front of the main altar, but, as half of the area has not been excavated, this incomplete pattern is not conducive to a comprehensive interpretation. There are no finds from the medieval floor layer, of which hardly anything is preserved, and the majority of the 44 registration numbers (83%) are either without context or from phases A or I. However, we do know that 12 out of the total of 44 registration numbers from the chancel are medieval, and all of these are fragments of window glass (39 objects). Of the post-medieval objects, 75% are potsherds and 23% are objects that might have been in use in the church (category 3).

The archaeological landscape under church floors  257

Figure 9.12 Dating of 301 non-numismatic objects.

Despite the small number of finds, the lack of medieval personal equipment in the chancel points to several possibilities. Firstly, it could be that the congregation did not access the chancel in the Middle Ages. Secondly, filling/dirt was neither brought in from outside to level new floors nor redeposited from the nave into the chancel, so there was no redeposition of layers in the medieval period. Thirdly, the levelling of new floors first happened in the early modern period when graves were established in the chancel and fillings were brought from the outside of the church. This could explain the presence and high proportion of post-medieval pottery. In addition, the relatively small number of coins from the chancel could be the result of massive redeposition of floor materials taken out of the chancel. The medieval coin finds in front of the main altar could indicate that this area was spared during the construction work. Finally, due to the presence of post-medieval personal artefacts, it could be argued that the congregation accessed the chancel in the early modern period.48 Stratigraphy and the formation of layers As usual in churches, most of the contexts have been disturbed or removed or redeposited, but some of the layers at Aggersborg can be identified in terms of both function and formation. In this study, the term “construction layer” is used in a broad sense, meaning many different kinds of layers created during building or reconstruction processes, that is to say, the waste of materials, tamped-down walking path surfaces and the infilling of foundation trenches. This is to distinguish “construction layer” from “levelling layer,” which is a layer specifically placed with the purpose of levelling out or heightening floor surfaces. Such levelling layers were either created by demolishing and reusing old contexts or by bringing (redepositing) dirt from the outside into the building. Both levelling layers and construction layers can be difficult to distinguish. There are layers in most excavations that cannot be identified with certainty, and some of these might be levelling layers.49 In Aggersborg, small parts of unidentified and perhaps even disturbed layers without finds may well be traces of levelling activities.

258  Henriette Rensbro and Jens Christian Moesgaard In the Aggersborg church, levelling layers occur in medieval phase F (c. 1100), in medieval phase E-chancel (mid-twelfth century) and in modern phase B1 (eighteenth– nineteenth centuries). The finds are generally scarce, only one residual medieval coin under a recent brick floor (phase B1) and three non-numismatic registration numbers (phase F): one potsherd, one whetstone and one window glass (the latter is medieval, the others are not dated). The scarcity of finds in these layers indicates that most of the dirt was brought into the church from outside, although clearly not from a settlement area. This is especially obvious regarding the levelling layer in phase B (eighteenth century onwards), as hardly any other layer is as void of finds as this one. But if the amount of finds in the levelling layers is non-existent or minimal, so are some of the preserved parts of the excavated original layers, so distinction is difficult. Therefore, perhaps we cannot stress this characteristic of “emptiness” as a definitive factor of dirt being brought into the church building from outside. The construction layer AL (phase D, c. 1300) is a mixture of construction layer and levelling layer, as it clearly consists of disturbed floor layers from the wooden nave (phase E, mid twelfth–thirteenth centuries) reused (phase D) as a levelling layer in the stone nave (phase C-nave, fourteenth–fifteenth centuries). The 10 registration numbers from AL and the rest of the construction of the stone nave50 differ from the 15 registration numbers from the medieval floor (Æ/CH) in the stone nave (phase C-nave, fourteenth–fifteenth centuries). The most striking difference is the dominance of personal equipment and brackets in the medieval floor and the dominance of potsherds and animal bones in the older construction layers. It seems logical that the floor should contain mostly personal equipment and clothing ornaments connected to clothing for worship, while the construction layers contain mostly potsherds and animal bones which were either brought into the church by older levelling layers or leftovers from the construction workers’ meals. The medieval floor (Æ/CH) in the nave was made of dirt, not clay or mortar. This factor probably affected the unusually high number of coins and non-numismatic finds in and on the floor. Seemingly, both the floor surface and the underlying construction layer (AL, phase D) was, from the start in the early fourteenth century, created from disturbed floor layers from the previous phases D–E, and the surface continued to grow by sediments of dirt well into the fifteenth century. Not only does the dating of the floor rely on coins but the story of its formation is also deduced from the chronology of the coins. The fact that the youngest coin of the construction layer created in preparation for the stone nave dates from the 1320s and the oldest coins from the subsequent floor date from 1200/25 is intriguing. Indeed, the representation of coins from the early and mid-thirteenth century in layer Æ/CH is astonishingly strong: no less than 13 out of 50 coins (the rest date from the late thirteenth to fifteenth centuries). As a matter of fact, the very existence of these early coins was one of the excavators’ arguments for dating the stone nave to the thirteenth century. However, following the re-dating of the nave to the early fourteenth century (Table 9.1), these early coins need explaining. Renovatio monetæ (recoinages at regular intervals) were often applied during the period in question, 51 so the thirteenth-century coins were hardly in circulation when the floor was in use after c. 1300. One reasonable explanation here is that the coins were in the floor of the previous wooden nave (phase E) and that this was disturbed (phase D) and reused as a levelling layer for the stone nave (phase nave-C). Because this levelling layer was also the new floor, some of the coins were included in the floor surface along with coins in use during the fourteenth and fifteenth centuries. This would also mean

The archaeological landscape under church floors  259 that the thirteenth-century coins in layer Æ/CH, rather than the newly deposited fourteenth- and fifteenth-century coins, must be considered dislocated. This should be taken into consideration in analysing the distribution plans in any later survey.

Case study: three micro-concentrations of finds in Aggersborg church A number of concentrations of coins have been documented in Aggersborg church. In the following case study, we will present three different clusters and discuss whether these concentrations relate to a medieval situation and attendant offering practices or relate to post-depositional disruptions. Both clusters and the absence of coins draw attention when distribution patterns are studied in detail. In Aggersborg, the most promising areas for further study are the south-east corner of the nave (1), the middle sections of the north side and the south side of the nave (2), and the south-west corner of the nave (3). The floor area just inside the entrance door from the south porch of the church was disturbed and excavated with less accurate methods. The records hold little information on the area inside the entrance door to the north, except for evidence of two lead melting pits, used when re-melting the lead plates used on the roof; such melting pits are usually found on this spot in churches.

Figure 9.13 T hree identified coin clusters. Map by Steinar Kristensen, UiO: Museum of Cultural History.

260  Henriette Rensbro and Jens Christian Moesgaard Cluster 1. South-east corner of the nave52 A nine square metre area53 in the south-east corner of the nave produced 49 coins, or 14% of the total number of coin finds in the church. The early medieval graveyard (phase G, late eleventh–early thirteenth centuries) was preserved in this part of the nave, and it is assumed that a medieval side altar was situated in this corner, presumably one dedicated to the patron saint, as in most medieval village churches. In this area are two archaeological features of importance: pit J54 and grave AI.55 The composition of coins in grave AI is heterogeneous, as most of the coins are c.  1300, a few are mid fourteenth and fifteenth century, and three are postmedieval (1627, 1652, 1722). Six non-numismatic registration numbers correspond to the general composition of finds in phase I graves (see Figure 9.14), except for the fact that most of them can be dated; only one is post-medieval. From pit J, all 12 coins are medieval, and the non-numismatic finds are medieval window glass, not-dated potsherds and a bracket. The location of pit J suggests that it was established after the proposed side altar was most likely replaced by a pulpit in the sixteenth century. The pit was presumably a part of the foundation of the pulpit, which has protected the corner from modern graves and further disturbances.56

Table 9.3 T he chronology of coin finds from graves, pit J and top layer (phase A) Refined Rough dating dating

19th C 18th C 17th C 16th C 15th C 14th C

13th C

12th C Total

Early modern and modern, 1541 onwards. Late medieval 1400– 1541 Medieval 1234– 1400 Early medieval up to 1234

Precise dating

Pit Grave Grave Graves in nave and J AI AE chancel (phase I) comprising coins: AS, BX, D, DUa , F, K, L, M, N/Å, W, X and Z

1722 1652 1627 1624 1524 1450–1500 1420–1435 1400–1420 1350 1330–1340 1320–1329 1310–1319 1300–1400 1300–1302 1290–1299 1286–1289 1247–1278 12411234–1241

1 1 1

1 1 1 4 1 3

1

2

1 1 2

1 1 1

2 4

Coins from not specified gravesb

1 1

1

10

9

2

6

1 17

1 21

Coins from top layer, phase A 10 7 19 6 63

1

1 1

1 2 12

12

6

a One coin, 1320–29 is either from grave DU, phase I or from ditch AW, phase E. b Most of these coin are found in the centre aisle of the nave.

105

The archaeological landscape under church floors  261

Figure 9.14  Dating of objects from modern and undated graves.

The absence of post-medieval finds suggests that this pit belonged to a context that was closed in the late medieval period and left undisturbed over the coming centuries (see Table 9.3 and Figure 9.13). Thus, pit J is the only feature presumed to be from the sixteenth century. Considering that 40% of all coins in the church as a whole are dated between c. 1234 and 1400 (see Figure 9.9), it is notable that 67% of all coins found in the southeast corner are dated from this period. This suggests direct evidence of extraordinary activity related to the use of money, most likely associated with votive practices, in the south-east corner of the nave during the Middle Ages. Normally, one could make a comparison between the archaeological finds from the south-east and the north-east corner of the nave, but this is not possible because the archaeological landscape under the floor in the north-east corner has been completely destroyed by a modern heating system. There is no study from Danish excavations that can support the Swedish and Finnish demonstrations of clusters of coin finds in the east corners of naves.57 On top of this, it is of great methodological interest to speculate whether this cluster would exist if pit J had not been established and if the cluster would have been located in the exact same spot.58 Cluster 2.  Middle section of the north side of the nave59 The middle section in the north side of the nave in many respects corresponds to the middle section of the south side. First of all, the medieval floor Æ was preserved and

262  Henriette Rensbro and Jens Christian Moesgaard

Figure 9.15 Distribution map of coins versus non-numismatic finds. Map by Steinar Kristensen, UiO: Museum of Cultural History.

registered in both areas, and this preservation was probably partly down to the absence of graves. The dirt surface of this floor yielded unusually high numbers of finds, if compared to floor surfaces in general. 60 Many of the finds are not registered in accordance with the square metre grid but instead with the layer Æ (see Figure 9.6b) As part of the recent survey and fabrication of distribution plans, those finds had to be linked to the parts of the layer, that are recorded on plans (see Figure 9.8), and this is obviously one of the reasons for the existence of centres of distribution on either side of nave (see Figure 9.13).

The archaeological landscape under church floors  263 However, there are some marked differences between north and south. The straight lines of the layer’s boundaries in the plans indicate that more of the floor’s south side was preserved and excavated in 1976 than its north side and that the registered parts were balks of dirt maintained by the archaeologists (see Figure 9.8). However, we cannot be sure how much more floor was preserved and how many finds from Æ might have come from unregistered parts of this layer. The excavation report offers no answer to these uncertainties. Furthermore, in the list of layers, a disturbed layer, CH, is mentioned (not recorded) in the same part of the church, presumably to the north of Æ, and possibly to the west. The content of this layer was similar to the medieval floor with regard to the amount and composition of finds, and it is very likely that CH was a disturbed part of Æ. As part of the recent survey, it was decided to link the finds from layer CH to a location between the two registered parts of Æ. This might have affected the appearance of a cluster in the middle section of the north side of the nave. These two clusters of finds, as seen on the distribution plans, were probably caused by methods of registration and do not reflect the interior design and use of the church in the Middle Ages. However, the total lack of post-medieval coins in the middle section of the north side of the nave requires an explanation. The non-numismatic finds are also either medieval or not dated. Perhaps differences in floor surface material in post-medieval times created this unique combination of finds. The absence of modern graves, however, supports a well-known picture of a post-medieval interior with pews along the side walls of the nave, with the exception of the eastern part of the south side, where there must have been at some stage a space of greater or lesser extent west of the free of fixed furnishings, where graves could be established. Cluster 3. West end of the nave, the south-west corner 61 This cluster of finds might have been artificially shaped by the demolition and possible removal of cultural layers in front of the southern entrance door. The south-west corner of the nave is only cursorily registered, and half the finds belong to the disturbed top layers: 8 coins dated late fifteenth to early eighteenth century, and 6 non-numismatic registration numbers comprising 11 objects, none of which are dated. The essential feature here is an unnamed layer situated below the modern brick floor (phase B, sixteenth– nineteenth centuries) and on top of wall trenches and the old foundations (phases D + F). This layer held three coins dated 1250 to 152662 and six non-numismatic registration numbers comprising 24 animal bones, 13 medieval potsherds and 2 medieval brackets for belts. The taphonomy of the animal bones has not been analysed, but this study assumes that the bones belonged to domestic animals.63 This unnamed layer is enigmatic. It probably served as a levelling layer underneath the modern brick floor, but was it created by using disturbed medieval floors? Or was it redeposited in the church from outside? The large number of animal bones and the fact that suchlike are not found elsewhere in the church64 suggest that at least part of the layer was brought into the church as dirt from outside. Yet, why would one take out disturbed floor layers and bring in some other dirt? In this case, as in other similar cases, it is difficult to interpret and recreate the history of formation of floor layers in medieval and early modern churches. However, the non-numismatic objects and the distribution patterns of finds and layers strongly suggest dirt from outside was used for levelling.

264 Henriette Rensbro and Jens Christian Moesgaard The reality of clusters To sum up the case study of the three clusters of finds, it is notable that clusters on distribution plans might well be the result of excavation and research methods as the reflection of medieval reality. Therefore, before analysing distribution plans, one should not only study the registration methods of the excavation but also take into account the general stratigraphy of the church, the relation of finds to specific layers and the formation history of layers.65 It is notable that, in church archaeology, it is often deemed that effort and time is spent most productively in the eastern part of the building, whereas the western part is excavated with less concern and interest; this might also have been the case in Aggersborg.

Acknowledgements The authors wish to thank Dr Xenia Pauli Jensen for many and detailed discussions without which this article would have been less comprehensible. And to thank the hard-working editors, The Religion and money project at the Museum of Cultural History, University of Oslo, and Kristin Bornholt, Sean O’Neill, and Kay Celtel who all made valuable contributions.

Notes 1 Henriette Rensbro, Spor i kirkegulve – De sidste 50 års arkæologiske undersøgelser i kirkegulve som kilde til sognekirkernes indretning og brug i middelalder og renæssance, unpublished PhD diss (Aarhus: Aarhus University, 2007). 2 In modern Denmark, 2,000 coins in 40 churches were found during the first six years of modern church archaeology (1952–58) See, Olaf Olsen, “Kirkegulvet som arkæologisk arbejdsmark,” Nationalmuseets Arbejdsmark (1958): 20 ff. Between 1963 and 1973, some 3,041 coins was found in 108 churches (Kirsten Bendixen, “Middelaldermønter i de sidste 10 års danske kirkefund,” Nordisk Numismatisk Årskrift (1972): 49–70). In 1975, the total number of church coins (stray finds from church floors, excluding hoards) since 1837 was 9,266 coins from 369 churches, c. 50% are medieval (Jørgen Steen Jensen, “Kirkegulvsmønter,” Hikuin 3 (1977): 295–302 (English summary). By 1994, the total number was 11,397, still half of them being medieval (Keld Grinder-Hansen, Kongemagtens Krise. Det danske møntvæsen 1241–1340 [Copenhagen: Museum Tusculanum Press, 2000], 165). In the years 1953–2005, some 8,250 coins were recorded in 269 churches out of a total of 506 parish churches investigated (cathedrals and monastery churches not included). Of the 269 churches, less than ten coins were found in 90 churches, and we do not know the number of coins from 15 churches (Rensbro, Spor i kirkegulve). 3 Some examples are: Nørre Nebel Church (Olsen, “Kirkegulvet,” 25), Sønder Vilstrup Church, Brusk hrd., Vejle amt (unpublished) and Bregninge church, Skippinge hrd., Holbæk amt (unpublished). 4 This is closer to a desire than reality for archaeologists, as in the years 1953–2003, a total number of 13 coins in 269 churches were found in contexts which allow the excavators to use the coins to date layers and features. The small number reflects primarily the scarcity of coins in floor surfaces and secondarily the excavation methods (Rensbro, Spor i kirkegulve, 160). 5 Some examples are: Fritze Lindahl, “Valdemarernes mønter, især ANNO DOMINI MCCXXXIIII-mønten belyst ved løsfund fra kirkerne,” Nordisk Numismatisk Årskrift (1963): 50–60; Kirsten Bendixen, “Mønternes vidnesbyrd. Numismatik – arkæologi – historie,” Festskrift til K.G. Kaaber. Fra Holbæk Amt 69 (1976): 21–32. The discussion of Christian J. Simensen and Bjørn Poulsen in Nordisk Numismatisk Unions medlemsblad

The archaeological landscape under church floors 265

6

7 8 9 10 11

12 13 14

15

16 17 18

19

20

(1986): 3–8 and (1987): 2. Bjørn Poulsen, “Hvilke informationer giver møntfund ikke,” Nordisk Numismatisk Unions medlemsblad (NNUM) 8 (1986): 150–51. Examples of research on non-numismatic artefacts from church floors are floor tiles from Saint Jakobs Church (Sankt Ib) in Roskilde 1961 (Olaf Olsen, “Sankt Ibs Kirke i Vindebode,” Fra Københavns Amt (1962): 81 ff), pins from Søndersted Church (Birgitte Bruun Jørgensen, “150 knappenåle i 6 m2 kirkegulv,” Fra Holbæk Amt 70 (1977): 83–93, and not finished report from excavation in Hørdum Church 1955. Jensen,“Kirkegulvsmønter.” Henrik Klackenberg, Moneta nostra: monetarisering i medeltidens Sverige (Stockholm: Almqvist & Wiksell International, 1992). Olaf Olsen, “Rumindretningen i romanske landsbykirker,” Kirkehistoriske Samlinger 7 (1965): 236–57; Olaf Olsen, “Middelalderkirkens møblering,” Skalk 5 (1966): 5–11. Elna Møller and Olaf Olsen, “Danske trækirker,” Nationalmuseets Arbejdsmark (1961): 35–58. Nils Engberg, Danmarks døde kirke – Middelalderens nedlagte kirker (Odense: Syddansk universitetsforlag 2018). Olsen, “Kirkegulvet,” 20. Birgit Als Hansen, “Gulve – i kirker,” in Bolig og familie i Danmarks Middelalder, edited by Else Roesdahl (Arhus Universitetsforlag, 2003), 247– 53; Knud Høgsbro Østergaard, “Arkæologiske undersøgelser i Viv Kirke,” Vejle Amts årbog (1957): 5–24. Svend Bruun Jørgensen, “Rapport fra undersøgelsen af nogle gulvlag i Søndersted Kirke,” Fra Holbæk Amt (1977): 71–79; Knud J. Krogh, “Gyldent alter, glasmalerier og andre fund fra Malt kirkes kor,” Fra Ribe Amt (1971): 513–47. Niels Knud Liebgott, Dansk Middelalderarkæologi (Copenhagen: Gad, 1989). Rensbro, Henriette, Nils Engberg and Jakob Kieffer-Olsen, “Kirken som gik i havet – Udgravninger i Mårup Kirke 1998, 2009 og 2015,” Nationalmuseets Arbejdsmark (2016): 130–43. Henrik Græbe, Birgit Als Hansen and Morten Aaman Sørensen, “Gundsømagle kirke. En bygningshistorisk undersøgelse,” Nationalmuseets Arbejdsmark (1990): 141–56. Birgit Als Hansen, “Arkæologiske spor efter døbefontens placering i kirkerummet gennem middelalderen,” Hikuin 22 (1995): 27–40. Birgit Als Hansen and Morten Aaman Sørensen, Ornamenterede middelalderlige gulvfliser i Danmark (Copenhagen: Det Kongelige Nordiske Oldskriftselskab, 2005) (English summary). Birgit Als Hansen, “Medieval Floor-tiles in Denmark,” Journal of the British Archaeological Association 153 (2000): 93–101. Birgit Als Hansen et al., “The Bistrup Project: A Comparison of Floor-tiles from Medieval Churches by Means of Neutron Activation Analysis,” in Second Nordic Conference on the Application of Scientific Methods in Archaeology, Helsingør, Denmark, 17–19 August 1981, edited by Olaf Olsen and Vagn Mejdahl (Strasbourg: Council of Europe, 1982), 383–39. Rensbro, Spor i kirkegulve, 102ff. Jes Wienberg, “Kirkearkæologi – Fra stil til kulturarv,” META – Medeltidsarkeologisk Tidskrift 3 (2006): 19–29. Aksel E. Christensen, “Danmarks befolkning og bebyggelse i middelalderen,” in Befolkning i Middelalderen, edited by Adolf Schück (Oslo: Aschehoug, 1938); Jes Wienberg, “Den gotiske Labyrint. Middelalderen og kirkerne i Danmark,” Lund studies in Medieval Archaeology 11 (Stockholm: Almqvist & Wiksell International, 1993); Poul GrinderHansen, “Kirkerne som kilder til middelalderens kulturlandskab,” in Mellem Fjord og Bugt, edited by Hans-Christian Eisen (Roskilde: Historisk Samfund for Roskilde Amt, 2009), 161–96. The role of the architects was and still is construction management; they function as the link between the parish vestry, contractors and antiquarians. Previously, they were often responsible for part of the registrations, especially the building archaeological registration of masonry, which today is the job of archaeologists (Birgit Als Hansen and Henriette Rensbro, “Kirkearkæologi fra pionertid til nutid – 50 års arkæologiske undersøgelser af kirkegulve,” Nationalmuseets Arbejdsmark (2007): 165–80. In Denmark, the latest large-scale church excavations are Slots Bjergby 1991 (Als Hansen and Rensbro, “Kirkearkæologi,” 165–80; Als Hansen, “Arkaeologiske spor efter dobefontens”) and Hedensted 2007–2008 (Michael Andersen and Hans Mikkelsen (eds.), Hedensted Kirke – undersøgelser og restaureringer, Nordiske Fortidsminder, serie B, bd.

266 Henriette Rensbro and Jens Christian Moesgaard

21 22 23 24 25

26 27 28

29

30 31 32

33 34

35

26, (Copenhagen: Nordiske Oldskriftselskab, 2016), including Gitte Tarnow Ingvardson, “Sjælefrelse eller lommespild – møntfundene fra Hedensted kirke,” 79–97. In Norway, the latest large-scale excavation was Ringebu stave church 1981–2 (Terje Hellan, “A Stave Church Revisited. Ringebu Stave Church, Ringebu Municipality, Oppland County,” in Coins in European Churches: Religious Practice and Devotional Use of Money. Schweizerische Numismatische Rundschau (SNR), edited by Benedikt Zäch and Svein H. Gullbekk 97 (2020): 225–43. All of them were conducted by experienced church archaeologists. Gullbekk et al., Wangsgaard Jürgensen and Ingvardson, this volume, chs. 1, 2 and 10. Risvaag, Myrberg Burström, Klackenberg and Jonsson, this volume, chs. 6, 11, 12 and 13. Jürgensen, this volume, ch. 2. Alex Wittendorff, Rejsen mod virkeligheden. Den europæiske forestillingsverden fra reformationen til nutiden (Copenhagen: Aschehoug, 1986). Else Roesdahl, “Aggerborg’s Location and History: Situation,” in Aggersborg – The Viking-Age Settlement and Fortress. Jutland, edited by Else Roesdahl, Søren M. Sindbæk, Anne Pedersen and David M. Wilson (Copenhagen: National Museum of Denmark, 2014), 17–30. Else Roesdahl, The Vikings (London: Penguin, 2016), 143–48. Jens Christian Moesgaard, “Møntproduktion i Ørbæk og kongemagtens interesse i Limfjordsområdet,” Årbog for Vesthimmerlands Museum 4 (2012): 31–41. In Fyrkat, another comparable Viking ring fortress, the excavation of the graveyard has yielded important information on the fort. See Else Roesdahl, Fyrkat. En jysk vikingeborg II: Oldsagerne of gravpladsen, Nordiske Fortidsminder, Serie B, bd. 4, (Copenhagen: Nordiske Oldskriftselskab, 1977). Jørgen Steen Jensen, “Mønterne fra Aggersborg Kirkes gulv,” in Aggersborg i vikingetiden – bebyggelsen og borgen, edited by Else Roesdahl, Søren M. Sindbæk and Anne Pedersen (Højbjerg: Jysk Arkæologisk Selskab, 2014), 468–73. A summary in English of these coin finds, without a coin list, is presented by Jørgen Steen Jensen, “Coins from the Floor of Aggersborg Church 1976. A Summary,” in Aggersborg. The Viking-Age, 418, appendix 2. Moesgaard, “Møntproduktion,” 31–41; Jens Christian Moesgaard, “Om den sande natur af 1000-tallets danske udmøntninger: nogle tanker inspireret af mønter fra Ørbæk,” Nordisk Numismatisk Unions medlemsblad (2013): 21–23. Andres S. Dobat, “Kongens borge: Rapport over undersøgelserne 2007–2010,” Jysk Arkæologisk Selskabs Skrifter 76 (2013): 182 ff. Simon Kjær Nielsen, “Middelalderborgen ved Aggersborg,” Vesthimmerlands Museum Årbog 2013 (2013), 37–44. Henrik Græbe, “Kirken,” in Aggersborg gennem 1000 år: fra vikingeborg til slægtsgård, edited by Felix Nørgaard et al. (Herning: Poul Kristensens forlag, 1986b), 171; Søren M. Sindbæk, “The Excavations: Strategy and Process,” in Aggersborg – The Viking-Age Settlement and Fortress, edited by Else Roesdahl, Søren M. Sindbæk, Anne Pedersen and David M Wilson (Copenhagen: National Museum of Denmark, 2014), 71. Else Roesdahl, Aggersborg Church 1976 (1977, unpublished). Excavation Report available at the Antikvarisk-Topografisk archive, The National Museum in Copenhagen. The excavation report has two separate texts by Else Roesdahl and Henrik Græbe, the first one includes a preliminary list of phases. There are descriptions of layers and contexts, a list of coins, a list of non-numismatic finds, 15 plans and 15 sections, a list of photos, registrations of each grave with grave-filling and skeleton and reports from analysis of human bones, layers and some finds, for instance window glass by Birgit Als Hansen, the book clasps of walrus ivory by Tage E. Christiansen, the human bones from graves in phase G by Berit Sellevold and the coins by Jørgen Steen Jensen, “Mønterne”. A list of medieval finds provided by Else Roesdal has been most helpful for this study. Two articles published 1977 and 1986 are used as a supplement to the report (Henrik Græbe and Else Roesdahl, “Kirken ved Aggersborg,” Nationalmuseets Arbejdsmark (1977): 13–26; Henrik Græbe, “Fra gård til købstad,” Hikuin 12 (1986a): 7–16. According to the excavation report “coffin nails and coffin brackets and the like made of iron were discarded at the excavation” before registration. Likewise, it is mentioned that only three out of several fragments of modern triangular floor tiles were registered. Post-medieval window glass was collected but does not appear in the find list and was

The archaeological landscape under church floors 267

36 37 38

39 40 41 42 43

44 45 46 47

48 49 50 51 52 53 54

55 56 57 58 59

probably discarded at the museum. Human bones, which are a common find type in church excavations, are missing in the list of finds. Animal bones are few, but this scarcity is real, as all animal bones were registered (Information by Else Roesdahl in May 2015). It is usual procedure for excavations to collect and register most finds from old and not-dated layers and features, whereas from obviously modern features mainly old artefacts and notdated finds are collected and registered. This strategy of registration potentially hampers comparative studies. It can be assumed that in Aggersborg many modern objects were not registered. Therefore, the bulk of objects in the find list are those which are often difficult to date by themselves, e.g. tools, black everyday ceramics, beads, brackets, casting waste and pins. Roesdahl, “Aggersborg Church.” Græbe, “Fra gård til købstad,” 7–16. As for the coins, this study relies on Jørgen Steen Jensen’s identifications supplemented by Grinder-Hansen’s more precise dating, and mint attributions for the coins of the period 1241–1340 (Keld Grinder-Hansen, Kongemagtens Krise). A few errors in the find numbers occur: coins 134–35=finds 262–63, coin 235=find 165, coin 336=find 189, coin 338=find 273. Coin 233 is from 1644 and coin 235 is Hede 151B. Bendixen, “Middelaldermønter,” 47–70. No thorough study of the typical composition of coin finds in churches has been conducted for several decades. Thomas Guntzelnick Poulsen, “Danmarks møntcirkulation i anden halvdel af 1500-tallet. Om betydningen af nordtyske mønter,” Nordisk Numismatisk Unions Medlemsblad (2007): 77–84. See further discussion in Jørgen Steen Jensen, “Mønterne,” 468 and 472. Jørgen Steen Jensen, “Tyske brakteater i danske fund,” Hikuin 6 (1980): 101–10. Thibault Cardon, “Les usages des monnaies (mi XIIe–début XVIe s.). Pour une approche archéologique, antropologique et historique des monnaies médiévale,” unpublished PhD diss. (Paris: Ecole des Hautes Etudes en Sciences Sociales, 2016). Hommedal, this volume, ch. 8. There are no objects connected to liturgy among the finds. Slag from iron making is absent in Aggersborg, and there is only one fragment of a horseshoe. An attempt is also made to split the non-numismatic finds into three categories: delicate, surprising and others. This has produced only one useful piece of information: most finds are neither delicate, nor surprising. Unfortunately, it has not been possible in connection with this re-investigation of the excavation to undertake a thorough study of the finds. Consequently, we do not know how many of the sherds show traces of wear from after they were broken and were, therefore, more likely to have been brought to the church in redeposited layers. On access to the chancel, see also Klackenberg, this volume, ch. 12. Rensbro, this volume, ch. 3. Finds which might come from the older phase E (construction of the wooden nave) are incorporated in these ten numbers. Keld Grinder-Hansen, “Møntfornyelse (Renavatio monetae) i Danmark fra år 1200 til Christopher II’s død i 1332,” NNUM (1997): 130–36. Plan 2 + 11, section III + VII (area DÅ + AP). Grids O-R/18-20. Pit J was registered as a grave but later described not to be. The excavation report offers no alternative explanation. J is still, for practical reasons, connected to phase I, despite the recent re-dating from “undated” to “medieval/early modern”. J was earlier thought to be younger than feature AI dated to after 1722 by coin 245. Grave AI is situated partly in the corner and partly to the west of the corner and was difficult to distinguish from J. Burials inside churches were forbidden in Denmark from 1805. Klackenberg and Jonsson, this volume, chs. 12 and 13. We might get a little closer to answering these questions by looking at the two modern graves, AE and AQ, in the west end of the nave. Plan 16, section IX (area DM + CØ).

