On The Surface - Ian Hodder

Download as pdf or txt
Download as pdf or txt
You are on page 1of 388

~ McDONALD INSTITUTE MONOGRAPHS

On the surface:
C::atalhOyiik 1993-95
Edited by Ian Hodder

IMERkO)

BRITISH INSTITUTE OF ARCHAEOLOGY AT ANKARA


• McDONALD INSTITUTE MONOGRAPHS

On the surface:
<:;atalh6yiik 1993-95
Edited by Ian Hodder

The <;atalhoyiik Research Trust IMERkO)


c;atalhoyiik Project Volume 1

BRITISH INSTITUTE OF ARCHAEOLOGY AT ANKARA


BIAA Monograph No. 22
Published jointly by:

McDonald Institute for Archaeological Research British Institute of Archaeology at Ankara


University of Cambridge c/o British Academy
Downing Street 20-21 Cornwall Terrace
Cambridge London
CB23ER NW14QP
(0)(1223) 339336

Distributed by Oxbow Books.


United Kingdom: Oxbow Books, Park End Place, Oxford, OX1 1HN.
Tel: (0)(1865) 241249; Fax: (0)(1865) 794449
USA: The David Brown Book Company, P.O. Box 511, Oakville, CT 06779, USA.
Tel: 860-945-9329; Fax: 860-945-9468

ISBN: 0-9519420-3-4
ISSN: 1363-1349 (McDonald Institute)
0969-9007 (BIAA)

© 1996 McDonald Institute for Archaeological Research

All rights reserved. No parts of this publication may be


reproduced, stored in a retrieval system, or transmitted, in any
form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without the prior permission of the
McDonald Institute for Archaeological Research.

Edited for the Institutes by Chris Scarre (Series Editor) and Dora A. Kemp (Assistant Series Editor).

Film produced by Gary Reynolds Typesetting, 13 Sturton Street, Cambridge, CB1 2QG.
Printed and bound by Short Run Press, Bittern Rd, Sowton Industrial Estate, Exeter, EX2 7LW.

ii
CONTENTS

Contributors v
Figures vi
Photographs X
Tables xii
Acknowledgements XV

Chapter 1 Re-opening <;:atalhoyiik 1


IAN HoDDER with a note on Building Complexity by TIM RITCHEY

Chapter 2 Preliminary Results of Geoarchaeological Investigations at <;:atalhoyiik 19


NEIL RoBERTS, PETER BoYER & RoMOLA PARISH

Chapter 3 The Konya Plain Survey: Aims and Methods 41


DouGLAS BAIRD

Chapter 4 Excavations at Pmarba:;a: the Early Stages 47


TREVOR WATKINS

Chapter 5 Topographic Survey of the <;:atalhoyiik Mounds 59


ToM POLLARD, CoLIN A. SHELL & DA vm R. TwiGG

Chapter 6 Systematic Surface Collection 73


RoGER MATTHEWS

Chapter 7 Surface Scraping and Planning 79


RoGER MATTHEWS

Chapter 8 Magnetometric Survey at <;:atalhoyiik East 101


COLIN A. SHELL

Chapter 9 Surface Pottery at <;:atalhoyiik 115


JoNATHAN LAST

Chapter 10 The Knapped Stone 173


JAMES CONOLLy

Chapter 11 Surface Material: Animal Bone and Worked Bone 199


LouiSE MARTIN & NERISSA RussELL

Chapter 12 Figurines, Clay Balls, Small Finds and Burials 215


NAOMI HAMILTON with a contribution by MEHMET ULUCEVIZ

Chapter 13 Trace Element Analyses of Bones and Teeth from <;:atalhoyiik 265
THEY A MoLLESON & PETER ANDREws

Chapter 14 Exploring the 1960s' Surface: the Stratigraphy of <;:atalhoyiik 271


WENDY MATTHEWS & SHAHINA FARID

Chapter 15 Multiple Surfaces: the Micromorphology 301


WENDY MATTHEWS, CHARLES FRENCH, TIMOTHY LAWRENCE & DAVID CUTLER

iii
Chapter 16 Luminescence Dating of Mud Brick from <:;atalhoyi.ik 343
RoMOLA PARISH

Chapter 17 Interim Dendrochronological Progress Report 1995/6 345


PETER KUNIHOLM & MARYANNE NEWTON

Chapter 18 <:;atalhoyi.ik: the Anthropology of an Archaeological Presence 349


DAVID SHANKLAND

Chapter 19 Conclusions 359


IAN HODDER

iv
Contributors

Peter Andrews Natural History Museum, London


Douglas Baird Department of Archaeology, University of Liverpool
Peter Boyer Department of Geography, Loughborough University
James Conolly Institute of Archaeology, University College London
David Cutler Jodrell Laboratory, Royal Botanic Gardens, Kew
Shahina Farid Museum of London
Charles French Department of Archaeology, Cambridge University
Naomi Hamilton Department of Archaeology, University of Edinburgh
Ian Hodder Department of Archaeology, Cambridge University
Peter Kuniholm Wiener Laboratory, Cornell University
Jonathan Last Archaeology Unit, Shire Hall, Cambridge
Timothy Lawrence Jodrell Laboratory, Royal Botanic Gardens, Kew
Louise Martin Institute of Archaeology, University College London
Roger Matthews British Institute of Archaeology at Ankara
Wendy Matthews McDonald Institute for Archaeological Research, Cambridge University
Theya Molleson Natural History Museum, London
Maryanne Newton Wiener Laboratory, Cornell University
Romola Parish Geography Laboratory, University of Sussex
Tom Pollard Department of Archaeology, Cambridge University
Tim Ritchey Department of Archaeology, Cambridge University
Neil Roberts Department of Geography, Loughborough University
Nerissa Russell Department of Anthropology, University of California, Berkeley
David Shankland Department of Archaeology, St David's University College, Lampeter
Colin A. Shell Department of Archaeology, Cambridge University
David R. Twigg Department of Civil and Building Engineering, Loughborough University
Mehmet Uluceviz Department of Archaeology, Bilkent University
Trevor Watkins Department of Archaeology, University of Edinburgh

v
Figures
1.1. c;atalhOyiik East. 2
1.2. c;atalhOyiik West. 3
1.3. Building complexity for levels I-III. 8
1.4. Building complexity for levels IV-V. 8
1.5. Building complexity for level VIA. 9
1.6. Building complexity for level VIB. 9
1.7. Building complexity for levels VII-XI. 10
1.8. Redrawn plan of Mellaart's level I. 10
1.9. Redrawn plan of Mellaart' s level II. 11
1.10. Redrawn plan of Mellaart's level III. 11
1.11. Redrawn plan of Mellaart' s level IV. 12
1.12. Redrawn plan of Mellaart's level V. 13
1.13. Redrawn plan of Mellaart' s level VIA. 14
1.14. Redrawn plan of Mellaart's level VIB. 15
1.15. Redrawn plan of Mellaart' s level VII. 16
1.16. Redrawn plan of Mellaart' s level VIII. 17
1.17. Redrawn plan of Mellaart's levels IX (top), XI and XII. 17
2.1. Late Quaternary sediments of the western Konya plain, and location of c;atalhOyiik. 21
2.2. Plan of c;atalhoyiik with core sites. 26
2.3. Core C:H.94-A: lithology and measured parameters. 27
2.4. Core c;H.94-B: lithology and measured parameters. 28
2.5. Core c;H.94-C: lithology and measured parameters. 29
2.6. Core C:H.94-D: lithology and measured parameters. 30
2.7. Core c;H.95-E: lithology and measured parameters. 31
2.8. Core C:H.94-F: lithology and measured parameters. 32
2.9. c;ar~amba palaeochannel PCI: recorded stratigraphy (1994), north-facing section. 34
2.10. c;ar~amba palaeochannel PCI: measured parameters for each stratigraphic unit. 34
2.11. Particle size ratios for 1993 and 1994 cores. 38
2.12. c;atalhoyiik: composite lithostratigraphy and correlation. 38
3.1. Sites in the c;atalhoyiik area. 43
3.2. C:ingere Hiiyiik showing distribution of collection squares and contours at 1 m intervals. 45
4.1. Map of the Konya plain, locating Pmarba~z relative to Kara Dag and c;atalhoyiik. 48
4.2. Plan of Site B- the rock-shelter: latest features. 49
4.3. Plan of Site B- the rock-shelter: the stone retaining wall. 50
4.4. Section on north side of Trench 1. 53
4.5a. Plan of Site A- the settlement on the peninsula. 54
4.5b. Plan of Site A- the burial in the sounding into the early prehistoric levels. 54
4.6. Sections in Site A. The locations of the sections are given in Figure 4.5a. 55
5.1. Plan of the c;atalhoyiik mounds and their immediate surroundings taken from a local map. 60
5.2. Plan of the east mound contoured at 0.5 m intervals. Scale 1:2500. 62
5.3. Generalized plan of the west mound at c;atalhoyiik with contours at 0.5 m intervals. Scale 1:2500. 64
5.4. Surface of the east mound digitally shaded from the northeast by a point source at 30° elevation. Scale 1:2500. 65
5.5. Northeast shaded surface of the east mound with 0.5 m contours superimposed. Scale 1:2500. 66
5.6. Northeast shaded surface of the east mound viewed from the east. 67
5.7. Approximately north-south profile of the east mound at c;atalhOyiik. 68
5.8. West-east profile across the main summit of the east mound at c;atalhoyiik. 69
5.9. West-east profile across the northern subsidiary summit of the east mound. 70
6.1. East mound: surface distribution of all artefacts. 74
6.2. West mound: surface distribution of all artefacts. 76
7.1. Plan of east mound showing location of all scraped squares. 80
7.2. Scrape areas on c;atalhOyiik East in 1993 as seen from the northeast of the site. 81

vi
7.3. North Area. Plan of entire area; historic phase walls in south are indicated by lighter shading. 82
7.4. Examples of detail observed after scraping on the northern eminence of c;atalh6yiik East. 83
7.5. North Area. General plan showing walls, reconstructed walls, burnt areas and space numbers. 84
7.6. Central Area (980,1080). 87
7.7. 1040,1240 plan. 89
7.8. 1040,1140 plan. 89
7.9. 1045,1125 plan. 90
7.10. 1040,1090 plan. 90
7.11. 940,1040 plan. 91
7.12. 990,1040 plan. 91
7.13. 1040,1040 upper plan. 92
7.14. 1040,1040 lower plan. 92
7.15. 1090,1040 plan. 93
7.16. 1090,1040 excavated burial plan. 93
7.17. 990,990 plan. 94
7.18. Plan of west mound showing location of scraped areas. 95
7.19. 480,1040 plan. 94
7.20. 580,1020 plan. 96
7.21. 640,960 plan. 96
7.22. Location of transects of scraped squares used in Figures 7.23 to 7.25. 97
7.23. Soil depths in scraped areas at 1000 mE. 98
7.24. Soil depths at 1040 mE. 98
7.25. Soil depths at 1040 m N. 98
8.1a. Plan of the c;atalhoyiik east mound: area geomagnetically surveyed and its relationship to the scrape areas. 102
8.1b. View of the east mound from the southwest: topographic position of the magnetic survey area. 103
8.2. Analysis of a static test on site of the FM36 fluxgate gradiometer. 105
8.3. Composite plot of the fluxgate gradiometer survey. 106
8.4. Plan showing the principle features visible in the magnetic survey, Figure 8.3. 107
8.5. Comparison between the magnetic results for scrape area (1040,1140) and its plan. 108
8.6. Comparison between the magnetic results for scrape area (980,1080) and its plan. 109
8.7. Comparison between the magnetic results for northern scrape area 2 on the northern mound and its plan.
8.7a. Magnetic survey with plan overlayed. 111
8.7b. Key to the serape plan. 112
8.7c. Magnetic survey. 113
9.1. Neolithic holemouth vessel rim forms from c;atalh6yiik East. 119
9.2. Decorated Neolithic rims and lugs from c;atalhoyiik East. 121
9.3. Neolithic open bowl rim forms from c;atalh6yiik East. 123
9.4. Other Neolithic bowl forms from c;atalh6yiik East. 125
9.5. Neolithic lugs, handles and base forms from c;atalh6yiik East. 127
9.6. c;atalh6yiik East surface collection -distribution of Neolithic holemouth and bowl rims. 130
9.7. c;atalh6yiik East surface scraping -location of different areas. 131
9.8. c;atalh6yiik East surface collection- total sherd numbers. 134
9.9. c;atalh6yiik East surface collection -Neolithic sherd numbers. 136
9.10. c;atalh6yiik East surface collection- percentage of sherds <2 em. 138
9.11. c;atalh6yiik East surface scraping- proportions of Neolithic sherds. 140
9.12a. c;atalh6yiik East surface scraping- northern eminence: density of all sherds. 142
9.12b. c;atalh6yiik East surface scraping- northern eminence: proportion of Neolithic sherds. 143
9.13. c;atalh6yiik East surface scraping- square 1020,1170: sherd distributions. 144
9.14. Wheelmade rim forms from c;atalhoyiik East. 147
9.15. Other wheelmade rim forms from c;atalhoyiik East. 149
9.16. c;atalhoyiik East surface collection -distribution of Hellenistic slipped wares. 150
9.17. c;atalhoyiik East surface collection -distribution of green-glazed wares. 151
9.18. Chalcolithic painted rim forms from c;atalh6yiik West. 155

vii
9.19. (:atalhoyiik West surface collection - Chalcolithic sherd numbers. 156
9.20. (:atalhoyiik West surface collection - wheelmade sherd numbers. 157
9.21. (:atalhoyiik West surface collection -distribution of dark painted wares. 158
9.22. (:atalhoyiik West surface collection -distribution of lugs and handles. 159
9.23. Chalcolithic lug forms from (:atalhoyiik West. 161
9.24. Chalcolithic painted and incised decoration from (:atalhOyiik West. 163
9.25. Handmade and wheelmade forms from (:atalhOyiik West. 165
9.26. Neolithic pottery from (:atalhoyiik East. 166
9.27. Chalcolithic pottery from (:atalhoyiik West. 167
10.1. Cores and core tablets- east mound. 178
10.2. Blades, retouched blades, retouched flints- east mound. 179
10.3. Histogram of blade widths: east and west mound top-scrape sample. 180
10.4. Retouched/utilized pieces - east mound. 182
10.5. Bifacesjprojectiles - east mound. 184
10.6. Large retouched flakes from cache on northern eminence. 185
10.7. Mean count of lithic artefacts in each west mound 2 x 2 unit, by northing. 186
10.8. Percentage of blades in each west mound northing. 186
10.9. Distribution of lithic artefacts: east mound. 187
10.10. Distribution of blades: east mound. 188
10.11. Distribution of chips: east mound. 189
10.12. Percentage of blades in each east mound northing. 190
10.13. Density of lithic artefacts by area: east mound. 190
10.14. Percentage of tools by area: east mound. 192
10.15. Percentage of chips and shatter: east mound. 192
10.16. 2 x 2 distribution of lithic artefacts in square 1040,1170. 193
10.17. 10 x 10 distribution of lithic artefacts in areas 2 and 3. 193
10.18. Percentage of blades by level. 193
11.1. The number of bone fragments collected from each 2 x 2 m square on the east mound. 200
11.2. The number of wheel made sherds collected from each 2 x 2 m square on the east mound. 202
11.3. The percentage of bone fragments <2 em length of the total number of fragments in each
2 x 2 m square on the east mound. 203
11.4. The distribution of identifiable medium-sized herbivore bones on the east mound. 204
11.5. (:atalhoyiik East; scraped square animal bones- fragments per m3 • 206
11.6. The percentage of burnt bone fragments for each 2 x 2m unit in scraped square 1020,1170. 207
11.7. The number of bone fragments collected from each 2 x 2 m square on the west mound. 208
11.8. Worked bone. 212
12.1. Examples of Neolithic figurine types from the 1960s' excavations at (:atalhOyiik East. 216
12.2. Figurines from (:atalhoyiik East. 234
12.3. Human figurine from (:atalhoyiik East. 235
12.4. Animal figurine from (:atalhoyiik East. 235
12.5. Stone artefacts from (:atalhoyiik East. 237
12.6. Miscellaneous small finds from (:atalhoyiik East. 239
12.7. Decorated pot stands from (:atalhoyiik West. 243
13.1. The occlusal surface of an adult mandible. 267
13.2. A) The occlusal surface of the mandible of a 4-5 year old individual; B) SEM micrograph showing the large
pits and strongly directional striations on the crushing facet of the first deciduous molar on the mandible. 268
14.1. Plan of sections recorded in 1993-95 in Mellaart's Areas A and Eat the southwest of the site. 272
14.2. Plan of upstanding walls within the 20 x 20m area of excavations renewed in 1995. 274
14.3. Area A, levels VIA/B-Ill. Sections 1-2. 279
14.4. Area A-E, levels VII-Ill. Section 3. 279
14.5. Area A-E, levels VII-VIA. Sections 4-5. 281
14.6. Area E, levels IX-V. Section 7. 281
14.7. Overall view of section 7. 282

viii
14.8. Composite of west-east sections through Area E, south facing. 283
14.9. Composite of east-west sections through Area E, north facing. 283
14.10. Shrine 1, levels VIII-VII. Section 40. 287
14.11. House 2, level VII. Sections 32-4. 283
14.12. House 4, level VII. Sections 10, 15 and 16. 288
14.13. House 5, level VII. Sections 11, 13 and 14. 288
14.14. House 6, level VII. Sections 19-21. 288
14.15. Shrine 8 antechamber, level VII. Section 22. 290
14.16. ?House/building and Shrine 10, levels VIII-VII. Section 8. 291
14.17. Shrine 10, levels VIII-VII. Sections 12 and 17. 291
14.18. House 12, level VII. Sections 28-31. 292
14.19. Sections 28-36 and buildings VIII:12, VII:2 and VII:16 in the area excavated by Mellaart. 293
14.20. House, Court and Shrine 25, levels XII-VIII and Court X/VIII.25/15 ?stable deposits. Section 6. 292
14.21. Plan of House 25, level XII floor. 293
14.22. Shrine 27, levels VIII-VII. Section 18. 295
14.23a. Chart illustrating depth of floors and height of walls in buildings which are sealed by later levels. 295
14.23b. Chart illustrating depth of floors as a percentage of the height of walls in buildings sealed by later levels. 295
14.24a. Area in m2 of the three major context types in sections 1-3 and 6-8. 299
14.24b. Percentages by area in m2 of walls, room fill and floor sequences within buildings (sections 1-3 & 6-8). 299
15.1. Microstratigraphy of two fire installations in section 2, level V, south of V.61. 305
15.2. Microstratigraphic columns illustrating the type, thickness and frequency of floors and deposits. 313
15.3. Microstratigraphy of platform and floors in the building south of VII.32, section 9. 314
15.4. Microstratigraphy of floors in the building east of VII.32, section 7. 315
15.5. Microstratigraphy of floors in Shrine 25, level VIII, section 6. 316
15.6. Phasing of floors in Shrine 25, level VIII, section 6. 316
15.7. Microstratigraphy of floors in a building in section 3, level VIA/V, east of VIA.42 Court or V.9. 318
15.8. Microstratigraphy of floors in a building in section 4, level VIB/A, north of VIA.46. 320
15.9. Preliminary guide to microstratigraphic attributes diagnostic of uses of space within buildings. 321
15.10. Preliminary guide to microstratigraphic attributes diagnostic of uses of space in open areas. 322
15.11. Microstratigraphy of ?stable deposits in Court X/VIII.25/15, section 6. 324
17.1. 570 year juniper ring-chronology from several occupation levels at Neolithic (atalhOyiik. 346
17.2. Two radiocarbon-dated samples (Lots A and B, ten rings each). Their mid-points are exactly 520 years
apart in the Catalhoyiik ring-sequence. 346
18.1. Imagery from (atalhOyiik on the cover of a catalogue of the Anatolian Civilizations Museum in Ankara. 353
18.2. Medal from the (umra melon festival. 353
18.3. Logo of the (umra youth radio station. 354

ix
Photographs
15.1. The microstratigraphic sequence in this building suggests there was a change in the use of space from:
i) domestic activities attested by relatively thick plaster floors and burnt occupation deposits [1], associated
with a plastered emplacement perhaps for a pot [2], to ii) cleanf?ritual activities attested by the sequence of
fine white and orange plaster floors [3] which are associated with a Bos jaw, horn cores and elaborate
plastered features [4]. Building in the south of section 3, level VIA/V. Scale= 20 em. 326
15.2. Block sample for micromorphological analysis in the process of being cut out from a sequence of multiple
layers of burnt plaster in a fire-installation. Building from level V in section 2, Area A. Sample (:h93.08,
H:13.5 em, W:6.5 em, depth 8-10 em. 326
15.3. Modern sample of soft lime deposits which occur within 5 km of (:atalhiiyiik and closely resemble materials
used in Neolithic white plasters. These deposits comprise up to 95 per cent pure carbonates of calcium
and magnesium, and are used today for plastering floors and walls in surrounding settlements. Sample
(:h94.50. Cross-polarized light (XPL). Frame width= 7.2 mm. 327
15.4. The first 41layers of a sample of wall plaster are visible here in sequence from left to right. These plasters
were laid as couplets comprising: [1] a pale brown silty clay preparatory layer, 24-720 J.Lm thick, with
vegetal stabilizers which have since decayed leaving voids in the sedimentary microstructure, and [2] a
finer denser finishing coat, 12-552 ).lm thick, which was keyed into the irregular surface of the preliminary
coat. The outer surfaces of these finishing coat are very regular and sharp, and were probably smoothed or
burnished. Sample (:h94.24. Collapsed fragment in Area E. XPL. Frame width = 7.2 mm. 327
15.5. Mud brick [1] and mortar [2] from the earliest building uncovered by Mellaart, XII.25. The mortar
comprises heterogeneous burnt and unburnt anthropogenic debris including fragments of charred plant
remains, bone, and volcanic glass. Sample (:h94.13. Area E, section 6. Plane polarized light (PPL).
Frame height= 7.2 mm. 328
15.6. Many fire-installation plasters have a coarser particle size than floor plasters, and closely resemble alluvial
deposits sampled in a palaeochannel adjacent to the site. Each plaster has been baked up to 1-2 em from the
surface [1]. Only discontinuous lenses of burnt remains <2 mm thick [2] survive between each layer of
plaster [3]. Sample (:h93.08. Area A, section 2, level V. PPL. Frame height= 7.2 mm. 328
15.7. Identified plants in thin section include: a) transverse section, charred wood, Quercus, deciduous white
oak type. Sample (:h93.12. Open area, Area A, section 2. PPL. Frame height= 7.2 mm; b) transverse
section, charred wood, Salix/Poplus, willow/poplar. Sample (:h93.12. Open area, Area A, section 2. PPL.
Frame width= 7.2 mm; c) transverse section, charred wood, Chenopodiaceae similar to Salsola and
Sueda for example. Sample (:h94.10. Open area, Area E, section 6, level IX. PPL. Frame width= 4.4 mm;
d) charred triquetrous stem of Cyperaceae, Cyperus/Carex/Scirpus. Sample (:h93.07. Open area,
Area A-E, section 3. PPL. Frame width= 1.76 mm. 329
15.8. Initial sequence of thick plaster floors [1 & 3] and occupation deposits [2] in section 3, probably
associated with domestic activities. These deposits include charred wood, siliceous plant remains,
calcareous ashes, bone, and burnt aggregates. The surface of the plaster floor [1] has sub-horizontal cracks
and is impregnated with pale yellow salts. Sample (:h93.03. Area A-E, section 3, southern building,
level VIA-V. PPL. Frame height= 7.2 mm. 330
15.9. A similar sequence of thick plaster floors [1 & 3] and occupation deposits [2] probably related to domestic
activities, includes bone and charred and siliceous plant remains, including a cereal-like remains. Sample
(:h93.01. Area A, section 4, building north of VI.46. PPL. Frame height= 7.2 mm. 330
15.10. Later sequence in section 3 of thin plaster floors [1] with finishing coats of white plaster [2] and
occupation deposits (<0.2 mm thick), associated with cleanj?ritual activities. Sample (:h93.03.
Area A-E, section 3, southern building, level VIA-V. PPL. Frame height= 7.2 mm. 331
15.11. Comparative sequence of thin plaster floors with multiple couplets of white plaster, and sparse occupation
deposits. Linear and curvilinear pseudomorphic voids from decayed vegetal stabilizers are clearly visible

X
in the plasters [1]. Sample (h94.21. Area E, section 7, building east of VII.32. XPL.
Frame height = 7.2 mm. 331
15.12. Microstratigraphic sequence of ?ritual deposits in the small hole in the top of the grave in section 3,
comprising lenses of: [1] thin white plaster, [2] silt loam ?plaster, [3] red ochre, [4] aggregates of white plaster,
[5] red ochre, [6] deposits with finely fragmented charred plant remains, and finally [7] a thick capping of
white plaster. Sample (h93.06. Area A-E, section 3. Reflected light (RL). Frame width = 7.2 mm. 332
15.13. Detail of probable domestic occupation deposits in section 3, including fragments of bone [1], cereal-like
husk [2], and siliceous grass epidermis [3]. Sample (h93.03. Area A-E, section 3, southern building,
level VIA-V. PPL. Frame width= 1.3 mm. 333
15.14. Deposits rich in organic staining [1], salts [2], aggregates of burnt fire-installation plaster [3], and
charred plant remains, east of the fire-installations in Shrine VIII.25. Sample (h94.02. Area E,
section 6. PPL. Frame width = 7.2 mm. 333
15.15. Thin layer of compacted fibrous deposits [1] below a thin layer of white plaster [2] in the later sequence of
deposits in section 3, southern building. Sample (h93.03. Area A-E, level VIA-V. PPL.
Frame height = 1.3 mm. 334
15.16. Thin lenses of sandy silt loam [2] and occupation deposits rich in charred 'reed'-like plant remains [3],
between white plaster floors in Shrine VIII.25 [1 & 4]. Sample (h94.07. Area E, section 6. PPL.
Frame width= 1.3 mm. 334
15.17. Interbedded layers of dung-rich deposits and salts in Area E, Court 25/15 level X/VIII. Sample (h94.15.
Section 6, west. Scale= 50 em. 335
15.18. Lenses of salt [1], dung pellet type 1 [2] and organic staining [3] in Court 25/15 level X/VIII. Sample
(h94.15. Area E, section 6, west. PPL. Frame width= 7.2 mm. 335
15.19. Detail of dung pellet in Photograph 15.18, illustrating partially digested plant remains, including
fragments of Gramineae stem and leaf, and ?tubers. Sample (h94.15. Area E, section 6, west,
Court 25/15 level X/VIII. (atalhoyiik. PPL. Frame height= 1.3 mm. 336
15.20. Detail of dung pellet, same view as Photograph 15.19 in cross-polarized light illustrating abundant
spherulites. The spherulites are 5-15 Jlm in diameter, and have a moderate birefringence with a cross
of extinction, in contrast to the isotropic plant remains. Spherulites have been found in the gut and dung
of animals such as sheep, goats, cattle, gazelle and pig, but further research is required, before any
associations can be confirmed. Sample (h94.15. Area E, section 6, west, Court 25/15 level X/VIII. XPL.
Frame height= 1.3 mm. 336
15.21. Scanning electron microscope image of a spherulite from a bulk sample of deposits from Court 25/15 level
X/VIII in collaboration with M. Canti, English Heritage. Sample (h95.85. Area E, section 6, west. 337
15.22. Sequence of dung-rich deposits in Court 25/15 level X/VIII. PPL. Frame height = 7.2 mm. 338
15.23. Comparative sample of modern sheep/goat stable deposits from the Konya region, collected by Seona
Anderson, University of Sheffield, illustrating similar but less compacted horizontal layering and
dung pellet fragments. PPL. Frame height = 7.2 mm. 338
15.24. Dung pellets type 2 [1] in debris lying on the latest floor in the southern building in section 3. These dung
pellets have a fibrous structure with few recognizable plant remains and do not contain any spherulites.
Some pellets are more compressed than others. Sample (h93.04. Area E. PPL. Frame height = 7.2 mm. 339
15.25. ?Omnivore coprolite type 3, with yellowish organic fine fabric, and fragments of bone, and charred and
siliceous plant remains and pseudomorphic voids. Sample (h93.16. Area A-E, open area in south of
section 3, level VIA-V. PPL. Frame height= 1.4 em. 339
15.26. Obsidian flake [1] and small tapered fragments of charred oak wood [2] in a layer in an open area with
charred 'reed' and willow/poplar fragments, which could perhaps represent remains from wood working and
?basketry/matting. Sample (h93.12. Area A, section 2. PPL. Frame width= 7.2 mm. 340

xi
Tables
2.1. Luminescence dose rate determinations. 24
2.2. Luminescence age determinations. 24
2.3. Principal sediment types and typical ranges for measured parameters. 35
2.4. Lower alluvium characteristics. 36
4.1. Radiocarbon dates. 52
9.1. Neolithic rim characteristics by building level. 117
9.2. Neolithic base characteristics by building level. 117
9.3. Neolithic lug characteristics by building level. 118
9.4. Neolithic body sherd characteristics by building level. 118
9.5. c;atalhoyuk East surface scraping- typological characteristics of the Neolithic rims. 122
9.6. c;atalhoyuk East surface scraping- typological characteristics of the Neolithic bases. 129
9.7. c;atalhOyuk East surface scraping- typological characteristics of the Neolithic lugs. 129
9.8. c;atalhOyuk East surface scraping- typological characteristics of the Neolithic body sherds. 129
9.9. c;atalhOyuk East surface scraping- assemblage characteristics by depth. 137
9.10. c;atalhOyuk East surface scraping- assemblage characteristics by area. 139
9.11. c;atalhOyak West surface scraping- Chalcolithic assemblages. 153
10.1. Debitage categories by raw material: east and west mound 2 x 2 sample. 177
10.2. Debitage categories by raw material: east and west mound top-scrape sample. 177
10.3. Blade and flake butt types: east mound top-scrape sample. 180
10.4. Blade dorsal scar patterning: east mound top-scrape sample. 180
10.5. Mean and standard deviation of obsidian blade widths: east and west mounds top-scrape sample. 180
10.6. Flake dorsal scar pattern: east mound top-scrape sample. 180
10.7. Mean and standard deviation of obsidian flake widths: east and west mound top-scrape sample. 181
10.8. Tools by raw material: east and west mound 2 x 2 sample. 181
10.9. Modified edge delineation by blank type: east mound top-scrape sample. 181
10.10. Modified edge delineation by blank type: west mound top-scrape sample. 181
10.11. Edge delineation and angle types: east mound top-scrape sample. 183
10.12. Edge delineation and angle types: west mound top-scrape sample. 183
10.13. Location of modification on blades and flake blanks: east mound top-scrape sample. 183
10.14. Location of modification by modification position: east mound top-scrape sample. 191
10.15. Functional classifications by raw material: east and west mound top-scrape sample. 191
10.16. Functional classifications by blank type: east mound top-scrape sample. 191
10.17. Numbers of blades, flakes and other debitage by area: east mound top-scrape sample. 191
10.18. Counts of bifaces/projectiles and cores and presence of thinning-flakes by structure. 196
11.1. The number of animal bone fragments identified to each taxon/category from the 2 x 2 m surface collection. 201
11.2. c;atalhOyuk East scraped squares, animal bone data (ordered north to south, west to east). 205
11.3a. The number of identified fragments per taxon/category for each scraped square assemblage. 209
11.3b. The relative proportions(%) of taxa for each scraped square assemblage. 209
11.4. Counts of proximal and distal humeri for cattle and medium-sized herbivores, and proximal and distal
tibiae fragments for medium-sized herbivores only, from all scraped deposits. 209
11.5. The number of identifiable animal bone fragments per taxon/category from the west mound 2 x 2 m
surface collection. 209
11.6. c;atalhoyak West scraped squares, animal bone data. 209
11.7. The number of identified fragments per taxon/category for each scraped square unit. 210
11.8. c;atalhoyuk East scraped squares, worked bone. 210
11.9. c;atalhoyak West scraped squares, worked bone. 210
12.1. Figurines by building. 217
12.2. Schematic and human figurines by level. 217
12.3. Schematic and human figurines by material and level. 217
12.4. Mellaart's figurines by fragmentation and context. 220

xii
12.5. Figurine fragments from the surface survey. 229
12.6. Distribution of clay ball fragments on the surface. 230
12.7. Beads: number of beads made of each material, by level. 247
12.8. Beads: number of bead groups containing each material, by level. 247
12.9. The sex of adult skeletons in the different levels according to Angel and Ferembach. 255
12.10. Juveniles by age. 255
12.11. Juveniles by building. 256
13.1. (atalhoyiik bones and teeth analyzed by ICP. 266
14.1. Height of the first floor in each building above base of walls. 275
14.2. Correlation of the absolute readings taken in 1993-95 of floors and architectural features visible today
in the surviving sections and the heights recorded by Mellaart on plans in Anatolian Studies 1962,
1963, 1964 and 1966. 277
14.3. Summary of section information recorded in 1995 (letters in column refer to Fig. 14.8). 284
14.4. Summary of section information recorded in 1995 (letters in column refer to Fig. 14.9). 285-6
15.1. Pollen analyses conducted by W. Eastwood. 309
15.2. Contextual variation in types of plant remains and dung. 311
19.1. The degree of continuity of elaborate and non-elaborate buildings between different levels. 364
19.2. The percentage of elaborate buildings in level n that continue as elaborate buildings (a) from level n-1,
and (b) to level n+ 1. 364

xiii
Acknowledgements

Funding for the field research was provided by the men I have met or seen anywhere. They come from
British Academy, the British Institute of Archaeol- the local village, Kiic;iikkoy, and are chosen and ably
ogy at Ankara, the McDonald Institute, and the Na- led by Ismail Ya~h. We are well looked after by
tional Geographic Society. The project also received Necati and Nazmiye Terzioglu and their family and
a 150,000 ECU development grant from the Euro- by Ismail Salmanc1 and protected by Sadettin Dural
pean Union, with which we have begun building the and Mustafa Tokyasm. To all these people, who have
'dig house' on the site. So far three laboratories have accepted us so readily and made us feel so much
been constructed and partially equipped. Another part of their lives so very quickly, we express the
wing of the courtyard building has been started. warmest gratitude.
The project is extremely grateful to the Minis- Special thanks are due to the Istanbul Friends
try of Culture, General Directorate of Monuments of <;atalhOyiik, especially Re~it Ergener, for their
and Museums (especially to the Director at the time generous energy and support, and also to Paul
Professor Dr Engin Ozgen, Kenan Yurttagiil, and the MacMillan of MPR/RPM, and Tomruk and Birgiil
late Osman Ozbek whom we hold in fond memory), Ozden of Meptur.
and to their representatives for their help and sup- The setting up of the project would not have
port. In 1993 the government representative was been possible without the tireless support of David
Remzi YagCI, and in 1994 Belma Kulac;oglu, and in Shankland who steered me through many poten-
1995 Mr Ali Ondah. Generous assistance was also tially troublesome waters in the initial phases. Start-
given by Ilhan Temiszoy, Director of the Museum of ing a project in a new country must always be fraught
Anatolian Civilisations at Ankara, by Konya Muse- with problems seen and unseen. My academic guides
ums (Director Erdogan Erol) and the <;umra Belediye and mentors in Turkey have been numerous, but I
Ba~kan. An enormous amount of logistical help and would like to thank specifically for their support,
advice was provided over the years by the directors honest advice and encouragement Professors Ufuk
and staff of the British Institute of Archaeology at Esin, Mehmet Ozdogan, Refik Duru and Berna
Ankara, to whom we are extremely grateful. Alpagut.
I would like to thank personally all those who The Board of Directors of the <;atalhoyiik Re-
have contributed to the field project over its first search Trust is Sir Mark Russell, Dr John Curtis, Lady
three years, but in particular Roger and Wendy Daunt, Professor Nicholas Postgate, Mr Sevket Sabanc1,
Matthews to whom I am deeply indebted. I wish Dr Andrew Sherratt, Mr George Warren and Dr
also to thank warmly Peter Andrews, Murat Asian, Trevor Watkins. The administrator is Amanda Cox.
Douglas Baird, Adnan Baysal, Dorothee Brill, Anne It is a great pleasure to acknowledge our debts
Butler, Dagmar Cee, James Conolly, Meltem Delibasx, to James Mellaart without whose generosity and
Burghardt Detzler, Nicholas Dooley, Martin Emele, warm kindness this new undertaking would not have
Shahina Farid, Daryl Fowler, Bekir Giirdil, Naomi been possible, and to his wife Arlette and their son
Hamilton, Christine Hastorf, Nurcan Kayacan, Emine Alan. It is because of James Mellaart's initial work
Kiic;iik, Jonathan Last, Louise Martin, Frank Matero, that we are so excited to be at the site again, and it is
Theya Molleson, Maryanne Newton, Aylin Orbash, because of his present support that we are able to be
Tom Pollard, Tim Ritchey, Neil Roberts, Nerissa there now.
Russell, Orrin Shane, Colin Shell, Constance Silver, I am grateful to our reviewer for pointing out
Lothar Spree, Mirjana Stevanovic, Tom Strasser, Ali many more mistakes than there should have been in
Tiirkcan, Mehmet Uluceviz, Ibrahim Uziinoglu, an earlier version of this volume.
Nurcan Yalman, Trevor Watkins. The project is sponsored by Visa International,
At <;atalhOyiik we have the best group of work- Glaxo-Wellcome, Merko, British Airways, and Shell.

