'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps 9781841712550, 9781407353180
A much-discussed topic in the studies of Alpine region lake-dwellings is their chronology and continuity. This study inc
148 76 33MB
English Pages [262] Year 2001
Polecaj historie
![The Tombs of the Middle Bronze Age II Period From the ‘500’ Cemetery at Tell Fara (South) (UCL Institute of Archaeology Publications) [1 ed.]
0905853059, 9780905853055](https://dokumen.pub/img/200x200/the-tombs-of-the-middle-bronze-age-ii-period-from-the-500-cemetery-at-tell-fara-south-ucl-institute-of-archaeology-publications-1nbsped-0905853059-9780905853055.jpg)
Table of contents :
Blank Page
Front Cover
Title Page
Copyright
Dedication
TABLE OF CONTENTS
List of figures
List of tables
Abstract
Preface
Acknowledgements
Abbreviations
Chapter 1. THE PAST AND PRESENT ALPINE REGION: A GENERAL OVERVIEW
Chapter 2. LAKE-DWELLINGS IN THE ALPS: DISCOVERIES AND SCIENTIFIC RESEARCH
Chapter 3. THE LAKE-DWELLING "PROBLEM"
Chapter 4. DATING THE ALPINE LAKE-DWELLINGS
Chapter 5. THE BRONZE AGE IN THE ALPS: CHRONOLOGY AND ARCHAEOLOGICAL PERIODISATION
Chapter 6. LAKE-DWELLINGS AND THE ARCHAEOLOGICAL RECORD: AVAILABILITY AND STATISTICAL ANALYSES
Chapter 7. THE BRONZE AGE LAKE-DWELLING PHENOMENON IN EACH OF THE SIX ALPINE COUNTRIES
Chapter 8. ON THE BRINK OF 15TH CENTURY BC: THE LAST EARLY BRONZE AGE LAKE-SETTLEMENTS
Chapter 9. THE MIDDLE BRONZE AGE LAKE-DWELLING HIATUS IN THE NORTHERN PARTS OF THE ALPS: AN ENVIRONMENTAL APPROACH
Chapter 10. LAKE-DWELLERS' COGNITIVE RESPONSE TO ENVIRONMENTAL CHANGE
Chapter 11. MIDDLE BRONZE AGE LAND SETTLEMENTS AROUND LAKE CONSTANCE AND LAKE ZURICH WITH POSSIBLE ORIGINS IN THE LACUSTRINE TRADITION
Chapter 12. CONCLUSIONS
REFERENCES
APPENDICES
Citation preview
BAR S968 2001 MENOTTI ‘THE MISSING PERIOD’: MIDDLE BRONZE AGE LAKE-DWELLINGS IN THE ALPS
‘The Missing Period’: Middle Bronze Age Lake-Dwellings in the Alps Francesco Menotti
BAR International Series 968 9 781841 712550
B A R
2001
'The Missing Period': Middle Bronze Age Lake- Dwellings in the Alps Francesco Menotti
N
LN
EBA
MBA
LBA
E!
BAR International Series 968 2001
Published in 2016 by BAR Publishing, Oxford BAR International Series 968 'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
© F Menotti and the Publisher 2001 COVER IMAGE Pfahlbau Museum, Unteruhldingen, D (Picture F. Menotti 2000)
The author's moral rights under the 1988 UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher.
ISBN 9781841712550 paperback ISBN 9781407353180 e-format DOI https://doi.org/10.30861/9781841712550 A catalogue record for this book is available from the British Library BAR Publishing is the trading name of British Archaeological Reports (Oxford) Ltd. British Archaeological Reports was first incorporated in 1974 to publish the BAR Series, International and British. In 1992 Hadrian Books Ltd became part of the BAR group. This volume was originally published by Archaeopress in conjunction with British Archaeological Reports (Oxford) Ltd/ Hadrian Books Ltd, the Series principal publisher, in 2001. This present volume is published by BAR Publishing, 2016.
BAR
PUBLISHING BAR titles are available from:
EMAIL PHONE
FAX
BAR Publishing 122 Banbury Rd, Oxford, OX2 7BP, UK [email protected] +44 (0)1865 310431 +44 (0)1865 316916 www.barpublishing.com
To my teachers John Campbell and Andrew Sherratt
Contents Page List of figures List of tables Abstract Preface Acknowledgements Abbreviations
vii xiv xv xvi xvii xviii
Chapter 1 THE PAST AND PRESENT ALPINE REGION: A GENERAL OVERVIEW 1.1
Introduction
1
1.2
The Alps: location and morphological structure
1
1.3
The hydrology of the Alps: rivers and lakes
1
1.4
The Alpine climate from the last glaciation to the present
3
1.5
Flora and fauna of the Alpine valleys in the past 11,000 years
4
1.6
Human occupation and mobility within the prehistoric Alpine region
6
1.7
Cultural transformations of the Alpine environment: subsistence and economy from the Neolithic to the Bronze Age
8
Chapter 2 LAKE-DWELLINGS IN THE ALPS: DISCOVERIES AND SCIENTIFIC RESEARCH 2.1
Introduction
11
2.2
The 19th century: the beginning of the lake-dwelling phenomenon
11
2.3
Lake-dwelling studies in the first half of the 20 th century
13
2.4
Second half of the 20 th century: two decades of lake-dwelling research 1950-1970
14
2.4.1
Lake-dwellingsvs non-lake-settlements:new research development
14
2.5
Scientific research on lake-dwellings in the 1970s and 1980s
15
2.6
Recent developments in lake-settlement research
16
Chapter 3 THE LAKE-DWELLING "PROBLEM" 3.1
Introduction
18
3.2
The beginning of the lake-dwelling phenomenon in the Alps
18
3.3
The Ober Meilen discovery
18
3.4
Ferdinand Keller's dogma
18
iii
3.5
The need for a typological differentiation
19
3.6
From lake-dwellings to lakeside-dwellings
20
3.7
The Pf ahlbauproblem is no longer a problem
20
Chapter 4 DATING THE ALPINE LAKE-DWELLINGS 4.1
Introduction
4.2
Relative dating: stratigraphy
4.3
The
4.4
Dendrochronology:
4.5
Thermoluminescence
4.6
Additional chronometric analyses as a help for chronology
29
4.7
Correlations between the various dating techniques
31
14
22 and pottery typology
22
24
C dating method
26
a revolution in precision vs pottery typology
28
Chapter 5 THE BRONZE AGE IN THE ALPS: CHRONOLOGY AND ARCHAEOLOGICAL PERIODISATION 5.1
Introduction
33
5.2
A chronological view of the Bronze Age in central Europe
34
5.3
Bronze Age in the Alps: absolute dating vs typology
35
5.4
The Bronze Age as an archaeological periodisation within the Alpine region
35
Chapter 6 LAKE-DWELLINGS AND THE ARCHAEOLOGICAL RECORD: AVAILABILITY AND STATISTICAL ANALYSES 6.1
Introduction
42
6.2
Sites state of preservation and discoveries: two important factors in the lake-dwelling chronological studies
42
Comparative analyses of the quantitative and qualitative distribution of the lacustrine sites within the whole Alpine region
45
Comparative analyses of the quantitative and qualitative Bronze Age lake-dwellings distribution on Lake Constance and Lake Zurich
47
6.3 6.4
Chapter 7 THE BRONZE AGE LAKE-DWELLING PHENOMENON IN EACH OF THE SIX ALPINE COUNTRIES 7.1
Introduction
72
7.2
Lake-dwellings in Switzerland
72
7.2.1
Archaeological research: policies and methods
72 IV
7.2.2
Swiss Bronze Age lacustrine settlements
73
7.3
The Pfahlbauten in Germany
75
7.3.1
Lake-dwelling archaeological research
75
7.3.2
South German Bronze Age lake-settlements
76
7.4
The Seeufersiedlungen in Austria
77
7.4.1
Austrian lacustrine settlements: history of research
77
7.4.2
Bronze Age lake-dwellings in Austria
77
7.5
The stations lacustres in France
77
7.5.1
The French lake-dwelling history of archaeological research
78
7.5.2
The Bronze Age lake-settlements of France
78
7.6
The palajitte lacustri in northern Italy
79
7.6.1
Italian lake-dwellings and the past archaeological research
79
7.6.2
The Bronze Age north Italian villaggi lacustri
79
7.7
The kolisca in Slovenia
81
7.7.1
Lake-dwelling archaeological research
81
7.7.2
Slovenian Bronze Age lake-settlements
81
Chapter 8 ON THE BRINK OF 15THCENTURY BC: THE LAST EARLY BRONZE AGE LAKE-SETTLEMENTS 8.1
Introduction
95
8.2
1500-12th century BC: more than three centuries oflake-villages absence in the northern parts of the Alps
95
8.3
Lacustrine sites immediately preceding the "missing period"
97
8.3.1
ZH-Mozartstrasse (Lake Zurich, CH)
97
8.3.2
Arbon-Bleiche 2 (Lake Constance, CH)
102
8.3.3
Bodman-Schachen 1 (Lake Constance, D)
105
8.4
Siedlung-Forschner
and Fiave: two middle Bronze Age "non"-lake-dwelling
sites
110
Chapter 9 THE MIDDLE BRONZE AGE LAKE-DWELLING HIATUS IN THE NORTHERN PARTS OF THE ALPS: AN ENVIRONMENTAL APPROACH 9.1
Introduction
115
9.2
Lack of continuity: the Bronze Age lake-dwelling occupational pattern in the Alps
115
9.3
Climatic changes in the northern Alpine region during the Bronze Age
117
V
9.4
Climatic and hydrologic conditions of Lake Constance and Lake Zurich: past and present
120
9.4.1
Short- and long-term lake level variations
121
9.5
GIS approach to lake levels fluctuations at ZH-Mozartstrasse, Arbon-Bleiche 2 and Bodman-Schachen 1: introduction to computer simulations
123
9.5.1
ZH-Mozartstrasse
126
9.5.2
Arbon-Bleiche 2
127
9.5.3
Bodman-Schachen 1
128
9.6
The Middle Bronze Age lake-dwelling hiatus: analysis of the environmental approach
129
Chapter 10 LAKE-DWELLERS' COGNITIVE RESPONSE TO ENVIRONMENTAL CHANGE 10.1
Introduction
141
10.2
Loss of tillable land as a possible cause of a socio-economic crisis
141
10.3
Trade and interaction as means of cultural diffusion: evidence from pottery typological analyses
143
Lake-settlements abandonment as a cultural choice: the late EBA exodus from the northern Alpine lakes
145
From environmental to cultural change
147
10.4
10.5
Chapter 11 MIDDLE BRONZE AGE LAND SETTLEMENTS AROUND LAKE CONSTANCE AND LAKE ZURICH WITH POSSIBLE ORIGINS IN THE LACUSTRINE TRADITION 11.1
Introduction
150
11.2
From "wet" to "dry" life: adaptation to a new environment
150
11.3
MBA land settlements around Lake Constance
152
11.4
MBA land sites in the surroundings of Lake Zurich
156
11.5
MBA hill sites on Lake Zug
159
Chapter 12 CONCLUSIONS 12.1
Middle Bronze Age lake-dwelling "missing period" in the northern Alpine region
162
References
165
Appendix: Database of the Alpine region lake-dwellings
185
Vl
List of figures Chapter 1 The expansion of the occurrence of the "lake-dwelling wheat" (Triticum aestivum) as an indication of the spread of the lake-dwellings in the Alps
7
1.7.1
Attempted reconstruction of the Zurich bay environment in the Early Neolithic
9
1.7.2
Attempted reconstruction
1.6.1
of the Zurich bay environment in the Late Neolithic
10
Chapter 2 2.2.1
Portrait of Ferdinand Keller
11
2.5.1
The first lake-dwelling underwater survey on Lake Constance in 1954
16
2.5.2
Underwater excavation at ZR-Kleiner Hafner on Lake Zurich in 1981
16
2.6.1
Marlot underwater expedition in 1854
16
Chapter 3 3.4.1
Keller's image of the pile-dwellings
19
3.5.1
Keller's theory about the lake-dwellings
19
3.5.2
Reinerth's theory about the lake-dwellings
19
3.6.1
Paret's theory about the lake-dwellings
20
3.7.1
The Pf ahlbauproblem as it is seen today
21
Chapter 4 4.2.1
4.2.2
Number of Lake-dwelling sites in the Alps dated and analysed using the four principal scientific techniques
23
Percentages of Lake-dwelling sites in the Alps dated using the three principal dating methods
24
4.3.1
Percentages of Lake-dwelling sites in the Alps dated using
4.4.1
The principle of chronology-building
4.4.2
Percentages of Lake-dwelling sites in the Alps dated using dendrochronology
28
4.5.1
Good and bad positions for pottery in an archaeological deposit in relation to TL sampling
29
Comparison of occupational patterns, lake level fluctuations and Alpine glacier movements (retreats and advances)
31
4.6.1
14 C
method
by tree-ring dating
25 26
Chapter 5 5.1.1
The "Danubian" and "Trans-Alpine"
Bronze Age axial routes
5.4.1
The chronology of Bronze Age lake-dwellings in the Alps
5.4.2
The chronology of Bronze Age lake-dwellings in north-eastern vii
33 38 Switzerland
38
5.4.3
The chronology of Bronze Age lake-dwellings in north-western
5.4.4
The chronology of Bronze Age lake-dwellings in southern Germany
39
5.4.5
The chronology of Bronze Age lake-dwellings in Austria
40
5.4.6
The chronology of Bronze Age lake-dwellings in eastern France
40
5.4.7
The chronology of Bronze Age lake-dwellings in northern Italy
41
5.4.8
The chronology of Bronze Age lake-dwellings in Slovenia
41
Switzerland
39
Chapter 6 6.3.1
Number of Lake-dwelling sites in Switzerland by period of occupation
49
6.3.2
Percentage of Lake-dwelling sites in Switzerland by period of occupation
49
6.3.3
Number of Lake-dwelling sites in Switzerland dated and analysed using the four principal scientific techniques
50
6.3.4
Number of Bronze Age Lake-dwelling sites in Switzerland by period of occupation
50
6.3.5
Percentage of Bronze Age Lake-dwelling sites in Switzerland by period of occupation
51
6.3.6
Number of Bronze Age Lake-dwelling sites in Switzerland dated and analysed using the four principal scientific techniques
51
6.3.7
Number of Lake-dwelling sites in Germany by period of occupation
52
6.3.8
Percentage of Lake-dwelling sites in Germany by period of occupation
52
6.3.9
Number of Bronze Age Lake-dwelling sites in Germany by period of occupation
53
6.3.10
Percentage of Bronze Age Lake-dwelling sites in Germany by period of occupation
53
6.3.11
Number of Lake-dwelling sites in Germany dated and analysed using the four principal scientific techniques
54
Number of Bronze Age Lake-dwelling sites in Germany dated and analysed using the four principal scientific techniques
54
6.3.13
Number of Lake-dwelling sites in Austria by period of occupation
55
6.3.14
Percentage of Lake-dwelling sites in Austria by period of occupation
55
6.3.15
Number of Lake-dwelling sites in Austria dated and analysed using the four principal scientific techniques
56
6.3.16
Number of Bronze Age Lake-dwelling sites in Austria by period of occupation
56
6.3.17
Percentage of Bronze Age Lake-dwelling sites in Austria by period of occupation
57
6.3.18
Number of Bronze Age Lake-dwelling sites in Austria dated and analysed using the four principal scientific techniques
57
6.3.19
Number of Lake-dwelling sites in France by period of occupation
58
6.3.20
Percentage of Lake-dwelling sites in France by period of occupation
58
6.3.21
Number of Lake-dwelling sites in France dated and analysed using
6.3.12
vm
the four principal scientific techniques
59
6.3.22
Number of Bronze Age Lake-dwelling sites in France by period of occupation
59
6.3.23
Percentage of Bronze Age Lake-dwelling sites in France by period of occupation
60
6.3.24
Number of Bronze Age Lake-dwelling sites in France dated and analysed using the four principal scientific techniques
60
6.3.25
Number of Lake-dwelling sites in Italy by period of occupation
61
6.3.26
Percentage of Lake-dwelling sites in Italy by period of occupation
61
6.3.27
Number of Lake-dwelling sites in Italy dated and analysed using the four principal scientific techniques
62
6.3.28
Number of Bronze Age Lake-dwelling sites in Italy by period of occupation
62
6.3.29
Percentage of Bronze Age Lake-dwelling sites in Italy by period of occupation
63
6.3.30
Number of Bronze Age Lake-dwelling sites in Italy dated and analysed using the four principal scientific techniques
63
6.3.31
Number of Lake-dwelling sites in Slovenia by period of occupation
64
6.3.32
Percentage of Lake-dwelling sites in Slovenia by period of occupation
64
6.3.33
Number of Lake-dwelling sites in Slovenia dated and analysed using the four principal scientific techniques
65
6.3.34
Number of Bronze Age Lake-dwelling sites in Slovenia by period of occupation
65
6.3.35
Percentage of Bronze Age Lake-dwelling sites in Slovenia by period of occupation
66
6.3.36
Number of Bronze Age Lake-dwelling sites in Slovenia dated and analysed using the four principal scientific techniques
66
Percentage of Lake-dwelling sites on Lake Constance and Lake Zurich by period of occupation
67
Number of Lake-dwelling sites on Lake Constance and Lake Zurich by period of occupation
68
Number of Lake-dwelling sites on Lake Constance and Lake Zurich dated and analysed using the four principal scientific techniques
69
Number of Bronze Age Lake-dwelling sites on Lake Constance and Lake Zurich by period of occupation
69
Percentage of Bronze Age Lake-dwelling sites on Lake Constance and Lake Zurich by period of occupation
70
Number of Bronze Age Lake-dwelling sites on Lake Zurich dated and analysed using the four principal scientific techniques
70
Number of Bronze Age Lake-dwelling sites on Lake Constance dated and analysed using the four principal scientific techniques
71
Number of Bronze Age Lake-dwelling sites on Lake Constance and Lake Zurich dated and analysed using the four principal scientific techniques
71
6.4.1
6.4.2
6.4.3
6.4.4
6.4.5
6.4.6
6.4.7
6.4.8
ix
Chapter 7 7.2.2.1
Bronze Age lake-dwellings on Lake Zurich (CH)
83
6.2.2.2
Bronze Age lake-dwellings on Lake Greifen (CH)
83
7.2.2.3
Bronze Age lake-dwellings on Lake Pfiiffikon (CH)
84
7.2.2.4
Bronze Age lake-dwellings on Lake Zug (CH)
84
7.2.2.5
Bronze Age lake-dwellings on Lake Constance (CH)
85
7.2.2.6
Bronze Age lake-dwellings on Lake Neuchatel (CH)
85
7.2.2.7
Bronze Age lake-dwellings on Lake Bienne (CH)
86
7.2.2.8
Bronze Age lake-dwellings on Lake Morat (CH)
86
7.2.2.9
Bronze Age lake-dwellings on Lake Geneva (CH)
87
7.3.2.1
Bronze Age lake-dwellings on Lake Constance (D)
87
7.3.2.2
Bronze Age lake-dwellings on Lake Feder (D)
88
7.3.2.3
Bronze Age lake-dwellings on Lake Starnberg (D)
88
7.4.2.1
Bronze Age lake-dwellings on Lake Att (A)
89
7.4.2.2
Bronze Age lake-dwellings on Lake Traun (A)
89
7.5.2.1
Bronze Age lake-dwellings on Lake Chalain (F)
90
7.5.2.2
Bronze Age lake-dwellings on Lake Clairvaux (F)
90
7.5.2.3
Bronze Age lake-dwellings on Lake Annecy (F)
91
7.5.2.4
Bronze Age lake-dwellings on Lake Bourget (F)
91
7.5.2.5
Bronze Age lake-dwellings on Lake Geneva (F)
92
7.6.2.1
Bronze Age lake-dwellings on Lake Varese (I)
92
7.6.2.2
Bronze Age lake-dwellings on Lake Ledro and Fiave (I)
93
7.6.2.3
Bronze Age lake-dwellings on and around Lake Garda (I)
93
7.7.2.1
Bronze Age lake-dwellings in the Ljubljana marsh (SL)
94
Chapter 8 8.3.1.1
Map of the eastern side of Zurich bay
97
8.3.1.2
Mozartstrasse site plan: (a) Mozart Street at the turn of the century (19th -20th ); (b) Mozart Street in the 1920s; (c) The Zurich Opera House and the Bernhard Theatre as they were joined together in 1987
98
8.3.1.3
The new building which joints the Zurich Opera House and the Bernhard Theatre
98
8.3.1.4
(a) Plan of the EBA settlement of ZH-Mozartstrasse; (b) Reconstruction of the EBA site of ZH-Mozartstrasse and its surroundings as they might have looked at the time
99
8.3.1.5
Plan of the two EBA villages. (a) Houses of group A (older) X
{b) Houses of group B (younger)
99
8.3.1.6
An attempted reconstruction
8.3.1.7
Ceramic assemblages found in the ZH-Mozarstrasse
8.3.1.8
Bronze objects found in the ZH-Mozartstrasse
8.3.2.1
The Arbon bay on the Swiss shore of Lake Constance
102
8.3.2.2
Plan of the Arbon-Bleiche 2 excavated area with dates of excavation
102
8.3.2.3
Schematic plan of the Arbon-Bleiche 2 1945 excavated area
103
8.3.2.4
Plan of the entire Arbon-Bleiche lake-dwellings area showing the Arbon-Bleiche 2 1990 and 1991 excavated sectors
103
8.3.2.5
Range of forms of different classes of pottery at Arbon-Bleiche 2
104
8.3.2.6
Some of the bronze objects found at the EBA site of Arbon-Bleiche 2 in 1945
104
8.3.2.7
Plan of the houses at Arbon-Bleiche 2: (a) Houses with reconstructable dimensions; (b) Houses difficult to define
104
Pfahlschuh technique of house-construction Attempted reconstruction of a house
105
8.3.2.8
at EBA houses at ZH-Mozartstrasse EBA layers
101
EBA stratum
102
at Arbon-Bleiche 2:
8.3.3.1
The occupational pattern of Bodman-Schachen
1
106
8.3.3.2
The location of Bodman-Schachen
8.3.3.3
Plan of the six houses in construction phase 1, layer A, 19th century BC
107
8.3.3.4
Plan of the three houses in construction phase 2, layer B, 17th century BC
107
8.3.3.5
Plan of the four houses in construction phase 3, layer B, 17th century BC
108
8.3.3.6
Plan of the four houses in construction phase 4, layer C, 17th century BC
108
8.3.3.7
Plan of the six houses in construction phase 5, no layer, 16th century BC
108
8.3.3.8
The 19th century pile-stabilising construction technique at Bodman-Schachen
8.3.3.9
Patterns of pottery decoration found in layer Cat Bodman-Schachen
8.3.3.10
The four different types of pottery found in layers A, B, C and on the surface of Bodman-Schachen 1 EBA site
109
8.3.3.11
Bronze objects found at Bodman-Schachen
110
8.4.1
Air-photograph
8.4.2
Neolithic and Bronze Age sites around the Lake Feder peat deposit
111
8.4.3
Photograph of the former Lake Carera basin
112
8.4.4
Air-photograph
112
8.4.5
Photograph and map of the "Dos Gustinaci" hill, the location ofFiave zone 3
113
8.4.6
Fiave: plan of zone 1 and its sectors
113
8.4.7
Fiave: section of the retaining walls constructed around the island (zone 1)
1 lake-dwelling in the Bodman bay
106
1
1 [layer C]
of Lake Feder
1
108 109
110
and map of the Fiave peat deposit
xi
8.4.8
in order to gain building space
113
(a) Layout of a house in sector IV, zone 1, Fiave 7 {b) Layout of a house in sector III, zone 3, Fiave 7
114
Chapter 9 9.3.1
Correlation between French Jura lake level fluctuations with occupational evidence and atmospheric 14C content of the past 10,000 years
119
Correlation between occurrence of lake-dwellings in the northern Alpine region from the Neolithic to the Iron Age and the atmospheric 14C content of the same period
119
9.4.1.1
Two different levels of transgression according to the sensitivity of the lakes
122
9.4.1.2
Seasonal fluctuations of Lake Constance at Arbon bay
131
9.4.1.3
Seasonal fluctuations of Lake Constance at Bodman-Schachen
131
9.4.1.4
Seasonal fluctuations of Lake Zurich at Zurich bay
131
9.4.1.5
Stratigraphy of the EBA "inland" site of Tagerwiler-Hochstrasse near the city of Constance
123
Locations of the Zurich bay (Lake Zurich, CH), Arbon bay and Bodman bay (Lake Constance, CH, D)
125
GIS computer simulation of the inferred EBA lake-level in the Zurich bay (402 m a.s.l.)