268 60 61 62 63

Henriette Rensbro and Jens Christian Moesgaard

Rensbro, Spor i kirkegulve. Fig. 13 area CZ. 1250–52, 1340–54, 1524–26. Animal bones in churches are generally either from birds, rats or mice or from domestic animals like cows and pigs. 64 There are only two registration numbers for animal bones from Aggersborg church. 65 Rensbro, this volume, ch. 3.

10 Coins in church contexts Hedensted Church, Jutland, Denmark Gitte Tarnow Ingvardson

Church archaeology has produced coins, jewellery, tools and liturgical objects hidden for centuries under the floor, in the walls or in the church interior. The objects mirror historic actions in the church room: deliberate, accidental, liturgical, personal or perhaps work-related actions. The focus of this chapter is to identify liturgical and other practices that lead to church depositions. The analysis is based on excavations of Hedensted Church (Eastern Jutland), where underfloor heating was planned in 2007. Therefore, a complete excavation of the church was necessary.1 Complete church excavations with modern archaeological methods are extremely rare in Denmark. Six thousand objects including 127 coins were excavated in different contexts (graves, dirt layers, work areas) connected to several construction and renovation phases. Artefacts in relation to context, activity phases, building structure and interior will be discussed in the chapter. The analysis will address the issue of accidental losses and deliberate offering, as well as pinpoint methodological problems related to the interpretation of church finds. The complete excavation of Hedensted Church utilising modern archaeological methods for documentation facilitates a nuanced discussion on if, how and when the material cultures beneath the church floor reflect the interior and more actively liturgical practice within the church.

Figure 10.1  Hedensted Church seen from south-east. Photo E. Nyborg, National Museum of Denmark (CC-BY-SA).

270  Gitte Tarnow Ingvardson

Figure 10.2 Hedensted Church seen from north-east with the fifteenth-century tower, before it was demolished in 1939. Photo from 1908 K. Hude, National M useum of Denmark (CC-BY-SA)

The city of Hedensted and Hedensted Church is situated in Hedensted parish in Eastern Jutland near one of the main thoroughfares of Jutland (cf. Chapter 1, Introduction, Figure 10.1). 2 Originally, the church was built on a marked ridge between bogs and streams approximately 66 m above sea level. The stone church of Hedensted was constructed with apse, chancel and nave around 1150, with a total exterior length of 27 m. Hedensted is the only church in Hatting district built in ashlar stone and not in limestone. However, the church is provided with interior triangular gables of limestone. In the second half of the fifteenth century, chancel and nave were vaulted, and at the same time, porch and tower were added to the church construction. The fifteenth-century tower was demolished in 1939 to give way for a larger tower (Figure 10.2). The facades of the church are decorated with a number of ashlar tablets with more or less elegantly formed animals, a rider and a scene depicting the weighing of souls (Figure 10.3).3 Wall paintings from five different periods have decorated the interior of the stone church. An apse decoration with Christ standing in a mandorla flanked by Peter and Paul is dated to c. 1175–1225, which is almost simultaneous with the construction phase of the stone church.4 The main part of the furniture is from the beginning of the seventeenth century including an alms post.5 The bell was cast in 1592. The only furniture from the early phase of the stone church is the Romanesque granite font, whose original position is not established (Figure 10.4).6 The building, frescoes and interiors of Hedensted Church are discussed and described in detail in Danmarks Kirker.7 In 2007–2008, the Church was excavated by the National Museum in corporation with Horsens Museum. The excavation was headed by Hans Mikkelsen. The results of the excavation, including analysis of activity phases, frescos and the around 6,000 archaeological excavated artefacts, are published in Hedensted Kirke.8 The analysis of this paper is based on the results of the excavation.

Coins in church contexts  271

Figure 10.3 The Archangel Michael and a demon are weighing souls. Ashlar tablet in the south-east corner of the nave in Hedensted Church. Photo A. Mikkelsen, National Museum of Denmark (CC-BY-SA)

Figure 10.4 Romanesque granite font from Hedensted Church. Photo A. Mikkelsen, National Museum of Denmark (CC-BY-SA)

272  Gitte Tarnow Ingvardson

How do changes in the interior and architecture affect the coin profile? Rebuilding, renovations and extensions have changed the interior and architecture of all extant Danish medieval churches. The churches had several building phases, and many churches stood as building sites for decades.9 In Klackenberg’s (1992) dissertation on Swedish church finds, it is demonstrated that coin finds are concentrated in chancels, the eastern part of naves and by entrances, corresponding to the areas, where main altars, side altars, images of saints and offertory trunks were placed. Klackenberg argues that this distribution supports the notion that the coin finds primarily represent accidental losses in connection with offerings in the church. It is further argued that the general decrease in the coin volume around 1500 should be seen as a result of changes in the liturgical practices after the Reformation.10 Complete excavations of churches with modern archaeological methods are extremely rare in Denmark. The main body of Danish church finds came to light in the 1950s and 1960s during large church restoration campaigns. Fortunately, the many church finds were saved from destruction, as all earth removed from the churches were sieved and objects, especially coins, were sent to the National Museum. However, the find spots were not documented and contextual information was not provided. With the 1970s came a growing awareness of the significance of in situ preservation in churches.11 Large-scale church excavations are therefore very unusual. Thus, the character of the main body of coin finds in Danish churches is not comparable with the results from the Swedish survey. The complete and modern excavation of Hedensted Church facilitates a detailed and nuanced study of coins in relation to contexts. During excavation, 6,000 objects were found and recorded, this according to Danish standards extremely high number of objects testify to the accuracy of the excavation. As every object is carefully recorded, the material culture can be related to a number of activities of both religious and secular nature in the church room, thus presenting changes in the interior and the architecture as a new dimension in the discussion and interpretation of the chronological and chorological distribution of coin finds. Furthermore, Hedensted is the only church in Denmark with a documented medieval plank floor12 and thereby constitutes an important contribution to the discussion on the relation between floor designs and coins losses. The aim of the paper is to discuss if and how changes in the interior and the architecture of the church affect the chronological and chorological distribution of the coin material. The well-documented archaeological results also support a discussion on some of the methodological problems related to the interpretation of church finds in general. The archaeological setting During excavation of Hedensted Church, 13 phases were observed. Some activity phases represent major changes in the interior and architecture of the church, while others did not affect the design of the church room. Thus, some phases are naturally more relevant to the present study than others. However, a short outline of the chronological sequences is needed as support for the analysis. The phases are based on observations made during the excavation and are described in the archaeological report.13 Phases 1 and 2 include structures from the Iron Age and a cultivation

Coins in church contexts  273 horizon. In phase 3, a three-aisled wooden structure measured a minimum of 15 x 7 m was constructed (Figure 10.5). The building was equipped with a clay floor, approximately 5 cm thick. The floor was covered by a stamped dirt layer, except for the area between the eastern roof-bearing posts. In the chancel, the floor showed no signs of wear and was 20 cm thick. This might indicate the position of an altar (Figure 10.6).14 Two identical formed wooden lined wells measuring 2 x 2 x 2 m were excavated approximately 5 m west of the wooden structure (Figure 10.5). The wells were taken out of use simultaneously, and the upper part was filled with stones. Charcoal found in the filling of the wells is dated to the eleventh century, and ceramic between the stone fillings dates to the end of the eleventh century. This indicates that the wells

Figure 10.5 Older structures within the frame of the Romanesque Hedensted Church. The function of the little pit between the two wells is unknown. Map: Steinar Kristensen, UiO: Museum of Cultural History based on data of N.H. Bendtsen, Horsens Museum.

274  Gitte Tarnow Ingvardson

Figure 10.6 The chancel in Hedensted Church is seen from the West. The thickened clay floor is visible in the profile. The white circles mark the position of the postholes in the east end of the wooden church. Photo National Museum of Denmark (CC-BY-SA)

functioned together with the wooden building, but were taken out of use half a century before the stone church was built. A boundary ditch was found just west of the wells (Figure 10.5). Ditch and wells may have been in function at the same time, but the ditch was constructed before the wells. The three-aisled wooden structure, the two wells and the ditch are by the excavators with caution interpreted as an assembled structure consisting of an ecclesiastical wooden building, two baptismal wells and a delimitation ditch of the cemetery.15 Shortly after the middle of the twelfth century (phase 4), the chancel and apse of Hedensted stone church were constructed on a limestone layer of 10 cm and on top of this a levelling layer of chippings of 25 cm. Both layers are probably residual products from the hewing of granite ashlars and limestones. The chipping layer was observed in several areas of the nave, but was here mixed with sand and only up to 4 cm thick. The thick levelling layers raised the floor in chancel and apse to a higher level than in the nave. The current stone church is built on top of the wooden church, the two wells and the ditch in a way that structure, wells and ditch are placed symmetrically beneath the church. One well is placed under the floor just inside the south entrance, the other well is placed just inside the north entrance, and the area with raised clay floor indicating the position of an altar is situated beneath the chancel (Figures 10.5 and 10.6). Even though the construction of the stone church concealed the wooden church, wells and ditch, the accuracy and correlation between the old wooden structures and their stone successor clearly signal that place continuity was emphasised during the construction of Hedensted stone

Coins in church contexts  275

Figure 10.7 Example of part of wooden plank floor preserved in Hedensted Church. The wood is surrounded by a layer of dirt, which was formed during the use of the plank floor. Photo N. Bendtsen, National Museum of Denmark (CC-BY-SA)

church. It also indicates that the wooden church was in use right up to the foundation of the stone church. The cemetery was established in phase 5, shortly after the construction of the stone church. Phase 6 includes the oldest floor of the stone church, which was a plank floor with north south going planks resting on longitudinal placed joists, constructed on top of the levelling layer of chipping described in phase 4 (Figure 10.7). During the next phase 7 from the middle of the thirteenth century to the beginning of the fourteenth century, the roof is among other things renovated, and lead casting pits in the church room can be connected to this phase. The plank floor was destroyed during this renovation, and as the final step of renovation, a brick floor in herringbone motif was established constituting phase 8. Remains of this brick floor were only preserved in the chancel and in the nave in Section 8. However, the screed layer of yellow sand beneath the brick floor was preserved/visible in all sections, and many very small fragments of red brick were present on the top of the screed layer. Foundations and scaffold postholes from phase 9 demonstrate that the church was vaulted in the first third of the fifteenth century. Again, the building activity had probably destroyed the floor, and the floor was removed and replaced with a floor of dark red square brick tiles. This floor was at some point removed in order to give way for another floor, which is not known today. The floor of square brick tiles was very carefully removed and was only preserved in the south-eastern corner of the nave in Section 11. Phase 10 includes the end of the seventeenth century and the eighteenth century, where many graves were established particularly in the centre aisled of the nave and in the chancel as well as in a grave chamber in the tower. No traces of floors from sixteenth century to nineteenth century were documented during excavation. A large church renovation also took place in the seventeenth century, and it is likely that older floors were removed at that time. Phases 11, 12 and 13 represent modern construction work as heating systems, the rebuilding of the tower in 1939 and instalment of

276  Gitte Tarnow Ingvardson modern flooring. It is presumed by the excavators that the rebuilding in 1939 damaged older layers. Around 6,000 objects were found during the excavation of Hedensted Church. The church room was divided into 13 sections by old heating pipes, and all objects were documented according to these sections (Figures 10.8a–b and 10.10). Almost all objects can be dated to the period of the stone church. Besides their significance for chronological considerations, the 6,000 artefacts are essential when interpreting the nature of the activities that took place in the church room. The artefacts from the excavations can be divided into four categories:16 Building debris: The largest group of objects; iron nails, fragments of window glass, lead foil and bronze foil, bricks and granite stone. These objects were formed and deposited during the building and renovation phases of the church. Liturgical objects: Objects related to the liturgical processes in the church as the bracket of a hymn book, fragments of censers, a washbowl and an ivory reliquary. Other objects that may be related to ritual actions are also included in the group, as fragments of oil bottles, a prunt of a krautstrunk beaker and nails probably from coffins. Personal artefacts: Primarily small objects as jewellery, buckles and knives. The personal objects are probably brought into the church room by the people who used the church for liturgical purposes or by people who worked in the church during rebuilding and renovations.

Figure 10.8a Hedensted Church just before excavation seen from the West. The modern church floor is removed, and heating pipes are dividing the church room into 13 sections.

Coins in church contexts  277

Figure 10.8b  Hedensted Church during excavation, seen from the East. Photos N. Bendtsen, National Museum of Denmark (CC-BY-SA)

Coins: The coin finds are also brought into the church room by the people who visited the church. The actions that lead to their deposition may be accidental losses, deliberate offerings/donations or a combination of both. As the main focus of this paper is the coin finds of Hedensted Church, a short presentation of the chronological and chorological distribution of the coins is given. Outline of the chronological and chorological distribution of coins One hundred twenty-seven coins in total including 110 medieval coins from the twelfth century to sixteenth century were recovered from the excavation of Hedensted Church. Two previously unknown bracteate types are present among the coin finds (Figure 10.9a and 9b). Two coins are not identified; the remaining 125 coins can be divided into five chronological groups.17 Group a. (c. 1150–1250) 14 coins: 13 Danish pennies minted during the reign of Valdemar I, Knud VI and Valdemar II c. 1150–1241 and a Norwegian bracteate minted c. 1200–1300. Group b. (c. 1250–1350) 71 coins: 68 Danish pennies minted c. 1241–1340, two Swedish Magnus Eriksson (1319–1364) bracteates and a sterling minted by Tomas de Bourlémont, Bishop of Toul (1330–1353). Group c. (c. 1350–1525) 24 coins: 17 North German bracteates minted in the second half of the fourteenth century to the beginning of the sixteenth century. From Lübeck a Viertel Witten (1375–1400) and a Dreiling (1392–1398). The Danish king

278  Gitte Tarnow Ingvardson

Figure 10.9a T he two best-preserved examples of the new bracteates found in Hedensted Church. Photo J. Lee, National Museum of Denmark

Figure 10.9b Church finds are a potential source for new unknown coin types. Two different variants are present among the four new bracteates, as one of the coins is decorated with four as opposed to three leaves. Coin no. 4 was found in the undisturbed layer of chippings in the chancel and is therefore decisive for the dating of the construction of Hedensted Church. Drawings P. Wöhliche, National Museum of Denmark (CC-BY-SA).

Erik of Pomerania (1396–1439) is represented with three Kronehulpenninge (1405– 1420) and a Kobbersterling (1420–35), and Christian I (1448–1481) is represented with a Hvid. Group d. (c. 1550–1700) 13 coins: Danish skillinge (pennies) from Christian IV (1588–1648), Frederik III (1648–1670) and Christian V (1670–1699); seven skillinge, four 2-skillinge and one 4-skilling. Group e. (1700–2000) three coins: Christian VII (1766–1808), skilling minted in 1979; Christian VIII (1839–1848), four rigsbankskilling minted in 1842; Fredrik IX (1947–1972), ten øre minted in 1961. During excavation, the coin finds were related to 13 sections, of which only Sections 4–11 and Section 13 are relevant in the analysis, as Sections 1–3 are situated in the

Coins in church contexts  279

Figure 10.10 Building phases of Hedensted Church. The excavated sections are marked with numbers 1–13. Map: Steinar Kristensen, UiO: Museum of Cultural History based on data of N.H. Bendtsen, Horsens Museum.

part of the church that was built in 1939, and Section 12 was not excavated as this would have destabilised the north wall of the chancel (Figure 10.10). All sections older than 1939 (4–11) hold coin finds. Section 7 situated just east of the south entrance and Section 13 covering chancel and chancel arch display the largest concentrations of coins with 14 and 29 specimens (Figure 10.11). Sections 5, 7 and 9 in the southern half of the nave and Section 8 in the northern half of the nave are dominated by coins from group b, which mainly consists of Danish pennies minted c. 1241–1340. Chancel and chancel arch are the only areas where early modern coins are predominant; in fact, all group d coins minted between c. 1550 and 1700 are found in Section 13.

280  Gitte Tarnow Ingvardson

Figure 10.11 T he chorological and chronological distribution of coins found in Hedensted Church. Map: Steinar Kristensen, UiO: Museum of Cultural History based on data of N.H. Bendtsen, Horsens Museum.

Before analysing the chronological and chorological distribution of coins outlined here, a short summary is given of the contexts in which coins were found during excavation. Furthermore, examples of recontextualised as well as in situ preserved coins are presented.

Coins in church contexts  281 Contexts The excavation demonstrated that the material culture found in the church can be connected to different contexts related to the construction, the renovations or the use of Hedensted Church. Contexts related to the construction of the stone church include the foundation and the layer of chippings, which formed the base for the first plank floor. Contexts related to the renovations of the church are numerous and include levelling layers of different types of floors, demolition and construction layers with building debris, scaffold postholes, vault foundations, lead casting pits and modern pits and trenches connected to heating systems. Contexts that may be directly connected to the liturgical use of the church are limited to graves and the dirt layer generated between and under the planks of the first floor in the stone church (Figure 10.7). The church room was transformed into a building site, during the major renovation activities. A clear evidence of redeposition of finds from different contexts was established in Section 5. In this area, two building layers18 formed in 1939 contained artefacts of medieval origin, and the composition of finds in the layers suggested that they had previously functioned as levelling layer for a brick floor of the fourteenth century and fifteenth century.19 The layers contained coins, waste of roof lead plates, fragment of a censer, window lead glazing and fragments of brick, lime and calcareous tufa. This combination of objects demonstrates that the layers were formed inside the church, and not brought into the church from the outside.20 When redeposited coins are found in layers with a similar combination of objects, it can be established that the coins were deposited in the church, but it cannot be established where the original depositions were made in the church, nor can the actions that lead to the depositions be established. The levelling layer of chippings in the chancel was almost without finds, except from a single animal bone, a Danish bracteate and some small iron fragments. In the apsis, however, the layer of chippings contained quite a few objects, among others five coins: a Danish bracteate of the same type as in the chancel, a Danish Valdemar II penny minted before 1234 and three Danish pennies minted respectively c. 1290–1299, 1320–1329 and 1330–1339. This marked difference in find intensity is due to later disturbances in the apsis, where younger objects have been redeposited in the layer of chippings.21 A similar concentration of coins in the apsis behind the main altar is observed at Høre Church. 22 The coin found in the undisturbed layer of chippings in the chancel is therefore decisive for the dating of the construction of Hedensted Church. The coin is a hitherto unknown bracteate type without inscription (Figure 10.9b, no. 4), which on the basis of material, weight and motif is ascribed to a group of bracteates minted in Northern Jutland during the reign of Valdemar I (1154/57–1182).23 The different find intensity in the layer of chipping in respectively apse and chancel illustrates that coin concentrations in the church room might not necessarily reflect liturgical actions. In the described example, the increased amount of finds in the apse are interpreted here as later disturbances in the area. The dirt layer that was formed under and between the planks of the first floor of Hedensted stone church contained 19 coins. 24 These coins are of special interest, as they represent some of the very few objects that can be classified as in situ preserved, and may be linked to the religious and devotional use of the church. These coins have

282  Gitte Tarnow Ingvardson not been moved and redeposited during the renovation and building activity in the church and were found at their original deposition place under the floor.

Context analysis The main focus of the analysis is to relate the coin finds of Hedensted Church to various building and restoration activities, in order to investigate if and/or how these activities influence the chronological and chorological profile of the coin evidence. The coin profile is related to the phases and activities documented during excavation of the church. 25 Contexts and chronology None of the coins can be connected to the earliest wooden church, and four major construction and renovation phases of the stone church have been chosen as the chronological framework for the analysis: 1 2

3

4

c. 1150–1350, includes the construction of the stone church shortly after the middle of the twelfth century and the use of the stone church with its first floor – the plank floor. This corresponds to the described phases 4 and 6. c. 1350–1500, includes the first major renovations of the church. Lead casting pits under the church floor can be connected to a renovation of the roof between the end of the thirteenth century and the beginning of the fourteenth century. A brick floor was established after the renovation work. This corresponds to phases  7 and 8. c. 1500–1650, includes the second major renovation of the church, where vault foundations and scaffold postholes demonstrate that nave and chancel were vaulted around the end of the fifteenth century. A new brick floor was established after the renovation work. This corresponds to phase 9. c. 1650–2000, includes the construction of graves in nave, chancel and tower corresponding to phase 10, and modern construction work as heating systems and rebuilding 1939, corresponding to phases 11–13.