XV
Chapter 1

Re-opening ~atalhoyiik

Ian Hodder
with a note on Building Complexity by Tim Ritchey

<::atalhoyiik 1 is an example of the important paintings, sculptures, textiles, wooden and ceramic
Anatolian contribution to the development of Medi- artefacts will then be placed on display in an on-site
terranean societies. A site of this importance for the museum, enhanced by virtual reality techniques and
Mediterranean heritage needs careful conservation interactive video. Some of the paintings will also be
and presentation to the public. It poses special prob- placed back in conserved houses on the site. Part of
lems of conservation of mud brick and wall plaster, the site will be covered so that the houses are pro-
and problems of site management. tected and so that visitors can walk around a
Given its international importance and its at- 'Neolithic town'. By providing a range of visitor
tractive and evocative art, <;atalhoyiik should be experiences the full heritage potential of the site can
an exciting and popular place to visit. Between begin to be exploited.
Cappadocia and the resorts of the south coast, the The Research Trust manages the three parts of
site could attract a large number of visitors. The the project.
museums in the nearby town of Konya receive 1.3 (1) Field research, involving excavation, environmen-
million visitors a year. And yet, at present, few make tal reconstruction and regional survey. The main
the trip to <;atalhoyiik. Most guide books warn that teams are from Cambridge, Edinburgh, Liverpool,
there is little to see. The trenches of the excavations Loughborough and Aberystwyth in Britain and from
are worn away, there are no paintings to see, there is Berkeley, California. In Turkey, students and staff
no museum and few visitor facilities. Apart from a are involved from the Universities of Ankara, Istan-
very helpful pair of guards, there is little more than bul, METU, Konya and Bilkent. The overall aim is to
an eroded mound. apply the latest scientific analyses to the archaeo-
The solution at <;atalhoyiik is full-scale mod- logical material in the field and in the laboratory.
ern archaeological excavation and conservation, and (2) Conservation and restoration. The conservation is
promotion of the site for visitor access. Archaeologi- coordinated by Dr Frank Matera, Director of the
cal excavation and conservation were started in 1993 Architectural Conservation Laboratory, University
by the <;atalhoyiik Research Trust, under the aus- of Pennsylvania and Constance Silver, Preservar Inc.,
pices of the British Institute of Archaeology at An- New York. The overall aims are to conduct research
kara, with a permit from the Turkish Ministry of into methods of conserving, displaying and restor-
Culture, and in close collaboration with the Middle ing wall paintings and sculptures and other materi-
Eastern Technical University and other universities als, using the latest scientific techniques and
and research institutions in Turkey. knowledge, and to treat and restore the paintings,
The ultimate aim is to provide the Turkish Min- and to monitor their condition over time.
istry of Culture with a well planned heritage site. (3) Heritage management. The Conservation Prac-
Visitors will be able to experience the site in a number tice, London coordinates with a team from the Sci-
of ways. A conservation laboratory will be built and ence Museum, Minneapolis, in close collaboration
the latest techniques applied. Visitors will be able to with the Directorate General of Monuments and
see this conservation at work. The conserved wall Museums in Ankara and Konya Museums, in order

1
Ian Hodder

to develop the site for tourism, including roofing of mound has been fenced by the General Directorate
parts of the site, the construction of a visitor centre of Monuments and Museums and Konya Museums
and museum, the provision of pathways, parking, and a guard is present at all times. This protection
shops etc.. The overall aim is to realize the potential has prevented large-scale looting. It has also led to
of the site as an attractive and informative place to the growth of vegetation on the surface of the east-
visit. ern mound and this has helped to stabilize erosion
in unexcavated areas. The western mound, on the
Research aims other hand, has not been fenced and has been used
as a communal village threshing and winnowing
<:;atalhoyuk consists of a pair of mounds in the Konya area, while agricultural fields have encroached onto
Plain in central Turkey. It was excavated by James its lower slopes. But perhaps the main damage to
Mellaart in the early 1960s (Mellaart 1962; 1963; 1964; both mounds since the 1960s is largely unseen. A
1966; 1967). The western mound proved to be massive programme of investment in irrigation in
Chalcolithic in date, but most attention has been the surrounding region has radically transformed
directed to the eastern mound which had levels dat- the landscape into a fertile plain, green with crops
ing to the eighth and seventh millennia BC. 2 No work and trees, and busy with earth-moving machines
took place at the site between 1965 and 1993 when and concrete-mixer lorries. The digging of canals
the present project began. has aided the recovery of sites partially or completely
There were many reasons for restarting work at buried by alluvium (see Chapter 3). But it has also
<:;atalhoyuk. Some of these are 'internal', relating to substantially lowered the ground water table.
problems posed by the site itself. Others are 'exter- Mellaart abandoned his deep sounding partly be-
nal', relating to the changing context within which cause his trench filled with water at about 3-4 me-
the site is situated. tres below the level of the plain. Today, the water
The first internal problem concerns damage to table is artificially controlled but is normally greater
the site since its excavation by James Mellaart and than 10 metres below the plain. Organic materials
the British Institute of Archaeology at Ankara be- which had been preserved in waterlogged condi-
tween 1961 and 1965. Trench walls have suffered tions are gradually being lost.
extensive erosion and collapse, Neolithic walls have The second set of internal problems concerns
fallen and large new areas of plaster have been ex- issues that have been raised about the interpretation
posed. The tops of Neolithic walls have become used of the site. The enormous importance of Mellaart's
as pathways for tourists. Fortunately the eastern work was in opening up wide areas of the eastern
mound so that its overall
r potential could be evalu-
ated and its importance
glimpsed . But his work
raised a host of new ques-
tions. Only 4 per cent of
the mounds at the site has
been excavated and the full
range of modern scientific
techniques has not been
applied. Work restarted at
<:;a talhoyuk in 1993 with
the aim of answering the
questions raised by the ear-
lier excavations. The main
research questions concern
palaeoenvironmental and
palaeoeconomic reconstruc-
tion, the early development
of the site, relationships
with other sites, the de-
Figure 1.1. c;atalhoyiik East. gree of centralization and

2
Re-opening <:;:atalhoyuk

specialization of production, the degree of social dif- 1990) to the skull ritual and statuettes at 'Ain Ghazal
ferentiation and its relationship with elaborate sym- (Rollefson 1986) there is an increasing understand-
bolic behaviour. ing of elaborate symbolism associated with early
The first external issue is that scientific tech- sedentism and the development of agriculture (Bar
niques in archaeology have changed since the 1960s. Yosef & Belfer-Cohen 1992).
Not only can conservation problems be tackled more In southeast Anatolia, early sites such as Hallan
adequately now, but a wider range of archaeometric <:;:emi and basal <:;:ayonu already demonstrate a com-
techniques are available to deal with the special chal- plexity in buildings and symbolism which is also
lenges posed by a site like <:;:atalhoyuk. For example, found in PPNB-type assemblages at Nevali <:;:ori,
one of the problems posed by the site is how to Gritille and Caferhoyuk (Ozdogan 1995; Hauptmann
detect activities that took place on floors in build- 1993; Voigt 1985; Cauvin 1989). In central Anatolia,
ings, since in most cases these floors have been swept the important new excavations at A~1kh Hoyi.ik (Esin
very clean. As is clear in Chapter 15, micromorpho- 1991) have revealed a large, densely packed, a ceramic
logy now provides a mechanism for identifying the site with elaborate and differentiated buildings in
micro-traces and trample left on floors even after the millennium prior to <:;:atalhoyuk, and apparently
sweeping. Analysis of organic and inorganic residues without domestication. There is now also increased
can be applied to floors as well as to ceramics and evidence of a continuity of elaborate symbolism in
ground-stone tools. And so on. The remarkable pres- the millennium after <:;:atalhoyuk, at sites such as
ervation of a wide range of materials at <:;:atalhoyl.ik Hoyli<;ek, Kura<;ay and Ko;;khoyi.ik (Ouru 1986; 1992;
allows an enormously wide suite of methods to be Silistreli 1989). Our understanding of the Neolithic
applied, although only a few are relevant to the ini- sequence has also changed in western Anatolia and
tial stage of research reported here. Marmara, where early sites have emerged and com-
Another external issue is that our understand- mon patterns identified with southeast Europe
ing of the Neolithic in Anatolia has changed consid- (Ozdogan 1995).
erably with the discovery and excavation of a large All this new evidence indicates clearly that
number of Neolithic sites in Anatolia and the Near <:;:atalhoyuk is no longer so distinctive. It is part of a
East. Many of these show a remarkable degree of wider trend. The excavated levels are not particu-
complexity, often at early dates which stretch back larly early. There are other large and complex sites
well before <:;:a ta lhoyuk. The <:;:a talhoyuk levels exca- at this time period. But having said that, in a sense
vated in the 1960s were radiocarbon dated to be- the uniqueness of the site has only been highlighted.
tween 8450 and 7670 yr BP. In Chapter 16 in this <:;:atalhoyuk is part of a trend, and elements of its
volume Parish reports OSL
dates from mud bricks
from level 3 I 4 and level12
which correlate well with
these dates, and in Chap-
ter 17 Kuniholm and New-
ton report on their initial
findings on the <:;:atalhoyuk
dendrochronology. The
<:;:atalhoyuk dates equate
with well-established Neo-
lithic sites in the Levant. It
is now clear that even at
the earliest sedentary sites
in Anatolia and the Near
East there is elaborate sym-
bolism associated with
houses, burial rituals, figu-
rines and bulls (e.g. Cauvin
1994; Hodder 1990). From
the complex houses at
Qermez Dere (Watkins Figure 1.2. <;ntalhOyiik West.

3
Ian Hodder

assemblage are found elsewhere. But none of the the site into its environmental and historical con-
recent work has identified the same degree of com- texts (Chapters 2, 3 & 4). How could the environ-
plexity. The site, and especially the art, still stand ment around the site have sustained a population of
apart as unique achievements. this size? What impact did the inhabitants have on
What the recent discoveries have begun to al- the environment, and what is the relationship be-
low us to do is to separate our understanding of the tween the site and alluviation in the plain? If we are
site from parallels with the Levant and southeast to understand the site as some sort of centre then we
Anatolia. As more sites are studied, the distinctive- must gain some knowledge of putative subsidiary or
ness of <::atalhoyiik can begin to be situated within a dependent settlements. If we can assert with confi-
Central Anatolian sequence separate from sequences dence that no such settlements exist then we will
to the east. 'Central Anatolia should neither be con- require a different view of <::atalhoyiik. In addition
sidered as a nuclear nor as a marginal zone to the we would like to know more about the antecedents
low lands of the Near East, but as a distinct cultural of <::atalhoyiik and those developments that imme-
formation zone, developing on different lines from diately follow the decline of this particular commu-
the Near East' (Ozdogan 1995, 54). As yet, however, nity on the plain. Only intensive survey and selected
the precise character of this regional tradition re- excavation can reveal the necessary information
mains to be defined. There are intimations of distinct about any sites that may be relevant to these issues.
lithic characteristics, a distinct ceramic sequence, The survey and excavation at other sites will also
closer links with European traditions, a greater em- address questions that relate to long-term environ-
phasis on non-domesticated plants and animals, and mental impact and interaction. Questions include
less social and ritual centralization. The new work at 'Did agricultural societies emerge in alluvial settings
<::atalhoyiik must contribute both to an understand- or on the hilly flanks of these plains?', 'How dense
ing of the uniqueness of the site itself within wider was settlement in different periods?', and 'What sort
trends, and to a definition of the distinctiveness of of settlement systems existed under regional and
the regional cultural sequence within which it oc- imperial states from the Hittites to the Ottomans?'.
curs. In terms of <::atalhoyiik itself, this volume cov-
A final external issue is that the wider context ers the first three years of surface work (both site-
within which the site is situated as part of global based and regional), and the work on the 1960s'
heritage has changed significantly. Part of Mellaart' s artefacts housed in the Konya and Ankara muse-
success was to bring the site to the attention of a ums. Although the <::atalhoyiik project is designed
wide audience. In our first years of research at to run for 25 years, the first three years were planned
<::atalhoyiik we have become aware of the seemingly primarily as a pre-excavation phase. The aim was to
endless stream of special interest groups who wish find out as much as possible about the site without
to be involved in some way in the site's interpreta- excavation- working on the various surfaces of the
tion. We have been visited at the site by Goddess mounds. This strategy was necessary as a prelimi-
Tours and have been involved in debates with nary to excavation, so that excavation strategies could
ecofeminists, Gaia Theorists and Mother Goddess be developed with the fullest possible information
cults. From quite a different angle come all those available. In addition, the strategy was designed to
interested in the origins of kilims and kilim designs redress the concentration in the 1960s on very lim-
and those researching the origins of textiles. Within ited parts of the mounds. The aim was to look at the
Turkey there are different sets of interests- nation- site as a whole, at least as far as information from the
alist versus Euro-centric, fundamentalist versus secu- surface could provide an overall picture.
lar, the state versus the local community (see Chapter The surface survey involved detailed contour
18). There has been so much speculation and contro- planning (Chapter 5), geophysical prospection (Chap-
versy about the site that it seemed important to pro- ter 8), the collection of surface artefacts in 2 x 2 m
vide some new and secure evidence on which the squares on a 20m grid, and surface scraping ('top
debates might develop. scraping') in 10 x 10m units, initially on a 50 m grid
(Chapters 6 & 7).
The surfaces of this volume Another important component of the surface
work was to examine the vertical surfaces, the sec-
<::atalhoyiik has long dominated our view of the ar- tions left by Mellaart but now sadly eroded and
chaeology of the Konya plain. The surveys and exca- collapsed. The aim here was to record information
vation in the region around the site allow us to put before it was lost in further collapse, to provide a

4
Re-opening <::atalhoyiik

basis for interpreting the 1960s' data and plans, and During the 1995 field season David Shankland
to understand the stratigraphy of the site as well as began an ethnographic study of the village closest to
possible before embarking on major excavation the site, Kii<;iikkoy (Chapter 18). As well as provid-
(Chapter 14). ing information on wider aspects of the village, this
Museum study visits were made to Konya Mu- work will look at the effect of the 1960s and new
seum and to the Anatolian Civilisations Museum at work at the site on the local community. Nearly all
Ankara by various specialists on the project ~bone, our workmen come from the village and the impact
lithic, ceramic, figurine, burial). The aim of such re- of our presence is considerable. The types of knowl-
search was to obtain comparative evidence for dat- edge about the site which circulate in the village are
ing, for differences between 'shrines' and non-'shrines', very different from our own. Yet symbols from the
differences between burials and house contents etc.. site are used - for example, by the local radio sta-
It was thought essential to try this approach despite tion to advertise and identify itself. Relationships
the enormous problems which severely limit the con- with the site are complex and multi-levelled and it
clusions that can be drawn from such data (Chapters will take sensitive and detailed study to provide a
9, 10, 12 & 13). full picture.
We are trying to answer other questions in very
different ways. For example, Mirjana Stevanovic from Some initial themes
Berkeley (to be joined by Ruth Tringham in 1996)
has begun work on experimental reconstruction of Three themes will dominate interpretation in this
the buildings. In this way we can experiment with volume. The first concerns site formation processes.
ideas about why the buildings were built in the way Figures 7.19-7.22 provide cross-sections through the
they were (e.g. with separate, rather than party walls). eastern mound, indicating depth of soil above un-
Using these and a wide range of analytical tech- disturbed archaeological deposits. The greater depth
niques we hope to get closer to technological and of disturbed soil down slope is to be expected, but
architectural questions. the impact of soil movement, erosion and weather-
In January 1995 Nurcan Yalman from Istanbul ing on the surface distributions of artefacts will be of
University and David Shankland, Acting Director of major concern throughout this volume (see espe-
the British Institute of Archaeology at Ankara, un- cially Chapter 9). Despite the considerable down-
dertook a preliminary trip to villages in the Karaman, slope movement of soil and artefacts, and despite
<::umra and Konya regions to look for traditional the localized selective processes linked to slope,
architecture that might produce useful parallels for vegetational cover and later occupation, can any-
<::atalhoyiik. In mountain villages such as Taskale thing be inferred about Neolithic behaviour on the
the roofs of houses are used as courtyards by houses mound surface? Further study of sedimentological
higher up the slope, but the houses are mainly made processes on the surface of the mounds will be car-
of stone. To get parallels for mud brick architecture, ried out in future by excavating transects across and
villages closer at hand are more useful and during down the slopes of the mound and out into the sur-
the 1995 summer field season N mean Yalman, helped rounding alluvium. But for the moment a range of
by Mirjana Stevanovic, began more detailed study of evidence will be discussed in this volume which
some villages close to <::umra (especially Uriinlii, relates to the formation of the site's surface.
Tiirkmen Camili and Gokhoyiik). Attention focused A second theme concerns the spatial and social
on methods of construction and use of mud brick organization of the eastern mound. For example,
and plaster walls. Some houses still have 'platforms' there has been considerable ambiguity about whether
and niches like those on the site, and the walls are the site is a village, town, urban centre or city (e.g.
richly decorated- not with paintings but with car- Mellaart 1967; Renfrew 1984, 90). Use of such terms
pets, kilims and other textiles. The parallels with the depends on the definitions used but also on the in-
Neolithic housing are obvious but so are the differ- terpretation of the activities taking place at the site.
ences - the plaster looks the same but has different We will, in this volume, use the information from
constituents, for example. A careful programme of the surface studies in order to address the size and
sampling bricks, mortars and plasters has begun in internal complexity of the site.
order to provide comparisons with the site architec- Another aspect of the second theme concerns
ture, and the potential for providing information Mellaart's (1967) suggestion that he may have exca-
relevant to the site, even if only by way of contrast, is vated a 'priestly quarter' containing 'shrines'. We
considerable. needed to discover whether elaborate ritual buildings

5
Ian Hodder

were found elsewhere on the site or whether they correspond to clear, separate horizons of destruction
were restricted to the southwest corner excavated by and rebuilding? Some of the levels had been de-
Mellaart. But in addition, we needed to evaluate the stroyed by fire, but were these general conflagra-
very concept of 'shrine'. The terms shrine or sanctu- tions or were they controlled, small-scale and
ary or cult building are notoriously difficult to de- piece-meal? Was the growth of the settlement or-
fine (Renfrew 1985), and certainly Mellaart had used ganic or structured? Do individual buildings go
a variety of criteria such as amount of sculpture and through changes of use, or is a 'shrine' always a
painting, numbers of burials and platforms etc .. We 'shrine'? Such questions necessitate detailed recon-
wanted to use contextual information in order to struction of stratigraphic sequences both at the gen-
evaluate building use. That is, we wanted to use eral level and in terms of detailed sequences of
variation between buildings on a number of differ- activities on replastered floors (Chapters 14 & 15).
ent dimensions in order to see how they might best
be differentiated without making a priori assump- Superficial meanings
tions about what a 'shrine' should look like. Thus we
wanted to compare variation in ceramics and lithics The research at <;:atalhoyiik over the long term will
etc., and gradually build up an interpretation of the allow the development of methodologies which
buildings depending on the nature of that variation. might be termed 'postprocessual' (Hodder 1991;
Our understanding of ritual specialization at 1992). The current phase in which full-scale excava-
<;:atalhoyiik would also throw light on the overall tion is being developed (from the 1996 season on) is
issue of the level of complexity of the site. Were we allowing experimentation with news ways of record-
dealing with ritual specialists in priestly quarters, or ing and working in the field. The key to this new
simply domestic houses with varying degrees of in- approach is contextuality or non-fixity. The aim is to
ternal decoration and elaboration? define typologies and terminologies not on a priori
As a starting point we needed to categorize grounds but on the basis of both general and local
buildings without using loaded terms such as 'shrine' information. Types and terms are seen as being sen-
and 'non-shrine'. The buildings in each level were sitive to context, capable of redefinition from differ-
quantified in terms of their architectural complexity ent perspectives. Contextuality leads to multivocality,
by counting numbers of platforms or bins, benches, interactivity and reflexivity.
pillars or posts, pits or basins, fire installations (ov- Such an approach is entirely dependent on hav-
ens, hearths etc.), mouldings, and subdivisions or ing contextual information- that is, houses, rooms,
rooms (termed 'spaces'). The number of paintings pits, burials in which things are associated. Indeed,
was not thought to be a reliable measure since many the main reason that I wanted to restart work at
paintings have collapsed into room fills. In addition, <;:atalhoyiik was that it appeared to offer rich contex-
our own research has begun to show that paintings tual patterning and the potential for 'thick' or 'deep'
might not be apparent on plaster surfaces but might description (Hodder 1991). The contextual associa-
be hidden within the multiple layers of plaster. Thus tions should allow insight into particular cultural
only the presence or absence of paintings was re- meanings. In Chapter 10 Conolly has begun to be
corded. When this information is plotted by level, reflexive about the terminologies normally imposed
using Mellaart's data, the results in Figures 1.3 to 1.7 in lithic analysis. But working on either the horizon-
are obtained (information collated by T. Ritchey and tal or vertical surfaces of the mound provides little
T. Pollard; see Note by T. Ritchey). As is clear from or no contextual information beyond that produced
all the figures, no clear break can be discerned be- by post-depositional processes. The erosional proc-
tween more or less complex buildings, and only ap- esses have erased or 'naturalized' the specific cul-
proximate correlations are obtained between a tural meanings and forced us to use universal and
building's position on the complexity graphs and its decontextualized methodologies. On the surface, we
labelling by Mellaart as 'shrine' or 'non-shrine'. Thus, were reduced for the most part to using type series
right from the start, the differentiation of buildings and terminologies developed at other sites or in ab-
is problematized. Understanding of the range of ac- straction. Or else we had to rely on the contexts (e.g.
tivities carried out in buildings of different architec- levels and buildings) provided by the 1960s' archive
tural complexity will be one of the recurring themes which were themselves built within very different
of this volume. questions and interests to our own and which pro-
A third theme concerns temporal change. To vided little of the sensitivity for which we were
what extent do the levels identified by Mellaart searching.

6
Re-opening <::atalhOyiik

The title of this volume thus refers at one level intra-level data are comparable for the most part,
simply to the fact that our research between 1993 inter-level comparison is not as secure, and these
and 1995 has concentrated on various surfaces of the should be seen as relative quantities.
mounds and of the region. But at another level it Information which was gathered included the
refers to our inability at this stage to go beyond the number of platforms, benches, pillars/posts, pits or
surface meanings of things in order to understand basins, ovens or hearths, mouldings, room spaces
situated and specific meanings at a greater interpre- and whether paintings were present in the building.
tive depth. In this volume we have been reduced, All these numbers, along with location information
because of the relative lack of contextual informa- were fed into a Microsoft Access database and dis-
tion, to orthodox methods and established descrip- played in an Excel spreadsheet. The goal of this analy-
tions. Such methods can easily be criticized as being sis was to provide a complexity scale which could be
little more than extensions of our own assumptions compared with artefact concentrations in the build-
and biases. Theories and methods replace each other ings. The scale was created as a stacked bar histo-
in a continual chain of signification, an endless play gram, with each feature type scaling to the number
of difference unrelated to anything beyond 'the sur- of features of that category found in the plans and in
face'. The play of signification can be confronted by the text of the publications. The presence of paint-
contextuality - by boundaries which localize and ings is indicated by the lighter single bar on top, and
specify meanings. Our theories and methods need to buildings identified by Mellaart as 'shrines' are de-
be sensitive to context, but until excavation has re- noted in the building numbers along the bottom of
commenced at <::atalhoyiik we are forced to follow the graphs.
largely traditional methods. In levels V, VIA and VIB, pillars seem to be a
Where multivocality, interactivity and reflexiv- better marker of complexity than paintings. But lower
ity can and should begin at this initial phase is in our down, in level VII, the connection between pillars
interactions with the world in which we are working and complexity begins to be lost. Instead, the number
in the Konya region and Turkey. Chapter 18 by David of platforms seems to be a good indicator of the
Shankland begins a process of evaluation of the im- more complex buildings.
pact of the archaeological work on the local commu- In the graphs there are no clear breaks between
nity. This research by a social anthropologist aims to more and less elaborate buildings. Overall, the de-
move beyond generalized statements about the rela- marcation between 'shrine' and non-'shrine' as pro-
tionship between archaeology and society towards a posed by Mellaart cannot be substantiated by
more contextualized understanding. The research architectural features alone. Except in levels where
also allows us to begin to respond and interact, de- there were too few buildings to get a good indica-
veloping a reflexivity in the face of the multiple in- tion, the clustering of 'shrines' does occur at the
terpretations of <::atalhoyiik. upper end of the complexity histograms; however,
this does not mean that 'shrines' can be easily cat-
Note: building complexity egorized since there is no obvious break in the ranked
by Tim Ritchey sequence of variation in architectural complexity.
There are essentially two issues to deal with. First, it
The data used to construct Figures 1.3-1.7 were gath- is not expected that there actually be a cut-off point
ered from the preliminary excavation reports and so that buildings can be placed into 'shrine' and
book published by James Mellaart (1962; 1963; 1964; non-'shrine' categories. The idea of 'shrine' versus
1966; 1967). While not ideal, the plans (redrawn here non-'shrine' is seen to be a much more blurry con-
as Figs. 1.8-1.17) provided the information neces- cept both theoretically and methodologically. Sec-
sary for quantifying the types of architectural fea- ond, even if we did expect a clear cut-off point
tures which were present in each building. In all, between the two, it would not be realistic to expect a
eleven of the twelve levels had plans which could be single variable such as architectural complexity to
used to draw information for the study. However, provide all the information needed for determining
the quality of plan varied between levels, and in to which set a building belonged. It is the architec-
some cases across levels where different plans were tural features in relation to the other artefacts which
found in different publications covering disparate will provide the finer scale analysis needed to un-
regions of the excavation trench. Therefore, while derstand variation between buildings.

7
18 paintings?

16 El no of spaces
14
"'~ D mouldings
;:s 12
-:o01
10 D fire installation
""'0
""'
.... 8
01 • pit/basin
.rl
e;:s 6
z II pillar I post
4
2 D bench
0
N Ol ~ (") '<!" 0 ~
• platform / bin
~ C':! "! Ltl "': ::! C':! Ltl ~
"! "': ~
f'-: ~

a:i <i <i a:i a:i a:i a:i <i "' "' <(
<i <i <i <i <i <i <i <i <i <i <i <i <i
ui ui ui
ui
Buildings

Figure 1.3. Building complexity for levels I-III. In this and the following figures, the prefix S was given by Mellaart to buildings he identified as 'shrines'. >--<
Pl
~
Other letter prefixes refer to areas of the eastern mound as defined by Mellaart.
00 ::c
0
0..
0..
(!)
>-t

25

20 • paintings?

"'~ no of spaces
2 15
"
~ Dmoldings
~
0
....
<II ml fire installation
..0
e 10
=
z • pit/basin

5 1111 pillar/post

Dbe nch
0
M ..,. M ..,. M ..,. M ..,. • platform / bin
~ M
2' "'
2' "'2' "'2' "' 2'"' "'
2' 2' 2'"
2' "' > >" ""z z
"" > > >
"' "'> ~ ~ "'> ""'> "'> >" >~ >" "'> >"' "'> > > >
w w w w w w w w w w w w w w >w >
w w w > uj
>
uj w w w w > uj
>
uj
w w w w w w .. w w w ..
ui "' " ui ui " .. ui ui ui ui ui ui
ui "' "'
Buildings

Figure 1.4. Building complexity for levels IV-V.


45

40

35

El paintings?
30
II no of spaces

0 moldings
25
Cl fire installation

• pit/basin
20
ID pillar I post
D bench
15
• platform/bin

10

:;o
roI
0
'"0
ro
:::s

crq
\0 Figure 1.5. Building complexity for level VIA. .(")
PJ
..,.
45
PJ
s:
Q:
40 '-<:
C:
7\'
35
• paintings?

• no of spaces
30
~

~
Dmoldings
=
] 25 Gl fire installation
0
Ill pit/ basin
i!e 20

z= Ill pillar /post


15 O be nch

Ill platform / bin


10

0 la ••••••••••••.•.•.•.•••.•.•. a.~.a.l.a.~.;.l.l.!.!.~.~.!.~
~ ~ ; ~
Buildings

Figure 1.6. Building complexity for level VIB.


25,---------------------------------------------------------------------------------------·
paintings?

20 no of spaces

Ill
"' Dmoldings
....
::s
-:.Ill 15
.... D fire installation
....0
....
Ill •pit/basin
..0
10
e::s
z II pillar I post

5 . Obench

• platform/bin

o ~~~~~~1 1 111i~~~~~8~1~1~1~Jjll ~
M M m 0
M ~ ~ ~ ~ !" ;;; N ~
~ ~ ~
~ ~ ~ ~ ~ x
gg~~ ~~~~ ~~~g ~g~~ g~~~ ~gg~ > > > > > >
uj
> > >
fli
> > > > > > > > > > > > > > > > > > >
"' '" '" '" "' '" '" "' '" "' '" "'
Buildings ......
Figure 1.7. Building complexity for levels VII-XI. Pol
::l
...... :I:
0 0
0..
0..
(!)
"1

N «1

10m

Figure 1.8. Redrawn plan of Mellaart' s level I.


Re-opening C::a talhoyuk

N '1

10m

Figure 1.9. Redrawn plan of Mellaart's level IT.

N '1

Figure 1.10. Redrawn plan of Mel!aart's level Ill.

11
Ian Hodder

......
.;

-"""\\,,
I \
\ ~
'I
\J

12
Re-o~pe=nmg
·~=-----
<:;:atalhoyuk ---------

,.:::·.

t
z
13
N-~ >-<
I))
- ::l
>--' ::c
0
*" lOrn 0...
0...
(1)
...,

Figure 1.13. Redrawn plan of Mellaart's level VIA.


N --
.......
Ul

::::;'
-or
Q:
'<
C:
7\

Figure 1.14. Redrawn plan of Mellaart's level VIB.


N..........-~
--
lOrn

OJ
-
::l
....... ::r:
a-- 0
0..
0..
(!)
"1

Figure 1.15. Redrawn plan of Mel/aart' s level VII.


Re-opening <.;:atalhoyi.ik

,,,, N..........--4~
'•
//

10m

N ,
10m

Figure 1.16. Redrawn plan of Mellaart's level VIII.

Figure 1.17. Redrawn plan of Mellaart's levels IX (top),


XI and XII.
Notes
References
1. A w id e range of spellings for <.;:atalhoyi.ik is found
in the literature. The spellin g most commonly Bar-Yosef, 0. & A. Belfer-Co hen, 1992. From foraging to
found in the international literature (<.;:atal farming in the Mediterranean Levant, in Tran sitions
Hi.iyi.ik) has become relatively rare in Turkey to Agricultu re in Prehistory, eds. A.B. Grebauer &
and the preferred spelling used here is that used T.D. Price. Madison (WI): Prehistory Press, 21-48.
on official permit documents by the present Turk- Cauvin, ]., 1989. La stratigraphie de Cafer Hoyuk-Est
(Turqu ie) et les origin es du PPNB du Taurus.
ish administration. However, other spellin gs,
Pnleorient 15, 75-85.
such as <.;:atal Hoyi.ik, are commonly found in Cauv in, J. , 1994. Naissnnce des divinites. Naissance de
Turkey. l'ngriw lture. Paris: CNRS.
Duru, R., 1986. Kurucay Hoyiigii Kazilari, 1984 Calisma
2. sc dates used in this volume refer to calibrated raporu. Belleten 50, 247-59.
dates; BP to radiocarbon years. Duru, R. , 1992. Hoyiicek Ka zilari 1989. Belleten 61, 551-66.

17
Ian Hodder

Esin, U., 1991. Salvage excavations at the pre-pottery site Mellaart, J., 1967. (:atal Hiiyiik: a Neolithic Town in Anatolia.
of A~nkh Hoyuk in Central Anatolia. Anatolica 17, London: Thames & Hudson.
123-74. Ozdogan, M., 1995. Neolithic in Turkey: the status of re-
Hauptmann, H., 1993. Ein Kultgebaude in Nevali Cori, in search, in Readings in Prehistory. Studies Presented to
Between the Rivers and Over the Mountains. Archaeologia Halet (:ambel. Istanbul: University of Istanbul, 41-
Anatolica et Mesopotamica, ed. M. Frangipane. Rome: 60.
University of Rome, 37-69. Ozdogpn, M. & A. Ozdogpn, 1990. Cayonu. A conspectus
Hodder, I., 1990. The Domestication of Europe. Oxford: of recent work, in Prehistoire du Levant II, eds. 0.
Blackwell. Aurenche & M.C. Cauvin. Paris: CNRS, 387-96.
Hodder, I., 1991. Reading the Past. Cambridge: Cambridge Renfrew, A.C., 1984. Approaches to Social Archaeology. Ed-
University Press. inburgh: Edinburgh University Press.
Hodder, I., 1992. Theory and Practice in Archaeology. Lon- Renfrew, A.C., 1985. The Archaeology of Cult. London:
don: Routledge. Thames & Hudson.
Mellaart, J., 1962. Excavations at <;atal Huyuk, first pre- Rollefson, G.O., 1986. Neolithic' Ain Ghazal (Jordan): ritual
liminary report, 1961. Anatolian Studies 12,41-65. and ceremony II. Paleorient 12, 45-52.
Mellaart, J., 1963. Excavations at <;a tal Huyuk, second Silistreli, U., 1989. Kosk Hoyuk Figurin ve Heykecikleri.
preliminary report, 1962. Anatolian Studies 13,43-103. Belleten 53,497-504.
Mellaart, J., 1964. Excavations at <;atal Huyuk, third pre- Voigt, M.M., 1985. Village on the Euphrates: excavations
liminary report, 1963. Anatolian Studies 14, 39-119. at Neolithic Gritille in Turkey. Expedition 27, 10-24.
Mellaart, J., 1966. Excavations at <;a tal Hiiyiik, fourth pre- Watkins, T., 1990. The origins of house and home? World
liminary report, 1965. Anatolian Studies 16,15-191. Archaeology 21, 336-47.

18
Chapter 2

Preliminary Results of Geoarchaeological


Investigations at ~atalhoyiik

Neil Roberts, Peter Boyer & Romola Parish

Geomorphological and geoarchaeological studies Background to the study


have made an important contribution to understand-
ing archaeological records at both landscape and <:;atalhoyiik lies on the Konya plain on the southern
site-specific scales, and in a wide range of contexts edge of the Anatolian plateau at an elevation of c.
(Davidson & Shackley 1976; Vita-Finzi 1978; Rapp & 1000 m asl. The climate of the plain is semi-arid, with
Gifford 1985). In the Eastern Mediterranean region, average precipitation below 300 mm p.a., and it ex-
they have included sediment-based investigations periences a substantial seasonal temperature range,
of cave sequences (e.g. Wilkinson & Duhon 1990), winter temperatures being around freezing and mean
past and present coastlines (e.g. Kraft et al. 1980), summer temperatures >20°C. The plain and its im-
alluvial histories (e.g. van Andel et al. 1995) and site mediate surroundings are treeless, except along river
formation processes (e.g. Courty et al. 1989). courses. By contrast, the watershed catchment area
This chapter relates to preliminary results of receives up to 1000 mm p.a. precipitation and is, or
geomorphological investigations carried out at, and was until recent deforestation, well-wooded. Most
in the immediate vicinity of, <:;atalhoyiik during a of this is in the Taurus mountains south and west of
preliminary reconnaissance in 1993 and in substan- Konya, where peaks reach 3000 m asl.
tial fieldwork seasons in 1994 and 1995. The work Hydrologically, Konya is a closed (i.e. non-out-
forms part of the wider KOP AL (KOnya basin let) basin, except for some karstic sink-holes (diiden).
PALaeoenvironmental research) programme, whose Most of the plain is today dry, although soils were
aim is to investigate the late Quaternary environ- waterlogged for much of the year prior to twentieth-
mental history of the Konya plain in relation to its century river regulation and irrigation. Shoreline
human occupation and, in particular, to the origin of depositional landforms and wave-cut cliffs are evi-
Neolithic agriculture. Our particular concern is to dence that a shallow but extensive lake formerly
examine the relations between changes in the natu- occupied the floor of the plain (Erol1978). As rivers
ral environment (climate, vegetation, geomorph- and wadis enter the Konya plain, most of their sedi-
ology, etc.) and the domestication of plants and ment load is deposited close to the basin edge as fan-
animals during a time period when global climate shaped masses of alluvium (Fig. 2.1). The largest of
was changing from glacial to interglacial conditions. the alluvial fans has been deposited by the <:;ar~mba
Ultimately, it is to be hoped that the current research river, which covers 474 km2• Although broadly fan-
programmes will lead to a reconstruction of an inte- shaped, its hydro-geomorphological characteristics
grated record of Holocene landscape change, both are today more akin to an alluvial floodplain than an
natural and cultural. The Konya plain is ideally suited alluvial-fan environment. Overbank deposition of
for this task, because rich archaeological and silts and clays over very low slopes has pushed these
palaeoenvironmental evidence is potentially avail- alluvial features towards the centre of the plain, on
able from sites in close proximity to each other. top of the lacustrine beds of palaeo-lake Konya. A

19
Neil Roberts, Peter Boyer & Romola Parish

distributary of the ~ar~mba river ran between the geo-archaeological work at the aceramic Neolithic
mounds of ~atalhoyiik east and west prior to recent site of Can Hasan III (Roberts 1991). This provided
artificial modification of the drainage system. evidence of changing late Quaternary sedimentary
The ~ar~mba alluvial fan is dotted with great regimes on the fan, tied to an archaeological chro-
numbers of ancient settlement mounds (hoyiik in nology. The mound of Can Hasan III was excavated
Turkish), many of which appear to have been sub- in 1969-70 by the British Institute of Archaeology at
ject to burial by Holocene alluviation. The original Ankara (French et al. 1972). Today it stands only 2.5 m
excavation by Mellaart (1964, 41ff.) also suggested above the general fan surface, but excavation uncov-
that this was the case at ~atalhoyiik. A deep sound- ered a greater depth of archaeological deposits bur-
ing below the floor level of an excavated room at ied by Holocene alluvium. Excavation revealed 4 m
~atalhoyiik was made during the 1963 excavations. of cultural deposit, and archaeological levels below
It penetrated to a depth of about five metres below this were sampled by a Dachnowsky corer at the
the modern ground surface, and Mellaart recorded time of excavation. A subsequent programme of hand
archaeological material as observable to the base of augering across the fan indicated that the Holocene
this (i.e. 'natural' was not reached). However, a dif- alluvium is fine-grained and moderately sorted, and
ferent conclusion was reached by Cohen (1970) who is underlain by a coarse-grained and/ or poorly sorted
claimed that there had been no post-Neolithic lower alluvium of late Pleistocene age. On the distal
alluviation in the area and that ~atalhoyiik was es- part of the fan the two alluvial units are separated by
tablished at the same level as the modern land sur- a wedge of lacustrine marl and silty sand of prob-
face. Cohen went on to propose that the waters of able deltaic origin. There was thus a sharp contrast
the Pleistocene Konya lake had receded only a short between the alluvial regime of the Ibrala fan during
time before the establishment of ~atalhoyiik, and the Holocene and in the late Pleistocene, which ap-
that, following rapid alluviation at the start of the peared to be related to major climate changes at the
Holocene, the early Neolithic farmers exploited newly- end of the last glaciation.
available soil and water resources - a kind of geo- In summary, Cohen's (1970) proposal that the
logical opportunism. The site of ~atalhoyiik certainly shrinkage of Pleistocene Lake Konya led to latent
lies on the former lake bed, and a prominent east- agriculturalists exploiting newly available soil and
west sand ridge (a beach spit of the palaeo-lake) is water resources, has not been supported by subse-
found about 6 km south of the mound (Fig. 2.1). quent work. Data do suggest, however, that impor-
Subsequent 14C dating of mollusc shells from tant environmental changes occurred in south central
this ridge and from other shoreline deposits (Roberts Anatolia prior to the Neolithic occupation. For ex-
1983) has indicated that the main fall in lake levels ample, an important shift in alluvial fan sedimenta-
occurred well before the end of the Pleistocene, prob- tion occurred prior to the appearance of the Neolithic
ably c. 17,000-16,000 14C yr bp (c. 19,000-18,000 Cal. site of Can Hasan III (Roberts 1991). On the other
BP). The site of ~atalhoyiik is therefore likely to have hand, the Can Hasan study did not have high chrono-
been free of permanent water through most of the logical resolution, and it would be premature to as-
deglacial climatic transition period, although it may sume the results from there would apply elsewhere
have lain at the southern edge of an extensive marsh in the Konya plain.
or shallow lake which was centred around Yarma The current field research programme at
during the terminal Pleistocene; (one of the 'second- ~atalhoyiik has been framed in the context of a
ary depressions' (Roberts 1983)). One of the authors number of objectives:
(NR) also undertook a brief assessment of alluviation • detailed measurement of the true depth of post-
at ~atalhoyiik in 1978, with a series of three hand- occupation alluvium and of cultural deposits at
augered cores (Roberts 1982), in an attempt to re- ~atalhoyiik, which should be a logical precur-
solve the differing views of Mellaart and Cohen. sor to new archaeological excavation;
This showed clearly that the site had been buried by • evaluation of the relationship between site
post-occupation alluvium; but a proper assessment stratigraphy and associated mound formation
of the depth of burial and reconstruction of the early processes, including post-occupation erosion,
Holocene landscape was not possible at that time flooding and alluviation;
without re-excavation of ~atalhoyiik. • assessment of the soil and water resource base
However, the lithostratigraphic sequence from available to the Neolithic population. Is the mod-
another of the fans in the Konya plain, that at Ibrala ern distribution of soil types (de Meester 1970),
near Karaman, was investigated in association with for example, a reasonable guide to those in the

20
IZ
">f11fiQL(J VJV:) jo UOLJ VJO ] puu ' UIVJd uliuo)[ WJJ5Jm Jlfl jo 5JU3ut! PCJ5 fiJVLUCJJVI1Q CJJV1 TZ am~!::l
SJl1Al~lU! WQQp l\1 SlllOlUO;)
Sl1Al1Jd pu11
pu11s Jll1W ~}[11'1
s~qsnw 'S~)[\1'1
j:..:::::. :j t:: j
0
lWI Sl 0
'{L--..r-~ - ----.....-----·
dl!W JO lr.llV D
~
.
{__ 11Jll:l(UV •
suoq e~ q saA UJ reJ~~oroae q ::)Jeoa~
Neil Roberts, Peter Boyer & Romola Parish

past (cf. French 1970)? Did <;atalhoyiik lie on the Core extrusion and sub-sampling, data logging,
<;an]amba alluvial fan at all, or was it located on sediment description and preliminary analyses were
the marl plain? Where was the river located at undertaken in a field laboratory set up in <;umra.
that time?; These included scanning of wet preparation raw sedi-
• investigation of the impact of geomorphological ment slides under x100 and x400 magnification un-
change on the development of Neolithic agri- der a transmitted light binocular microscope, and
culture. Did hydro-geomorphological changes measurement of whole core magnetic susceptibility
at around the time of the Pleistocene-Holocene using a loop sensor and a Bartington MS2 meter
transition lead to the creation of alluvial soils linked to a portable computer. Magnetic data were
suitable for agriculture, and was this associated imported into Excel spreadsheets. After fieldwork,
with a switch in archaeological site locations sediment samples were stored in core boxes and
from the margins of the plain to the alluvial returned to the UK for laboratory analysis.
fans?; In 1995, topographic data were obtained by sur-
• assistance in understanding the functioning of veying using a high precision Leica system 200 GPS
<;atalhoyiik's economy, including agricultural (Global Positioning System) and associated SKI and
processes such as irrigation and manuring, and LISCAD software, operated by Dr David Twigg. This
the sourcing of raw materials for ceramic and proved highly successful in providing horizontal and
mud brick manufacture. vertical data over a range of 10 x 20 km, with a
While in no sense arguing for environmental precision of +I- 1 em. In the 1993 and 1994 field
determination, testing of these ideas will indicate if seasons, topographic data were derived from the
newly available environmental resources were a nec- main <;atalhoyiik Total Station survey carried out by
essary pre-condition for the successful adoption of Tom Pollard.
farming as a mode of production in south central
Turkey. Laboratory
Four analytical measures are used here to character-
Methods ize sediments, namely organic matter and carbonate
content, magnetic susceptibility and particle size dis-
Field tribution. Fluctuations in these parameters reflect
Sediments of lacustrine, alluvial and archaeological changing sediment source, transport and depositional
origin were studied and sampled mainly by coring, characteristics. Organic matter content, for example,
although some deposits (palaeochannel, Pleistocene is typically high in sediments deposited in anaerobic
lake shorelines) were exposed as sections in quarries standing water, while in terrestrial sediments high
and dry irrigation ditches; (pumping has caused a concentrations may reflect organic matter accumula-
dramatic lowering of the water table in the Konya tion during soil formation processes (Stein 1984;
plain since the mid-1980s). Coring mainly employed 1992). Authigenic lake carbonates are associated with
an Eijkelkamp corer and Cobra motor, with exchange- calcareous organisms such as ostracods, or precipi-
able sampling heads 100 em in length but of varying tation out of solution. Organic matter and carbonate
diameters. Sediments from open gouge heads were content of the sediment samples were calculated us-
sub-sampled in the field into plastic bags; those from ing the loss on ignition technique, at 550°C for two
lined sampling heads were either extruded into hours and 950°C for one hour respectively, as out-
clingfilm inside downpipe cut lengthways or left in lined by Dean (1974). Assuming that the carbonate
the core liner to be opened in the UK. Sediment cores in each sample was purely calcium carbonate, the
taken with this system retained good stratigraphic calcium carbonate equivalent can be calculated by
integrity as indicated by features such as horizontal dividing the C02 content by 0.44 (the fraction of C02
layering, and involved minimal sample contamina- in CaC03) (Stein 1984). When adjusted, the CaC03
tion. During the 1993 reconnaisance season, two ex- (equiv.) figures for alluvial sediments at <;atalhoyiik
ploratory cores were taken with a hand auger and a are very similar to those reported by Roberts (1982)
25 em-long 'Dutch' sampling head. While providing who used a gas-volumetric method for measurement.
a useful initial guide, this coring system does not Magnetic susceptibility is a measure of the re-
provide reliable stratigraphic detail, and the two ex- action to an induced magnetic source (Dearing 1994)
ploratory cores have therefore been replicated in sub- and can be used to differentiate between deposits
sequent field seasons. The results of the 1993 cores and provide some indication of their mineral com-
will only be referred to briefly in this report. position, and thus source. When measuring using a