132
GIS computer simulation of the inferred MBA lake-level in the Zurich bay (403 m a.s.l.)
132
GIS computer simulation of the inferred MBA lake-level in the Zurich bay (404 m a.s.l.)
133
GIS computer simulation of the inferred MBA lake-level in the Zurich bay (405 m a.s.l.)
133
GIS computer simulation of the inferred MBA lake-level in the Zurich bay (406 m a.s.l.)
134
GIS computer simulation of the inferred MBA lake-level in the Zurich bay (407 m a.s.l.)
134
GIS computer simulation of the inferred EBA lake-level in the Arbon bay (392 m a.s.1.)
135
GIS computer simulation of the inferred MBA lake-level in the Arbon bay (394 m a.s.1.)
135
GIS computer simulation of the inferred MBA lake-level in the Arbon bay (395 m a.s.1.)
136
GIS computer simulation of the inferred MBA lake-level in the Arbon bay (396 m a.s.1.)
136
GIS computer simulation of the inferred MBA lake-level in the Arbon bay (398 m a.s.1.)
137
GIS computer simulation of the inferred MBA lake-level in the Arbon bay (400 m a.s.1.)
137
9.3.2
9.5.1
9.5.1.1
9.5.1.2
9.5.1.3
9.5.1.4
9.5.1.5
9.5.1.6
9.5.2.1
9.5.2.2
9.5.2.3
9.5.2.4
9.5.2.5
9.5.2.6
Xll
9.5.3.1
9.5.3.2
9.5.3.3
9.5.3.4
9.5.3.5
9.5.3.6
GIS computer simulation of the inferred EBA lake-level in the Bodman bay (392 m a.s.1.)
138
GIS computer simulation of the inferred MBA lake-level in the Bodman bay (394 m a.s.1.)
138
GIS computer simulation of the inferred MBA lake-level in the Bodman bay (396 m a.s.1.)
139
GIS computer simulation of the inferred MBA lake-level in the Bodman bay (397 m a.s.1.)
139
GIS computer simulation of the inferred MBA lake-level in the Bodman bay (398 m a.s.1.)
140
GIS computer simulation of the inferred MBA lake-level in the Bodman bay (400 m a.s.l.)
140
Chapter 10 10.3.1
10.3.2
10.3.3
10.4.1
Bodman-Schachen 1: (a) Arbon culture pottery found in stratum B (b) Arbon culture pottery found in stratum C
144
Pottery of the EBA Arbon type indicative of contacts along the 16th century Trans-Alpine route
144
Embossed ridges and "zigzag" engravings: two main patterns of the EBA lacustrine pottery in the northern Alpine region (a) ZH-Mozartstrasse; (b) Arbon-Bleiche 2; (c) Bodman-Schachen
1
The beginning of the EBA-MBA cultural interaction between western and eastern northern Alpine lacustrine groups
145
147
Chapter 11 11.3.1
Locations of the MBA land settlements in Lake Constance and the Lake Zurich region
152
The MBA sites ofHilzingen "Unter Schoren" (D): (a) Plan of the excavated area; (b) A picture of the excavated area
153
The MBA site ofRielasingen-Worblingen "Riedern" (D): (a) Cobbled floor; (b) Postholes of a house
153
The MBA site ofMiihlhousen-Ehingen (D): (a) Hearth; (b) House-plan (black dots indicate postholes); (c) Photograph of the post holes of a house
154
11.3.5
The cobbled floor of the MBA site ofBodman-Breite
155
11.3.6
The Kreuzlingen valley near Constance (CH)
155
11.4.1
Location of the most important MBA land sites in the Lake Zurich region
156
11.4.2
The excavated area of the MBA site ofDietikon-Vorstadtstrasse
157
11.4.3
Part of the stone floor of Dietikon-Vorstadtstrasse
11.4.4
Excavated area of the MBA site of Erlenbach: (a) stone floor made of gravel and pebbles; (b) The hearth from Erlenbach-Obstgartenstrasse
158
Excavated area of Pfaffikon-Hotzenweid:
158
11.3.2
11.3.3
11.3.4
11.4.5a
(D)
32
(a) Stone floor
xiii
32
157
11.4.5b
Excavated area of Pfiiffikon-Hotzenweid: {b) Plan of stone floor and hearth
159
11.5.1
Stone floor made of gravel and pebbles found at the MBA land settlement of Cham-Oberwil Hof
160
The MBA land settlement of Cham-Oberwil Hof: part of the stone floor with traces of wooden planks
160
Cham-Oberwil Hofland settlement: (a) Plan ofa MBA house; {b) House layout showing large post holes
161
11.5.2
11.5.3
NB. In indicating the origins of illustration, "after" implies reproduction of an entire figure; "from" indicates partial reproduction; "modified from" indicates that the illustration has been altered.
List of tables Table 1.6.1 The main Neolithic cultures in the Alpine region
XIV
8
Abstract One of the most discussed topics in studies of Alpine region lake-dwellings is their chronology and continuity. Although lacustrine settlements started to appear around the Alps in the Neolithic, and disappeared permanently at the beginning of the Iron Age, the lakeshores were certainly not continuously settled. Phases of occupation alternated with phases of abandonment. The Bronze Age in the northern Alpine foreland, for example, was characterised by two periods of occupation and two hiatus phases. Although the second phase of abandomnent, which occurred during the Middle Bronze Age (15th-1i 11 centuries BC), was not the longest, it is unquestionably the most intriguing and challenging. The objective of this research is threefold. First of all, to determine whether the MBA occupational hiatus in the northern Alpine foreland was a real phenomenon in need of explanation, and not simply a lack of archaeological evidence, which is likely to be "filled" with further discoveries. Secondly, to clarify the causes that forced the MBA lacustrine communities of the northern Alpine lakes to abandon the lakeshores towards the end of the 16th century BC, resettling them more than three centuries later. Thirdly, to find out where those groups went, once they had deserted the lacustrine environment. The study is approached from two different, but related angles: environmental and cultural. The environmental argues for an abrupt change in climatic conditions, which altered the hydrological balance of the northern Alpine lakes, affecting their levels. These lake transgressions demonstrably influenced the lacustrine villages, forcing the lakedwellers to abandon the area. The cultural approach, on the other hand, analyses the lake-dwellers' cognitive responses to environmental change. Three EBA lacustrine sites are particularly important for the study of the northern Alpine MBA occupational gap: ZH-Mozartstrasse (Lake Zurich), Arbon-Bleiche 2 and Bodman-Schachen 1 (Lake Constance). Not only do these Bronze Age settlements have striking similarities regarding their location, climatic setting and various cultural aspects, but, most importantly, they were occupied and abandoned in the same period. They were the very last ones to be deserted in the last decade of the 16th century BC; just before the MBA hiatus began. The analysis centres on lake-level fluctuations around the three above-mentioned sites where the effects of changing water levels on the settlements' surroundings are simulated and displayed using GIS computer applications. Having defined the cause of abandonment, the final objective of the project is to identify possible location where the lake-dwellers went after they abandoned the lakeshores. Recent discoveries of MBA "land" sites in the very vicinity of Lake Constance and Lake Zurich have raised the possibility that they could have been settled by fonner lacustrine communities. In fact, because of the geographical location of these settlements and the nature of their material culture, and layout, as well as the timing of their appearance, it is most plausible that they had lacustrine origins. The disappearance of a whole class of settlements might give the impression of a catastrophe. But, no matter how severe the environmental variability, it is very unlikely that an entire community would vanish without a trace. The MBA lake-dwellers of the northern slopes of the Alps certainly did not evaporate. Hydrological factors forced them to move away from their premises and settle elsewhere until the lacustrine environment became safe again.
xv
Preface As the title of the book implies, there is an apparent lack of lacustrine settlements on all of the lakes in the northern Alpine region during the Middle Bronze Age. The lakeshores seem to have been abandoned towards the end of the 16th century BC and not reoccupied until the beginning of the 12th century BC. The purpose of this work is to confirm the reality of this interruption in settlements and to clarify the possible cause of that hiatus by suggesting a plausible explanation for the exodus from lacustrine environments. Although the research is mainly approached from an environmental perspective, with a special emphasis placed upon the change in climate as the cause of the exodus, it does not adopt a simplistic climatic determinism; the book also explores the cultural aspect in terms of the lakedwellers' response to environmental variability. Not only did the change in climatic conditions directly alter the lacustrine ecosystem, but it also influenced the socio-economic stability of the lake-dwelling communities and forced them to seek alternatives to the now "hostile" environments around the lakes. Chapter one explores the natural and cultural aspects of the prehistoric Alpine environment, focusing in particular on the wetlands situated on the fringe of the Alpine range. It considers the morphology of the region, the flora, the fauna and the climatic oscillations, which modified the natural and cultural aspect of the area throughout the millennia that followed the last glaciation. Chapter two presents an overview of the historical aspect of the lake-dwelling research from the first discovery (1854) until the present with a view to assessing the reliability of the archaeological record. The chapter is chronologically divided into phases: the 19th century and Keller's pile-dwellings, the first half of the 20th century and the beginning of scientific archaeological research, the first two decades after the discovery of the 14C dating technique and finally the development of the various scientific methods of analysis in the past 30 years. A more detailed description of the history of research is given in chapter three in connection with the intensively debated Pfahlbauproblem (the lake-dwelling problem). The initial issue of whether the lacustrine villages were built on stilts in the water or directly on the surface of the shore was finally solved in the 1980s when scholars agreed that the building-techniques of the lake-settlements did not follow a single style, but that various types depended on the morphological structure of the lakes surroundings. Once the "lake-dwelling problem" was no longer an issue, the main concern of Alpine lacustrine research became chronology and patterns of human occupation. The advent of 14C, dendrochronology and the application of other scientific dating methods revolutionised the entire Alpine chronological chart, especially those parts concerning wetland sites. The variety of the Alpine micro-environments, along with the different research policies of the six countries which are part of the Alpine, region have contributed to the complexity of different archaeological periodisations in the area. Chapters four and five address this issue by analysing the various dating techniques applied the most important lacustrine sites of the Alpine foreland ( chapter 4) and the consequent chronological relationship, which resulted from their application ( chapter 5). Chapter six considers some of the most crucial aspects of lacustrine research such as preservation, discovery, statistical distribution and availability of archaeological evidence, which allows us to determine whether or not, the MBA hiatus in the northern Alpine region was a real phenomenon, or just a misinterpretation of the archaeological evidence. The comparative analyses of the six Alpine countries' most important Bronze Age lake-settlements discussed in Chapter seven leads to the identification of the three EBA lacustrine villages which are most relevant for the understanding of the MBA occupational hiatus in the northern part of the Alpine chain. A detailed description of these three sites, namely ZH-Mozartstrasse, Arbon-Bleiche 2 and Bodman-Schachen 1, is given in Chapter eight. This will provide enough background information to begin discussing the possible cause of the occupational gap. The environmental evidence relating to the lacustrine villages disappearance through climatic deterioration and lake-level fluctuations is more fully discussed in Chapter nine. In addition to providing a detailed description of the palaeoclimatic conditions in the Alpine region during the Bronze Age, the chapter demonstrates, through GIS analyses and computer simulations, how the above-mentioned EBA lacustrine sites' surroundings changed as the floods advanced. It is suggested that an increase in lake levels meant loss of tillable land around the settlements, which consequently triggered economic crises. This environmental variability seems not only to have had a direct impact on many existing settlements, but also to have generated a cultural attitude of suspicion towards lakeside environments even in areas that were not directly threatened. This negative attitude towards the lacustrine life spread rapidly around the region, causing a general abandonment of the lakes (Chapter 10). The exodus from the lakeshores did not obviously mean that those lacustrine groups disappeared without a trace. Chapter eleven, in fact, discusses various MBA settlements found in the vicinity of the lakes, which, because of their timing and characteristics of construction, may have lacustrine origins. They were, in other words, occupied by former lake-dwellers. XVI
Acknowledgements There are many people to whom am very grateful for providing me with invaluable assistance during the preparation of this work. First of all, I would like to express my sincere gratitude to Dr. Andrew Sherratt who provided me with encouragement, useful suggestions and above all, his remarkable knowledge. I would like to acknowledge a number of scholars who were extremely helpful during my field work in the Alpine region: Dr. Pierre Petrequin and Dr. Michel Magny (Centre National de la Recherche Scientifique, Besanr;on), Dr. Pierre Corboud ( Universite de Geneve ), Dr. Helmut Schlichtherle, Dr. Bodo Dickmann and Dr. Andre Billamboz (Landeskenkmalamt Baden-Wurttemberg), Dr. Helmut Schroder (Seeforschung, Langenargen), Dr. Joachim Koninger (Archiiologische Dienstleistungen, Freiburg), Dr. Gunter Schobel (Pfahlbaumuseum, Unteruhldingen) Prof. Michel Egloff (Universite de Neuchatel), Dr. Urlich Ruoff (Stadtarchiiologie, Zurich), and Dr. Eduard GrossKlee (Kantosarchiiologie, Zurich), Erwin Rigert and Dr. Urs Leuzinger (Amt fur Archiiologie, Frauenfeld), Dr. Marcel Joos, Prof. Stefanie Jacomet and Prof. Jorg Schibler (Seminar fur Ur- und Fruhgeschichte, Basel), Dr. Stefan Hochuli and Ursula Gnepf (Kantonsarchiiologie Zug) Dr. Calista Fischer (Universitiit Zurich), Dr. Franco Marzatico (Servizio Beni Culturali Trento), Prof. Leone Fasani (Universita degli Studi di Milano), Dr. Johann Offenberger and Dr. Elisabeth Ruttkay (Naturhistorisches Museum, Wien), Gillian Wallace (University of Cambridge) and finally, Dr. Anton Veluscek (lnstitut za arheologijo, Ljubljana). I am also indebted to friends and colleagues such as Patrick Daly, Andre Tschan, Vuk Trifkovic and in particular Tyler Bell for giving me useful insights about GIS computer applications. A warm thanks goes to the staff of the Institute of Archaeology, Oxford and the Kantonsarchaologie Zug, Switzerland for being so nice and helpful during my research and writing up of this book. Finally, intimate thanks are due to my parents for their support throughout and beyond my studies.
xvii
Abbreviations N LN EBA MBA LBA EI BC BP NAP LBK SD TL CH D A F I SL AG BE SG TG VD
vs ZH
Neolithic Late Neolithic Early Bronze Age Middle Bronze Age Late Bronze Age Early Iron Age Before Christ (Cal.) Before Present (Cal.) Non-Arboreal Plants Linearbandkeramik Standard Deviation Termoluminescence Switzerland Germany Austria France Italy Slovenia Canton Argau (Switzerland) Canton Bern (Switzerland) Canton St. Gallen (Switzerland) Canton Thurgau (Switzerland) Canton Vaud (Switzerland) Canton Valais (Switzerland) Canton Zurich (Switzerland)
xvm
Chapter 1 THE PAST AND PRESENT ALPINE REGION: A GENERAL OVERVIEW mean temperature decreases 1° C. Given this, it is easy to realise why lower locations are preferred in order to make ordinary crops profitable. The valleys and plains between mountains have a wide variety of slope length and width. Valley complexes, which run parallel with the longitudinal axes of extensive mountain chains, are often very large, whereas those that run in a transverse direction are relatively short. The longest valley in the Alps, and indeed in Europe, is that of the Rhone, which up to the point where it enters the plain of Lyon is 370 km long. Next is the valley of the Drave, which is 335 km long up to Varazdin; the valley of Inn as far as Rosenheim is 300 km long; the valley of the Save as far as Sisak is 230 km; and the valley of Rhine as far as Bregenz is 185 km. Other notable valley systems in the Alpine region are those near the Ligurian coast (northern Italy), the Po valley district just before ( coming from the south) the whole pre-Alpine region of northern Italy; the Etsch or Adige district (south Tirol); the valley complex of the Adriatic coast (north-eastern Italy) and finally the whole valley system of the Danube district which, in addition to the aforementioned valleys of the Inn, Save and Drave, also includes the Raab valley near Hungary, the Enns on the north-eastern slopes of Austria and the valley of the Mur some 40 km south of that of the Raab valley.