The coin finds display an uneven distribution within the four major construction and renovation phases of the stone church. The earliest coins appear during the construction of Hedensted stone church shortly after 1150. A group of seven coins minted during the reign of Valdemar I (1157–1182) constitute the oldest coins. Four of these represent the hitherto unknown bracteate type, dated to Valdemar I (Figure 10.9a

Table 10.1 Coin finds related to building phases of Hedensted Church Construction and renovation phases

Amount of coins

Average yearly coin loss

1) c. 1150–1350 2) c. 1350–1500 3) c. 1500–1650 4) c. 1650–1000

85 (67%) 25 (20%) 13 (10%) 3 (2%)

0,43 0,17 0,09 -

Coins in church contexts  283 26

and  9b). One of these bracteates is directly connected to the construction of the stone church, as it was found in situ in the chancel in the layer of chippings that formed the base for the first plank floor. Interestingly, the vast majority of coins belong to the construction of the early stone church (phases 4 and 6) after c. 1150 and before the first large renovation around 1350. In this first phase, the church had a wooden floor. The excavation revealed a dirt layer under and between the wooden planks (Figure 10.7). The layer held large numbers of objects, which had been lost and not recovered by the parishioners of the church. This interpretation is underlined by the fact that most of the small personal objects can be dated to the twelfth century and the thirteenth century, during the use of the plank floor. Then, after c. 1350, the roof of the church was renovated. The findings of four lead casting pits in the church room demonstrate that the wooden floor most certainly was destroyed during this renovation activity (Figure 10.12). Following the renovation, a new floor was established, and from an archaeological point of view, the new floor was unfortunately made out of bricks, which do not hold the same advantageous characteristics with regard to the preservation of artefacts. There was a better chance to recover lost coins and other artefacts on a brick floor than on a plank floor. Thus, the relatively high number of finds of coins minted between c. 1150 and 1350 is probably partly due to these changes in the interior of the church and is not solely a reflection of an increased availability of coins in society in general. The majority of coins minted between 1350 and the beginning of the sixteenth century are found in disturbed contexts related to building and renovation activities. Seven coins were found in levelling layers of various floors, five were found in grave fillings, four were found in scaffold postholes, three were found in the dirt layer beneath the plank floor or in levelling layers, two were found in the vault foundation, two were found in nineteenth-century pits and one is without defined context. It is difficult to decide if these coins are lost in connection with the building and renovation activities, or if the coins were lost/deposited in the church prior to the activity, and then moved and redeposited.27 Layers younger than the sixteenth century have been removed during renovation and remodelling of the church, and all seventeenth-century and later coins are exclusively preserved in grave fillings. No coins can with certainty be categorised as grave goods. Thus, the vast majority of coins minted after 1350 had probably been moved and redeposited within the church room either in connection with building activities or in connection with the establishment of graves. Contexts and chorology The degree of preservation within the different sections varies greatly. Some layers and structures hold large amounts of artefacts, while other layers and structures are almost empty. As demonstrated in the following analysis, the degree of preservation and the extent of different layers are decisive factors in the interpretation of coin distribution in the church room. Section 7 in the southern half of the nave holds one of the largest concentrations of coins. Especially, coins from group b, dominated by Danish pennies minted 1241–c. 1340, are numerous. The section is situated just east of the south entrance, which is a likely location of an offertory trunk. The 22 coins in Section 7 were found in different types of contexts. Two were found in a levelling layer just above the foundation of the

284  Gitte Tarnow Ingvardson

Figure 10.12 One of four lead casting pits found during excavation. A residue of lead is visible in the centre of the pit. Photo N. Bendtsen, National Museum of Denmark (CC-BY-SA)

stone church, five were found in layers and a posthole connected to building activity, one is found in the dirt layer beneath the plank floor, eight were found in a levelling layer of the fourteenth-century brick floor and four were found in a levelling layer of a fifteenth-century floor. The context of the last two coins is uncertain. 28 As the contexts reveal, the main part of coins found in Section 7 was deposited during phase 4/6, between the construction c. 1150 and the first renovation c. 1350. Section 7 is one of the best-preserved areas of the church, and this must be taken into account when interpreting the coin concentration. The sections that flank the north portal of the church (Sections 4 and 6) are on the contrary some of the poorest preserved areas.29 Hardly, any original floor levels were preserved in Sections 4 and 6, as they were removed prior to the building and renovation activity. Instead, the area holds many layers and structures related to the first major renovation of the church in the first half of the fourteenth century, as, for example, lead casting pits (Figure 10.12). The differences in preservation make it difficult to establish if the high numbers of coin finds in Section 7 and the low amount of coins in Section 6 reflect differences in the interior use of the church, or the degree of preservation (Figures 10.10 and 10.11). However, the concentration of coins by the south portal most likely represents traces of an offertory trunk, and if so the coin finds may indicate accidental losses in connection with the parishioners’ donations. Section 10, which means the north-western corner of the nave, is the section most disturbed by modern activity. A tile stove, a chimney and a heating grate were situated here, and as a result, medieval layers were only preserved in situ in the south-western corner of the section. In spite of the disturbances, no less than 14 coins were found in Section 10, as older floor levels were redeposited in the modern pits and trenches. Only 2 of the 14 coins were found in a medieval context (dirt layer under the plank floor). Besides coins, the redeposited layers contained coffin fittings, a fragmented lock and fragments of window glass, lead foil and frescoes, and these finds support that the layers were formed in the church. Even though the amount of coins is not significantly higher than else were in the church, Hans Mikkelsen, head of excavations, interprets

Coins in church contexts  285

Figure 10.13 These three rare Valdemar I, 1154–1182, coins were found north of the chancel arch and possibly constitute a small offering at a side altar. Photo J. Lee, National Museum of Denmark (CC-BY-SA)

the 14 coins as evidence of a northern side altar in the nave. The argumentation is based on a high number of coins despite the disturbances and the relatively high portion of the oldest coin group (four Valdemar I, 1152–118230) (Figure 10.11).31 Three of the coins in Section 10 are a variant of a rare coin type minted by the Danish king Valdemar I, 1154–1182 (Figure 10.13).32 At present, eight single finds of the type are known in Denmark, of these two are of the same variant as the Hedensted coins. Another fragment is known from a little hoard of seven coins from Østerild Church in Northern Jutland.33 Because of the modern disturbances in Section 10, it cannot be established if the coins represent one collected deposit or if the coins were deposited one by one. However, the rarity of the coins points towards a small hoard of tree coins deposited at the same time. The small hoard may be interpreted as a personal offering by an altar, thus supporting the interpretation of the existence of a northern side altar in Section 10. A similar concentration of coins, 14 specimens, was also found south to the chancel arch in Section 11. The medieval layers in this section were well preserved, and 6 out of the 14 coins were found in the dirt layer formed under the plank floor. The composition of coins in Section 11 is fairly similar to the composition of coins in Section 10, though with a slightly younger coin profile (Figure 10.11). The excavations at Hedensted Church did not produce direct evidence of side altars. However, when the numismatic evidence is taken into account, it seems likely that the three rare Valdemar I coins should be interpreted as a personal offering, by a side altar at the north-eastern corner of the nave. Sections 10 and 11 form an interesting case with regard to how building and restoration activities influence the chorological distribution of coins. If there existed side altars in the north-eastern and south-eastern corners of the nave, the composition of the coin assemblages from Section 10 to Section 11 might provide evidence for now vanished interiors and its use in this area of the church. In addition, coins can be used

286  Gitte Tarnow Ingvardson as proxies to map post-depositional processes such as the movement of soils. In this particular case, it is interesting to note that both sections hold the same amount of coins, 14 specimens each with a comparable chronological profile. In Section 10, later activities disturbed the layers to a larger extent, whereas layers in Section 11 remained undisturbed. This particular context in Hedensted suggests that even though coins may have been redeposited, they move not far from the original spot they once were used and discarded. The dirt layer that was formed under the floorboards of the first floor of the stone church – the plank floor – is of special interest as it is the only context with documented in situ reserved coins. The dirt layer is present in large areas of the nave in Sections 5, 7–11 and thereby also constitutes a good starting point for an analysis of the distribution of coins in the nave. It must, however, be stressed that the degree of preservation of the layer varies in the different section. The dirt layer was only preserved in a small part of Section 10 and the layer was not preserved in Section 6, when this is taken into account the coins seem to be evenly distributed in the nave east of the entrances. An exception is in Section 7, where only one coin is found in the dirt layer. However as argued above, it is likely that the main part of coins in Section 7 was deposited during the phase 4/6 during the lifetime of the plank floor. It is difficult to determine if the coins in the dirt layer represent accidental losses, where the owner was not aware that he/she dropped a coin, accidental losses in connection with offerings by altars and collects or deliberate offerings between the floorboards. However, the chronological distribution of small personal objects supports that at least some coins represent accidental losses. The main part of small personal objects dates to the twelfth century and the thirteenth century, which corresponds to the time, when Hedensted Church had a plank floor. It seems likely that the small personal objects represent accidental losses during the congregation’s visits to the church, and likewise, some coins should probably also be interpreted as accidental losses. However, jewellery, buckles and knives may also have been offered in the church. The coin concentration in Section 7 and the small hoard in Section 10 also indicate accidental losses in connection with offerings by altar and offertory trunk. Floor designs and coin profiles The main part of coins found in Hedensted Church can be ascribed to the time of the plank floor (c. 1150–1350). Supported by the chronological distribution of small personal object, which is also concentrated to this period, it is likely to interpret the plank floor as a decisive factor in the chronological distribution of coins. The chronological distribution of medieval coins in Hedensted Church resembles the general distribution of coin finds in Danish churches.34 There is a limited amount of coins minted before 1241, a large group of pennies minted between c. 1241 and 1340, coins from the later part of the fourteenth century to the beginning of the sixteenth century are fewer, but with a clear dominance of hulpenninge (small bracteates), while the amount of later coins are very limited. If the change from plank floor to brick floor had a decisive influence on the coin profile in Hedensted Church, this may also be the case in other churches. This questions if church coins are representative of the general monetary development, as argued by Klackenberg.35 Danish numismatists are blessed with a vast and fast-growing body of coin finds by metal detector surveys, and this find group serves as an excellent mirror of coin use

Coins in church contexts  287

Figure 10.14 The chronological distribution of all 1461 coins, which were declared treasure trove from May 2012 to May 2013 at the National Museum of Denmark. Based on data facilitated by Line Bjerg, National Museum of Denmark.

in non-ecclesiastical contexts, as the main part of detector found coins is located in the countryside. A complete and up-to-date record of the detector finds in Denmark is not available, but the general trends are visualised by the inventory of coins, which were declared treasure trove at the National Museum of Denmark from Maj 2012 to Maj 2013 (Figure 10.14).36 The large numbers of Danish pennies minted c. 1241– 1340 stand out as particularly rich. In this period, millions of debased pennies were minted in Denmark, and the waste number of coins in circulation is clearly reflected in the coin body both inside and outside the churches.37 In Hedensted Church, 86% of the coins are minted c. 1241–1340, and among the treasure trove coins, 59% are minted in this period. Thus, the very high proportion of coins minted c. 1241–1340 in Hedensted Church is likely due to changes in the general monetary situation as well as changes in the floor design from plank floor to brick floor. Another noteworthy trend is the total lack of Danish pennies minted during the reign of Valdemar I, Knud VI and Valdemar II c. 1150–1241 among the detector finds from Maj 2012 to Maj 2013 (Figure 10.14). In Hedensted Church, around 17% of the coins are minted in this period. Even though a national coinage was initiated by Cnut the Great around 1020, the Danish monetary process was still in its youth around 1150, when the construction of Hedensted stone church was initiated. Surveys on detector, church and excavation coin finds in eastern Denmark show that very few coins were in circulation in the first half of the twelfth century, and coins were most common in city and church contexts.38 Around 1150, there is a marked rise in coin use, and coins do now also appear in rural contexts. However, outside the main urban centres and their hinterland, the main part of the coin finds is still from churches. The last trend that will be approached here is the presence of hulpenninge in Hedensted and other churches and the lack of the same among the detector finds. This

288  Gitte Tarnow Ingvardson discrepancy has earlier been explained by the fact that hulpenninge were minted on small and very thin planchets, which are difficult to locate with metal detectors and are easily destroyed in the fields.39 New surveys on coin finds from two redundant churches do, however, partly dismiss this explanation. The coin material from the redundant church sites had been disturbed by ploughing in the same way as many other detector sites, but still contained hulpenninge, though not in as large numbers as in church finds. This demonstrates that the lack of hulpenninge on detector sites is not exclusively due to the condition of preservation, but in fact reflects a difference in coin use in respectively ecclesiastical and secular contexts.40 So far, surveys have only been applied on two redundant churches, but this promising new method gives hope that future research will solve the question on why medieval hulpenninge are almost exclusively found in hoards and churches in Denmark. Thus, with the present evidence, it can be concluded that despite the marked drop in the numbers of coin finds after the change from plank floor to brick floor in Hedensted Church, the fifteenth-century hulpenninge are over-represented in churches in comparison with non-ecclesial sites. Finally, a comment should be made on coins minted after the Danish Reformation in 1536. Both in Denmark and in Sweden, the volume of church coins is diminished after the beginning of the sixteenth century, which by Klackenberg is interpreted as the result of changes in the liturgical practices after the Reformation.41 Unfortunately, a comparison to the Danish metal detector finds is not possible, as coins younger than 1536 are not declared treasure trove, and is therefore unfortunately not handed in and recorded at the National Museum of Denmark. It seems likely that the changes in the liturgical practices after the Reformation are the main reason for the marked drop of coins after the beginning of the sixteenth century. However, the archaeological evidence from Hedensted Church does point to the fact that layers younger than the sixteenth century were removed during the renovation in 1939 and thereby not viable for interpretation.

Church coins in/out of contexts As the investigation of Hedensted Church has demonstrated, the chronological and chorological coin profile is formed and influenced by building and renovation activities, changes in floor designs, the interior design of the church as well as the general monetary development and liturgical and perhaps also other practices in the church room. How the coin profile is influenced by building and renovations activities and is dependent on the nature of activities. It has been exemplified that even though coins often are recontextualised by the establishment of graves, the survival rate may be larger in areas with graves than without graves, as illustrated by the concentration of early modern coins, dated to c. 1550–1700 (group d.) in the chancel.42 With regard to the spatial distribution of coins, graves as well as later modern disturbances as the tile stove, the chimney and the heating grate in Section 10 had recontextualised the coins, but probably not moved the coins far away from their original deposition place. This is demonstrated by the similar coin profiles in respectively the very disturbed Section 10 and the undisturbed Section 11, located on each side of the chancel arch. Levelling and building layers may on the other hand heavily affect both the chronological sequence and the chorological distribution of coins, as demonstrated by the conditions in Section 5, where the layer formed during the renovation activities in

Coins in church contexts 289 1939 carried a large number of medieval objects. Finely, does the removal of earth of course have a marked influence on both the chronological and chorological distribution of coins. This is demonstrated by the low amount of coins in Sections 4 and 6 in the northern half of the nave, where earth was removed in connection with the establishment of lead casting pits, and perhaps is the lack of coins later than the beginning of the sixteenth century also partly due to the removal of floor layers during the renovations in the seventeenth century and the twentieth century. Supported by the chronological distribution of small personal objects, it is substantiated that changes in the floor design influence the coin profile and that one may expect a larger percentage of coins from churches with plank floor than from churches with brick floors. Thus, in Hedensted, the large numbers of thirteenth- and fourteenth-century coins may partly be due to the fact that the church was equipped with a plank floor in this period. However, when related to the general monetary development illustrated by the metal detector material, it is clear that trends in the bulk of coins in society in general are manifested in the church finds. Thus, a general rise in coin use is evident both inside and outside churches from the middle of the thirteenth century to the middle of the fourteenth century. However, in the comparison between church coins and metal detector coins, it is clear that churches do form a special environment with regard to coin use in the Middle Ages. The difference is clear among coins minted before the middle of the thirteenth century and among the group of fifteenth-century bracteates hulpenninge. Both coin groups are evidently more widespread in church contexts than secular contexts. This questions if church coins are the ideal material evidence in surveys concerning the general monetary development. It has proven very difficult to use the coin material as evidence of the interior design and liturgical practices in the church room. Exceptions are the coin concentration in Section 7 that implies the presence of an offertory trunk at the southern entrance, and the small hoard of rare Valdemar I coins that may be interpreted as a personal offering by a side altar. As younger layers were removed during restorations in the seventeenth century and twentieth century, it is impossible to use the Hedensted coins to either support or reject the interpretation made by Klackenberg that the general lack of coins minted later than c. 1500 is due to changes in the liturgical practices after the Reformation. The complexity of the well-documented archaeological evidence from Hedensted is striking. No interpretation seems straightforward. In some areas, coin concentrations may be due to liturgical action in other areas due to building activities and yet in other areas due to modern disturbances. The chronological sequences are influenced by building activities, monetary trends as well as liturgical and other practices in the church room. The analyses presented in this contribution clearly demonstrate that contexts are essential when interpreting church coins. Thus, the evidence of Hedensted Church calls for caution when interpreting the waste majority of Danish church coins, which are in fact characterised by their lack of contexts.

Notes 1 The results from the excavation in Hedensted Church are published in Hedensted Kirke. Undersøgelser og restaureringer, edited by Michael Andersen and Hans Mikkelsen, Nordiske Fortidsminder, serie B, vol. 26 (Odense: Syddansk Universitetsforlag, 2016). The

290

2

3 4 5 6 7 8 9 10 11 12 13

14 15

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

Gitte Tarnow Ingvardson

volume contains a discussion of the coin finds Gitte T. Ingvardsson, “Sjælefrelse eller lommespild – møntfundene fra Hedensted kirke,” 79–97. “Hedensted kirke, Hattings herred,” in Danmarks kirker, Vejle amt, vol. 18, edited by Ebbe Nyborg and Niels Jørgen Poulsen (Copenhagen: Nationalmuseum, 2014), 1825–82. A detailed description of the construction, interior and frescoes of Hedensted Church is also available online on Danmarks Kirker hosted by The National Museum of Denmark: http://danmarkskirker.natmus.dk/vejle/hedensted-kirke/ “Hedensted kirke, Hattings herred,” 1824, fig. 17–18. “Hedensted kirke, Hattings herred,” 1846–53, fig. 25–28. “Hedensted kirke, Hattings herred,” 1866, fig. 48. “Hedensted kirke, Hattings herred,” 1864, fig. 45. “Hedensted kirke, Hattings herred,” 1828–70. Andersen and Mikkelsen (eds.) Hedensted Kirke. Undersøgelser. Jes Wienberg, Den gotiske labyrint. Middelalderen og kirker i Danmark (Stockholm: Almqvist & Wiksell International, 1993, 9–12). Henrik Klackenberg, Moneta Nostra. Monetarisering i medeltidens Sverige, Lund Studies in Medieval Archaeology, vol. 10 (Lund: Almqvist & Wiksell International, 1992), 35–38. Rensbro, this volume, ch. 3. Rensbro, this volume, ch. 3. Nanna Holm Bendtsen and Hans Mikkelsen, Rapport fra arkæologisk undersøgelse i Hedensted kirke, Hatting Herred, Vejle Amt i perioden 4. oktober 2007–14. marts 2008 (Copenhagen: Nationalmuseet, 2009). For an overview of the sections mentioned in the text, see Figure 10.10. “Hedensted kirke, Hattings herred,” 1828–29. Baptismal wells are known from other early Scandinavian church buildings and should probably be seen in connection with adult baptisms during early Christianisation, Maria Cinthio, “Trinitatiskyrkan, gravarna og de första lundaborna,” in Kristendommen i Danmark før 1050: et symposium i Roskilde den 5.-7. februar 2003, edited by Niels Lund (Roskilde: Roskilde Museums Skrifter, 2004), 168–69; Nanna Holm Bendtsen and Hans Mikkelsen, “En trækirke i Hedensted?” in Hedensted Kirke, 9–28; Göran Tagesson, “Vreta före Cistercienserna. Huvudgård, kyrka och en (o)möjlig dopanläggning?” in Fokus Vreta kloster. 17 nya rön om Sveriges äldsta kloster, edited by Göran Tagesson, Elisabet Regner, Birgitta Alinder and Lars Ladell, “The Museum of National Antiquities, Studies 14,” Riksantikvarieämbetet Arkeologiska Undersökningar Skrifter 77 (2010): 103–104; “Hedensted kirke, Hattings herred,” 1829–30. Nanne H. Bendtsen and Hans Mikkelsen, “Kirkens brug – et genstandskatalog,” in Hedensted Kirke.Undersøgelser, 67–76. A complete list of coins in Ingvardson, “Sjælefrelse eller lommespild,” 78–96. Bendtsen and Mikkelsen, Rapport fra arkæologisk, layer A205 and A208. Rensbro and Rensbro and Moesgaard, this volume, chs. 3 and 9. Bendtsen and Mikkelsen, Rapport fra arkæologisk, 27. Hans Mikkelsen, “En romansk stenkirke bliver til og forandres – arkæologiske resultater,” in Hedensted Kirke, Undersøgelser, 31–65. See Risvaag, this volume, ch. 6. Ingvardson, “Sjælefrelse eller lommespild,” 79–97. Coin list x-no.: 153, 320, 326, 364, 365, 390, 391, 392, 422, 437, 445, 446, 497, 658, 415, 509, 520, 521 and 522 in Ingvarson, “Sjælefrelse eller lommespild.” Mikkelsen, “En romansk stenkirke,” 31–65. Ingvardson, “Sjælefrelse eller lommespild,” 86–90. Coins found in scaffolding post and in grave fillings as indicators of building and burial activities are discussed by Myrberg Burström, this volume, ch. 11. Bendtsen and Mikkelsen, Rapport fra arkæologisk, 30. Bendtsen and Mikkelsen, Rapport fra arkæologisk, 26–28. Ingvardson, “Sjælefrelse eller lommespild,” Coin list x-no: 322, 324, 334 and 373. Bendtsen and Mikkelsen, Rapport fra arkæologisk, 32. Peter Hauberg, Myntforhold og Udmyntninger i Danmark indtil 1146 (Copenhagen, 1900), Table III no. 58.

Coins in church contexts 291 33 Ingvarson, “Sjælefrelse eller lommespild”; Jørgen Steen Jensen, Kirsten Bendixen, NielsKnud Liebgott and Fritze Lindahl, Danmarks middelalderlige skattefund c.1050–c.1550 (Copenhagen, 1992), 298. 34 Kirsten Bendixen, “Middelaldermønt i de sidste 10 års danske kirkefund,” Nordisk Numismatiske Årsskrift (1972), 54; Keld Grinder-Hansen, Kongemagtens Krise. Det danske møntvæsen 1241–1340 (Copenhagen: Museum Tusculanum Press, 2000), 165–67. 35 Klackenberg, Moneta Nostra, 34–40. 36 I am very grateful to Line Bjerg for allowing me access to this unpublished data. 37 Grinder-Hansen, Kongemagtens Krise. 38 Peter Carelli, en Kapitalistisk anda. Kulturella förändringar i 1100-tallets Danmark (Stockholm: Almqvist & Wiksell International, 2001), 192–201; Cecilia von Heijne, “Särpräglat, Vikingatida och tidligmedeltida myntfnd från Danmark, Skåne, Blekinge och Halland (ca 800–1130),” Stockholm Studies in Archaeology 31 (Stockholm, 2004): 146– 47, 156; Nils Hybel and Bjørn Poulsen, The Danish Resources c. 1000–1550, Growth and Recession (Leiden – Bosten, 2007), 330–31; G. Ingvardson, Møntbrug – Fra vikingetid til vendertogter (Aarhus, 2010), 41–44, 58–61; Bjørn Poulsen, Møntbrug i Danmark 1100– 1300, Fortid og Nutid XXVIII (1979): 283–85. 39 Keld Grinder-Hansen, “Dansk møntcirkulation i anden halvdel af 1300-tallet – nogle metodiske bemærkninger,” in Ord med mening. Festskrift til Jørgen Steen Jensen. 1. november 1998, edited by Jens Christian Moesgaard and P. Nielsen (Taastrup, 1978), 38; Keld Grinder-Hansen, “Mønter,” in Vor Frue Kloster – et benediktinernonnekloster i Randers, edited by Hans Mikkelsen (Randers: Jysk Arkæologisk Selskabs Skrifter, 2002), 68. 40 Jens Christian Moesgaard, “The Redundant Churches of Oldrup and Uld, Jutland, Denmark. Geomagnetical Surveys and Metal Detector Finds,” Journal of Archaeological Numismatics 8 (2018): 175–206. 41 Klackenberg, Moneta Nostra, 38. 42 See also similar conclusion Myrberg-Burström, this volume, ch. 11.

11 Building for glory Coins in the houses of the Lord in Gränna and Arby, Sweden Nanouschka M. Burström

Introduction This chapter focuses on two general issues of relevance to the investigation of coin finds in church buildings based on case studies of two medieval1 churches, Arby and Gränna, in the province of Småland, Sweden.2 Firstly, there is the issue involving the relationship between coin use in churches and the ownership of the church, which could be in the hands of either private patrons or parish congregations. What differences in coin use might have resulted from the different ways a church might be managed? Can the coins help us to see such changes? Of central importance to this line of investigation is the notion that coins reflect and are part of their historical context, but also affected people’s everyday and religious lives (e.g. used for indulgences, masses, embellishment of church space, opening up interior space and power relations). Secondly, there is the relationship between coin finds from churches and general coin use in society. How should differences in coin find patterns and fluctuations in the intensity of deposition be interpreted? To what extent are patterns from coin finds in churches useful for extrapolating and for interpretations of the role of coins outside of that specific context? Here, evaluations of the depositional processes and of the composition of the coin finds are of great consequence. This line of enquiry seeks to push beyond the common view that the church is “a society in miniature,” partly accepting it, but also underlining the unique character of the ecclesiastical context and of the events taking place in it, and thus the particular conditions for artefacts found there. The results of this study of two Småland churches indicate, firstly, that coin use inside churches in Sweden only took off in a significant way with the formation of parishes in the thirteenth century, when congregations took over the responsibility for the church. Secondly, that coin use in churches was largely connected with offerings, and with certain events, such as the celebration of a designated saint’s day and fundraising for building projects. This raises the question of whether, and in what ways, fluctuations in the coin material found in ecclesiastical contexts may be used to discuss phenomena outside of the church walls.