22
Geoarchaeological Investigations

magnetic susceptibility meter, a figure is given which the presence of pollen and diatoms, predominantly
is a measure of the ratio between the magnetization from marl deposits but also from fragments of
produced in a sample and the magnetic field im- ceramics, mud brick and some other sediments. As-
parted. This is termed the volume magnetic suscep- semblages from the marl could provide environmen-
tibility, represented by the Greek letter K, and is tal data for the period of deposition, and assemblages
measured on a scale of J0-5• To allow for different from ceramics could be used to source clays utilized
densities between samples, a measure of mass spe- (Battarbee 1988).
cific magnetic susceptibility (c) may be made where
the mass and volume of the sample are known: Radiometric dating
x=K+p Good chronological control is obviously a pre-requi-
where p = bulk density of sample (kg m-3 ). site for accurate correlation between on- and off-site
This provides a more accurate figure, particularly stratigraphic sequences. It was known from previ-
useful in determining subtle trends within groups of ous investigations (Roberts 1982) that most alluvial
readings. For this reason this was the figure calcu- deposits at <::atalhoyiik contained sparse (or derived)
lated in this work. Mass specific magnetic suscepti- organic matter, and that radiocarbon dating was
bility is expressed in units of m 3 kg-1• A significant therefore unlikely to be always applicable. Conse-
value is one of 0.1 m 3 kg-1, because (with a few excep- quently a programme of luminescence dating has
tions) samples with readings below this tend to be been been built into the research programme, and is
dominated by low susceptibility paramagnetic min- being carried out by one of the authors (RP) at the
erals such as carbonates and silicates, whereas sam- University of Sussex Dating Laboratory. The OSL
ples with readings above this are dominated by much (Optically Stimulated Luminescence) dates reported
higher susceptibility ferrimagnetic iron minerals here derive from samples obtained during the 1994
(Dearing 1994). This value is shown on Figures 2.3-2.6. field season. A number of further samples from 1995
The data presented here are based on single are currently in the process of luminescence and
samples, not on whole core measurements. Samples radiocarbon dating, and the chronology outlined here
were prepared by oven drying at 35°C for a mini- should be regarded as tentative and provisional.
mum of 72 hours, and were crushed using a pestle Luminescence dating is a technique proven for
and mortar into a fine powder, or until all of the inorganic sediments up to 200 ka (Aitken 1994). The
particles were disaggregated. Measurements were technique is based on the measurement of trapped
made using a Bartington Instruments MS2 magnetic charge held in defects of quartz and feldspar crys-
susceptibility meter with a low frequency sensor. tals. Exposure of the crystals to light or heat detraps
Repeat readings were taken on a high resolution 0.1 some of the charge which recombines with 'holes'
K interval scale on weighed and filled 10 ml plastic with the emission of light (luminescence). This emp-
sample pots. This allowed the calculation of mean tying of the crystals is known as the 'zeroing event'.
magnetic susceptibility taking account of background After burial, the crystals accumulate trapped charge
interference. with time; the charge is emitted as a product of the
Particle size analysis utilized the same samples decay of naturally occurring radioisotopes of U, Th
prepared for magnetic susceptibility work, although and K. The event that is dated is the last exposure to
because of the time taken for fine particle size analy- sunlight (or heating) of the crystals. The age limit of
sis, only selected sediments were analyzed. This the dating technique is determined by the saturation
involved initially dry sieving the dried and dis- of the crystals with trapped charge. The age is deter-
aggregated samples through a series of sieves of mined by the Equivalent Dose (ED) I Annual Dose
mesh sizes, 8 mm, 4 mm, 2 mm, 1 mm, 500 j..tm and Rate. The former is determined by laboratory meas-
250 j..tm. The residue held in each sieve was then urement and represents the radiation dose adminis-
weighed and the proportion of material greater than tered by an artificial source which gives rise to a
250 j..tm calculated. The fine sediment passing through signal of equivalent intensity to that measured for
the 250 j..tm sieve was analyzed using a Micromeritics the 'natural' signal (i.e. that emitted from the sample
Sedigraph. Samples were measured down to parti- when it is first measured). The latter was determined
cle sizes of between 2 and 0.2 j..tm, according to the by laboratory measurement of the U, Th and K con-
nature of the material. The results from the sedigraph tent or the alpha, beta, gamma and cosmic activity of
were combined with the manually sieved data to pro- the bulk sample.
duce a full particle size distribution for each sample. As noted above, samples were taken in the field
A number of sub-samples were taken to test for in two different ways; namely, by coring and from

23
Neil Roberts, Peter Boyer & Romola Parish

open sections. In the former case, an opaque plastic remove the outer layer affected by external alpha
liner (5 em diameter) was inserted into the sampling dose, and then cleaned in HCl, distilled water and
head, from which a section was cut off in the labora- acetone, dried, resieved and washed again. Alu-
tory. For open sections, short (15 em) lengths of minium discs were prepared of the quartz fraction
opaque plastic tubing were inserted into the sedi- by adhesion using silicone spray.
ment unit, removed and wrapped up in black plas- The ED was determined by the additive dose
tic. For both section and core samples, the end 3-4 technique. Four discs were measured of the natural
em at each end of the sample tubes, which might signal, and four discs for each of 5 additive dose
have been exposed to light during sampling, were points. Irradiation was by a manually-operated Sr90
removed in the luminescence laboratory and used source, delivering 14.3 Gy I min. Samples were meas-
for dosimetry measurements. The central part of each ured in a Rise automated reader, using a preheat of
sample was unexposed to light during sampling and 220°C for 5 minutes for each disc and normalization
used for OSL measurements. by a 0.5 second exposure to the green light source.
All samples, except CP6c, were fine-grained The polymineralic fine grain samples, and the coarse
preparations. Preparations were carried out in a dark grain quartz fraction of CP6c, were measured using
room lit by dimmed red lights. Samples were pro- 30 seconds exposure to a green halogen lamp to stimu-
tected from any unnecessary or extended exposure late the OSL signal. Detection was by a filtered photo-
to this light source. The procedure followed entailed multiplier tube with a combination of GG420 blue/
deflocculation using Calgon, the removal of carbon- green filters for excitation and U340 for detection.
ates using concentrated HCl and the removal of or- The dose rate (Table 2.1) was determined by
ganic material using Hydrogen Peroxide. Samples alpha counting and thick source beta counting to
were then washed through a 43 !liD sieve with dis- determine the U and Th contents and the total beta
tilled water, retaining the fraction which passed activity. The K% content was derived from the U
through the sieve. This was centrifuged as 2000 r.p.m. and Th ppm (Nambi & Aitken 1986). For all fine
for 5 minutes and the supernatant was decanted. grain samples, the estimated a-value used was
The sample was resuspended in distilled water, thor- 0.075±0.005. Water content was measured as a per-
oughly mixed and centrifuged again. This proce- centage of dry weight. For CP6c, the mean grain size
dure was repeated 4-6 times to ensure the sample after etching was c. 200 !liD. The cosmic component
was free from traces of the previous chemical treat- was calculated according to Prescott & Hutton (1988).
ments. Finally, the sample was washed in IMS 2-3 The OSL ages for eight sediment samples from
times to remove traces of water, and dried. Each two series are listed in Table 2.2. The first series
sample was then re-suspended in acetone, and the relates to 'natural' sediments lying beneath
size fraction 2-11 11m was isolated by suspension in <;atalhoyiik, recovered from two cores taken through
6 cm3 acetone. The fraction kept was that
produced from the supernatant follow- Table 2.1. Luminescence dose rate determinations.
ing 2 minutes settling, and the sediment Site ref. Lab ref. U-ppm Th-ppm K% Cosmic Water% Total annual
after 20 minutes settling. This fraction was (J.!Gy/a) dose (mGy/a)

diluted and used for preparing sample C:H94A(V)


C:H94A(VIII) CH2
CHI 3.91±0.67
3.37±0.62
9.52±2.31
12.24±2.12
2.67±0.64
3.37±0.59
138±14
94±9
20.6S
29.69
4.34±0.70
4.56±0.61
discs. Twenty-four aluminium discs, 1 em C:H940(VIII) CH3 3.9S±0.78 12.79±2.66 3.3S±0.74 94±9 16.32 S.37±0.84
C:H94D(IX) CH4 3.5S±0.7S 14.06±2.S7 3.87±0.70 73±7 22.5 S.53±0.77
diameter, were cleaned for each sample
PCI, unit 4 CP4 3.22±0.S3 9.S6±1.8S 2.66±0.49 20S±21 16.71 4.3S±O.SS
and placed in flat-bottomed glass vials. 1 PCI, unitS CPS 3.04±0.SS 9.62±1.90 2.67±0.S3 20S±21 8.27 4.69±0.66
ml of fine grained suspension was added PCI, unit 6a
PCI, unit 6c
CP6a
CP6c
1.04±0.82
4.98±0.89
21.8S±2.81
1.48±0.2S
S.77±0.7
0.64±0.11
20S±21
20S±21
18.13
3.96
6.2S±0.7S
1.94±0.29
to each vial and evaporated off in a cool
oven.
Sample CP6c, being coarse-grained, Table 2.2. Luminescence age determinations.
was first cleaned of carbonates and or- Site ref. Lab ref. ED(Gy) age± (years) Random Systematic %error
ganics as before. The sample was dried error error

and sieved, retaining the 180-2251-lm size C:H94A(V)


C:H94A(VIII)
CHI
CH2
38.S8±4.01 8886±1664
S0.76±10.54 11136±2219
17.89
21.92
S.54
S.56
18.73
22.62
fraction. This fraction was cleaned, and C:H94D(VIII) CH3 47.90±11.3 8910±220S 2S.24 S.57 2S.SS
C:H94D(IX) CH4 4S.S1±4.39 8233±1370 IS.68 s.ss 16.64
separated using heavy liquids (sodium
PCI,unit4 CP4 19.78±2.S6 4S4S±80S 16.83 S.SI 17.71
polytungstate, s.g. 2.62 g/ cm 3 ) and PCI, unitS CPS 31.04±6.73 6613±1S66 23.04 s.so 23.69
cleaned again. The quartz fraction was PCI, unit 6a
PCI, unit 6c
CP6a
CP6c
17.9S±2.S8
12.89±1.42
2872±472
6629±1109
IS.67
IS.82
4.98
S.4S
16.44
16.73
etched in 40 per cent HF for 40 minutes to

24
Geoarchaeological Investigations

the mound. The second series derives from the fill of beneath the building layers. Beneath 561 em cultural
a former channel of the C::ar~amba river, exposed in deposits were followed by grey-brown alluvial silt-
an irrigation ditch section south of C::atalhoyiik. The clay, overlying a dark organic clay (<680-687 em)
context and significance of these dates will be dis- and in situ lake marl (690->733 em). Two OSL dates
cussed below. It should be noted that luminescence have been obtained for the alluvial silt-day from this
ages are directly comparable with calender years, core (Table 2.2).
and that comparison with radiocarbon dates requires C::H.95-E: (645 em long; see Fig. 2.7) Taken off
calibration of the latter. The error function for lumi- the southwestern side of the east mound, this re-
nescence dating is calculated on a percentage basis peated a hand-augered core sequence undertaken in
(in these samples between 15 and 25%). 1993 (C::HE.93-2), but with proper stratigraphic con-
trol and sediment sampling. The core recovered 241
Lithostratigraphy em of alluvial silt-day, alternating with bands of
Coring was undertaken at six sites on, or adjacent to, fluvial sand between 137 and 241 em, over cultural
the main (east) mound at C::atalhoyiik. In 1994, this fill, at least some of which appeared to be in situ
involved four locations on a north-south transect on (241-318 em). Below this lay alternating fluvial
the east mound; in 1995 a further two sites were sands/ gravels and silts/ clays (318-610 em), and an
cored, both of them lying between the east and west olive-grey marly silt (610-640 em).
mounds (Fig. 2.2). Core stratigraphies may be sum- C::H.95-F (700 em long; see Fig. 2.8) and parallel
marized as follows: core C::H.95-G (540 em) These cores were taken be-
C::H94-A: (701 em long; see Fig. 2.3) Taken near tween C::atalhoyiik east and west, and C::H.95-F com-
the base of one of Mellaart' s trenches, this confirmed prised 189 em of brown alluvial silt over fluvial
a hand-augered core sequence undertaken in 1993 sands/ gravels with some silt layers and occasional
<C::HE.93-1), but with proper stratigraphic control and blackened bone and shell (189-295 em). Beneath this
sediment sampling. The core commenced 15.2 m be- was very dark grey or black, soft organic silt, with
low site datum and encountered 454 em of in situ occasional sand and gravel, pottery and animal bone
Neolithic deposits, underlain by 98 em of 'natural' (295-400 em), and very dark grey day-silt with abun-
alluvial clay and 141 em of lake marl, pale grey- dant cultural material including potsherds, locally
yellow in colour, separated by c. 8 em layer of very sandy (400-535 em), and finally olive grey clay or
dark grey organic clay. OSL ages have been obtained silty lake marl (535-700 em). The cultural material
for both the alluvial clay and the upper part of the was not in situ.
marl from this core (Table 2.2). In addition to these cores, examination of irri-
C::H.94-B: (350 em long; see Fig. 2.4) This core gation channel sections in the immediate vicinity led
was taken at the northwestern edge of the main to the discovery of a number of informative sections,
mound, in the area of the new dig house. No cultural including several buried channel-like features cut
material was encountered (either derived or in situ), into the underlying lake marl. In 1994 a large, buried
the sediments comprising brown silty clay (0-127 palaeochannel of the C::ar~amba river was found c.
em) and heavy 'backswamp' clay (127-286 em) over 500 m south of C::atalhOyiik (PCI). This section was
lake marl (268->350 em). cleaned, surveyed, recorded and the sediments sam-
C::H94-C: (583 em long; see Fig. 2.5) Immedi- pled. Further samples from this palaeochannel were
ately north of the present bekc;i's house, this core did collected in 1995, and a second feature west of PCI in
not encounter any apparently in situ Neolithic de- the same irrigation ditch was also cleaned, sampled
posits, although the top 200 em included late (?Byz- and recorded (PCII). Other features were visible east
antine) cultural material. Beneath this were grey to and north of C::atalhOyiik, including a sand and gravel
brown silts and clays of alluvial/ colluvial origin, filled 'channel' containing pottery (Stephen Holmes,
mixed with derived cultural material (notably char- pers. comm.). They will be studied further in future
coal), and with silty marl below 529 em. The basal field seasons.
part of the day-silt contained a well-preserved dia- The former river channel PCI was 42.5 m wide
tom flora. at the top of the marl horizon, and was filled with
C::H.94-D: (733 em long; see Fig. 2.6) Further fine- to medium-grained clastic sediments; organics
upslope and c. 2m higher than core 94-C, the core at were notably absent although they may have been
this location encountered apparently in situ Neolithic lost through oxidation. The palaeochannel deposits
deposits. Between 475-500 em was a layer of lake are represented by Units 4 to 6 (below). In order
marl including charcoal, apparently used as fill from the base up, the following lithostratigraphic

25
IN

z
~
:;::::1
0
o"
(\)
::;.
,Cfl
>-;;
(\)

ro-
....
t-J 95 .F/G OJ
0\ 0
• '<
(\)
....
~
:;::::1
0
3
0
5l
>-;;
til
:::!.
(fl
::>"

Figure 2.2. Plan of c;atalhoyiik with core sites.


Geoarchaeological Investiga tions

1005 0
3 -1
~: ............. OJ Jliil kg
L •L •&. •L

.............:'-
..................
.............. ~.

1004 1
...........
... 2

1003 2

....... . ..
...........
6"' 1002 ...........
...
......... r-.. 3
CIS ...
..........
...
6
'-"
..c
6 ....
0..
'-" Q)
s:: 0
0
....CIS ~

- > 0
Q)
1001 u 4
Cultural

-,-
~
Deposits
...... .
~
5 Alluvium
1000 L., L., L• L
'-., '-., L., L
L., L., L• L
................. _j_
Marl
0
--- 6
999
G ---
---·
---·
998 7 I
I I I I
0 5 10 20 40 60 0 0.3 0.6 0.9
% % xtf
Organic CaC0 3
Matter Equivalent ( JlrJ kg" I)
Content

Wctay
ts::::J
millsitt
~
I::) Marl 1:::::1 Charcoal Organic Clay

I Plaster 0 Olive
2 Redeposited Marl G Grey

Figure 2.3. Core (H.94-A: lithology and measured parameters.

27
Neil Roberts, Peter Boyer & Romola Parish

1003.5
0
........
........
1003 ........
....... 0.5

·-..:-..:-..:. ..
•L.•L•L.•L.
•L.•L•L.•
•L. "L "L "L.
•L •&.•L.•
1002.5 0.1 ,.ur? kg-1
...........
... 1.0

-
vi
1002
. .... ....... ...
.. ....
.... .. .. ,-._
6 I.
....... ....
L L. L a
~
. ....... . 1.5
L L L L. '-"
..c::
6
'-"
. ......
L. L. L. L.
L L. L. L
;.J
0..
0)
c .. ... .... ....
L
L
L. L. L.
L. L. L. a
.s L. L L. L.
~
;.J
<:':l
;> 1001.5
L. L
......
L. L
0

- u
L L L L
0)

~
.. .... ....
... ... L. L.
L. L L. L.
L. L L. L
2.0
.......
L L L L
L L
.... L. L.

... .... ..... ......


L. L L. L

L. L L. L.
L. L. ... L.
1001 ~~~~':~~&.
L. •L. •L. •L. ..
•&. •L. •a. •L
•&. •L "L"
2.5
•L. •L. •L •L.
L. "L"L •L. •
"L."L"L"L.
"L •L. •L •
•&. •&.. •L •L
..... •L. .......
•L. "L •L "L
•L. ..........
•L "L •L. •L. Alluvium
"L "L •t. •
1000.5
Marl
3.0

1000
3.5

999.5 I I I I
0 5 0 20 40 600 0.2 0.4 0.6
% % x1r
Organic CaC03 3 -1
Matter Equivalent (Jlm kg )
Content

Clay r::::::l silt


l:d t:: :I Marl

Figure 2.4. Core (H.94-B: lithology and measured parameters.

28
Geoarchaeological Investiga tions

.........
.........
...... ...
.........
.......
1004
... ... ..
.........
.........
......... 0.1 J.1ffi3 kg'
.........
.........
.........
.........
.........
.........
.........
.........
.........
...........
.........
......
......
.........
-.

......... ~~.

...... ...... ......


.........
1003 .........
.........
.........
.........
..... ... +
.........
.. ~ .. !'" ~ 2

~
1002
6"'
«i
6
...... 3
......
~

c
0
~
>-
ClJ
......
u:i 1001

...... 4
. ~· ~· ~

..
·~· ~· ~
~· ~· ~

..~· ~· ~
~· ~· ~
1000
..
~· ~· ~
~· ~· ~

..
~· ~· ~
~· ~· ~

.. D
~· ~· ~
~· ~· ~ 5
.
~· ~· ~
~· ~· ~
Cultural Deposits
--- ---
Silty+Marl
999

I I I I I I I
0 2 4 60 10 20 30 0 0.2 0.4 0.6 0.8
% % x1f
Organic CaC03
Matter Equivalent ( J.lm3 kg·')
Content

Figure 2.5. Core (:H.94-C: lithology and measured parameters.

29
Neil Roberts, Peter Boyer & Romola Parish

1007

... ... ... .


1006
r r r

... ... ... ...


r r r
r r r
r r r

.. .. .. ..
r r r
r r r
1005
... ... ... ...
r r r
r r r
r r r
. ~. ~. ~.
2
.. .. .. ..
r
r
r
r
r
r
1004 .. .. .. ..
r
r
r
r
r
r
.. .. .. ..
r
r
r
r
r
r
.... r
r
r
r
r
r
3
.. .. .. ..
r
r
r
r
r
r
....
0
1003 r r r
8"' ... ... ... ... 8......,
r r r
r r r
i r r r
g r r r
-5
v0..
4
r:: : "'! "'! --: 0
.~ : '-: "': "'!
~ : L: L: L.: ...v
0
;..
v 1002 : L: L: L: u
Ul : L: L: L:
:L:L:L:

:L.:L:L: 5
:L:L.:a.:
: L: L: L: Cultural
: "'! "'! "'!
1001
-~r~
: "'! "': ""!
................
L
.. .. ..
L L

... ... ... ...


L L L
L L L 6
L L L
Alluvium
.. .. .. ..
L L L
L L L

-•-t
1000 L L L

B-
Marl
7

999 ~~~~1~1~1~1-+1~1 ~~~1--~~1--~1


0 2 4 6 8 10 0 20 40 60 0 0.1 0.3 0.5 0.7
% % xtr
Organic CaCD.J 3 1
Matter Equivalent ( j..Ull kg" )
Content

Clay Silt f~~3 Marl []charcoal Plaster Organic Clay

c Coarse Marl
Figure 2.6. Core <:;H.94- D: lithology and measu red parameters.
30
Geoarchaeological Investigations

1004

0.1 ~ 3 kg -l
0 I

1003
....
..
1

1002

2
......._
,....; Alluvium
<ll
1001
8 t
cd Cultural
8
'-" ......._
Deposits
c::
0
•.jj
8 3 !
CIS ..c::
....
> 1000 Alluvium
0
......
~ 8:
4

999

998

6
Marl
997 +
0 0.4 0.8 1.2
%Organic xtr
Matter (Jl.Il13 kg -1)

Clay Marl

Figure 2.7. Core (H.95-E: lithology and measured parameters .