1.1 Introduction Because of its morphology, the Alpine region constitutes a peculiar microenvironment within the central European continent. As a result, all different natural aspects such as climate, hydrology, flora and fauna retain special features developed and adapted through time to suit the whole ecosystem. This slow process of development and adaptation was not entirely based on natural phenomena, since people have also contributed a great deal. By clearing forests, domesticating animals and plants they altered the delicate balance of the natural ecosystem. People eventually settled the whole Alpine region, establishing a fairly efficient communication network even before the Roman expansion.
1.2 The Alps: location and morphological structure Located in the heart of Europe between Nice and Vienna, the Alpine region falls within a rectangle of about 500 x 200 km. It lies between 44° and 48° North, is 1200 km in length, and has a surface of nearly 200,000 km2 • Its orientation is almost west east throughout Switzerland, where the ranges are less than 100 km in total width and throughout Austria where they again broaden to 150 km. The Alpine chain occupies parts of six countries: France (18% of the Alps surface area), Italy (32%), Switzerland (13%), Germany (4%), Austria (30%) and finally Slovenia (4% ). The Alps surround the north ofltaly like a girdle, broadening out to the east where the chain diverges and extends from the Mediterranean Sea to the Middle Danube and the northern Karst mountains of the Balkan Peninsula where they approach the Adriatic Sea. The Alpine region is also continuous with the Apennines in the southwest and, by means of the Karst Mountains, connected to the chain of the Balkan Peninsula in the southeast. Finally, the mountains of France, Germany and the Carpathian closely surround them respectively in the southwest, north and east. Except in the two places where they join (namely the Apennines and the Karst Mountains), the Alps are bounded by plains and river valleys: on the south by the Mediterranean Sea and the plain of the Po in northern Italy; on the north by the Rhine and Lake Constance, the Suabian-Bavarian tablelands and the Danube; on the west by the Rhone and on the east (at least the spurs) by the Hungarian plain and the Danube.
The Alpine and pre-Alpine lake complex, along with the river network, account for the complexity of the Alpine hydrological system, which is the topic of the next part of this chapter.
1.3 The hydrology of the Alps: rivers and lakes With a river network, high altitude glaciers and lakes, the Alpine region includes the largest reserves of fresh water in central Europe. Ahnost every Alpine valley system has its own water supply, which is provided mostly by rivers and lakes. The sources of the Alpine rivers lie at varying altitudes from the foot of a mountain to the flank or even near the sunnnit. Most of them emerge as springs, while others have their origin in glaciers and ice fields. The latter dry up in winter and have greatest flow in the hottest part of the summer. Ahnost every valley in the Alps is crossed by a river whose size and importance depend upon the size and location of the valley itself. In the southern part of the Maritime Alps (southwest) for example, rivers, which from the Alpine chain flow towards the Ligurian coast, are the Arrosia, the Taggia, the Nervia and the Roja. These rivers, like most in the area, are all part of a relatively simple valley system. One exception is the Var, which rises in Mont Pelat and is fed by a number of tributary streams such as the V oire, the Tinee and the V esubie before emptying into the
Considering the Alpine morphological structure from a human point of view, valleys are of far greater importance than mountains. Even among the hilly lands and low mountains, the agricultural value of the valleys is as a rule greater than that of the hills. This is particularly true in the Alps proper where altitude plays a crucial role: on average, for every 170 metres of increased height the
1
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
Mediterranean west of Nice. Moving north in the Rhone district the main river is indeed the homonymous river Rhone which, from the slopes of the St. Gothard range, flows for 370 km before meeting the Saone in Lyon. On its way to France, the Rhone is fed by various tributaries before flowing into Lake Geneva (near Montreux) which it leaves around Geneva City heading for the Haute Savoie and eventually Lyon.
In the southern part of the Alps the most important river
is without doubt the Po, which rises, in the western slopes of the Alpine chain and cuts through the whole Po Plain flowing into the Adriatic near Porto Tolle. The River Po has a large number of tributaries including the V araita, the Maira, the Sturn and the Tanaro at the beginning of its course, the Ticino coming from Lake Maggiore, the Adda from Lake Como, the Oglio from Lake d'Iseo and the River Mincio an estuary of Lake Garda. The southern side of the Alps also called South Tyrol has, in particular, one major valley system, the Etsch or Adige valley with the River Adige. The sources of the Adige are formed by several small lakes in the Reschen Scheideck region, the river first flows south through the Malser then it takes a south-easterly course from Merano to Bolzano and south again towards Verona where it turns left flowing parallel with the Po until the Adriatic Sea.
The Rhine district is a particularly complex valley system. The the main river is the Rhine which, similarly to the Rhone, rises near the St. Gothard, but in this case it flows eastwards. Before flowing into Lake Constance (near Bregenz) the Rhine takes two names, the VorderRhein and the Hinter-Rhein which both cross the entire Adula Alps. The Rhine leaves Lake Constance from the lake's western end also called Untersee and it receives the waters of three other fairly important rivers the Aare, the Reuss and the Limmat before running towards Basel and leaving the Alpine region. The sources of the Aare lie in the Oberaarhom and Lateraarhom near Lake Morat. On its way to the Rhine it forms Lake Brienz and Lake Thun, flows through the Bernese Alps and it is joined by the Saane just before Kerzers. The Reuss also comes from the St. Gothard and, flowing through the central Swiss Plateau, forms Lake Lucerne before joining the Rhine. Finally, the Limmat flows out of Lake Zurich in Zurich itself. In its upper course it is called Linth, and is carried, by means of a canal, into Lake Walen, which is also connected with Lake Zurich.
The Alpine lakes are divided into two classes: border and mountain lakes. The border lakes, as the name implies, are found on the border of the Alps and are of considerable size, but are situated at no great altitude above the sea level, whereas the mountain lakes lie among the Alpine chain and are relatively small. Valley lakes are a sort of transition form between border and mountain lakes and retain the characteristics of both. Large lakes which belong to the border lake category all lie round the central group of plains and valleys which under various names are bounded by the parallel chain of the Jura. These lakes generally have elongated basins, lying sometimes southwest and northeast and sometimes southeast and northwest. Large basins at the foot of the Limestone Alps and the Jura, such as Lake Neuchatel, Lake Morat and Lake Bienne corrispond to the first description and Lake Brienz, Lake Sam and the lakes of Engadine also have the same direction. The large lakes on the Italian side lie almost parallel with the Jura chain and lakes. Lake Constance, Lake Zurich, Lake Sempach, Lake Zug and Lake Thun, on the other hand, lie southeast/north-west. Northern Alpine border lakes which have different orientations are Lake Bourget, Lake Annecy and Lake Chalain in the French Jura, Lake Walen and Lake Baldegg in Switzerland, Lake Chiem in Germany, Lake Mond, Lake Wolfgang, Lake Millstatt, Lake Wissen, Lake Worth and Lake Neusiedl in Austria and fmally, three minor lakes in the southern part of the Alps namely Lake Varese and Lake Ledro (Italy) and Lake Lugano in Switzerland. The remaining border lakes such as Lake Stamberg and Lake Ammer (southern Germany) and Lake Att, Lake Traun and Lake Allstatt (Austria), scattered around the Alpine foreland lie north/south similar to that of the main northern Italian lakes.
The Danube valley system and the River Danube itself are the northeastern margin of the Alpine region. Nevertheless the Danube is important because many of the northeast Alpine rivers form part of its hydrological system. One of the most relevant is the Inn which rises in the Grison, runs through the Engadine and further on towards Austria crossing the Kitzbilhler Alps and towards the Bavarian Plateau where it is joined by the Salzach before itself joining the Danube. Another important river is the Traun, which is fed by the Kammer, Toplitz, Gundel, Alt-Aussee and Oden lake regions, flows southwest, and fills the basin of Lake Hallstatt at the northern foot of the Dachstein, then runs northwards as far as Ischl, where it is joined by the River Ischl flowing from Lake St. Wolfgang, then north-east, it broadens out into Lake Traun, and leaves the Lower Alps below Gmunden, joined at Lambach, by the waters of Lake Fuschel, Lake Mond and Lake Att in Austria. The longest river in the eastern Alpine valleys is the Drave (about 335 km). Its sources are in the Tyrolean Puster valley and, it loses its Alpine characteristics and joins with the River Mur 40 km east of V arazdin after a j oumey of ca 370 km through the Carinthian-Styrian Alps. Another relevant group of rivers which, like the majority in the area, flow from the eastern Alpine complex towards the north-eastern and eastern Danubian district contains the Enns, the Inn, the Rienz, the Raab and the Save.
The majority of the aforementioned lakes were dotted with prehistoric lake-dwellings and will therefore be taken into consideration in the following chapters of this thesis.
2
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
1.4 The Alpine climate from the late glacial to the present
(1977) and Patzel (1977) started to formulate chronological charts of climatic fluctuations during the Late Holocene. Research progressed and results became more reliable in the 1980s when Gamper (1981), Renner (1982) and Suter (1982) attempted to correlate four types of scientific analyses (14C, dendrochronology, ice coresampling and pollen analysis) in order to see whether the results were similar or they followed completely independent patterns. The study confirmed remarkable similarities, although regional diversity had to be carefully taken into account especially as far as the southern and northern pre-Alpine areas are concerned. Additional contributions to the Alpine palaeoclimate were made by Burga (1991a, 1991b, 1993, 1995) in the 1990s. Magny (1993, 1995b, 1995c) along with Gross-Klee and Meise (1997) went even further considering the changes in insolation as one of the many possible causes of past climatic fluctuations. Thanks to three decades of Alpine palaeoclimatologic research, it is now possible to produce a chronological chart of the Alps' climate subdivided in specific periods evidenced by absolute dates starting from the Younger Dryas (ca 11,000 BP), up to the present.
The Alps have a mainly temperate climate. The northwestern part is influenced by an Atlantic maritime climate, the south by the proximity of the Mediterranean Sea, while the northeast has a continental climate. The contrast between the climate of the pre-Alpine belt and the continental climate of the intro Alpine valleys is an important factor. Some scholars subdivide the Alpine range in five zones each characterised by a specific climate: the Mediterranean zone (Liguria, Provence and Maritime Alps) with its marked dry seasons typical of Mediterranean-type climates, the External Southern Alps with a Insubrian-type climate (warm and wet), the External Northern Alps with a sub-oceanic-type (cold and wet), the Internal Eastern Alps (continental cold and wet climate), and finally the Central Valleys with a continental dry climate (Aeschimann and Guisan 1995). Before considering the regionality of the Alpine and preAlpine climate in more detail, attention will be focused on the past climatic variations, which occurred between the late- glacial (11th millennium BP) and the present. Climatic fluctuations are the result of global, regional and local geological changes that influence atmospheric and hydrological circulation-patterns (Rousseau et al. 1994). Thanks to multidisciplinary studies, palaeoclimatology provides powerful tools for understanding the possible consequences, which climatic and environmental changes might have had on human settlements during prehistoric times. Various methods are used to study past climatic conditions. The most common and well-developed ones are, for example, pollen analyses and stratigraphic investigation of soil deposits (sedimentology). Biological evidence can also be obtained from diatom and plant microfossil analyses, fossil insect remains, marine and non-marine molluscs and even from animal remains such as bones and horns. Extremely rich palaeoecological data are retained by lake deposits on which all the abovementioned analytical techniques can be applied to reconstruct both palaeoclimatic and palaeoenviromental conditions (Burga and Furrer 1982; Lowe and Walker 1997; Rapp and Hill 1998). An important aspect of the mineralogical and biological proxy data is that they can be dated using quite reliable absolute dating methods such as radiocarbon, varve chronology, dendrochronology, lichenometry and cryology with a particular emphasis on the annual layers in glacier ice (see Chapter 4 for a more detailed description of some of these dating methods). It is indeed thanks to these scientific techniques that a palaeohistory of the Alps' climate in the last 10,000 years can be fairly accurately reconstructed.
The Younger Dryas (11,000 - 10,000 BP) in the Alps is characterised by the last Late Wfumian climate deterioration clearly shown by the expansion of glaciers and by the tree-line depression registered all over the Alpine region. During the Preboreal (10,000 - 9000 BP) climatic conditions improved, the most important glaciers retreated and consequently the tree line rose especially on the southern slopes of the Alps. The climate became much milder than in the previous phase (Magny 1995b). In the Boreal phase (9000 - 7500 BP) the temperature increased even further ( especially up to 8400 BP) and the glaciers became smaller reaching the same dimensions as they had in 1920. Vegetation also improved and some species such as the spruce started an immigration process following suitable microclimatic areas. Favourable climatic conditions were however limited to some zones and in fact, the glaciers in the Wallis area experienced an expansion between 7700 and 7400 BP. The Older Atlantic phase (7500 - 6000 BP) was a quite long and stable climatic phase with another increase in climatic conditions within which temporary cold phases took place. These periods of climatic deterioration are registered between 7250 and 7190 BP, 7020 and 6960 BP and fmally 6550 and 6480 BP. From 6400 to 5750 BP the Alpine region experienced two different kind of climatic regime: a deterioration in the eastern part (Austria and Tyrol) and a stable phase in the western slopes (northwestern Italy, France and Switzerland) (Gamper and Suter 1982; Burga 1991). In the next climatic phase, the Younger Atlantic (6000 - 4700 BP) the climate became more unstable with alterations of relatively cold and temperate periods. According to dendrochronology, warm phases are registered from 5600 to 5525 BP and from 5065 to 4960 BP whereas cold periods occurred from 5745 to 5695 BP; from 5275 to 5150 BP and from 4960 to 4855 BP. The glacier-fluctuations became more
Palaeoclimatic studies within the Alpine region began in the late 1960s and beginning of the 1970s when, following accurate results obtained from pollen analyses and glacier-ice core sampling, scholars such as Schindler (1971) Zoller (1977) and Rothlisberger (1986), Furrer
3
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
irregular with a general tendency to expansion between 5200 and 4400 BP (this period is also knomi as the two little glaciations, Piora 1 and 2), but with some phases of retreat as well (Renner 1982). During the Subboreal phase (4700 - 2700 BP) and more precisely after the two aforementioned little glaciation periods, Piora 1 and 2, a long phase of fairly mild climate began. The most favourable climatic conditions occmTed between 4400 and 3600 BP (Gamper 1981; Burga 1991). This climate did not last and, between 3340 and 3175 BP we have the coldest phase of the entire Last post-glacial time. Glaciers expanded, precipitations increased and the tree line dropped throughout the whole Alpine region (Renner 1982; Magny 1995c). The climate started to improve at the beginning of the Subatlantic phase (ca 2700 BP Present), but for a short period. In fact, dendrochronologic analyses show cold phases right between 2700 and 2640 BP, and from 2570 to 2490 BP, intervalled by only two warm periods: from 2640 to 2570 BP and from 2490 - 2430 BP confirming a general cold climate. The climate kept fluctuating also during the historical time, from the Roman to the Middle Ages with a tendency in cool conditions or at least not as warm as in the first part of the Subboreal phase. The Alpine glaciers experienced a more stable period although cryological analyses show some transgressions between 1950 and 1800 BP; from 1200 to 1000 and lastly from 900 to 700 BP (13th century AD) (Gamper and Suter 1982).
well as the hydrological conditions of Lake Constance and Lake Zurich will be taken into account. 1.5 Flora and fauna of the Alpine valleys in the past 11,000 years.
The Alps, in consequence of their location in the interior of Europe, of their climatic conditions, and partially also of their geological conditions, have a characteristic flora and fauna with an obvious distinction from other central European areas. It might be remarked, however, that the flora has, generally speaking, a more specific stamp than the fauna, which is far more resilient to climatic variations. As it was the case for the climate, also the Alpine vegetation was never constant in the past. Thanks to pollen analyses carried out in various zones within the Alps, it has been possible to reconstruct a chronological chart of flora changes over the last 15,000 years. Unfortunately, due to problems of pollen preservation and derivation, sedimentological unconformities and topography, the number of "complete" palaeovegetational records is fairly limited (Bintz et al. 1991; Burga 1988). Furthermore, not all areas in the Alps have been thoroughly investigated; the country with the highest number of pollen analytical diagrams in regard to palaeovegetation is without any doubt Switzerland, followed by eastern France, southern Germany, northern Italy, Slovenia and Austria. Following the same chronological criteria used in subchapter 1.4 for the palaeoclimatic variations, a general description of some of the main vegetational taxa and their local migrations and changes within the Alpine region will be provided. A particular emphasis will be placed upon the pre-Alpine lacustrine areas, which were settled by prehistoric populations from the Neolithic to the Late Bronze Age (Subboreal - beginning of the Subatlantic ).
Having considered the Alps' climatic variations in the past, it is now important to emphasise their regionality and seasonality in the present Alpine region. In temperate Alpine and pre-Alpine regions most ecosystems change far more from winter to summer and from one microenvironment to another than they do from year to year. The division of the Alps into five different zones, namely Mediterranean, External Southern Alps, External Northern Alps, External Eastern Alps and the Intro Valleys is sometimes not enough to classify the remarkable diversity of the Alpine climate. The climate can be totally different from one valley to another within an hour's wallc This contrast is more pronounced between the northern and southern slopes and in particular in the vicinity of the St. Gothard Pass where two different kinds of climate may be encountered within only a few kilometres. On the other hand, this contrast is not so obvious in the eastern part of the Alps between Austria and the south Tyrol. The Tyrolean Alps have themselves another peculiar characteristic; for particular geomorphological and aeolian reasons this area has 15% more sunshine throughout the year and also the precipitation is much lower than in the External southern and northern Alps (Barry 1992). This factor ce1tainly influences the hydrology of that specific Alpine microenvironment with a repercussion on the local water basins. A more detailed consideration of this phenomenon related to different levels of precipitation in very limited areas will be given in Chapter 9 where the regionality as
Younger Dryas (11,000 - 10,000 BP) vegetational evidence in the Alps is generally well documented. During this period pine (Pinus) usually decreases, whereas the NAP percentages increase, especially sage (Artemisia), joint-pine (Ephedra), and goosefoot (Chenopodiaceae). The climatic deterioration during the Younger Dryas is generally expressed by a tree-line depression of about 200-400 m, and in several localities the typical Late Wfumian steppe elements increase their representation. A comparison between pollen analytical evidence for this climatic fluctuation and the distance of the investigated areas from glacial deposits of that period, shows a good correspondence between climatic and vegetational change.
During the Preboreal phase (10,000 - 9000 BP) the analysed sediments in the High-mountain regions are mainly inorganic, and it is therefore problematic to obtain reliable radiocarbon dates. In the Alpine foreland, on the other hand, pollen analytical studies and dating processes
4
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
are facilitated by the deposits' fairly high concentration of organic sediments. Despite the Preboreal phase in the central Alps and the pre-Alps being in general characterised by a massive expansion of spruce (Pinus sylvestris) and birch (Betula), the vegetation at the altitude of the present tree- line remained very scattered. At lower altitudes and especially in the southern slopes of the Alpine region larch started to become more frequent and a first appearance of mixed oak forests is recorded in southern Switzerland.
Human impact on vegetation became more evident during this phase, the lacustrine environment started to be occupied and larger surfaces of woodland were cleared for tillable soil. Traces of agricultural activity and in particular wheat and barley production are found around Lake Zurich and Lake Constance during the Egolzwil and Cortaillod periods (Gross 1987; Liese-Kleiber 1987; Jacomet et al. 1995), in the Lake Feder area in southern Germany (Liese-Kleiber 1995) and in the Lake Varese district in north-western Italy (Banchieri 1981, 1991).
The rich organic sediments in the Boreal period (9000 7500 BP) show a rapid expansion of the hazel (Cory/us) especially in the northern latitude pre-Alpine areas. The hazel spread took place also in the northern pre-Alps although in some regions such as the Wallis and the Grisons (Switzerland) its presence was fairly limited. High altitude vegetation was mainly formed by larch (Larix) and, during this period, there was also an initial migration of the spruce from the eastern Alps of Austria to the Swiss regions. Spruce also dominated the southeastern slopes of the Alpine chain whereas the southwestern parts were mainly characterised by species of mixed oak forest.