Studying Arby and Gränna churches and their coin finds – some starting points The investigated sites Arby church (Figure 11.1), probably founded sometime around AD 1100, is situated close to Arby village in Småland, on the eastern coast of Sweden. The area is

Building for glory – Gränna and Arby  293 characterised by its position between the inland woods and the Baltic Sea coast, and in the Middle Ages, it depended economically on agriculture, cattle breeding and the production and transport of iron. The area is rich in ancient monuments from all periods, not least from the Viking Age, and the high number of medieval round (e.g. Hagby, Voxtorp) and fortified churches (e.g. Arby, Halltorp, Kläckeberga) is a distinctive feature of the area.3 Arby church is beautifully situated on the riverside, while the village is further away from the water, probably moved from its original downstream position at some point in the Middle Ages. The distance between the village, in both its present and its assumed earlier locations, and the church indicates that the latter was originally built as a private church.4 The name Arby, meaning “the village by the river,” was formed in the Viking Age, and the church follows the common countryside pattern of being named after the community where it was built. The earliest written record of this name for the church is from 1346.5 Gränna church (Figures 11.2 and 11.3) was initiated around AD 1150 and is situated in the small, eponymous, seventeenth-century town, just beneath a fault scarp on the shore of Lake Vättern, one of Sweden’s largest bodies of water. The church sits on a slope where it overlooks the town, the lake and a small harbour as well as the ancient main road leading from the mid-eastern Swedish provinces to the south. The cultural landscape is rich in heritage from both prehistory and history. During the Middle Ages, exceptionally large tracts of land in the area were in the possession of large landowners, the king and the Church. In Lake Vättern, the island of Visingsö boasts several early medieval stone churches and a medieval royal castle, probably built on or near the site of the older royal estate, Husaby. The husabyar formed an extensive network of sites in Scandinavia, of which most seem to have been established in the Viking Age, although the formalised institution may belong to the twelfth century.6 They were part of the economic assets of the early royal institution and used well into the thirteenth century as bases for the king when travelling, a necessary means to uphold royal power.7 The spot where today’s

Figure 11.1 Exterior of Arby church (southern façade) in 2005. Photograph by the author

294  Nanouschka M. Burström

Figure 11.2 Exterior of the present Gränna church. Photo: Gränna parish. Used with permission

Figure 11.3 Gränna church in 1875. In the foreground, the later farmstead on the medieval Husaby site, on the stretch/plateau between the scarp (in the background) and Lake Vättern. Xylography by Wilhelm F. Meyer. Photo: KMB. Public domain

Building for glory – Gränna and Arby  295 Gränna grew was another such manorial site. While gränna means something like “the lane” and originally denoted the whole narrow stretch of seashore beneath the scarp, the main farmstead was likely called Husaby in the early medieval period, since this name is retained for certain plots on the town grounds adjacent to the church, which was probably built on Husaby land. 8 The earliest written records for “Gränna” are as follows: denoting the area, 1248–12669; in relation to the Husaby estate, 131510 ; denoting the parish, 1375; and in direct connection with the church, twice in 1470.11 Comparing the two sites, Arby is peacefully situated along a river near the eastern coast of Sweden, and Gränna is located more dramatically on the shore between a major inland lake and a scarp. At a local level, those areas have a somewhat different history with regard to their sustenance and proximity to the central powers of king and Church, and they are also divided by a tract of highland and woods. But, in other respects, they share history; for example, in the Middle Ages, they were part of the same diocese, Linköping, and judicial province (lagsaga), Östergötland. These are important aspects for the interpretation of the composition of coin finds. Both churches were built in well-established sites in the Early Middle Ages, and seem to have been initiated by individual landowners or local elite. Only with the written evidence of the fourteenth century do we know more securely their names and what community they served at that time. The finds from these two churches (out of many possible ones) were selected for the present study mainly based on considerations about their setting (i.e. medieval Swedish mainland parish churches) and the thoroughness of the original excavations and documentation.12 The quality and scope of extant documentation and excavation reports were of utmost importance so as to enable the objectives of this study.13 A fair number of finds and the presence of early (i.e. twelfth- and thirteenth- century) coins were also important, to trace the process of change between the Early and High Middle Ages, and to increase the reliability of the conclusions drawn while allowing analysis within the available time frame. Enabling comparisons to be made between the two sites, they will be presented and discussed both parallelly and jointly, with summaries and conclusions provided throughout the text. The excavations Arby church was excavated and the walling documented in the 1970s, in connection with extensive restoration works. A total of 142 coins and about 250 other finds were registered (inventory no. SHM 30040) (Figure 11.4). The church was excavated in its entirety and carefully documented in ground plans and photography (Figure 11.5). There was no recording of coordinates or squares but rather of contexts, which were described and marked on the plans, e.g. “mortar layer” and “posthole” (my translations). Vertical data or stratigraphy was not recorded as a rule, but height indications for, e.g., foundation boulders and the bottom of postholes were noted on the plans. Rather, detailed information was given as to the precise find context (e.g. to building layers or particular features such as postholes).14 As the coins and other finds were listed separately, there are sometimes two records (one coin and one “other” find) with the same find number. Furthermore, standard procedure in the coin cabinet (KMK) is to reorder the coins according to the numismatic principles; every coin thus has one excavation find number and

296  Nanouschka M. Burström

Figure 11.4  Selected finds from Arby church (SHM 30040). Left: mounting, date?, find no. 52. Top right: clasp, c. 1300, find no. 36; below: hymn book clasp with cherub’s head, seventeenth–eighteenth centuries, find no. 31. Not to scale. Photo: Ola Myrin, SHM. Used with permission.

one museum coin number. This chapter refers to find numbers throughout. Unfortunately, the detailed descriptions of the graves were not included in the published excavation report; also, it was not possible to locate this information in the ATA at the time this article was written. Therefore, the exact position of coins in the graves could not be established for present purposes, though this would have contributed an increased understanding of the relation between the coins and individuals. The excavator, however, comments that “many coins lay in the filling […] this regards in particular coins found in and adjacent to late graves.”15 If the coins were found mainly in the filling, the exact positions of coins in the graves may not be decisive to the argument here. Gränna church was excavated in 1987. Like in Arby, this was prompted by restoration works, and the aim was to better understand the layout, building history and use of the medieval church. A total of 126 coins were found, and there were 422 records registered in all; the finds are now divided between the JLM, the KMK and the SHM (Figures 11.6 and 11.7). Most of the foundations and lower parts of the medieval walling were found under the floor of the standing nineteenth-century church and could be documented. The surfaces of the interior were not extensively excavated, but several trenches were opened and excavated to various depths, and vertical (“z”) data were recorded (Figure 11.8). The excavation was documented in ground plans and photographs. Finds positions were mainly indicated by trench (in most cases c. 2 × 2 m 2) or profile, and by layer and height level (find list in ATA). Many finds are also

Building for glory – Gränna and Arby  297

Figure 11.5 a)  Arby church, excavation plan. Map: Steinar Kristensen, UiO: Museum of Cultural History after Andersson & Svennebring (1989, Figure 43); b) Arby, church, coin finds 1220-1584 and 1632-1827. Map by Steinar Kristensen, UiO: Museum of Cultural History after Andersson & Svennebring (1989, Figure 43).

individually discussed in the report. Excavation find numbers include coins and other finds in the same series.16 The above considerations of the conditions for the study of sample, choices and quality of documentation are vital when attempting to understand past excavation sites and to translate old excavation data into modern digital analysis and presentation. The aim here is to use those investigations, different as their original purposes may have been, to shed light on the research issues of the present study. While it is generally rewarding to have a fresh look at old excavations and coin finds and approach them with new questions, new historical knowledge or new theoretical frameworks, the quality and outcome of the enquiry also depend on factors such as those mentioned above.17

298  Nanouschka M. Burström

Figure 11.6  S elected finds from Gränna church (SHM dnr 198/87). Top left: candleholder, date?, find no. 167. Top right: ring clasp, date?, find no. 261. Below: comb, date?, find no. 247. Not to scale. Photo: Ola Myrin, SHM. Used with permission

Figure 11.7  Gränna church, fifteenth-century crowbar. JLM dnr 198/1987, find no. 286. Photo: Göran Sandstedt, JLM. Used with permission.

Building for glory – Gränna and Arby  299

Figure 11.8  Gränna church, foundations of the Romanesque (brown) and present church, and the excavated trenches. Map: Steinar Kristensen, UiO: Museum of Cultural History after Varenius (1993, Pl. 2 and 3).

The coin finds One obvious difference between Swedish medieval coins and the post-Reformation ones is the shift of size (diameter and weight), which increased rapidly from the ­sixteenth century. From small coins of about 0.2–0.5 g and 1–1.5 cm in diameter, coins of c. 2.5 cm and 2.5–6.0 g became common, some coins weighing even 13 or 14 g (Table 11.1). Also, the material and visual impression of coins changed, as the medieval coins are all struck in silver while, for the most part, the post-Reformation coins are copper. These are general observations that are not just valid for the coins

300  Nanouschka M. Burström Table 11.1 Coin finds from Arby church Fno

KMK Country no.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

92 16 61 64 58 109 43 89 144 121 35 34 2 114 107 11 47 119 111 120 108 124

23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

40 128 27 52 100 104 72 41 65 126 48 44 115 96 13 14 74 105 50 125

Sweden Denmark Baltic? Sweden Sweden Sweden Sweden Sweden Sweden Denmark Sweden Sweden Denmark Sweden Sweden Sweden Sweden Sweden Sweden Denmark

43 44 45 46 47

29 9 135 51 94

Sweden Germany Sweden Sweden

Mint/area

Issuer

Denomination Period/year

Avesta

Fredrik I Magnus Eriksson Karl XI Karl XI Karl XI Karl XIV Johan Kristina Fredrik I

1 öre SM Penning 1/6 öre 1/6 öre 1/6 öre 1/4 skilling 1/4 öre 1 öre KM

1750 1354–1363 1668 1666 1666 1820 1636 1720

Skilling Penning Penning Valdemar Penning Örtug Gustav IV Adolf ½ skilling Penning Kristina 1/4 öre Örtug Karl XIV Johan 1/4 skilling Hans Hvid Karl XIV Johan ½ skilling Erik of Pomerania Penning

1520 1400–1450 1400–1450 1250–1275 1420–1440 1802 1400–1425 1641 1420–1440 1825 1481–1513 1827 1405–1424

Johan III Kristian II

1584 1513–1523 1360–1520 1666 1807 1803 1660–1697 1635 1670 1448–1481 1635 1635 1420–1440 1721340–1360 1320–1340 1671808 1671 1396–1439

Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden

Avesta Avesta Avesta Avesta Nyköping/Säter Stockholm

Denmark Germany Germany Sweden Denmark Sweden

Gotland, Visby Mecklenburg Mecklenburg Götaland Gotland, Visby Avesta

Sweden Denmark Sweden Denmark Sweden Denmark

Säter Gotland, Visby Avesta Malmö Avesta Naestved and Lund Stockholm Malmö Avesta Avesta Avesta Avesta Nyköping/Säter Avesta Malmö Avesta Nyköping/Säter Gotland, Visby Stockholm

Avesta Avesta Stockholm Naestved and Lund

Pomerania Stockholm Stockholm

Fyrk Klipping

Karl XI Gustav IV Adolf Gustav IV Adolf Karl XI Kristina Karl XI Kristian I Kristina Kristina

1/6 öre ½ skilling 1/12 skilling 1/6 öre 1/4 öre 1/6 öre Hvid 1/4 öre 1/4 öre Örtug Fredrik I 1 öre KM Magnus Eriksson Penning Magnus Eriksson Penning Karl XI 1/6 öre Gustav IV Adolf 1/12 skilling Karl XI 1 öre Erik of Pomerania Penning Penning Penning Karl XI Fredrik I

1 öre 1 öre KM

1300–1325 1300–1325 1300–1399 1672 1724

Building for glory – Gränna and Arby  301 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89

70 99 84 123 67 87 97 129 106 59 143 66 63 69 77 76 55 112 62 53 73 75 86 71 82 101 81 88 83 95 103 90 18 142 60 98 28 85 10 3 78 7

90 91 92 93 94 95 96 97 98 99

5 1 4 45 91 38 42 46 39 137

Sweden Sweden Sweden Denmark Sweden Sweden Sweden Denmark Sweden Sweden

Avesta Stockholm Stockholm Naestved Avesta Stockholm Stockholm Malmö Avesta Avesta

Karl XI Fredrik I Ulrika Eleonora Erik of Pomerania Karl XI Ulrika Eleonora Fredrik I Kristian II Gustav IV Adolf Karl XI

1/6 öre 1 öre KM 1 öre KM Penning 1/6 öre 1 öre KM 1 öre KM Klipping 1/4 skilling 1/6 öre

1676 1725 1719 1405–1439 1673 1720 1724 1513–1523 1799 1666

Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden

Avesta Avesta Avesta Stockholm Stockholm Avesta Avesta Avesta Avesta Avesta Avesta Stockholm Avesta Stockholm Avesta Stockholm Stockholm Stockholm Stockholm Avesta Stockholm

Karl XI Karl XI Karl XI Karl XII Karl XII Karl XI Karl XIV Johan Karl XI Karl XI Karl XI Karl XII Ulrika Eleonora Karl XI Ulrika Eleonora Gustav IV Adolf Ulrika Eleonora Ulrika Eleonora Ulrika Eleonora Fredrik I Gustav IV Adolf Ulrika Eleonora Magnus Eriksson

1/6 öre 1/6 öre 1/6 öre 1 daler SM 1 daler SM 1/6 öre 1/4 skilling 1/6 öre 1/6 öre 1/6 öre 1/6 öre 1 öre KM 1/6 öre 1 öre KM 1/4 skilling 1 öre KM 1 öre KM 1 öre KM 1 öre KM 1/12 skilling 1 öre KM Penning

1671 1668 1675 1715 1715 1666 1825 1668 1666 1666 1708 1719–1720 1680 1719 1802 1719 1720 1719 1724 1802 1719 1354–1363

Sweden Sweden

Avesta Stockholm

Karl XI Fredrik I

1/6 öre I öre SM

Ulrika Eleonora

1 öre KM Penning Penning 1 daler SM Half penning

1666 1725 1375–1400 1719 1300–1325 1250–1275 1715 1275–1290

Penning Penning Penning 1/4 öre 1 öre SM ½ öre 1/4 öre 1/4 öre 2 öre Pfennig

1250–1275 1220–1250 1250–1275 16-1739 1564 1635 16-1572 1300–1399

Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden

Stockholm Götaland

Säter Avesta Stockholm Nyköping/Säter Säter Stockholm

Valdemar Karl XII Magnus Birgersson Valdemar Erik Eriksson? Valdemar Kristina Fredrik I Erik XIV Kristina Kristina Johan III Teutonic Order

(Continued)

302  Nanouschka M. Burström Fno 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144

KMK Country no. 33 139 23 24 56 110 25 ? 102 79 140 68 8 80 30 57 138 113 37 20 17 136 21 130 116 118 22 133 36 12 134 122 54 127 31 117 132 19 131 32 6 141 93 15 49

Mint/area

Issuer

Denomination Period/year

Germany Germany Sweden

Mecklenburg Pomerania?

Sweden Sweden

Avesta Avesta

Karl XI Karl XIV Johan

1/6 öre 1/4 skilling

Sweden Sweden Sweden

Götaland Avesta Stockholm

Valdemar Gustav IV Adolf Ulrika Eleonora

Penning 1/4 skilling 1 öre KM

Sweden Sweden Sweden

Avesta

Karl XI

Stockholm

Ulrika Eleonora

1/6 öre Penning 1 öre KM

Avesta Elbing Gotland, Visby Västerås

Karl XI

1/6 öre

Sweden Germany Denmark Sweden Sweden Sweden Sweden Denmark Denmark Denmark Sweden Germany Germany Sweden Germany Sweden Sweden Norway Germany Denmark Germany Sweden Germany Germany Sweden Sweden Sweden Sweden Sweden

Malmö Gotland, Visby Gotland, Visby Parchim Mecklenburg Stargard Gotland Avesta Mecklenburg Gotland, Visby Wismar Wismar Mecklenburg

Stockholm

Penning Pfennig Penning

Örtug ½ örtug Penning Penning Pfennig Penning Klipping Örtug Örtug Magnus Eriksson Penning Viertelwitten Penning Birger Magnusson Penning Pfennig Penning Karl XI 1/6 öre Hans Hvid Penning Örtug Viertelwitten Magnus Eriksson Penning Viertelwitten Penning Valdemar Penning Magnus Eriksson Penning Fredrik I 1 öre KM Magnus Eriksson Penning Karl XI 2 öre Sten Sture t.e. Magnus Eriksson Magnus Eriksson Teutonic Order Magnus Eriksson Kristian II

1400–1450 1400–1499 1300–1325 1475–1500 1666 1825 1300–1325 1250–1275 1806 1719 1700–1799 1674 1300–1325 1719 1400–1425 1666 1457 1390 1470–1503 1354–1363 1354–1363 1300–1399 1354–1363 1513–1523 1420–1440 1420–1440 1354–1363 1389 1400–1450 1290–1318 1300–1399 1220–1250 1667 1481–1512 1400–1450 1420–1440 1379 1319–1364 1379 1400–1450 1250–1275 1354–1363 1721 1320–1340 1664

Building for glory – Gränna and Arby  303 Arby church, coins per century (∑=142) 40 35 30 25 20 15 10 5 0

34 27

9

27

24

1

7

12

1

Figure 11.9  Number of coins per century, Arby church.

from the sites included in this study, and they are helpful to bear in mind when looking at worn or corroded coins, thus enabling an educated guess as to their rough age. The 142 coins (SHM 30040) found in Arby church (Table 11.1) span a time frame from the 1220s to 1827 (c. 660 years), an average number of 0.2 coins per year (i.e. one coin every five years). In reality, however, the deposition or loss rate was not that even; rather, there are obvious clusters of coins to take into account (Figure 11.9). For example, 24% (34 specimens) of all the coins found originated in the seventeenth century; 42% (60 specimens) of the coins were medieval (in practice, c. 1220 until c. 1550); and of those, 45% (27 specimens, 19% of the total coin finds) originated in the fourteenth century. Actually, what stands out is the relatively low numbers deriving from the thirteenth, sixteenth and eighteenth centuries. The first instance could be due to a generally low number of circulating coins, the period being rather early from a monetisation point of view. The nineteenth century, on the other hand, was surely monetised, but as there are only coins present from the first quarter of the century, the number remains low. The low numbers from the sixteenth century remain to be explained. Kenneth Jonsson demonstrates a general decrease in coin circulation in the fifteenth century; this, however, is not visible in Arby.18 The earliest coins found were from the Swedish king Erik Eriksson (1220–1250) (Figure 11.10:5)19 and one possibly from Gotland with about the same dating (Table 11.1, nos. 91, 131). The remaining thirteenth-century coins are from Swedish kings Valdemar Birgersson (1250–1275) (nos. 16, 90, 93, 95, 110, 143), Magnus Birgersson (1275–1290) (no. 89) and Birger Magnusson (1290–1318) (no. 132). 20 It is thus fair to say that coin use in the church took off in a significant way around AD 1250, but it should be kept in mind that in reality, it mainly pertains to the coins from one particular issuer (Valdemar, 1250–1275). Of the 60 medieval coins in the find, most (29 specimens, 48% of the medieval coins) originated in mainland Sweden. They are supplemented by 13 coins from Germany, nine from Gotland and nine from Denmark. The post-Reformation coins were all Swedish. The latest coin was one of several from King Karl XIV Johan, dated 1827 (no. 21). Three coins from the eighteenth century were struck on slightly earlier coins (nos. 8, 53 and 79). There were also two perforated copper flans – spangles – not

304  Nanouschka M. Burström

Figure 11.10 A selection of coins found in Arby church. From top left to bottom right: 1: Sweden, Avesta mint, Karl XI, 1/6 öre 1676 (Cu, 6.16 g., no. 48); 2: Sweden?, spangle, reused coin? (Cu, 2.77 g., perforated, no. 58); 3: Sweden, Avesta, Karl XIV Johan, 1/4 skilling 1825 (Cu, 3.69 g., no. 65); 4: Denmark, Malmö mint, Kristian II, klipping 1513–23 (square, Cu, 1.99 g., no. 55); 5: Sweden, Erik Eriksson?, penning 1220–50 (Ag, 0.14 g., perforated, no. 91); 6: Sweden, Gotland?, penning 1220–50 (Ag, 0.08 g., no. 131); 7: Sweden, Valdemar, penning 1250–75 (Ag, 0.14 g., no. 90); 8: Sweden, Magnus Birgersson Ladulås, halvpenning 1275–90 (Ag, 0.06 g., no. 89); and 9: Germany, Mecklenburg, pfennig 1400–50 (Ag, 0.20 g., no. 139). Nos. 91, 90, 89 and 139 are bracteates. Photo: Ola Myrin, KMK

listed as coins, but which, quite obviously, were made out of coins or token money (German Rechenpfennige) (nos. 9 and 58). In Gränna church, 126 coins (KMK 102 490) were found, ranging from about AD 1065 until 1844, a period of c. 780 years (Table 11.2). This produces an average of 0.16 coins (remaining) per year, or one coin every six years. Also noteworthy in this church is the fact that coins were deposited less evenly (Figure 11.11). A few early coins were found: a Norwegian Viking Age coin (dated c. 1065–1080; Table 11.2, no. 1); a French coin from the twelfth century (1157–1200; no. 21) (see Figures 11.12:5, 11.12:2); and one Gotlandic coin from the thirteenth century (Table 11.2, no. 10), possibly from c. 1220 to 1250. The numbers increase only from the 1250s onwards, with coins from King Valdemar (1250–1275) (nos. 12, 13, 78, 86, 102). Most datable coins found were minted in the fourteenth century (24% of the total finds) or in the seventeenth century (21% of the total), while the numbers from the thirteenth, fifteenth, sixteenth and eighteenth centuries are lower and remain on a fairly stable level (c. 7–11% of the total). The series ends with five coins from the nineteenth century, the last one dated 1844. The finds also include several coins that could not be dated, but in all cases except one, they could at least be attributed to either the Medieval or post-Reformation period, according to their material qualities as indicated above. Seventy medieval coins (56% of the total coin finds) were found in Gränna church, of which the major part (≥61% of total medieval coins) were of Swedish origin. The second largest group was from Germany (≥10% of the medieval coins), supplemented by a few Danish, Gotlandic and Norwegian coins and the one from France (Table 11.2). All coins dating to the period after c. 1450 were Swedish.

Building for glory – Gränna and Arby  305 Gränna church, coins per century (∑=126) 35 30

30

25

26

20 15

14

10 5 0

1

1

10

14 9

13 5

2

1

Figure 11.11 Number of coins per century, Gränna church.

Figure 11.12 A selection of coins found in Gränna church. From top left to bottom right: 1: Sweden, Avesta, Karl XI, 1/6 öre 1666 (Cu, 6.37 g., no. 55); 2: France, Lyon, denier after 1200 (Ag, 1.30 g., no. 21); 3: Germany, jeton, eighteenth century (?) (brass, 0.89 g., perforated, no. 51); 4: Gotland, Visby, gote 1420–50 (Ag, 0.78 g., no. 108); 5: Norway, Olaf Kyrre (?), penning c. 1065–80 (Ag, 0.45 g., no. 1); 6: Sweden, Valdemar, halvpenning 1250–75 (Ag, 0.17 g., no. 12); and 7: Sweden, Uppsala?, penning 1260–75? (Ag, 0.23 g., no. 35). Nos. 12 and 35 are bracteates. Photo: Ola Myrin, KMK

Of particular interest is a thirteenth-century coin (no. 35) of uncertain date and origin: a bracteate with a cross-and-banner over a cross-hatched shield (Figure 11.12:7). The inscription reads ARVS (= Aros, the name meaning “river estuary” and used for several Scandinavian towns thus situated), and it was probably minted according to the Svealand 21 regional standard, hinting at an origin in either medieval Uppsala (“Eastern Aros”) or Västerås (“Western Aros”). The coin was unique when it was found during the excavations in 1987, but one die-identical specimen has since been

306  Nanouschka M. Burström Table 11.2 Coin finds from Gränna church Fno

KMK no.

Country

1 2 3 4

102 100 9 7

Norway Denmark Sweden Sweden

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

112 18 21 121 118 93 111 3 4 103 33 28 23 37 104 29 114 122 26 119 97

Germany Sweden Sweden

26 27 28 29 30 31 32 33 34 35

14 58 8 123 117 115 12 31 13 1

Sweden Sweden Sweden Sweden

36

86

Sweden

37 38 39

79 69 92

Sweden Sweden Sweden

Stockholm Avesta Avesta

Gotland Germany Sweden Sweden Norway Sweden Sweden Sweden Sweden Norway Sweden France

Mint/area

Lund/Naestved Erik of Pomerania Götaland Magnus Birgersson Kalmar Valdemar/Erik Birgersson Magnus Eriksson Magnus Eriksson

Visby? Götaland Götaland Oslo

Scania, Lund

Sweden Sweden Sweden

Avesta Götaland

89

Sweden

19

Sweden

Period/year

Penning Penning Half penning Half penning

1065–1080 1400–1420 1275–1290 1275

Penning Penning Penning

1319–1363 1340–1354

Penning Penning Valdemar Half penning Valdemar Half penning Håkon V Magnusson Half penning Unknown Magnus Eriksson Half penning Magnus Eriksson Penning Penning Håkon V Magnusson Half penning Magnus Eriksson Half penning

Magnus Eriksson

Sweden

40

Denomination

Lyon

Sweden

41

Issuer

Svealand, Uppsala?

Magnus Eriksson Smek Birger Magnusson Karl XI Magnus Birgersson

Penning

1220–1290 1250–1275 1250–1275 1299–1319 1300–1325 1350 1340–1354 1370–1380 1299–1319 1350 1157–1200 1354–1363 1332–1360

Obol 1/6 öre s.m. Half penning

1290–1318 1666 1275–1290

Penning Half penning Obol Penning

1290–1318 1300–1360 1290–1318 1260–1275

1 öre k.m.

1719–1750 1718 1677 1844

Gustav IV Adolf

1 daler s.m. 1/6 öre 1/6 skilling banco ½ skilling

1800

Magnus Eriksson

Penning

1340–1354

Birger Magnusson Magnus Eriksson Birger Magnusson Valdemar Birgersson/ Archbishop? Ulrika Eleonora/ Fredrik I Karl XII Karl XI Oskar I

Building for glory – Gränna and Arby  307 42 43 44 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63

15 17 27 27 65 59 54 67 38 107 126 73 72 64 60 62 74 63 61 68 75 77 87

Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Germany Germany Sverige Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden

64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92

76 46 80 24 10 99 11 20 109 105 110 22 101 36 6 56 82 83 85 78 94 48 2 40 98 88 84 81 120

Sweden Sweden Sweden Sweden Sweden Denmark Sweden Sweden Germany Germany Sweden Denmark Sweden Sweden Sweden Sweden Sweden Sweden Sweden Gotland Sweden Sweden Sweden Denmark Sweden Sweden Sweden

Avesta Avesta Avesta

Birger Magnusson Magnus Eriksson Magnus Eriksson Magnus Eriksson Karl XI Karl XI Kristina Karl XI Unknown

Lübeck Avesta Avesta Avesta Avesta Avesta Avesta Avesta Avesta Avesta Avesta Avesta

Karl XI Karl XI Karl XI Karl XI Karl XI Karl XI Karl XI Karl XI Karl XI Karl XII Karl XII Ulrika Eleonora/ Fredrik I Avesta Karl XII Stockholm Johan III Stockholm Ulrika Eleonora Magnus Eriksson Götaland Magnus Birgersson Lund/Naestved Erik of Pomerania Götaland Magnus Birgersson Magnus Eriksson Mecklenburg Mecklenburg Malmö Götaland Stockholm Stockholm Stockholm Stockholm Avesta Visby Stockholm Götaland Lund Avesta Stockholm Stockholm

Magnus Eriksson Hans Unknown Valdemar Karl XI Ulrika Eleonora Ulrika Eleonora Fredrik I Karl XII Johan III Valdemar Unknown Erik of Pomerania Gustav IV Adolf Ulrika Eleonora Ulrika Eleonora

Half penning Penning Penning Penning 1/6 öre s.m. 1/6 öre s.m. 1/4 öre 1/6 öre s.m. Penning Penning

1290–1318 1319–1340 1350 1350 1673 1666 1633–1644 1676 1470–1500 1300–1500

1/6 öre s.m. 1/6 öre s.m. 1/6 öre s.m. 1/6 öre s.m. 1/6 öre s.m. 1/6 öre s.m. 1/6 öre s.m. 1/6 öre s.m. 1/6 öre s.m. 1/6 öre s.m. 1/6 öre s.m. 1 öre k.m.

1681 1671671 1666 1667 1686 1668 1666 1676 1707 1715 1719–1750

1/6 öre s.m. 2 öre 1 öre k.m. Penning Half penning Penning Half penning Penning Penning

1708 1573 1719 1340–1363 1275–1290 1400–1420 1275–1290 1340–1354 1300–1500 1200–1300 1400–1500 1340–1353 1481–1513 1370–1380 1250–1275 1688 1719 1719 1720 1718 1420–1450 1575–1592 1250–1275 1500–1520 1405–1420 1805 1720 1719

Penning Penning Hvid Penning Penning 1 öre 1 öre k.m. 1 öre k.m. 1 öre k.m. 1/6 öre s.m. Gote Fyrk Penning Penning Sterling 1/12 skilling 1 öre k.m. 1 öre k.m.

(Continued)

308  Nanouschka M. Burström Fno

KMK no.

Country

Mint/area

Issuer

Denomination

Period/year

93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 168 419a 419b 422

90 55 71 47 51 49 52 50 45 5 116 108 34 39 16 96 44 106 32 41 43 57 113 30 42 66 25 91 95 35 53

Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden

Avesta Reval Avesta Stockholm Stockholm Stockholm Göteborg Stockholm Västerås Götaland

Karl XIV Johan Kristina Karl XI Johan III Karl IX Karl IX Karl IX Karl IX Gustav I Valdemar

1/4 skilling 1 öre 1/6 öre s.m. ½ öre 1 öre 1 öre 1 öre 1 öre Fyrk Penning

1821 1649 1671575 1611 1610 1610 1611 1523–1531 1250–1275

Germany Sweden Sweden Sweden Gotland Sweden

Lübeck

Penning Penning Penning Penning Gote 4 penningar Half penning Half penning ½ örtug Fyrk ½ öre k.m.

1300–1500 1363–1370 1470–1500 1319–1340 1420–1450 1523–1531 1275–1325 1300–1360 1512–1520 1523–1534 166-

Kristina

Half penning Örtug 1/6 öre Penning 1/4 skilling Gote Penning 1/4 öre

1350 1524–1527 1673 1354–1363 1802–1830 1420–1450 1363–1370 1633–1644

Karl XI

1/6 öre s.m.