31
Neil Roberts, Peter Boyer & Romola Parish

997

996 I I I I I I I I t-1.L.....-.-+-----lll-------41
0 10 20 30 0 10 20 30 0 0.4 0.8 1.2
%Organic % CaC03 X If
Matter Equivalent (1J.m3 kg-1)

~~~~I Marl
Figure 2.8. Core C:H.94-F: lithology and measured parameters.

32
Geoarchaeological Investigations

units were recorded (Figs. 2.7 & 2.8): 186-384 em: Dark greyish brown (2.5Y 4/2-3/2 to
PCI SY 4/2) clay ['lower alluvium'; unit 3]
(Pre-palaeochannel) 384-444 em: Olive grey to greyish brown (SY 5/2
Unit 1: Light olive-grey (SY 6/2) lake marl (maxi- to 2.5Y 4.2) clay, with occasional gravel
mum thickness>150 em) [unit 4]
Unit 2: Dark (2.5Y N3/) organic clay (max. 20 em) 444-519 em: Dark greyish brown (2.5Y 4/2) sand,
Unit 3: Heavy, dark greyish brown (2.5Y 3/2) locally silty, with some gravel [unit 5]
'backswamp' clay (max. 160 em); ['lower al- 519->550 em: Olive grey to dark greyish brown (SY
luvium'] 4/2-3/2 to 2.5Y 4.2) stiff clay with
shell fragments and calcareous nod-
(Channel fill) ules [unit 6].
Unit 4: Dark greyish brown (2.5Y 4/2) clay with
manganese staining (max. 30 em) The main lake marl, which was visible at the channel
Unit 5: Matrix-supported gravel, mean particle size margins, was thus not recorded in this central pro-
8-32 mm, sub-angular to sub-rounded, lo- file, either because it had been truncated completely
cally part-cemented (max. 30 em) by the channel or (more likely) because the base of
Unit 6: Fluviatile silts, sands and gravels, coarsen- the channel fill was not reached. The potsherds from
ing from west to east (max. 220 em) the basal part of the 'lower alluvium' were identified
Sub-unit 6a: Red-brown (10 YR 4/3) well-sorted as water-abraded early Chalcolithic (J. Last pers.
silts comm.). The animal bone sampled from the same
Sub-unit 6b: Buff (10 YR 4/ 4) well-sorted quartz- level comprised 1 medium herbivore sized long-bone
rich silts shaft fragment, 1 Bos proximal radius fragment
Sub-unit 6c: Grey to orange, medium to coarse, (fused, with evidence of carnivore gnawing), and 1
moderately to well-sorted fluviatile large (?Bos, ?Cervus) radius/ulna shaft fragment (L.
sands containing large freshwater Martin pers. comm.).
bivalve shells
Sub-unit 6d: Fluviatile sandy gravels Sediment analysis and palaeoenvironmental
interpretation
(Post-palaeochannel)
Unit 7: Brown (lOYR 5 /3) alluvial silt-day with man- Using a range of indicators, including colour, tex-
ganese staining (max. 230 em). This 'upper ture, organic matter and carbonate content, and mag-
alluvium' caps all the underlying units. netic susceptibility, it is possible to identify and
Little cultural material was found in these deposits characterize a number of principal sediment types,
(one piece of fired pottery in Unit 6). Four sediment shown in Table 2.3. Preliminary particle size analy-
samples from the fill of palaeochannel PCI have so ses have been completed on some of the 1993 and
far been OSL dated (Table 2.2). 1994 samples, and downcore gravel-sand-silt-day
ratios are illustrated in Figure 2.11. It is clear that the
The second channel-like feature (PCII) was 89 m finer (clay and silt size) sediments are dominant,
wide at the top of the marl; it did not have a fill of with the clay content often approaching 50 per cent,
coarse-grained clastics (e.g. sands), but was infilled except in lotic (moving-water) fluvial deposits, such
largely with the lower alluvial clay. The contact sur- as in core <;HE.93-2. This preliminary core was taken
face with the underlying marl was also irregular between the mounds and shows a sequence of coarser
rather than sharp, and potsherds and bone were and finer fluvial sediments.
found near the base of the channel fill. A section Most of these sediment types also represent
cleaned in the centre of the channel revealed the lithostratigraphic units. The oldest of them, the lake
following stratigraphic sequence, with the part be- marl, is fine-grained and carbonate-rich (40-60 per
low 404 em being recovered by augering through cent at most sites); authigenic CaC03 having been
the base of the ditch. (Note that the numbered units precipitated out of lake waters and deposited on the
for PCI and PCII do not correspond.) lake bed. Cores <;H.94-C and <;H.95-E and F show
PCII lower carbonate values (25-40 per cent). The marl
0-22 em: Topsoil analyzed for mass specific magnetic susceptibility
22-186 em: Brown alluvial silt-day ['upper allu- produced readings systematically below the 0.1J.Lm3
vium'; unit 2] kg-1 level, suggesting a dominance of paramagnetic

33
Neil Roberts, Peter Boyer & Romola Parish

Horizontal Distance (m)


0 10 20 30 40 50 60 70

1006
781.81Et
472.26N
1005

~
1004
</i
E 1003
::: ::::::: ::: :::::::: ::::::::::: .7.:::::::::::::::::::::::::::::
~
g
c
0
".::l 1002
«l
>
0)

iii
1001

1000

999

I; ;~Gravel []sand D Silt Nclay


u D
Upper
Alluvial
Clay
~==~Marl II Organic
Clay
Bivalve
[ ] Mollusc
Shells
(See text for details of lithological units )

Figure 2.9. (ar~amba palaeochannel PCI: recorded stratigraphy (1994), north-facing section. (Note exaggerated
vertical scale.)

6c

6b
I
~
·c;;
8.
6a ~
Q3
c
c
6d "'
..c:
u
8
~

·a 5 "'
<a
p..,
:::>

3
t
2b

2a

0 2 4 6 8 10 12 0 10 20 30 40 50 0 0.4 0.8 1.2 1.6


% Organic Matter Content % CaC0 Equivalent Mass Specific Magnetic Susceptibility
3
(~1m 3 kg- 1)
Figure 2.10. (ar~amba palaeochannel PCJ: measured parameters for each stratigraphic unit.
34
Geoarchaeological Investigations

minerals in the deposits, presumably with some cent), it is possible that it represents a lateral equiva-
diamagnetics as well. This correlates well with the lent to the marsh marl, located towards the southern
high carbonate readings, as calcium carbonate is periphery of the Yarma secondary depression. In the
weakly diamagnetic with a mass specific magnetic PCI palaeochannel section an additional sub-unit (2b)
susceptibility (X) of -0.0048 J..lm3 kg-1 (Dearing 1994). was recorded in the upper part of the dark organic
A number of paramagnetic silicates (0.01-0.95 J..lm3 clay, and which had a mass specific magnetic sus-
kg- 1) may also have been present, as well as dolo- ceptibility in excess of 1.6 J..lm3 kg-1• This may have
mite (CaMg(C03 ) 2) <x = 0.011 J..lm3 kg-1). A marl sam- been a palaeosol developed on the underlying dark
ple from core <;H.94-A has given an OSL age of organic clay unit. Topsoil, in particular, exhibits in-
11,136±2219 years. This is significantly younger than creased magnetic susceptibility compared with
14C ages for the last main high lake phase (17,000- unweathered sediments (Oldfield 1983; Thompson
23,000 14C yr BP) on Dreissena shells from shoreline & Oldfield 1986).
facies. This discrepancy may be explained by sub- Above the marl and organic clay, there was a
aerial reworking of the upper part of the marl dur- clear increase in x, indicative of a change to allu-
ing and after lake regression. Reworking, which has vium dominated by ferrimagnetic minerals from the
been previously suggested by de Meester (1971) on catchment. The lower alluvium is a heavy silt-day
the basis of fossil wormcast evidence in the upper which forms prismatic blocks with slickensides upon
metre of the marl, would 're-bleach' sediments and drying, and which may correspond to the swelling,
result in younger luminescence ages. smectite-rich Karkm soil series of Dreissen & de
The next youngest stratigaphic unit is the dark Meester (1969). It is notably lacking a coarse fraction,
organic clay, up to 30 em in thickness, which is found and can have organic matter values >5 per cent,
in the palaeochannel sections and two of the cores suggesting deposition in the 'backswamp' environ-
from the east mound. Its absence in other sections/ ment of flood basins. This alluvial unit, which off-
cores is explained by its subsequent removal due to site is generally c. 150 em thick in the <;atalhoyiik
river incison or other erosion. This clay is probably a area, was observable in three of the east mound
marsh or shallow lake deposit. A lentic (standing cores~ In terms of sedimentary characteristics, it was
water) depositional origin is supported by its high very similar here and in unit 3 of the palaeochannel
clay and low sand content, particularly from the PCI (PCI) section, a close correlation which can be seen
palaeochannel section sediments, and even compared from Table 2.4. The lower alluvium has three OSL
with the underlying marl. A dark-coloured, organic ages from cores <;H.94-A and 0, where it is c. 100 em
'marsh marl' is described by de Meester (1971) from in thickness and lies beneath in situ Neolithic cul-
the Yarma and Hotam1~ depressions, respectively tural strata. Three dates give consistent ages of
north and east of <;atalhoyiik. This 'marsh marl' was around 8000-9000±2000 calender years BP (Table 2.2).
formed due to a short-lived re-advance of lake wa- Given that these samples must pre-date the Neolithic
ters and has been dated to 12,000-11,000 14C yr BP layers at <;atalhoyiik, they indicate that alluvial depo-
from mollusc shells. Although the dark organic clay sition began soon after the start of the Holocene (c.
at <;atalhoyiik does not contain aquatic mollusc shells 11,000 Cal BP). It is uncertain when deposition of this
and is not a marl (CaC03 equiv. values are <15 per unit ceased, although this was not later than the

Table 2.3. Principal sediment types and typical ranges for measured parameters.

Sediment type Organic Matter % CaCO, (equiv.) % Magnetic Susceptibility (X) Clay% Silt% Sand + Gravel %
mlkg-t <2J.I 2-SOJ.I >SOJ.I

Cultural fill, 4-8* 10--30* 0.1-1.0• 5-60• 35-85• 2-13•


in situ
Brown upper alluvial 3-5 18-30 0.4-0.8 23-48 37-56 14-21
silt-day
Fluvial silts, sands 1-5 5-25 0.2-1.0 11-35 22-66 3-70
and gravels
Dark grey-brown lower alluvial 3-7 10--20 0.1-0.4 42-58 29-54 2-12
'backswamp' silt-day
Dark organic clay 5-12 3-15 0.1-0.2 61-72 27-36 0-3
Pale olive-grey calcareous 2-5 25-60 <0.1 37-56 41-56 0-7
lake marl

• highly variable stratigraphically

35
Neil Roberts, Peter Boyer & Romola Parish

Table 2.4. Lower alluvium characteristics. alluviation then buried the old land surface (which
we shall be examining in case of possible soil devel-
Organic Carbonate Mass Spec. Colour Texture opment). Compared to the lower alluvium, the up-
Content(%) Content(%) Mag.Sus. (X) Range
<;:H.94·B 4.90-5.48 13.52-19.36 0.21-0.38 2.5Y 3/2-10YR 3/2 Clay-silt
per alluvium has a higher carbonate content, elevated
<;:H.94-D 4.22-6.89 5.86-17.48 0.17-0.29 2.5Y 3/2-2.5Y 4/2 Silt-day magnetic susceptibility values and a significant coarse
<;:H.94-A 3.92-5.24 10.66-23.18 0.15-0.27 2.5Y 4/2-2.5Y 6/2 Silt-day
PCl 7.04 17.16 0.25 2.5Y 3/2 Clay-silt fraction. This suggests a contrast in the catchment
source area between the early and late Holocene,
perhaps from the Bey~?ehir-Sugla basin, which has
overflowed into the <;ar~?amba river in the past
mid-Holocene (see below). To judge from the pres- (Roberts 1983). The ultimate cause of this difference
ence of early Chalcolithic pottery in the lower part of in sedimentary regime is likely to be climatic and I or
the fill of channel PCII, it is probable that deposition anthropogenic; for example, related to enhanced top-
continued during and after the human occupation at soil erosion following Bronze Age and later defor-
both <;atalhoyiik east and west. Assuming channel estation. Whatever the cause, the <;ar~mba alluvial
PCII to be of fluvial origin, then it may have been cut fan in the vicinity of <;atalhOyiik would have been
and occupied by the <;ar~amba river contemporary less waterlogged during recent millennia than it was
with the occupation at <;atalhoyiik. Its subsequent during the early Holocene.
infilling was certainly partly coeval with deposition Both in situ and reworked archaeological de-
of the lower alluvium. posits were encountered. The former are character-
The fluvial sands and gravels represent rela- ized by very erratic curves in x, organic matter,
tively high-energy environments within the river carbonate and grain-size, reflecting the wide range
channel itself. They therefore testify to the presence of differing deposits and processes represented in
in the past of one or more channels of the <;ar~mba the archaeological record. Other cultural deposits,
river running south of <;atalhoyiik and between the not apparently in situ, had either been subsequently
east and west mounds. Palaeochannel PCI, which reworked by mound erosion or were deposited into
separates the upper and lower alluvia, has three OSL active or abandoned river channels (e.g. the organic-
dates of mid-Holocene age (7000-4500 Cal BP) (a rich deposit recorded in core <;H.95-F, probably an
fourth date of 2872 BP may be too young because of old channel used for refuse disposal). One particu-
re-bleaching). These, of course, correspond to the larly unusual deposit was found in the lower part of
age of abandonment of this river channel, and it may core <;H.94-C. Overlying the marl was a silty clay
have been hydrologically active prior to, as well as containing a great deal of re-deposited cultural ma-
during, this time period. On the other hand, the OSL terial and a rich diatom assemblage. This was domi-
age of the (thin) basal layers (PCI units 4 and 5) is nated by Nitzschia frustulum and N. liebetruthii but
not older than the substantive channel fill of me- also with smaller numbers of Amphora coffeaeformis,
dium- to coarse-grained clastics (PCI unit 6), which A. veneta and N. palea/N.capitellata and occasional N.
suggests that palaeochannel PCI is unlikely to have aff halophila (C. Sayer pers. comm.). Most of these
been active much before the mid-Holocene, and was diatoms are eutrophic or halophytic species and sug-
therefore probably not contemporary with <;atal- gests standing water which was inundated by or-
hoyiik. Some of the fluvial sands and gravels en- ganic waste and subject to high evaporation. A
countered between the two mounds presumably number of opaline phytoliths were also observed.
represent the northward continuation of the same This core lay towards the base of a gully emanating
mid-Holocene river channel, but others are likely to from near the top of the east mound (Fig. 2.2), down
be earlier in date and match the period of site occu- which drainage runoff and waste may have been
pation. Further work is needed to verify this. transported. No alluvium was present in core C,
The upper alluvium caps all of the other non- presumably because this and the top of the marl had
archaeological strata in the <;atalhoyiik area to a been deliberately removed, for example, as a result
depth of c. 150-250 em. It post-dates palaeochannel of digging for marl as a raw material for ceramic
PCI and by inference is late Holocene in age. The production or mud bricks. This resulted in the for-
clear contact between upper and lower alluvial units mation of a hollow which became filled with water;
suggests a break in continuity of deposition; that is, indeed, it is therefore feasible that a small pond or
there may have been a phase of stability during which depression for liquid outfall existed at the north-
alluvial deposition presumably took place elsewhere west of the east mound. Alternatively there may
on the fan, away from the occupation area. Renewed have been a ditch around the site.

36
Geoarchaeological Investigations

Core correlation and discussion reconciled if Mellaart' s sounding had encountered a


ditch dug around the outside of the Neolithic settle-
One important point of correlation which can be ment. The current work has encountered a filled
made between each of the cores and palaeochannel depression to the northwest of the mound. This, how-
sections is the olive grey to white lake marl at the ever, was at the site edge and appeared to be filled
base of each sequence. The elevation of the top of the with redeposited cultural material, whereas Mellaart' s
marl horizon, however, varies by over 3m at differ- sounding was below stratified in situ archaeological
ent sites (Fig. 2.12). In several cases (i.e. CH.94-C, deposits. The 1994 cores did not encounter the grey
CH.95-E and F, and palaeochannels PCI and PCII) clay band recorded by Mellaart, although lake marl
this was clearly because the marl had been cut into (which is often light grey in colour) was found within
by later river erosion or by digging. Nonetheless, cultural deposits at different elevations, presumably
this still leaves a discrepancy of about 1 m between used as a fill when dwellings or floors were rebuilt.
the height of the top of the marl at sites on and off The height of the basal in situ Neolithic levels ap-
the east mound. This might be explained by the marl pears to vary across the mound, with the lowest
surface not being sub-horizontal, and observation of deposit recorded at 1000.26 masl in core CH.95-E,
the top of the marl in ditch sections shows that this beyond the present limit of the mound. The spatial
surface is - perhaps surprisingly - often rather extent of the east mound beyond its visible limits
irregular. The height discrepancy, however, also ap- also appears to vary. On the northwest corner of the
plies to the overlying dark organic clay layer. If this site, in situ Neolithic levels are restricted to the mod-
is, as suggested, a marsh deposit, it is more difficult ern hoyiik, whereas to the southwest they are found
to explain in terms of erosion of the pre-Holocene outside it beneath the alluvium.
surface. An alternative is that the weight of cultural The new results confirm that <::atalhoyiik has
deposit has caused compression loading below the been partially buried by post-Neolithic alluvium, but
mound, and has depressed the underlying sediment only by c. 3 m of river sediment, a smaller depth
layers by a metre or more. If so, this would compli- than suggested by an earlier geoarchaeological re-
cate calculations of the depth of alluvial burial at the connaissance study (Roberts 1982). Alluvial burial is
site and the thickness of cultural deposit to be exca- complicated by the height discrepancy between units
vated. Away from the site itself, the contact between on and off the mound (possibly linked to compres-
upper and lower alluvial units provides another im- sion loading), so that although there has only been c.
portant stratigraphic marker, of mid-Holocene date. 3m of post-Neolithic alluviation at <::atalhOyiik, there
Observations elsewhere indicate that these two allu- may be more than this of in situ archaeological de-
vial units are widespread over the outer part of the posits below the modern level of the Konya plain.
<::ar~amba fan. The 'natural' layers underneath <::atalhoyiik are of
The sediment sequences at and around <::atal- alluvial origin, and indicate that the site was cer-
hoyiik help to determine the extent and thickness of tainly located on the <::ar~mba fan as at present,
the Neolithic levels on the southwest side of the east although the soils were heavier and wetter than those
mound. It appears that c. 4.5 m of cultural material which exist today. The existence of the river channel
lie beneath the base of the deepest trench left by deposits between the two mounds and of palaeo-
Mellaart, prior to the present excavations (Fig. 2.3). channels in nearby irrigation sections indicates that
In his deep sounding at the end of his 1963 season, a branch of the <::ar~mba river has flowed here dur-
Mellaart (1964) reported that he encountered house ing much of the Holocene. So far, the only channel to
levels down to 1002.42 m, below which was a brown have been dated radiometrically (PCI) is of mid-
humus layer on top of a 30-cm-thick band of grey Holocene age, and therefore younger than the pe-
clay, interpreted by him as a flood deposit. These riod of occupation at <::atalhoyiik. On the other hand,
sediments were in turn underlain by a further 370 the presence of Neolithic-Chalcolithic cultural de-
em 'lower fill' containing stone and bone artefacts, bris in channel PCII and in fluvial sediments from
plaster floors and collapsed mud brick. The bottom cores taken between the east and west mounds, of-
of the archaeological deposits was apparently not fers strong support to the idea that a distributary of
reached in this deep sounding (estimated basal el- the <::ar~mba river also ran next to the site during
evation of 998.3 m). The lowest recorded cultural the early Holocene. Confirmation of the contempo-
deposits in the 1994 cores lie 2-3 m above the base of raneity of these river deposits with the Neolithic
Mellaart's deep sounding (Fig. 2.4). The discrepancy occupation at <::atalhoyiik requires further fieldwork
between these two sequences might conceivably be and radiometric dating. It is, however, likely that the

37
Neil Roberts, Peter Boyer & Romola Parish

<:H 94 -A
1004

<:H94-D

1003
<:H 93-2
<:H 94- B

1002
<:H 94- C
v;
6
od
g 1001
c:
0
";:l

"'>
(!)

t:li
1000 0 50 100
%

999 0 50 100
0 50 100 %
% 0 50 100
%
998
0 50 100
Delay Silt Sand Gravel %

Figure 2.11. Particle size ratios for 1993 and 1994 cores.

1007

1006

1005
94-C

1004 /
94-B
1003
---:-
~

E
i c:
1002

-~ 1001
til
1000

999

998

997

996

F:'W Gravel
~ b\\l Sand 1::::::~ Silt ~Clay
~
[::3 Marl -Organic
IIIII
Clay ~ Cultural
In situ m Redeposited
Cultural
Deposits Deposits

Figure 2.12. (:ntalhdyiik: composite lithostratigraphy and correlation .


38
Geoarchaeological Investigations

main mound at C::atalhoyiik was founded on the east- the early Holocene, because the site has been buried
ern bank of the C::ar~mba river, which appears to by 2-3 m of post-occupation alluvium. Most of this
have been cut into and built up to form terrace-like comprises an upper alluvial silt-clay, which was de-
features. It might be expected that the mound(s) of posited during the later part of the Holocene, when
C::atalhoyiik would have been established on raised the plain appears to have been somewhat less wet
levee deposits adjacent to the river channel, as van than at the time of human occupation. Consequently,
Andel & Runnels (1995) suggest similar locations for there are certainly several metres of in situ archaeo-
Neolithic mounds or magoula in Greece. However, logical layers lying below the general modern level
the alluvium underlying <::atalhoyiik lacks a fine sand of the plain. Because of a height discrepancy be-
component and is too clay-rich to be a true levee tween sedimentary units on and off the mound, pos-
deposit. sibly related to compression loading, cultural layers
are likely to continue deeper than would be pre-
Conclusion dicted from burial by alluviation alone.
The geoarchaeological programme at C::atal-
The sediment cores and samples obtained so far pro- hoyiik has also been planned as the central reference
vide the basis for understanding the changing natu- site for analysis of alluvial stratigraphy on the
ral and cultural landscape around C::atalhoyiik since C::ar~mba alluvial fan as a whole, spanning the en-
the time of the last glaciation. The site was founded tire Holocene period. A three-dimensional recon-
on an alluvial plain, which had started to accumu- struction of the alluvial fan stratigraphy is in
late around the start of the Holocene on the bed of progress using cores and sections taken at, or near
the former Konya lake. It is likely that a branch of to, a number of the archaeological sites being sur-
the C::ar~mba river ran next to the site at the time of veyed by Dr Douglas Baird and his team from Liver-
site occupation, with heavy backswamp soils stretch- pool University (Chapter 3). The results of this will
ing away behind. On the basis of the modern river be published elsewhere. Other planned future work
flood regime, the backswamp areas would have been includes an analysis of the relationship between al-
covered by floodwaters during early spring and pos- luvial stratigraphy and site formation processes, and
sibly also in winter. It is probable that alluvial depo- identification of the buried Neolithic ground surface
sition continued during and after the human along with evidence of manuring and other cultural
occupation at C::atalhoyiik. There is also evidence of activities.
off-site human activity. A nearby buried channel fea-
ture, now exposed in section (PCII), contains early Acknowledgements
Chalcolithic pottery and animal bone, and other cul-
tural debris has been found in cores taken between The 1994 and 1995 KOPAL field seasons received
the east and west mounds. River channels, active or financial support from the National Geographic So-
abandoned, may therefore have been used for refuse ciety. We are also grateful to the Turkish Ministry of
disposal. On the northwest corner of the main mound Culture for research permission, to the British Insti-
there was an artificial pond or ditch used for wash- tute of Archaeology at Ankara, and to our repre-
ing or animal watering, probably created by excava- sentatives from the Konya museum, Muzaffer Tun<;
tion of lake marl for building material, which (1994) and Necip C::ay (1995).
subsequently became infilled mainly by material The fieldwork involved the following personnel:
eroded off the mound. Given the very poor drainage 1993- Dr Neil Roberts (Loughborough University),
of the area at the time of occupation, building the Dr Henry Lamb and Mr Warren J. Eastwood (both
Neolithic settlement upwards (to form an artificial University of Wales, Aberystwyth)
mound) rather than outwards would have had obvi- 1994 - Dr Neil Roberts, Dr Hakan Yigitba~10glu (An-
ous advantages. A sloping site, raised above the plain, kara University), Dr Philip A. Barker (Lancaster Uni-
would have been less likely to suffer from river flood- versity), Dr Romola Parish (Sussex University), Mr
ing, and both rainwater and human waste would Warren J. Eastwood and Mr Carl D. Sayer (Laugh-
have run off the site onto the surrounding land or borough University).
into drainage ditches. 1995 - Dr Neil Roberts, Dr Hakan Yigitba~10glu, Dr
In contrast to the suggestion of Cohen (1970) Romola Parish, Dr David Twigg and Mr Peter Boyer
the current land surface is not the same as that (both Loughborough University) and Mr John Tibby
exploited by the inhabitants of C::atalhoyiik during (Monash and Loughborough Universities).

39
Neil Roberts, Peter Boyer & Romola Parish

References Kraft, J.C., I. Kayan & 0. Erol, 1980. Geomorphic recon-


structions in the environs of ancient Troy. Science
Aitken, M.J., 1994. Optical dating: a non-specialist review. 209, 776-82.
Quaternary Science Reviews (Quaternary Geochronology) Mellaart, J., 1964. Excavations at <;:atal Hiiyiik, third pre-
13,503-8. liminary report, 1963. Anatolian Studies 14, 39-119.
Battarbee, R.W., 1988. The use of diatom analysis in ar- Nambi, K.S.V. & M.J. Aitken, 1986. Annual dose conver-
chaeology: a review. Journal of Archaeological Science sion factors for TL and ESR dating. Archaeometry 28,
15, 621-44. 202-5.
Cohen, H.R., 1970. The palaeoecology of south central Oldfield, F., 1983. The role of magnetic studies in
Anatolia at the end of the Pleistocene and the begin- palaeohydrology, in Background to Palaeohydrology,
ning of the Holocene. Anatolian Studies 20, 119-37. ed. K.J. Gregory. Chichester: Wiley, 141-65.
Courty, M.A., P. Goldberg & R. MacPhail, 1989. Soils and Prescott, J.R. & J.T. Hutton, 1988. Cosmic ray and gamma
Micromorphology in Archaeology. Cambridge: Cam- ray dosimetry for TL and ESR dating. Nuclear Tracks
bridge University Press. and Radiation Measurements 14, 223-7.
Davidson, D.A. & M.L. Shackley (eds.), 1976. Geo- Rapp, G. & J. Gifford (eds.), 1985. Archaeological Geology.
archaeology. London: Duckworth. New Haven (CT): Yale University Press.
de Meester, T. (ed.), 1970. Soils of the Great Konya Basin, Roberts, N., 1982. A note on the geomorphological envi-
Turkey. Wageningen: Centre for Agricultural Pub- ronment of <;:atal hiiyiik, Turkey. Journal of Archaeo-
lishing and Documentation. logical Science 9, 341-8.
de Meester, T., 1971. Highly Calcareous Lacustrine Soils of Roberts, N., 1983. Age, palaeoenvironments and climatic
the Great Konya Basin, Turkey. Wageningen: Centre significance of Late Pleistocene Konya lake, Turkey.
for Agricultural Publishing and Documentation. Quaternary Research 19, 154-71.
Dean, W.E., Jr, 1974. Determination of carbonate and or- Roberts, N., 1991. Late Quaternary geomorphological
ganic matter in calcareous sediments and sedimen- change and the origins of agriculture in south cen-
tary rocks by loss on ignition: comparison with other tral Turkey. Geoarchaeology 6, 1-26.
methods. Journal of Sedimentary Petrology 44, 242-8. Stein, J.K., 1984. Organic matter and carbonates in ar-
Dearing, J.A., 1994. Environmental Magnetic Susceptibility. chaeological sites. Journal of Field Archaeology 11, 239-
Using the Bartington MS2 System. Oxford: Bartington 46.
Instruments Ltd. Stein, J.K., 1992. Organic matter in archaeological con-
Dreissen, P.M. & T. de Meester, 1969. Soils of the (:umra texts, in Soils in Archaeology, ed. V.T. Holliday. Wash-
Area, Turkey. Wageningen: Centre for Agricultural ington (DC): Smithsonian, 193-216.
Publishing and Documentation. Thompson, R. & F. Oldfield, 1986. Environmental Magnet-
Erol, 0., 1978. The Quaternary history of the lake basins of ism. London: Allen & Unwin.
central and southern Anatolia, in The Environmental van Andel, T.H., K. Gallid & G. Toufexis, 1995. Early
History of the Near and Middle East Since the Last Ice Neolithic farming in a Thessalian river landscape,
Age, ed. W.C. Brice. London: Academic, 111-39. Greece, in Mediterranean Quaternary River Environ-
French, D.H., 1970. Notes on site distribution in the <;:umra ments, eds. J. Lewin, M. Macklin & J. Woodward.
area. Anatolian Studies 20, 139-48. Rotterdam: A.A. Balkema, 131-43.
French, D.H., G.C. Hillman, S. Payne & R.J. Payne, 1972. van Andel, T.H. & C.N. Runnels, 1995. The earliest farm-
Excavations at Can Hasan III 1969-1970, in Papers ers in Europe. Antiquity 69, 481-500.
in Economic Prehistory, ed. E.S. Higgs. Cambridge: Vita-Finzi, C., 1978. Archaeological Sites in their Setting.
Cambridge University Press, 181-90. London: Thames & Hudson.

40
Chapter3

The Konya Plain Survey: Aims and Methods

Douglas Baird

The Konya plain survey is sponsored by the Uni- centre of the basin less than 200 mm per annum. This
versity of Liverpool and the British Institute of Ar- places the centre of the basin beyond the limits of
chaeology at Ankara. It is funded by the University reliable dry farming and at the edges of the basin
of Liverpool, the British Institute of Archaeology at dry farming is likely to occasion a notable degree of
Ankara, the British Academy, and the Society of An- risk of crop failure. Around <::umra, on the <::ar§amba
tiquaries of London. alluvial fan in the southwest of the basin, reports
This contribution represents a statement of suggest that in the recent past only once in 10 years
• objectives for the survey has there been sufficient moisture for the growth of
• and methods developed and tested in order to winter wheat (Cohen 1970, 122). Within such an in-
achieve those objectives. land drainage waterlogging and salinity may both
The purpose of the project is to provide a de- be problems. The relative aridity and these other
tailed reconstruction of the settlement history of the specific environmental features make the Konya ba-
Konya plain. Such a reconstruction will have an im- sin a particularly interesting setting in which to study
portant impact on our views of the development of the development of settlement. Our detection of
human settlement not just in Anatolia, but on semi- changing population densities and variation in the
arid and arid alluvial plains throughout the Near distribution and nature of settlements are likely to
East. The settlement history of such plains, the foci inform about significant changes in land manage-
for some of the most significant developments in ment (Roberts 1990, 58) and the nature of human
human history, remains poorly understood. Because interaction with the environment amongst other fac-
the Konya plain is relatively small we can survey a tors.
significant proportion and thereby achieve a mean- <::atalhoyiik has long dominated our view of
ingful understanding of its settlement history. The the archaeology of the Konya plain. The survey in
Konya plain may offer, in microcosm, a reflection of the region around the site will put <::atal in its con-
more widespread developments within the Near East. temporary and historical context. If we are to under-
Essential to such a wider contribution will be an stand the site as some sort of centre then we must
understanding of those features peculiar to the envi- gain some knowledge of putative subsidiary or de-
ronment of the plain and of the development of set- pendent settlements. If we can assert with confi-
tlement upon it. The survey project is therefore dence that no such settlements exist in its environs
undertaken in close cooperation with a team of then we will require a different view of <::atalhoyiik.
geomorphologists, directed by Neil Roberts of Laugh- In addition we would like to know more about the
borough University. They are studying the develop- antecedents of <::atalhoyiik and those developments
ment of past environments on the plain and, in that immediately follow the abandonment of this
particular, its history of alluviation. particular setting for occupation on the plain. Only
The Konya plain is an area of inland drainage intensive survey can recover information about any
at an altitude of around 1000 m above sea level. sites that may be relevant to these issues.
Major alluvial fans characterize the southern and In addition the survey will attempt to address
southwestern parts of the basin. Rainfall in the basin the questions that relate to the long-term develop-
is less than a mean of 300 mm per annum and in the ment of settlement on the plain. This will include the

41
Douglas Baird

influence of the environment of the plain on such of tells and the plain is relatively flat, the locating of
development and the impact of such developments a significant number of ancient sites on the plain is
on the environment. In what ways did the earliest straightforward. They stand out as distinct topographi-
agricultural societies develop? Did they emerge in cal features in this otherwise featureless landscape.
such alluvial settings or on the hilly flanks of such The methodological problem for the survey
plains? Was the rapid expansion of agricultural soci- project is our ability to identify, at the surface, low
eties related to the colonization of such plains? How tell sites (French 1970, 141), sites of very limited area,
much did population growth or redistribution, or and non-tell sites and the variety of sites completely
the emergence of key centres have to do with the rise buried by alluviation. We know that there is 1m of
of urban societies? When do these appear on the sediment underlying the earliest occupation levels
alluvial plains? How dense was settlement in differ- at <;atalhoyiik and overlying Pleistocene lake marls.
ent periods? What sort of settlement systems existed Another 3 m of sediment has been deposited around
under regional and imperial states from the Hittites <;atalhoyiik since it was first occupied (Roberts this
to the Ottomans and how did these externally based volume). In some areas alluvial deposits may be 4 m
powers affect settlement in such areas? deep or more. Clearly many small and early sites
The plain is flat making interpretation of the may be buried. Unless we have some method for
interrelationships of settlements and the working of recovering such sites the settlement record which
past settlement systems simpler, because we do not we will document will be very partial. In particular
have to consider the role of topography in the devel- the early settlement history of the plain will remain
opment of settlement. Models allowing us to infer obscure.
the economic structure underlying settlement distri-
butions were first developed for such plains. 2. Site comprehension
The density of sites in the plain has been known Tell sites are frequently complex palimpsests of oc-
for some time (French 1970, 139). This is particularly cupations of different periods and types. It is regu-
true in the area of the alluvial fans which border the larly the case that occupations of different periods
plain, the <;ar~amba fan forms one of the main foci of are located in different areas of particular sites. Such
the survey. Our work in 1995 indicated that in some occupations are also likely to be subject to different
areas of the fan site density is at a minimum of 1 site degrees of survival and exposure. To state only the
per 6 km2 (Fig. 3.1). Given the multi-period occupa- most obvious processes: the earliest, smallest occu-
tions on most of these tell sites we are clearly dealing pations are most likely to be buried, the latest eroded.
with a relatively well-preserved settlement record. Fortunately, due to local weathering and dis-
Mellaart, French, and Hall surveyed sites in turbance processes and the nature of the local veg-
this area in 1958. They visited a significant number etation the surfaces of a large proportion of tells, at
of sites in the plain, including some of the larger. least on the <;arf?amba fan, are littered with artefacts.
The sites they visited still only represented a small This has provided the opportunity to assess, to some
proportion of those present (French 1970, 139-41). In degree, the area of the different phases of occupation
addition, their relatively rapid visits and unsystem- on tells and the detection of some early occupations
atic collections are unlikely to have documented fully masked by later deposits. Additionally it has offered
the occupations represented on these sites (French us the opportunity to discover the presence of dis-
1970, 139). Systematic sampling of sites between 1993 tinctive activities and uses of certain sites or parts of
and 1995 indicates that they indeed, quite under- sites at particular periods.
standably, failed to identify some periods of occupa-
tions present at sites they visited. In particular, they, Methods
unsurprisingly, failed to detect prehistoric occupa-
tions, represented by low density material, buried 1. Site location
by later settlement. Due to major agricultural development in the area
the plain is transected by large and small irrigation
Issues canals, some of which are regularly redredged. New,
large canals are currently being cut. By walking these
1. Site location canals (Fig. 3.1), one team member on each bank, we
Recently revised 1:25,000 topographical maps pro- can inspect the canal sides for evidence of occupa-
vide base maps for the survey. Because a large tion deposits and the upcast for evidence of artefacts
number of ancient settlements on the plain are part deriving from such occupation. Since most of the

42
Abditdu
K1slar Hi.i
M29-C1.10
~
/lba 1006.9m IUJ!' Karaca II
I <;:ingene
'>, Karaca Hi.i @) 1005.3 m
Q Karhane Hi.i 'f> M29-C2.12 Kosel! o
Q M29-C1.17 M29-C1.13
1015.5m 1007m M29-C2.7 0 Sofular
M29-C2.11

~ Ovakavag1
Avratham Hi.i
0 M29-Cl.3
1.'\ Tekke Hi.i
'-.J M29-Cl.2 K!zll Hi.i Sam1k
Kmd1rahk <;:mill 1'\
1017.3m \) M29-C1.12 0
1/' 1002m J . M29-C2.6 :::::
<;:m11i '<:
0 Pol
• "' Beskilise Yayla "1:1
M29-C2.3 OJ
1005.3m ::;·
D Vl
Kap,dalli Hi.iO 1
AgadamiJ Hi.i c....,
M29-Cl.6 Ki.i<;i.ikkoy <
M29-C1.18 ro
.;:. M29-C1.15 '<:
w
Hala<; M29-C2.4 :»
<;:atalhoyi.ik East ( 1008m fill Plfasanb §"
<;:atalhoyi.ik West () M29-Cl.7 (fl

M29-C1.8
0 1020.3m Pol
:::::
_/"\
Ku~lu Q) Km1k<;1 0..
M29-C2.9./ ° Ku~lu 1005.3m
M29-C2.10 ~
~ ro
Alemdar .. t) g.
Smak Hi.i Uri.imdi.i Hi.i 0
l002.5m 0 M29-C1.9 Tell?0 0..
M29-Cl.ll (fl

1007.5m
t) Dedeli Hi.i (§)
M29-Cl.S Araboglu \<$J'l-'o«' Alanh
1008m M29-C2.2 o /'
0 Torundede Hi.i 1004.6m __./
M29-C1.14

After Konya M29-Cl / C2 1:25000


G•;dN~ 0 Skm

Figure 3.1. Sites in the (:atalhoyiik area.


Douglas Baird

largest canals cut down into the Pleistocene lake areas and thus relative size of, at least some of the
marls we hope to gain an indication of the presence occupation periods on these multi-period sites.
of small tell sites, non-tell sites and those sites com- Supplementary collections of the most diagnos-
pletely or substantially buried beneath alluvium. tic material from the general surface of sites are un-
At this stage, limited surface prospection has dertaken. This material may offer the opportunity to
been conducted only in the immediate vicinity of provide more precise dating for certain periods of
known sites, partly as a means of verifying their occupation.
limits, but also to test the possibility of identifying The evolution of these methods was the result
non-tell sites or small tells through field walking. It of trials conducted in the reconnaissance seasons in
is intended to expand this programme once the ini- 1993 and 1994.
tial results of surface inspection around tells and
canal surveys have indicated areas most likely to Preliminary results of methods
repay such prospection. A project to develop the
remote sensing of sites using satellite imagery may 1. Site location
help define areas suitable for such field walking. Canal walking has proved effective in the location of
This is a Ph.D. project being undertaken by sites not on topographic maps. 9.5 km of canal were
Stephen Holmes at Edinburgh University. Terres- walked in map square M29C2 (Fig. 3.1). Three sites
trial survey has identified sites not located on topo- not on the topographic maps were located. These
graphic maps. If these can be distinguished on included:
satellite images we may be able to provide criteria
for the identification of other such sites. We will test • a small tell site Kuslu II (M29C2.9). This covered
such putative site locations in future seasons. The an oval area 100 m by 80 m and was less than 1
survey thus provides a testing ground for a new m high. Such small sites are likely to have been
type of site location procedure. missed by topographers.
• 1 non-tell occupation site, Kopruyeri, (M29C2.10),
2. Site comprehension c.40mx40m;
In order to gain a better understanding of individual • 1 non-tell site, Sofular (M29C2.11), a diffuse arte-
sites the following procedures have been developed fact spread. This may represent general activity
and tested. in the landscape or seasonal settlement.
Each site is the subject of a detailed contour
survey. A site grid is established and intensive arte- In addition, field walking and discussion with local
fact collection is undertaken from regularly spaced people located three sites not on topographic maps.
sample squares (Fig. 3.2). The 3 x 3 m squares are These included 1 buried tell site projecting c. 0.5 m
spaced between 10 and 50 m apart depending on the above the present ground surface.
size of the site. Most commonly the squares are Also located in this manner were two artefact
spaced 30 or 40 m apart. This provides coverage of concentrations close to, but quite separate from neigh-
all areas of a site. bouring tell sites, Kizil I (M29C1.12) and Kizil II
The vegetation from the surface of each square (M29C2.3). These artefact concentrations may have
is removed and each square is scraped with a shovel. been outlying habitations or may represent evidence
The fine sediments and artefacts thus collected are of off-site activity. One such instance includes obsid-
sieved through a 5 mm mesh. This ensures standard ian blade fragments whose attributes clearly provide
retrieval conditions for each part of each site. Con- evidence of such off-site activity in the prehistoric
sistent collection is thus enhanced. The smaller arte- period.
fact categories are also well represented by this A number of individual finds were retrieved by
method. These include smaller components of the canal and field walking, including prehistoric
chipped stone assemblages, particularly important chipped stone from the alluvium. These finds may
for identifying the presence of some, especially ear- represent more generalized and diffuse activities in
lier prehistoric, occupations. Also thus retrieved are the landscape or transportation of such material by a
more representative samples of beads, fine bone ar- variety of possible mechanisms.
tefacts, and fine, thin ceramics.
By looking at the distribution of chronologi- 2. Site comprehension
cally diagnostic artefacts from square to square across Systematic collection methods are already revealing
the site it is hoped we can map the approximate relatively promising details about sites and variations

44
Konya Plain Survey: Aims and Methods

c;ingene Hiiyiik ••••• I


0 lOOm

D Collection squares

D D D D

Figure 3.2. (:ingere Hiiyiik showing distribution of collection squares and contours at 1 m intervals.

in the nature of occupations through time. of particular settlements. Thus C::ingene (M29C2.12)
Thus intensive collection has revealed sporadic grew from a small Hellenistic settlement c. 3.1 ha to
artefacts from buried prehistoric occupations on a slightly larger Roman settlement, to a Byzantine site
large, later period sites (Fig. 3.1). Examples include: twice as large as its Hellenistic predecessor at least c.
6.3 ha. This pattern may be repeated at other sites.
• Late Chalcolithic occupation on Iron Age, Hellen- Intensive sampling is also recovering material
istic and Roman-Byzantine Hala~ (M29C2.4); pointing to specific kinds of activity on sites. Thus
• an Epipalaeolithic type assemblage from the site we have recovered the residues of certain industrial
of Kizil I (M29C1.12) almost certainly derived processes (this material awaits analysis). We have
from redeposited material; also recovered human bone from disturbed cem-
• an Early Bronze occupation at C::ingene (M29C2.12). etery areas on a number of sites.
It is premature to talk of 'results' other than the
The Epipalaeolithic material would not have been proven value of our strategy and tactics. It is clear
recovered without sieving. that distinct patterns are likely to emerge. Certain
We have been able to document the quite var- periods are well represented, others less so. This was
ied size of settlements at particular sites at different already appreciated by the earlier investigators
periods, precise areas await computation. Thus at (French 1970, 142). It is also clear that certain periods
Koseli (M29C2.7) an Early Bronze I-II site was twice are much better represented in some areas than oth-
the area of the later Iron Age site there - Early ers. For example, around C::atalhoyiik Early Bronze
Bronze site area c. 1.5 ha, Iron Age site area c. 0.8 ha. I-II sites are very common. Further east Iron Age
In some cases we seem able to document the growth sites are well represented, Early Bronze occupations

45
Douglas Baird

less so. One major issue we will need to investigate, References


once we have defined this phenomenon more pre-
cisely, is why this should be so. Cohen, H., 1970. The palaeoecology of south central
The survey also involves major cooperative ven- Anatolia at the end of the Pleistocene and the begin-
tures. As I have already indicated we are working in ning of the Holocene. Anatolian Studies 20, 119-37.
close cooperation with a project studying the geo- French, D., 1970. Notes on site distribution in the <;umra
area. Anatolian Studies 20, 139-48.
morphology and palaeoenvironments of the plain, Roberts, N., 1990. Human-induced landscape change in
directed by Neil Roberts of Loughborough Univer- south and southwest Turkey during the later
sity. By identifying occupation material interstratified Holocene, in Man's Role in the Shaping of the Eastern
with alluvial and other depositional units we will Mediterranean Landscape, eds. S. Bottema, G. Entjes-
help provide a chronological framework for this Nieborg & W. Van Zeist. Rotterdam/Brookfield: A.A.
palaeoenvironmental record. Balkema, 53--68.

46
Chapter4

Excavations at Ptnarba§t: the Early Stages

Trevor Watkins

The cluster of sites at Pmarba~1 was noted in 1993 in prehistoric occupations, because of their uniqueness
the first, preliminary season of survey work in the on the central Anatolian plateau and the potential
region around <;atalhoyiik in the Konya plain, car- they have for shedding light on the processes of
ried out by Dr Douglas Baird and the present au- increasing sedentism and the early development of
thor. Attention had been drawn to the location by Dr sedentary societies, the adoption of cultivation and
David French, who had first visited the location and herding. The programme is part of the <;atalhoyiik
picked up a small amount of chipped stone while he Research Project, because of the potential for shed-
was working at Can Hasan in the 1970s. There was ding light on what may be termed the longer term
already a known occupation of a cave there in the prehistory of <;atalhoyiik. The excavations were
Byzantine period, documented by a rock-cut inscrip- funded by the British Academy, the British Institute
tion, but the presence of early prehistoric remains of Archaeology at Ankara, the Society of Antiquar-
had not previously been noted. A preliminary search ies of London and various funds at the University of
of the locality in 1993 noted that, as well as a series of Edinburgh. Our collaborator on the excavation (car-
small-scale occupations, constructions and tombs ried out jointly by the University of Edinburgh and
dating to various periods ranging from the Late the Karaman Museum) was Cengiz Topal, Assistant
Chalcolithic or Early Bronze Age to the Byzantine, Director of the Karaman Museum. The team's
there was evidence of substantial early prehistoric archaeo-botanist is Mark Nesbitt; the zoo-archaeolo-
occupation in several rock-shelters and on an open gist is Dr Nicola Murray; and the study of the lithic
village site below the cliffs and at the very edge of a assemblages is the responsibility of Dr Douglas Baird.
small, spring-fed lake. It was also clear that there This paper consists of a brief account of the
had been substantial recent damage to the site cluster of sites at Pmarba~1 and a short report on the
through the search for and looting of a number of work of the first two seasons, in 1994 and 1995.
stone-built tombs.
In view of the importance of a number of rock- The sites at Pmarba!it
shelters and an open village site of early prehistoric
date in that part of Turkey, and because of the dam- In the southern part of the Konya plain there is a
age that was threatening them, our report at the end great volcanic massif, Kara Dag, that thrusts up more
of the season to the Turkish Directorate-General of than a thousand metres above the surrounding plain.
Museums and Antiquities drew particular attention At the northwest margin of the zone of volcanic up-
to the problem. The Directorate-General called on thrust, a ridge of limestone hills was pushed up. The
the Karaman Museum (the sites in fact lie just within northern tip of the ridge of hills takes the form of a
the territory of the newly formed province of cliff. Immediately below, there are springs that feed
Karaman, and not in Konya vilayet) to confirm the a permanent shallow lake. Around the quite small
situation, and the present author, representing the lake and extending far to the north there are large
Department of Archaeology, University of Edin- stretches of seasonal water and reed-marshes, where,
burgh, and the Karaman Museum made a joint ap- at the end of the Pleistocene period, there was a
plication to undertake rescue excavations at the site, short-lived, shallow lake, one of five such basins on
beginning in 1994. the Konya plain at that time.
The excavation work is concentrated on the early At the base of the cliffs at the end of the limestone

47
Trevor Watkins

• @)
<;:ATALHOYUK

••
Figure 4.1. Map of the Konya plain, locating Pznarba~1 relative to Kara Oag and (atalhoyiik.

48
Excavations at Pmarba~n: the Early Stages

ridge there are several rock-shelters. Below the shel- large limestone block, somewhat lower than the main
ters there is a slope down to the edge of the lake. line of hills, but still possessing cliffs five or six me-
Extending into the lake is a short peninsula that tres high. On the flat top of this massive block there
terminates in a pile of massive limestone boulders. are some architectural remains, and probably the
On the neck of this peninsula there are masonry disturbed remains of tombs. In its east-facing, verti-
tombs, which have been subject to recent looting, cal side, there are three more rock-shelters or shal-
and a surface scatter of pottery dating to various low caves. The southernmost, called Site C, is a small
periods. When the site was first visited in September chamber with a chimney-like formation. The cham-
1993, a recent looters' pit close to the track through ber has been modified, and there is a rock-cut step or
the site showed that there was a deposit of a differ- bench close to its present ground level, several foot-
ent character beneath the surface stratum. From the holds leading upwards into the chimney, and a small,
heap of spoil from this pit it was easy to collect a eroded, Christian inscription in indecipherable Greek.
significant handful of chipped stone that included At the mouth of the cave there is a bank of deposit
recognizable blade lets and very small-scale debitage, that has almost certainly been pushed out of the
and a small number of quite distinctive microliths. It cave when it was modified in Byzantine times. From
therefore appeared that, beneath later occupations the bank there are eroded small-scale chipped stone
on the peninsula, there was an early prehistoric oc- pieces. All those that have been collected are of ob-
cupation. This site has been called Site A. sidian, but since obsidian is so much more easy
The rock-shelter that had been first recognized to spot than flint, the observation is probably not
by Dr David French had
also been the scene of re-
cent interference. A large
pit had been dug against
the rock-face beneath the
over-hang, and partly re-
filled . This site also yielded
G
a number of small-scale
chipped stone pieces, but
no pottery was in evidence.
The first rock-shelter, which
faces west, has been called
Site B. Before excavation, it
consisted of a flat surface
beneath an overhang of the
cliff-face. Several huge
boulders projected from the
surface of the soil, fallen at
o _ __ ___ _ l_ __ _____ ___ ---
some time in the past from
the cliff above. To the south
of the flat surface, there
were two more huge boul-
ders on the surface. These
had been compounded with
smaller stones to form a
barrier, which is continued
as a dry-stone wall around
the rest of the area. The dry- I
stone wall has an entrance 0
about two metres wide, and L - - ·- - - - - '2 !? ? - - - - _Q~ - - - - - - - - - - -':C. - - - - - -

then curves around to meet


the rock-wall to the north
of the rock-shelter. Figure 4.2. Plan of Site B- the rock-shelter: latest features. The trench is aligned on
Opposite the rock- magnetic north. The southern trench is Trench 1; the northern trench, begun in 1995,
shelter Site B, there is a is Trench 2.
49
Trevor Watkins

significant in terms of the date of the deposit. What and stretching back five or six metres into the rock.
is more likely to be meaningful is the observation The crevice is partly roofed with rock. The nearly
that all the collected material belongs in the bladelet flat surface of the soil in this area is dark and fine.
scale of chipped stone. Once again, small-scale chipped stone pieces have
Immediately to the north of Site B, there is an- been collected from the surface, both within the crev-
other opening in the vertical rock surface. It resem- ice and outside it and down the slope to the north.
bles the peak of a shallow cave or rock-shelter. If that This possible rock-shelter has been called Site E.
is what it is, then it has become almost entirely filled Finally, around the corner from Site E, there is a
with deposit. A number of small-scale pieces of ob- low vertical face of rock that faces north and looks
sidian have been noted on the surface, apparently out onto the peninsula and the lake. A very small
eroding out of an archaeological deposit. It is possi- amount of chipped stone was found here, too, but of
ble that material that has eroded out of the bank of greater importance is an area of mortars of different
material from Site C, which is immediately above sizes formed in the natural surface of the rock where
this area, has slipped down and is confusing us. For it forms a shelf immediately below this vertical face .
the moment, however, we are calling this possible It is by no means clear whether this area forms a
cave or rock-shelter Site D. separate and distinct site, or whether it is a particu-
Only a few metres further north again is a sub- lar adjunct of another, adjacent site, either Site A, the
stantial crevice in the rock, about four metres wide settlement on the peninsula, or Site E, the rock-shel-
ter that is only a few metres
away around the corner.
For the present we are call-
ing this site Site F.

The excavations

The purpose of the first sea-


son qf excavation, in Sep-
tember 1994, was to test the
two components of the site
that showed signs of recent
interference, Sites A and B.
The season was intended as
an assessment exercise. We
I
needed (a) to learn how
I deep the stratigraphy is on
L -
each site, and how well de-
tailed and rich it is, (b) to
obtain cultural assemblages
from both areas of the site
in order to form some idea
of the periods of occupation
and the quality of the avail-
-·,
·,
able archaeological material,
and (c) to obtain carefully
controlled samples of bo-
tanical and zoological ma-
terial in order to assess the
I research potential for learn-
L_ - ·- ·- - · - · - - - - - - - - - - - - -
ing about the environment
and its exploitation. In view
2m of the uniqueness of these
early prehistoric sites on the
Figure 4.3. Plan of Site B- the rock-shelter: the stone retaining wall. plateau, close to <:;:atalhoyiik,

50
Excavations at Pmarba~;a: the Early Stages

we urgently needed (d) to obtain samples for radio- material is recorded as it is re-packed ready for trans-
carbon dating. In September 1995, work was concen- port to the laboratory for sorting and analysis.
trated on Site B, the rock-shelter site, with a view to The heavy residue from the process is drained
doubling the area of the 1994 sounding, and pen- and dried on meshes at the site, and is then sepa-
etrating further into the stratified deposits. rated into various fractions. In 1995 we introduced a
Over both seasons a sampling strategy and re- 10 mm sieve in order to segregate the material that is
covery methodology were developed and tested. All >10 mm. A 5 mm mesh and a 3 mm mesh are also
archaeological deposits removed from the trench are used, so that we have a fraction that is >5, <10 mm,
quantified approximately through the use of a stand- another that is >3, <5 mm and a third that is >1, <3
ard-sized bucket of known capacity. At first, a mini- mm. In 1994 portions of the smallest fraction, the >1
mum of 30 or 60 litres (depending on the assumed mm, were sorted, but it was concluded at the end of
size of the new unit) was removed for wet-sieving the season that the amount of useful information
and flotation. In the second season this was modi- that was being recovered in terms of chipped stone
fied, for simple practical reasons of portability, to a or animal remains at >3 mm was not worth the very
standard sample of 40 litres. A second sample of 40 large amount of time that is taken sorting such a fine
litres is taken later from another part of the same residue. In 1995 the smallest fraction was simply
context. Further samples are then taken, where there weighed and retained in case the archaeo-botanist
are reasons to do so, and as long as there is the wishes to sort it for carbonized plant remains that
capacity to deal with the product. Where greater failed to float. The >3 mm, >5 mm and >10 mm
precision on the location of the samples is required, fractions were weighed and then sorted at the exca-
samples were taken from a particular square metre, vation work-rooms. In 1994 the results of the sorting
allowing us to quantify the variability within a de- of the >3 mm and >5 mm heavy residues were care-
posit to a certain degree. The general objectives are fully compared. It seemed to us that the critical
to obtain the maximum amount of wet-sieved mate- threshold was around 5 mm, and that there is only a
rial that can be dealt with, and to screen all the limited amount of useful data in the smaller fraction.
deposit with at least a 5 mm mesh. Everything that is Consequently, we have standardized on the proce-
not sent for wet-sieving is dry-sieved at the side of dure of sorting only 20 per cent of the heavy residue
the trench. The rocking frame is equipped with a fraction that is >3 mm.
wire mesh of approximately 5 mm. Comparison of
the dry-sieved and the wet-sieved product has shown Summary of early excavation results
the difference in general terms, and those differences
will be quantified and standardized through study of Excavations were begun in 1994 with soundings in
the faunal remains and chipped stone next season. one of the rock-shelters, called Site B, and in the
The flotation and wet-sieving are carried out open area settlement on the peninsula, called Site A.
using a modified Siraf-type machine, made locally The work on Site B was continued in 1995.
under the supervision of Mark Nesbitt. In 1994 we
used the British Institute in Ankara machine, which Site B - the rock-shelter
had been built by Mark Nesbitt. In the 1995 season, a Our first sounding in this area was set immediately
new machine was built in Karaman to the same de- to the S of the damaged area. It took the form of a
sign as that built for (atalhoyiik. At Pmarba~n we trench, 4.5 x 2 m wide, at right angles to the rock
have had the great advantage that the flotation and face. In 1995 the area of the sounding was doubled
wet-sieving machine can be set up beside the dwin- by adding an identical trench parallel to the first,
dling, spring-fed lake. In 1994 the floated material and enclosing the area of the damage. Michael
was caught in two sieves, the first of 0.3 mm and the Rawlings, the supervisor of this excavation site, de-
second of 1 mm. In 1995, in common with (atal- fined five stratigraphic phases in 1994, and two more
hoyiik, a single fabric mesh of 0.3 mm was used for phases have been added to the sequence in 1995. The
catching the floated materials. The wet-sieving is trench begun in 1994 is now Trench 1, while that
carried out using a 1 mm flexible mesh, through opened in 1995 is Trench 2.
which the fine residues are encouraged to drop by The large pit dug illicitly before September 1993
the agitation of the water entering the tank and in the back of the rock-shelter was mostly within the
manual assistance. The floated residues are left to area of the new trench. The pit was emptied and the
drain at the site, and are then air-dried at the excava- sides cleaned; it was found to be about 1 metre deep,
tion work-rooms. The gross weight of the unsorted and dug entirely through stratified archaeological

51
Trevor Watkins

deposits. The rock-face exposed at the back of the deposit of clay. The shallow pits or scoops, and the
rock-shelter was seen to be blackened by smoke, and fire-installation had been dug into a layer of grey
the slope on the surface of the rock was still dipping ashy material which lay all over Trench 1 and the outer
backwards. These observations suggest that there is part of Trench 2. This material was up to 0.15 m in
still some considerable depth of archaeological de- depth and was excavated from most of the trench.
posit below the bottom of the pit. The large quantity and splintered nature of the ani-
The surface deposit across both trenches was a mal bone from this deposit, and the copious amounts
thin layer of trampled animal dung and dead veg- of charcoal, may suggest a build-up of occupation
etation. When that was removed, clearly articulated material but there were no surfaces defined within
archaeological features could be identified. Four shal- the deposit.
low, roughly circular pits or scoops, covered at the The situation in Trench 2 was complicated be-
top with stones, and, at least in one case, marked out cause, in addition to the recently dug pit against the
or even lined with carefully set stones, are among rock-face, there seemed to be earlier disturbance of
the latest features on the site, two in each trench. A the deposits immediately adjacent to the pit. There
number of the stones covering these pits or set around were some Byzantine potsherds and a few, scattered
their edges were found to be burnt. Below their stony fragments of human bone, suggesting that perhaps
surfaces, these pits contained significant amounts of there had been a shallow burial of Byzantine date
small wood charcoal. These features would seem to that had been badly disturbed. At the east end of
be small hearths or cooking installations. A radio- Trench 2 there was a further complication in the
carbon date from one of these pits placed its use form of a structure of stone blocks forming three
somewhere early in the fourth millennium sc sides of a square, with the edge of a huge, half-
(uncalibrated), or early fifth millennium sc in calen- buried boulder forming the fourth, western side. The
dar years (Table 4.1). excavation of this construction is not complete and
The digging of one of the pits had cut a basin- will continue, we hope, to a more positive conclu-
shaped 'fire installation' or hearth recorded in the sion next season.
northeast corner of the trench. This had a base of Below the grey, ashy deposit in Trench 1 was
baked clay that was actually bonded to the rock-face found the irregular top of a curving wall made up of
at the east and was probably the initial lining of the large blocks of limestone, and, against the north side
feature. The whole installation was placed within a of the trench, the fill that had been tipped into the
cut and consisted of a series of ashy or silty fills with structure. The wall is crudely built of three or four
at least one instance of re-lining indicated by another courses of very large stones, its west end butted
against the rock wall at the
back of the shelter. At the
Table 4.1. Radiocarbon dates.
western end of the exposed
Lab. Ref. Context Material Age •• (oc) Calibrated age; 95.4% confidence; BC part of the wall was a single
OxA-5499 ABJ charcoal 9050±80 (7100±80) 8331-8310 (0.04); 8255-8223 (0.03); 8215-7943 (0.93) large limestone orthostat
OxA·5500 ABR charcoal 9290±80 (7340±80) 8582-8572 (0.01); 8521-8509 (0.01); 8496-8089 (0.98)
OxA-5501 ABU charcoal 9140±80 (7190±80) 8352-8015 (1.00) which had been recorded
OxA-5502
OxA·5503
BAl
BAT
charcoal
charcoal
5725±65 (3775±65)
7145±70 (5195±70)
4764-4741 (0.02); 4725-4454 (0.97); 4416-4402 (0.01)
6159-6144 (0.01); 6122-6085 (0.07); 6069-6064 (0.92)
within the trench area at the
OxA-5504 BBA charcoal 7450±70 (5500±70) 6415-6160 (0.95); 6143-6123 (0.02); 6086-6064 (0.03) start of the excavation as
Note: The six dates were obtained from carbonized wood fragments recovered by flotation in the 1994 season. Contexts it was projecting slightly
beginning with the letter' A' are in Site A, the area on the peninsula; contexts beginning with the Jetter 'B' are in Site B, the
rock-shelter. The raw ages Br (before present- i.e. before AD 1950) are complemented by the equivalent uncalibrated ages
above the pre-excavation
Be (in parenthesis), in order to facilitate comparisons with dates in the published literature. In the final column are the ground surface. To the north
calibrated age-ranges, as prepared by the Oxford University Radiocarbon Accelerator Unit, using the agreed procedures
that are incorporated in the OxCal computer program, and based on the 1993 bi-decadal calibration curve. The ranges are of the orthostat was a gap
quoted at two standard deviations (95.4 per cent probability). Where the calibration process encounters 'wiggles' in the
calibration curve, it may cross the curve two or three times; the OxCal calibration program calculates the probabilities
in the wall, which may be
against each encounter with the curve, expressing that probability (for example, 0.95) as a fraction of 1.00. part of an entrance. A close
ABJ -a thin stratum of fine soil, chocolate brown in colour, into which the early prehistoric grave had been cut. ABJ examination of the south-
overlies ABR. ABJ produced an assemblage that contains no material other than the early prehistoric chipped stone.
ABR -a thin stratum of fine soil, reddish brown in colour, underlying ABJ and overlying ABU, and containing exclusively
facing section at this point
early prehistoric chipped stone assemblage. indicates a slight change in
ABU -the lowest stratum reached so far, a stratum of fine soil, charcoal grey in colour, underlying ABU, and associated
with various stony or stone-built features. the colour of the deposits
BAI -a lens full of carbonized wood below a capping of stones in a shallow pit. Together with another similar pit, the
latest sealed stratified context in the sequence, immediately below the shallow surface layer.
here. The wall shows no
BAT - deposit of deliberate fill within the area enclosed by the curving wall. BAT is high in the series of sloping deposits sign of any bonding mate-
with lenses of charcoal.
BBA -deposit of deliberate fill within the area enclosed by the curving wall; BBA is stratigraphically below BAT. rial having been used. It
seems to be a dry-stone

52
Excavations at Pmarba;a: the Early Stages

1m

G
Figure 4.4. Section on north
side of Trench 1. The radio-
carbon date on charcoal from
BAJ (see Table 4.1) came
from the pit at the top of the
section; the other two
radiocarbon dates were on
samples from the tilting
charcoal lenses below.

construction and its purpose was to act as a revetment and below the stratified deposits of ash and bone
against the soft deposits of rubbish into which the described above, the light brown, gravelly material
structure had been cut. appeared as a ridge, which had been cut away to
The fill within the wall was a series of tips and form hollows to the southeast and the west. These
dumps of material, dipping quite steeply in from the hollows await further investigation next year, when
inner edge of the wall towards the centre of the area the trench can be extended, and when the wall of the
enclosed by the wall. Some of the thinner tip-lines stone structure is dismantled.
were very rich in charcoal; other, thicker tips were It is clear that this light-coloured, gravelly de-
more silty but still had a high ash and charcoal con- posit, full of frost-fractured stone, is mostly of natu-
tent. It appears that the filling of the structure was a ral origin. At present we do not know how thick it is.
rapid process, during which the deposits did not A small test was dug into this material at the base of
have time to weather or erode. At the base of the fill the large stone structure. This test showed that this
there was no sign of a floor or trampled surface. The material contains some animal bone and chipped stone.
fill lay directly on the earliest stratum so far encoun-
tered. The wall seems to have been based on a ledge Site A- the settlement on the peninsula
cut into this earlier deposit, and the floor of the Initially, a square area, 3 x 3 m, was opened on the
enclosed space was a shallow, saucer-like depres- site, immediately adjacent to the looters' pit whose
sion within the retaining wall. spoil had provided the first indications of an early
Excavation around the west and south of the prehistoric settlement in this area . When a pithos
large stone structure revealed a succession of thin burial was encountered in the later deposits, a small
strata that all contained considerable amounts of car- extension was added at the northwest corner, and it
bonized plant remains (mostly wood charcoal) and was decided to continue the sounding in a restricted
animal bones, and are putatively of earlier date. 1 m wide strip along the southern side of the square.
Amounts of chipped stone varied from one layer to The maximum depth of deposit explored is a little
the next, but a larger sample of chipped stone has over 0.90 m, and there is no sign of the base of the
certainly resulted. Preliminary analysis of the chipped archaeological deposit. The supervisor of this area of
stone has resulted in tables of statistics that await excavation, Catriona Gibson, on whose report this
detailed analysis against the stratigraphic evidence, section of the paper is based, provisionally defined
but first impressions are that these strata represent nine stratigraphic phases.
rubbish disposal, and the chipped stone may well rep- The lowest three of the nine phases seem to be
resent an earlier phase in the occupation of the site. early Neolithic in date. They produced significant
The wall was found to be standing on a deposit quantities of well-preserved animal bone and diag-
of light brown soil and gravel, mixed with frost- nostic chipped stone, both in obsidian and flint, in
fractured limestone that originated from the roof of particular a substantial number of micro lithic pieces.
the rock-shelter. The lower part of the depression In addition a decorated stone shaft straightener was
that formed the structure had been dug into this found in these earliest levels, reminding us in its
deposit, which is obviously quite deep. This under- decoration of a very similar piece in the same soft,
lying stratum of light brown and stony material was fine-grained stone and of another shaft straightener
also found at a higher level outside the stone struc- with different decoration, both found in the surface-
ture. Here, to the southwest of the stone structure collected material in 1993.

53
Trevor Watkins

I I
~ - - - - - - - -- -- -- --- t>~- - --- -
~
I
L - - - -- Fr£ ... -- 0 I

rY --
I .
a/?D
• _ I
:0'~ ~V
~~~~CL ___ ~9~_ti~& __ _j
I . . . . . . · .. .. . . . I