During the Subboreal phase (4700 - 2700 BP) human impact on the Alpine vegetation intensified even fmther especially from the Neolithic to the Late Bronze Age and traces of woodland clearance and agricultural activity involving all sorts of crop production were to be found almost every where in the Alps (Rachoud-Schneider et al 1998). As far as natural woodland was concerned, in both the pre-Alps and the central Alps alder achieved a wide extension invading, in some cases, deforested areas, spruce reached its fittest extent to the west as far as Lake Geneva and the French Jura, and finally, beech became quite common in the northern Alpine foreland in particular around Lake Zurich and the northern part of Lake Constance.
In the Older Atlantic (7 500 - 6000 BP) the main forest vegetation of both the Plateau and the pre-Alps was a mixed deciduous forest with abundant hazelnut, whereas in the continental central Alps pine (Pinus) and birch (Betula) occurred with some elements of mixed oak woodland, as in the Valaisian Alps. The migration of spruce from Eastern Europe, which started in the Boreal phase, continued reaching as far as Lake Lugano in southern Switzerland. In some areas of the northern Alpine region Foreland considerable percentages of Alder (Alnus) are recorded for the first time. Human impact on the Alpine vegetation occurred much earlier than what was previously thought. In fact, recent studies with more accurate pollen analyses show that the first evidence of deforestation and a related beginning of agricultural activity took place between the Older and the Younger Atlantic about two millennia before the first lacustrine settlement (Erny-Rodmann et al. 1997). Figure 1.5.1 shows the various places where some of the oldest traces of plant cultivation were found within the Alps and preAlps.
In the Subatlantic period (2700 BP - Present) climate variations have no longer been the main causes of flora transformations in the Alpine region. Human exploitation and management of woodland areas have reached extremely high levels especially from the Roman time; large surfaces of forest have been cleared for buildings and ski slopes grounds as well as an ever-increasing road network. Pollen analyses from Late Bronze Age, and Iron Age peat bog profiles show an intensification of beech, alder and pine, and in a number of sites also a high percentage of oak pollen related to the use of acorns to feed pigs. The chestnut, which is quite common in the whole northwestern Italian pre-Alpine region, is not endemic, but it was introduced in the area by the Romans (Burga 1988). Finally, the majority of spruce and pine forests, which can be seen in the Alps today, are humanly managed monocultures to satisfy the high demand of the wood industry. Turning from the vegetation kingdom to the animal kingdom, one can notice that, with regard to its descent and its original home, the Alpine fauna is composed of three elements. The first may be considered as the altered remnant of an earlier tropical and sub-tropical species, which inhabited Europe in the Eocene. Another consists of species such the marmot and the ptarmigan which have reached here from the north and taken refuge in the mountains, finally a third consists of those from northern and central Asia which arrived in the Alps after the last glaciation; two good examples are the lemming and fieldfare. Animal life appears in the Alps at heights corresponding to those at which vegetation is found and the number of species generally increases the lower the
The Younger Atlantic (6000 - 4700 BP) is characterised by the first extension of the tree line to an altitude of about 2300 metres a.s.l. and a snow-line of over 3100 metres in the central Alps. During this period we have a general migration of the beech (Fagus) from west to east with the first forest of fir and beech settling the hilly preAlpine areas of the Swiss plateau, the northern foreland and some low altitude areas of the southern Alpine slopes. The spruce migration reached Lake Constance and part of southern Germany in the north, and as far as the Locamo area on Lake Maggiore in the south where the main vegetation of the period was a mixed oak forest.
5
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
altitude above the sea level. In terms of numbers, mammals are far exceeded by birds.
economy. As a result, some species such as the bear, the wolf and the bison became extinct (Schibler and HiisterPlogmann 1995). Some of these extinct species have recently been reintroduced and they are now protected in the numerous national parks scattered around the Alpine region.
The species of wild mammals of the relatively low Alpine valleys are confined to several kinds of bats, marten, wild pig and numerous species of mice and hares. Woodland areas are rich in bird species, some of the most common are ptarmigan, raven and crow plus a large number of hedge sparrows, robins, blackbirds and starlings. The owl family is represented by the great and little homed owl, the dwarf-eared owl and some varieties of screech owl. In addition to these nocturnal predators, there is a large number of those which hunt by day such as the goshawk, falcon and eagle. A fairly high percentage of the Alpine fauna is to be found in, on and around these water basins. Trout, pike, perch and salmon populate the underwater environment whereas waterfowl such as the wild duck, water hen and heron have settled the majority of lakes, rivers, swamps and other minor areas surrounding the water basins. A considerable number of frogs, lizards, salamanders and other small reptiles such as the viper and asp are also to be found. Needless to say, the number of fresh water molluscs, spiders, as well as all sorts of insects, is enormous although it varies seasonally.
1.6 Human occupation and mobility within the prehistoric Alpine region
Considering the harsh climate as well as environmental aspect of the Alpine region one might thinl( that our remote ancestors did not dare to penetrate it until, archaeologically speaking, fairly recent times. On the contrary the very first evidence of human beings roaming around the edges of the Alps dates back about a million years: human remains belonging to Homo erectus have been found in the southern part of the Maritime Alps and more precisely at "Grotte du Vallonet" on the Riviera at Roquebrune-Cap-Martin a few kilometres east of Monaco (De Lumley 1963). More recent traces of human occupation (ca 600 kya) have been unearthed in the same area at Terra Amata near Nice and at Cava Vecchia near Quinzano, Verona (Italy), but in these two cases the occupation was also limited to the Mediterranean outskirts of the Alpine region (De Lumley 1966). The real settling of the Alps began in the Riss/Wurm Interglacial that is some 100,000 years ago. To this period date the finds from the "Drachenloch" above Vattis in the Tamina valley (2445 m a.s.l.) near the River Rhine in Switzerland. More evidence, from 70,000 to 40,000 years ago, comes from the Simmen valley near Bern and from the Wildkirchli on the Santis. Traces of human occupation during the Palaeolithic are not only present in the Swiss Alps, but also the north-eastern Alps were settled, with the maj01ity of caves which hold human remains being found in the Austrian province of Styria. It is important to clarify that despite this Palaeolithic period in the Alps being named as the "Alpine Palaeolithic", this designation does not imply a cultural unity, but is a matter of a few individuals who had only temporarily left the favourably situated Alpine fringes and went to the mountains (Mottl 1975).
As the altitude above sea level increases, the number of animals becomes more limited and only a few species are to be found around the tree line and above. Mammals include mole, marten, marmot, chamois, red deer and wild goat whereas as far as birds are concerned there are some rock eagles, crows and ptarmigans. Insects are confined to several kinds of flying beetle, bumblebee and wasp; the omnipresent earthworm is apparently the only member of its family, which ascends as high as the upper snow region. In conclusion, a few words must be devoted to the domestic animals, which, like the wild ones, have also some peculiarities. For example, the superiority of the Swiss, Austrian and Tyrol cattle is recognised within the whole Alpine region and even in central Europe. Goats and sheep also have special Alpine characteristics with the most distinctive being those found in the northwestern Italian Alps. As mentioned earlier, the Alpine fauna is not influenced by climatic variations as much as the flora; the main species of wild animals have adapted themselves so well to the various and sometimes harsh environments of the Alps that even an abrupt change in climate (unless very drastic) does not jeopardise their survival. Apart from some species, which became extinct, the Alpine fauna has maintained a similar aspect throughout the whole Holocene. The above-mentioned extinctions were almost totally due to human influence and in particular to hunting. People have always hunted in the Alps since they first colonised them, but the practice intensified considerably from the Early Neolithic and, in spite of the increase of agriculture development throughout the Bronze Age and Iron Age, hunting remained an important activity for the Alpine populations' subsistence and
People started to live in the Alps permanently after the last glaciation and more or less at the beginning of the Holocene, although occupation was still limited to the lower valleys of the northern and southern Alpine foreland. In fact, Epipalaeolithic and Mesolithic high altitude camps such as those of the Sellajoch (2214 m a.s.l.), Reiterjoch (1980 m a.s.l.) and Plan de Frea (1930 m a.s.l.) in southern Tyrol (Lunz 1981) and the more recently discovered rock shelter of "Schwarze Tschugge" near the Matterhorn in Switzerland (Curdy et al. 1998) were used by hunters and shepherds only on a seasonal basis. Although various populations had already settled within the Alpine region by the end of the Mesolithic (ca 6000 BC), it was not until the Neolithic that specific regional groups started to develop and external human
6
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
influence from southern and northern Europe began. This so-called "neolithicising" process reached the Alpine region by two routes: from the south, that is from the Mediterranean and from the southeast overland diagonally across the Balkans. This period in central Europe is characterised by the expansion of the Bandkeramik or Linear Pottery culture. This culture was based on agriculture; there was a constant search for fertile land. When LBK groups arrived at the fringe of the Alps, they took no interest in the harsh mountain terrains. As a result, distribution maps of settlements with Bandkeramik show clearly how this culture by-passed the Alps in the north where, in the region between Basel and Lake Constance, it actually did not cross the Rhine at all towards the south (Pauli 1984).
the Alpine region was completed by reaching Bavaria in southeastern Germany and the Kammergut region in Austria towards the beginning of the 4th millennium BC (Fig. 1.6.1) (Schlichtherle 1997a). Although this expansion shows mobility within the Neolithic Alpine region, it does not prove cultural uniformity amongst the lake-dwellers. In fact, at least 24 different Neolithic cultural groups have been recognised within the entire Alpine foreland (see Table 1.6.1).
A new phenomenon known as the "lake-dwellings" started to develop in the Alpine area during the Neolithic. These prehistoric populations settled close to lakes, swamps and other wetland environments within the whole pre-Alpine and Alpine region. A number of theories have been formulated regarding the origins and spread of the lake-dwelling phenomenon, but the most plausible one seems to be that which argues for a southwestern provenance. This hypothesis is based on palaeobotanic analyses of a specific kind of wheat also called the "lakedwelling wheat" (Triticum aestivum ), which is commonly found, on most of the wetland sites around the Alps. Apparently, the origins of this wheat are to be located in the Mediterranean area and in fact, traces of it have been found in the stratigraphic deposits of some of the Catalan lacustrine environments near Banyoles in south-eastern Spain and on Lake Bracciano (central Italy) which date back to the 6th millennium BC. The same cultivated cereal taxa appear in the Po Plain lake deposits namely Lavagnone and other settlements around Lake Garda in northern Italy and in particular on the numerous sites on Lake Chalain in the French Jura in the 5th millennium BC. The northwards expansion continued in two directions at the end of the 5th millennium BC. One from eastern France and western Switzerland to the Swiss Plateau and southern Germany reaching first Egolzwil on Lake Wauwil (CH), then Lake Constance (CH/D) and Lake Feder (D). The other northwards shifting started from the southern part of the Alps moving to the Ljubljana marsh in Slovenia. The spread of "lake-dwelling wheat" around
Fig. 1.6.1 The expansion of the occurence of the "lakedwelling wheat" (Triticum aestivum) as an indication of the spread of the lake-dwellings in the Alps (after Schlichtherle 1997a: 12)
Migration and acculturation processes made these lacustrine groups disappear during the Bronze Age, but new cultures emerged and the lakeshores were continuously occupied until the beginning of the Iron Age. Some of the best known Bronze Age cultures within the Alps are the Rhine culture in France and western Switzerland, the Arbon and Bodman cultures (eastern Switzerland and southern Germany), the Polada, Fiave and a few others (northern Italy) and, fmally, the Urnfield culture which covered the whole northern pre-Alpine region, except Austria, in the Late Bronze Age (Strahm 1997).
7
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
CULTURES
COUNTRIES e/CH
Chalain Chalain recent Clairvaux Ferrieres Corded Ware Saone-Rhone Lilscherz Goldberg III Horgen Vinkovci Vucedol lg (a&b) Maharski-Prekop (a&b) Cortaillod Pfyn Altheim Bajc-Retz Schussenried Aichbilltl Homstaad Resnikov-Prekop Lagozza Vasi a Bocca Quadrata
* * *
*
*
w/CH
* * *
*
A
s/D
e/F
nil
SL
* * * *
* * * *
* * * *
* *
* *
* * *
* *
*
Table 1.6.1 The main Neolithic cultures in the Alpine region (e/CH= eastern Switzerland; w/CH= western Switzerland; s/D= southern Germany; A= Austria; e/F= eastern France; nil= northern Italy; SL= Slovenia)
The fact that the single cultures covered fairly large areas and their goods were distributed to various localities in the Alps, especially during the Bronze Age, proves that different groups were certainly in contact and that such contact was an important factor in this period. As it will be seen in more detail in sub-chapter 10.3, pottery and other goods were traded along various routes in the Alpine region. Generally speaking, this exchange activity was on a regional basis, but in some cases the Alps were influenced by very long distance trade routes such as the so-called "Alpine route" from the east Mediterranean to the Baltic passing through the central Alps in the Middle Bronze Age (1500-1300 BC) and Late Bronze Age (see also Fig. 5.1.1).
1.7 Cultural transformations of the Alpine environment: subsistence and economy from the Neolithic to the Bronze Age
From the Neolithic onwards, climatic change was no longer the only factor affecting the vegetational aspect of the Alpine environment. Human influence on the wild flora became more and more relevant and the consequences were in some cases fairly drastic: deforestation, landscape morphological transformation, hydrological balance alteration and erosion. Plants started being cultivated, some wild taxa genetic make-up was altered and, as a result, new "domesticated" crops began to be used. A good tool for gauging all these humanly caused environmental changes is, once again, palynology. In fact, not only do pollen analyses give the archaeologists the possibility of reconstructing palaeoclimates and palaeoenvironments in general, but they also provide evidence concerning human use of the land such as agriculture, pastoralism and forest management.
Small and large-scale trade connections as well as migrations within, and immigrations from outside the preAlps and Alps increased during the Iron Age. In addition to bronze and copper, iron and salt also became important elements of exchange and the Alps became an important and strategic crossing point between north and south Europe. The Late Iron Age saw the development of the limited number of cart-roads, which existed before the Romans, into a fairly efficient communications network.
Some kinds of pollen are characteristic of specific habitats and according to the presence of these various pollen taxa, one can reconstruct the vegetation of a particular area. For example, typical pollen indicators of
8
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
woodland clearance phases are: Epilobium agustifolium, Alnus viridis, Selaginella selaginoides, Botrychium and Pteridium aquilinium and similarly, permanent settlements are dotted with pollen residues of Ligustrum, Crataegus, Prunus and Berberis. Also, pre-Alpine and Alpine pasturage activity can be easily identified, with the most characteristic indicators being Plantago lanceolata, Selaginella selaginoides, Alnus viridis, Artemisia and Thalictrum. Another indicator of human occupation is without doubt the presence of weeds. The areas around dwellings and cultivated areas are usually covered in weed taxa such as Rumex acetosa, Urtica, Polygonum aviculare, Centaurea cyanus and Convolvulus arvensis.
in many villages the intake of meat obtained from wild game was much higher than that of domesticated animals. Despite the dwelling location in the proximity of lakes, fish never occupied a major part of the lake-dwellers diet. In fact, fishing equipment as well as fish bones has always been found in very limited numbers within the anthropogenic layers of the sites and in some cases they were completely non-existent. In order to create more space for the expanding settlements, agricultural activity and pasturage, forest clearance increased even further in the Late Neolithic and during the Bronze Age. Good evidence comes, once again, from the lacustrine settlements where, because of the peculiar geomorphology of the areas around the lakes, people did not have much land suitable for agriculture and consequently the search for more tillable terrains made the anthropogenic transformation of the environment even more evident. This is particularly true in areas which show continuity in occupation such as the sites of Auvernier and Hauterive-Champreveyres (Lake Neuchatel, CH), the whole Zurich bay (Lake Zurich, CH), the Bodman area near the Espasinger Plain (western extreme of lake Constance, D), the Arbon bay on the Swiss shore of Lake Constance and the entire Carera basin the location of Fiave Neolithic and Bronze Age lake-dwellings (northern Italy). In fact, thanks to thorough pollen investigations, it has been possible to reconstruct part of the Zurich bay environment showing the transformation it underwent from the first Neolithic settlements to those of the Early Bronze Age; large areas had been deforested and cultivable as well as pasturage land reclaimed from swampy terrains (Figs. 1.7.1 and 1.7.2) (Jacomet et al. 1990; Brombacher and Jacomet 1997).
Pollen identifications, such as the appearance of cerealiatype, Plantago lanceolata and clearance phases, suggest that the earliest anthropogenic influence on vegetation in the Alps took place between 7000 and 4500 BC (ErnyRodmann et al. 1997). It has to be pointed out though, that this is only true as far as the lowland pre-Alpine areas are concerned. In fact, high altitude agricultural activity did not start until the second half of the 5th millennium BC. Some examples of early mountain cultivation are to be found at Schwarzmoos (1770 m a.s.l.), Lai da Vons (1991 m a.s.l.) and Simplon-Hobschensee (2017 m a.s.l.) (Burga 1995). Agricultural activity in the Alpine region started to be well established with the advent of the first Neolithic lacustrine villages such as those of the Cortaillod culture, Pfyn and Egolzwil groups. Although it varied from place to place, it can be said in general terms that their economy was essentially based on cereal cultivation and animal husbandry: cattle, pigs, goats and sheep. Hunting was also an important activity in the daily economy and
Fig. 1.7.1 Attempted reconstruction of the Zurich bay environment in the Early Neolithic (after Jacomet et al. 1990: 88)
9
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
Fig. I. 7.2 Attempted reconstruction of the Zurich bay environment in the Late Neolithic (after Jacomet et al. 1990: 89)
Although subsistence and economy among the lacustrine settlements were fairly similar, percentages and mode of production varied from site to site and of course from region to region. For instance, while cereal production occupied a major part within the Alpine lake-dwelling economy, at Fiave it appears to have played a subordinate role serving mainly to provide winter feed for cattle. In fact, taking the Carera basin as a whole and considering the demography at the time of occupation, the maximum area of cultivated land is reckoned to have been only 17.5% of the usable ground around the settlement (Perini 1987). A similar case occurred at Arbon-Bleiche (Lake Constance, CH) where the scarcity of agricultural land is reflected by the high percentage of wild game hunted during the period of settlement: almost equal to the percentage of domestic animals. An opposite case is noted at the Late Neolithic and Early Bronze Age Village of ZH-Mozartstrasse (Lake Zurich) where the domestic animal meat consumption was more than 90% of the total (Schibler and Suter 1990). As the time went by, agriculture became less and less important within the Alpine region economy. In fact, with the disappearance of the lake-dwellings during the Iron Age and, even more, in more recent historical time, cattle breeding became the major factor of the whole pre-Alpine and Alpine region and, as it was pointed out in subchapter 1.5, Swiss and Austrian cattle are today recognised as one of the best hybrids in central Europe.
10
Chapter 2 LAKE-DWELLINGS IN THE ALPS: DISCOVERIES AND SCIENTIFIC RESEARCH rapidly spread around the entire Alpine region. By the tum of the century, hundreds of sites were discovered and a number of scholars had already begun to study them.
2.1 Introduction
During the 19th century the lake-dwellings of the Alpine region emerged as one of the classic forms of well preserved archaeological monuments, in the same way as "hillforts" or "shellmiddens". Their state of preservation and characteristics made them famous throughout Europe. Although a number of artefacts belonging to these lacustrine villages had already been found along some lakes within the Alpine region as early as the beginning of the 18th century, the lake-dwelling phenomenon was not recognised until Ober Meilen discovery in 1854 (Coles and Coles 1989). Ferdinand Keller who, basing his assumptions on exotic ethnologies, argued that the lacustrine villages stood on stilts above the water promptly studied the site. Keller's theory was first contested only about seventy years later, abandoned in the 1950s, and eventually compromised with other theories in the 1980s. With the Pfahlbauproblem another important issue in lake-dwelling studies was the development of scientific techniques of excavation and analysis which started in the 1920s, improved in the 1950s with the discovery of 14C dating method and boomed in the 1970-1980s with the application of dendrochronology. To complete the improvement in understanding prehistoric lacustrine sites, a crucial role is being played by underwater archaeology, which has made possible to excavate those sites that, because of the rise of the lake level, are presently situated off shore below the lake water surface.
Fig. 2.2.1 Portrait of Ferdinand Keller (after Speck 1981: 101)
Generally speaking, the future of lacustrine settlements studies looks fairly positive even though it should be noted that the stage of development of wetland archaeology is not homogeneous in all of the six Alpine countries. This disparity might cause problems at the level of a comparative analysis of different areas within the whole Alpine region, if it is not explicitly recognised.