1677

Sweden Sweden Sweden Sweden Germany Sweden Sweden Sweden Sweden Sweden Gotland Sweden Sweden

Unknown Unknown Magnus Eriksson Visby Stockholm

Gustav I

Stockholm Stockholm Avesta

Magnus Eriksson Sten Sture d.y. Gustav I Karl XI

Stockholm Avesta Avesta Visby

Magnus Eriksson Gustav I Karl IX Magnus Eriksson Karl XIV Johan

124 125 Sweden

Avesta

found in St Olof church in the medieval town of Sigtuna, sustaining the coin’s Svealand provenance. 22 Based on iconography, inscription, weight and intrinsic value, it has been suggested to be an archiepiscopal coin from the end of the thirteenth century. 23 In Gränna church, there were also several coins from the first part of the eighteenth century, some struck on coins from the preceding ruler (nos. 80, 81, 82, 90), and a perforated brass disc (German Rechenpfennig), perhaps used as a spangle (no. 51).

Some conclusions to draw from the coin finds The coin finds from the two investigated churches visibly increase in number from c. 1250. Both churches have many finds from the fourteenth and seventeenth centuries,

Building for glory – Gränna and Arby  309 but while these numbers are very high compared to other centuries in Gränna, this is not really the case in Arby. In Arby, there is instead a “dip” in the sixteenth century, which is visible but not as pronounced in Gränna (see Figures 11.9 and 11.11). In Arby church, there were marginally fewer medieval coins than the post-Reformation ones, while in Gränna, they were slightly more numerous. In both churches, the Swedish coins dominate the medieval finds, but more obviously so in Gränna, while in Arby, they were about half the group. The origin of non-Swedish mainland coins is comparable: the largest group consists of German coins, while Danish and Gotlandic coins are next. The number of non-Swedish coins is a little higher in Arby, possibly because of it being further from the central Swedish monetary area and closer to the Baltic Sea commerce than was Gränna.24 The finds from Gränna, however, include two early and “exotic” coins such as the Viking Age coin from Norway and the French twelfth-century coin, the Norwegian coin predating the construction of the church (see further below). All in all, one may thus point to some general trends emerging from the two sets of finds, albeit cautiously (remembering that only one of the churches was excavated in extenso): a few early coins, not to be considered really part of general monetary use; a rapid increase in coin use from about 1250; a high number of coins deposited in the fourteenth and seventeenth centuries; a “dip” in the sixteenth century; and the use of coins or token coins for spangles. It is of general interest to compare the coin finds from these two churches and to consider a common explanatory model within coin studies, which points to the small size of the medieval coins as a possible reason for the coins being lost and left in the churches. This hypothesis does not hold up well, however, once the fact is taken into account that there are also as many coins from the post-Reformation period (in these two sets of finds, the two periods also comprise just about the same length of time, c. 300–350 years), when coins were generally much larger than in previous periods, and thus should be easier to hold on to and to retrieve if lost.

The building history In order to use the coin finds to say something about the church and medieval religious practices, and to use those to understand the evidence of coins found there, an important prerequisite for this study was to understand the detailed context of specific coin finds, the building history of the church and how specific features of the site were (or may be) interpreted. Some features mentioned are included to help construct an understanding of the patterns of movement and the light conditions inside the church. The sources used are the excavation reports and extensive archival material as indicated above; here, however, it will only be possible to present a summary of results relevant for the argument. A sketch of the building history will be followed by an analysis of the finds. Arby church – from wood to stone When Arby church was excavated in 1971, it became apparent that it was one of only a few cases in Sweden where the earliest history of church building could be archaeologically demonstrated. The traces of a small (c. 4 × 8–10? m) wooden church were discovered beneath the floor of the (still standing) thirteenth-century stone church.

310  Nanouschka M. Burström This was indicated in the bottom mortar and limestone floor layers by postholes and a furrow left by a wooden sill for the wall (see Figure 11.5). The construction was made with upright posts secured in a stone-packed posthole, but it was not possible to ascertain the wall type25 or the precise position of the gables. This wooden church was built at the beginning of the twelfth century. At the turn of the century (AD c. 1200), a stone chancel and an apsis (still standing) were connected to the east end of the wooden church, possibly replacing a small wooden chancel. They are unusual in that the apsis is internal; i.e., the east gable wall is flat when viewed from the outside, apart from the usual small opening for the east window, and the chancel was provided with an unusual solution of stairs to directly access the upper floor or attic. The chancel and apsis marked the first steps in transforming the building into one made entirely of stone. Scaffolds (evidenced by numerous postholes) were at the same time erected for the new building project around the wooden church, in what was to become the new nave. The inner wooden building probably continued to work until the stone church was completed around it. This happened not later than 1250, at which point the wooden church was dismantled and the posts were extracted from the ground. 26 In this process, several small finds – like coins – from the floor fell into the resulting holes and gaps, an important factor to keep in mind whenever using coins to date structures. The new church (Figure 11.13) was provided with a broad and imposing doublehooded west bell tower, built with the nave but completed slightly later. There were

Figure 11.13 A rby church in 1747 exhibited still with medieval windows, doors and towers. Details are not to scale, but measures are indicated in writing on the drawing, showing that the towers are in reality twice as broad in the E–W direction. Drawing by Petrus Frigelius. Photograph: RAÄ/KMB. Public domain (cf. layout at fig. 11.3 and 11.16).

Building for glory – Gränna and Arby  311 three entrances on the south side of the church: one to the chancel, one to the nave and one to the ground floor chamber of the tower, but none on the north side. There was a doorway between the tower and the nave. In the chancel and in the bell tower, there were stairs inside the walls, leading up to the attic or a tower above the chancel, and to the chambers and garrets of the west tower.27 A vestry was built adjacent to the chancel in the fourteenth century. The tower was taken down in the 1770s, after a period of repeated attempts of restoration, and its lower part was integrated with the nave. A new, western, entrance door was opened in the former tower in the 1820s. The small medieval windows were then also replaced with larger and lower ones, and the southern entrances to chancel and tower were rebuilt into windows. The church was pewed no later than 1600, 28 and a balcony for women was built in 1736. The chancel screen was permanently removed in 1770, when the chancel wall opening was broadened.29 The main, Romanesque, block altar in the apsis was replaced in the eighteenth century30 and a rood altar in front of the chancel arch was removed in the seventeenth century to make space for a large grave chamber. There were at least two side altars, one on each side of the chancel arch, which were both taken down in 1609. A few medieval movables are still left in the church, notably a baptismal font from the latter part of the thirteenth century31 and a rood from about the same time. 32A suffering Christ from the fifteenth century may have been previously placed on one of the side altars, 33 and the existence in the church of a mounted St George is mentioned in 1667.34 The church bells, which have occupied a detached bell tower since the seventeenth century, 35 may be dated 1403, and display an icon of St Mary, St Olof or St Sigfrid, as well as an inscription that may mention St Olof, St Mary and Christ (in that order). This could point to St Olof as the patron saint of the church.36 Gränna church – from royal estate to market town congregation The Romanesque church in Gränna was probably built around AD 1150 or slightly after that. A few burials precede the church on the site: one adult female was found under the foundation walls37 during the excavations in 1986, and one individual (only a badly preserved skull remained) was found in the construction layer. 38 However, no traces of a wooden church have been found, and the chronologically subsequent burials are probably contemporary with the twelfth-century building. 39 One conspicuous indication of early Christian burial grounds in these provinces is ornamented stone cists, of which fragmented examples were found in nearby Kumlaby and possibly also Ströja churches, but none in Gränna. Two rune-stones may have been erected on the site,40 but both are now lost. In any case, rune-stones are not indicative of churchyards, but are instead likely to have been associated with the old roads, such as the one passing through the area, and Viking burial grounds, of which there are a few around Gränna. This pattern fits well with what is observed at Husaby, where the church was built. While, based on excavations and documentation works in 1986, the outline of the Romanesque church is largely established, knowledge of its interior is less complete. The excavations mainly comprised the outline of the walls and six trenches covering a total area of c. 60 m 2 (see Figures 11.8 and 11.14).41 The excavation results and some early photos and drawings also tell us quite a lot about the interior and major fittings all through the period, and the main events will be traced below. A major remodelling of the church in the seventeenth century, when walls were demolished to enlarge the

312  Nanouschka M. Burström church, created a very different space in the nave. Also, the church burnt down almost entirely in 1889 (Figure 11.15) and was rebuilt in 1890 with a new chancel. Since these phases are rather distinct, and the last one did not yield any coin finds at all, the focus here is on the Medieval period up to the time of the seventeenth-century church. The Romanesque Gränna church was unusually large for its time, about 28 m long, even without a tower (e.g. it was about three times the size of the otherwise comparable Ströja church,42 or the estimated length of the wooden twelfth-century church in Arby). This is only one of the indications that Gränna church was of particular importance to its neighbourhood or to the functions of the royal estate. Certain exterior decorative details on the chancel and apsis, a (possible) stone canopy over the

Figure 11.14 Gränna church, plan from report with added structures, burials and coin finds. Map: Steinar Kristensen, UiO: Museum of Cultural History after Varenius (1992).

Building for glory – Gränna and Arby  313

Figure 11.15 Gränna church, interior of the chancel and apsis after the fire in 1889. Photo: Carl Curman, RAÄ/KMB. Public domain

south-west entrance and a (possible) private stand or gallery inside the building add to this picture. Despite these prestigious features, it appears that the building was left without a tower until the Late Middle Ages or even the post-medieval period, which seems odd alongside the multitude of variously embellished towers that are otherwise found in the province, and since these towers might be regarded as signifying elite donors and worldly power. General access to the medieval church was through a portal in the south-western part of the nave, which was probably decorated with a canopy on the outside.43 The chancel could be accessed directly through a portal on the south side. There were originally no entrances from the north. The floor was likely of earth, mixed with sand and gravel.44 At the west end, two remnants of posts testify to an elevated wooden structure, possibly a stand of wood erected for a patron.45 As the original walls were later torn down, nothing can be said about wall benches or windows in the nave, but windows were likely small and high up, surely present on the south side and possibly (but generally less common) on the north side. The chancel seems to have had a window only on its south side.46 The apsis had three small windows; the frame of one window remains, and the recess is now used to house a votive sculpture.47 The present high altar may be the original one.48 It is likely there were at least two altars in the nave, one on each side of the triumphal arch.49 After about 200 years, several changes occurred. The wooden structure in the west end was taken down, a north portal to the nave was opened, and new windows were probably opened in the north wall at the same time.50 A larger porch replaced the canopy of the south entrance, possibly around AD 1400, used not only to enter the church but also for burials in the course of several hundred years. The ceiling was

314  Nanouschka M. Burström vaulted (in the 1420s?) and beautifully painted (towards the end of the century) by Amund, one of the great masters of the time.51 We know only the paintings in the chancel and apsis, a Credo motif (Figure 11.15, 11.16).52 In most of the church, a floor of slate slabs was laid (before 1450), but a wooden floor was preferred for the west end of the building. A wooden bell tower was probably added in the west during the fifteenth century, as was a roof turret over the eastern part of the nave. These changes point to a modernised church space, where the early medieval expressions of power and patrons (like the canopy and the private stand) were downplayed and replaced with embellishments and comfort for the congregation. Various changes that required significant cost and effort, and greatly improved conditions for those attending church, point to the expanding role of the congregation in managing the church: new access to the nave, a better floor and possibly new window openings that provided more light; at the same time, the addition of colourful paintings with the Credo motif underlines the importance of the congregation’s faith. The tower and turret (if they were indeed built in this period) may also have been added as part of this communal effort. The medieval church was thoroughly altered at the end of the seventeenth century, when the south and north long walls of the nave were torn down, the vaults demolished and the church widened, and the church was thus transformed into a spacious, cruciform Baroque building. Chancel and apsis were provided with upper storeys. Larger and lower windows in the new walls brought more daylight to the interior. These construction works followed changes in the status of Gränna, which

Figure 11.16 Gränna church, Credo paintings in the apsis of the Romanesque church. Drawing by Nils Månsson Mandelgren (published 1862, Pl. XIV Figure 4). Photo: RAÄ/KMB. Public domain

Building for glory – Gränna and Arby  315

Figure 11.17 Seal matrices from Gränna church: one from the sixteenth century (cross keys, right), the other from the seventeenth century (church, left). Photo: Torgny Lundqvist. Used with permission

transformed from village and country parish to a privileged merchant town. This was largely based on the initiative of the Brahe family, who had been lords of Visingsö since the mid-sixteenth century, and from the seventeenth century including Gränna and other areas in the district within their realm of power. Per Brahe (the younger, 1602–1680) “founded” Gränna town in 1651, and although he rebuilt another church (Ströja) for himself and his family on Visingsö, he also took an interest in the church in Gränna, for example by allegedly sponsoring the construction of its stone tower. The medieval church, once large for its time, was no longer considered sufficient when the market town grew. The present church, rebuilt after the fire in 1889, is mainly rebuilt in the image of its predecessor from 1690. A few movables remain from the medieval and Baroque churches, e.g. a sandstone baptismal font from the thirteenth century, a fifteenthcentury tabernacle, fragments of late medieval sculpture (a Rood and a Mary), two seventeenth-century seals53 (Figure 11.17) and parts of an altar screen from 1722. An inventory list from 1828, before the fire, also registers many other types of items such as books, an organ, paintings, chests and textiles, providing an image of what was at the time kept in the church and considered to be the church treasures.54 Some conclusions based on the building history The changes made to the churches and sketched out above are only a few of the many makeovers this kind of ancient “body” normally experiences over the centuries. The picture is in part very similar to that of most medieval Swedish churches, following the wider changes in population, liturgy and fashion, while some events are due to specific conditions and shaped according to individual tastes. Many of the medieval additions of vestry, porch, vaults and frescoes follow the fashions of their time and are not unique in kind, although they may be so in terms of detail. The movable furnishings and liturgical equipment provide clues as to aspects of the interior and the use of the church, such as celebrations of saints’ days or why graves were placed in certain positions.

316  Nanouschka M. Burström Arby and Gränna churches thus had rather different life histories but also share some traits, such as the indications of a private patron in the early structures of the Romanesque churches, although those are expressed in different ways. For example, at Arby the honorary entrance led to the tower, where the first floor formed a private balcony. In Gränna, it was a communal entrance, but it was an enhanced and impressive portal leading to the nave, and possibly also to a private stand or platform in the west end of the nave. The differences between the two churches are more accentuated in the post-Reformation period, when Gränna was reshaped to fit a new and modern context as it became centre for a town and its congregation rather than a village community.

The detailed contexts of the coin finds, and some other archaeological features With only scant historical evidence to draw upon, the use and function of coins in churches must be investigated based on material traces. Accordingly, what follows is an analysis of the coin finds in relation to the archaeological features where they were discovered, such as individual layers and graves. In addition, some selected features and contexts of relevance to the discussion will be highlighted, such as altars and postholes, even if they did not yield coin finds. Arby church: coins and archaeological features Arby church was excavated in its entirety. The upper layers of the floor, in particular in the chancel, were considerably disturbed by burials and floor reconstructions, by burial chambers built in in the seventeenth and eighteenth centuries and by installation of heating in the 1940s. Also, the floors had already been removed down to medieval layers in the 1770s. Thus, the main focus for the excavations was the medieval mortar layers – the bedding for a partly preserved twelfth-century limestone floor and for a floor from the thirteenth century – through which other structures such as graves and posts had been dug (30 postholes from the wooden church and from scaffolding were registered) (see Figure 11.5).55 Some of the postholes contained coins, revetment or remains of wood. This mortar layer could be followed through the church. A groove in the mortar indicated the south wall of the wooden church, and a deep posthole adjacent to the wall was interpreted as the remains for a hoisting device from the building of the stone church. One large posthole in the central stone apsis was one metre deep and contained wooden residues which could be dated to AD c. 1100. It was tentatively interpreted as a support for the building of the apsis. In the apsis, a “negative trace” of the medieval high altar (c. 1 × 1.2 m) in the mortar layer was also found, which included at least 20 coins (find nos. 119–138) mainly from the fourteenth and fifteenth centuries, the oldest one being a coin from the early thirteenth century (Sweden/Gotland, no. 131) and the latest one from 1667 (Sweden, no. 132), which may have been deposited on or around the altar. The medieval altar layer was covered, possibly in the 1770s, by a metre-thick layer of stones. In the southern part of the apsis, one fifteenthcentury coin (Germany, no. 139) was found in a stone-packed posthole which may indicate the place for the sanctus bell. The stone chancel (constructed with the apsis) was dendrochronologically dated to the first decades of the thirteenth century based

Building for glory – Gränna and Arby  317 on two features: a post from after 1171, of which parts were left in the chancel attic, and an upper wall plate from after 1176. The earliest window opening was stylistically dated to the first part of the thirteenth century. The early thirteenth-century coin from Gotland (no. 131), found under the later altar fundament, fits in well with this evidence, while the later German coin (no. 139) may have entered its posthole only when the post was taken away, and thus dates the end of use for the post rather than the start. When the wooden church was dismantled inside the stone nave, the posts supporting its roof were pulled out; thus, no wood was left to date. 56 But, some coins from c. 1250 to 1290 (nos. 87, 89) may have ended up in two of the postholes when it was done, again dating the dismantling of the church rather than its time of construction. Another coin (no. 140) from the same time was found by the triumphal arch (north side) on the mortar floor of the stone church. Also, the earliest datable grave (grave VIII, a female) contained in its filling coins from about the same time (nos. 90, 91, 92, c. 1220(?)–1275), in this case indicating the terminus post quem of the burial. The grave was situated in the north-east corner of the stone-church nave, just in front of the side altar (see Figure 11.5). Out of 12 excavated medieval interments, this grave was not only the earliest, 57 but also the richest, containing items such as an elaborate and gilded silver brooch from the beginning of the fourteenth century. 58 It may thus be suggested that the coins from 1220(?) to 1290 found in the north-east corner were deposited on the side altar, possibly before 1300, while the burial took place after 1300, including some of the fallen-down altar coins in its filling. Possibly, the coin from Erik Eriksson (no. 91; see Figure 11.10:5) belonged to the dress of grave VIII since it apparently was used as a spangle, and with its crown imprint would have been a small but pretty high-status symbol. It was, however, c. 100 years old at the time of the burial, so an alternative suggestion could be that it was originally attached to the dress of an image of St Mary on the altar itself. Three graves (I, II and III, not containing coins), more or less contemporary with grave VIII, 59 were situated in a corresponding position on the southern side of the nave, i.e. in front of the south side altar. Just south of grave II, adjacent to the wall and foundation, some large stones were found, and on the wall above it, some fragments of a fresco canopy were found, together suggesting an additional altar or some kind of grave marker. An additional medieval grave contained coins in its filling: grave IX, close to the south portal, which included a Swedish coin from c. 1300 to 1325 (no. 102) and two German coins from post-1400 (nos. 100 and 101). In the same area were found a few stray coins from the fourteenth and fifteenth centuries (nos. 81, 103, 114, 143), but also two modern coins (nos. 142 and 144). Also, post-medieval graves contained medieval coins in the filling, e.g. one grave in the west end of the nave60 (nos. 117, 118, 141), just a few metres from grave IX, which supports the supposition that coin finds in churches may end up in secondary positions such as grave fillings, but still reflect the approximate position where they were first deposited. The grave chambers dating to the post-Reformation period in the east end of the nave and in the chancel contained mainly coins from the seventeenth and eighteenth centuries,61 into the first years of the nineteenth century (in total 42 coins).62 As the chambers are closed contexts for bodies and coffins, and not filled up with earth, it is more difficult to say how the coins ended up there. One interpretation is that coins fell from the floor into the chambers as those were repeatedly opened for new burials; another is that they were deposited with the bodies as “Charon’s obols” (small coins intended to pay

318  Nanouschka M. Burström the ferryman of Greek mythology), a revived tradition in the post-medieval period. Without detailed grave descriptions, these remain suggestions. Quite a few coins, from c. 1250 until the seventeenth century, were also found in the western part of the nave, on the floor and in the filling between stones in the stone foundations. An area of concentrated finds in the north-west corner was suggested by the excavator to reflect the position of the baptismal font (see Figure 11.5). It is, however, uncertain how (and, indeed, if at all) coins were used in connection with baptismal fonts in Scandinavia,63 and based on analogy with the eastern end of the church, an altar could be a better explanation. Also, there were medieval graves in this corner, and the spot (with no north portal) would have been a calm corner of the church, yet immediately visible when entering the nave from the south portal. Finally, three coins from 1572 to 1635 were found in the filling of a post-medieval burial at the bottom floor of the former tower (nos. 96–98). Several coins from Arby church could thus be associated with archaeological features, in most cases supporting or suggesting dates for the building history of the church, of the end of use for certain features (like posts) or of the time after which they must have been dug or built (like the burials). The coins also suggest the existence of an altar in the north-west end of the nave from c. 1250, although this was not otherwise documented. The similar positions of graves on the northern and southern sides of the triumphal arch, the finds in the north grave (VIII) and the coins included in its filling indicate that the altars were localising factors for burials from c. 1300. The differences in time between coin and burials further suggest that the coins were not deliberately deposited with the bodies during the Scandinavian Middle Ages, but were associated with the altars. Gränna church: coins and archaeological features Gränna church was excavated less extensively than Arby, but several trenches intended to provide a broad picture of the building history were opened (see Figure 11.8). Forty burials were documented, 33 of which were found in the south-west corner (the medieval porch). Of particular interest is grave no. 8 (trench 4), underlying the foundation for the triumphal arch and thus predating the church; it may therefore indicate an earlier burial ground or a preceding church, which could not be otherwise established. The grave contained no coins. Another burial (no number, trench 5) was cut just after the foundation walls but yet under a mortar layer from the building phase. The grave continued under a fifteenth-century pilaster and was too fragile to be further examined. Other early burials were graves 31, 32 and 34, deposited in and adjacent to the foundations for the canopy of the south-west portal. One of these graves had a worked sandstone slab by the head, but none contained coins. The first graves containing coins were dug in the fourteenth century. Grave 2 (in trench 1), a coffin grave in the north-east corner of the medieval nave, had a small sandstone slab by its head and contained in its filling the oldest coin found in the church (no. 1, Norway, AD c. 1065–1080) as well as two from the fourteenth century (nos. 33 and 34). Above this grave was another (grave 1, trench 1), from the fifteenth century, which included 18 coins in the filling, ranging in date from the twelfth century (no. 21) to the 1360s (no. 23). The burials could be broadly dated from the stratigraphy and arm positions. The excavator argues that the fillings reflect

Building for glory – Gränna and Arby  319 the chronologically mixed composition of finds from the spot where the graves were dug.64 The coins here therefore seem to support the presence of a side altar in the north-east corner of the nave. The Viking Age coin could be interpreted as an exotic souvenir, or perhaps as deriving from a burial predating the Romanesque church, while the French twelfth-century coin fits in well with the presumed date for the foundation of the church. However, apart from these two foreign coins and one from Gotland (no. 10, also found in trench 1), possibly minted as early as c. 1220/1225, the coin series starts only c. 1250 with the coins of (Swedish) King Valdemar (1250– 1275), when the church had been in use for about a hundred years. The north-west corner of the nave is of interest for the handling of coins in the last period of the Romanesque church. The numbers of coins increase in general in the church in the seventeenth century, and many of those are found in trench 5, inside the north portal. Here, remnants of a painted brick pilaster, likely part of a fifteenthcentury inner vaulted ceiling construction, were also found. There is no particular reason why a vault pilaster should attract numerous coins, unless it was adjacent to an altar or possibly an offering box. According to Karlsson,65 original side altars were often taken down when vaults were built, and were sometimes used as foundations for the pilasters, in which case they might be traced archaeologically, but sometimes completely removed. New altars were also often built, attached to the pilasters above the floor level. Remnants of such altars by pilasters were often mentioned in antiquarian descriptions of churches, while they may be difficult to trace today if not apparent in the surface of the pilaster. As coins were continuously deposited in this area both before and after the building of the north portal and the vaults,66 an altar (or altars) and offering box remain the most likely explanation. The porch was a new and important feature of the late medieval church. Most of the coin finds come from this area (starting with two specimens from the thirteenth century), which was continuously used for burials until it ended up under the floor of the newly built south aisle of the Baroque church. Thirty-four adults and an unusually high number of children and infants (21 individuals) were buried in the porch area. Four burials were probably older than the porch itself. The majority of adults were buried in the porch during the fifteenth century, while the upper layer of graves, containing almost exclusively children, was added in the post-Reformation period.67 Coins were not obviously used in burials in the porch as they were found solely in the fillings, but the continuous activity in the restricted space created disturbances that make it difficult to ascertain what the coins actually were used for. One possible reason for the extensive use of coins in the area could be the presence of an altar or side chapel,68 although it cannot at present be confirmed. Another explanation could be an offering box at the entrance. From the fourteenth century, coins seem to play a bigger role than before in the church (see Figure 11.11), but in the fifteenth century, the numbers decline. This could be related to a new stone slab floor laid in large parts of the nave, in particular as most of the late fifteenth-century coins were found in the porch – where such a floor was not laid – and in trench 5 where the stone floor met with a contemporary wooden floor. This seems to sustain the often-voiced opinion that coins will be fewer where there is a stone floor. The matter is not settled here, however, since three coins are recorded as deriving from the sand layer underneath the slabs, two of which (nos. 76 and 85) are presumably more recent than the slab floor. Apparently, coins could

320  Nanouschka M. Burström trickle down between slabs like between floor boards.69 It would be tempting to suggest that the numbers of coins from all the church would be comparable to those from the porch, had the entire church been provided a wooden floor. Still, this would be jumping to conclusions, judging from trench 5, where wood floor and stone floor met. This spot yielded almost no coins from the fifteenth and sixteenth centuries at all, despite its wooden floorboards, yet there were again numerous coins deposited there in the seventeenth century. Compared with the profile of coin finds from Arby, it also demonstrates that the declining numbers in the sixteenth century may belong to a more general pattern, such as a general decline in the availability of coins. Accordingly, floor covering is only one of the factors affecting the numbers of coins lost and found. Coins and archaeological features: some conclusions drawn from Arby and Gränna churches From the above, it seems that coins were rarely, if ever, deliberately deposited with buried bodies, but appear in the filling, and they thus primarily reflect what was on the floor and in the layers around the spot where the body was interred. Coins could also end up as filling in postholes when posts were removed, which means that they, in those cases, can be used mainly to date the removal of posts, not as an indicator of when they were erected.70 It has also been shown that flooring material (wood or stone) only partially explains the presence or absence of coin finds or differences in coin deposition rate over time. As indicated by the many coins found in the chancel and apsis in Arby, the medieval high altar could probably be accessed by the congregation in order to place coin offerings on it,71 or some other ritual, such as the priest blessing the offerings, may have taken place there. The majority of the coins from the location were struck in the fourteenth and fifteenth centuries, thus giving a broadly late medieval dating of the practice.72 It may be suggested that side altars in the north-west corner of the nave were typical in the churches, at least in those of this diocese, judging from the concentrations of coins in the north-west corners of both churches.73 Early graves were found in front of the east side altars in both churches: in Arby, to both the north and south, and in Gränna, confirmed in the north (but the relevant area in the south remains unexcavated). The position in front of those side altars may be considered exclusive or honorary, an observation based not least on the fashionable grave goods belonging to the woman in grave VIII at Arby. Still, coins do not seem to have been part of the grave goods. The evidence from the two churches thus supports the hypothesis that in Sweden, finds of medieval coins in churches mainly result from their use on, or adjacent to, altars.74 In the post-Reformation period, side altars supposedly became fewer and offering boxes may have replaced some of them, in line with the Lutheran focus on doing good deeds for your fellow rather than offering to God through saints on altars. Coins were then also collected directly in the pews using collection bags, a practice still in use, and coins may have dropped on the floor in this process as well. When coins are found in postholes or in graves, it is as secondary filling, not as intentional depositions, and this is reflected in subsequent building and grave-digging activities.75 The coin evidence here also suggests that burial inside the church was not a regular practice in the earliest phase, but started in the early fourteenth century.76