Figure 4.Sb. Plan of Site A- the burial in the sound-


I I
ing into the early prehistoric levels.
E l_ _ _ _ _ _ - - - - - - - - - ------------I F
Th e latest deposit within this earliest phase is a
dark brown sandy deposit. This layer is important
because it appears to be the latest deposit that con-
tains no ceramic finds (with the slight exception in
I the northern part of the trench, where there was
L -- - · - - -- - - -- - - - - - - ------ - _j
0 c observed animal disturbance). It also seals the burial
and as~ures us of its early date, and it seems to
represent the soil development during an occupa-
Figure 4.5a. Plan of Site A- the settlement on the tional hiatus lasting thousands of years. Above this
peninsula: pithos burial and cist (?)at the base of the level, the next phase is mad e up of a badly degraded
third millennium BC levels. mud-brick horizon containing evidence of mud-brick
collapse. As well as a few sherds of pottery there is
The ea rliest phase so far encountered consists quite a change in soil types. The layers from this
of dark humic loamy layers, associated with three point upwards are much lighter and grittier I siltier
possible structures (with radiocarbon dates in the and generally less humic and loamy than those in
late eighth millennium BC uncalibrated, or late ninth the earlier phases.
millennium calibrated in calendar years). Some This phase also includes several incidences of
ground-stone tools, particularly smoothers and disturbance from a higher level, including a pithos
pounder fragments, came from this phase. Parts of a burial and what may be a cist burial. The probable
stony curvilinear feature were found, and two fur- cist only projected a short distance into the trench on
ther possible structural elements were also noted, its west side, and was not excavated. The pithos
formed from smallish and medium-sized, very well burial was set into a very tight fitting, roughly oval
compacted stones and pebbles, concreted together cut, measuring 1.15 x 0.61 m and was made up of
in a mud matrix. Another potential structural ele- parts of three separate pithoi . The burial within the
ment consists of at least three courses of rectangular pithos was that of a juvenile, but the bones were
blocks of white marl sitting in the east facing section much disordered by animal disturbance. Associated
of the 1 m sondage (on section E-F, Fig. 4.6) . with the pithos burial, and lying immediately above
The next phase consisted of a small grave in it, were some fine red painted pottery sherds and a
which a small child had been inhumed . The grave complete goblet in fine, red slipped and burnished
was sub-ovoid in plan with vertical sides and con- Early Bronze Age fabric .
tained the remains of a crouched child, lying on its Above this phase, we could identify a fragmen-
left side with arms folded . Its grave goods included tary plaster floor surface that sealed both the pithos
a nicely rounded, bun-shaped lump of red ochre, burial and the cist, and portions of the bases of a
placed just to the south the inhumation's folded arms. mud-brick wall. The pottery found in these levels is
Against the eastern side of the grave lay what ap- monochrome, red-surfaced and black at the core. It
pea rs to be a damaged adze, re-used as a smoothing seems likely that it is a limited repertoire of Early
implement, a part of a sub-spherical basalt object, Bronze Age ceramics, but at present we have too
and a fragment of a basalt grinding stone. little material to form a good idea of the corpus.
54
Excavations at Pmarba~r: the Early Stages

Finally, immediately below the surface soil, there The faunal material
was a linear feature roughly 0.65 m wide and run-
ning west-east across the trench. It appears as a The animal bone material recovered both from wet-
shallow ditch with a stony. Amongst the finds were and dry-sieving is generally in extremely good con-
high densities of fine and coarse ware pottery, prob- dition, and in most of the contexts sampled in this
ably of Roman or Byzantine date, chipped stone season seems to be present in very good quantities.
and two bronze coins, one of which is definitely Some preliminary analysis of some samples was un-
Roman. dertaken during the season, but without access to
good reference collections and more detailed study
Chipped stone it is not possible to make definitive statements. Good
samples of material of Neolithic date from the rock-
Dr Douglas Baird has been able to give a brief ex- shelter were available from a very early stage, the
amination to the material excavated in 1994, but the ratio of diagnostic to non-diagnostic bone being very
1995 material from Site B awaits proper analysis. In encouraging. There are animals of all sizes, ranging
the rock-shelter, Site B, the chipped stone pieces that from Bas and equids down to small birds, small mam-
are chronologically diagnostic indicate that all the mals, tortoise and snake, and including ovicaprids,
deposits so far excavated are Neolithic. One or two fox and hare. There are bones of larger birds, up to
sherds of pottery from the later phases and the ob- the size of goose, presumably taken as winter mi-
sidian point fragments in many of the excavated grants on the lake, but rather fewer fish bones than
contexts suggest that the latest occupation is pottery might have been expected. The equids are particularly
Neolithic, close to the date of c;:atalhoyuk itself, .Q
around 6000 to 5600 BC. The earlier deposits, includ-
ing the fills of the large curvilinear stone structure,
have point types and bi-face fragments that suggest
an earlier, seventh or early sixth millennium date.
The technology of some of the chipped stone also
indicates that there is probably an earlier, aceramic
Neolithic component in the rock-shelter, that is, some-
thing a little earlier than the currently known mate-
rial from C::atalhoyuk. Microliths and elements of a
small-scale, bladelet-based industry are also present,
in very small quantities and presumably as residuals,
in these Neolithic deposits in the rock shelter. It is
noteworthy that they are different from those from
the earliest levels in Site A, and they are presumed
to derive from a substantially earlier prehistoric oc-
cupation in or around the rock-shelter, but of a date
different from that of Site A.
In Site A, the settlement on the peninsula, there
is no chipped stone material of the Neolithic date of
the kind found in the rock-shelter. The earliest de-
posits so far excavated in this area have yielded
what appears to be (on the basis of very small sam-
ples) a relatively uniform assemblage, characterized
by quite distinctive microlithic types. Unlike the
Neolithic in the rock-shelter, a lot of flint was used
alongside the obsidian. The products of the industry
are of very small size, and the raw material was
clearly intensively and exhaustively utilized, brought '"'
as it was over considerable distances. Classic geo-
metric forms are very rare but there was a tendency Figure 4.6. Sections in Site A. The locations of the
to produce a microlith of distinctive character that sections are given in Figure 4.5a. The three radiocarbon
was an elongated, asymmetric triangle or arched dates (Table 4.1) come from the two lowest strata defined
backed piece. at the base of section C-0.
55
Trevor Watkins

prominent, and Equus hydrontinus is apparently well on the plain. Were the people who occupied the
represented. rock-shelter in the sixth millennium Be using it sea-
More difficulty was experienced with Site A. In sonally, for example, as a hunting station for killing
the Early Bronze levels, to judge from the amounts and butchering wild equids? Or were they a small
of residual early prehistoric chipped stone present, permanent community at the margins of the society
the bone samples are likely to be very mixed in date. that occupied the plain, earning their living by sup-
In the early prehistoric levels that were encountered plying, for example, meat of wild animals? Detailed
first, the amounts of bone were rather small and it study of the cultural assemblage, the animal bones
was mostly heavily splintered. In the deeper depos- and the botanical evidence will be needed to help us
its very good bone samples were retrieved. These to address this question.
have not yet been sorted even in a preliminary man-
ner, but it is quite clear that they represent a wide Environmental and economic data
range of species across the whole size spectrum from
Bas and Equus down to small mammals, birds of all The botanical material is still being examined by
sizes and a few fish bones, a spectrum not at all Mark Nesbitt at the Institute of Archaeology in Lon-
dissimilar to that of Site B. don. The great bulk of the botanical material seems
It is of interest to note that a single Dentalium to consist of wood charcoal, often in microscopic
shell, the species very much used in the Epipalaeo- pieces, and there are very few seeds. There is a strik-
lithic Natufian culture of Israel and western Jordan, ing difference in the quantities produced by the two
was recovered, but from a mixed context in Site A. sites. Site A has produced very little carbonized plant
Four Mediterranean seashells from the Epipalaeo- remains, whereas, in the 1995 season, flotation of
lithic deposits in the same area further attest inter- about one metric tonne of deposit produced just over
esting long-distance contacts and exchanges. 10 kg of carbonized plant materials from Site B. Plans
are now in hand to begin the study of the wood
The cultural assemblage from Site B charcoals.
The zoological material is also now being stud-
Given the radiocarbon dates (in the sixth millen- ied under the supervision of Dr Nicola Murray in
nium BC, uncalibrated), and the clear resemblance Edinburgh University. It is very well preserved, and
between the chipped stone tools in the latest levels is proving to be rich in potential information. This
at Pmarbaiin and those from (atalhoyiik, the cultural season, a further 34 kg of excellently preserved ani-
assemblage from the late levels of the rock-shelter at mal bone was recovered. What is immediately ap-
Pmarbaiin is distinctly eccentric. In particular it is parent is the very large quantities of bones of an
deficient in both pottery and heavy ground-stone extinct equid species, especially in the late Neolithic
tools. The late Neolithic chipped stone assemblage is deposits from Site B. Other species represented in-
very heavily dominated by obsidian, and flint is a clude very large wild cattle, sheep and goat, fox,
considerable rarity. It is striking that, although cores hare, many birds and some reptiles. The study of the
of obsidian are very rare, there are plenty of prepa- late Neolithic material will be of particular interest
ration flakes. This suggests that the occupants of the in conjunction with the nearly contemporary mate-
site were at least resident long enough to need cores rial now being excavated at (atalhoyiik. It is already
in order to make tools. The absence of other tools apparent that there is a considerable degree of dif-
than those of chipped stone, especially food prepa- ference in the bone assemblages from the late
ration equipment such as mortars, pestles and querns, Neolithic in the rock-shelter and the early Neolithic
suggests that they either did not use cereals and in Site A, the open village settlement beside the lake.
pulses or that they brought them to the site already
prepared. There is a distinct shortage of bone tools, Discussion
also. So far, only one fragment of spatula and one
bone point of a needle have been recovered. Two The two seasons of excavation have produced an
very small disc-shaped beads complete the very lim- excellent representation of the material from the late
ited repertoire of the site in the late Neolithic period. Neolithic, and possibly earlier, occupations in the
The central question relating to the later rock-shelter, Site B, and a smaller amount of mate-
Neolithic occupation concerns the nature of that oc- rial from the early Neolithic village on the penin-
cupation and the relationship of the occupants to the sula. In addition, we have a small sample of the
communities that lived on hoyiiks like (atalhoyiik later, Early Bronze Age, occupation in Site A. We

56
Excavations at Pmarba~n: the Early Stages

have the first, small set of radiocarbon dates (Table back at least into the middle of the eighth millen-
4.1), which place the early occupation of Site A in the nium Be. This sequence of occupation should be
second half of the eighth millennium BC (uncali- placed in the context of some knowledge of the local
brated), and the last significant phase of occupation environment when Dr Neil Roberts' team has com-
of the Site B rock-shelter in the sixth millennium Be pleted work on the core from the bed of the lake at
(uncalibrated), with later re-use in the early fourth Siileymanhaci, only about 4 km from the sites at
millennium BC. We now know, therefore, that small- Pmarba~1.
scale, bladelet-based, microlithic industries contin- Contacts were made in 1995 with the interna-
ued beyond the Epipalaeolithic into the Neolithic tional team, led by Prof. Yalcinkaya of Ankara Uni-
period in Anatolia, following a different develop- versity and Prof. Marcel Otte of Liege, which is based
mental history from the Levantine industrial se- at Kara'in, near Antalya, and is working in that cave
quence, which has been the only one known in detail. and at the nearby site of Okuzini. It is hoped that in
The sampling strategy and recovery methodology future we shall be able to exchange research visits
have produced excellent samples of carbonized plant with this team and work towards ensuring direct
remains, and plans for extensive research on the comparability of the results from the two teams. The
animal bones, the seeds and the wood charcoals, and opportunity to co-ordinate work and working meth-
the chipped stone industries are now in hand. ods between the important stratigraphic sequences
The complexity of the stratified deposits, and from Epipalaeolithic to early Neolithic near Antalya
the need to move at the speed at which the materials 'and those on the southern plateau is exciting, as is
can be processed mean that the pace of excavation the prospect of working in direct comparison with
will necessarily be slow. However, the potential ex- the contemporary material coming from <;atalhoyiik
ists to explore a long sequence of occupation reaching itself.

57
ChapterS

Topographic Survey of the ~atalhoyiik Mounds

Tom Pollard, Colin A. Shell & David R. Twigg

The two C::atalhoyiik mounds are situated on the description and spatial study of the sites and
Konya alluvial plain some fourteen kilometres north their immediate environment using a computer-
of the town of C::umra, and about two kilometres based Geographical Information System (GIS),
south of the village of Kii<;iikkoy, the road to which and also provide possible insight into the sur-
passes by the north side of the west mound. The face taphonomy of the mounds.
relict course of the C::ar§amba river passes between
the mounds (Fig. 5.1). The environs are intensively Methods
cultivated for cereals and horticultural crops, includ-
ing the well-known tomatoes of the C::umra region. The first task of the survey was to establish an over-
Development of the irrigation system is bringing all site grid. Being one of the more elevated features
raised concrete channels across the surrounding land- in the Konya plain, the larger east mound has a
scape and with it some reorganization of nearby triangulation point on its summit, the height of which
fields and ditch sytems. The area between the mounds is marked on the 1:25,000 map as 1020.3 m a.s.l.,
is occupied by established small orchards and a plan- some sixteen metres above the surrounding plain.
tation of poplar, no doubt taking advantage of the The UTM Zone 36N coordinates for this datum were
higher moisture regime of the old stream course. determined subsequently to be 484,844 mE, 4,168,967
The surrounding alluvial plain is at about 1004 m m N, using a Global Positioning System (GPS) -see
a.s.l.. Appendix 1 for a listing of equipment employed in
the survey. The datum point, from which the survey
Objectives originated, was assigned the nominal coordinates of
1000 m E, 1000 m N in the site grid, which was
It was clear from the outset that for a project of the intended to encompass both the east and west
scale anticipated for C::atalhoyiik, one of our primary mounds. Magnetic north measured from this point
concerns should be to obtain an accurate record of was found to lie almost in the direct line of the
the surface topography of both the east and west minaret of the mosque in the village of Kii<;iikkoy.
mounds before they were disturbed by further exca- The centre line of this landmark from the datum
vation or other archaeological activity. point was adopted as the site grid's north reference
There were four main aims for the survey: direction.
1) that it would generate an accurate model of the The initial placement of the grid pegs was de-
surface of the hoyiiks; termined in the first instance by the requirements of
2) that the model would provide a secure frame of the surface collection programme: a 2 x 2 m square
reference for the systematic surface sampling ex- was to be investigated systematically at 20 m inter-
ercises and the excavations which would even- vals across the entire site. Where appropriate, more
tually proceed down from the surfaces; substantial permanent markers were put in at 50 m
3) that it would be sufficiently detailed to highlight intervals to aid the precise placement of trenches.
micro-topographical variation which might re- The east mound was gridded during the first days of
late to sub-surface features (e.g. buried walls) the 1993 season, while the west mound was covered
and also the more recent history of the mounds; early in the 1994 season once the surface collection
4) that the model would form a basis for the on the east mound was complete.

59
Tom Pollard, Colin A. Shell & David R. Twigg

4170000

4169000

484000 485000
Figure 5.1. Plan of the c;atalhoyiik mounds and their immediate surroundings taken from a local map. Contours at 0.5
m intervals above 1004.00 m. Scale 1:8000.

60
Topographic Survey of the <:;atalhoyiik Mounds

Surveys trenches. The area was recorded in a strictly topo-


graphic way, not on the systematic 2 m grid, taking
The east mound (1993-1995) as many readings as necessary to describe the fea-
The east mound was the main focus of the topo- tures.
graphic survey, as it was here that the bulk of the The survey data were downloaded at the end
archaeological work would take place. It is enclosed of each day from the datalogger to a computer run-
by a fence, erected following Mellaart's excavation ning the necessary software for the equipment being
of the site in the 1960s (Mellart 1967). The area en- used (the different total stations recording in differ-
closed measures some 500 m north-south, and 300 ent formats). The data were edited and 'cleaned up'
m east-west. The mound is in fact composed of three (i.e. had additional or unecessary digits or leaders
sub-mounds or eminences: the main, or southern, removed) before transfer to a contour generation
eminence, which rises to about 17m above the sur- program. Daily plots were produced of the points
rounding plain and was the location of Mellaart's taken, and the contours generated to check that the
work; a smaller northern eminence, rising to about surface was accurately recorded. Occasionally, where
12 m above the plain; and a much lower plateau to the recording was found inadequate we were able to
the east of the main eminence. Other than the occa- return the following day to a specific location to
sional fox-hole and some modern trackways and rectify any omission or refine the detail. Each day's
other disturbance, the mound appeared to be uni- data were incorporated into a larger site map and
formly smooth and featureless. the progress of the survey could be monitored. Over
The desired topographic detail could only be the two seasons of survey, approximately 25,500 read-
achieved through a dense coverage of readings. Just ings were taken on the east mound: we believe an
how dense that coverage should be was a topic of unprecedented coverage of a site of this size.
debate, but a 2 m spacing was settled on as a prag- Major features are fully delineated in a contour
matic balance between the desire for thorough re- plot of the mound at 0.5 m intervals (Fig. 5.2). The
cording of detail and the limitations of time. It was extent of Mellaart's deep excavations are apparent
accepted that while the smallest surface features on the west side of the main southern mound, with
would likely be overlooked at this resolution, the the associated long linear dump at the western edge
more substantial ones (if they existed) would almost and three substantial spoil tips on the southwestern
certainly be picked up to some degree. In this it slope of the mound. The north-south axiality of both
should be recalled that the rooms excavated by the southern and the subsidiary northern mounds is
Mellaart ranged in size from two metres upwards. well defined. The low plateau (1100,1040) to the east
The technique for the detailed survey was of the main mound appears to comprise of a later
straightforward. One 20 m square at a time was smaller phase superimposed on a larger low area
marked out with poles, which acted as guides for the extending a further 50 m to the east. The cutting
staff-holder, who walked each square boustrophed- down of the surface between the summits by the
on in lines approximately 2m apart, with readings track running northwest to southeast is clearly vis-
taken by electronic total station every two metres. In ible, as also is the line of the track on the east side of
this way up to 1500 measurements could be taken the mound running southwards from by the water
during a working day, with a potential three-dimen- tower and pumping plant (1120,1280). The site of a
sional precision of the order of 5 em. The grid was small irregular excavation with a dump to the north
not always rigidly adhered to, particularly where side is revealed at (990,1230) on the northwest slope
there was modern disturbance, or lack of detail off of the north mound. A more detailed study of the
the mound. For instance, in the northwest corner of micro-topography and the evidence for the
the fenced area is a low seasonally marshy area weathering of the mound are discussed separately
which, though enfenced, is off the mound and al- below.
most completely flat. Lower density recording was
possible here at approximately 5 m intervals with- The west mound (1995)
out a significant loss of detail. This area, inciden- The west mound is very different from the east,
tally, is the location of the new dig-house, where the being a low spread structure. The site is unfenced,
excavation of the foundations confirmed the absence but is nevertheless clearly delimited by modern fea-
of archaeological material. Mellaart' s excavations re- tures; the northwest side is curtailed by a road, while
quired more attentive treatment, owing to the much the northeast side is truncated by a large modern
more varied topography of old archaeological drainage channel. The mound clearly extends into

61
Tom Pollard, Colin A. Shell & David R. Twigg

130

125

1200

1150

1100

1050

100

950

850 900 950 1000 1050 1100 1150 1200

Figure 5.2. Plan of the east mound contoured at 0.5 m intervals. Scale in metres.
Topographic Survey of the <;atalhoyiik Mounds

adjacent fields at its southern end, where consider- grid to provide a reference point for the portable
able quantities of Chalcolithic pottery were found receiver with which readings were made by the 20m
during the surface collection, and similarly to the grid pegs, at significant surface features and along
northwest beyond the modern road. Like the main the bounds of the site. The data from the two receiv-
east mound its margins are almost certainly covered ers were then processed at the excavation labora-
to some degree by later alluvium, masking its full tory, and further upon return to the UK. The result
extent. Although much lower than the east mound was a generalized record of the centre of the 9 ha site
(the summit is only 6 m above the plain), the west from about 400 readings. The GPS survey also estab-
mound is still in excess of 300 m in length and lished the orientation of the site grid on the west
breadth. mound with respect to UTM north. The local grid is
The site has suffered significant damage from oriented 5°35' west of UTM north - allowing the cor-
recent farming practice. A deep trackway cuts across rect positioning of the east mound plan in Figure 5.1.
the middle of it, effectively dividing it in half. The In the plan of the west mound contoured at 0.5
site has been used for many years for the storage of m intervals (Fig. 5.3), the severe dissection of the
straw for fodder and mud-brick manufacture, which mound by the north-south track is seen, with the
is kept in large rectangular-based piles 2 m or more surface cut down by at least 0.5 m. The axis of the
in height. Frequently, shallow trenches are dug mound appears to have been more in an east-west
around these piles to channel winter rain away from orientation, and as on the east mound, there is evi-
them, and the excavated soil mixed with water and dence for a specific construction phase where there
applied over the heap to form a waterproof coating. was an extension of the mud-brick settlement in a
The result of all this activity is that much of the particular direction, in this case to the south, creat-
mound is deeply rutted, and in places covered by up ing a low plateau.
to 0.5 m of decaying straw.
It was not intended that the west mound would Micro-relief and surface taphonomy of the east
be subject to the same degree of archaeological scru- mound
tiny as the east mound. However, it was initially felt
that both mounds should be recorded to the same When first visiting <;atalhOyiik in 1958, Mellaart
level of accuracy, as a vital part of our role as archi- (1967, 27) noticed that the remains of burnt build-
vists. In the event it seemed, given the already sig- ings were exposed by surface erosion on the south-
nificant alteration of the surface, that an accurate west side of the east mound. Elsewhere the mound
survey would probably require, if anything, a much was covered and stabilized by vegetation. The sub-
higher, and prohibitive, density of readings than was sequent fencing of the site has maintained the veg-
necessary for the east mound. It was decided that etation cover. The evidence for the survival of houses
such a level of detail, of what is effectively an en- near the surface of unexplored parts of the mound,
tirely modern surface, was somewhat superfluous, and in particular their form and shape, was a princi-
and that for the present, an outline topography would pal concern at the beginning of the new study.
be obtained which could be elaborated in the future The detailed topographic survey was under-
if required. As an outline, it was felt that measure- taken to discover if the surface relief, itself, pre-
ments based on the 20 m site grid already in place, served any evidence for structures and their form in
with significant breaks of slope recorded, would give the unexcavated areas of the site. To seek such evi-
an adequate picture of the general form of the mound. dence, the topographic survey data were examined
As the total station was fully occupied in the with software able to visualize the finer features
recording of the excavations on the east mound, the present. The data were gridded using a quintic inter-
survey was conducted using the Global Positioning polation at a 0.5 m interval from the triangular ir-
System (GPS) receivers. The equipment consists of regular network (TIN) defined by the array of points,
two antennae linked to dataloggers, one at a fixed and a lattice constructed that could be viewed in the
location, the other recording the detail of the site. GIS software, Arc/Info. Fine detail may be extracted
The receivers monitor a series of satellites which and enhanced from a dataset by high-pass filtering
transmit radio signals modulated with precise time techniques (Lillesand & Kiefer 1994, 553). Alterna-
and position data, allowing the relative positions of tively, using appropriate software, a model surface
the two receivers to be calculated to within 2 em (see of the mound can be viewed and illuminated from
Twigg in press). The fixed receiver was placed on any direction, so that features may be revealed by
the top of the east mound at the origin of the site low oblique lighting digitally equivalent to that

63
Tom Pollard, Colin A. Shell & David R. Twigg

41691~0

UUI lone J6N Ccxrd IMtoa

CrotOU""a ot o. Sm Intorvo le

~'.6Q100
All levola rolatlvo to QOOI....Ind level
at bOlO crt p II I Qf' en Eoet Mou'ld
with oaaUIT»d revel ot t020 . Jm

~....

4168050

Figure 5.3. Generalized plan of the west mound at c;atalhoyuk with contours at 0.5 m intervals. UTM coordinates in
metres.

successfully used in aerial photography to discover include the vestiges of pits dug for brick making, as
sites of low relief. Figure 5.4 is a plan view of the east recorded by Mellaart (1967, 32), though those of the
mound shaded by low illumination from the north- Hellenistic and earlier periods could be expected to
east, and Figure 5.5 is an image illuminated from the have largely silted up in the intervening years. The
northeast with contours and grid superimposed. Fig- numerous slight mounds and hollows from colonies
ure 5.6 is a view of the shaded surface from the east. of burrowing European suslik (citellus citellus) are
In addition to the major features defined in the not resolved. Finally, the remaining most obvious
contour plan (Fig. 5.2) the shading shows a number features are aligned on the survey grid and repre-
of modern elements which confirm the sensitivity of sent artefacts of the survey method where smaller-
the technique. A 10m square from the scraping exer- rors of a few centimetres in the transfer of the level
cise is clearly visible at (1040,1240). On its north side from one surveyed area to the next have introduced
is a narrow cutting, associated with the water sup- step edges aligned north-south and east-west. Within
ply, running west southwest across the site from the these areas there is further smaller striping associ-
water tower. Visible also are the barrow runs from ated with slight alignment errors in the holding of
the 1960s' trenches to the three spoil tips on the the detail pole. These artefacts verify the high sensi-
southwest side of the mound. A further 10m scraped tivity of this mode of displaying the data, and con-
square is visible at (1075,1175), immediately to the firm that there is no obvious fine-scale detail visible
north of which is a fox-hole. The irregular hollows on the mound that relates to the immediately sub-
from collapsed similar burrows, are probably the surface structures- a conclusion not entirely unex-
most common feature across the site, but they may pected given the material similarity of the room fills

64
Topographic Survey of the <;atalhoyiik Mounds

Figure 5.4. Surface of the east mound digitally shaded from the northeast by a point source at 30° elevation. Scale 1:2500.

65
Tom Pollard, Colin A. Shell & David R. Twigg

Figure 5.5. Northeast shaded surface of the east mound with 0.5 m contours superimposed. Scale 1:2500.

66
>-l
0
'1j
0
OQ
...,
tll
'1j
::;'
;:;·
if...,
<
ro
'<
0
......
....,.
::;'
ro
..()
tll
....,.
tll
s:o:
'<
C:
:>;""'

3:
0
c
::l
0.
ffl

Figure 5.6. Northeast shaded surface of the east mound viewed from the east.
Tom Pollard, Colin A. Shell & David R. Twigg

Figure 5.7. Approximately north-south profile of the east mound at c;atalhoyiik across the summits.

68
Topographic Survey of the <:;:atalhoyuk Mounds

Figure 5.8. West-east profile across the main summit of the east mound at (atalhoyiik .

69
Tom Pollard, Colin A. Shell & David R. Twigg

Figure 5.9. West-east profile across the northern subsidiary summit of the east mound.

70
Topographic Survey of the <;atalhoyiik Mounds

and the mud-brick walls themselves. Area and volume of the mounds
On the macro-scale, however, there may be one
aspect of the shape of the main mound that relates to An accurate estimate can be made of the areas and
the mud-brick buildings. The general orientation of volumes of the mounds from the present surveys.
the structures so far excavated is north-south/ east- The west mound has a surface area of 6.59 ha within
west. On the southeast side of the mound there is a the bounds of the surveyed area, and a volume of
change in the general curvature of the contours at 157,000 m 3 above the 1004.00 m level of the sur-
the position of a line running southeastward from rounding landscape. However, if the dense spread
the summit towards (1040,950). This could represent of artefacts in the surrounding level alluvium is in-
the slightly higher resistance of the mound to ero- cluded, the area of the west mound is 8.5 ha. The
sion associated with the aligned corners of mud- east mound has a calculated surface area of 13.56 ha
brick walls in this area. There is, though, a slight within the boundary fence, and an estimated vol-
concavity in the contours to the west of this line, ume of 677,700 m 3 above the 1004.00 m level.
confirmed by the shading, which may be the re-
mains of a past surface mud slide in this region, Conclusion
which, itself, may be the reason for the slight ridge
to the east. A similar slight hollow is also apparent Whilst the detailed topographic survey of the main
on the northwest slope of the mound at (960,1110). east mound at <;atalhoyiik failed to reveal evidence
for the underlying structures, it has provide a com-
Erosion and long-term stability plete frame of reference for not only the ongoing
The present form of a prehistoric tell represents a investigation of the site, but also is a digital founda-
combination of the original structural development tion for the study of the site and its environs. The
of the mound from the successive building and re- survey contributes also a valuable conservation
placement of mud-brick houses, and the subsequent record of the state of the mounds at the beginning of
erosional processes. The erosion will depend not only the new major programme of research at this inter-
on the climate and vegetation cover, but is mediated nationally important site.
also by both the original form of the buildings and
nature of their materials- stone footings to walls Acknowledgements
may for example add stability to a mound. On present
evidence the east mound at <;atalhoyiik appears to be We would like to take this opportunity to thank
entirely constructed of mud brick derived from local those people who suffered so long at the other end
alluvium, with the surface virtually free of stone of any of the survey operation, in particular to Mehmet
size except for quem fragments, and its consequent Uluceviz and Ali Tiirkcan. Also to Tom Strasser,
shape is, as expected, generally smooth and uniform. Naomi Hamilton, Nurcan Yalman, Bekir Giirdil,
The erosion processes on tells are a well estab- Ibrahim Oziinoglu, James Conolly and Murat Asian.
lished interest (Davidson 1976; Rosen 1986), which Many thanks are due to Chris Burgess and the De-
draws on the extensive studies of agricultural and partment of Archaeology, University of Edinburgh,
natural soil erosion. The formulation of a Universal for the loan of the Wild TC1000 total station in the
Soil Loss Equation by Wischmeier & Smith (1978) 1993 season.
was a first step in general model quantification. The
predicted long-term form of a uniformly eroded
mound of essentially homogeneous alluvium is Appendix 1
Gaussian in profile (Rosen 1986, 28), with eroded
soil collecting in extending surfaces around the Survey equipment and data processing
mound. From the present survey, detailed profiles
across the summit of the east mound and its north- 1993 Wild TC1000 total station with datalogger.
ern hump (Figs. 5.7-5.9) show that it is close to this Data processed through Gif 12 and Microsoft
state and is likely to survive without significant Excel, then Surfer and Autocad.
change other than slow deflation if the present veg-
etation cover is maintained. There are no deep con- 1994 Geodimeter 420 with Psion Organiser II as
cavities in the surface of the mound where significant datalogger, running Geolog datalogging pro-
rainwater erosion gullies can be expected to form gram. Data processed in Map400, then Surfer
and threaten its long-term stability. for Windows, Autocad and Arc/Info.

71
Tom Pollard, Colin A. Shell & David R. Twigg

1995 Sokkia SET SE, with Psion Organiser II, run- Image Interpretation. 3rd edition. New York (NY):
ning Toposet datalogging program. Thereaf- John Wiley.
ter Surfer (Win32), Autocad, Arc/Info, AVS. Mellaart, J., 1967. Catal Huyuk: a Neolithic Town in Anatolia.
Wild System 200 single frequency GPS, with London: Thames & Hudson.
Rosen, A.M., 1986. Cities of Clay: the Geoarcheology of Tells.
data processed with SKI software, thereafter
Chicago (IL): University of Chicago Press.
LISCAD, Microstation, Autocad and Arc/Info. Twigg, D.R., in press. GPS and its use for terrain mapping
and monitoring, in Landform Monitoring, Modelling
References and Analysis, eds. S.N. Lane, K.S. Richards & J.H. Chan-
dler. (BGRG Symposium Series.) Chichester: Wiley.
Davidson, D.A., 1976. Processes of tell formation and ero- Wischmeier, W.H. & D.D. Smith, 1978. Predicting Rainfall
sion, in Geoarchaeology: Earth Science and the Past, Erosion Losses: a Guide to Conservation Planning. (U.S.
eds. D.A. Davidson & M.L. Shakley. London: Department of Agriculture, Agriculture Handbook
Duckworth, 255-75. no. 537.) Washington (DC): U.S. Government Print
Lillesand, T.W. & R.W. Kiefer, 1994. Remote Sensing and Office.

72
Chapter6

Systematic Surface Collection

Roger Matthews

Introduction and methodology may be established. Thirdly, degrees of fragmenta-


tion of all categories of artefact, in absolute and rela-
A programme of collection of surface artefacts from tive terms, can be established and examined across
2 x 2 m squares at 20 m intervals was executed the mound. Fourthly, we hope to establish the extent
across both the east and the west mounds. A uni- to which patterning of artefacts on the surface of the
formly dense coverage of vegetation across the mound relates to the distribution and nature of sub-
mounds rendered surface collection difficult and surface artefacts and features. These patterns are
unreliable, so that the technique was devised of clear- looked for in the full knowledge that they are certain
ing vegetation from each 2 x 2m square before tak- to be the result of a complex web of cultural,
ing from that square a set volume, 36litres, of surface depositional and post-depositional factors which are
earth to be passed through a 5 mm mesh from which unlikely fully to be apprehended in our interpreta-
all artefacts were retrieved. Initial experiments with tions.
a 2 mm mesh soon established that its use was pro- Surface and subsurface work at a number of
hibitively time consuming and that no significant Near Eastern sites and elsewhere, however, has dem-
improvements in artefact recovery were thereby onstrated that surface artefact patterns may signifi-
achieved. Artefacts recovered from each 2 x 2 m cantly relate to patterns of occupation in deposits
square were recorded on set forms under the head- immediately underlying the mound's surface and,
ings of pottery, bone, building materials (divided in a more general sense, have demonstrated the va-
into daub, tile and plaster), obsidian, flint, ground lidity of employing a systematic approach to the
stone, shell, and other materials. Each category of collection and analysis of surface material. The first
object was divided into two size classes: below 2 em explicitly systematic attempt to deal with surface
(largest dimension) and above 2 em. Within these artefacts was made in the 1960s by Binford at the
classes each artefact category was then counted and Native American site of Hatchery West, where clus-
weighed. For the obsidian and flint, the use of three ters of classes of artefacts were interpreted as traces
size classes - below 1 em, 1-2 em, and above 2 em - of specific occupations or activity loci. Subsequent
provided finer resolution. All pottery sherds were excavation of the site, however, detected no signifi-
further divided into wheelmade or handmade cat- cant correlations between surface and subsurface
egories. materials and features (Binford 1972).
The aims underlying these fairly labour inten- The prehistoric site of Ayngerm in southeast-
sive procedures are manifold, but have in common a ern Anatolia was treated in a similar manner by
quest for significant patterning in the nature of sur- Whallon, again in the 1960s, with artefacts being
face artefacts. Firstly, we are looking for absolute collected from 10 per cent of the site's surface. Mate-
densities of artefacts, in total and divided into cat- rial was collected and analyzed from a total of 99
egories, across the mound and, secondly, for the squares, each measuring 5 x 5 m. The results were
relative frequency with which one or more types of presented in a series of contour maps, showing den-
artefact occur in association with other types. Pat- sities of different types of artefact across the mound's
terns of obsidian distribution or of handmade sherd surface, with the aims of indicating chronological
distribution, for example, may thus be detected, and variations in occupation, the differential distribution
correlations between bone and pottery, for example, of artefact types and the correlation between various

73
Roger Matthews

1300

1200

.£i 1100
8
~
s
1000

900

800

900 1000 1100 1200


metres east

0 50 100 150 200 250


number
Figure 6.1. Enst mound: surfnce distribution of n/1 nrtefncts.