Although interest in the lake-dwellings expanded to the whole Alpine region and even further, the highest number of discoveries were made in Switzerland. People from all possible professional backgrounds were involved in the search for any kind of prehistoric lacustrine deposits. Being familiar with the lake environment, fishermen, for instance, were of great help in locating new sites. Likewise, famous antiquarians such as Colonel Schwab made a significant contribution by collecting lakedwelling material, which subsequently filled various museums around the entire Helvetia Confederation and even beyond its borders. Finally, "scientific" studies carried out by scholars such as Keller, Pallmann, Desor and many others shed some light on the understanding of these prehistoric populations. Within a few years after the first discovery at Ober Meilen, various other sites had been found on almost every Swiss lake. Around Lake Zurich, for example, one could count tens of lacustrine settlements, although, as Keller (1854,1858) states in his reports, not all of them could have been classified as lake-
2.2 The 19th• century: the beginning of the lakedwelling phenomenon
In the winter of 1853-54, due to a long drought and extreme cold, Lake Zurich receded to its lowest recorded level exposing the remains of a prehistoric settlement. The discovery was made by a schoolteacher from Ober Meilen named Aepply, who promptly reported the fact to the Antiquarian Association at Zurich. The Ober Meilen prehistoric lacustrine dwelling was soon studied by the Swiss naturalist Ferdinand Keller (Fig. 2.2.1) who by the end of 1854 published a detailed site report (Keller 1854). It was indeed with this publication that interest in the lake-dwellings began. The quest for lake-settlements 11
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
dwellings. Keller includes only the most relevant settlements, which yielded a sufficient variety of artefacts for the scholars to be able to classify them typologically and, in some cases, to establish the approximate archaeological period they belonged to.
large number of lacustrine settlements was found and soon after studied. On Lake Morat, for instance, we have the sites of Sugiez, Guevaux, Greng and Vallamond. Lake Bienne yielded those of Sutz, Lattrigen, Morigen and Nidau Steinberg. The biggest of the Swiss Jura lakes, Lake Neuchatel, has the highest number of prehistoric lake-dwellings. Some of the most relevant ones are the Bronze Age stations of Cortaillod-Les-Essert, Font, Hauterive, Concise, Corcellettes and Auvernier-Nord (Desor 1865). Numerous lacustrine settlements were also found on Lake Geneva (Lac Leman), the largest of the northwestern Alpine region lakes. As it is the case of Lake Constance, also Lake Geneva is shared by two countries. The northern part belongs to Switzerland and the southern to France.
In addition to the famous site of Ober Meilen, other well known lake-dwellings discovered and studied around Lake Zurich in the second half of the 19th century include those of Kleiner and Grosser Hafner, discovered in the 1850s, Mannendorf which was already known in the 1840s, but not investigated until later, Latten and Haumessergrund found in the 1860s and the Neolithic settlement of -0-tikon discovered in the 1880s (Mumo 1890). Lake-dwellings in eastern Switzerland were not only discovered on Lake Zurich. We have in fact the lacustrine settlements of Hochdorf-Baldegg on Lake Baldegg, Riedikon on Lake Greifen and the Neolithic/Early Bronze Age sites of Mariazell, Kleine Inseln and Schenken on Lake Sempach. There are also three sites on Lake Zug namely Derschbach, Koller and St. Andreas (Keller 1866) and fmally, the famous settlements of Riedbiihl and Robenhausen on Lake Pfiiffikon thoroughly studied by the landowner H. Messikommer (1868).
On the French shore lake-dwellings were discovered at Messery, Pointe de la Bise, Thonon and Evian, whereas on the Swiss part of the lake stations were found at Morges, Bellevue, Geneva, Nyon, St. Prex, Cully and many others. Some 26 settlements were already identified as early as the beginning of the 1860s (Coles and Coles 1989). As mentioned earlier, the lake-dwelling phenomenon spread rapidly around the entire Alpine region. The first official discovery of a prehistoric pile-dwelling in Germany, as Lisch (1865) reports, was made at Wismar in the Mecklenburg region in 1863. Following this important discovery and those on Lake Constance, other German lakes and peat moors were explored in the quest for lacustrine dwellings. The search proved to be successful in the Bussen Marsh, at Riedschachen, Aichbilhl and Schussemied in the Federsee peat moor (Lake Feder), at Mindli Marsh, on the shores of Lake Milnchenbuch, on Lake Olzreuth and fmally on the Island Der Rosen on Lake Starnberg also known as Wilrmsee from which the last glacial (Wurm) takes its name.
Northeastward of Lake Zurich is the second largest lake of the northern Alpine region, Lake Constance (Bodensee ), which for ahnost its entire extension divides Switzerland from Germany. As with other Alpine, it also has prehistoric lake-dwellings dating from the Neolithic to the late Bronze Age. Amongst those, which were discovered in the 19th century, there are important settlements such as Arbon-Bleiche, Altnau, Milnsterlingen and Kreuzlingen, on the Swiss shores and those of Haltnau, Fischbach, Hagenau-Burg and Manzell on the German part. The narrow and entirely German northwest part of Lake Constance, sometimes called Lake Uberlingen, is also dotted with prehistoric lake-dwellings. Some of the best known are: Sipplingen, BodmanSchachen, Unteruhldingen and Wallhausen-Liltzelstette. More lacustrine settlements are to be found on the Untersee, a lake section that connected to the Lake Constance only by the river, which flows through the city of Constance. Some examples of lake-dwellings on the Untersee are those of Oberstaad, Feldbach and Werd (Little Island) on the Swiss shore and Reichenau, Langerein, Gundolzen and Markelfmgen on the German side (Keller 1866; Pallmann 1866; Munro 1890).
Austrian lakes also yielded some lacustrine settlements, but the lake-dwelling phenomenon was not as marked as elsewhere in the Alps. The first discoveries were made on Lake Worth in the Carinthia and at Ohnilss on Lake Gasrohrer in the 1850s. Two decades later, a few lakedwellings were also found at See and Scharfmgen on Lake Mond, at Traunkirchen on Lake Traun and lastly at the peat moor ofNeusiedl not far from Lake Neusiedl, the largest lake in Austria (Mumo 1890). Influenced by the remarkable finds at the Swiss lakes, the search for pile-dwellings also extended to France and in particular on the western part of the Alps namely Savoie and Haute Savoie. Bronze Age lacustrine settlements were discovered at Le Pieure, Annecy and Roselet on Lake Annecy. In Haute Savoie, Lake Bourget also yielded lacustrine dwellings such as those of Chatillon, Gresine, Meimart and Les Fiollets (De Mortillet 1867). Other settlements of great importance were found on Lake Clairvaux, Lake Luissel and Lake Aiguebelette.
The search for prehistoric pile-dwellings also took place in the western part of Switzerland. In the region of the Swiss Jura lakes of Bienne, Morat and Neuchatel a large number of settlements was discovered. The non-stop quest was facilitated by the correction of the three lakes' levels in order to prevent them from flooding in a vast surrounding area called "Great Moss" in 1868 (Schwab 1973). As the levels of the lakes declined, lakebeds were exposed together with pile structures and artefacts. A
12
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
A few years after the discovery of the Ober Meilen station on Lake Zurich scholars such as Keller, Desor and De Mortillet suggested that the lake-dwelling phenomenon was not restricted to the northern Alpine region. The numerous Prealpine Italian lakes might also hold similar lacustrine settlements. The quest for pile-structures around the lakes of the southern part of the Alps began in the early 1860s and indeed a first site was promptly identified by Gastaldi (1860, 1861) at Mercurago in the Novara province. During a visit to Lombardy in 1863, De Mortillet and Desor, together with Stoppani, explored the shores and surroundings of Lake Varese quickly identifying various lake-dwellings immediately (Stoppani 1864). Some of the stations they discovered, namely Bodio and Bardello, were even named after them in recognition of their contribution. In addition to the sites on Lake Varese, more lacustrine settlements were discovered on other lakes, peat moors and marshes throughout the entire southern Alpine region. Stations were found on Lake Garda namely Isola Cecchi, Pacengo and Peschiera. Or again on the numerous morainic lakes situated a few kilometres from the southern shores of Lake Garda. Here the discovery of pile-dwellings such as those of Bande di Cavriana, Cattaragna and Polada helped to spread the wetland site "fever" to the whole of northern Italy and, by the turn of the century (almost 50 years after Ober Meilen), the lake-dwelling phenomenon had become of great interest also in the southern slopes of the Alps.
were being discovered and some guesswork interpretations of those prehistoric populations' way of living started being formulated. Scientific research had just begun. 2.3 Lake-dwelling studies in the first half of the 20th • century There were no major changes in research on lakedwellings during the first half of the 20th century. New sites came to light, but systematic research using scientific methods was yet to form part of archaeological research. As in the 19th century, so also in this period, that after a site was discovered it was recorded and, if saved from looters, the finds were sent to museums or private collections. In Switzerland the main discoveries of this period were those of Staflingen and Retschwil-Seezopf on Lake Baldegg, Arbon- Bleiche (Lake Constance) MaurSchifflande (Lake Greifen) Seengen (Lake Hallwil) Pfaffikon-Burg (Lake Pfaffikon) ZH-Utoquai, ZHSeehofenstrasse, ZH-Alpenquai, ZH-Feienbach and Horgen-Scheller on Lake Zurich and Herwil-Inseli, Galgen and Sennweid on the shores of Lake Zug. In the western part of the country the lake-dwelling discoveries were more limited; only Lake Neuchatel yielded some sites, with the most important ones being those of Chabrey-Montbec and Cornaux-Les-Sauges. In Austria and France new lake-village discoveries in this period were basically non-existent, only in the eastern part of France were some fmds made on Lake Chalain in 1903 and at Charavines on Lake Paladru in the 1920s. In northern Italy new sites came to light at Colderio peat moor in Piedmont (Castelfranco 1905, 1906), Molino Casarotto (Lake Fimon), Cisano and Lavagnone on Lake Garda and above all the Bronze Age site of Molino di Ledro on Lake Ledro (Battaglia 1943). As far as Germany was concerned, a fruitful area of discovery was the peat moor of Lake Feder in the Oberschwaben region. Sites such as Dullenried, Wasserburg-Buchau and Siedlung-Forschner were unearthed and it was indeed in these localities that Reinerth (1937) and Schmidt (1937) started to argue Keller's theory of houses standing on stilts above the water. This was in fact the most debated issue concerning the lake-dwellings. Archaeological research on lacustrine sites began being based on systematic and scientific grounds and, stratigraphy became an essential part of the excavation, which started to be a multidisciplinary study involving geology, sedimentology, botany and archaeozoology.
The last part of the Alpine foreland to be influenced by the search for lacustrine villages was the present state of Slovenia. The reason for the initial lack of interest was the absence of any major lakes in the area. In fact, the most important concentration of wetland settlements was found in a swampy area known as the Ljubljana marsh. Although this vast marshy area, not far from Ljubljana, had already been explored in the 1850s, relevant discoveries did not occur until the 1870s when the German scholar Deschmann (1875) thoroughly excavated the Isca area. Deschmann's work was the first, and remained the most significant, for Slovenian lakedwelling research of the 19th century. Although scientific research was limited to a simple typological classification of the material culture, and thus not enough to provide an adequate picture of these lakesettlements, the charismatic figure of Ferdinand Keller created a "false" image of them. Basing his argument upon ethnographic material collected by travelers and explorers, Keller (1854) stated that the prehistoric houses around the lakes had stood on stilts above the water surface. It was not until the 1920s, with the help of scholars such as Reinerth (1922) and Vouga (1928), that the romantic idea of those lacustrine settlements standing on piles on the lake surface started to be questioned.
Using all these techniques Reinerth (1937) was able to prove that various so-called lake-dwellings were actually built directly on the soft ground of the shore and not on piles above the water level. Similar studies were also carried out by Bosch (1939) at the site of HochdorfBaldegg, contributing in that way to sustain Reinerth and Schmidt's argument. In the meantime other lacustrine
However, the 19th century ended with great interest still being shown in the lake-dwelling phenomenon. New sites
13
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
sites scattered around the northern Alpine region started being thoroughly investigated. Some examples are those of Sipplingen on Lake Constance in Germany (Reinerth 1932), Egolzwil in the Wauwil peat moor, Zug-Sumpf Lake Zug (Speck 1928) and Auvernier-Nord on Lake Neuchatel in Switzerland.
studies were concerned. It was during this time that the second Juragewiisserkorrektion (lake level regulation) took place bringing to light new sites such as Twann 2 (Lake Bienne), Morat (Lake Morat) and Autavaux, Chable-Perron, Gletterens, Cheyres and Port Alban 1-2 on Lake Neuchatel (Schwab 1973). There were also new discoveries in the north-eastern part of Switzerland where traces of lacustrine settlements came to light at Egolzwil 5-6 in the Wauwil peat moor, at Rlischikon-Rorly and Horgen-Dampfschiffsteg on Lake Zurich and at Oberrisch-Alther (Lake Zug).
Apart from the two above-mentioned areas of Switzerland and the Federsee in Germany, the lake-dwelling research in the remaining parts of the northern Alpine region was fairly neglected in the first part of the 20th century. A similar situation occured in the southern slopes of the Alps where lake-dwelling studies were not especially active in that period. In Italy for instance, the only reasonably thorough excavation was carried out by Battaglia (1943) at the Bronze Age settlement of Molina di Ledro in the late 1930s. A few more settlements were studied in Slovenia; the site of Notranje Gorice was excavated by Schmid (1910) at the beginning of the century and Blatna Brezovic Neolithic lacustrine village by Jesse in the early 1940s. With the outbreak of the Second World War, archaeological research on lakedwellings, as well as various other academic disciplines, was ahnost totally interrupted until the beginning of the 1950s.
Apart from scattered findings on Lake Annecy (France) and Lake Garda (Italy), new fmdings of lake-dwellings were not particularly prolific in other parts of the Alps during the 1950s and the 1960s. An exception was the Ljubljana marsh, where old as well as recent discoveries started to be systematically excavated. A good example is the work carried out by Korosec (1963) at Kanmik pri Krimom and Blatna Brezovica in the late 1950s. This two-decade-period (1950s-1960s) was very important for lacustrine-settlements research: more and more sites such as those in the Zurich bay (Lake Zurich) and some of Lake Constance began to be analysed using advanced scientific methods. The use of 14C, for example, became an essential part of the site dating process and along with meticulous typological classifications contributed to create the first reliable chronological sequence for lakedwellings.
2.4 Second half of the 20th century: two decades of lake-dwelling research 1950-1970
After about a decade of interruption the studies of the Alpine lake-villages restarted with the Pfahlbauproblem still unsolved. The question of whether the lake-dwellings were built above the water or on the shore was still of great interest amongst wetland researchers. The centenary of the first lake-settlement discovery was approaching and an answer was desired. A conclusion was reached with an important collection of papers called Das Pfahlbauproblem published by W.U. Guyan in collaboration with H. Levi, J. Speck, H. Tauber, J. Troels-Smith, E. Vogt and M. Welten right after the jubilee was celebrated in 1954. All these scholars produced scientific evidence to show that a large number of lake-dwellings were actually built on the soft ground near the lakeshore, and not above the water (Guyan et al. 1955). So, ironically, the centenary of the pile-villages was celebrated by denying their existence.
2.4.1 Lake-dwellings vs non-lake-settlements: research development
new
The second half of the 20th century began with a new perspective in the lake-dwelling research. The Pfahlbauproblem was partially solved and new chronological charts started to take form revealing major gaps in occupation on the Alpine foreland lakes. Scholars began to wonder where the lake-dwellers went during those occupational hiatus and the research was also turned towards non-lacustrine sites. Terrestrial settlements in the Alpine region were known well before the lake-dwelling phenomenon, but after Ober Meilen (the first official lake-dwelling discovery) the studies of dry land sites became of secondary interest and, in particular in the northern paits of the Alps, the efforts concentrated on the more prolific wetland settlements. The reason was mainly due to the state of preservation: terrestrial settlements were not very well preserved and the chance to find them was much lower than that of the wetland sites.
The lake-dwelling research continued with new discoveries being made and new questions raised. An important fmding, which, once again, revolutionised the Pfahlbauproblem, occurred at Fiave on Lake Carera peat moor (northern Italy) in the late 1960s. During the first excavation campaign, Perini (1971, 1975) came across a large number of extremely long piles (3-4 metres), which, as the research proceeded, proved to be the support of dwellings, which were probably situated above the water.
New road development which took place after the Second World War brought about the discovery of a fairly large number of terrestrial settlements which, thanks to the advent of the 14C dating method, could now be given a more accurate calendric date within the chronological chart. A major change in research orientation occurred in
In northwestern Switzerland, the 1960s was a particularly prolific period as far as lake-dwelling discoveries and
14
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
the notihern Alpine region and in particular in Switzerland and southern Germany where the lacustrine studies were more influential (Rychner 1998). The archaeological period upon which was put more emphasis was the Early and the Middle Bronze Age (23 rd -12th centuries BC), which, up until the 1980s, was believed to have been the longest lacustrine occupational hiatus of the entire lake-dwelling phenomenon. Important EBA terrestrial sites such as those of Spiez (BE) (Hafner 1995, 1998) and Bavois (VD) (Voruz 1989) in Switzerland, the cemetery of Singen am Hohenwil (Germany) and the tomb of Le Petit Chasseur 1 (VS, CH) (Gallay 1986) were discovered and excavated in the 1950s and 1960s. Because of this boom in land settlement studies, also long-time-known Bronze Age terrestrial sites such as Wittnau (AG) (Bersu 1945), Sonterswil (TG) and Oberriet (SG) (Frei 1954/55), which were discovered in the first half of the century, were re-excavated more systematically in the 1960s and 1970s (Hochuli 1998b).
dating technique had already been in use in some parts of the United States in the 1930s, it was not until the 1960s that this method also started being successfully applied to wetland archaeology in Europe. In fact, a fairly reliable Neolithic and Bronze Age lake-dwelling chronology based on oak was created in the northern Alpine region about two decades later (Fletcher 1978). The co-operation of various tree-ring laboratories such as those of Besanc;on (France), Hemmenhofen (Germany) and Zurich and Neuchatel in Switzerland has helped archaeologists not only concerning the chronological aspect of the lakevillages occupational patterns, but also to draw a detailed picture of climatic as well as local environmental conditions during the last few millennia BC (Billamboz 1997). In the southern part of the Alps, namely northern Italy, on the other hand, dendrochronological research was still in an embryonic state in the 1980s. It is only recently that, thanl(s to pioneer work done by Aspes (1982) and Fasani (1980) around Lake Garda, some treering dates have started to be produced.
As the lake-dwelling research continued, the EBA-MBA hiatus (23 rd - lih centuries BC) became shorter and shorter and it was limited (and still is today) to the Middle Bronze Age (15th -li 11 centuries BC). As a result, the terrestrial settlements research became more focused on those sites, which were situated reasonably near the lakes and, chronologically included between the 15th and the 12th century BC. The reason for this new development in terrestrial sites studies was to see whether there was a relation between those settlements and the former EBA lacustrine villages. In other words, to notice if the land settlements in the vicinity of the lakes were built by former lake-dwellers who were abandoning the lakeshores, or they always existed as typical terrestrial settlements. The answer was not to be found until the 1990s, when, thanks to new discoveries at Bodman-Breite (D) (Schlichtherle 1995c), the various MBA sites in area near Kreuzlingen (Rigert 1998, 1999), Fallanden (ZH) (Bauer 1992; Fischer 1997) and the number of sites around Lake Zug (Hochuli 1995; Gnepf et al. 1996) in Switzerland, it was possible to notice that these land settlements had lacustrine origins ( see also Chapter 11).
The increasing number of archaeological projects and the advent of underwater archaeology facilitated a more thorough exploration of "new and old" lacustrine sites throughout the 1980s. In southern Germany for instance, sites such as Sipplingen (Kolb, 1995), Bodman-Schachen 1 (Koninger 1995a), Unteruhldingen (Schobel 1995a) and Insel Mainau (Koninger and Schlichtherle 1995) on Lake Constance and Roseninsel (Schmid 1995) on Lake Starnberg were excavated using diving equipment provided by the Bodensee-Oberschwaben project (PBO) in collaboration with the Landesdenkmalamtes BadenWurttemberg. In Switzerland and Germany, underwater archaeological activity started in the 1950s and 1960s (Fig. 2.5.1), but it became an essential part of the lakedwelling research only in the 1980s. As a result, various and well-known underwater sites such as ZR-Kleiner Hafner on Lake Zurich (Fig. 2.5.2) and Boschen (Lake Greifen) were finally investigated (Ruoff 1997). A similar situation occured in Austria where, thanl(s to a sports diving group called "UTC Wels", many submerged Neolithic and Bronze Age lacustrine villages within the Salzkammergut region (Lake Att, Lake Mond and Lake Traun) were and still are recorded (Ruttkay 1990).