Building for glory – Gränna and Arby  321

Building for glory The church, the parish, the altars and indulgence In Scandinavia, early churches were built from the tenth century AD, as far as can be gleaned from the existence of wooden churches, which preceded stone buildings built from the eleventh century onwards. The phenomenon largely coincides in time with other noteworthy developments, such as the most intensive phase of rune-stone erection (most of which demonstrate their patron’s Christian faith) and the start of domestic coinage. Presumably, the aim of producing domestic coinage was, at least in part, to facilitate a more regular, everyday and monetary use than was the case for earlier coinage, where foreign coins were activated mainly as bullion and for ritual and decorative purposes. There are only a few contemporary coins in the earliest churches, but they do appear occasionally from about 1050.77 In some cases, they may derive from older graves on the site,78 but some may be older coins that were still in circulation, and which were deposited in the church during the twelfth century. Coins seem to be present in medieval Swedish and Finnish churches in significant numbers79 from about AD 1200, with a notable increase around 1250–1300.80 This is reflected in the two churches of this case study, which in this respect fall into the general pattern. This pattern is likely the result of the formation of parishes, i.e. the formalisation of the physical extent of a congregation, which in the thirteenth century succeeded the twelfth-century introduction of tithes. The formation of parishes in Scandinavia took place at slightly different times and rates in different areas, but was probably completed in Denmark and some central areas of Scandinavia by AD 1200, while in other areas it took another 50 years or so; in all, it was probably a process that unfolded over two centuries.81 A general development may be outlined where aspiring royal powers and the Church during the eleventh and twelfth centuries cooperated in building a first few stone churches in important places and centres, parallel with, or preceded by, private initiatives in smaller villages or elite farmsteads building preferably in wood. Such small private churches on elite estates already existed on the continent and in England from AD c. 500. When the Scandinavian parishes formed in the twelfth and thirteenth centuries, possibly after English model, either they were connected to an existing church (a previous private or royal church) or a new one was built on common or granted lands.82 Private churches that were not integrated with the parish system eventually fell out of use. This general outline would obviously have to be assessed critically for specific cases, but seems to fit well with the churches focused upon here: Arby, probably a private church built in wood c. 1100, and after a century rebuilt in stone for a larger congregation; and Gränna, probably built c. 1150 to serve the royal estate, but turned into a general parish church about 1250 when the Husaby system went out of use and the estate owner Queen Katarina left the area.83 The change would have entailed an increased number of participants in the services and affected how services were carried out. Olav Tveito proposes that in Scandinavia, remunerations for services such as baptism, weddings and requiems were, as a rule, made to the church in the period before tithes were introduced, and that those payments were made as offerings on altars.84 In that case, one would have reasonably expected a higher number of coins predating AD c. 1200 and a clearer connection with the earliest church buildings.85

322  Nanouschka M. Burström It seems safe to claim that coin use in churches proliferated when the parishes formed and the congregation took over control of the finances and the building, and gained increased access to the inner space. This is sustained by the suddenly increasing numbers of coins. But, what was the actual use for coins inside the medieval church? Preserved accounts give an idea as to what incomes and expenses were normal in a medieval Swedish parish church when it comes to the fabrica, the fund for the maintenance of the church, and the running expenses of the cult. Incomes from tithes (in money and in kind) and rents from tenants were the main part,86 supplemented with incomes from offering trunks and collection boxes, from legacies, from renting cattle to the tenants and from lending, e.g., money, iron and corn. Recurrent expenses were for wax (for candles), wine, wafers, incense and myrrh; for fees to the cathedral; for movable effects such as chalices, candlesticks, silk, cloth and sculptures; and for building and maintenance.87 Some of the payments made in coin could have been handled in the church, but it may also be reasonably suggested that many of these transactions took place outside of the church building, e.g. in the presbytery.88 Of the donations to the maintenance and embellishment of the church made in wills, many would be in kind and not in coin.89 This same pertains with regard to the mensa, i.e. the fund and land at the priest’s personal disposal. The only payments involving coins that surely were made inside the church are those put in offering boxes (truncis),90 or placed on collection tables (tabula) and directly on altars. Wangsgaard Jürgensen91 has also pointed to the temporary nature of some collections, like specific devotional objects that could be placed differently during the year. A study from late medieval Kent showed that the acquisition of artwork (in a broad sense) for the parish churches was done in several ways: through donations in kind (e.g. of silverware to be reworked), through cash donations (which were most likely paid outside of the church) and through fundraising by the church, where multiple donations and offerings were important factors.92 Only in the last case may we expect to see any quantity of the payments involved to materialise as coin finds in churches. There were often several altars in the nave, apart from the two by the triumphal arch, which most often survive into later periods. Several saints, international and local, were venerated in even a small parish church; in town churches, there could be numerous altars including private ones paid for by guilds.93 To give just a few examples, there were at least 15 altars in the church of Jönköping in 150194 and several altars apart from St Mary’s in Risinge church in 1513,95 and at least six different altars are specified in Halmstad church between 1455 and 1533.96 In Lund cathedral, there were at least 68 medieval altars, and in St Olof’s church, there were possibly nine.97 Apart from private practices connected with indulgence, direct offerings may have been made in connection with masses for the dead, as indicated, e.g., by Norwegian wills where money is bequeathed for this specific purpose.98 The practice of offering in the church was (in particular during the later Middle Ages) closely related to the practice of indulgence, redemption from sin in return for certain performances, including offerings of money and candles. The extent of the indulgence and the expected performance were specified in letters issued by church officials such as the Pope, cardinals or bishops. Such letters are therefore a major source of information about the patron saints of different churches, which saints and saints’ days were celebrated there, the efforts expected from the sinner, what the money was meant for and the sums involved. The reward could be greater on certain days than others, e.g. higher on the day of the patron saint than on other days, and

Building for glory – Gränna and Arby  323 the money could be used for different purposes but was often explicitly intended for the fabrica.99 Usually, a specific church and a specific object (mainly sculptures or the Host) or altar would be indicated, and the letters of indulgence were therefore a great economic asset to the churches. Some churches could have some 40 indulgence days annually; most common were the days of Christ and Mary, and the consecration day of the church, but many other saints appear, including domestic and local saints.100 Examples of what performances qualified for indulgence include attending mass, payment to the fabrica, praying, participating in building works, walking around the cemetery praying for those buried there, accompanying the priest to the sick and participating in funeral services. Apart from what could be gained directly in monetary terms and for the embellishment of the church, the indulgence letters were thus also an important tool for the church to increase the number of participants in services.101 Robert Swanson underlines, based on English sources, that the surviving written documentation of indulgences, such as letters, granted privileges, collecting licences and other records, is fragmentary and spread over numerous institutions, and the number of collections made (and the amounts involved) is therefore often greatly underestimated. It follows that it is important to include in the discussion material data such as church paintings, art, architecture, prayer beads, seals and inscribed objects as direct or indirect evidence,102 and to these examples of material data, we may certainly add coins. Pennies for heaven In the cases of Gränna and Arby, the practice of offering coins on altars or in adjacent boxes is reflected and confirmed through the coins found in the filling of graves, which were dug in positions in front (west) of known altars to the north and south of the triumphal arch, and, in Arby, in front of the Rood and in relation to the medieval high altar (which was not investigated in Gränna). Assuming that this correlation has general value, clusters of coins could be used to suggest further, otherwise undocumented, altars in the north-west corner of the nave (both churches) and in the porch (Gränna). The continued, post-Reformation, use of coins in the same positions may suggest continued use of the altars or, more likely, of offering boxes adjacent to sculpture in the same places, in addition to the collection bags passed around the pews. The coins continue to end up in graves and grave chambers until no more burials are allowed in the church.103 A few late coins may indicate that some burials were allowed in old (family) chambers even after the formal decision was taken to put an end to the custom. This prohibition seems to be a more meaningful factor in the appearance of coins than whether the floor was laid in wood or stone. Surely coins were collected in Gränna after 1844, too, but this is not reflected in the coin finds since no in-church burials took place after that time.104 Similarly, the latest coin in Arby is from 1827 despite the fact that a wooden floor was also used after that. Relating to the first question of this chapter – whether different managements of the church resulted in different coin uses – it thus seems fair to claim, based on this case study, that coins appear in Swedish churches mainly after the parish “reform” when a larger group of people could access the building, when the buildings often were rebuilt for this purpose and when priest and church were no longer maintained by a single lord or family, but through the mensa, fabrica and offerings made by the congregation. Increasingly, taxes were to be paid to the Church superstructures of

324  Nanouschka M. Burström diocese and papacy. It is suggested that at this point, offerings of food and for hosts and candles were not enough and could not be stored satisfactorily, which may explain the rapid increase in coin use inside churches. This is reflected in the coin finds, which may be used to trace this process while also pointing at previously undocumented points of worship in the church space. The increase of coin use continued in the next century,105 demonstrating the usefulness of coins for those purposes as well as the increased availability of coins for the broader population. The conclusion here is that the main reason why coins were deposited in churches is that they were used for religious offerings according to their time’s practice, on altars, in offering boxes and in collection bags and that they ended up under the floors not by intentional deposition, but mainly as a result of grave-digging and other works inside the church. Regarding the second question of the chapter – the relation between coin finds from churches and general coin use in society – the composition of the finds allows us to think that the coins inside the churches to some extent, although not fully, reflect the monetary situation outside. In Arby, situated in the periphery of the medieval Swedish state, but close to the main water trading routes of the Baltic Sea, more than half of the medieval coins are foreign, most from Germany, but significant portions also come from Gotland and Gotland/Denmark.106 These were found in some quantities in the chancel (around the medieval altar), making it less plausible that the foreign coins derive from individual worshippers passing by. This situation remains all through the Middle Ages, and Swedish official coinage dominates totally only from the Reformation onwards.107 In Gränna, in the core of the Swedish state, a larger part of the coins were Swedish all through the Medieval period, and exclusively so from c. 1450. Here, however, foreign coins played a role in the early and high Middle Ages. The differing sizes of the excavated areas of the two churches should not affect the validity of the samples. These differences between the sites seem to fit in with what could be imagined based on their known historical situations and would thus reflect a normal, circulating, body of coins. But it is noteworthy that although the numbers of coins differ between the centuries, in most cases the differences are very small compared to what would presumably reflect the situation in the increasingly monetised society outside of the church. For example, in Arby, the same number of coins (27) was found for the fourteenth century and for the eighteenth century, and likewise in Gränna, almost the same number of coins (14/13) derives from the thirteenth century and the eighteenth century. This underlines the fact that coin use in the church does not necessarily mirror coin use outside the church; instead, the religious context follows its own logic and needs.108 It is possibly of importance for the interpretation of the function of coins in churches that in Gränna, the largest coin groups (Magnus Eriksson’s coins from the later part of the fourteenth century and those of Karl XI from the later part of the seventeenth) immediately precede in time the two major rebuildings recorded before the nineteenth-century fire. For a long time, the silver penny (penning) was the only denomination in use, but it is often assumed that depreciated, small or foreign coins would have been used primarily in churches, as the act of offering would have been more important than the actual value of the coin.109 As contemporary hoards are scarce, and settlement finds even scarcer, it is difficult to make conclusive comparisons by contrasting the types of finds. Keeping in mind the intended uses of the coins (paying for candles, wine,

Building for glory – Gränna and Arby  325 masses, construction works, etc., as exemplified above), valueless coins probably would not have been very popular with the priests.110 Low-value coins, however, may well have been used, in particular considering the small amounts of cash assumed to have been available to the ordinary rural population, including women, children and servants, to whom the small coins may have been the only available option.111 In the post-Reformation period, there are more denominations and a bi- or tri-metallic system was in use, so what stands out quite clearly is that silver coins are few in the finds (and gold almost non-existent), while copper coins become most frequent in the church finds. Again, those are the less valuable coins, even if they are much larger than the medieval ones. The interpretation here is in no way straightforward, but two possibilities might be considered: either that the low denominations reflect the common and plain character of the practice and the practitioners or that low denominations were preferred based on religious reasons, such as recalling the piety of the biblical poor widow’s mite.112 The traditional explanation that small coins dominate the finds mainly because they are more easily lost than big ones seems less valid when considering the numerous finds of copper coins, which often weigh ten times or more than medieval or contemporary silver ones. The case study of Arby and Gränna churches has shed light on several theoretical and methodological topics of relevance for cultural history, numismatics and (church) archaeology. However, it should not be forgotten that these cases fundamentally also shed light on the interplay between manifestations of worldly and divine power, and on how building, rebuilding and embellishment of the church are acts of worship and reverence in themselves. The materiality of coinage played a very active role in this. Through the coins, contact with the divine could be established, personal redemption promoted and spiritual life of the community maintained. Though the building itself was only a focal point for this contact with the Lord, it was still of major importance to common and individual devotional practices. It was the duty of worldly lords to set the example by contributing large amounts to building and embellishment, and the duty of the congregation to contribute what they could. Together, they created a glimpse of heaven on earth.113 In Arby, a member of the local elite probably had the first church built, while the village or members of the new parish substituted that with a more impressive one in stone, suited to host the congregation and the local lord, as well as the Lord above. In Gränna, the church may have been initiated by local elite or the royal family, but was also meant to host a much larger group of people from the outset, possibly villagers and inhabitants of the Husaby estate. The building, magnificent for its time, was subsequently rebuilt even grander by the local lord in the seventeenth century to fit the needs and status of the newly established town, the practices of the reformed church and a modern feeling for how to express God’s glory on earth.

Acknowledgements I wish to thank my colleagues in the “Religion and Money” research team for feedback, encouragement and sharing of expertise: in particular, Svein H. Gullbekk, Christoph Kilger, Henrik Klackenberg, Henriette Rensbro, Jon-Anders Risvaag and Håkon Roland for their prepared workshop comments on parts and drafts of this chapter; and Svein H. Gullbekk and Giles Gasper for inviting me to the project. I would also like to thank one anonymous reviewer for positive and constructive

326

Nanouschka M. Burström

feedback on the present text. Special thanks are due to Björn Varenius for checking up long forgotten details about the Gränna excavations, to Anders Andrén for generous help with references, to Ann-Catherine Bonnier, Henrik Klackenberg, Sven-Erik Pernler and Göran Tegnér for helpful responses to questions regarding Swedish seals and patron saints, to Torgny Lundqvist and Bertil Badman for the photographs of Gränna church and seals and to the staff of the ATA, JLM, KMB, KMK and SHM for help with their archives, finds and photographs. Any remaining errors are my own.

Notes 1 “Medieval” in this context refers to c. 1050–1520, i.e. from the end of the Viking Age to the end of the Catholic period. “Early medieval” in Swedish/Scandinavian terms indicates c. 1050–1250. 2 The term “medieval Sweden” here encompasses the historical provinces of Dalarna, Dalsland, Gotland, Gästrikland, Hälsingland, Närke, Småland, Södermanland, Uppland, Värmland, Västergötland, Västmanland, Ångermanland, Öland and Östergötland. From the twelfth century until the beginning of the nineteenth, it also included several provinces in present-day Finland. The remaining provinces of present-day Sweden, such as Jämtland in the north or Skåne in the south, were from time to time parts of Norway or Denmark. 3 Karin Andersson, “Kalmarkustens kyrkor under tidig medeltid,” Hikuin 9 (1983): 189–90. 4 Karin Andersson and Otto Svennebring, Arby kyrka: Södra Möre härad, Småland (Stockholm: Almqvist & Wiksell International, 1989), 15–16. For details about private churches in Scandinavia, see KHLNM XIII: 462–67 (entry Privatkyrka) and 136–44 (entry Patronatsrätt); Arvid Bäckström, Utredning rörande de i vårt land utom Skåne, Halland och Bohuslän förefintliga patronatsrättigheter och därmed jämförliga kallelserätter (Stockholm: K.L. Beckmans boktryckeri, 1914), 9–19; Gunnar Smedberg, Nordens första kyrkor: en kyrkorättslig studie (Lund: CWK Gleerups förlag, 1973). 5 DMS (Medieval Sweden) 4:1: 28–36; SDHK (Swedish medieval charters, online edition) 5275, 5276. http://riksarkivet.se/sdhk 6 Stefan Brink, “Nordens husabyar – unga eller gamla?” in En bok om husbyar, edited by Michael Olausson (Stockholm: Riksantikvarieämbetet, 2000), 68, 71; Åke Hyenstrand, Centralbygd-randbygd. Strukturella, ekonomiska och administrativa huvudlinjer i mellansvensk yngre järnålder (Stockholm: Almqvist & Wiksell, 1974), 118; Asgaut Steinnes, Husebyar (Oslo: Grøndahl, 1955), 136; Keith Wijkander, Kungshögar och sockenbildning. Studier i Södermanlands administrativa indelning under vikingatid och tidig medeltid (Nyköping: Södermanlands museum, 1983), 136–38. 7 KHLNM VII: 94–96. For a more recent compilation of ideas on husabyar, see Olausson, En bok om husbyar. 8 In analogy to Swedish and Norwegian examples quoted by Brink, “Nordens husabyar,” 68–70, it may be suggested that Gränna was the original name of the estate/village and is retained in the area and church. 9 SDHK 606. 10 Johan Baer donates a part of the Husaby estate to Alvastra monastery, SDHK 2704. 11 “The holy church in Grenna” (translation by author), Ortnamnsarkivet entry Gränna; Nils Fadersson donates parts of Husaby to Gränna church, SDHK 29044. 12 The coin finds from most medieval Swedish churches (excluding town churches, monasteries and provinces not part of medieval Sweden) were evaluated for their quality of documentation in Henrik Klackenberg, Moneta nostra: monetarisering i medeltidens Sverige (Stockholm: Almqvist & Wiksell International, 1992). An additional 17 church finds from medieval Sweden are listed in Kenneth Jonsson, “Myntfynden i landsortskyrkor i det medeltida Sverige,” Myntstudier 11 (2011): 1–16. An overview of coin finds from all provinces of present-day Sweden may be found in the series Sveriges Mynthistoria Landskapsinventeringen, published by the Royal Coin Cabinet (KMK), Stockholm.

Building for glory – Gränna and Arby

327

13 The documentation and data used for this study consist mainly of the primary documentation kept in Antikvarisk-Topografiska Arkivet (ATA) and Jönköpings Läns Museum (JLM), and excavation reports; see Andersson and Svennebring, Arby kyrka; Nina Edenstig, “Barngravar i Gränna kyrka, senmedeltid – 1600-tal. En osteologisk analys av skelettmaterialet från vapenhuseet,” Seminar paper in historical osteology, (Stockholm: Stockholm University, 1994); Lisa Hartzell, “Osteoarkeologisk analys av gravar i Gränna kyrka och kyrkogård från medeltid och framåt,” Seminar paper in osteoarchaeology (Stockholm: Stockholm University, 2004a); “Osteologisk rapport över gravar från Gränna kyrka och kyrkogård,” Unpublished report (Stockholm University, 2004b); Cecilia Ring, Arby kyrkogård. Arby församling, Växjö stift, Kalmar län. Kulturhistorisk inventering av kyrkogårdar/ begravningsplatser i Växjö stift 2006 (Kalmar: Kalmar läns museum, 2006); Björn Varenius, Arkeologisk undersökning Gränna kyrka, Archaeological report 1992, no. 22 (Jönköping: Jönköpings läns museum, 1993). The coins and most of the finds were examined and photographed by the present author in the KMK and in the Swedish History Museum (SHM). Some finds from Gränna are kept in the JLM, which kindly provided photographs of selected items. The KMK kindly provided earlier photographs of coins. Also important were the Web resources Kulturmiljöbild (KMB) with photographs published online by the Swedish National Heritage Board, and Sök i Samlingarna containing information about the finds and made available by the SHM. Several overviews include information on Arby and Gränna churches; see in particular Marian Ullén, Medeltida träkyrkor 1, Småland samt Ydre och Kinda härader i Östergötland (Stockholm: Riksantikvarieämbetet, 1983); Andersson and Svennebring, Arby kyrka; Markus Dahlberg and Kristina Franzén, Sockenkyrkorna: kulturarv och bebyggelsehistoria (Stockholm: Riksantikvarieämbetet, 2008). Specific aspects of the coin material were discussed in Monica Golabiewski, “Kompletteringar till publicerade svenska medeltidsmynt,” Nordisk numismatisk unions medlemsblad 7 (1980): 146–52, and Lars O. Lagerqvist, “Myntet med korsfanan,” in Flaggor – från fälttåg till folkfest, edited by Leif Jonsson (Lidköping: Läcköinstitutet, 1993), 56–7, “Myntet från Gränna: myntet med korsfanan,” in Det nära förflutna: om arkeologi i Jönköpings län, edited by Mikael Nordström and Linnéa Varenius (Jönköping: Jönköpings läns museum, 1997), 190–93, and “Grännamyntet ännu en gång,” in Myntstudier: festskrift till Kenneth Jonsson, edited by Tukka Talvio and Magnus Wijk (Stockholm: Swedish Numismatic Society, 2015), 177–80. 14 The date when finds were discovered was also stated, which proved helpful for the author as less clear indications could be interpreted based on an understanding of which part of the church was being excavated on a particular day. 15 ATA, unpublished excavation report draft, 42 (my translation). Although the original report could not be located in the ATA (2014), copies of the find lists are kept there. Parts of the report, unfortunately not including detailed grave descriptions, were kindly made available by Dr Henrik Klackenberg as copies and excerpts made in the ATA in the 1980s. 16 The coin list included in the published report (Varenius, Arkeologisk undersökning, Appendix 1) does not include all coins (missing the last page of the KMK coin list, nos. 109–26). 17 Rensbro and Moesgaard, this volume, ch. 9. 18 Jonsson, “Myntfynden,” 4, 13. 19 Note the perforations to transform it into a spangle. 20 No. 132 is included with the fourteenth-century coins in the diagram. 21 Svealand is a large region in Middle Sweden, notably including Lake Mälaren and the major medieval cities in its vicinity such as Sigtuna, Stockholm, Strängnäs, Uppsala and Västerås. 22 KMK dnr 711-755-2005. The coin is cut in half and was found in 2002. See Jonsson, “Ytterligare ett exemplar av Grännabrakteaten,” Svensk Numismatisk Tidskrift 6 (2003): 133. 23 Lagerqvist, “Myntet från Gränna,” and “Grännamyntet ännu en gång”; Jonsson, “Grännabrakteaten,” 133, suggests that it might be earlier, considering the style and the fact that no other archiepiscopal coins postdating the 1220s are known from the area. The excavation data from Gränna provide no conclusive evidence, but perhaps hint at an earlier dating than Lagerqvist’s since the coin was found in layer 16, the oldest floor layer,

328 Nanouschka M. Burström

24 25 26

27 28 29

30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

underneath layer 13, which at the bottom contained a Gotlandic coin dated 1220–1250, and also included two coins from 1275 to 1290 (Varenius, Arkeologisk undersökning, 19, Appendix 2). For the dating of the Gotlandic coin, see Nanouschka Myrberg, Ett eget värde. Gotlands tidigaste myntning, ca 1140–1220 (Stockholm: Stockholm University, 2008), 140. Jonsson, “Myntfynden,” fig. 16. Andersson and Svennebring, Arby kyrka; Ullén, Medeltida träkyrkor 1. This process is not unique to Arby church; one illustrative example of such a sequence of events is provided by Silte church on Gotland. See Mats Bergman, Sveriges kyrkor. Gotland. Silte kyrka (Stockholm: Almqvist & Wiksell International, 1992); Gustaf Trotzig, “En stavkyrka i Silte,” Gotländskt arkiv 44 (1972): 73–88. The first-floor chamber of the west tower is characterised as a balcony by the excavator. The oldest extant pew list is from c. 1600. The medieval opening was c. 3 m wide. A chancel screen was surely present in 1613 when a new one was made, but it cannot be established whether that was a continuous medieval practice. Anna Nilsén, Focal Point of the Sacred Space. The Boundary between Chancel and Nave in Swedish Rural Churches from Romanesque to Neo-Gothic (Uppsala: Uppsala University, 2003), argues that chancel screens were commonly removed with the introduction of the Gothic style in the fourteenth century, but reintroduced after the Reformation of the mid-1500s. The red limestone cover for the original relics is reused and inset in the present altar. The original reliquary (a lead box) should be kept in the Kalmar county museum but could not be located (2015). From Gotland, a so-called “paradise” font; see Andersson and Svennebring, Arby kyrka, 77–79 for the font’s history. The rood has close parallels in Gotland (Öja and Fröjel churches). Andersson and Svennebring, Arby kyrka, 67–68. Now in the Kalmar county museum but probably originally belonging to Arby church. Andersson and Svennebring, Arby kyrka, 65, 67, 70. Nils-Gustaf Stahre et al., Rannsakningar efter antikviteter, band III (Stockholm: Kungl. Vitterhets Historie och Antikvitets Akademien, 1992), 71. Ring, Arby kyrkogård, 12. Sven-Erik Pernler pers. comm. 11/1/2015. An alternative reading of the inscription is “Johan gave me [to the church].” The belfry is too frail to allow closer examination of the bells; see Andersson and Svennebring, Arby kyrka, 96–97. Beneath the triumphal arch, trench 4, grave 8. Hartzell, “Osteoarkeologisk analys,” 17; Varenius, Arkeologisk undersökning, 44–45. Varenius, Arkeologisk undersökning and Björn Varenius, “Gränna kyrka – ett historiskt dokument,” in Det nära förflutna: om arkeologi i Jönköpings län, edited by Mikael Nordström and Linnéa Varenius (Jönköping: Jönköpings läns museum, 1997), 184. Tage Grennfelt et al., Gränna–Visingsö historia (Stockholm: Liber Förlag, 1980), 27; Rannsakningar III, Arby och Hagby socknar, 71. Varenius, Arkeologisk undersökning, 5–7. Ing-Marie Nilsson, “Ströja – en aristokratisk medeltidskyrka,” in Grevars och bönders tempel. En bok om Brahekyrkan på Visingsö, edited by Robin Gullbrandsson (Jönköping: Jönköpings läns museum, 2013), 83. Varenius, Arkeologisk undersökning, 30–34. Varenius, Arkeologisk undersökning, 17–19, Appendix 2. Varenius, Arkeologisk undersökning, 57–58. A window may have existed also on the northern side of the chancel, at some point bricked up to form the niche directly opposite the south window. The window frame was taken out of a secondary position in the sacristy in 1927. The elaborated decorations indicate that it may have been the east window of the apsis, and it has been suggested that it was made by the well-known artisan Bestiarius (ATA dnr 2999/1927); Lise Gotfredsen and Hans Jørgen Frederiksen, Troens billeder. Romansk kunst i Danmark (Herning: Systime, 1987), 49–53; Tage Grennfelt et al., Gränna – Visingsö historia, 42. The apsis window openings were slightly widened, probably already in the Middle Ages.