74
Systematic Surface Collection

artefact categories, all with the intention of provid- mound of 19.79 artefacts per m 2 • At the Neolithic site
ing information upon which to design an excavation of ~ayonii 15,000 artefacts were retrieved from 2100
strategy (Whallon 1980). The sites of ~ayonii and m 2 (84 squares, each 5 x 5 m), an average density of
Girikihaciyan, both in southeastern Anatolia, were 7.14 artefacts per m 2, while at the Halaf period site of
also subjected to systematic surface treatment, this Girikihaciyan 15,000 artefacts were recovered from
time by Redman and Watson, in the 1960s. In both 2725 m 2 (109 squares, each 5 x 5 m), an average
cases 10 per cent of the mounds, in 5 x 5 m squares, density of 5.5 artefacts per m 2 • The higher density of
was collected. Contour maps of artefact densities artefacts at ~atalhoyiik may be at least partly ac-
show ratios between artefact categories. Test trenches counted for by the use of sieves in the recovery
at both sites established a very strong correlation procedure.
between the nature of surface and subsurface arte- The surface distribution of all artefacts is de-
fact assemblages, to a depth of up to 0.5 m (Redman picted in Figure 6.1, showing a fairly even spread
& Watson 1970; Watson & Leblanc 1990). across much of the mound, with notably higher oc-
The programme of surface collection at ~atal­ currences on and around the three eminences. The
hoyiik is designed as an integral element of a tripar- most consistently high densities occur on the south-
tite strategy involving surface work, subsurface eastern slopes of the mound, while the lowest fre-
scraping and excavation. To some extent, each ele- quencies are to the west, north and east of the
ment is regarded as informative and significant on northern end of the mound. These distinctions are
its own terms while at the same time providing data largely the result of the distributional pattern of pot-
which will shape the wider strategy. Thus, surface tery, generally wheelmade pottery, which reflects a
studies have provided important information which high concentration of post-Neolithic occupation in
can stand by itself, but which also has relevance for the southeastern area of the mound.
subsurface scraping and for full-scale excavation.
West mound surface collection
East mound surface collection
Artefacts were collected from a total of 213 sample
In all, a total of 298 squares, each measuring 2 x 2 m, squares on the surface of the west mound, at 852 m 2
were collected from the east mound. This represents representing 1.00 per cent of the 8.5 ha of the mound.
a total area of 1192 m 2, or 0.88 per cent of the total Recovered objects have the following make-up:
surface area of the mound if calculated at 13.5 ha.
The fact that the sampled proportion is not 1.00 per Item Number recovered %age of total Density per m 2
cent (each 2 x 2m square being 1.00 per cent of the 20 Pottery 17,124 81.73 20.10
Bone 2356 11.24 2.77
x 20 m squares which comprise the sample grid) is Clay I daub/tile/plaster 413 1.97 0.48
due to the fact that collection was not carried out on Obsidian
Flint
922
82
4.40
0.39
1.08
0.10
squares in the area of Mellaart's excavations, where Ground stone 20 0.10 0.02
Other (metal, slag, shell, glass) 36 0.17 0.03
large spoil heaps and upcast material would not
Total 20,953 100.00 24.59
have provided accurate indications of surface at-
tributes. Although not in absolute terms a large pro-
portion, we are confident in the validity of the Artefact density across the west mound, de-
conclusions given the relatively close spacing, at picted in Figure 6.2, is thus notably higher than for
20 m intervals, of collection squares and the system- the east mound, and is largely accounted for by the
atic use of 5 mm sieves throughout. The overall arte- greater quantities of pottery, sherds being almost
fact category make-up is as follows: twice as dense per m 2 on the west mound. This dif-
ference indicates the increased use of pottery during
Item Number recovered o/vage of total Density per m 2 the Chalcolithic period. Were it not for the Classical
Pottery 12,834 54.42 10.77 occupation on the Neolithic east mound, the density
Bone 7055 29.91 5.92
Clay I daub/tile/plaster 1851 7.85 1.55 of pottery there would be much lower still. The sec-
Obsidian 1547 6.56
Charred remains 120 0.51
1.30
0.10
ond major distinction is the greater density of bone
Flint 61 0.26 0.05 per m 2 on the east mound, where bone is more than
Ground stone 50 0.21 0.04
Other (metal, slag, shell, glass) 67 0.28 0.06 twice as common per m 2 than on the later west
Total 23,585 100.00 19.79 mound. Explanations for this clear distinction are
not immediately obvious, but it may be that the
Thus, a total of 23,585 artefacts was recovered extent of agricultural activity across the west mound
from 1192 m 2, giving an average density across the has led to increased disintegration of bone. It may
75
Roger Matthews

1100

1000

900

800

400 500 600 700


metres east

0 50 100 150 200 250 300 350 400


number

Figure 6.2. West mound: surface distribution of all artefacts.

also be that the high density of bone on the east Chalcolithic inhabitants of the west mound may have
mound is a result of dumping kitchen refuse in open taken their refuse off-site to areas now buried by
areas within the Neolithic settlement, while the recent alluviation.

76
Systematic Surface Collection

References Site in Southeastern Turkey. (UCLA, Institute of Ar-


chaeology Monograph 33.) Los Angeles (CA): UCLA,
Binford, L.R., 1972. Hatchery West: site definition- sur- Institute of Archaeology.
face distribution of cultural items, in An Archaeologi- Whallon, R., 1980. The systematic collection and analysis
cal Perspective, by L.R. Binford. New York (NY): of surface materials from a prehistoric site in south-
Seminar Press, 163-81. eastern Anatolia, in The Joint Istanbul Chicago Uni-
Redman, C.L. & P.J. Watson, 1970. Systematic, intensive versities' Prehistoric Research in Southeastern Anatolia,
surface collection. American Antiquity 35, 279-91. eds. H. <;:ambel & R.J. Braidwood. Istanbul: Istanbul
Watson, P.J. & S. Leblanc, 1990. Girikihaciyan: a Halafian University Faculty of Letters, 207-19.

77
Chapter 7

Surface Scraping and Planning

Roger Matthews

Introduction and methodology only a few centimetres below the modern surface of
the mound.
During the 1993-1995 seasons an extensive pro- In essence, there are three stages in the tech-
gramme of surface scraping was conducted at nique of scraping. Firstly, after removal of vegeta-
<;atalhoyiik. As with surface collection, this pro- tion all surface artefacts are collected and kept for
gramme is very much an integral part of the total study. Secondly, a spit of surface earth is removed
strategy, while at the same time providing signifi- using pick-axes, shovels and baskets. All artefacts
cant results in its own right. In overall terms, surface from this earth are again kept separately. Thirdly,
scraping in the 1993-1995 seasons has greatly en- the resulting surface, which is generally some 0.1 m
hanced our knowledge of occupation at <;atalhoyiik, below the modern surface, is scraped clean with
yielding new and important data as well as indica- wide-bladed hoes. These hoes are not available in
tors for future work. The main aim of scraping is to the region of <;atalhoyiik, but the local blacksmith
investigate and establish patterns of architectural and was able to manufacture some first class tools for
occupational activity across the site, upon which our use. The blades are wide and very sharp, in
chronological and functional distinctions can be order to cut the mound's surface and scrape back all
based. This information, often extremely detailed, loose earth which is then removed by shovels and
may then be used as a prelude to full excavation in baskets. If after this scrape no architecture or soil
selected areas. changes are visible it may be necessary to remove a
The technique of subsurface scraping is one further spit of earth before rescraping. Again, all
that has long been employed by Near Eastern ar- artefacts from the scraping itself are kept separately.
chaeologists, often purely as a prelude to excava- In general, visibility of surface architecture is excel-
tion. Its use as a full-blown technique in its own lent at <;atalhoyiik after removal of surface earth
right, in order to provide detailed information with- and scraping. Detectable features include walls,
out excavation, took its most coherent and explicitly which are frequently defined by white plaster faces
strategic shape at the Sumerian city of Abu Salabikh and often with internal brick edges and mortar lines
in southern Iraq, where a substantial proportion of clearly visible, while interior features such as fire
the surface architecture was planned in this way installations and platforms are regularly discernible.
(Matthews & Postgate 1987; Postgate 1983). Pits and rubbish spreads are also common.
At <;atalhOyiik, the density of surface vegeta- Once the square has been scraped clean, all
tion and possible soil build-up initially raised some architecture and features are planned and recorded
doubts about the applicability of the scraping tech- in detail using standard forms and soil description
nique. Once in use, however, it was soon established methods. In this volume the results of the scraping
that, on the whole, the site is greatly suited to scrap- operation are presented in descriptive form with as-
ing, with architecture and features clearly visible at sociated plans of scraped areas. The very detailed
depths varying from 0.05 to 0.3 m below the surface. deposit descriptions, carried out using standard soil
In general, greater amounts of soil had to be re- science methods, are not included here but are held
moved as scraping progressed further down the in the <;atalhoyiik Research Trust archive of field
mound's slopes (Figs. 7.22-7.25). In several areas, how- records.
ever, extremely clear architecture could be traced The original intention had been to scrape

79
Roger Matthews

850 900 950 1000 1050 1100 1150 1200


Figure 7.1. Plan of east mound showing location of all scraped squares.

80
Surface Scraping and Planning

sample areas measuring 20 x 20 m within a 100 m Neolithic buildings were exposed immediately un-
grid across the site. Several factors persuaded us to derlying the mound's surface. In essence, surface
change the sample area to 10 x 10m within a 50 m scraping strongly supports the conclusion of sys-
grid . Firstly, the size of the east mound, 13.5 ha tematic surface collection that the east mound of
visible above the modern plain, meant that the sepa- C::atalhoyuk is almost entirely a Neolithic mound
ration of sample squares from each other by 100m in with a scattering of Hellenistic, Roman and Byzan-
each direction suggested we might end up completely tine occupation in restricted localities. The following
missing vital areas of the mound. Secondly, sub- discussion presents the salient results of scraping in
stantial soil build-up on the lower slopes of the all the investigated areas, commencing with the most
mound would have meant a large investment in productive area at the north of the east mound.
time and labour in order to clear 400m 2 (20 x 20m),
whereas an area of 100 m 2 (10 x 10 m) could be North Area
cleared and scraped in one or two days, an impor-
tant consideration given the relatively short dura- Investigation of the North Area began with the clear-
tion of each season. ance of sample square 1040,1190 in the 1993 season.
The policy from the start was to expand those It was immediately clear that the Neolithic architec-
scraped sample areas which appeared worthy of fur- ture in this part of the mound had not suffered any
ther investment in time and labour. Thus, two ex- considerable post-Neolithic disturbance and that the
tended areas were uncovered on the east mound, architecture was readily accessible only a few centi-
one at the north end and the other near the centre. metres below the modern mound surface. Accord-
The former of these proved to be, by some way, the ingly in 1993 the scraped area was expanded to a 20
most informative and significant of all the areas in- x 20m square with its southwest corner at 1030,1180.
vestigated. In total33 squares, each covering 10 x 10 In 1994 a further twelve 10 x 10 m squares were
m, were scraped and planned on the east mound, cleared, expanding the North Area into a 40 x 40 m
giving a total of 3300 m 2 or 0.33 ha of planned sur- area with its southwest corner at 1020,1170. Finally,
face architecture and features. This total represents in 1995 three further squares were scraped to the
2.44 per cent of the mound's visible surface area. The south of the 1994 area, giving a total of nineteen 10 x
distribution of these scraped squares is shown in 10m squares cleared and planned in the North Area.
Figure 7.1. In addition, three 10 x 10m squares were By these means we have recovered a detailed and
scraped and planned on the west mound, where largely coherent plan of over thirty Neolithic struc-
special problems were encountered, as detailed be- tures over an area of 1900 m 2 .
low.
In general terms, sub-
surface scraping at C::atal-
hoyiik in 1993-1995 es-
tablished that Neolithic
dwellings and occupation
debris occur at almost all
points of the east mound
so far investigated, but also
that in some areas these lev-
els have been substantially
disturbed by much later oc-
cupation, rendering the
Neolithic occupation not
comprehensible without
full-scale removal of the
overlying features and de-
posits, which date to the
Hellenistic, Roman and
Byzantine periods. At other
points, however, more or
less completely undisturbed Figure 7.2. Scrape areas on (atalhdyiik East in 1993 as seen from the northeast of the site.

81
Roger Matthews

I
r
....
' -

~
m

0 5 10

Figure 7.3. North Area . Plan of entire area; historic phase walls in south are indicated by lighter shading.
82
Surface Scraping and Planning

The plan of these structures is illustrated in Secondly, the density of artefacts recovered from
Figure 7.3, showing relatively little in the way of the scraped area does not necessitate the existence of
later disturbance of the Neolithic deposits. One re- an overlying phase from which such artefacts might
cent feature is a pair of arcs of ditch in the north- have originated. It might be argued that some types
western quadrant of the cleared area, which also of artefact, especially the rather fragile handmade
showed up well in the magnetometer survey. Ex- Neolithic pottery, may have eroded and disappeared
amination of a 1960s' photograph of <:::a talhoyuk, from the archaeological record at the same pace as
taken from the northwest, reveals that these ditches their architectural contexts, but this argument could
were almost certainly dug around a tent pitched on not apply to more durable categories such as obsid-
that part of the mound during Mellaart's excava- ian and polished stone. All types of artefact are scarce
tions (Mellaart 1967, pl. 1). Otherwise, post-Neolithic across the North Area, and excavation of spaces 70-
features are restricted to a few scoops and shallow 72 in 1995 supported the interpretation that most
pits of uncertain date. At the southern limit of the objects are contained within collapsed room fill and
scraped area, however, traces of a substantial late within and between walls rather than in primary
period building with associated burials were de- contexts on floors. In sum, there is no evidence for a
tected, as supported by finds of Classical pottery density of artefacts in the North Area which might
and extensive spreads of post-Neolithic rubbish in indicate the existence of a now-deflated overlying
large pits cutting the Neolithic architecture. occupation level.
The clearance and planning of the Neolithic Thirdly, the evidence points to a strong posi-
buildings, when considered in concert with the pre- tive correlation between pottery collected from the
liminary results of the 1995 excavation of one of surface of the North Area, before scraping, and that
these structures (spaces 70-72 in Fig. 7.5), throw new recovered during both scraping and excavation, with
light on the complex issue of erosion and site forma- all lines of evidence indicating a dating to around
tion processes at C::atalhoyuk. In sum, the weight of level V in Mellaart's terms and only small amounts
the evidence suggests that the architectural level as of material from later levels.
planned by us immediately underlying the modern Both scraping and recent excavation in the
surface of the mound may well represent the very North Area have demonstrated that many of the
latest level of occupation in this part of the mound, rooms planned in this area are now very close to
and therefore that the degree of post-abandonment floor level. Plastered floors have been detected on
erosion may be considerably less than anticipated. the mound's surface at several points within the
There are several threads to this line of evidence. complex of buildings, and excavation of spaces 70-72
Firstly, the general lack of
intrusive features cutting
through the planned archi-
tecture suggests that little
activity took place in this
part of the mound after the
abandonment of the visible
Neolithic buildings. It is
known from Mellaart's ex-
cavations that the business
of living, and dying, in the
Neolithic levels at C::atal-
hoyuk involved a consid-
erable amount of digging
beneath occupied houses,
particularly in order to
bury the dead. Yet in the
North Area there is so far
no solid evidence of pits or
more especially graves cut
from a now-eroded occu-
pation level overlying the Figure 7.4. Examples of d.etail observed after scraping on the northern eminence of
level planned by us. c;atalhoyiik East.
83
Roger Ma tthews

42

0 lOrn

1020,1160

Hellenistic/
Byzantine walls N

Figure 7.5. Nor th Area . Genera l plan showing walls, recons tructed wa lls, burnt areas and space numbers.
84
Surface Scraping and Planning

in 1995 established that floors here were no more large complex of buildings. Continuity of walls and
than 0.5 m below the modern surface. Not surpris- deposits across most of the scraped area suggests a
ingly, much or all of the surviving room fill consists strong degree of contemporaneity of these build-
of mud-brick rubble and debris originating from the ings, with the possible exception of spaces 19, 42 and
upper destroyed parts of the building. If buildings 44-50, which may belong to an earlier level exposed
originally stood to a height of 2 m or more, based on by erosion of the western slopes of this part of the
evidence from Mellaart's excavations, and if at their mound.
abandonment the roof and upper parts of the walls In general terms, and in many of the details,
were tipped into the empty building, as suggested these structures bear close comparison to the
by the room fill in spaces 70-72, then the Neolithic Neolithic levels excavated on the southwestern slopes
buildings of the North Area may have had an imme- of C::a talhOyiik in the 1960s. All walls are constructed
diately post-abandonment standing height of around of unbaked mud brick, usually of large dimensions
1 m. These tentative interpretations suggest that, at (up to and over 1m in length), and there is an aver-
least in the North Area of the east mound, no more sion to the use of party walls between buildings.
than 0.5-1.5 m of archaeological deposits have been This frequent use of double, occasionally triple, walls
eroded from the surface of the mound since its aban- is also matched in the Aceramic Neolithic and
donment in the Neolithic period. The variability in Chalcolithic architecture of Can Hasan III and I
slope and microclimates across the mound may mean (French 1962). There are almost no true right angles
differing rates of erosion at other points. and the feeling is of an organic, cellular agglomera-
A discussion of the extent of erosion has rel- tion of buildings over time rather than a unified
evance also for the consideration of the movement of planned layout. Frequent use is made of white wall
settlement around the mound at <::atalhoyiik. Both plaster, generally with multiple applications visible
the pottery and the lithic evidence from the recently in plan. Lavishly plastered features or niches also
excavated building in spaces 70-72 in the North Area occur, and there is every indication that many of
point to a date comparable to Mellaart' s level V or these rooms have qualities typical of the so-called
VI. If no upper levels ever existed over the North 'shrines' of the 1960s levels, including a grand pair
Area as revealed by surface scraping, the indication of in situ cattle horns in space 89.
is of a substantial shrinkage of settlement through Buildings appear to consist of one large room,
the course of the Neolithic period on the east mound, often approximately square in plan, with or without
with occupation dwindling to a small outpost on the additional smaller rooms, but the lack of evident
eastern skirts of the mound before complete aban- doorways between rooms renders difficult a precise
donment at some time late in the Neolithic period. division of all rooms into discrete house units. Brick
The spatial organization of the Neolithic build- and plaster platforms are detectable in some of the
ings in the North Area is most readily comprehended large rooms, such as spaces 53, 63, 66, 71 and 82.
in Figure 7.5, where walls have been depicted in Room sizes also compare well to the 1960s' build-
solid black with additional walls, indicated in out- ings. A provisional analysis of the sizes of large rooms
line only, which have been provisionally recon- in the North Area and from successive levels of the
structed in order to make sense of walls which, for a 1960s' excavations shows an overall compatibility:
variety of post-construction reasons, are only par-
Level Large room average area Sample size
tially visible. Areas of burning are indicated by stip-
1990s' North Area 22.00 m' 17
pling and space numbers are designated. The 1960s' III 17.93m2 3
complex of Neolithic buildings planned in this part 1960s' IV 24.26 m' 9
1960s' V 21.16 m' 6
of the mound is of great significance and is now 1960s' VI A 19.22m' 26
1960s' VI B 19.31 m' 30
considered in more detail. 1960s' VII 20.28 m' 17
The general alignment of these buildings is of 1960s' VIII 24.40 m' 8
1960s' III-VIII 20.94 m' 99
note, with a gentle but distinctive radial sweep. In
the central strip of the cleared area walls are aligned Discrete areas of burning were detected across
approximately north-south and east-west, while to the scraped area, as indicated on Figure 7.5. In some
either side there is a marked swing in the major cases entire or almost entire rooms, such as spaces
alignments. The cause of this radial effect is unknown, 38, 63, 82 and 89, are filled with heavily burnt mate-
but may relate to the presence of streets or passage- rial, while elsewhere more restricted patches of burn-
ways into the centre of the settlement. One such ing occur, sometimes in association with ovens or
street passes through the scraped area (spaces 1, 20, hearths, as in spaces 17,32 and 95. In some instances
43 & 58-60), forming a clear western boundary to a it may be feasible to associate burning with the
85
Roger Matthews

deposition of rubbish in open areas, such as spaces important implications for any attempt to under-
30, 32-3 and 60, but in other cases there is the possi- stand the overall structure and layout of the site.
bility of destruction of particular rooms, perhaps Although these issues require much future work, a
deliberately, prior to rebuilding or abandonment. tentative first impression is that at any one time the
There is, in any case, no convincing evidence here settlement at ~atalhoyiik did not consist solely of
for any extensive destruction by fire across the whole densely packed agglomerated buildings with every
building complex, but rather the impression of strictly available part of the mound occupied, but that large
controlled use of fire within coherent and discrete open spaces existed between built-up areas. Parts of
architectural contexts. A comparison of burnt areas these open areas may have been encroached upon in
in the scraped area with the results of the magnet- times of high population density, but generally they
ometer survey shows a high degree of correlation. were used for rubbish disposal, the rubbish always
Artefacts from these buildings were very sparse, burnt in order to reduce smell and keep away dogs.
agreeing with the interpretation that the complex Perhaps most significantly, the presence of ex-
did not suffer from a mass destruction, by fire or tensive open areas has important bearings on future
otherwise, but was routinely and carefully cleaned demographic study of the site, pertaining to issues
prior to rebuilding. Pottery, bone and lithics were of population size and density. Large-scale burnt
not found in great quantities. There was a steady rubbish deposits of probable Neolithic date occur in
representation of burnt clay balls, usually in frag- areas 1040,1040; 1045,1125; and parts of 980,1080.
ments (for the interpretation of these see Chapter Current attempts to address demographic issues at
12). A total of ten figurines, or fragments thereof, ~atalhoyiik are likely to be greatly assisted by future
was found in the scraped area, including humans, work at the site, but some initial ideas can here be
animals and birds. The relative dearth of artefacts explored, despite the evidently simplistic nature of
from this dense area of buildings can be contrasted the process at this stage. Let us firstly assume that
with the very high frequencies of Neolithic artefacts during a given time of the Neolithic period the east
found in other areas of the site where architecture is mound was in its entirety as densely and uniformly
not visible but where rich burnt rubbish deposits occupied by buildings as the North Area. At a mini-
occur, such as squares 1045,1125 or 1040,1040. Al- mum of 30 buildings in 1600 m 2 (the approximate
though strict contemporaneity between houses and density in the northern scraped area), each hectare
rubbish remains to be established, the impression is would contain 187 buildings, giving a total of at least
that the Neolithic residents of ~atalhoyiik carefully 2530 buildings for the 13.5 ha mound. Leaving aside
and regularly cleaned their buildings, depositing their for now all question of variety in building function,
rubbish in discrete areas of the mound. we may further assume that each structure repre-
The existence of this practice has several impli- sents the living space of, as a purely speculative
cations. Firstly, it is clear that excavation of these figure, 4 persons, giving a population total of over
clean buildings will have to contend with a signifi- 10,000. The presence of large open areas, however,
cant lack of in situ artefacts upon which to base strongly suggests that at any given time the mound
interpretations of room function and social and eco- was not universally densely occupied. If we assume
nomic issues, thus underlining the importance of a that 33 per cent of the settlement was given over to
rigorous sampling and micromorphological pro- open areas, then we may estimate a total of 1687
gramme for the recovery of minute traces of activi- buildings housing 6748 people. 50 per cent of open
ties within rooms. Other areas of the site, however, area gives 1265 buildings and 5060 persons. In addi-
may yield extensively destroyed Neolithic levels, tion, not all the buildings need have been occupied
such as Mellaart's level VI A (Mellaart 1964, 40), in at the same time, and some buildings may not have
which many in situ artefacts may occur. Secondly, it been houses. Settlement density is also likely to have
is hoped that careful excavation of Neolithic rubbish been considerably reduced during the later Neolithic
dumps, again involving micromorphology, will pro- when surface evidence suggests that only the east-
vide detailed and extensive information concerning ern part of the east mound was occupied.
the range and types of activities conducted within No evidence has yet been detected for large-
built-up areas, and give some diachronic understand- scale public facilities at ~atalhoyiik, such as temples,
ing, even at the seasonal level, of the rhythms of palaces or communal storage areas (see, however,
domestic and agricultural life at the site. 'central area' below). The evidence hitherto suggests
Thirdly, the presence of large open areas of that at the gross scale the Neolithic settlement
burnt rubbish within the Neolithic settlement has comprised only two elements: houses and open

86
Surface Scraping and Planning

spaces, these latter generally filled with extensive of 0.15 m of earth. All walls were built of the same
spreads of rubbish. This simple settlement constitu- fine-textured pale orange bricks, with a maximum
tion, with scope for a restricted range of social func- measurement of over 0.6 m, but with little detectable
tions, supports an interpretation of the site as an wall plaster. At least two stretches take the form of
elaborate village rather than an urban or proto-ur- double walls. The alignment and construction of all
ban settlement. Future work at the site will certainly walls in this area strongly suggest a Neolithic date.
yield data to facilitate a more substantive approach Large spreads of burnt debris, probably largely
to these important issues. Neolithic in date, have obscured much of the surface
architecture. An unusual feature is the series of six
Central Area parallel walls spanning squares 980,1080 and
980,1090. These are all Neolithic in date, as attested
A 20 x 20 m area was cleared and planned near the by their brick sizes and their attachment to the series
centre of the east mound, at the point where the main of Neolithic rooms to the east. The finding of a fine
southern eminence dips down to the north (Fig. 7.6). clay stamp seal fragment in square 990,1080 further
The southwest corner of this area was at point supports a Neolithic date for these structures. These
980,1080. A largely coherent plan of a building, on a unusual walls are difficult to interpret but they
north- south alignment, was revealed after the removal suggest that we cannot preclude the notion of

~
- 0

I
/
0 4m

0
a
0 •.

980, 1080

Figure 7.6. Central Area (980,1080).

87
Roger Matthews

substantial non-domestic buildings in the central part large bricks, maximum dimension 0.8 m, of probable
of the mound. The extensive spread of burnt rubbish Neolithic date. Other features, including pits and
in the southeastern part of square 990,1080 corre- part of a circular structure are of uncertain date.
lates well with a large area of burning detected in In square 940,1040 (Fig. 7.11),located only 20m:
magnetometer survey. to the north of the northern edge of the 1960s' trenches
The architecture of the Central Area is much on the western slopes of the mound, considerable
less susceptible to description and interpretation than amounts of earth had to be removed before the ap-
that of the North Area. There is little use of wall pearance of any features. Even after the removal of
plaster and the walls are more poorly preserved. some 0.45 m of homogeneous earth across the square
Without excavation in the Central Area it is not pos- only extremely fragmentary remains could be de-
sible unequivocally to state that these buildings rep- tected, including one short stretch of wall, constructed
resent a standard or mode of Neolithic living differing of large silty bricks, running east-west along the
in some substantive way from that betokened by the southern edge of the trench. The size of these bricks,
buildings exposed in the North Area and those exca- at least 0.6 x 0.25 m, suggests a Neolithic date. Apart
vated by Mellaart in the 1960s. from faint bricky patches elsewhere in the square,
the bulk of the deposits here were clearly affected by
Sample areas on the east mound severe Classical disturbance, as indicated by large
quantities of Hellenistic and later pottery.
In addition to the North and Central Areas discussed Square 990,1040 (Fig. 7.12), just to the north of
above, a total of ten sample squares, each 10 x 10m the summit of the east mound, yielded some coher-
in dimensions, were scraped and planned with vary- ent features after the removal of 0.15 m of surface
ing results. The following discussion treats each earth. A further 0.05 m was removed, as an experi-
square in turn, starting at the north end of the mound ment, in the northwest quadrant of the square. Sev-
and giving the co-ordinates of the southwest corner eral pits, at least two of which are brick-lined, and
of each square. most of which may be graves, dot the square. The
Features and walls in 1040,1240 (Fig. 7.7), lo- series of walls in the southeastern corner form at
cated towards the northern limits of the east mound, least three sides of a heavily burnt room, with singed
did not show clearly even after the removal of 0.4 m wall plaster and burnt room fill. Parts of in situ pots,
of earth. Deposits were partly covered in wash mate- all wheelmade, and the size of the bricks, here and in
rial from higher up the mound and partly disturbed all walls in this square, with a maximum dimension
by modern trenching. The dating of the few visible of less than 0.45 m, indicate that all visible architec-
remains is uncertain, but post-Neolithic material was ture in this square belongs to the Classical periods,
well in evidence. although their precise dating is uncertain.
In square 1040,1140 (Fig. 7.8), close to the north- In square 1040,1040 (Figs. 7.13-7.14), features
ern eminence of the east mound, some walls and were firstly planned after removal of some 0.25-0.3
features were visible. The clear wall in the north- m of earth (Fig. 7.13), and a second plan was made
western corner, of bricks 0.4x 0.25 m, is of probable after the removal of a further 0.1 m of earth (Fig.
Classical date, but other walls, all unclear in detail, 7.14). Apart from an extensive series of pits and
may be earlier. Just to the southeast, magnetometer scattered heavily burnt deposits, rich in animal bone,
survey had revealed a series of rectilinear alignments the main features are two east-west walls constructed
suggestive of a large structure immediately under of large, fine textured, pale orange bricks, some over
the mound's surface. Clearance of square 1045,1125 1m in length, of Neolithic type. A semicircle of laid
showed that the entire square was covered by rub- bricks was also found in the southwestern corner.
bish deposits (Fig. 7.9), all Neolithic and yielding Little coherent architecture was revealed in
much in the way of pottery, animal bone and obsid- square 1090,1040 (Fig. 7.15), immediately to the west
ian. It seems that the magnetometer had detected of the eastern eminence of the east mound. Most
strips of rubbish burnt to varying degrees, rather features took the form of pits, some of which are
than mud-brick walls. likely to have been burials. One definite burial, in
After the removal of 0.2 m of earth from square the northwestern quadrant of the square was ex-
1040,1090 (Fig. 7.10), no clear features were revealed, posed and fully excavated (Fig. 7.16). The burial was
so that a further 0.1 m was removed from the capped by two terracotta tiles, each with diagonal
southern half of the square. The wall running to the finger-marks, which covered the burial of a small
northwest from the southeastern corner is built of infant without grave goods. The body was aligned

88
Surface Scraping and Planning

/
/
/

/ /
/ /
/
/

/ /
/
/
/
/ '\
/

/
/
") \ I

d
CJ
I /

/ N

~
/
I

0 2m

1040,1240

Figure 7.7. 1040,1240 plan.

~
0 2m

1040,1140

Figure 7.8. 1040,1140 plan.


89
Roger Matthews

L_ N

\
~
_ __,-~2m
0

"J~
N

CJ CJ
Oo
Q
~
D
~
c,
' '
'
',,

0 2m

1040,1090

Figure 7.10. 1040,1090 plan.


90
Surface Scraping and Planning

I
I
I
I
I I
L _ _ _ _I
,-~ - -- _, ----,
:
' '
:
·-------·-·--"

'
--- .1..

---
~
0 2m

940,1040

Figure 7.11. 940,1040 plan.

8
I/

0 2m

990,1040

Figure 7.12. 990,1040 plan.


91
Roger Matthews

0 2m

1040,1040

Figure 7.13. 1040,1040 upper plan.

~
(J 0 2m

B
CJ
D
1040,1040

Figure 7.14. 1040,1040 lower plan.


92
Surface Scraping and Pl annmg
.

2m

Figure 7.15. 1090 ' 1040 pan.


I

N
a b

0 20crn

Figure 7 ·16 · 109 0,1040 excava ted buna


· 1 plan.

93
Roger Matthews
.

,--,~~> ~
, __
--·
~- --

(
/

\_-- I
\
~
D
990,990

Figure 7.17. 990 ' 990 plan.

~----t /

~
480,1040
I
Figure 7.19. 480 ' 1040 plan.
94
Surface Scraping and Planning

1100

1000

l
900

400 500 600 700


Figure 7.18. Plan of west mound showing location of scraped areas.

east-west, with the head to the east. The nature of bricks. The small size of many of the pits suggests a
th e tiles suggests a Byzantine date for this burial. high incidence of child burial here. In square 1000,950
Two squares near the summit and southern no architectural features at all were detected after
slopes of the main eminence of the east mound the removal of up to 0.2 m of surface soil. Some
yielded little in the way of prehistoric results. In alignments of stones and faint traces of pits may
square 990,990, immediately to the southwest of the indicate the presence of Byzantine graves. Pottery
mound's summit, again no architecture was detected. recovered from this square supports our previous
A series of non-overlapping pits, almost certainly interpretation, based on collection of surface arte-
Byzantine graves with an east-west alignment, was facts, that the entire southern reaches of the east
cleared and planned in the eastern half of the square mound are extensively overlain and disturbed by
(Fig. 7.17). Some of these pits were lined with mud Classical and Byzantine deposits.

95
- -
---- Roger Ma tthews

0 0 N

()
)
580,1020

.
Ftgure 7 ·20. 580,1020 plan.

0
640,960

Figure 7.21. 640 ' 960 pian .


96
Surface Scraping and Planning

()

1300

1250

1200

1150

1100

1050

1000

950 N

900

850

850 900 950 1000 1050 1100 1150 1200


contours at lm intervals

Figure 7.22. Location of transects of scraped squares used in Figures 7.23 to 7.25.

97
Roger Matthews

Soil depths at 1000mE

1023 25

20
1018 6
~
;; 15
~

E
1013
10
=""
"'
1008 ·s"'
"'
1003
0 0 ,..,
0 0 0 0 0 ,..,
0
,..,,..,
0
"'
00 "
00
"' "'"' "'"' "',..,
0 ",..,
0 "',..,,..,
northing

Figure 7.23. Soil depths in scraped areas at 1000 m E.

Approximate soil depths at 1040mE

1018 50
45
40
1013 35 ]
30 ..c
25 Q.
20 .,"'
1008 15 ·c;
10
5
"'
1003 0
... ... "'..... ........ "'..... "'..... "'.......... ............. "'.......... "'.......... ...
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

"' "'"' "' "'"' "'"' "' "'"' "' "'"' "'"' "'..... "'"'..... "'..... "'"'..... "'"'..... "'.....
0 0 0 0 0 0
0
..... 0 0 0 0 .....
.....
northing

Figure 7.24. Soil depths at 1040 m E.

Approximate soil depths at 1040mN

40
1023 35

1018
30 a
~

.6
"iii
1013
25
20
.;::
p.
01
15 '1::1

1008 10 ·s
5 "'
1003 0
0 0
0
0
....
0 0 0 0
0
0
....
0 0 0 0
0
0
.........
0 0 0 0
0
"""" 0\
"'
0\ 0\
"'
0\
""
0\ 0
..... "'.....
0 0
..... "'.....
0 "".....
0 .....
..... "'.......... ..... "'.......... "".......... "'.....
easting

Figure 7.25. Soil depths at 1040 m N.

98
Surface Scraping and Planning

The west mound on the east mound. The graves are dug into a homo-
geneous soil with few or no clear features, and the
Brief investigations of the west mound in the 1960s interpretation must be that by far the greater part of
(Mellaart 1965) demonstrated the almost exclusively the disturbance to the upper levels of the west mound
Chalcolithic date of the mound, and also suggested was caused by agricultural activities in the centuries
that prehistoric architecture might be difficult to de- preceding the digging of the graves, perhaps during
tect here. Our three 10 x 10 m squares (Fig. 7.18) Hellenistic or Roman times. Clearance of two other
indeed confirmed that subsurface architecture is ex- squares on the west mound, 580,1020 and 640,960,
tremely difficult to detect on the west mound, the also revealed no architecture, but several linear
main reason almost certainly being the extent and plough marks and series of pits, some of which are
intensity of ploughing and agricultural use of the probably graves (Figs. 7.20-7.21).
mound in the millennia since its abandonment. The
west mound is lower and more level than the east References
mound and therefore much more suited to agricul-
tural exploitation. Even today much of the mound's French, D.H., 1962. Excavations at Can Hasan, first pre-
surface, while not actually ploughed, is disturbed by liminary report, 1961. Anatolian Studies 12, 27-40.
activities involving the production and storage of Matthews, R.J. & J.N. Postgate, 1987. Excavations at Abu
chaff. Salabikh, 1985-86. Iraq 49,91-119.
In the westernmost sample square on the west Mellaart, J., 1964. Excavations at C::atal Hiiyiik, third pre-
mound, 480,1040, the clearest features were brick- liminary report, 1963. Anatolian Studies 14,39-119.
lined graves with alignments not far from east-west Mellaart, J., 1965. C::atal Hiiyiik West. Anatolian Studies 15,
135-56.
(Fig. 7.19). These graves, like those on the east mound,
Mellaart, J., 1967. (:a tal Hiiyiik: a Neolithic Town in Anatolia.
are almost certainly Byzantine in date. Other fea- London: Thames & Hudson.
tures include plough marks which are certainly later Postgate, J.N., 1983. The West Mound Surface Clearance.
in date than the brick-lined graves which they cut. (Abu Salabikh Excavations 1.) London: British School
Such plough marks have never been detected by us of Archaeology in Iraq.

99
ChapterS

The Magnetometric Survey at ~atalhoyiik East

Colin A. Shell

The pioneering excavations of James Mellaart (1967) human occupation. The plain alluvium is of varying
revealed a continuous sequence of succeeding struc- iron content and consequent colour: nearby to
tures in the west side of the east mound at C::atal- C::atalhoyiik the soil is yellow-buff, indicating crudely
hoyiik, investigating an area where originally, some presence of iron and possibly higher organic
through soil erosion, the tops of mud-brick walls content, whilst several hundred metres to the north
and evidence for burning had been visible at the it is a very white, low in iron, sediment, strikingly
surface. The nearness of features to the surface sug- visible from the mound in the spoil heaps of the
gested that not only could the top level of archaeol- large irrigation canals currently under construction.
ogy be investigated by shallow scraping of the topsoil Often the mud brick was not manufactured from
in other areas of the mound, but that structures might raw alluvium, but in preference partly made from
also be detected by geophysical survey. Whereas the occupation sediments (cf. courtyard material), possi-
investigation of the surface levels is possible by scrap- bly for its improved properties and not merely its
ing in small areas (10m square) systematically across more convenient proximity. In these circumstances
the mound, it was clear that if geophysical tech- some mud-brick walls may have a significantly
niques could detect the signal from the differing fills higher iron as well as organic content than some of
of the rooms, and ideally resolve mud-brick walls the cleaner fills of the rooms. Alternatively, the rooms
themselves, it might be possible to undertake a sur- fills may include burnt features, such as ovens and
vey that would assist in understanding the nature of fire settings or burnt fills from a destruction phase,
buildings over a large area of the mound. In plan- that will be more magnetically enhanced than the
ning the progamme for the new study at C::atalhoyiik, walls themselves. In either case, in optimal condi-
it would be a distinct advantage to know whether or tions a sensitive magnetic survey could be expected
not there are changing patterns in the type of build- to define divisions between rooms and buildings,
ing in different parts of the mound, that might indi- even if the mud brick itself may not be differenti-
cate differing use of the area or a later phase of ated. In contrast, ground resistivity measurement,
construction. The opportunity arose to integrate a the other common geophysical technique, was ex-
geophysical investigation into the initial reconnais- pected to have difficulty in defining structures, given
sance phase of the new project and for geophysics to that the matrix material, the altered alluvium, is very
possibly contribute to the overall project design, a uniform in structure. Resistively, the walls and fills
still relatively rare occurrence (Shell 1996). It was are likely to be very similar, though the variation in
particularly advantageous that the scrape survey also organic content might affect the moisture retention
commenced in the first season in 1993, providing an properties. On this basis, it was decided that resis-
immediate and direct comparison for the geophysi- tivity surveying would be of limited application on
cal results and aid in their interpretation. the site, especially as the survey would be under-
taken during the excavation season in late summer
Technique and method when the surface of the mound is very dry and con-
tact resistance would be particularly high.
The material of the mounds at C::atalhoyiik is almost Following the initial indications in 1993 that a
entirely derived from the alluvium of the immediate Geoscan Research FM36 fluxgate vertical gradiometer
neighbourhood, though modified by the activities of could detect consistent variation in signal on the

101
Colin A. Shell

850 900 950 1000 1050 1100 1150 1200

Figure 8.1. a) Plan of the c;atalhoyiik east mound showing in outline the area geomagnetically surveyed in the 1993-
95 seasons and its relationship to the scrape areas; b) view of the east mound from the southwest showing the topo-
graph ic position of the magnetic su rvey area.