Within two decades (1950-1970) of scientific lacustrine research in the Alpine region, it became clear that the land settlements (especially those situated in the lakes' surroundings) were in some cases crucial for the understanding of the lake-dwelling occupational patterns; the separation between the terrestrial and lacustrine sites archaeology in the Alpine region became less and less evident.
From a lacustrine research point of view, the importance of the studies from the early 1970s to the late 1980s lies in the fact that a large number of relevant sites discovered in the last century could eventually be systematically excavated. The application of advanced digging techniques and scientific methods has helped archaeologists understand better the occupational patterns of these prehistoric settlements. For instance, a site could have been occupied more than once during a long period of time, but the distinction between the layers is not obvious to the naked eye. An apparently single-phase settlement might hide traces of more occupational periods can only be identified with special scientific analyses. Some examples are to be found at Bodman-Schachen
2.5 Scientific research on lake-dwellings in the 1970s and 1980s A crucial point in the archaeological study of lacustrinevillages was achieved by the application of dendrochronology at the beginning of the 1970s. Although, thanks to A.E. Douglass, the tree ring-based
15
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
(Lake Constance), ZH-Mozartstrasse (Lake Zurich) Auvernier (Lake Neuchatel) and Fiave (Lake Carera).
2.6 Recent developments in lake-settlement research
Apart from a very limited number of new discoveries, the lake-villages research of the past eight years has essentially been based on more multidisciplinary-based excavations of already known wetland sites around the major Alpine and pre-Alpine lakes. Nowadays, a lakedwelling excavation involves a number of scientific analyses from sedimentology to palynology and in some cases even microbiology. Dating techniques have also improved considerably with 14C and dendrochronology being the two most commonly used. In the northern Alpine region, especially northern Switzerland and southern Germany, dendrochronological charts for lacustrine settlements are more extensive covering most of the Neolithic and a fairly large part of the Bronze Age (Billamboz and Martinelli 1996; Billamboz 1997). Unfortunately this is not the case for the southern slopes of the Alps, where tree-ring dating is still at its early stage of development. Consequently, as the dendrochronological sequence is too fragmentary, the extremely limited number of dates has to be compared with the 14C calibration chart through the technique of "wiggle-matching" (Aspes et al. 1997) (see also Chapter 4). An important role in present-day lake-dwelling research is played by underwater archaeology. Diving techniques and methods of excavation have changed enormously since the first attempt carried out by Fore 1, Morlot and Troy on at Morges on Lake Geneva in 1854 (Speck 1981) (Fig. 2.6.1 ). As pointed out earlier, modem underwater research in Alpine and pre-Alpine lakes started at an amateur level in the late 1950s, and it was then thanks to Swiss scholars such as M. Egloff and U. Ruoff, and the Germans H. Schlichtherle J. Koninger and G. Schobel, that it has reached professional levels integrating itself within the archaeological discipline.
Fig. 2.5.1 The first lake-dwelling undenvater survey on Lake Constance in 1954 (after Schabel 1995b: 52)
Fig. 2.5.2 Undenvater excavation at ZR-Kleiner Hafner on Lake Zurich in 1981 (after Speck 1981: 68)
Fig. 2. 6.1 Marlot underwater expedition in 1854 (after Speck 1981: 130)
The 1980s ended in an atmosphere of optimism as far as the lake-dwelling studies were concerned. Twenty years of systematic and scientific research solidified the basis for future work, which was and still is to be improved in the last decade of the 20th century.
Today, there are numerous professional archaeological diving groups in each of the six Alpine countries and, despite the high cost of the diving equipment, underwater
16
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
excavations have substantially increased in the past few years, becoming an essential part of wetland archaeological research. Some examples of recent lakedwelling underwater excavations around the Alpine region are to be found at Sipplingen (Kolb 1995), Bodman-Weiler 1 and Bodman-Schachen 1 (C. Schmid and Koninger 1995; Koninger 1995b), Insel Mainau (Koninger and Schlichtherle 1994, 1995), Unteruhldingen (Schobel 1995a) on Lake Constance (Germany) and at Roseninsel (W. Schmid 1995) on Lake Starnberg in Bayem. Intense lake-settlement underwater research is also being carried out in Austria, in particular on Lake Mond, Att and Traun in the Salzkammergut region where, because the of local hydrological conditions, most of the prehistoric lacustrine sites (even the Neolithic ones) are presently underwater (Offenberger and Ruttkay 1997). Another countiy where diving archaeological activity is increasing is Italy. For example, a number of underwater lake-villages is being explored on Lake Garda and some on Lake Iseo (Fasani 1994; Fasani pers. com. 1997). In conclusion it can be said that the scientific quality of lake-dwelling research has definitely improved within the whole Alpine region in the 1990s. In some countries such as Germany, Austria, Italy and Slovenia lake-villages studies are increasing and in others (Switzerland and France) they have remained steady since the 1980s. A positive aspect of Alpine lacustrine research is the collaboration within the various countries: comparative analyses of different regional areas have become more and more common facilitating the understanding of this fascinating phenomenon as a whole. As a concluding remark for this chapter, it can be said that the lake-dwelling investigation around the Alps is on a fairly solid basis and the positive perspective for future studies is certainly a good sign for improving our comprehension of those prehistoric groups which populated the Alpine and pre-Alpine lakeshores and marshes from the 5th to the 1st millennium BC.
17
Chapter 3 THE LAKE-DWELLING "PROBLEM" 3.1 Introduction
lacustrine origin were also reported on the German shore of Lake Constance, in the Salzkammergut region in Austria and in Peschiera (Lake Garda, Italy) in the southern part of the Alps.
Following F. Keller's work, the first image given to the lacustrine settlements of the Alps was that of wooden constructions standing on piles above the lake water surface. This picture persisted from 19th century to the 1920s when scholars such as Schmidt and Reinerth began to question the validity of the theory. Two decades later, having more scientific evidence available, Paret was able to prove that a large number of the so-called lakedwellings were built on the lakeshore, and their wooden floors lay directly on the ground.
3.3 The Ober Meilen discovery In the winter of 1853-54, due to a long drought and
extreme cold, Lake Zurich regressed to its lowest recorded level, exposing the remain of a prehistoric human settlement at Ober Meilen. The discovery was made by a school teacher named Aepply, who promptly reported the fact to the Antiquarian Association at Zurich. The Ober Meilen prehistoric lacustrine dwelling was soon studied by the Swiss naturalist Ferdinand Keller who, by the end of 1854, published a detailed site report (Keller 1854). It was indeed with this publication that the lakedwelling phenomenon was recognised. The quest for lake-settlements rapidly spread around the entire Alpine region and within a couple of decades, all major lakes were dotted by known prehistoric lacustrine settlements.
Research continued and the "pile-dwellings" theory was eventually abandoned completely in the middle 1950s. The centenary of Keller's lake-dwellings, celebrated in 1954, ironically signaled the end of his theory: lakesettlements became lakeside-settlements and in the following years the argument which had intrigued many scholars for about a century seemed to be resolved. Yet, today, both theories are accepted although in different circumstances. Archaeologists are more and more convinced that the choice of whether to construct houses on piles, or directly on the soft and marshy ground of the lakeshores was a matter of necessity ( environmental conditions) rather than stylistic cultural traditions.
3.4 Ferdinand Keller's dogma In the aftermath of the Ober Meilen discoveries a lake-
dwelling fever resulted in the discovery of a large number of lake-settlements all over the Alps within a few years. Thanks to that, Keller was able to publish seven more reports, and he soon became the authority in lakedwelling research.
3.2 The beginning of the lake-dwelling phenomenon in the Alps Although the notional date for the recognition of the Alpine lake-dwelling phenomenon is set after the fortunate discovery of Ober Meilen in 1854, traces of lacustrine settlements had been found well before. For example, evidence of the now famous lake-dwelling of Nidau on Lake Bienne (Switzerland) was first reported in 1472 and subsequently mentioned by A. Pagan in his writings three centuries later in 1767. Prehistoric wooden piles standing in the shallow lake water near the shore were also noticed by the humanist Vadin around Arbon and Rorschachen (Lake Constance) in the 16th century (Speck 1990a: 9). Bronze weapons such as daggers and spear points were also found on the shore of Lake Luissel in 1791 and around Lake Sempach in 1806. Finally, in his report, Albert Jahn (1865) wrote that the prehistoric lacustrine wooden structures of Moring en (Lake Bienne) had already been noticed in the first half of the 19th century, but never seriously taken into consideration.
An obvious question, which Keller had to answer, was whether those ancient lacustrine villages were built on the ground near the lake or directly above the water using wooden piles as a support. Keller opted for the latter hypothesis, basing his argument upon a main factor: the lake level fluctuation. Comparing the sites of Ober Meilen (Lake Zurich) to that of Nidau (Lake Bienne ), he argued that the prehistoric villages must have stood on piles above the water because neither Lake Zurich nor Lake Bienne was likely to have been influenced by major water level fluctuations. It was soon realised that according to Keller's statement, Lake Bienne must have fluctuated more than two metres for the Neolithic settlement of Nidau to be located in the water those days (Speck 1981: 105). Ignoring this contradiction, Keller diffused the idea that the lacustrine villages stood on piles above the water with a general platform supporting various houses (Fig. 3.4.1). He called the dwellings Pf ahlbauten and the term was promptly translated into the other two main languages around the Alpine region, French and Italian (palafittes and palafitte).
Isolated lake-dwelling findings prior to 1854 were not only restricted to Switzerland. Ancient objects of
18
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
Prehistory and the Heritage Office of Tiibingen (Germany), systematically and scientifically based excavations started to be carried out in the Lake Feder peat deposit. Schmidt and Reinerth continued the pioneering work of Frank (1876) and, with the help of various scientific disciplines such as geology, climatology and botany, they came to the conclusion that Keller's term Packwerkbau was no longer appropriate and it was changed into Moorbau or marsh land houses pointing out a strong possibility for their floors being built directly on the ground (Schmidt 1930). Fig. 3.4.1 Keller's image of the pile-dwellings (after Keller 1854: plate 1)
KELLER'S THEORY
LAKE
The success of Keller's doctrine was not only the result of his charismatic figure (the "father oflake-dwellings"), but it had some, although ephemeral, scientific help. For instance, support came from the geologist Arnold Escher, a prominent scholar involved in the new subject of limnology in Switzerland. Further support was provided by a few ethnological accounts of exotic societies namely the water-dwellers of the Malaysian Archipelago who seemed to have also had some kind of wooden piledwellings constructed right above the water.
I
SHORE
Fig. 3.5.1 Keller's theory about the lake-dwellings
With such a basis and very little scientific evidence, the romantic image spread all over the Alpine region. All prehistoric settlements found in a humid environment were classified as Keller's "lake-dwellings", neglecting some times the evidence, which could have proved the opposite.
A turning point in the lake-dwelling research occurred when Reinerth began to develop a new pile-settlement theory. Contrary to Keller's Pfahlbauten which were supposed to stand on piles in the water all year around (Fig. 3.5.1), Reinerth argued that the immediate surroundings of the houses ( still on wooden stilts) were only seasonally flooded. In other words, the purpose of the raised floors was that of preventing the dwellings from getting wet during the lake's seasonal transgressions (Reinerth 1922) (Fig. 3.5.2). A confirmation ofReinerth's theory came with his excavation at Sipplingen (Lake Constance) in 1929-30. By constructing a wooden caisson he managed to dry out a surface of about 500 m2 and excavate the submerged Neolithic dwelling as it was on dry land (Reinerth 1932).
3.5 The need for a typological differentiation
As lake-dwelling research progressed, new sites were discovered and it soon became evident that the typology of the constructions varied from place to place. In particular, a new type of lake-dwellings in peat deposits and marshes was discovered. Some of the best known wet environments of this kind which were first excavated were those of W auwil, Robenhausen and Niederwil in Switzerland, the entire surroundings of Lake Feder in Germany and Mercurago, Mombello and Bardello near Varese (Italy).
REINERTH'S
House floors came to light for the first time, but because it could not be proved that they were directly built on the ground level, Keller's theory could not as yet be contrasted. Those dwellings with floor evidence were simply treated as new type of wetland constructions and Keller himself named them Packwerkbauten or houses built on wooden floors.
I.AKE
THEORY (19205)
I SHORE
HIGH WATER LEVEL
t
Lacustrine settlement studies continued throughout the whole second half of the 19th century, more and more sites were being discovered and excavated, but it was not until the 1920s that, thanks to the Research Institute of
Fig. 3.5.2 Reinerth 's theory about the lake-dwellings
19
The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
A great result was that Reinerth could state that the village was not built in the water, but on the shore, which was seasonally inundated by the lake fluctuations. The importance of this discovery was even more accentuated by the fact that the settlement lay on the shore of one of the main Alpine lakes. This irrefutable evidence finally proved that lacustrine villages built on semi-dry terrains did not only occur around marshy environments as it was previously thought, but also on the shores of large glacial lakes.
A first sign of thaw on the matter occurred with the excavation ofEgolzwil 3 in the Wauwil peat moor carried out by the Swiss scholar Emil Vogt. Thanks to an accurate excavation Vogt (1951) found out that pieces of tree bark were situated between the anthropogenic stratum and the ancient calcareous lake deposits being placed there as a means of insulation against ground humidity. Moreover, there were traces of land plants growing on the lake deposits before the bark chips were laid which prove a dry environment well before the occupational period.
3.6 From lake-dwellings to lakeside-dwellings
More evidence sustaining Paret's hypothesis came from the Late Bronze Age lacustrine site of Zug-Sumpf on Lake Zug (Switzerland). In addition to the structure of the house, which is typical of a land settlement, Zug-Sumpf stratigraphic analyses show signs of water-covered terrain long before the village was constructed, but not during the period in which the settlement was inhabited (Speck 1955).
The concept of lacustrine villages built on stilts in a permanently wet environment was strongly challenged during the 1920s and 1930s. Reinerth's theory of a semidry setting had gained more and more credibility, but for those in favour of a totally dry building ground these results were not satisfactory. A decisive attack was delivered by Oscar Paret in the early 1940s. In response to Bosch and Vogt's "biased" excavation analyses of Hitzkirch-Seematt and Hochdorf-Baldegg Neolithic and Early Bronze Age lake-dwellings (Lake Baldegg, Switzerland), Paret (1941, 1942) argued that the houses were situated neither in a permanent aquatic environment nor were they periodically washed out by the lake level transgressions, but they were built on a completely dry ground (Fig. 3.6.1). The fact that the stratigraphy showed the presence of lake marl underneath and above the anthropogenic layer was due to lake fluctuations with consequent lake marl deposition well before and after the villages were built.
Faced with such irrefutable evidence, the majority of Swiss researchers accepted Paret's theory and the term "lake-settlements" was finally changed into "lakesidesettlements" in the early 1950s. The 100-year-jubilee of Keller's pile-dwellings was celebrated, ironically, by denying their existence (Vogt 1955). 3. 7 The Pfahlbauproblem is no longer a problem
The beginning of the second half of the 20th century determined an important turning point in the lacustrine settlement research. The advent of 14C and dendrochronology along with other scientifically based archaeological analyses brought a better understanding of the excavated sites. The question of whether the villages stood on piles in the water or were built directly on dry land lost its fascination. Emphasis was now placed on occupational patterns, chronology, subsistence and economy of those prehistoric agglomerates in order to have a better picture of the lake-dwellers' life.
Of course this theory was promptly rejected by the Swiss scholars. Keller-Tarnuzzer (1944, 1945) for example argued that both lake-villages of Breitenloo near Pfyn (Breitenloo peat deposit) and the various dwellings of Arbon-Bleiche (Lake Constance) were defmitely piledwellings. A similar stand was taken by Pinosch (1947) with the Burgaschi-Ost site in the homonymous peat deposit.
The more sites were excavated, the more it became evident that the location of the settlements was mainly related to the environment morphology without necessarily following a specific and characteristic style of constructions (Fig. 3.7.1). In fact, Reinerth's type oflakedwelling were found by Christian Strahm (1973) at Yverdon-Avenue des Sports (Lake Neuchatel) in the 1970s and by Stefan Hochuli (1994) at Arbon-Bleiche (Lake Constance) in the 1980s and 1990s. Similarly, Paret's lake-dwellings directly constructed on dry ground came to light at Egolzwil 5 (Wauwil peat deposit) in Switzerland, at Ehrenstein near Ulm (Germany) and, very recently, at Castellaro Lagusello peat moor near Lake Garda in northern Italy (Aspes 1997). Constructions of this kind were not only found around peat deposits or marshes, but on large lakes. Good examples are the settlements of Auvernier (Lake Neuchatel), Twann (Lake
PARET'S THEORY (19405)
LAKE
l
SHORE
HIGH WATER LEVEL
t
LOW WATER LEVEL
Fig. 3.6.1 Paret's theory about the lake-dwellings
20
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
Bienne), ZH-Mozartstrasse (Lake Zurich) and the numerous villages scattered around Lake Zug (Hochuli 1995: 74).
underwater excavated by Urlich Ruoff in the late 1970s (Ruoff 1981: 49). The last forty years have signed an important stage of the lake-dwelling research. More advanced scientifically based analyses have revealed that there is not necessarily a single style of lacustrine construction in the Alpine region, but they mainly depend upon cultural and environmental factors. In other words, the lake-dwellers adapted themselves to the natural surroundings by choosing the kind of construction, which suited them most.
THE PFAHLBAUPROBLEM LAKE
••------
I
I
SHORE
REINERTH'S THEORY (1920S)
CURRENT INTERPRETATION
I
PARET'S THEORY. (1941)$)
(2000)
Fig. 3. 7.1 The Pfahlbauproblem as it is seen today An important discovery, which defmitely closed the
Pfahlbauproblem discussion, was that made at Fiave (former Lake Carera) in northern Italy in the 1960s. The site excavated by Perini shows all types of lakedwellings; from the classic Keller's pile-settlements to the land-built villages argued by Paret. The lay-out of Fiave lacustrine dwellings consists of three zones and it follows a chronological occupation pattern which goes from the Neolithic to the beginning of the Late Bronze Age. It has to be pointed out though that not all three zones were occupied in the same period. Zone one, for example, was settled from the Late Neolithic to the beginning of the early Bronze Age (Fiave 1 and 2) then abandoned for about three centuries and reoccupied from the Middle Bronze Age (Fiave 6) to the beginning of the Late Bronze Age (Bronzo recente, Fiave 7) about at the end of the 12th century BC. Zone 2, on the other hand, was occupied between the two above-mentioned phases e.g. from Fiave 3 to Fiave 5 (1700 to 1400 BC). Finally, zone 3, situated on the little hill south of the first two zones, was occupied in the last phase of Fiave lake-settlement (Fiave 7, ca 13th and 12th centuries BC) (Perini 1984, 1987). If one analyses the above mentioned chronology along with the various typological divisions of houses, it can easily be understood how the building structure was strictly related to environmental demands. In fact, while Neolithic buildings were situated on dry land, demographic increase, which occurred towards the 15th century BC, forced Fiave lake-dwellers to expand their villages encroaching to the lake. This water wards movement is clearly visible in zone 1 and 2 where Middle Bronze Age houses were built on stilts in a permanent water environment. Two similar examples are also to be found in the northern part of the Alps, namely F eldmeilen-V orderfeld (Lake Zurich) and the site of Storen-Wildsberg on Lake Greifen
21
Chapter 4 DATING THE ALPINE LAKE-DWELLINGS that the various artefacts showed different characteristics, which, although suitable for either one of the two abovementioned periods, pointed out different cultural aspects within the archaeological period itself. In other words, the periods themselves (Stone Age and Bronze Age) could have been subdivided into a number of distinct cultural groups. This "problem" urged an immediate solution, which was provided by typological sequences. Following the pioneering work of the Swedish scholar Oscar Montelieus, the lacustrine material culture started to be arranged in sequences based on various styles and designed in relation to the place of origin (Montelieus 1903). The technique was obviously challenged by many researchers and eventually refined by the German prehistorian Paul Reinecke (1911-1965) whose system for the European Bronze Age is still in use (see also Chapter 5).