Building for glory – Gränna and Arby

329

48 Varenius, Arkeologisk undersökning, Pl. 2. 49 The south altar was still there and used to support the pulpit in the 1980s, but was covered with cement shortly after Torgny Lundqvist (pers. comm. 11/1/2015). The north altar is indicated by coin finds; see below. 50 A pointed, “gothic” arch window frame (Varenius, Arkeologisk undersökning, 53, fig. 46), found during excavations in the churchyard on the north side of the church (original position unknown), may indicate either that the north wall was provided with windows at about the same time as the north portal was opened or that they were remade then. 51 Bengt Ingmar Kihlström, “Den apostoliska trosbekännelsen i vår medeltida kyrkokonst,” Fornvännen 47 (1952): 129–52; Bengt Söderberg, De gotländska passionsmålningarna och deras stilfränder. Studier i birgittinskt muralmåleri (Stockholm: Wahlström och Widstrand, 1942), 193–226. Wangsgaard Jürgensen and Kilger, this volume, chs. 2 and 5. 52 Nils Månsson Mandelgren, Monuments Scandinaves du moyen age avec les peintures et autres ornements qui les décorent (Paris, 1862); Söderberg, De gotländska passionsmålningarna, 206–209. 53 I thank the Rev. Torgny Lundqvist for providing me with photographs of the seals, and Henrik Klackenberg, Inga-Lill Pegelow and Göran Tegnér for sharing with me their professional considerations regarding the dating and iconography of the seals. 54 ATA, inventory list dated 17 April 1828; see also the ATA photograph archive. The list includes many other post-Reformation items and pays much interest to books, drawings and sepulchral tablets. 55 Rensbro, this volume, ch. 3. 56 An upper wall plate from the nave of the stone church was dendrochronologically dated to “after 1183”; Andersson and Svennebring, Arby kyrka, 43. 57 C-14 dating indicated that the skeleton was buried in 1130, but this seems unlikely judging from the grave goods, the arm position and the general context of the burial. 58 Similar brooches were found in Tölö church, western Sweden (SHM 2461), and in Rösta church, northern Sweden (SHM 9007). 59 Grave I was possibly dug a little later as it seems to cut the feet of grave II. 60 Post-medieval graves were not numbered on the excavation plan, and since the full grave descriptions could not be found and used for this investigation, I avoided giving graves new numbers that might create confusion later on. 61 Burial inside the church was prohibited in 1775. 62 On concentrations of post-medieval coins in the chancel, see Ingvardson and Jonsson this volume, chs. 10 and 13. 63 Dr Mona Bramer Solhaug pers. comm. 6 November 2014. Mona Bramer Solhaug, “The Architectural Setting of Baptism: Rituals, Norms and Practices in Scandinavia c. 1050– c. 1250,” in Coins in European Churches – Religious Practice and Devotional Use of Money, edited by Benedikt Zäch and Svein H. Gullbekk. Schweizerische Numismatische Rundschau 99 (2019): 99–123. 64 Varenius, Arkeologisk undersökning, 43. 65 Mattias Karlsson, Konstruktionen av det heliga. Altarna i det medeltida Lunds stift (Lund: Lund University, 2015), 208–209. 66 Coins from c. 1250 until c. 1720. 67 Edenstig, “Barngravar i Gränna kyrka”; Hartzell, “Osteoarkeologisk analys” and “Osteologisk rapport över gravar från Gränna kyrka”; Varenius, Arkeologisk undersökning, 46. 68 Karlsson, Konstruktionen av det heliga, 220–23, quotes several examples of altars and side chapels in porches in his thesis on medieval altars in the diocese of Lund. 69 Varenius, Arkeologisk undersökning, 18, and find lists in ATA. 70 Risvaag, this volume, ch. 6, on Høre stave church where coins could more securely be connected with the erection of the church, as in the stave churches from Ringebu, Kinsarvik and Uvdal. 71 Risvaag, Roland, Klackenberg and Jonsson, this volume, chs. 6, 7, 12 and 13; Anna Nilsén, Kyrkorummets brännpunkt. Gränsen mellan kor och långhus i den svenska landskyrkan, från romantik till nygotik (Stockholm: Kungl. Vitterhets Historie och Antikvitets Akademien, 1991), 79–80; Jürgensen, this volume, ch. 2. 72 Klackenberg, this volume, ch. 12. 73 Eeva Jonsson, this volume, ch. 13.

330 74 75 76 77

78 79 80

81 82

83 84 85 86

87 88 89 90 91 92 93 94 95 96 97 98 99

100 101

Nanouschka M. Burström Klackenberg, Moneta nostra, 34–38. Gitte Tarnow Ingvardson, this volume, ch. 10. Two instances of early burial in Gränna seem to precede the church. For Norwegian examples, see Svein H. Gullbekk, “The Church and Money in Norway c. 1050–1250: Salvation and Monetisation,” in Money and the Church in Northern Europe 1000–1200: Practice, Morality and Thought, edited by Giles E.M. Gasper and Svein H. Gullbekk (Farnham: Ashgate, 2015), 234–35. Coins were sometimes included in graves during the Viking Age (AD c. 750–1050). Less than 100 coins in total; Klackenberg, Moneta nostra, Catalogue (in extenso). Klackenberg, Moneta nostra, 68, 93, 118–19, 144, 167; Kilger, this volume, ch. 5. Notably, the rural finds in Uppland are fewer and the general increase in number appears later than in the southern parts of Sweden, despite the central character of the area with important trading sites and towns such as Sigtuna and Stockholm. KHLNM XVI, 374–85; Stefan Brink, Sockenbildning och sockennamn. Studier i äldre territoriell indelning i Norden (Uppsala: Almqvist & Wiksell International, 1990), 369– 70; Jón Viðar Sigurðsson, Kristninga i Norden 750–1200 (Oslo: Samlaget, 2003), 74–78. KHLNM XIII, 462–67; Brink, Sockenbildning och sockennamn, 369, “Sockenbildningen i Sverige,” in Kyrka och socken i medeltidens Sverige: en samling uppsatser, edited by Olle Ferm (Stockholm: Riksantikvarieämbetet, 1991), 115–18; Gunnar Smedberg, Nordens första kyrkor. En kyrkorättslig studie (Lund: CWK Gleerups förlag, 1973), 162–66, 179–81, 184–91. SDHK 642; Tage Grennfelt et al., Gränna – Visingsö historia, 53. Olav Tveito, “Mynter i messen. Kirkefunnene som bidrag til kunnskap om offerpraksis og kirkeskikker (11.–17. årh.),” Historisk Tidsskrift 94 (2015): 389–90. Ingvardson, this volume, ch. 10; Klackenberg, this volume, ch. 12; Risvaag, this volume, ch. 6. Two churches in one case study got 85% of their incomes from these two entries during a 15-year period in the fifteenth century. See Göran Dahlbäck, “Stockapenningar, tjärtunnor och beläten. Något om den senmedeltida sockenkyrkans ekonomi,” in Kyrka och socken i medeltidens Sverige en samling uppsatser, edited by Olle Ferm (Stockholm: Riksantikvarieämbetet, 1991), 359. In the same case study, the building maintenance and the investments in movables were the largest expenses, 21–40% and 23–39%, respectively. Göran Dahlbäck, “Stockapenningar, tjärtunnor och beläten,” 373. But Wangsgaard Jürgensen (this volume, ch. 2) is referring to tithes as being offered, entirely or symbolically, to the priest in the chancel arch on specific “offering days.” Judy Ann Ford, “Art and Identity in the Parish Communities of Late Medieval Kent,” in The Church and the Arts, edited by Diana Wood (Oxford: Blackwell, 1992), 227–32. For details about truncis in Scandinavia and medieval coin finds made in those, see KHLNM XIII, 172–75. Wangsgaard Jürgensen, this volume, ch. 2. Ford, “Art and identity,” 230–34. Lars Bisgaard, De glemte altre. Gildernes religiøse rolle i senmiddelalderens Danmark (Odense: Universitetsforlaget, 2001); Tveito, “Mynter i messen,” 391, 402, 405. SDHK 34376. SDHK 37489. SDHK 31575, 32885, 33156, 35243. Karlsson, Konstruktionen av det heliga, 195, 431–35. Tveito, “Mynter i messen,” 404–405. Carl-Gustaf Andrén, “De medeltida avlatsbreven – instrument för kyrkans verksamhet,” in Investigatio memoriæ patrum: libellus in honorem Kauko Pirinen, edited by Aimo Halila and Pentti Laasonen (Helsinki: Societas Historiae Ecclesiasticae Fennica, 1975), 202–206; Anders Fröjmark, Mirakler och helgonkult. Linköpings biskopsdöme under senmedeltiden (Uppsala: Uppsala University, 1992), 137–40. Andrén, “De medeltida avlatsbreven,” 208–11. Andrén, “De medeltida avlatsbreven,” 212–15 and Carl-Gustaf Andrén, “Avlatsbreven från Västgötadelen av Skara stift,” in Avlatsbreven från Västgötadelen av Skara stift, edited by Johnny Hagberg (Skara: Skara stiftshistoriska sällskap, 2006), 23–28.

Building for glory – Gränna and Arby

331

102 Robert N. Swanson, Indulgences in Late Medieval England. Passports to Paradise? (Cambridge: Cambridge University Press, 2007), 77–109 [in particular 79–81, 107–109], 266–77, 422–39. 103 A Swedish law prohibiting in-church burials was passed in 1783, but the practice continued in most churches for some time after that, while families still held rights to the chambers. Göran Lindahl, Grav och rum: svenskt gravskick från medeltiden till 1800-talets slut (Stockholm: Almqvist & Wiksell, 1969), 200–202, 204, 206. 104 The last burials found in the excavations were found in trench 4 and were probably dug in the early nineteenth century. Varenius, Arkeologisk undersökning, 39. 105 Klackenberg, Moneta nostra, 68, 93, 118–19, 144, 167. 106 The island of Gotland had its own coinage all through the Middle Ages, but was in periods under Danish control; see Myrberg, Ett eget värde. The finds from Denmark, Germany and Gotland/Denmark belong to the fourteenth and fifteenth centuries, not the Early Medieval period. 107 Klackenberg, Moneta nostra, 87. 108 Ingvardson, this volume, ch. 10. 109 Jonsson, “Myntfynden,” 5. 110 Georg Galster, Unionstidens udmøntninger. Danmark og Norge 1397–1540. Sverige 1363–1521 (Copenhagen: Danish Numismatic Society, 1972), 220, note 70a; Jørgen Steen Jensen, De skriftlige kilder til Danmarks middelalderlige møntvæsen: et udvalg 1085–1500 (Copenhagen: Nationalmuseet, 1989), no. 670. 111 Wangsgaard Jürgensen, this volume, ch. 2. 112 Gullbekk et al. and Wangsgaard Jürgensen, this volume, chs. 1 and 2. 113 Andrew Martindale, “Patrons and minders: the intrusion of the secular into sacred spaces in the late Middle Ages,” in The Church and the Arts, edited by Diana Wood (Oxford: Blackwell, 1992), 143–46.

12 Coin finds and coin use in the medieval round church at Klåstad, Östergötland, Sweden Henrik Klackenberg

Introduction The Klosterstad estate is situated c. 4 km east of Vadstena in Östergötland (Figure 12.1), in an open agricultural landscape where views stretch for mile upon mile. This is one of the most fertile areas of Sweden with many traces of a prehistoric past. Archaeological remains and place names suggest early settlement of the region, which

Figure 12.1 Map of southern Sweden. Map: Steinar Kristensen, UiO: Museum of Cultural History

Coin finds, coin use – a medieval round church  333

Figure 12.2 The seal matrix of the Knight Abjörn Sixtensson, AD c. 1295. Photo: by permission SHM

in the Middle Ages became part of the diocese of Linköping as well as a core area of Östergötland and of the entire kingdom.1 The Klosterstad estate, which received its present name in the eighteenth century, was originally named Klåstad and is first mentioned in written sources in 1296, in a document relating to an inheritance settlement. Then, Klåstad was a curia, i.e. a manor owned by an aristocrat. The owner of this and various other estates was the knight Abjörn Sixtensson of the Sparre family (Figure 12.2), councillor of the realm (riksråd) and later the highest ranking official (drots) of Dukes Erik and Valdemar, and thus a member of the social elite of the realm. In the 1380s, the estate was donated to the monastery of Vadstena by his grandchildren. It remained a part of the monastic estate until the 1540s when it was transformed into a royal estate as one of the consequences of the Reformation. 2 What makes this estate interesting and relevant for a study of the relationship between the medieval population, money and the church? Initially, the intriguing answer lay in a field slightly east of the estate buildings. In this field, there was previously a large clearance cairn, adjacent to which limestone fragments of an early Christian grave monument had been found (Figure 12.3). In 1997, archaeologist Rikard Hedvall, from the National Board of Antiquaries, conducted a minor investigation of the cairn searching for more fragments. To his astonishment, he unveiled a hitherto unknown round stone church, concealed under the cairn. The building had been demolished almost to ground level and through the centuries stones from the surrounding field had been deposited on this barren patch creating the clearance cairn, and thus burying the memories of the medieval church. The following five-year archaeological investigation was predominantly financed by the Berit Wallenberg Trust. The excavations were led by Rikard Hedvall, and a small round church with an apse was recorded (Figure 12.4). The building had already been preliminarily dated to the

334  Henrik Klackenberg

Figure 12.3 The clearance cairn at Klosterstad. Photo: Rikard Hedvall

Figure 12.4 The remains of the round church at Klåstad. Photo: Rikard Hedvall

Coin finds, coin use – a medieval round church  335 Early Middle Ages due to its shape, and this was corroborated by the dateable finds, especially the coins.3 The story of the church under the cairn does, however, not end there, since traces of a predecessor were found immediately north of the cairn. Eight postholes and wall ditches indicated a rectangular wooden building with a narrower chancel (Figure 12.5). The nave measured 8.5 x 5.5 metres, while the sides of the square chancel were just over three metres long. A cemetery with c. 240 burials was also associated with the two churches. Several fragments of early Christian burial monuments were found, e.g. an in situ gravestone, covering a burial, dated to the second half of the eleventh century. Through radiocarbon dating, as well as indirectly through the types of graves and coin finds, the wooden church is presumed to have been erected during the first half of the eleventh century and abandoned when the round church was built in the twelfth century. The story continues, although in the case of the round church it ends in the 1560s when it was burnt and not restored. This presumably took place during the Nordic “seven-year war” (1563−1570) when a Danish army, commanded by Colonel (Fältöverste) Daniel Rantzau, ravaged the province of Östergötland. This is supported by the tithe records of 1567, which state that “the tithe from the entire vicarage (prästgäll) of Vadstena was burnt and ravaged by the enemy and not a single grain of crop was delivered to the tithe barn aforementioned year.”4 The royal cadastre records from the 1540s state that Klåstad parish consisted of three farms in each of the villages of Klåstad, Granby and Egeby, and one in Klåstatorp, all within a kilometre of the church. Ownership was almost exclusively in the hands of the monastery at Vadstena: only one of the ten farms in the parish was owned by nobles. Already in 1380, the parish was taken over by the monastery whose appointed vicar administered pastoral care.5 The Klåstad case provides excellent conditions for a numismatic analysis of coin usage in a medieval parish church. It is an example of a well-executed and documented archaeological investigation of two churches, in use from the eleventh century

Figure 12.5 The traces of the wooden church at Klåstad. Photo: Rikard Hedvall

336  Henrik Klackenberg until the Reformation. Unlike most medieval churches still standing, these churches have not been damaged by later rebuilds, burials or renovations, and after the round church was destroyed, the remnants lay untouched under a protective layer of stone. During the well-documented investigations, 104 coins were retrieved. All coin finds were recorded within a coordinate system based on square metres, and in many cases, the exact positions of the coins are known. The chronology spans from c. 1040 to 1520. It is these coins and their use that form the focal point for my analysis. In this way, the church becomes a stage for the “monetary drama.”6 The purpose of the investigation is to provide answers to the following questions, all linked to the question of how coins were used in this medieval church: Where in the church were the coins found and what does this tell us about their use? When were coins first introduced in the church and how did this influx change over time? Do the changes reflect changes in liturgy or something else? How did the coin use change when the church ceased to be a manorial church and instead formed part of the Vadstena monastery estate?

Excavation and documentation The church which was revealed and documented during the excavations was round with a semicircular chancel and an entrance in the south-west. The limestone walls were one-metre thick and built on granite foundations. The length of the church was 15 m and the internal diameter of the round nave was c. 9 m, while the internal measurements of the small chancel were c. 4 x 3.5 m. The documented recorded permanent fittings included benches along the nave walls, the base of the high altar in the chancel, and two limestone and brick side altars by the chancel arch in the easternmost part of the nave. Later, the building was used as a quarry; hence, only three tiers of the wall remain, and even fewer in the chancel. In the western part of the nave, several fragments of a baptismal font were found, which were stylistically dated to the thirteenth century. Only 3 round churches are known in Östergötland, and no more than 13 are known from medieval Sweden. It has been suggested that these churches were modelled on and inspired by the Holy Church of the Sepulchre in Jerusalem although there is no clear evidence for this. Jerusalem was the goal of many pilgrimages and crusades, and the ideas may have been transmitted this way, either directly or via other round churches in Europe. The Order of the Knights Templar, who were closely linked to the Holy Land, had several round churches built on the Continent and in England. What seems to be clear is that the round churches in Scandinavia were built by the social elite.7 In spite of extensive damage to the original stratigraphy, caused by at least 38 medieval burials inside the walls of the church (Figure 12.6), several floor layers were identified. In the nave, three different layers were documented. The earliest was a lime-mortar layer by the wall benches and chancel arch. In the associated layers, 30 coins were found. The oldest coin dated from the latter part of the twelfth century, although several were of much later origin due to the disturbed stratigraphy and ensuing difficulties of defining the floor layer. Primarily in the western part of the nave, traces of a wooden floor were found (Figure 12.7), resting on beams set in grooves in the lime-mortar layer. Traces of such beams were found in several places within the nave, and 16 coins were associated with this floor layer, the oldest of which dates to the fourteenth century. The latest floor was made of red limestone and which

Coin finds, coin use – a medieval round church  337 presumably covered the entire floor space of the nave. Only fragments remained, however, and it is believed to be of fourteenth-century origin. In the chancel, several floor layers were also recorded. A well-preserved ­herringbone-patterned brick floor (Figure 12.8) was preceded by two different layers, one wooden, and one older lime-mortar one, just as in the nave. A layer of bedding sand on top of a presumed wooden floor formed the base for the brick floor. This floor was contemporary with the red limestone floor of the nave, and an earlier height difference between the chancel and the floor of the nave was levelled at this time. Also here, different coins could be linked to various floor levels although it was difficult to distinguish the coins deposited on the oldest lime-mortar floor from the ones relating to the deteriorated wooden floor. Of the nine coins associated with these layers, the

Figure 12.6  T  he church ruin at Klåstad: early and later graves. Dark green signifies burials associated with the older wooden church, while light green signifies burials associated with the round church. Plan by P. Carlsson, Riksantikvarieämbetet, and Steinar Kristensen, UiO: Museum of Cultural History.

338  Henrik Klackenberg

Figure 12.7 The church ruin at Klåstad: coin finds from AD 1150−1300. Plan by P. Carlsson, Riksantikvarieämbetet, and Steinar Kristensen, UiO: Museum of Cultural History.

oldest were minted around AD c. 1300. In the bedding sand under the brick floor, the oldest coins dated from the 1360s. With the exception of the topsoil, the church was manually excavated and recorded using single contexts. A small number of layers were dry sifted, while the stiff clay layers were examined using screens. On several occasions, metal detectors were used. The finds were documented according to layer, level and position, either precisely or related to a square-metre grid. In addition to the 104 coins, a large number of nails were found, as well as a smaller amount of window glass, lead mullion, candle holders, pins, fastenings, buckles, etc. A total of 255 items were found, nails and coins being the most commonly occurring. The window glass fragments were mostly found in the chancel, while the few pins came from the northern part of the nave.

Coin finds, coin use – a medieval round church  339

Figure 12.8 The church ruin at Klåstad: coin finds from AD 1300−1400 and fragments of wooden floor. Plan by P. Carlsson, Riksantikvarieämbetet, and Steinar Kristensen, UiO: Museum of Cultural History.

The coin distribution within the church Since the excavation was documented using a grid system, it is possible to analyse where in the church the coins were found. For a third of the coins, it was impossible to determine where they were found more precisely than within a particular squaremetre grid; hence, their position in the plan is marked in the middle of their respective grid square, while the position of the other two-thirds of the coins is documented in detail. Coins were found throughout the church, but were more concentrated in certain areas. The most obvious concentration was in the chancel where 33 coins, i.e. close to a third of the total number, were found within a metre of the altar base. A second concentration was associated with the chancel arch and the side altars. Another, less obvious, pattern is that more coins were found in the northern part of the nave than in

340  Henrik Klackenberg the southern part. Noteworthy is also that not a single coin was found in the wooden church. One coin, from the first half of the eleventh century, found in a burial inside the round church may, however, be linked to the wooden church due to its orientation. New distribution patterns emerge through chronological stratification of the material (see Figures 12.7–12.9). This demonstrates a change in the chancel around the year AD 1300. Here, only a few coins from earlier periods have been found. The concentration of coin finds from the period after AD c. 1300 in the area by the high and side altars can be expressed as parts of the whole. Thus, only 1 coin of 11 (9%), from the period 1150−1300, was found within one-metre distance of any of the altars. The corresponding figure for the period AD 1300−1400 is 30 out of 51 (59%), and for the period AD 1400−1520, 24 out of 36 coins (67%). The one-metre distance is arbitrarily chosen, but it is not implausible to assume a link between activities near the altars and coins found so close to them. This change in distribution patterns raises questions regarding changes to the coin use in the church, which will be further addressed below. The archaeological documentation also enables an analysis of the coin distribution in the different layers and contexts. Linking the coins to the stratigraphical matrix and sequence of layers in the archaeological report does not provide a clear result, since older floor layers may contain younger coins and vice versa. Therefore, the oldest floor layer of the nave contains coins from the thirteenth century together with fifteenth-century ones, while the layer formed when the church was demolished in the sixteenth century contains coins from the twelfth century. There are several explanations for such obvious anomalies, the most important being the disturbances caused by the large number of burials in the church, where coins in older floor layers may have been mixed with coins from a younger floor layer. When the graves were backfilled, later coins may have ended up lower down in the floor. Additionally, there is the − for field archaeologists − well-known problem of precisely defining the various layers and firmly linking each find to a specific layer.8 Lastly, it has to be pointed out that the desired numismatic dating of anonymous bracteates such as these, regarding both time of minting and circulation, is imprecise. Some coins can only be dated within a 50-year period.

The chronological distribution of the coins and their provenance The oldest coin found in the round church belonged to a different and older context. This coin is a German penny minted 1030−1050, found in a grave belonging to the cemetery of the older wooden church.9 The oldest coins which can be linked to the round church are three bracteates minted in Norway and Götaland between 1150 and 1200. The early thirteenth century is only represented with one coin, which cannot be identified with certainty. The number of coins increases in the latter half of the thirteenth century, and the five coins in question are all bracteates from Götaland minted during the reigns of the Swedish Kings Valdemar and Magnus Birgerson. A bracteate from Reval may, however, also belong to this period.10 The fourteenth and fifteenth centuries are represented by a greater number of finds. The 16 coins that can be linked to the first half of the fourteenth century are all Swedish and minted for Kings Birger Magnusson and Magnus Eriksson.11 Most of these are double-sided pennies, which were not previously minted in medieval Sweden. The practice of minting bracteates, however, was re-established in the latter half of the

Coin finds, coin use – a medieval round church  341 century, and bracteates dominate the coinage of the time. Three Norwegian coins were among the 25 coins from this period.12 The fifteenth- and early sixteenth-century coins are predominantly foreign in origin. The first half of the fifteenth century is dominated by the crown bracteates of Eric of Pomerania (King of the Kalmar Union), and the latter by bullhead bracteates from Mecklenburg. While the minting period of the crown bracteates is known, in 1400−1420, the bullhead bracteates are more problematic and have here all been assigned to the period after 1450, although a few may be of older date. Only one Swedish bracteate was found among the 19 coins dating from the first half of the fifteenth century.13 The bullhead bracteates, which are difficult to date, dominate the period after 1450 although three Swedish bracteates minted 1470−1520 also belong to this period. In total, 21 coins have been assigned to this period. The latest coin found at Klåstad is a klipping minted c. 1520 by the last King of the Kalmar Union, Kristian II.14

The coin use in Klåstad: interpretation and discussion The background account of where and how the coins were found allows us to return to the previous questions regarding the use of coins in the church and how this was linked to church rituals. It is clear that coins have been deposited on the floor of the round church throughout the Middle Ages, in smaller numbers already before the thirteenth century, and in ever-increasing numbers from 1250. During the latter half of the fourteenth century, one coin was deposited every other year on average. This number was slightly lower in other periods. Coins were found across the entire floor area, in both the chancel and the nave, although the concentration of late medieval coins is highest around the high and side altars. I have previously argued that the bulk of these coins were inadvertently dropped as the congregation made offerings on altars or in offertory trunks. Offertory wastage was the term I then chose for this phenomenon.15 With this in mind, let us analyse the developments in the Klåstad church. The three bracteates from the latter half of the twelfth century found in the nave show that coins were already handled in the church at that time, although this limited material makes it difficult to draw any further conclusions. The distribution pattern for the thirteenth-century coins does not allow for any other or further conclusions, other than that also in this period coins were handled in the church. It is interesting that no twelfth-century coins and only two from the thirteenth century were found in the chancel (Figure 12.9). Does this mean that the chancel was not constructed until AD c. 1300, or that coins were not handled here before the fourteenth century, which if so would imply a change in the way coins were handled during mass? There is every reason to believe that the chancel and nave were built at the same time since they share foundations. The excavator’s interpretation is that the chancel floor was more thoroughly cleaned, but in my opinion, there is room for a different interpretation: before AD 1300, the congregation did not have access to the chancel during the time of offering. The question of congregational access to the chancel at various stages of the Middle Ages has been debated by theologians, art historians and archaeologists and their differing views have been dependent on the type of source material used.16 In my opinion, the coin finds at Klåstad show that the chancel was not available for the monetary offerings of the congregation until AD c. 1300. The greater number of coins from the fourteenth and fifteenth centuries provides a basis for more detailed

342  Henrik Klackenberg analyses of the coin use within the church, since it is at this time that areas around the high and side altars stand out as areas where coins were handled and also dropped. This applies to the time after AD 1300 (see Figures 12.7 and 12.8) and relates to my earlier interpretation of offering wastage. It was at the altar or in an adjacent offertory trunk that the congregation made their monetary offerings, and coins were then inadvertently dropped to the church floor and never retrieved. The coin concentration by the chancel arch and the side altars can be explained in the same way, since the coins from the congregation were collected either in offertory trunks or directly in the hand or maniple of the priest, which led to some being lost. In 1514, the Bishop of Linköping recorded the seven days which were considered the right ones for making offerings (Christmas, Candlemas, Easter, Whitsun, All Saints’ Day, St. Michaels’ mass and the Day of the Church patron), when the peasant and his wife should offer one penny each. In addition, everyone who received the Easter Holy Communion should offer one penny at the altar and one directly to the priest.17

Figure 12.9 The church ruin at Klåstad: coin finds from AD 1400−1520 and fragments of brick floor. Plan by P. Carlsson, Riksantikvarieämbetet, and Steinar Kristensen, UiO: Museum of Cultural History.