102
Magnetometric Survey at <::atalhoyiik East

103
Colin A. Shell

site, a sytematic survey of the north side of the main and processed in the evening on a portable micro-
mound was commenced extending northwards over computer, and paper output produced to monitor
the subsidiary summit (Fig. 8.1). The fluxgate progress. The initial processing was by PC-based
gradiometer is a common instrument (Clark 1990, software of Jan Hartmann (1988). The data were later
69-82), which has the advantage of both data log- analyzed and displayed using other data visualiza-
ging and an automatic trigger. The trigger allows the tion and image-processing software. With the low
continuous recording of readings whilst the opera- signal range often encountered in a square, typically
tor is walking along the grid lines. It is necessary to no more than 10nT, when there was no significant
ensure that an audible signal from the trigger is contribution from burning of the sediments, the in-
heard to coincide with the metre positions on each strument noise and particularly the longer term drift
line. Given the expected width of the mud-brick walls created problems in matching the output from adja-
on the site, the reading interval was 0.25 m along the cent 20 m squares, even when employing an edge
grid lines oriented on magnetic north, with an east- matching procedure as proposed by Haigh (1992).
west separation between the lines of 0.5 m. For con- This limited the building up of an overall magnetic
venience, the survey was carried out in a sequence picture of the site, but was mitigated by separate
of 20 metre squares. histogram equalization of the data in each square
One of the concerns with the survey was the and median filtering where appropriate to reduce
ability of the instrument to consistently differentiate noise. The filtering has also the tendency to remove
between materials with small differences in mag- the contribution from small, effectively point, fea-
netic properties. Although the instrument records to tures in the survey record, and gives more emphasis
the nearest 0.05nT, the expected accuracy is limited on the deeper, larger features. In consequence there
by the electronic noise in the measuring circuitry, is also a smoothing of the signal, tending to reduce
and the tendency of the zero to drift. The latter is the sharpness of edges. More subtle aspects of the
corrected at intervals by reference to a fixed zeroing data are apparent when viewing the data un-
point. The stability of the FM36 is discussed more equalized, but the principal features of the survey
fully in Appendix 1, where the results of a static test are best revealed by the equalized plots. Figure 8.3 is
show not only high frequency noise and long-term an overview of the magnetic survey at the end of the
drift, but also a short-term fluctuation in the signal 1995 season, by which time 2.3 hectares had been
(Fig. 8.2). The precision of the readings is not better surveyed, involving some 90 kilometres of walking
than 0.2nT. This in combination with the reading with the instrument.
interval defines the magnetic and spatial detection
limits of the survey. Analysis and comparison
The principal contribution to the fluxgate meas-
urement is from the first half metre or so of the Whilst the detail in Figure 8.3 is such that only the
ground. Being designed as a differential device to major features are visible, all aspects of the survey
compensate for the diurnal variation in the Earth's are represented. Figure 8.4 is a summary plan to
field, but with a detector separation of only 0.5 m, a illustrate this, and the features represent also the
contribution to the gradient from a deep feature can three components that contribute to the overall mag-
only be expected if it is an anomaly with a strong netic measurement. Principally, the signal represents
signal. Mellaart's sections showed that the topsoil changing values in the magnetic susceptibility of the
graded to a deeper thickness at the perimeter of the sediments, enhancing the field where the suscepti-
mound, and it was expected that this would limit bility is higher. The large amorphous areas (e.g. in
the value of surveying the margins of the mound, square 1000,1080) have a consistently higher signal
once the topsoil thickness exceeded 0.4 m. As the from what proved to be extensive areas of later ashy
thickness of the uniform topsoil increases there is a tip extending across much of the north slope of the
gradual loss and diffusion of the signal from the main mound. Also apparent in the background are
underlying features. A further problem can be the rectilinear features that almost certainly correspond
presence of highly magnetic metallic iron, but the to the lines of mud-brick walls, either revealed di-
<::atalhoyiik east mound is fortunately very clean in rectly, where the contrast with the adjacent material
this regard. The principal interference in the meas- may show the walls to be of either lower or higher
urements tended to come from the stray nails left in magnetic susceptibilty, or where there are clear rec-
the field by the systematic surface pick-up survey. tilinear areas that correspond to fills confined to
The data from each day's survey were dumped rooms of which the walls themselves are not visible.

104
Magnetometric Survey at <;atalhoyuk East

2 .5

1 .5

0.5

~ ~ ~

0 0 0
C\J
"
reading number "'

10 interval moving average

2.5

1.5
f;:

0 .5

0
~ ~ ~ ~ ~ ~ ~

0 0 0 0 0 0 0
C\J cry 1'-
"
reading number
LO
"'

50 interval moving average

1 .5

f;:

0 .5

0
~

0
~

0
~

0
~

0
~

0
~

0 ;; ~

0
cry 1'- <Xl
C\J
"
reading number
LO
"'

Figure 8.2. Analysis of a static test on site of the FM36 fluxgate gradiometer, showing the noise and longer term
trends in the recording of data . Readings were made at a rate of about 4 per second.

This is discussed further below in the comparison well defined.


between the magnetic data and the detail revealed in Just to the northwest of the oval enclosure is
the scrape areas. A large oval feature is defined in the line of a narrow trench for the modern water
the northwest corner of the survey at (1015,1215). supply, which was also recorded in the topographic
With the dimensions of some 20 x 12m this appears survey. The signal of a low reading to the south of a
to be a later enclosure superimposed on the underly- higher one is characteristic of the signal produced
ing Neolithic. The line of the track running southeast when the fluxgate gradiometer is carried over a hol-
between the main and subsidiary mounds is also low in the ground so that the lower detector is

105
Colin A. Shell

1040 1080
1240

~
1200

1140

' 1100

980 1000

Figure 8.3. Composite plot of the fluxgate gradiometer survey, with the data histogram-equalized to emphasize the
principal features of the survey. The rectangles of uniform contrast are unsurveyed areas of modern tip material. Dot
density plot with high readings dark, and low lighter in colour.
Mag ne tome tric Survey a t <:;a talhoyiik Eas t

1000 1040 1080


1240 1240

I I
II ,_B _ _I
N
~I

~ 1200 1200

A Ashy tip
B Burnt J F- I --
_j I I
S Scrape
U Unsurveyed
/'
,,
It- .'

1140980
r--------1' ~ [ill 1140

'\.\ -=;L=-
-
-I _,
1
I::- , II
940
/' ' :J
B . - --
1120r----------'- - [
J

.····.
··· .. :
1100
~n
ULJ A

.. ..
.... ,.• ····· ····
A
·······
1060

/ B >
:
" ~
A
:
·.:
/' -
1020 I 1020
920 960 0 1050

980 1000

Figure 8.4. Pla n showing the principle features visible in the magnetic su rvey, Figure 8.3.

107
Colin A. Shell

momentarily further from the surface of the soil. In The scrape area (1040,1140) when compared
consequence, its measurement is lower and the mag- with the magnetic image (Fig. 8.5.) shows a direct
netic gradient reduced at the point over the trench, relationship between the major mud-brick walls to
and the reading recovers beyond it. The sensitivity the middle and north of the square and dark, high
of the instrument to step changes in the surface is susceptibilty, linears in the magnetic survey. In con-
also clear where the survey was made over area trast, the mud-brick wall in the southeast corner has
(940,1040) at the beginning of the scraping exercise a lower reading than the adjacent material to the
there. A clear east-west line defines a 15-20 em step north, as predicted might be the case given the na-
in the trench . The overall smooth surface of the ture of the materials from which the mud bricks
mound precluded these step features from occur- were made. Although the magnetic features have
ring elsewhere, except occasionally as a few point necessarily diffuse edges, the intervening spaces are
readings over the entrance to an active fox set. well defined and the magnetic plot suggests the con-
The third and most distinct component of the tinuation of walls that were not revealed in the sur-
magnetic image is the thermo-remanent magnetiza- face scraping. In the centre of the lower left space is a
tion of the sediments where the ancient burning in region of higher magnetic signal where the fill mate-
situ has remained undisturbed and the sediment has rial of the room may have been subject to some
retained the enhanced magnetization that it received burning, possibly by a small hearth at a lower depth
on cooling through the Curie Point of the iron-con- in the space.
taining materials following the firing. Repeatedly Similarly, direct correspondences between the
refired structures such as hearths and ovens can ac- scrape plan and the magnetic survey are found in
cumulate a particularly high remanent field. Thermo-
remanent magnetic features have a characteristic
central high reading where the magnetization rein-
forces the current Earth's field, and a trough in the
reading to the side of the feature that corresponds to
the ancient direction of magnetic north. Where the
installation is repeatedly used, and also when it be- 1155
comes quite large the signal may resemble more a
peak with a surrounding halo of lower readings. At
(1040,1080) are adjacent areas of high signal from
extensive burning which are divided by lines which ,::;
suggest the presence of intervening walls. In a 0
~ 1150
number of places, such as at (1032,1168), are pairs of £;
adjacent high thermo-remanent signals which can "'
E
be equated with the pairs of ovens found by Mellaart
to be set in the walls of certain rooms. The example
mentioned here is just at the south edge of a set of
rectilinear lines that probably correspond to the walls 1145
of the Neolithic houses with which they are associ-
ated.

Comparison with scrape areas


Much of the above interpretation would have been
the subject of some uncertainty but for the opportu- 1042.5 1047.5
nity to directly compare the magnetic results with metres cast
the plans of the 10m square and larger areas subse-
quently investigated by scraping. The ability to make ~-,:,}$r,
a direct and almost immediate comparison with the I I I I I I I I
-5.0 -3.R -2.5 -1.2 0.0 1.2 2.5 3.8 5.0
scrape plans enhanced not only the interpretation of
nT
the magnetic surveys but in some cases also allowed
from the magnetic data an understanding of the im-
mediately surrounding features adjacent to the scrape Figure 8.5. Comparison between the magnetic results
areas. for scrape area (1040,1140) and its plan. Scale 1:200.

108
Magnetometric Survey at C::atalhoyuk East

1080
980 985 990 995
metres cast

1- '
-15
r
-10
I
-5
~,:.
I
0
I
5
I
10
I
15
nT

Figure 8.6. Comparison between the magnetic results for scrape area (980,1080) and its plan. Scale 1:200.

109
Colin A. Shell

scrape area (980,1080). Figure 8.6 shows the outline the direct comparison with the scrape areas shows
plan overlayed on the magnetic survey, with the that the variation in the survey, in the great majority
interpreted walling outlined below. The group of of cases, directly relates to buried archaeological fea-
parallel walls equates to a region of high magnetic tures at <::atalhoyiik . In combination with the scrape
field. The relationships with the walls are not quite plans it provides the means for a fuller interpreta-
as clear as in scrape area (1040,1040), but again de- tion of the use of the space and its differentiation.
fine spaces within which localized burning has oc-
curred. Few features were recorded in scrape area Conclusion
(940,1040) but the lower susceptibilty mud-brick wall
running along the east end of the south edge of the The large survey is capable of further detailed analy-
square is clearly visible in the magnetic survey. The sis, but the point must ultimately be reached where
adjacent scrape area (990,1040) also shows the corre- the interpretation is limited by the survey's spatial
sponding relationships between walls, spaces and and magnetic resolution: the rectilinear relationship
magnetic features. between short segments in the magnetic image, es-
The confirming evidence for the direct relation- pecially where there is smoothing of the data, cannot
ship between the magnetic survey and the features always repesent the corner of a mud-brick room.
immediately below the topsoil is provided in the The survey has shown, however, that there are areas
large scrape area on the northern mound (Fig. 8.7a- of clear trend in the orientation of such segments
c). As in the above examples, the alignment of that they must represent the locations of similarly
walls and the filling of certain spaces with the high aligned buildings. The orientation of these areas
magnetic signal from burnt material are apparent changes between different parts of the mound but is
throughout the scrape area. Two significant features generally following the line of the slope.
should be noticed. Firstly, the path or street running Further geophysical approaches at <::atalhoyi.ik
down the west side of the scraped area appears in may be rewarding in the light of this survey.It would
the plan to be wider at the north than the south, provide a useful test site to study the varying effect
where it becomes restricted between the walls of of the magnetic properties of a uniform sediment on
adjacent houses. The magnetic survey, however, measurements with ground-sounding radar, an ap-
shows that at the northern end there is a line of proach which, if it were successful, would make a
higher magnetic signal that corresponds to the south- considerable contribution to the excavation of the
ern width of the path continuing, with the space to rooms themselves.
the west defined by an area of lower signal, and one
can assume that some form of boundary wall would Acknowledgements
be found at a slightly lower depth. Secondly, to the
west of the path are some slight walls which align The surveys were assisted in the heat of the day by a
directly with features extending into the magnetic number of students and colleagues from Turkey and
survey in the squares to west, especially if the plan is the UK, who probably would prefer to remain anony-
registered slightly further to the south. Continuity is mous. The surveys would not have been possible
also evident to the north and east. The confinement without their support.
of the higher magnetization from burning to a par-
ticular space is well demonstrated in the adjacent Appendixl
spaces 66 and 67 (1032,1185).
Sometimes the burnt areas were recorded in Stability and precision of the FM36 fluxgate gradiometer
the plans if a particularly dark, ashy, deposit was 800 points were read continuously at a fixed posi-
found. This often is the result of incomplete reduced tion on the north mound, at the normal recording
firing with some surviving charred organic material, rate of rate of approximately 4 readings per second,
which has a lower magnetic signal that the fully to monitor both the short-term fluctuation and long-
oxidized firing that produces, in low iron sediments, term drift of the Geoscan Research FM36 fluxgate
a paler yellow to red colour. Consequently, there are gradiometer. The raw data are plotted in Figure 8.2,
areas of high magnetic burnt signal that are not re- and also with narrow and broad moving average
corded in the plan. Although the magnetic survey's smoothing to show the nature of the medium-term
resolution is such that the interpetation of structural variation in the measurement and the longer-term
feature on the mound is using the survey at its limit, drift. The basic descriptive statistics are as follows:

110
Magnetometric Survey at <:;atalhoyiik East

1210

1200

..!:
t:
0
~
(f)
<lJ
1190
""'
'a)
E

1180

1170

1160
1000 1010 1020 1030 1040 1050 1060 1070
metres east

-10 -8 -5 -2 0 2 5 8 10
nT

Figure 8.7. Comparison between the magnetic results for northern scrape area 2 on the northern mound and its plan:
a) magnetic survey with plan overlayed; b) key to the scrape plan; c) magnetic survey. Scale 1:200.

Mean 1.136 Mean 0.607


Median 1.10 Median 0.60
Mode 1.05 Mode 1.00
Standard Deviation 0.427 Standard Deviation 0.285
Minimum 0.10 Minimum 0.10
Maximum 2.40 Maximum 1.40

In the short term, for 80 readings taken over 20 sec- The analysis confirms that, although the instrument
onds, the time for a single line to be walked, the high readings are recorded to the nearest 0.05nT, there is
frequency variation and noise has the following de- at least 0.2nT uncertainty in the value and that signal
scriptive statistics: differences must be greater than 1.0nT for significant

111
Colin A. Shell

42

~
m

0 5 10

112
Magnetometric Survey at C::atalhi:iyiik East

1210

1200

-=...,
0
t::
'Jl 1190
...,
~

Q)
s

1180

1170

1010 1020 1030 1040 1050 1060 1070


metres east

-10 -8 -5 -2 0 2 5 8 10
nT

features to be defined. At C::atalhi:iyiik the signal can Hartmann, J.L.H., 1988. Interpreting earth measurements
vary of the order of 2nT over short distances (1-5m) in image enhancement techniques, in Computer and
in areas where there has not been extensive burning Quantitative Methods in Archaeology 1988, ed. S.P.Q.
of the sediments, and so the FM36 fluxgate Rahtz. (BAR International Series 446.) Oxford: BAR,
223-47.
gradiometer is able to differentiate features mag-
Mellaart, J., 1967. (:a tal HUyilk: a Neolithic Town in Anatolia.
netically, but has been used throughout at close to London: Thames & Hudson.
its measurement limits. Scollar, I., A. Tabbagh, A. Hesse & I. Herzog, 1990. Ar-
chaeological Prospecting and Remote Sensing. Cam-
References bridge: Cambridge University Press.
Shell, C.A., 1996. Appropriate geophysics and excavation
Clark, A.J., 1990. Seeing Beneath the Soil: Prospecting Meth- strategy: from mud brick to masonry in the east
ods in Archaeology. London: Batsford. Mediterranean region, in Archaeological Sciences 1995:
Haigh, J.G.B., 1992. Automatic grid balancing in geophysical Proceedings of a Conference on the Application of Scien-
survey, in Computer and Quantitative Methods in Archae- tific Methods to Archaeology, eds. A.J. Sinclair, E.A.
ologJ; 1991, eds. G. Lock & J. Moffett. (BAR Interna- Slater & J. Gowlett. (Monographs in Archaeology
tional Series 577.) Oxford: Tempus Reparatum, 191-6. 64.) Oxford: Oxbow, 333-42.

113
Chapter9

Surface Pottery at ~atalhoyiik

Jonathan Last

Chronology of the Neolithic pottery handles. In level II more developed forms like disc
bases and tubular lugs appear.
Summary of published information Deeper soundings subsequently revealed ear-
The published reports on James Mellaart's excava- lier pottery, continuing down as far as level XII, the
tions at the Neolithic site of c;atalhoyiik provide lowest reached (Mellaart 1964; 1965). This ware is
little more than a general discussion of the ceramic more primitive in character: less well fired with thick
material found (Mellaart 1962; 1963; 1964; 1966: the black cores and usually vegetable tempered. The ear-
information scattered through these reports has been liest pottery is a heavy buff, cream or light grey
summarized by Stucki (1984)). It may be supposed ware, already burnished and frequently having mot-
that the main interests of the excavators lay else- tled surfaces. Shapes are not dissimilar to the later
where, and indeed Mellaart (1963, 101) states that 'if forms, but deep bowls are more common than
one single category of finds at c;atal Hiiyiik might be holemouth jars.
described as relatively rare and unimportant, then it Few vessels were illustrated by Mellaart; putting
is pottery', although he did recognize the signifi- together these drawings by building level (Adnan
cance of the ceramic material from the lower levels Baysal pers. comm.), one can see how the thick-
of the site, dating to the later seventh millennium BC, walled squat bowls of the early levels give way to
which is still among the earliest known in the Near more globular and closed forms in level VIII, fol-
East (Moore 1995, 40). lowed by the re-emergence of open forms and new
In comparison to Late Neolithic/Early Chalco- features like footed bases (Fig. 9.5:12) in the upper-
lithic sites like Hacllar (Mellaart 1970), and indeed most levels. However, one of the priorities for the
c;atalhOyiik West (Mellaart 1965; and seep. 152 be- renewed investigations at c;atalhoyiik was an as-
low), pottery is not abundant at c;atalhoyiik. A thin, sessment of chronological variation within the sur-
dark-faced pottery appears in small quantities in face assemblages, a task for which the published
level VIII and increases through the upper levels. pottery was inadequate. Further typological and
Forms are simple with holemouth jars (Fig. 9.1: note chronological information, particularly of a quanti-
that illustrated pots are all from 1993-1995 investi- tative nature, was only available by returning to the
gations) and open bowls (Fig. 9.3) predominating. finds excavated by Mellaart.
The pottery is generally not fired to a high tempera-
ture and often has grey cores. Inclusions are mineral; The archive pottery
there is never any organic temper. The prevailing The sherd material is stored in Konya Museum; com-
surface colour is brownish-black and many vessels plete and restorable vessels were taken to Ankara
are finely burnished. In addition some red slip is after excavation, although many were subsequently
found in the upper levels, but painted decoration is returned to Konya. These proved to be of little value
virtually absent, and only a handful of pots have because of 27 Neolithic vessels only 17 were recorded
incised decoration below the rim (Fig. 9.2:8). Ac- by level, and just two of those were assigned to a
cording to Mellaart (1962), light-coloured pottery particular structure (both in level III). The initial
(cream, orange and pinkish) appears in level V and problem with the extant sherd material, on the other
increases in quantity up to level I. The prevailing lug hand, was to assess how representative it is of the
form also changes, from perforated knobs to ledge total excavated assemblage. Mellaart (1966) reports

115
Jonathan Last

that some 300 sherds were found in the 1965 sound- for VIa to IV. Hence relatively narrow, more pointed
ing from levels Vlb to XII. The sherd bags yielded rims are found at the end of the sequence. The high-
312 items from these levels, including a few not from est proportion of flat rims occurs between levels X
that sounding. Therefore it appears that all sherds and Vlb, indicating a clear morphological difference
recovered from level Vlb and below are present in between the earlier and later levels. However, levels
Konya. The upper levels are more problematic: IV to XI and XII have lower values again so changing rim
VIa are represented by numerous body sherds as shape is not merely a reflection of decreasing wall
well as diagnostics so may be more or less entire thickness.
assemblages, minus some complete or restorable ves- Rim diameter is less variable, despite Mellaart's
sels; for levels II and III there are no bodies and far statement that vessel size increases from level V. In
fewer sherds overall, so these clearly represent only fact there seems to be a greater range of sizes in the
a selection from the excavated assemblage. No finds upper levels, including miniature vessels. Many of
at all from levels 0 and I were deposited in Konya the diameter measurements are estimates from rather
Museum. small sherds and should be treated with caution,
The complete vessels provide a little chrono- especially as the repertoire includes oval forms and
logical information on specific forms: for instance, the complete pots show that mouths are often rather
one bowl from level V had a lug shaped as an animal irregular in shape. However, it does appear that- if
(probably cattle) head. A similar feature was found anything - mean rim diameter decreases between
on a sherd, also from level V. It is notable that these levels XII-IX and VI-III, with levels VIII and VII
are the only naturalistic decorations on any of the more variable, and a further decline in level II indi-
Neolithic pottery. Two other such lugs, also on open- cating a greater proportion of small vessels at the
mouthed vessels, were not assigned to a building end of the sequence.
level. One of these probably depicts a ram. Oval and Changes in vessel form may also be significant
oblong pots come from levels II and IV; an unusual in terms of chronological sequence. A simple dis-
cuboid vessel, probably a pot stand, is assigned to tinction, based on rim angle, is that between closed
level IV. There is an early four-footed base from forms (mouth diameter less than maximum vessel
level V, a vessel with a footring from level II and a diameter) and open forms (maximum vessel diam-
deep red burnished bowl with vertically placed lugs eter at mouth). Neutral shapes (i.e. straight-sided
from level VI. vessels) are included in the latter category; most of
the bowls fall into this group since flared profiles are
Rim forms rare. In general terms this classification distinguishes
The lack of contextual information about the com- plain bowls from holemouth jars, although there is
plete vessels, however, compels us to focus on the no necessary correlation between mouth form and
sherd assemblages. Despite the variation in the quan- vessel shape. Bowls and jars are typically defined by
tity of finds through levels XII to II, many general the ratio of diameter to depth. At ~atalhoyiik
trends in the ceramic material can be traced. At- holemouth bowls and open-mouthed jars are also
tributes of the rim sherds show varying degrees of known, particularly from the lower levels. Hence
change over time (Table 9.1a). The first two columns the shorthand used here, distinguishing 'bowls'
chart mean sherd thickness, both of the vessel wall (open) and 'holemouths' (closed), should not be taken
and the rim itself. The former shows that walls are as an indication that vessel forms fall into just two
thickest at the start of the sequence in level XII and categories; rather it reflects an attempt to use the
decrease slightly in size through levels XI and X and information from rim sherds to distinguish broad
again, more sharply, in VIII and VII. This no doubt vessel families which might reflect different sets of
relates to improving ceramic manufacturing and fir- practices, perhaps relating to storage on the one hand
ing technologies. From level Vlb onwards - which and cooking or serving on the other. The lack of
accounts for the majority of the sherds- wall thick- complete profiles precludes a more detailed analysis
ness is constant. Rim thickness follows a similar pat- of vessel form although rim angles, particularly of
tern, with a general decrease through the early levels holemouths, do show some variation.
and stability after level VII, although rims become Just over 50 per cent of vessels in levels XII and
narrower again from level III, which ties in with XI are open forms, and the proportion increases in
other typological changes at this time. The propor- levels X and IX. The early levels have small samples
tion of flat or squared-off rims reflects this too, with and the two lowest are focused on a single room,
a lower figure for levels III and II compared to that so these may not be representative assemblages.

116
Surface Pottery

Subsequently there is a gradual increase in the West (Mellaart 1965). These everted bowls (Fig. 9.4:3)
number of holemouths from level VIII, peaking in increase steadily after level V as well, while slightly
level V at around 70 per cent, before the proportion inturned rims (Fig. 9.3:2) are more stable in their
of open forms increases again up to level II. The representation.
open:closed vessel ratio may therefore be a useful
chronological indicator for the surface assemblages. Base forms
Rim angles of holemouth forms vary from The base sherds also show metrical and morphologi-
nearly vertical to about 45°; some have straight pro- cal variation (Table 9.2). The proportion of bases to
files, others have inturned rims or a slight neck (Fig. rims is variable, but generally high until level VIII,
9.1). A sample of rims from levels VII to II was low in levels VII and VI, and moderate thereafter.
considered in more detail (Table 9.1b), which shows This may reflect the preferential selection of diag-
that the necked forms (Fig. 9.1:5-6) only become nostic rim sherds in the later assemblages, although
common from level III while inturned forms (Fig. this was probably not the case for levels VII and Vlb.
9.1:2) are most common in levels IV and III. This It could therefore represent a real difference in the
variation indicates differences between the assem- relationship of base diameter and rim diameter, and
blages of levels VI-IV, III and II, which are con- hence reflect a change in vessel shape from squarer
firmed by measurements showing an overall increase forms in the early levels to smaller, more rounded
in mean rim angle from level V. bases later. The mean angle between wall and base
The open vessels (Table 9.1c) vary from simple certainly appears steeper in levels XII to IX than
hemibowls to more angular profiles in the later lev- thereafter, and the widest bases are found in levels X
els (Figs. 9.3 & 9.4), which prefigure the Early to VII. However, it is also clear that better defined
Chalcolithic carinated shapes found at C::atalhoyuk bases (with an angular rather than a rounded june-
tion) increase proportionately in the later levels. In
Table 9.1. Neolithic rim characteristics by building level. addition, more pronounced bases (Fig. 9.5:10) and
footed vessels (Fig. 9.5:12) are not found before level
a: Rim morphology
VI and are only common from level III. Small foot-
Level Total Rims Wall Rim Diameter %Open %Flat
sherds thickness thickness (mm) forms rims
rings appear in level III as well. Base thickness fol-
(mm) (mm) lows a similar pattern to rim thickness with the
II 82 61 5.5 4.2 122 83.6 44.8 chunky early forms (levels XII to VIII) giving way to
Ill 48 34 5.3 3.9 149 58.8 46.9
IV 334 125 5.5 4.8 155 44.7 60.0 thinner vessels later, via a transitional phase per-
v 741 207 5.4 4.9 144 30.7 55.1
haps covering levels VII to Vlb (although this is a
VIa 359 110 5.6 5.1 137 35.8 55.0
Vlb 36 28 5.1 4.7 152 32.1 74.1 very small sample).
VII 41 16 7.6 5.3 168 53.3 60.0
VIII 80 29 9.3 6.5 146 62.1 74.1
IX 30 13 12.1 6.9 170 75.0 66.7
X 59 20 11.8 8.0 176 68.4 78.9
Lug forms and decoration
XI 31 10 13.1 7.7 175 55.5 50.0 Lugs are more restricted chronologically but still
XII 35 12 14.6 10.5 197 54.5 58.3
show some variation (Table 9.3). They are found
b: Holemouth forms only very rarely before level VI and no definite lugs
Level Hole- Necked Straight In turned Angle Angle Angle
are known below level IX. The proportion of lugs to
mouths (%) (%) (%) <60 60-75 >75 rims is low until Vlb, peaks in VIa and V and de-
(%) (%) (%)
clines again from level IV. This reflects the changing
11 22 31.8 40.9 36.3 18.2 45.4 36.4
III 12 25.0 8.3 66.7 25.0 66.6 8.3
IV 54 7.4 38.9 53.7 40.8 57.4 1.9
v 43 2.3 48.8 48.8 55.9 41.9 2.3 Table 9.2. Neolithic base characteristics by building level.
VIa 71 12.6 40.8 46.5 10.6 86.3 4.5
Vlb 20 0.0 70.0 30.0 15.0 80.0 5.0 Level Bases: Bases %Angular % Rounded % with Thickness Angle Diameter
Vll 5 20.0 60.0 20.0 20.0 60.0 20.0 rims junction junction feet (mm) (degrees) (mm)
(%)
c: Bowl forms II 32.8 20 44.4 11.1 38.9 6.3 40 67
III 26.5 9 12.5 12.5 75.0 4.4 30 25
Level Bowls lnturned Straight Everted Flaring IV 40.0 50 49.0 24.5 22.5 6.6 37 77
(%) (%) (%) (%) v 30.0 62 72.2 20.4 7.4 7.4 35 81
VIa 15.5 17 25.0 62.5 6.3 4.3 0 0
II 33 18.2 39.4 27.3 15.1
Vlb 7.1 2 50.0 0.0 50.0 7.5 30 65
III 25 20.0 52.0 28.0 0.0 8.0
VII 25.0 4 2.0 2.0 0.0 38 110
IV 16 18.8 68.8 6.3 6.3
VIII 79.3 23 44.4 55.6 0.0 12.2 35 117
v 8 0.0 75.0 0.0 25.0 IX 61.5 8 33.3 66.7 0.0 13.4 47 110
Via 26 34.6 53.8 3.8 7.6 X 70.0 14 38.5 53.9 7.7 12.8 57 115
Vlb 6 33.3 33.3 16.7 16.7 XI 40.0 4 25.0 50.0 0.0 11.0 60 55
VII 8 12.5 87.5 0.0 0.0 XII 58.3 7 40.0 40.0 0.0 10.6 65 71

117
Jonathan Last

Table 9.3. Neolithic lug characteristics by building level. Table 9.4. Neolithic body sherd characteristics by building level.
Level Lugs: Lugs % Width Height Pointed Flaring Straight Level Bodies %Dark %Light Thickness Size
rims Perforated (mm) (mm) (%) (%) ('Yo) surfaces surfaces (mm) (mm)
(%)
II 0
II 0 Ill 0
III 14.7 5 80.0 42.0 14.2 50.0 50.0 0.0 IV 112 55.0 11.0 5.7 61.3
IV 16.0 22 90.0 56.2 22.9 32.0 60.0 8.0 v 403 60.2 19.3 5.3 52.5
v 26.1 54 83.8 55.6 19.9 30.0 40.0 30.0 VIa 105 63.8 11.4 5.1 43.3
VIa 24.5 27 87.5 45.9 17.2 52.6 5.3 42.1 Vlb 2 59.4 3.1 5.0 59.5
Vlb 14.3 4 100.0 29.3 22.8 VII 21 68.4 15.8 6.2 45.1
VII 6.3 1 VIII 28 14.3 82.9 8.6 44.0
VIII 0.0 0 IX 7 19.2 80.8 11.9 48.7
IX 3.3 1 0.0 30.0 25.0 X 18 31.4 62.7 12.3 61.5
X 0 XI 17 8.7 82.6 11.4 50.5
XI 0 XII 16 28.6 45.7 11.7 52.8
XII 0
Note: surface colour percentages for levels Vlb to XII refer to all sherds, not only
bodies.

popularity of holemouth forms on which lugs are


mainly but not exclusively found. The complete ab-
sence of lugs in level II probably reflects the selec- Figure 9.1
tion policy, since at least one of the restored vessels 1. 854/16 Handmade jar rim, slightly pointed; diam.
from this level is lugged. From levels VI to III the c. 260 mm, thickness 5-8 mm; ext. burnished and
proportion of perforated lugs is consistently between striated, dark grey brown; int. burnished dark grey;
80 per cent and 90 per cent, so the change to ledge fabric grey, moderate mineral. Unit 854, area 5;
handles and unperforated forms reported by Mellaart Neolithic.
(1962) must occur in level II. The size of lugs appears
to increase somewhat between levels VIa and IV but 2. 853/1 Handmade jar rim; diam. 110 mm, thickness
there is little consistent variation in other attributes 4-5 mm; ext. and int. burnished dark brown; fabric grey,
like the diameter of the perforation or the distance of moderate mineral. Unit 853, area 5; Neolithic.
the lug below the vessel rim. The prevailing form of
lugs and handles also serves as a chronological indi- 3. 854/21 Handmade jar rim, thickened and slightly
cator. The three main lug shapes (in profile) may be flattened; diam. c. 110 mm, thickness 4-6 mm; ext.
summarized as pointed (Fig. 9.5:4), flaring (Fig. 9.5:5) burnished dark grey brown; int. burnished dark brown;
and straight (Fig. 9.5:2). The pointed form is found fabric brown/grey, moderate med.flarge mineral. Unit
in all levels but the flaring lugs are generally later 854, area 5; Neolithic.
than the straight ones, with an overlap in level V. Of
the rarer forms, vertically placed lugs (i.e. width less 4. 601/4 Handmade jar rim, thickened; diam. 100
than height) and basket handles (Fig. 9.5:6) show no mm, thickness 3-5 mm; ext. burnished dark brown; int.
consistent pattern while double perforations (Fig. burnished dark greyish brown; fabric dark grey, sparse
9.5:4) are not found after level V. small mineral inc. mica. Unit 601, area 7; Neolithic.
The only provenanced animal head lugs come,
as mentioned, from level V; these two examples (and 5. 527/12 Handmade jar rim, necked; diam. c. 140
two with no provenance) represent the only plastic mm, thickness 3-6 mm; ext. burnished red with darker
decoration found on C::atalhoyiik Neolithic pottery. mottles; int. burnished mid brown by rim, darker lower
Incised decoration is also rare: sherds with between down; fabric grey, moderate mineral. Unit 527, area 7;
three and five parallel horizontal incisions below the Neolithic.
rim (Fig. 9.2:8) occur in levels V, IV and III. Part of a
unique small vessel with three incised lines just above 6. 818/7 Handmade jar rim with slight neck; diam.
the base was found in an unstratified context in 220 mm, thickness 6-8 mm; ext. burnished reddish
1995. brown with darker mottles; int. burnished light pinkish
brown; fabric grey, moderate med.flarge mineral inc.
Body sherds mica. Unit 818, area 2; Neolithic.
The body sherds may also furnish some quantitative
information, particularly in terms of colour and wall 7. 504/3 Handmade jar rim, flattened and with slight
thickness (Table 9.4). Colour groups were picked out neck; diam. 260 mm, thickness 5 mm; ext. burnished
visually (without using Munsell charts) and are based light brown; int. burnished reddish brown; fabric grey,
on external surface colour only, so they are to some sparse small mineral. Unit 504, area 7; Neolithic.

118
Surface Pottery

/ 2

4
5

'
6

Figure 9.1. Neolithic holemouth vessel rim forms from c;atalhoyiik East. Scale 1:2.

119
Jonathan Last

extent subjective. Nevertheless, the proportion of Figure 9.2


'dark' (dark grey or dark brownish grey, rarely a 1. 504/2 Handmade jar rim, slightly pointed, with
true black) sherds shows a clear change between remains of lug; diam. 300 mm, thickness 9-11 mm; ext.
levels VIII and VII when they increase greatly in burnished reddish grey to black; int. mottled black to
number. This point also marks a change from pot- light pinkish brown; moderate smallfmed. mineral inc.
tery with at least some organic temper (93 per cent of mica. Unit 504, area 7; Neolithic.
rims in level VIII) to a purely mineral-gritted fabric
(75 per cent of rims in level VII; 96 per cent in Vlb). 2. 854/7 Handmade jar rim with remains of lug;
The lack of bodies from levels III and II prevents a diam. 200 mm, thickness 5-6 mm; ext. burnished grey;
quantitative analysis of later surface colours. int. burnished. Unit 854, area 5; Neolithic.
Mean body thickness more or less mirrors the
measurements from the rim sherds although the 3. 854/13 Handmade jar rim, slightly thickened, with
gradual decline in size from levels XII to IX is not remains of lug; diam. 190 mm, thickness 5-6 mm; ext.
apparent in the former, so seemingly indicative of burnished dark grey brown/dark brown; int. dark brown,
changes in rim form within this early phase (p. 116). mottled; fabric grey, sparse small mineral. Unit 854,
Analysis of sherd size is more problematic since col- area 5; Neolithic.
lection and selection policies have not been made
explicit. It is of more use when comparing assem- 4. 269/1 Handmade jar rim with pierced lug handle;
blages within a level; nevertheless there is an overall diam. 160 mm, thickness 6-8 mm; ext. and int. bur-
pattern, with levels IX-VIa generally having smaller nished light brown greyish brown; fabric grey, moderate
sherds than earlier or later phases. This may be re- mineral inc. mica; lug square, 20 x 18 x 13 mm; perf.
lated to changes in wall thickness and firing tem- diam. 2 mm. Unit 269, area 10; Neolithic.
peratures which would affect the susceptibility of
vessel fragments to further breakage. The very thick 5. 808/1 Handmade jar rim, slightly flattened, with
walls before level IX and the better fired fabrics after pierced lug handle; diam. 210 mm; thickness 5-6 mm;
level VI may have been more resistant to processes ext. burnished red to light reddish brown; int. burnished
such as trampling. One of the principal aims relating pinkish brown to pink; fabric red, sparse small mineral
to pottery recovered from excavation in future sea- inc. mica; lug vertically placed 25 x 45 x 18 mm, perf.
sons will be an analysis of patterns of vessel break- diam. c. 10 mm. Unit 808, area 2; Neolithic.
age and deposition.
6. 67/6 Handmade jar rim, pointed, with 'vestigial'
Surface assemblages lug; diam. 200 mm, thickness 5-7 mm; ext. and int.
The chronology derived from the museum assem- burnished pinkish grey; fabric pinkish grey, moderate
blages may be compared against the material from mineral inc. mica; lug pointed, 40 x 12 x 6 mm. Unit 67,
surface collection and scraping in order to assess area 10; Neolithic.
variation within and between different areas of the
site. The 2 x 2 m surface collection (see Chapter 6) 7. 506/2 Handmade jar rim, flattened, with lug handle;
provided few diagnostic pieces but the small number diam. 120 mm; thickness 3-4 mm; ext. burnished dark
of holemouth and open bowl rims do show different reddish brown; int. grey; fabric grey, sparse small mineral;
patterns, with the former evenly distributed across lug flaring, 32 x 11 x 8 mm, with cavity in upper surface
the main mound and northern eminence but the but not pierced through. Unit 506, area 7; Neolithic.
latter concentrated on the eastern slope of the main
mound and the eastern plateau (Fig. 9.6). Given that 8. 504/1 Handmade jar rim; diam. 80 mm, thickness
bowl forms increase proportionally towards the end 3-4 mm; ext. burnished, mottled black and light reddish
of the sequence, this small sample suggests that oc- brown, with 3 parallel horizontal incised lines below rim,
cupation in the eastern part of the site may well 1-1.5 mm wide; int. burnished light reddish brown;
endure rather later than on the rest of the site. No fabric grey, sparse small mineral inc. mica. Unit 504,
distinct patterning of other attributes of the rim area 7; Neolithic.
sherds, such as thickness or Munsell value (distin-
guishing dark and light shades) was evident. 9. 196/28 Handmade jar rim, everted with internal,
The larger assemblages from the 10 x 10 m pierced ledge for lid; diam. 100 mm, thickness 4 mm; ext.
scraped squares provide more detailed 'windows' burnished light brown; int. pinkish grey; fabric light
within this distribution (Fig. 9.7). The scraped material brown, sparse mineral. Unit 196, area 11; Late Neolithic.

120
Surface Pottery

)I

2
3

4 5

t 7

'
Figure 9.2. Decorated Neolithic rims and lugs from (atalhOyiik East.
121
Jonathan Last

was divided into thirteen spatially distinct assem- Figure 9.3


blages. For the typological analysis the adjacent ar- 1. 601/19 Handmade bowl rim, slightly everted; diam.
eas 3 and 4 were combined, as were 8 and 9, because 150 mm, thickness 4-6 mm; ext. and int. burnished red
of the similarity of the assemblages. Area 13 gener- brown; fabric light reddish brown with grey core,
ally lacks diagnostic pieces. The areas compared here moderate mineral. Unit 601, area 7; Neolithic.
are therefore:
2. 506/1 Handmade bowl rim, slightly inturned;
1. downslope to north of northern eminence (100 diam. 150 mm, thickness 5-6 mm; ext. and int. bur-
sq.m.); nished dark brown; fabric dark grey, sparse smallfmed.
2. upper part of north slope of northern eminence mineral. Unit 506, area 7; Neolithic.
(1600 sq.m.);
3 I 4. top of northern eminence (400 sq.m.); 3. 601/15 Handmade carinated bowl rim; diam. 130
5. southeast slope of northern eminence (100 sq.m.); mm, thickness 5-9 mm; ext. light brown, dark grey
6. top of gully between main mound and north- brown by rim; int. mid brown; fabric grey, sparse small
ern eminence (100 sq.m.); mineral. Unit 601, area 7; Late Neolithic.
7. northern slope of main mound (400 sq.m.);
8/9. western slope and upper part of main mound 4. 710/3 Handmade bowl rim; diam. 200 mm, thick-
(200 sq.m.); ness 7 mm; ext. burnished light brown to dark brown;
10. eastern slope of main mound (100 sq.m.); int. light pinkish brown; fabric light brown with grey
11. eastern plateau (100 sq.m.); core, moderate med. mineral and numerous small voids
12. summit of main mound (100 sq.m); probably from calcareous inclusions. Unit 710, area 2;
13. south slope of main mound (100 sq.m.). Neolithic.

Details of the assemblages from these areas are given 5. 501/1 Handmade bowl rim, rounded and slightly
in Table 9.5. As a whole the pottery adds no new everted; diam. 180 mm, thickness 4-5 mm; ext. bur-
features to those exemplified by Mellaart's pottery, nished dark reddish brown; int. burnished pale pinkish
with the possible exception of a bowl rim with di- brown; fabric probably grey, but encrusted; sparse small
agonal incised lines from unit 12 in area 9 (Fig. 9.4:9). mineral inclusions. Unit 501, area 7; Neolithic.
Most of the pottery is burnished to a greater or lesser
degree: of rim sherds from the 1994 contexts 62 per 6. 504/4 Handmade bowl rim, flattened; diam. 310
cent were burnished inside and out, and a further 12 mm, thickness 8 mm; ext. burnished dark greyish brown;
per cent had a light burnish; 12 per cent were bur- int. dark greyish brown; fabric grey, moderate small/
nished on the exterior only (most holemouth vessels med. mineral inc. mica. Unit 504, area 7; Neolithic.
had at least some bur- .-------------------------------....,
Table 9.5. c;atalhOyak East surface scraping- typological characteristics of the Neolithic rims.
nishing inside the rim)
and 5 per cent on the a: Rim morphology
interior alone; 10 per cent
Area Rims Thickness Rim thickness Rim shape Diameter
lacked evidence of bur- (mm) (mm) %rounded %flat %pointed (mm)
nishing, but these sur- 1 7 6.6 5.4 33.3 50.0 16.7 190
2 68 5.9 5.0 31.1 55.7 13.1 151
faces were also smooth. 3/4 66 6.0 4.9 39.7 41.3 19.0 150
Some of the non-bur- 5 65 6.3 4.6 53.3 36.7 10.0 152
6 58 6.7 5.5 58.9 26.8 14.3 172
nished internal faces 7 155 6.5 5.4 53.7 32.9 13.4 161
8/9 41 7.3 5.7 61.0 14.6 24.4 163
showed evidence for 10 145 6.1 4.9 40.6 44.9 14.5 159
scraping. The burnish- 11 142 6.6 5.5 48.2 29.2 22.6 166
12 15 7.4 6.5 46.7 53.3 0.0 138
ing was sometimes done 13 0
with a narrow tool which
has produced a non-uni- b: Vessel forms
form lustre. For the 1994 Area Holemouths Holemouths Bowls
(%of rims) %necked %straight % intumed Rim angle % intumed %straight %everted
contexts, of those rim
1 71.4 0.0 40.0 60.0 57 0.0 50.0 50.0
sherds where burnish- 2 45.6 13.3 46.7 40.0 58 9.1 66.7 24.3
3/4 67.7 12.2 41.5 46.3 57 23.8 47.6 28.6
ing lines could be seen, 5 60.0 13.9 47.2 38.9 60 17.4 69.6 13.0
70 per cent showed a 6 54.4 64.5 35.5 0.0 60 0.0 30.8 69.3
7 61.4 25.8 49.4 24.7 60 14.0 66.0 20.0
horizontal pattern, 19 8/9 68.3 54.2 41.7 4.2 62 16.7 25.0 58.4
10 69.2 15.1 46.5 38.4 64 0.0 67.5 32.5
per cent vertical and 11 11 49.6 79.4 20.6 0.0 69 0.0 16.2 83.8
per cent diagonal. 12 73.3 27.2 63.6 9.1 57 0.0 33.3 66.6
Surface Pottery

'
. ~-

....
1 2

..·;

4
3

' I 5

Figure 9.3. Neolithic open bowl rim forms from c;atalhoyiik East .