4.1 Introduction
Dating methods within the Alpine lake-dwelling research have undergone a remarkable development since the mid19th century; from chronologies purely based on relative dating to sophisticated chronometric techniques developed recently. Because of their specific environmental context, and the large number of relative and absolute dating techniques, only a few are particularly suitable for placing the prehistoric lacustrine settlements into a fairly accurate chronological order. Relative dating, essentially based on stratigraphy and pottery typology, is the oldest and still the most common dating method within the lake-settlement studies in the Alps. The second half of the 20th century was characterised by the development of various absolute dating techniques with 14C being the most widely used. The radiocarbon revolution (Renfrew 1973) along with the advent of dendrochronology has changed most of the old lake-dwelling chronologies. Tree-ring dating not only allows archaeologists to differentiate clearly between the various occupational phases, but it is even possible to separate the phases of construction within a single occupation. With the invention of thermoluminescence in the 1960s, researchers tried to use it for dating the large amount of pottery found on Alpine lacustrine sites in the previous century of archaeological studies. Unfortunately, due to lack of precision, the employment of thermoluminescence in accurate dating processes was soon abandoned and it is now used only in museum studies to prove the authenticity of pottery. A great help for the Alpine region lacustrine sites chronologies comes from climatically determined frameworks such as pollen sequences and glacio-lacustrine varves. In addition to offering a support for other chronometric dates such as dendrochronology, these dating techniques allow archaeologists to reconstruct environmental chronologies ofvegetational and climatic variations. A final and crucial aspect of all the above-mentioned dating techniques is that they can be correlated and compared to one another giving scholars the opportunity of improving accuracy and reliability of chronologies of the Alpine lakedwelling dating process.
Thanl(s to the high number of artefacts collected from the increasing number of lake-dwelling discoveries, the process of typological classification and seriation proved to be quite successful throughout the whole Alpine region. A shortcoming which archaeologists were concerned about was the lack of stratigraphy within the archaeological records. In fact, most of the prehistoric lacustrine objects were collected from different sites without recording any stratigraphic sequence. What made it even worse was the fact that some artefacts of different assemblages were not infrequently mixed together, thus causing confusion. Fairly accurate analyses started towards the end of the 19th century and, by the tum of the century (20th century), a few sites such as those of Haumessergrund (Lake Zurich), Riedikon (Lake Greifen) (Ruoff 1980), the station of Hochdorf-Baldegg on Lake Baldegg (Wey 1990b), the famous settlements of Riedbiihl and Robenhausen on Lake Pfafiikon (Eberschweiler 1990a), Sipplingen, and Unteruhldingen (Lake Constance) (Schobel 1996), Sugiez (Lake Morat), Sutz, Lattrigen and Morigen on Lake Bienne, CortaillodLes-Essert, Hauterive, Concise and Auvemier-Nord (Lake Neuchatel) (Egger and Gassmann 1985) and finally, in the southern part of the Alps those of Bodio and Bardello (Lake Varese) (Banchieri 1991) and Bande di Cavriana, Cattaragna and Polada near Lake Garda (Aspes 1997) had already produced some fairly reliable typological classifications.
4.2 Relative dating: stratigraphy and pottery typology
Relative dating was the only method archaeologists had available to place the Alpine lake-dwellings in a chronological order after the first discovery in the 19th century. The Alpine lacustrine villages were initially divided into two archaeological periods according to the content of their material culture: assemblages which consisted essentially of stone artefacts were classified as belonging to the Stone Age and those with bronze objects were placed within the Bronze Age. It was soon noticed
Because of the extremely large quantity of pottery found on all lacustrine sites, typological sequences based on ceramics became very popular and they began to be divided into regional groups named after the places where the assemblages were found. As the research progressed, new sites were being discovered and a fairly clear picture of the various lacustrine cultures in the Alps started to be drawn. 22
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
Hafuer and ZH-Mozartstrasse (Lake Zurich) that were occupied in six distinct phases from the Neolithic to the Late Bronze Age. Thanks to the clear distinction between anthropogenic layers, it was fairly easy to separate the various cultural groups and their characteristic pottery styles placing them in the correct chronological order.
At least 24 different cultures were, and still are, thought to have occupied the Alpine lakeshores in the Neolithic. Out of the total number, 17 belong to the northern Alpine region, and seven are located in the south and southeastern parts of the Alps (see Table 1.6.1). The Bronze Age groups namely the Rhone culture in France and western Switzerland, the Arbon and Bodman cultures (eastern Switzerland and southern Germany), the Polada, the Fiave and a few others (northern Italy) were, on the other hand, fewer and, to a certain extent, more regionalised. An exception was the Urnfield culture, which covered the whole northwestern pre-Alpine region in the Late Bronze Age (Strahm 1997).
Typological classifications have become a standard part of archaeological analyses and, dating. Some countries have more firm traditions than others within archaeological analyses (see also Chapter 7) and despite the numerous chronometric methods available nowadays, they still rely on the relative dating technique of typology. Switzerland and Germany for instance, the two most advanced countries in the Alpine region in terms of lakedwelling absolute dating methods, still place an enormous emphasis on pottery typology which is compulsorily included in a very detailed form in all of the published and unpublished archaeological reports. Finally, the importance of typological classifications, especially of pottery, is also confirmed by the fact that all the studied lacustrine villages have been cross-dated using relative dating techniques, in order to integrate the absolute dates for cultural groups for the production of chronological charts ( see the lake-dwelling database in the appendix and also Fig. 4.2.1 and Fig. 4.2.2).
Despite the fairly reliable pottery classifications, there was still a problem, which haunted archaeologists before the advent of 14C and that was the diachronic position of the single cultural groups within chronological sequences. Classifying pottery styles by their aesthetical appearance was not sufficient to place them in a correct chronological order. In fact, as the example of Arbon-Bleiche shows, two quite different types of pottery could be contemporary (Hochuli 1994). This is obviously difficult to detect without absolute dating techniques if the stratigraphic analysis is not reliable. A partial solution was offered by undisturbed sequences found on sites with multi-occupational phases such as those of Kleiner-
Lake-dwelling dating methods in the Alps
550 500 450 400 350 300 250 200 150 100 50 0 1,C
p
D
Ty
Tot.# Sites
Fig. 4.2.1 Number of Lake-dwelling sites in the Alps dated and analysed using the four principal scientific techniques {14C, D: Dendro, P: Pollen Analysis, Ty: Typology)
23
The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
Lake-dwellings in the Alps: percentages of sites dated with the indicated methods
Series 1 Percentage of sites dated vvith the indicated dating technique
■ series
1
■ seies2
14C
D
Series 2 Percentage of sites non-dated with the mdicated dating metllod
Ty
Fig. 4.2.2 Percentages of Lake-dwelling sites in the Alps dated and analysed using the three principal methods Dendro, Analysis, Ty: Typology)
4.3 The
14 C
f 4C, D:
unclear at the beginning. In some cases such as for example the distinction between the Pfyn and Horgen cultures in Switzerland, the radiocarbon method was sufficiently precise because these two cultures covered a large time span, which was not longer than the limits of resolution of the SD of radiocarbondates. Complications started when cultures such as the Ltischerz (western Switzerland), had to be distinguished from the W estem Horgen culture and the Saone-Rhone group. In fact, the Ltischerz culture occupied a fairly short period between the two latter cultures and the 14C method was unfortunately not able to defme it. It was, in other words, difficult to determine whether the Ltischerz group had a specific chronological position between the Horgen and the Saone-Rhone cultures or it was contemporaneous to one of them. Improvement occurred with the advent of dendrochronology (Becker et al. 1989; Becker 1993) (see also sub-chapter 4.4). This very precise dating method provided the tool to calibrate 14C dates proving that Hessel de Vries' (1958) suggestion, that radiocarbon years and solar years should not be assumed to be equivalent, was actually right. The "second radiocarbon revolution" as pointed out by Colin Renfrew (1973) had begun. The 14C method became much more reliable even though the standard deviation remained too high to be able to distinguish short multiple occupations within a single cultural group. However, the dating process of the Alpine lake-dwellings continued and the various Neolithic and Bronze Age lacustrine groups (see subchapter 4.2) started to have their calender years chronological place within the prehistory of the Alpine region.
dating method
As we have seen so far, dating the Alpine lake-dwellings was essentially based on relative methods such as stratigraphy and typological sequences until the end of the first half of the 20th century. With the advent of radiocarbon dating (14C) invented by the American chemist Willard Libby in the 1940s, the lacustrine sites could then be given a chronological position in calender years within European prehistory. It is now recognised that atmospheric 14C levels have not been constant (though they are very nearly so), but require calibration by comparing with a known (dendro) sequence. As it happened for the rest of Europe, also in the Alps and in particular the numerous lacustrine sites, the radiocarbon revolution brought forward a lot of controversies. A large number of old chronological sequences did not coincide with those of 14C and, as a result, some archaeologists, particularly the older generation, began to question the overall validity of the radiocarbon dating method. However, the rapidly mounting evidence of the general correctness of the technology in a broad outline changed discussions within the professional archaeological community from a question of validity to a question of accuracy of the 14C method values. The Stone Age and Metal Age became conceptually more distinct, but the chronological order of the various regional cultural groups within a single period, remained
24
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
Percentages of 14C dated lake-dwellings in the Alps
Series 1 14C dated
F
Series 2 Other
■ series 1 fflseries2
D
CH
A
SL
Fig. 4.3.1 Percentages of Lake-dwelling sites in the Alps dated using 14 C method
The replacement of relative chronologies by ones based on l 4C dates was a long and, in some countries, slow process. Even in Switzerland, which was the most advanced country in lake-dwelling research, the deeply rooted pottery typology tradition prevailed for a long time and to a certain extent is still very influential. A change started in the 1980s when, thanks to a more intense application of absolute dates, new discoveries and more thorough archaeological analyses, the aspect of the Swiss prehistory chronology, especially that of the lacustrine villages, started to assume a new aspect. A relevant contribution to this chronological revolution was made by systematic analyses of sites such as ZH-Mozartstrasse (Lake Zurich), Arbon-Bleiche 2 (Lake Constance) and Twann (Lake Bienne ). A breakthrough in the lakedwelling research was the beginning of underwater archaeology with the excavations of the well-preserved submerged sites of Erlenbach-Widen, Kleiner Hafner, Meilen-Rorenhaab, Meilen-Schellen (Lake Zurich), Boschen (Lake Greifen) and Egg
Radiocarbon dating is an essential aid to dendrochronology since many sequences (at least initially) were "floating" and could only be tied down by 14C. In fact, even if most lacustrine sites retain large quantities of well preserved wood, some regional dendrochronological sequences cannot be developed. It is in this case that isolated tree-ring dates are wigglematched with 14C dates in order to create what is known as floating sequences (Fasani 1994; Fasani and Martinelli 1996; Aspes et al. 1997) (see also sub-chapter 4.4).
Obere-Gull, Bodman-Schachen 1, Sipplingen and Unteruhldingen on the German shores of Lake Constance (Koninger 1995b, 1997; Ruoff 1997). Old chronological charts of the prehistoric Alpine region based on typological divisions of pottery and other artefacts changed not only in Switzerland and southern Germany, but all across Europe. Sites where, due to lack of dendrochronological sequences, 14C played an important role in determining the Alpine region lake-dwelling absolute chronology were Fiave (northern Italy) (Perini 1987), Ljubljana marsh (Slovenia) (Harej 1986), the lacustrine villages in the Salzkammergut region in Austria (Ruttkay 1990) and initially also the French Jura lakes, in particular Lake Chalain (Petrequin 1988).
At more than four decades from its first application, we have seen how 14C has revolutionised all the established typology-based chronological sequences for lakedwellings in the Alps. It has placed the various cultural groups in a fairly precise chronological order and in some cases even distinguished between multi-occupational phases within a single cultural period. In summary, it can be said that, despite not being as accurate as dendrochronology, thanks to its flexibility of application 14C is and will probably remain the most common absolute dating technique in the Alpine lake-dwelling studies for quite a long time in the future.
Statistical analyses show that, despite the increasing number of dendrochronologically dated sites, 14C still remain the most common and widely used absolute dating method within the Alpine lake-dwelling research. The country with the highest percentage of 14C dated lacustrine sites is Austria (60% ), which is followed, by France (55.2%), Switzerland (52.2%), Italy (39%), Slovenia (36.8%) and finally Germany with only 26.2% (Fig. 4.3.1).
25
The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
that permits the identification in different tree of rings that were formed contemporaneously with one another" (Dean 1986: 133-134). Chronology building is the process of constructing long and absolutely dated ring sequences from many individual cross-dated ring series (Fig. 4.4.1). Chronologies are extended beyond the range of living trees by cross dating in ring series from progressively older archaeological wood. Apart from the example of the California bristlecone pine (Pinus longaeva) sequence which dates back more than 4000 years (the oldest living tree species on earth), all the other long sequences are usually constructed in pieces, in other words, floating sequences are linked to each other, usually starting from living trees. Since isolated sequences are the result of archaeological discoveries, the chances of filling the gaps between them are usually related to further discoveries.
4.4 Dendrochronology: a revolution in precision
Dendrochronology, the science that uses tree rings for dating past events and reconstructing past environmental conditions, has undergone a period of explosive growth within Alpine lake-dwelling research in the last three decades. Firmly grounded in the principle of cross-dating, using aspects of ring morphology to identify contemporaneous rings in different trees, dendrochronology provides absolute dates accurate to the calendar year, and quantitative reconstructions of environmental variations on seasonal to century scale. Archaeological applications of dendrochronology fall into three categories: chronological, behavioural and environmental. Chronological analysis involves both concrete and abstract units of archaeological investigation: in other words, the development of calendar sequences within short or long periods and the theoretical aspect of archaeological dating behind them (Dean 1997). Tree-ring studies also provide a broad spectrum of information on past human behaviour, including the treatment of trees as a natural resource of wood for raw material, sources of timbers, season of wood procurement and finally the working technologies related to different tools. Dendrochronology also offers a fairly reliable environmental information such as abrupt change in climatic conditions, short-term temperature oscillations and vegetational transformation due to natural as well as human influence.
Because of the abundance of well preserved wooden structures found on prehistoric lacustrine sites, archaeologists have been able to construct quite reliable sequences which, as in the case of the northern Alpine region (in particular north-eastern Switzerland and southern Germany), date back to 5100 BC (Wyprachtiger 1990). The first tree-ring analysis on a lacustrine settlement was successfully carried out by Bruno Huber (1941) at Wasserburg-Buchau (Lake Feder, D) in the late 1930s. Dendrochronological studies continued and, in addition to Wasserburg-Buchau, three more lakedwellings had already been partially analysed by the end of the 1960s. The boom of lake-dwelling dendrochronological research came in the 1980s, when relatively long sequences started to be developed. The
There are two basic principles of dendrochronology namely cross-dating and chronology building. The first one is defined as "the characteristic in tree-ring structure
l
ld :
\11I 1~11! I 1111ii111111 !11111111)
111111H 111111111111~~111
l , 10: J\ I
'
I
I
I
l
illl l =, .
I
111 i\1111111111111 iiI ii 111111 I
I
I
I
1/\
•
I
I
-
"\AA M h 9v~oi~g~ AAr1\ IL, Tm•""'w"" n~VQ,v11mm \mm ,Q7'~'l'J ~:I~ I
I
1
I
I
I
14-->1 Overlapping
Overlapping
Fig. 4.4.1 The principle of chronology-building by tree-ring dating
26
•
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
or not the 1i 11 century BC occupation had one or two phases. In fact, tree-ring analyses pointed out very clearly that the occupation definitely had two phases (Hochuli 1994). Finally, at the Neolithic lake-village of Charavines-les-Bagneurs on Lake Paladru (F), dendrochronology was even able to determine in which season the first posts were cut. In fact, thanks to analyses of the soil chemical content in relation to the tree-ring width, archaeologists came to the conclusion that in both the first and the second year of constructions, trees were felled in the non-vegetative season about October or November.
first northern Alpine region oak sequence was constructed by Becker in the early 1980s and, a few years later, some of the most relevant Neolithic and Bronze Age lacustrine villages of Switzerland, southern Germany and France namely Auvernier and Cortaillod (Lake Neuchatel, CH), Chalain (Lake Chalain, F), ZH-Mozartstrasse, Kleiner Hafuer and Meilen-Schellen (Lake Zurich, CH), Siedlung-Forschner (Lake Feder, D) and Boschen (Lake Greifen, CH) could be given precise calendar dates which, in some cases, reached the accuracy of+/- 1 year. Dendrochronology not only revolutionised the greater part of the lacustrine village chronology, but it made it possible to separate the various occupational phases within a single cultural period even without a reliable stratigraphic analyses of the anthropogenic layers. Depending on the presence of the cambium (also called bark ring - the outermost tree ring), archaeologists are even able to distinguish between the different construction phases (if more than one) and tell whether and when the dwellings were being renewed or fixed during occupation. Finally, if the sequence is reliable and limited to a fairly small region, it could also be determined in which season the trees were felled and the construction of the dwellings took place. Good examples of the above-mentioned potentials of dendrochronology are to be found in the entire northern Alpine region with the most popular being Hornstaad-Hornle (Untersee Lake Constance, D) (Billamboz and Becker 1985), ArbonBleiche 2 and Bodman-Schachen 1 (Lake Constance, D), Siedlung-Forschner and Wasserbur-Buchau (Lake Feder, D), ZR-Kleiner Hafuer, ZH-Mozartstrasse, ZH-Kan San (south), ZH-Kan San (north), ZH-Pressehaus and MeilenRorenhaab (Lake Zurich, CH), Boschen (Lake Greifen,CH), Lattringen (Lake Bienne, CH), Auvernier and Cortaillod (Lake Neuchatel, CH), Charavines-les Bagneurs (Lake Paladru, F) and finally the numerous settlements on Lake Chalain (F). Amongst these lacustrine sites there are four in particular which reflect the benefits obtained from dendrochronological analyses. The first is ZH-Mozartstrasse whose numerous and closely-related phases of occupation could eventually be separated and given precise calendar dates (Gross 1987a, 1987b) ( see also sub-chapter 8.3 .1). The second is Bodman-Schachen 1 when, during the occupational phase 2, two tree-felling and three construction phases were detected within two anthropogenic layers (Koninger 1996a). The third example is the Bronze Age lacustrine site of Arbon-Bleiche 2 where dendrochronology eventually solved the various disagreements on whether
As briefly mentioned at the beginning of this sub-chapter, tree-ring sample collections can be used to infer various aspect of past human behaviour such as forest economy, woodland maintenance (cutting of primary or secondary forests), agriculture and even animal husbandry. For example, the use of ehn leaves as cattle fodder in the French Vercos during the Neolithic (a practice still in use in the region today) is clearly reflected in the extremely limited width of the tree rings. In some cases, the availability and provenance of wood (primary or secondary forest) even influenced the structure of the houses. In fact, as Billamboz (1997) notices, northern Alpine lake-villages were normally built using secondary forests wood during the last settlement phase before the MBA occupational gap (15th-lih centuries BC). The shift from primary to secondary forest wood was due to the high exploitation of the former during the 17th centuries BC. The great potential of dendrochronology is sometimes limited by natural factors, with the most relevant (at least in the northern Alpine region) being the regionality of its applications. In such a diverse environment as that of the Alps, climate and vegetation vary a lot from place to place according to latitude, altitude and landscape morphology. Since tree rings are highly influenced by all these natural factors, dendrochronological sequences are also limited to a specific regional microclimate. Consequently, if for certain reasons the tree-ring sequence of one of those areas cannot be developed, the archaeological finds of wooden objects of that area cannot be dendrochronologically dated. A possible solution to this problem is the use of inter-regional sequences, whereby a single sequence is applied to two or more areas with similar climatic and environmental characteristics.
27
The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
Percentages of dendrochronologically dated lake-dwellings in the Alps
Series 1 Dendro dated
F
D
Series 2 Other
■ series 1 Clseries2
CH
A
SL
Fig. 4.4.2 Percentages of Lake-dwelling sites in the Alps dated using dendrochronology In case of a plentiful availability of well-preserved wood and a complete absence of any dendrochronological sequence, archaeologists adopt a dating technique called wiggle-matching. This method consists of creating floating tree-ring sequences from radiocarbon dated wooden structures and matching them with the 14C calibration curve. A shortcoming of this technique is that these sequences cannot be related to living trees (usually the beginning of a reliable sequence) and therefore they are much less accurate.
main tree-ring laboratories namely those of Zurich (CH), Besam;:on (F) and Hemmenhofen (D). Verona (I) and Ljubljana (SL) have recently joined them and contributed to the development of new chronological sequences. Dendrochronology is certainly the most accurate dating technique available at the moment and it will not be too presumptious foreseeing it as the most applied dating method of the future lake-dwelling research.