Coin finds, coin use – a medieval round church  343

Figure 12.10 Portable offertory plate (lat. tabula) from Hemse church, Gotland, fifteenth century. Photo: by permission ATA.

It is harder to explain the presence of coins in other areas of the church, as well as linking them to liturgical practices on any detailed level. A collection of coins in portable offering plates (Figure 12.10), which is known from other parts of the Linköping diocese, could result in coin finds from all floor areas.18 This is a possible explanation, but in my opinion, it is enough to conclude that coins were brought and handled in church since the twelfth century and that what we see are traces of the congregational will to make offerings to the glory of God.

A regional comparison with the parish churches at Rogslösa and Ödeshög Until now we have examined how changes in church rituals can explain the distribution patterns exhibited through the coin finds, but other explanations may also be worth considering. One such explanation concerns whether a change in the status of the church is mirrored in the coin material. The churches of Klåstad were built as part of the private estate, in the time before the parishes had been established. In the

344  Henrik Klackenberg early thirteenth century, at the latest, a small parish was created out of a few villages and farms within a 1 km radius of the church.19 In 1296, we know that the manor at Klåstad, and presumably also the church, belonged to one of the most powerful families in the realm. In the 1380s, the estate, including the church and parish, was transferred to the Vadstena monastery. Is this change reflected in the coin finds? In order to answer this question, a comparison with similar coin finds in the area is necessary. In western Östergötland, within 30 km of Klåstad, the parish churches of Rogslösa and Ödeshög are situated. Both have been excavated and medieval coins found, although the documentation does not allow for spatial analyses of coin concentrations with the same precision as at Klåstad. 20 The churches in Rogslösa and Ödeshög underwent complete investigations in the 1950s, but unfortunately, the coin finds were only documented in relation to the part of the church in which they were found, rather than to a grid pattern. The pattern which is distinguishable, however, is a concentration of coins in the eastern part of the nave, and in the chancel, i.e. the same pattern as at Klåstad. As no discernible differences between the coin distributions in the churches seem to be present, it may be worth comparing the composition of the coin finds to investigate possible differences between the coin use at Klåstad and that of the other churches. Here, the only clear difference is the number of medieval coins retrieved from each church: Klåstad (103), Rogslösa (44) and Ödeshög (24). This is, however, most likely a result of the varying quality and extent of the excavations and thus bears no relevance in this case. The assemblage of the coin finds from the other churches is similar to the one at Klåstad: a few coins from Gotland and Götaland from the period 1150−1250, Swedish coins and a substantial increase in number after AD 1300, especially after 1350, and finally a majority of Danish and German coins from the time of the Kalmar Union in the last century of the Middle Ages. There is thus nothing to suggest a different pattern of coin use at Klåstad compared to the other nearby rural churches. The coins were all part of the same circulation and used in the same church rituals. The fact that Klåstad was originally a small manorial church of the nobility, and later a part of the Vadstena monastery estate, does not seem to have influenced the way the congregation used coins.

Regional comparison between the town church of Skänninge and the monastery at Vadstena Two other cumulative coin finds may be of relevance for comparative purposes, although they are not derived from rural churches. These are the coins found during the excavations at the Allhelgona (All Saints’) church in the town of Skänninge and those of Vadstena monastery. 21 While the finds from the Allhelgona church were documented using a grid pattern, the level of documentation at Vadstena was regrettably low. This was mainly because the investigation focused on masonry and architectural history and no record was made of the locations of coins and other artefacts retrieved. In the case of Vadstena monastery, the majority of coins were not found in the excavations of the monastic church, which makes them more indicative of the type of coins circulated in Vadstena monastery, rather than church offerings. 22 The Allhelgona (All Saints’) church is significantly larger than the round church at Klåstad; nevertheless, there are similarities between these sites, as the Allhelgona predecessor was a wooden church built in the twelfth century as a manorial church

Coin finds, coin use – a medieval round church  345 for nobility, and which later became an annex church for a monastery.23 During the excavations, which were conducted in several stages between 1959 and 1961, 65 coins were found, clearly concentrated in the chancel and the eastern parts of the nave. The composition of coin finds is similar to that at Klåstad, apart from one significant detail: the oldest coin is German and minted 1014−1024. It is probably, just as the one from Klåstad, derived from a burial in the old wooden church.24 The composition of the coin assemblage is similar also for the period 1150−1400: there were early medieval coins from Norway, Gotland and Götaland, followed by a large percentage of Swedish coins from the thirteenth century, which drops significantly in the fifteenth century. The large number of Danish coins, which characterises the fifteenth century at Klåstad, is lacking in the Allhelgona church. This decrease could possibly be explained by the church’s change in status in the Late Middle Ages when it lost its role as the town’s congregational church. This role was instead taken by the new Vårfrukyrkan (Church of Our Lady) around AD 1300, and since the Allhelgona church was annexed to the St. Martin’s monastery in Skänninge in 1326, it became less important.

Figure 12.11 Coins found in the church ruin at Klåstad (not to scale). The find numbers refer to Table 12.1. Photo: Ola Myrin, layout by Frida Roland

346  Henrik Klackenberg Table 12.1 Coin finds from the church ruin at Klåstad (KMK 711-609-2005). Find number in bold means that the coin is pictured in Figure 12.11 Find no.

Value

Provenance

Minted

Reference

Room

Context

259

Penny

Sweden

1150−1200

Nave

K 12 planning layer

43

Penny

Sweden

1167−96

NNÅ 1991 p. 130 LL XII:1b

Nave

7

Penny

Norway

1170−1205

Nave

5

Penny

Sweden?

1200−1300

Nave

K 12 planning layer

308

Penny

Gotland?

1210−80

Schive VIII:72−75 Unidentified bracteate Myrberg II?

K 1 layer of collapse/K 6 floor layer K 6 floor layer

21 26

Penny Penny

Sweden Sweden

1250−75 1250−75

LL XVII:1a LL XVII:4a

305

Penny

Reval

1265−1332

31

Penny

Sweden

1275−90

Haljak 2010:10 LL XVIIIB:1b

39 257 9 20 302 303 304 1

Penny Penny Penny Penny Penny Penny Penny Penny

1275−90 1275−90 1290−1318 1290−1318 1290−1318 1290−1318 1290−1318 1290−1410

8 41 262 268

Penny Penny Penny Penny

Sweden Sweden Sweden Sweden Sweden Sweden Sweden Teutonic Order Sweden Sweden Sweden Sweden

273 348 275

Penny Penny Penny

Sweden Sweden Germany?

1300−25 1300−25 1300−1500

22 34

Penny Penny

Sweden Sweden

1319−40 1319−40

BM KrH ÄI BM KrH ÄI Unidentified coin LL XXVI:4 LL XXVI:1a

45

Penny

Sweden

1319−40

LL XXVI:3a

Nave

278 279 281

Penny Penny Penny

Sweden Sweden Sweden

1319−40 1319−40 1319−40

LL XXVI:1a LL XXVI:1a LL XXVI:1a

Chancel Chancel Chancel

297

Penny

Sweden

1319−40

LL XXVI:3a

Chancel

10 24

Penny Penny

Sweden Sweden

1340−64 1340−64

LL XXVII:11 LL XXVII

Nave Nave

277

Penny

Sweden

1340−64

LL XXVII:11

Chancel

1300−25 1300−25 1300−25 1300−25

LL XVIIIB:1a LL XVIIIB:1a LL XXIII:10 LL XXIII:1 LL XXIII:10 LL XXIII:10 LL XXIII:8 Waschinski 113 BM KrH ÄI BM KrH ÄI BM KrH ÄI BM KrH ÄI

Chancel K 86 foundation/ drainage Nave K 6 floor layer Nave K 1 layer of collapse/K 6 floor layer Nave K 12 planning layer Nave

K 1 layer of collapse/K 6 floor layer Nave ? Nave K 12 planning layer Nave K 6 floor layer Nave K 6 floor layer Chancel K 53 floor layer Chancel K 53 floor layer Chancel K 53 floor layer Nave K 1 layer of collapse Nave Nave Nave Nave

K 6 floor layer K 1 layer of collapse K 55 burial K 7 lime-mortar floor/K 12 planning layer Chancel K 15 brick floor ? ? Chancel K 53 floor layer Nave Nave

K 6 floor layer K 4 wooden floor/K 6 floor layer K 4 wooden floor/K 6 floor layer K 53 floor layer K 53 floor layer K 48 layer of collapse K 46 layer of collapse K 6 floor layer K 4 wooden floor/K 6 floor layer K 53 floor layer

Coin finds, coin use – a medieval round church  347 274 16

Penny Penny

Norway Sweden

1350−1400 1354−63

25

Penny

Sweden

1354−63

48

Penny

Sweden

1354−63

269 271 272 256 283 14 29 290

Penny Penny Penny Penny Penny Penny Penny Penny

Sweden Sweden Sweden Sweden Norway Sweden Sweden Sweden

1354−63 1354−63 1354−63 1360−64 1360−64 1360−64 1360−64 1360−64

307 258 266 286 299

Penny Penny Penny Penny Penny

Sweden Sweden Sweden Sweden Sweden

1360−64 1363−70 1363−70 1363−70 1363−70

301

Penny

Sweden

1363−70

261

Penny

Sweden

1363−80

285 3 36 44

Penny Penny Penny Penny

Sweden Mecklenburg Mecklenburg Mecklenburg

1363−80 1364−1500 1364−1500 1364−1500

289

Penny

Mecklenburg

1364−1500

13

Penny

Sweden

1370−80

280 270 306

Penny Penny Penny

Sweden Sweden Sweden

1370−80 1370−80 1370−80

28 37 12 18

Penny Örtug Penny Penny

Norway Gotland Denmark Denmark

1380−87 1380−1420 1400−20 1400−20

27

Penny

Denmark

1400−20

38 40

Penny Penny

Denmark Denmark

1400−20 1400−20

42 260 263

Penny Penny Penny

Denmark Denmark Denmark

1400−20 1400−20 1400−20

267 287

Penny Penny

Denmark Denmark

1400−20 1400−20

Skaare 290 Chancel K 53 floor layer LL XXVIII:4a Nave K 6 floor layer/K 12 planning layer LL XXVIII:3a Nave K 6 floor layer/K 7 lime-mortar floor LL XXVIII:2a Nave K 1 layer of collapse/K 4 wooden floor LL XXVIII:4b Nave K 50 burial LL XXVIII:2a Nave K 67 burial LL XXVIII:2a Nave K 50 burial LL XXIX Nave K 36 burial Skaare 282 Chancel K 47 bedding sand LL XXX:4a Nave K 6 floor layer LL XXX:1a Nave K 6 floor layer LL XXX:1b Chancel K 46 layer of collapse LL XXX:4a Nave K 38 burial BM KrH ÄII:d Nave K 35 burial BM KrH ÄII:d Nave K 6 floor layer BM KrH ÄII:d Chancel K 47 bedding sand BM KrH ÄII:d Chancel K 46 layer of collapse BM KrH ÄII:b Chancel K 46 layer of collapse BM KrA Äb Nave K 7 lime-mortar floor/K 12 planning layer BM KrA Äb Chancel K 47 bedding sand Saether II:16 Nave K 14 floor layer Saether II:13 Nave K 4 wooden floor Saether II:13 Nave K 1 layer of collapse/K 3 burnt layer Saether II:13 Chancel K 46 layer of collapse BM KrH ÄIIi Nave K 3 burnt layer/K 6 floor layer BM KrH ÄIIh Chancel K 53 floor layer BM KrS Äa Chancel K 53 floor layer BM KrS Äa SW of K 1 layer of collapse Nave Skaare 277 Nave K 6 floor layer LL XXXV:3 Nave K 6 floor layer Galster 5 Nave K 4 wooden floor Galster 5 Nave K 4 wooden floor/K 6 floor layer Galster 5 Nave K 4 wooden floor/K 6 floor layer Galster 5 Nave K 4 wooden floor Galster 5 Nave K 4 wooden floor/K 6 floor layer Galster 5 Nave K 1 layer of collapse Galster 5 Nave ? Galster 5 Chancel K 46 layer of collapse Galster 5 Nave K 12 planning layer Galster 5 Chancel K 47 bedding sand (Continued)

348  Henrik Klackenberg Find no.

Value

Provenance

Minted

Reference

Room

Context

291

Penny

Denmark

1400−20

Galster 5

293

Penny

Denmark

1400−20

Galster 5

294

Penny

Denmark

1400−20

Galster 5

295

Penny

Denmark

1400−20

Galster 5

296

Penny

Denmark

1400−20

Galster 5

298

Penny

Denmark

1400−20

Galster 5

309 6 2 11 15 17 19 30 46

Penny Penny Penny Penny Penny Penny Penny Penny Penny

Denmark Lübeck Mecklenburg Mecklenburg Mecklenburg Mecklenburg Mecklenburg Mecklenburg Mecklenburg

1400−20 1400−1500 1400−1500 1400−1500 1400−1500 1400−1500 1400−1500 1400−1500 1400−1500

Galster 5 Jesse 187 Saether IV:21 Saether IV:20 Saether Saether IV:21 Saether IV:21 Saether IV:22 Saether IV:21

47

Penny

Mecklenburg

1400−1500

Saether IV:21

264 265 282 292

Penny Penny Penny Penny

Mecklenburg Mecklenburg Mecklenburg Mecklenburg

1400−1500 1400−1500 1400−1500 1400−1500

Saether Saether IV:20 Saether IV:20 Saether IV:20

288

Penny

Germany?

1400−1500

4 32

Penny Penny

Sweden Denmark

1410− 1424−

Unidentified bracteate BM KrS YI Galster 8

Chancel K 46 layer of collapse Chancel K 46 layer of collapse Chancel K 46 layer of collapse Chancel K 46 layer of collapse Chancel K 46 layer of collapse Chancel K 46 layer of collapse Nave K 21 altar base Nave K 19 burial Nave K 13 wooden floor Nave K 6 floor layer Nave K 13 wooden floor Nave K 6 wooden floor Nave K 13 wooden floor Nave K 6 floor layer Nave K 4 wooden floor/K 6 floor layer Chancel K 1 layer of collapse/K 15 brick floor Nave K 12 planning layer Nave K 42 burial Chancel K 47 bedding sand Chancel K 46 layer of collapse Chancel K 47 bedding sand

23

Penny

Sweden

1470−1500

284 35 300 276

Penny Penny Skilling Penny

Sweden Sweden Denmark ?

1470−1500 1500−20 1518−22 ?

33

Coin?

Nave Nave

K 5 layer of collapse K 1 layer of collapse/K 4 wooden floor BM KrH YIIe Nave K 7 lime-mortar floor/K 12 planning layer BM KrH YIIIe Chancel K 47 bedding sand BM KrH YV Nave K 4 wooden floor Galster 42 Chancel K 53 floor layer Unidentified Chancel K 53 floor layer bracteate Unidentified Nave K 1 layer of collapse/K 7 planning layer

References/abbreviations: BM = Malmer, B, Den senmedeltida penningen i Sverige, Stockholm 1980; Galster = Galster, G, Unionstidens udmöntninger, Köpenhamn 1972; Haljak = Haljak, G, Livonian Coins XIII–XVIII century, Tallinn 2010; Hävernick = Hävernick, W, Die Münzen von Köln, Köln 1936; Jesse = Jesse, W, Der wendische Münzverein, Braunschweig 1967; LL = Lagerqvist, L-O, Svenska mynt under vikingatid och medeltid samt gotländska mynt, Stockholm 1970; Myrberg = Myrberg, N, Ett eget värde, Stockholm 2008; NNÅ = Nordisk numismatisk årsskrift 1991 (1992); Saether = Saether, O, Mecklenburgske oksehodebrakteater i norske kirkefunn, Nordisk numismatisk unions medlemsblad 1965; Schive = Schive, C.I., Norges mynter i middelalderen, Christiania 1865; Skaare = Skaare, K, Norges mynthistorie, Oslo 1995; Waschinski = Waschinski, E, Brakteaten und Denare des Deutschen Orden, Frankfurt am Main 1934.​

Coin finds, coin use – a medieval round church  349 The excavations of the monastic area of Vadstena revealed traces of a different type of coin circulation with a larger proportion of high-value coins. A total of 327 coins, of which 200 were medieval, were retrieved from excavations conducted between 1957 and 1969. Apart from one Anglo-Saxon coin minted in 978−1016 and one from Götaland dating from 1250 to 1275, the coins are all late medieval. Unsurprisingly, there was an emphasis on the fifteenth century, which was the high point of the monastery. The significant difference in the composition of the coin assemblages of the fifteenth century is that the one in Vadstena contains, apart from the expected Swedish, Danish and German bracteates, larger denominations such as örtugar from Gotland, and a vast amount of Danish hvider and sterlinger. Such larger coins were, with a few exceptions, not found at Klåstad. This probably reflects a medieval reality where the coin use in the monasteries of the aristocrats differed from that of peasants in their parish churches, although the source-critical aspects related to the excavation techniques must be borne in mind. 25

Conclusion The archaeological excavations of the medieval remains underneath the clearance cairn at Klosterstad gave unexpected and rich results, especially regarding early medieval church construction and organisation. These investigations were also interesting from a numismatical and church historical perspective, since the numerous and well-documented coin finds provide good foundations for further interpretations of the coin use in a small country church in one of the heartlands of the kingdom during the Catholic era. The coins found at Klåstad fit in well with my previous research regarding medieval coin use, published in my doctoral thesis Moneta nostra: monetarisering i medeltidens Sverige (Moneta nostra: Monetisation in Medieval Sweden), and in addition providing greater detail. Klåstad provides an opportunity to study the entire medieval period, from the eleventh-century wooden church to the sixteenth century, when the twelfth-century round church was demolished by fire. The general picture of how Östergötland was monetised is not changed by the Klåstad excavations, but it is strengthened and clarified. Coin use in churches began already in the second half of the twelfth century, and volumes did not significantly increase until after AD 1250, with continued increases in the ensuing century. The domestic coins dominated until the 1400s, when Danish and German coins took over as a result of the Kalmar Union. The only significant deviation from this pattern is the low number of Gotlandic coins from the twelfth and thirteenth centuries, as these were frequently occurring elsewhere in the diocese of Linköping during this period.26

Notes 1 Anders Kaliff, “Skepnader i historiens gränsland: identitet, centralitet och externa influenser i västra Östergötland under järnåldern,” in Skuggor i ett landskap: västra Östergötlands slättbygd under järnålder och medeltid: resultat av ett tvärvetenskapligt projekt, edited by Anders Kaliff (Stockholm: Riksantikvarieämbetet, 2009), 12–45. 2 Rikard Hedvall, “Kyrkorna i Klåstad,” in Människors rum och människors möten: kulturhistoriska skisser: Berit Wallenbergs Stiftelse 50 år: vetenskapligt symposium på Nationalmuseum 14 november 2005, edited by Anders Perlinge and Gunnar Wetterberg (Stockholm: Berit Wallenbergs stiftelse, 2007), 146–47; Carl Silfverstolpe, Vadstena klosters uppbörds- och utgiftsbok 1539–1570 (Stockholm: Ivar Hæggström, 1895).

350 Henrik Klackenberg 3 The excavation is presented in Hedvall, “Kyrkorna,” 143–64. The archaeological documentation (Hedvall’s manuscript) is a work in progress, which he has graciously made available to me. All details about the construction of the church, stratigraphy and contexts are derived from this material, and no further specific references to this will be added. 4 Daniel Rantzau, Daniel Rantzau dagbok, edited by Gunnar Lindqvist (Linköping: Östergötlands länsmuseum, 1987), 23: 2. 5 Landskapshandlingar: Östergötlands handlingar, 1545:19 and 1548:4, Riksarkivet (National Archives), Stockholm. Lars-Arne Norborg, Storföretaget Vadstena kloster (Lund: Gleerup 1958), 39, 253. 6 The coins are presented in Table 12.1 and Figure 12.11. They have been classified by the author with the aid of Kenneth Jonsson, Monica Golabiewski Lannby and Frédéric Elfver. At present, the coins are in the care of the Royal Coin Cabinet, Stockholm, diary number KMK 711-609-2005. 7 Hedvall, “Kyrkorna,” 157–58; Jes Wienberg, “Iøjnefaldende arkitektur. Nordens middelalderlige rundkirker. Conspicuous architecture. Medieval round churches in Scandinavia,” KUML (2014): 205–44. 8 This problem is adressed by Rensbro, this volume, ch. 3. 9 Coin no. 358. 10 Coin no. 7, 43, 259, 308, 21, 26, 31, 39, 257, 305. 11 Coin no. 8, 9, 20, 22, 24, 34, 41, 45, 262, 268, 273, 278–79, 281, 297. 12 Coin no. 1, 13, 14, 16, 25, 28, 29, 48, 256, 258, 261, 266, 269–72, 274, 280, 283, 285, 286, 290, 299, 301, 306, 307. 13 Coin no. 4, 12, 18, 27, 32, 37, 38, 40, 42, 260, 263, 267, 287, 291, 293–96, 298, 309. 14 Coin no. 2, 3, 6, 11, 15, 17, 19, 23, 30, 35, 36, 44, 46, 47, 264, 265, 282, 284, 289, 292, 300. 15 Henrik Klackenberg, Moneta nostra: monetarisering i medeltidens Sverige (Stockholm: Almqvist & Wiksell International, 1992), 34−38. 16 Anna Nilsén, Focal Point of the Sacred Space: The Boundary between Chancel and Nave in Swedish Rural Churches: from Romanesque to Neo-Gothic (Uppsala: Uppsala University, 2003), 263; Olav Tveito. “Mynter i messen—Kirkefunnene som bidrag til kunnskap om offerpraksis og kirkeskikker (11.–17. årh.),” Historisk tidsskrift 3 (2015): 396. On this topic, see also contributions by Risvaag and Roland, this volume, chs. 6 and 7. 17 Carl-Martin Edsman, “Offer,” Mottaker-Orlogsskib, KHLNM 12 (1967), col. 514–25; Tveito, “Mynter i messen”; see also Gullbekk et al. and Wangsgaard Jürgensen, this volume, chs. 1 and 2. 18 Bengt Stolt, “Kyrkorum och kyrkoskrud,” in Mässa i medeltida socken: En studiebok, edited by Sven Helander et al. (Skellefteå: Artos, 1993), 161. For a survey of medieval collection plates from Gotland, see Pia Bengtsson Melin and Kenneth Jonsson, “Medeltida kollekttavlor från Gotland,” Myntstudier 2 (2019): 1–92. In the written sources, these offering plates are addressed in Latin as «tabulae», see Kilger, ch. 5 this volume. Late medieval accounts from Kumla Church in the Swedish province of Närke make a distinction between coins offered on collecting plates, addressed as «taffla paennigha», and coins given to offering trunks as «stokka paenningha» “Medeltida kollekttavlor,” 6. 19 Herman Schück, Ecclesia lincopensis: studier om Linköpingskyrkan under medeltiden och Gustav Vasa (Stockholm: Almqvist & Wiksell, 1959), 199; Mats Anglert, “Kristianisering, kyrkor och centralitet i Östergötland,” in Skuggor i ett landskap: Västra Östergötlands slättbygd under järnålder och medeltid; resultat av ett tvärvetenskapligt projekt, edited by Anders Kaliff (Stockholm: Riksantikvarieämbetet, 2009), 231. 20 Klackenberg, Moneta nostra, finds 53 and 58. 21 Sveriges mynthistoria Landskapsinventeringen vol. 1, Östergötland, finds 86 and 134. 22 On the use of coinage in monastic contexts in medieval Norway, see Hommedal, this volume, ch. 8. 23 Hedvall, “Ett kyrkligt centrum,” 65−75. 24 On interpretation of late Viking Age coins in church and early Christian burial contexts, see Gullbekk et al. and Risvaag, this volume, chs. 1 and 6. 25 Klackenberg, Moneta nostra, 92; Elisabet Regner, Den reformerade världen: Monastisk och materiell kultur i Alvastra kloster från medeltid till modern tid (Stockholm: Stockholm University, 2005), 122. 26 Klackenberg, Moneta nostra, 89–90.

13 Jomala church, Åland Islands Coin offerings to the Virgin Mary and the long Reformation Eeva Jonsson1

Figure 13.1  T  he church of Jomala. Exterior from the north. Photo: MrFinland, W ­ ikiCommons (CC-BY-SA)

Jomala church revisited The rural church of Jomala is located in the middle of the Åland Islands, c. 10 km north of the capital of Mariehamn. It is the earliest stone church in modern-day Finland, dating to the thirteenth century, and the only one in Finland built in the ­Romanesque style. An archaeological excavation in the medieval nave and the choir was carried out in 1961. It yielded a total of 594 coins of which at least 134 are medieval. Thanks to the large number of medieval coins and the relatively high quality of the excavation documentation, the Jomala church provides particularly useful insights into the different contexts for coin offerings made during the medieval period.2 The aim of the present paper is to analyse the coin finds, both medieval and post-Reformation, by applying modern digital techniques. It will explore the church space from a liturgical perspective and makes extensive use of both coin finds and related historical sources. Particular attention is given to devotional practice, folk religion and long-term change. The religious practice and its continuity, as it is reflected

352  Eeva Jonsson in coin use, will be followed over five centuries from the thirteenth century to the eighteenth century, with a focus on the cult of the Virgin Mary. In the present chapter, the Reformation and the process of change in the religious life will be emphasised. The coin finds from the church of Jomala reflect these processes step by step. Although the church edict in 1540 demanded a thorough purification of both church doctrine and ritual, the coin material reveals the existence of a continuing Catholic practice. The efforts of the Lutheran Church to regulate the religious life and private worship carried out in Finnish churches are also visible in the written documents, some of which are presented below.3 In addition to the coin offerings and non-monetary use of coins, a possible foundation treasure in the chancel will be discussed. The connection between coin finds and liturgy was regarded as uncertain and open to debate for many years. The 1992 PhD thesis Moneta nostra: Monetarisering i medeltidens Sverige [Moneta nostra. Monetisation in medieval Sweden] by Henrik Klackenberg marked a turning point in scholarly approaches to the subject because of the holistic approach of the study. In order to analyse why medieval coins are found in the soil of a church, and how representative they are for coin circulation in general, Klackenberg documented the distribution of the coins.4 He was able to demonstrate that the majority of the medieval coins found beneath the church floors were intended as offerings by the altars, in the offertory chests and so forth, but were lost in the process of offering. Recent surveys of the medieval finds of coins in Swedish and Finnish churches are largely based on the results published by Klackenberg.5 The post-Reformation coin finds from churches have recently become the subject of intensified scholarly interest in Sweden and in Finland.6 These overviews are of particular interest when tracing fluctuations in the practices of making offerings during different periods. So far, the post-Reformation coins have been treated as a homogenous group, representing grave goods and accidental losses, as well as possible offerings. It is suggested here, however, that the distribution of the post-Reformation coin finds in the Jomala church has a discernible connection to the celebrat