Lack of dendrochronological sequences is a major problem in the southern part of the Alps. In fact, most of the dating process of lacustrine site in the Lake Garda region (I) has been carried out using the above-mentioned wiggle-matching method (Fasani 1994; Fasani and Martinelli 1996; Aspes et al. 1997). The situation is even worse at Fiave (Lake Carera) where, despite the large quantity of well-preserved wood, the site chronology still relies on pottery typology and a few 14C dates.
Although thermoluminescence has been researched for the past two centuries, it was not until the 1960s that its possible application for dating archaeological artefacts started to be developed. A great advantage this technique has over 14C is that it can date the most common inorganic material usually found on archaeological sites: pottery. As it is the case for many other chronometric dating methods, also thermoluminescence is based upon radioactive decay. The alpha, beta and gamma radiations emitted by the radioactive decay bombard the crystalline structure displacing electrons, which remain trapped within minerals. The clock is set to zero any time the material is heated rapidly to 500 °C. In the case of pottery for instance, the clock is reset at the time when the object was fired. The age of the object is obtained by reheating the artefact to the same temperature (> 500 °C), freeing the trapped electrons and calculating the amount of TL emitted. In other words, the intensity of TL measured from the original firing (when the object was made) to the reheating in the laboratory is the age of the object (Lowe and Walter 1997; Rapp and Hill 1998). It is important that the radioactivity from impurities within the sample is distinguished from "external" sources of radiation such as those of the soil in which the sample was found.
4.5 Thermoluminescence vs pottery typology
Dendrochronological analyses on the Alpine lakesettlements have proliferated considerably in the past three decades, but the contrast between the northern and the southern part of the Alps is still very marked. The three leading countries with the highest percentage of dendrochronological dates are France (49.3%), Germany (47.7%) and Switzerland (41.4%) followed in a long distance by Slovenia (26.3%) and Italy (16.9%). A striking exception is Austria, where, despite the fact that it is situated north of the Alpine chain and it has a similar vegetational and climatic environment to that of Switzerland and southern Germany, dendrochronology is basically non-existent (Fig. 4.4.2). The hundreds of dates, which have contributed to transform all the lacustrine chronological aspect, have been produced by the three
28
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
those "non-classifiable" pottery assemblages coming from various unknown lacustrine sites. It was soon realised though, that the accuracy of the method, even in perfect conditions as listed above, was too poor for an accurate chronological division of the various Alpine region lacustrine cultures. In fact, the humid environment, which preserves organic material so well, turned out to have a negative influence on pottery samples preservation for thermoluminescence dating. Another factor that was totally against the method was the fact that a large quantity of the lacustrine pottery was not fired and therefore the "clock" could not be reset when the artefact was made.
One of the biggest shortcomings of thermoluminescence is its limited accuracy. Sources of error in TL dating include systematic errors associated with the calibration of laboratory radiation sources, which means that a precision of better than +/- 5% (with 1 sigma = 68% probability) is very unlikely to be achieved. Furthermore, the place and state of pottery preservation play a crucial role in the date level of accuracy. Disequilibrium in the uranium decay chain, migration of radioelements through surface sediments and estimation of past water content of the deposits all pose problems for the establishment of the date. Because gamma rays coming from the buried soil provide part of the radiation dose-rate, it is important that the object is completely buried and no other radioactive element is near it. For instance, a piece of pottery found on the surface is a bad sample because it is influenced by other "external" sources of radiation and light. The same situation occurs if a lump of rock, with different level of radioactivity to that of the filling, is situated near the object to be dated. The ideal situations when the pottery is homogeneously surrounded by at least 0.3 m. of soil with the same radioactivity and the object is examined (with the surrounding soil) as soon as it is extracted in the course of excavation (Fig. 4.5.1) (Aitken 1997).
With the beginning of systematic and multi-disciplinary based archaeological excavations, thermoluminescence became less and less popular within the Alpine lakedwelling research. The simple fact that organic materials were always extremely abundant on evety lacustrine site and that even the least precise 14C date was more accurate than those obtained from thermoluminescence, made archaeologists reluctant to use this dating technique. The decisive factor that made thermoluminescence completely redundant was the development of dendrochronology in the 1980s when, thanks to the development of reliable tree-ring sequences, the majority of pottery assemblages belonging to the various lacustrine cultures were accurately cross-dated and given fairly precise positions within the Alpine lake-settlement chronology.
BAD
---v---
__11).
Present Surface
-----
POOR
Today, despite being no longer part of the lacustrine villages archaeological dating process, thermoluminescence has become an important tool for museums and private antiquity collectors. It is in fact employed to detect whether the pottery is original or just an imitation.
Old Surface
GOOD
As we have seen the "competition" thermoluminescence vs pottery typology has never been significant within the Alpine lacustrine settlement studies; accuracy, availability of other dating methods and also timing played an important role in the adoption of this technique. Had it been invented much earlier than other chronometric dating methods, thermoluminescence would have perhaps been an essential part of the lake-dwelling dating process.
POSSIBLY
Fig. 4.5.1 Good and bad positions for pottery in an archaeological deposit in relation to TL sampling
4.6 Additional chronometric analyses as a help for chronology
Because of the large quantity of pottery collected from the numerous lacustrine sites, especially those without a reliable stratigraphic context, archaeologists were very tempted to introduce thermoluminescence as an absolute dating technique as soon as it was developed in the 1960s. The majority of pottery analysed came from Lake Constance and Lake Zurich and it was all collected and stored in museums or private collections without having a reliable chronological order. After an evaluation of the first results, the method was revealed to be indeed quite helpful for the identification and chronological placing of
Most of the dating techniques discussed so far utilise a physical change occurring within the material of the sample, which allows the determination of age, which may then be used as a basis for stratigraphic or archaeological correlation. These physical clocks may be based on radioactive decay, building up of radiationstimulated effect, or on chemical change. The date is intrinsic to the sample and is dependent, essentially, on laboratory measurements. A complementary approach to archaeological chronology particularly used in the lake-
29
The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
dwelling research is the placement of sites within a climatically determined framework, for example by stratigraphy. Two of the most widely used climatic frameworks are pollen analyses and glacial varves chronology.
the prehistoric Alpine lacustrine cultures in the years to come. The second climatically determined framework is the glacial varve chronology. Large supplies of sediment are deposited in postglacial lakes as a result of rapid spring and summer ice melt. The coarsest particles are deposited first on the bottom of the lake, leaving the finer clay particles in suspension. These suspended particles gradually settle out in winter producing a clay lamina that contrasts markedly with the coarser summer layer. Annual strata can, in this way, be distinguished and studied from a chronological point of view. The potential of glaciolacustrine sediment as a means of dating was first recognised by the Swedish geologist Gerard De Geer (1912) who noticed that as the last ice sheet wasted northwards across Sweden, it left behind a complex of moraines including lakes showing laminated sediments called varves. De Geer measured each laminar pair, on the assumption that they were produced annually and comparisons between sequences enabled correlations to be established (using similar principles to those now employed in matching of ring-width series in dendrochronology). By methodically extending his master-chronology across Sweden, De Geer (1940) compiled a long varve-chronology which extended back continuously to the beginning of the deglaciation, beyond 13kBP.
Pollen sequences are of great help in understanding ancient environments as well as providing information for relative dating. Pollen grains and spores are produced in great abundance by almost all species of plants and, because of their remarkable degree of preservation through time, especially in acid soils and peat bogs, they allow palynologists to reconstruct past vegetational patterns related climatic variations, in a chronological order. It is known that plants and trees are climatesensitive and therefore characteristic of specific environments, hence the possibility of determining any climatic variation in that area by analysing the various types of pollen found there. Palynology is not only used to reconstruct natural phenomena, but also contributes to the understanding of human activities integrated in a natural ecosystem. In other words, it creates chronological patterns of human influence on the environment which can eventually be given absolute dates. Palynological studies are very well developed within the Alpine lake-settlement research. They started playing an important role in the Pfahlbauproblem in the first half of the 20th century, when sites such as Sipplingen (Lake Constance, D) (Reinerth 1932), Egolzwil 2 and 3 (former Lake Wauwil, CH) (Vogt 1951), Hochdorfand HitzkirchSeematt (Lake Baldegg, CH) (Bosch 1939), Zug-Sumpf (Lake Zug, CH) (Speck 1955) and Arbon-Bleiche 2 (Lake Constance, CH) (Keller-Tamuzzer 1945) were intensively researched and careful pollen analyses were carried out in order to test Keller's theory (see also Chapter 3). Once the Pfahlbau problem was no longer an issue, pollen studies became mainly focused on climatic and vegetational aspects in relation to natural and cultural transformations including subsistence and economy. Lacustrine sites which, without doubts, played an important role in the achievement of the above-mentioned multipurpose palynological studies were those of Homstaad-Hornle (Untersee, Lake Constance, D), ZHMozartstrasse (Lake Zurich, CH), Bodman-Schachen 1 (Lake Constance, D), Arbon-Bleiche 2 (Lake Constance, CH) and most of the lacustrine villages of the entire Lake Feder (D) and Lake Chalain (F) basins. In some cases such as ZH-Mozartstrasse for example, pollen analyses were carried out so meticulously that not only a progressive cultural transformation of the site's surroundings could be reconstructed (see sub-chapter 1.5 and Figs. 1.7.1 and 1.7.2), but also the demographic aspect behind it could be considered. Palynological studies have become a crucial part of the lake-dwelling research in the past 70 years and, enhanced by the recently developed scientific techniques, they will definitely continue to contribute to the understanding of
If the deposits are not disturbed by erosion, the same varve principle could be applied to glacial lakes within the Alpine and pre-Alpine region. Unfortunately, because the lakes in the Alps lie in different micro-climatic zones and they do not have a similar hydrology, sequences are most of the time floating sequences without the possibility of correlation. Nevertheless, if such annual deposits are found on archaeological sites right before or straight after the anthropogenic stratum, they can be good evidence of climatic variations and in optimal conditions they can even be cross-dated and calibrated with dendrochronology. Although stratigraphic studies on Alpine lake deposits have remarkably increased in the past three decades, the development of a glacio-lacustrine varve chronology has as yet not been achieved. A successful application though, is the correlation of glacier retreats and expansions with the various lake level fluctuations based on elastic deposits on Lake Zurich. Thanl(s to core sampling on the shores as well as in the lake and archaeological evidence from the numerous lacustrine sites, it has been possible to prove that the glacier retreats corresponded to the lake high water level (Fig. 4.6.1). In other words, the augmented quantity of water received by the lake was the result of the glacier melting ice. This hydrologic characteristic of Lake Zurich is also reflected by the seasonal variations of the lake level which is in fact higher in late spring - early summer when the snow melting process is more intense ( see also sub-chapter 9.4).
30
'The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
and therefore, if we know the growth patterns and no much time has elapsed from the surface exposure and the lichens colonisation, it is possible to calculate for how long they have been growing. This time span is the surface exposure period in other words, the lichens' age. Once again, the accuracy of the date depends upon the area (in particular climate) and time (the older the less precise). In optimal conditions the date can be pushed back to 4500 years, but in more temperate or low altitude zones the range is much less. Although this method cannot be applied to the lake-dwelling dating, it has been employed in the dating of glacier recessions in the Alps and correlated it with the lake level fluctuations.
Seospiogolstaod
w,
"'""
,,,. t:: ·w tl!OO
i"
"' w :,
" [
""" As we have seen, although pollen analyses and glaciolacustrine varve chronology are not directly used as chronometric tools for the lake-settlement dating process, they are important integrating methods which can clarify and support the most commonly used dating techniques adding some vital information to the final lake-dwelling chronological development.
'"°"
4. 7 Correlations techniques
between
the
various
dating
The correlation of different dating methods is considered to be one of the most promising and useful avenues in chronology. Different dating techniques used in support of one another can bring much more accurate results than what can be achieved by blindly rely on one single method. A good example is dendrochronology, which has been used to calibrate 14C dating. The same observation is true as far as the relationship between relative and absolute dating is concerned. In fact, although the majority of dates are now provided by absolute dating methods, much of the internal consistency of those dates and chronological sequences comes from the framework provided by relative dating e.g. pottery typology.
[llJl atchi!ologisch.
belegta Si•dluogsphaseo (Lake-dwelling
Ill! Schlagphasen
(Tree-felling phases)
Because of the various dating techniques employed and the regional aspect of the lake-dwelling research, correlating the different methods has become a crucial integrating procedure within the prehistoric lacustrine sites dating process. The most common correlation is the comparison of tree-ring sequences of different regions. As pointed out in sub-chapter 4.3, tree-rings are climate sensitive and their width depends on weather and soil conditions. This is the reason why the comparison can only be made if the two or more regions considered have similar characteristics. For instance, the fairly reliable oak sequence created by the dendro-laboratory of Hemmenhofen (D) can easily be used all around Lake Constance and in the Oberschwaben region as far as Lake Feder and sometimes even compared with the neighbouring sequences (Billamboz et al. 1996). In exceptional circumstances, as it is the case of the comparative sequence analysis attempted by the dendrochronological laboratories of Verona (I) and Hemmenhofen (D), correlations can even be made
occupation)
Fig. 4. 6.1 Comparison of occupational patterns, lake level fluctuations and Alpine glacier movements (retreats and advances) (after Gross-Klee and Ritzmann 1990: 168)
Another dating technique that is strictly related to the above-mentioned glacier fluctuations is lichenometry. Lichens are complex organisms consisting of algae and fungi living together symbiotically. The principle of lichenometry, pioneered by Beschel (1973), can be summarised as follows: once lichens start colonising a surface, they keep growing for several thousands years
31
The Missing Period': Middle Bronze Age Lake-Dwellings in the Alps
between regions which are situated very far from one another. In fact, a correlation between the newly developed Bronze Age Lake Garda floating dendrochronological sequences and the oak sequence of southwestern Germany has been tried recently; studies are still in progress and final results are yet to be published (Billamboz pers. com. 1999). As mentioned earlier, a classical example of correlation between different dating techniques is the use of dendrochronology to calibrate 14C dates. Although also in this case, accuracy depends upon regional contexts and single situations, this method has been of great help for dating lacustrine sites, which, despite being located in areas covered by dendrochronological sequences, they were lacking of wooden remains, and the only dating technique which could be applied was 14C. Further applications of the symbiotic collaboration of dendrochronology and 14C is the wiggle-matching technique, whereby floating dendrochronological sequences are developed by carbon dating the rings of lacustrine wooden post found in archaeological excavations matching them with a precisely calibrated 14C curve. Once various floating sequences are created, they can be connected together (if they are close enough) eventually forming a master sequence which starts from the present days.
preservation with a consequently ahnost complete lack of organic materials, the only way of giving them a place in the chronological chart is to correlate the pottery typology with some low-accuracy 14C dates (Dickmann 1988). Correlations between the different absolute and relative dating techniques can be extended even further to those methods included in a climatically determined framework. As shown in sub-chapter 4.6, the two most common climatically determined dating techniques ( excluding dendrochronology) are pollen sequences and glacio-lacustrine varve chronology. Since these methods are determined by meteorological variations ( and the former being an important factor in determining vegetational transformations), they are often employed to correlate occupational patterns with climatic variations. Moreover, since climatic changes are all reliably recorded in the tree rings (by variation in width), dendrochronological sequences automatically become evidence of palaeoclimatic alterations and, as a result, different types of chronologies (such as palaeoclimatic, palaeovegetational and anthropogenic) can be obtained. Unfortunately, such analyses require a fairly precise set of data, which can only be obtained with an accurate multidisciplinary-based research. It is therefore not surprising if a complete and general chronological aspect of the prehistoric Alpine region is yet to be developed. There are though a few good examples of reliable regional chronologies covering both the environmental and the human aspect. Some of them are to be found around the Zurich bay area (CH), in the Bodman region (the western extreme of Lake Constance, D) as well as other limited areas on Lake Constance in general ( CH, D ), the Neolithic site of Hornstaad-Hornle (Untersee, Lake Constance, D), the numerous sites situated on the western shore of Lake Chalain (F), at Egolzwil (CH) and finally the entire well researched Lake Feder.
Absolute dates are often used in conjunction with relative ones. This is particularly true for 14C dates of lacustrine sites with a multi-phase occupational pattern. In fact, especially if dendrochronology is not available, accurate 14C dates in relation to the stratigraphic analyses of the anthropogenic layers are vital to reduce the 14C standard deviation. An excellent example of this application is the EBA (1 i 11 century BC) lacustrine occupation of ArbonBleiche 2. Meticulous comparative analyses of the apparently single anthropogenic layer dated with 14C, proved the existence of two short-term settling phases that were eventually fully confirmed by dendrochronology (Hochuli 1994) (see also sub-chapter 8.3.2). Similar examples are to be found all over the Alpine region where lacustrine settlements were occupied more than once throughout a limited period of time. Among them, the most relevant are: Auvernier and Cortaillod (Lake Neuchatel, CH), ZH-Mozartstrasse (Lake Zurich, CH), Bodman-Schachen 1 (Lake Constance, D), some of the numerous lacustrine sites on Lake Chalain (F) and Fiave (former Lake Carera, I).
The lake-dwelling research in the Alps is progressing very fast these days: comparative analyses and correlations between the various dating techniques are becoming more and more important and, since the main objective of archaeologists is to understand the Alpine lake-settlement phenomenon as a whole, they will certainly undergo further development in the future.
Another important correlation between absolute dates and relative sequences is that between chronometric dates and typological analyses (in particular pottery chronologies which is the oldest, but still the most used dating technique within the lake-dwelling studies). As it is explained in more details in Chapter 11, this kind of correlation can be vital in extreme cases such as the dating of MBA (15th -13th centuries BC) "in-land" sites located in the proximity of lakes and believed to be of lacustrine origins. Because of the very poor site
32
Chapter 5 THE BRONZE AGE IN THE ALPS: CHRONOLOGY AND ARCHAEOLOGICAL PERIODISATION 5.1 Introduction
periodisation is likely to split up into local interpretations which suit various particular local conditions and regional traditions of research. It is consequently vital to be as precise as possible in terms of location when the metal Age is being discussed.
Before attempting a chronological exposition of the Bronze Age we must defme how this term has evolved. The concept of an European Bronze Age dates from the first half of the 19th century when the Dane C.J. Thomsen divided prehistory into three main phases: a Stone Age (later divided into Palaeolithic and Neolithic), the Bronze Age and fmally the Iron Age. Archaeologically speaking the term "Bronze Age" defines a specific span of time in which various cultures made use of alloyed metal in their everyday life for the first time (Coles and Harding 1979; Sherratt 1992).
The first use of copper-alloy atiefacts on large scale occurred in the Near East and southeast Europe towards the end of the fourth millennium BC whereas the adoption of tin-bronze reached the Balkans and Carpathian basin by 2500 BC (Sherratt 1994). Within a few centuries, much of the rest of Europe including the western Mediterranean area and also central Europe were influenced by the innovative alloy, and distinct traditions of bronze use began to develop in specific regional contexts. These regional groups were linked by two distinct transcontinental routes, namely the Danubian Route and the Alpine Route, which were used alternatively between the 23 rd and the 7th century BC. The Danubian Route was dominant from 2200 to 1600 BC and from 1300 to 900 BC whereas the Alpine Route prevailed from 1500 to 1300 BC and from 900 BC to the end of the Bronze Age (Fig. 5.1.1).
The initial adoption of metal in prehistoric times was not uniform everywhere. In other words, if we want to use the Bronze Age as a technical archaeological periodisation, it is necessary to specify the geographical area in question, and the particular systems or sub-divisions in use in that region. There is for instance a substantial difference between the southeast European Bronze Age and that of Central Europe, or between the Mediterranean Bronze Age and the northern European one. Furthermore, in a geographically and linguistically diverse area such as the Alpine region, as it will be shown later, any system of
·,
Salik
Danube
~
c'l'lst M&diternm&an
2500 oc
?000
I
Alps
I
."
Satlit:
Baltic
•
B~f!:G
•
I
Oa1;ub9
East M;di!Hr-Ml&M
~
East •
Italy
M~