123
Jonathan Last

Surface colour was also assessed on the basis of Figure 9.4


the 1993 assemblages. Within this sample 17 per cent 1. 854/14 Handmade carinated bowl rim, pointed; diam.
of rims are dark grey or dark grey-brown (Munsell c. 140 mm, thickness 3-6 mm; ext. encrusted, probably
N41 to 10YR 312), 12 per cent are dark brown (7.5YR burnished red brown; int. probably mid brown; fabric
312 to 10YR 313), 24 per cent mid-brown (lOYR 413 grey, moderate mineral. Unit 854, area 5; Late Neolithic.
to 7.5YR 5 I 4), 14 per cent light brown (around lOYR
614), 20 per cent red or red-brown (5YR 514 and 2. 812/1 Handmade bowl rim, pointed and slightly
redder) and 14 per cent of mixed or mottled colour. everted; diam. 100 mm, thickness 4-7 mm; ext. bur-
Only about 1 I 3 of vessels therefore fit Mellaart's nished dark brown; int. burnished mid brown to dark
'brownish-black' description. greyish brown; fabric grey, moderate small mineral inc.
mica. Unit 812, area 2; Neolithic.
Rim sherds
Early, organic tempered sherds were not found on 3. 196/7 Handmade bowl rim, everted, pointed; diam.
the surface, implying all this material is the equiva- 180 mm, thickness 4-6 mm; ext. and int. burnished red;
lent of level VI or later, but some typological differ- fabric pinkish grey, sparse mineral. Unit 196, area 11;
ences between the assemblages are immediately Late Neolithic.
apparent. Taking the rim sherds first: a high propor-
tion of holemouth forms is found in areas 1 and 3 I 4 4. 196/10 Handmade carinated bowl rim, everted,
(on the northern eminence) and areas 819, 10 and 12 pointed; diam. 100 mm, thickness 5-7 mm; ext. and int.
(around the summit of the main mound); intermedi- burnished light pinkish brown; fabric reddish brown;
ate numbers occur in the squares around the gully moderate mineral. Unit 196, area 11; Late Neolithic.
between the two main parts of the mound (areas 5, 6
and 7); and low proportions are found in areas 2 and 5. 196/30 Handmade carinated bowl rim, slightly
11. Area 13 produced no rim sherds. Hence on the everted, pointed; diam. 140 mm, thickness 4-8 mm; ext.
main mound (areas 6 to 12), the eastern and north- and int. burnished mid brown; fabric light brown,
ern areas appear, perhaps surprisingly, to have a moderate mineral. Unit 196, area 11; late Neolithic.
later assemblage than the summit. However, the large
number of open forms from area 11 does confirm the 6. 501/2 Handmade bowl; rim slightly pointed, diam.
pattern detected in the 2 x 2m surface collection. On 100 mm, thickness 4-7 mm; base flat, rounded junction,
the northern eminence there is a similar decline in diam. 30 mm, thickness 15 mm; vessel height 41 mm;
the number of holemouths away from the summit, ext. and int. burnished light brown with some charring
with the exception of the small assemblage in area 1. around ext. of base; fabric grey, sparse small mineral inc.
Overall, the contexts on the main mound (areas mica. Unit 501, area 7; Neolithic.
7, 819, 10 & 12) include 65.9 per cent closed vessel
forms; those on the northern eminence (areas 1 to 5) 7. 269/7 Handmade miniature bowl; rim diam. 50
have 58.0 per cent; and the eastern slopes (areas 6 & mm, thickness 5-7 mm; base flat, angular junction,
11) only 51.0 per cent. This may reflect chronological diam. 30 mm, thickness 8 mm; vessel height 28 mm; ext.
differences between the latest occupation of the dif- and int. light pinkish grey; fabric grey, moderate mineral
ferent regions of the site but the mixed nature of the inc. mica. Unit 296, area 10; Neolithic.
assemblages should induce a degree of caution.
In fact other details of the rim sherds serve to 8. 904/12 Handmade bowl rim, pointed, with pierced
modify this pattern somewhat. Holemouth rim an- lug handle; diam. 70 mm, thickness 2-6 mm; ext. very
gles, which gradually increase from level V to II (i.e. lightly burnished medium brown; int. dark brown,
the vessels become more open) show a rather similar smoothed; fabric grey, moderate small mineral; lug
pattern in terms of the difference between the slope vertically placed, rounded rectangular, perf. diam. 4 mm.
and summit assemblages but do not distinguish the Unit 904, area 3; Neolithic.
northern eminence and main mound as a whole.
Mean angle is greatest for area 11, while areas 5 to 10 9. 12/13 Handmade bowl rim, slightly pointed; diam.
on the slopes of the main mound and the south side 100 mm, thickness 6 mm; ext. burnished light greyish
of the northern eminence all have generally steep brown, incised with horizontal line below rim and 4
rims. The northern eminence and main summit as- parallel diagonal lines; int. burnished light greyish
semblages have more inturned forms. The propor- brown; fabric grey, moderate mineral inc. mica. Unit 12,
tion of necked holemouths (Fig. 9.1:5-6), which should area 9; ?Neolithic.

124
Surface Pottery

2
1

4
5

7
6

\ l ~l 9
8

Figure 9.4. Other Neolithic bowl forms from (:atalhoyiik East. Scale 1:2.

125
Jonathan Last

Figure 9.5 Unit 906, area 3; Neolithic.


1. 527/2 Handmade body sherd with lug handle;
thickness 7 mm; ext. burnished dark greyish brown; int. 8. 906/4 Handmade base sherd, flat, angular junction;
burnished dark greyish brown; fabric dark grey, sparse diam. 60 mm, thickness 4-8 mm; ext. smoothed mid/dark
small mineral inc. mica; lug flaring, 50 x c. 30 x 17 mm, brown, lighter under base; int. burnished dark brown;
unperf. Unit 527, area 7; Neolithic. fabric light brown/grey, moderate med.flarge mineral.
Unit 906, area 3; Neolithic.
2. 504/6 Handmade body sherd with pierced lug
handle; thickness 5-7 mm; ext. burnished dark greyish 9. 818/6 Handmade base sherd, slightly concave,
brown; int. burnished reddish brown; fabric grey, sparse angular junction; diam. 60 mm, thickness 4-10 mm; ext.
small mineral; lug straight, 30 x c. 12 x 13 mm; perf. mottled pale brown to dark grey; int. black; fabric grey,
diam. 5-9 mm. Unit 504, area 7; Neolithic. frequent small to large (3-4 mm) mineral inc. mica.
Unit 818, area 2; Neolithic.
3. 51/2 Handmade body sherd with pierced lug
handle; body thickness 6 mm; ext. burnished light 10. 808/2 Handmade base sherd, flat, pronounced
greyish brown; int. light greyish brown; core light junction; diam. 80 mm, thickness c. 8 mm; ext. and int.
greyish brown, sparse mineral inc. mica; lug straight, c. burnished pinkish brown; fabric orange, moderate med./
45 x 27 x 20 mm; perf. diam. 9 mm. Section no. 1; large mineral inc. mica. Unit 808, area 2; Neolithic.
Neolithic.
11. 603/3 Handmade base sherd, flat, pronounced
4. 263/1 Handmade body sherd with double pierced junction; diam. 120 mm, thickness 5-8 mm; ext. bur-
lug handle; body thickness 6-7 mm; ext. and int. bur- nished mottled black to reddish brown; int. light brown;
nished dark greyish brown; fabric dark greyish brown, fabric buff with grey core; moderate med. mineral. Unit
sparse mineral; lug pointed, 45 x 24 x 16 mm; perf. 603, area 12; Neolithic.
diams. 7, 7 mm. Unit 263, area 2; Neolithic.
12. 601/3 Handmade base sherd, footed; diam. c. 90
5. 295/14 Handmade body sherd with pierced lug mm, thickness 7 mm; foot curved, length 53 mm, height
handle; body thickness 5-7 mm; ext. burnished mid 8 mm; width at base 7 mm; ext. burnished light red to
brown; int. dark reddish brown; fabric light pinkish light brown, abraded in places to pinkish grey surface;
brown, moderate mineral; lug flaring, 90 x 20 x 27 mm; int. encrusted, probably light brown; fabric grey, moder-
perf. diam. 14 x 11 mm. Unit 295, area 1; Neolithic. ate small to large mineral. Unit 601, area 7; Neolithic.

6. 1020/5 Handmade jar rim with basket handle; 13. 504/8 Handmade base sherd, footed; diam. 60 mm,
orientation and diam. uncertain, thickness 6mm; ext. thickness 6 mm; foot curved, length 35 mm, height 8
burnished light yellowish brown; int. burnished light mm, width at base 6 mm; ext. and foot burnished red
yellowish-brown with darker mottles; fabric mid brown with black mottles; int. burnished mid brown; fabric
with grey core, moderate small mineral; handle elliptical grey, sparse smallfmed. mineral inc. mica. Unit 504,
section, ext. surface flattened, 40 x 14 mm. Unit 1020, area 7; Neolithic.
unstratified; Neolithic.
14. 124/sf38 Handmade oval bowl; rim slightly
7. 906/3 Handmade base sherd, flat, rounded junc- flattened, diam. c. 80 x 40 mm, thickness 3-4 mm; base
tion; diam. 70 mm, thickness 5-7 mm; ext. burnished flat, angular junction, diam. uncertain, thickness 3 mm;
mid brown, paler under base; int. smoothed, dark grey ext. and int. grey; fabric grey, frequent mixed mineral
brown; fabric light brown/grey, moderate small mineral. inc. mica. Unit 124, area 10; Neolithic.

126
Surface Pottery

. ........... . . ....__,
· <itr; :,~;;(,;,\•:i!iM?{.~ · -·· ~ceo
'· · --~ - -c ...;,:>;J.:~;(~;JJj)~;~C.~

4
5

~LJ
'-...L_/. \__ I

6
9
co 14

-----'------~
~r
11

13

12

Figure 9.5. Neolithic lugs, handles and base forms from c;atalhoyiik East. Scale 1:2.

127
Jonathan Last

increase over time, is lowest on the northern emi- bases represent a similar proportion of the total of
nence, rather higher on the main mound, and great- the northern eminence and main mound assemblages
est in areas 6 and 11 downslope to the east. So in (around 6 per cent).
terms of this index, the northern eminence is earlier The ratio between angular (Fig. 9.5:8) and
in date than the main mound, while the late position rounded (Fig. 9.5:7) base junctions should also have
of the eastern plateau is confirmed. Among the open chronological significance, with a higher figure ex-
vessels, the later (everted) forms (Fig. 9.4:3) again pected for later assemblages. In this case areas 819
predominate in areas 6, 819 and 12 on the main mound, and 12 lack rounded bases altogether, 7 and 11 have
and particularly in area 11 (the eastern plateau). very high figures, and area 3 I 4 has the lowest ratio.
The rim shape indices therefore suggest, a little Rounded bases represent 28 per cent of the northern
ambiguously, two main chronological differences: eminence assemblage but only 9 per cent on the
that the slopes of both the northern eminence and main mound. As with rim shapes, therefore, there is
main mound have later material than the summits; some evidence that the main mound assemblages
and that the northern eminence assemblage as a are more recent than those on the northern emi-
whole is earlier than the main mound, while the nence.
eastern plateau is consistently late. So simple Base thickness and diameter appear to have
open:closed ratios do not tell the whole typological less chronological value after level VII and may, as
story. It is highly probable that this index reflects a with rim sherds, reflect functional differentiation.
degree of functional variation, while the more subtle Here the thickest bases occur in areas 819, 5, 6 and
differences in form and angle, which have no neces- 11; thinner ones in 2, 10, 314, 12 and 7. Smaller
sary bearing on the properties of the containers, are diameters are found in 8 I 9, 3 I 4 and 2; larger ones in
of greater chronological significance. I prefer there- 12, 11 and 7. Hence the patterns are less clear than
fore, to conclude that the northern eminence has the with the rims, but the smallest pots again seem to
earliest assemblage, followed by the main mound, occur on the northern eminence (areas 2 & 3 I 4) and
while the eastern plateau is latest. the largest ones on the eastern plateau (area 11).
Other indices may reflect functional as much as Clear differences are apparent, however, between
chronological differences: mean rim diameter and adjacent areas such as 819 and 12, or 314 and 5. It is
wall thickness provide information on the size of possible that base size and thickness were controlled
vessels and in Mellaart' s collection show little con- less strictly by potters than were rim thickness and
sistent variation over time after level VII. Among the mouth diameter- but variability in all these indices
scraped assemblages, however, there are some dif- is relatively high, and no doubt largely reflects the
ferences. Thicker-walled vessels occur in areas 819 mixed nature of the assemblages.
and 12 on the top of the main mound, while thin
vessels are found on the northern eminence and in Lugs and handles
area 10. Rim diameter is, with the exception of the A further area of difference concerns the lugs and
small assemblages in areas 1 and 12, greater on the handles, both in terms of numbers and types. From
main mound than on the northern eminence. Hence Mellaart's assemblage it seems that the proportion
on both scales, the pots from the northern eminence of lugs to rims decreases from level V, as the number
are somewhat smaller. of holemouth vessels on which they principally oc-
cur declines. Among the scraped square assemblages
Base sherds the proportion of lugs is highest - by some way -
Looking at the base sherds, the most significant vari- in area 5; and higher on the northern eminence as a
ation concerns the proportion of footed (Fig. 9.5:12) whole than on the main mound, consonant with the
and pronounced (Fig. 9.5:11) bases. These appear to suggestion that this pottery is typologically earlier.
increase over time from level V. Here they are most However, the lugs give a less consistent result for
common in areas 314 and 6, moderate in 7, 10 and the eastern plateau. Straight lugs (Fig. 9.5:2), which
11, and scarce in 2, 5, 819 and 12. On the main are rare in Mellaart's assemblage after level V, are
mound therefore, the more recent assemblages are most common in areas 3 I 4 and 11 but absent from 5
again further from the summit - as with the pro- and 6 and scarce in area 10. Flared lugs (Fig. 9.5:5),
portion of holemouth rims- but the eastern plateau which may increase over time, are rarest in 2, 6 and
is less distinct. On the northern eminence there is 11 but more common in areas 5 and 10. This con-
greater variability with, conversely, more feet on the firms the earlier position of the northern eminence but
top of the eminence than downslope. Overall, footed area 11, strangely, also looks like an early assemblage,

128
Surface Pottery

Table 9.6. (atalhoyiik East surface scraping- typological Table 9.8. (atalhoyiik East surface scraping- typological
characteristics of the Neolithic bases. characteristics of the Neolithic body sherds.

Area Bases % % % % Thickness Diameter Area Body sherds %Dark %Light %Red Thi<kness
Straight Rounded Defined Feel (mm) (mm) (mm)

1 0 I 51 56.9 43.1 0.0 6.1


2 23 47.8 47.8 4.3 0.0 7.0 76 2 484 51.0 45.5 3.5 5.8
3/4 28 39.3 42.9 7.1 10.7 7.7 72 3/4 502 47.0 48.4 4.6 5.9
5 36 66.7 27.8 0.0 5.6 8.8 87 5 229 58.5 34.5 7.0 6.7
6 17 35.3 35.3 23.5 5.9 8.6 87 6 462 40.3 53.5 6.3 7.2
7 57 68.4 17.5 0.0 14.0 8.1 99 7 859 51.3 38.2 10.5 6.9
8/9 14 76.9 15.4 7.7 0.0 9.2 69 8/9 281 48.4 37.0 14.6 7.2
10 55 43.6 41.8 10.9 3.6 7.6 88 10 472 61.7 26.7 11.7 6.4
11 86 59.3 22.1 10.5 8.1 8.4 109 11 492 46.5 41.3 12.2 7.1
12 10 90.0 0.0 10.0 0.0 7.8 123 12 86 44.2 44.2 11.6 7.2
13 2 50.0 50.0 0 0.0 6.5 80 13 14 28.6 57.1 14.3 7.9

Table 9.7. c;atalhOyiik East surface scraping- typological characteristics of the Neolithic lugs.

Area Lugs Lugs: Rims % % % % Perforated Width Height Protrusion Below rim
(%) Straight Pointed Flaring Handle (%) (mm) (mm) (mm) (mm)
I 2 28.6 0.0 0.0 100.0 0.0 100.0 90 22 23 0
2 10 14.7 12.5 62.5 0.0 12.5 100.0 44 21 14 20
3/4 21 31.8 30.0 40.0 25.0 0.0 70.0 43 18 15 28
5 16 55.4 0.0 31.3 43.8 18.8 41.7 49 22 21 27
6 11 19.0 0.0 54.5 9.1 18.2 11.1 37 18 10 21
7 30 19.4 20.0 40.0 33.3 6.7 56.0 47 19 14 18
8/9 7 17.1 14.3 14.3 28.6 28.6 0.0 57 20 16 0
10 14 9.7 7.1 50.0 42.9 0.0 76.9 57 20 13 24
11 22 15.5 27.2 9.1 4.5 54.5 45.5 30 22 11 30
12 3 20.0 33.3 33.3 33.3 0.0 0.0 100 5 59 16
13 0

while areas 5 and 10, relatively close to it, appear to perforation (Fig. 9.5:4) and an animal head, both
be the latest. However, the lack of lugs in the level II considered early forms.
assemblage (p. 118) and the small number from level Size indices also show some interesting
III means there is no comparative material for an patterning: area 11 has very small lugs, as does area
assemblage as late as that from area 11 appears (on 6; the other areas on the main mound have generally
the basis of rim forms) to be, and it is therefore larger lugs than the northern eminence- which is
possible that lug morphology changes in a different consistent with the vessel size information. Lug
manner in the uppermost layers. height varies little, indicating that the largest exam-
This may be confirmed by the high proportion ples are also more elongated; protrusion (i.e. maxi-
of unperforated forms on the eastern plateau, which mum distance above vessel surface) is more variable
is probably a late feature. However, area 6 and the and generally proportionate to size. The area 11 lugs
small assemblage from area 8/9 consist almost en- show another difference in that they are generally
tirely of unperforated lugs (Fig. 9.5:1), and area 5 set farther below the rim of the vessel than are those
also has large numbers. None of this reflects the from the other areas.
sequence derived from Mellaart's material and implies Hence the lug data are rather complex but seem
either that these are also late assemblages or that the to show a distinct assemblage on the eastern pla-
west part of the main mound where Mellaart exca- teau, which in some respects looks early but need
vated is not representative of the site as a whole- at not be so. The northern eminence is rather different,
least in terms of certain aspects of the ceramics. In a also appearing relatively early but with some dis-
relatively simple ceramic repertoire, lugs and han- tinctive features. Another distinctive assemblage oc-
dles could represent an important way of signalling curs in area 6 (rather small lugs, very few perforated,
social or functional differences. nearly all pointed) while areas 5 (despite the large
Of the rarer types, basket handles (Fig. 9.5:6) number of lugs) and 10 appear latest in terms of the
are twice as common in area 11 as anywhere else, typology derived from Mellaart's data. All the
reaffirming the difference of the eastern plateau from squares on the eastern side of the site therefore have
all other assemblages. Similarly, the only vertically- late or idiosyncratic assemblages. Chronological trends
placed lugs come from the northern eminence, as seem to be more complex than the quantifications
do the only examples from the surface of a double from Mellaart's area (Tables 9.1-9.4) would imply.

129
Jonathan Last

1300

1200

1100
£....
2
~
s

1000

900

800

900 1000 1100 1200


metres east

grey = bowl rim, black = holemouth rims

Figure 9.6. (atalhdyiik East surface collection- distribution of Neolithic ho/emouth and bowl rims.

130
Surface Pottery

1300

1250

1200

1150

1100

1050

1000

950

900

850

800 L-,_-----,-------,------------~~~~~--,------,------~
850 900 950 1000 1050 1100 1150 1200
contours at 1 metre intervals

Figure 9.7. c;ata/h oyiik East surface scraping -location of different areas.

131
Jonathan Last

Body sherds eminence, in terms of rim, base and lug form, al-
Analysis of trends in the surface colour of body though the extremely high proportion of rim sherds
sherds was handicapped by the lack of evidence for with lugs is indicative of a specialized assemblage.
levels II and III (p. 120). However, there is some Post-depositional disturbance here is probably less
suggestion that the proportion of dark coloured than elsewhere on the site (p. 140), and this material
sherds declines somewhat after level VI. For the may consequently be less mixed. Areas 2 and 3/4
scraped squares the lowest proportion of dark col- are distinct from everywhere else in terms of
ours (by some way) is in area 13, while area 6 has holemouth rim shapes and angles, vessel thickness
relatively low numbers. The highest proportions of and diameter, and ratios of rounded to angular bases,
dark sherds are in areas 1, 5 and 10. A more detailed but lug proportions are rather different. Clearly some
analysis of rim sherds from the 1993 squares (areas typological features are more chronologically con-
4, 6, 7, 8/9, 10 & 11) generally confirms this picture sistent than others. The proportions of footed bases
of more dark vessels on the northern eminence, across the site, for instance, does not fit a purely
lighter colours on the north part of the main mound temporal patterning, at least not that derived from
(areas 6 & 7) and more red vessels on the rest of the Mellaart' s assemblage.
mound and the eastern plateau. Since Mellaart sug- The overall picture suggests that the final occu-
gests red slipped (or burnished) vessels are more pation of the northern eminence occurs earlier than
common in the latest levels this pattern appears to that on the main mound, for which the latest assem-
support a broad chronological distinction between blage derives from the eastern plateau. Despite the
the northern eminence (early) and the eastern pla- ambiguity due to the mixed nature of all these as-
teau (late), with a more mixed (intermediate) assem- semblages and the problems associated with
blage on the main mound. Such broad spatial Mellaart' s archive, this broad chronological pattern
differentiation may also imply local differences in is relatively consistent. Only the proportion of
ceramic production or post-depositional factors such open:closed forms (due to the relatively large number
as the frequency of secondary burning. of bowls in area 2) would put the northern eminence
Body sherd thickness confirms the pattern indi- later, but the relatively early date of the archaeology
cated by the rims with the northern eminence hav- here has to some extent been confirmed by initial
ing thinner-walled vessels (mean 6.0 mm) than the excavation in 1995 within area 2, which produced a
main mound (6.9 mm) or eastern plateau (7.1 mm). small assemblage probably equivalent to level V or
Within the northern eminence the thickest sherds VI (Last n.d.). In key indices, however, such as the
come from area 5, while on the main mound the proportion of holemouths (80 per cent), this also
summit, south and west (areas 8/9, 12 & 13) have appears earlier than the surface assemblage. It is
thicker sherds than the north and east (6, 7 & 10). quite possible that the topsoil material represents in
part the erosion of a later level of occupation in this
Conclusions area of the site, but equally it is inconceivable that
The ceramic material from the scraped squares shows given continuing occupation 100m to the south some
a number of differences across the site, and certain of this material was not deposited on the northern
distinct assemblages can be recognized. The eastern eminence. Further understanding of the site forma-
plateau stands out in particular in terms of vessel tion processes in operation at <;atalhoyiik, the de-
shapes (more sinuous profiles, steep rim angles), a gree to which material has moved downslope or is
high proportion of open forms, and a large number representative of the underlying building levels, can
of basket handles. At least the first two of these be gained only by considering a further set of at-
imply a relatively late date for activity in this area. In tributes of the surface assemblages relating to the
terms of rim form the other late assemblages come density and preservation of potsherds of all periods.
from areas 6 and 8/9, which also share a low propor-
tion of perforated lugs. Areas 7 and 10, lying be- Future work
tween these squares, have slightly smaller vessels The study of the surface pottery and the re-analysis
but in other respects are rather mixed: 7 looks later of Mellaart' s assemblage has also helped to set re-
in terms of bases, 10 in terms of lugs. A rough chrono- search questions and priorities for the coming sea-
logical order of 7-10-8/9-6 makes little spatial sense sons. Clearly the major task is to test the devised
and implies again that the slope assemblages are ceramic chronology, and its lacunae, against
generally very mixed. stratigraphically closed assemblages from contexts
Area 5 looks later than the rest of the northern on different parts of the site. Questions include:

132
Surface Pottery

Which features of the pottery have chronological matrix as objects are trodden into the ground, and
significance across the mound as a whole? Do differ- abrasion is a side effect of all these activities. The
ent parts of the site have distinct features reflecting actions of burrowing animals, including earthworms,
social or functional variation? The recovery of com- and plant roots within the soil can produce horizon-
plete pots or restorable profiles is also essential in tal and vertical movement (up or down) as well as a
order to link the different morphologies (rim, base, degree of abrasion and breakage. Finally there are
lug, body) within a more holistic vessel typology. the major processes of post-occupational human
A further element of the pottery assemblage modification of the site, involving the movement of
which requires more detailed study is fabric. Be- soil and artefacts by excavation (of pits, graves, etc.)
yond the broad change in inclusion type mentioned or ploughing.
above (p. 120) fabrics appear visually very similar, The second set of processes is natural in origin.
in colour and inclusions (usually sparse to moderate These are generally of low energy and primarily
mixed small or medium mineral). Some sherds, how- climatic or topographic. The former include: rain-
ever, clearly contain more mica or calcareous parti- wash, producing horizontal movement downslope
cles, which could relate to different clay sources. and abrasion of exposed material; wind, with simi-
Fabric analysis based on more detailed visual and lar effects; and freeze-thaw which can produce abra-
thin-section studies is essential in order to recognize sion and vertical movement through cracks, etc .. The
distinct fabric groups and to address questions of latter include related processes of soil creep or ero-
technological development and clay provenience. sion which are basically gravity effects and produce
horizontal or vertical movement.
Formation processes Further complexities can be imagined: for in-
stance, dense vegetation moving objects vertically
Method and theory within the topsoil can also impede downslope move-
The above discussion (pp. 120-32) is premised on ment and mitigate against trampling effects; slopes
the assumption that assemblages from the topsoil and prevailing winds will mean artefacts in some
are direct reflections of the underlying archaeology. locations will be more vulnerable to natural proc-
The distinct differences between the pottery from esses. In addition, different artefact types or size
different parts of the site shows that typological vari- categories will be more susceptible to different fac-
ation is indeed recognizable. But the mixed nature of tors. It is the smaller pieces that are more likely to be
many of these assemblages is equally noticeable, even moved by soil animals or roots, but the bulkier ones
when considering only their Neolithic components, (especially pieces such as handles which do not lie
let alone the Classical material with which they are flat) that are more likely to be kicked.
mixed. It is vital to recognize that a set of dynamic Here is not the place for a comprehensive dis-
processes has operated across the site since the end cussion of all the processes and combinations of proc-
of the Neolithic occupation, serving to muddle and esses that can alter the surfaces of mounds like
blur any chronologically or functionally related <;atalhoyiik. The gross morphology of mound al-
patterning originally present. teration has been discussed by Rosen (1986). Here I
These processes are of two main causative types: wish merely to note that all these processes could
a) organic have operated to some extent on different parts of
b) inorganic the site, and that patterning in potsherd densities
and can have four kinds of effect on ceramic material: and size distributions will be to some degree the
1) horizontal movement (that is across the mound, result of these post-depositional effects. On the other
primarily downslope); hand, such patterning may also be diagnostic of some
2) vertical movement (that is within the plough- of these processes and aid our understanding of how
zone or topsoil); the mound surface has developed and altered. Dif-
3) breakage; ferential effects of gravity, prevailing winds and later
4) abrasion. anthropogenic activity may all produce characteris-
The first set involves living organisms, whether hu- tic signatures. This section of the text is an attempt at
mans, animals or plants. They are generally high understanding something of the degree and nature
energy processes. The passage of animals across a of the post-depositional effects which have operated
surface produces primarily horizontal movement at <;atalhOyiik and is offered as a preliminary dis-
(from kicking) and breakage (from trampling), al- cussion of the patterns rather than a comprehensive
though vertical movement is also possible in a soft explanation. It is hoped that comparison with finds

133
Jonathan Last

900 1000 1100 1200


metres east

0 25 50 75 100 125 150 175 200 225 250


number

Figure 9.8. (ata lhoyiik East surface collection - total slrerd numbers.

134
Surface Pottery

from less disturbed underlying contexts, and sec- parts of the mound with relatively fewer sherds on
tions through topsoil deposits in different places on the steeper parts of the slope. Perhaps the basic ef-
the site will in the coming seasons of work enhance fect of gravity means that material moving down-
our understanding and provide more general con- slope is less likely to come to rest here.
clusions that may be of use to projects undertaking It seems, therefore, as if the original distribu-
similar mound surveys in the future. tion of Classical material on the main mound has
been heavily modified by processes of erosion and
Surface collection slopewash. There appears to be a similar pattern
The relatively low number of diagnostic pieces from around the northern eminence as well, although in-
the surface collection means these assemblages are volving smaller quantities of material and with less
of only limited value in the reconstruction of occu- directional movement.
pation sequences, but the controlled method of re- The handmade sherds lack such a clear focus
covery (sieving a measured volume of soil from a 2 x on a single area of the site - unsurprisingly given
2m square) provides invaluable information on sherd that the whole mound is a Neolithic feature, what-
densities and weights which is of use in elucidating ever its subsequent modifications. In other respects,
aspects of site formation processes. however, the pattern is rather similar. On the main
Total sherd numbers (Fig. 9.8) show a clear fo- mound the squares with large numbers of hand-
cus of material around the summit of the main mound made sherds tend to have a lot of wheelmade pot-
and on the southern slopes from the southwest to tery too, and the reduced densities of material on the
the southeast. The northern side of the main mound middle slopes compared with the upper and lower
and the northern eminence yielded relatively few parts is still visible. Presumably, therefore, a certain
sherds while the shallow slopes lower down on the amount of Neolithic pottery has moved downslope
northern and eastern sides of the site have moderate with the later sherds. On the other hand the pre-
densities. But this pattern reflects almost entirely the dominance of finds to the south and southeast is
distribution of the numerically predominant wheel- lacking, while regions of high density in the north
made pottery on the mound surface, masking the and east of the site are not mirrored in the wheelmade
Neolithic assemblages. Only in a band across the distributions. The overall density of Neolithic mate-
northern slopes and northern eminence, and in a rial is neither any greater nor any less in those parts
small area south of the eastern plateau, do Neolithic of the site with significant concentrations of Classi-
sherds comprise about half the assemblage. If the cal material, which seems to imply that, in general,
wheelmade sherds are removed from the plot en- later activity has neither brought to the surface nor
tirely then the picture is rather different and the masked a substantial Neolithic assemblage. An ex-
patterning much less distinct (Fig. 9.9). The Neolithic ception is the southeastern slope of the main mound
pottery is relatively dense on the southern part of where Neolithic densities fall off rather sharply, im-
the northern eminence, the eastern plateau and a plying that earlier artefacts may have been masked
number of squares on the slopes of the main mound, by the fan of Classical material here. This pattern
but is not focused on one side. seems even more pronounced for the animal bone
A possible interpretation would be as follows. (see Chapter 11).
Classical activity concentrated on the summit of the Further local concentrations are visible within
main mound. Erosion over the last 1500 years or so the Neolithic sherd distribution. One lies southwest
has led to the movement of much of this material of the top of the northern eminence at the head of a
downslope but prevailing winds from the north or pronounced gully which runs off to the northwest.
northwest and the associated rain water run-off have The erosion processes which formed this feature may
caused this movement to be primarily in a southerly have brought material to the surface here. Towards
direction. Only in a narrow band approximately be- the base of the gully there is a slightly higher density
tween two and five metres below the summit does in those squares aligned with it than in the adjacent
there appear to be a lower density of material. This ones, indicating a limited amount of downslope
may represent a gap between the exposed but in situ movement as well. The wheelmade pottery shows a
material near the eroded summit and the 'fan' of similar, although less marked, pattern, and the same
accumulated soil on the lower slopes, in effect a can be recognized in the distribution of animal bone.
place where the processes of erosion and accumula- The second concentration of Neolithic sherds is
tion cancel each other out. There is a similar though on the flatter ground of the eastern plateau; again
less distinct pattern around the northern and eastern there is a less distinct but similar pattern in the bone

135
Jonathan Las t

1300

1200

-5 1100
g
~
s
1000

900

800

900 1000 1100 1200


metres east

0 5 10 15 20 25 30 35
number

Figure 9.9. c;atalhoyiik East surface collection - Neo lithic sherd numbers.

136
Surface Pottery

distribution. This would be a region where material without sieving. I divided all contexts into four cat-
moving downslope might be expected to accumu- egories: surface collection (level 0), topsoil removal
late, although it lies away from the prevailing direc- (level 1), second spit or cleaning (level2) and further
tion of movement of the wheelmade pottery. scraping (level3). Not all squares include level2 and
Typological analysis indicated this area has the lat- 3 assemblages since building plans often showed up
est Neolithic material (p. 132), and the distinctive- at a higher level, but some patterning is evident
ness of the assemblage does suggest that it is not (Table 9.9). The percentage of Neolithic sherds is
merely the product of downslope movement of ma- approximately the same in levels 0-2 (i.e. these are
terial from the summit of the mound. However, a all equally disturbed by later activity) but much
third, less pronounced concentration of handmade higher in level 3, which usually starts some 20 em
sherds at the base of the northern eminence may below the surface. Where there is a deep deposit of
indicate that the primary direction of downslope topsoil, therefore, there is less Classical material lower
movement for the Neolithic material has indeed been down. The sherds are also size-sorted by depth, with
northeasterly rather than southerly. Perhaps this re- the mean weight of Neolithic sherds (as with animal
flects changes in prevailing winds over the last 7000 bone) increasing steadily through the four levels.
years. The sherds from level 0 are particularly small in
A further source of information is from sherd comparison, indicative of the long-term effects of
size, which was measured in two ways. Firstly the weathering and trampling as well as the likelihood
proportion of small sherds less than 2 em long (Fig. that animal activity would tend to bring smaller
9.10) is clearly much greater in the northern half of sherds to the surface. That much of this patterning
the site, but this may largely be due to the lower by depth is comparatively recent in origin is shown
numbers of (generally larger) wheelmade sherds by similar figures for the wheelmade sherds, although
here. A band around the base of the steeper slopes in their mean weight declines slightly in the lowest
the north of the site also appears to have larger sherds, level, which relates in part to variability within the
corresponding with a relatively higher proportion of small number of level 3 assemblages. Two contexts
Classical pottery here. Perhaps more significantly in particular (269 & 277) provided most of the level3
there are also generally more small sherds on the wheelmade pottery: in one case mean size is smaller
lower than the upper slopes of the main mound. than for the layer above and in the other it is
This supports the picture of significant downslope greater.
movement: both because smaller (lighter) sherds are This patterning relative to depth confirms that
more likely to move and because weathering proc- post-depositional factors have had at least some
esses will tend to abrade sherds as well as move effect on the distribution and condition of sherd ma-
them. However, a similar pattern is not visible in the terial but perhaps also demonstrates that the sub-
faunal remains. surface assemblages (levels 1-3) are less affected by
It seems, therefore, that the major differences in erosion than the material from level 0 or the 2 x 2m
sherd density are related to a combination of the surface collection discussed above (p. 135). As with
differential focus of late (Classical) activity and the typological data, the larger assemblages from the
mound formation processes, particularly involving scraped squares provide a more detailed gloss on
downslope movement of material. Any spatial dif- the patterns revealed by the surface collection. I used
ferences identified in typological terms may not there- the same areas as above (p. 122; Fig. 9.7) but sepa-
fore accurately reflect the exact location of different rated areas 8 (west slope of main mound) and 9 (top
phases or types of activity. of northern slope) because of the difference in eleva-
tion and gradient of these squares (Table 9.10).
Scraped squares
Despite the reduced control of recovery rates the 10 x
10m scraped square assemblages shed further light
Table 9.9. c;atalhoyiik East surface scraping- assemblage
on these patterns because of: the much greater quan- characteristics by depth.
tities of material; the investigation of the whole depth
of topsoil, not just the very surface; and the chance Level Total sherds % Neolithic Mean weight Mean weight Mean sherds
Neolithic (g) wheelmade (g) per square
to correlate these assemblages with architectural re-
0 2350 28.6 8.8 11.2 81.0
mains revealed underneath. I 11,257 29.8 15.4 18.1 331.1
2 3744 25.1 17.4 27.4 249.6
The scraped square material was collected in a 3 1158 60.5 18.1 22.9 144.8
series of spits, though not of uniform depth and

137
Jonathan Last

900 1000 1100 1200


metres east

0 20 40 60 80 100
percentage
Figure 9.10. c;atalhoyiik East surface collection- percentage of sherds <2 cm.

138
Surface Pottery

Sherd numbers reflect the better preservation of Neolithic architec-


Beginning from the overall density of ceramic mate- ture here. It seems that whereas the surface distribu-
rial (i.e. the approximate number of sherds per cubic tions reveal a general pattern which is largely the
metre of soil), there is a much greater concentration result of mound erosion, individual squares have
of pottery on the summit of the mound than any- undergone a variety of processes contributing to the
where else on the site. The next highest densities visibility or otherwise of Neolithic artefacts. It is
(but only about 15-20 per cent of this figure) occur in noticeable that on the northern eminence Neolithic
areas 8, 10 and 11 on or below the steep eastern and densities are lower both where there is good archi-
western slopes of the main mound. Moderate densi- tectural preservation (area 2) and where there is a
ties occur in areas 1, 6, 7, 9 and 13 on the shallower great depth of disturbed topsoil (area 1). Across the
northern and southern slopes of both the main site as a whole, the correlation between numbers of
mound and the northern eminence. This vague cor- Neolithic sherds and numbers of animal bone frag-
relation with degree of slope (r = 0.45) may indicate ments per square is r = 0.60, so a slight relationship
the effects of slopewash over a lengthy period, with is evidenced.
material from the eroded top of the mound moving In contrast to the east-sided distribution of
more rapidly down the steeper slopes. The numbers Neolithic pottery on the main mound, wheelmade
more or less match the surface distributions although sherd densities show a clear fall-off with distance
it should be noted that no scraping was carried out from the summit: the nearer assemblages (8, 10 & 13)
lower down on the southern slope where surface have higher densities than those further away (6, 7 &
densities are highest. In addition, the figures for the 11), irrespective of elevation (since area 7 is higher
steepest areas (8 & 10) appear rather higher than than 8 or 10). This suggests that erosion and weath-
might be expected from the density of finds in the ering have been responsible for some movement of
relevant 2 x 2 m squares. This may reflect a tendency material away from the focus of Classical activity on
for material brought to the surface here to subse- the top of the mound. The extent of erosion here is
quently be carried further downslope. shown by the high density of graves in area 12, the
Neolithic densities are highest on the eastern skeletons now apparently close to the ground sur-
plateau (area 11), relatively high on the eastern slope face. Undoubtedly downslope movement to the
(6 & 10) and the summit, moderate on the northern south, not investigated by scraping, has been greater.
and western slopes (7 & 8) and very low on the Lower densities of pottery are found on the
southern slope (area 13). This patterning may be northern eminence as a whole, but particularly in
influenced by the intensity and nature of Classical area 2. This is probably related to the good preserva-
activity in different areas. In areas 11 and 12 the tion of building plans in this part of the site: the sorts
digging of later graves has probably brought a lot of of post-depositional processes which have disturbed
Neolithic material to the surface, while the Classical the soil matrix and brought material to the surface
architecture in 9 and the depth of topsoil in squares elsewhere did not occur to such a degree here. In
8 and 13 may have masked Neolithic material. The addition, as excavation may be starting to confirm,
other squares with high densities revealed both the building fills of the relatively early phase repre-
Neolithic and Classical features, indicating a range sented on the top of the northern eminence may not
of processes, while the intermediate values in area 7 have contained large quantities of pottery. I also

Table 9.10. t;atalhiiyuk East surface scraping- assemblage characteristics by area.

Area Total sherds %Neolithic Mean weight Mean weight Volume (m 3 ) Total sherd Neolithic Wheelmade Mean
Neolithic (g) wheelmade (g) density sherd density sherd density slope

I 883 7.9 11.1 14.1 20.0 44.2 3.5 40.7


2 1048 57.3 9.6 11.3 147.7 7.1 4.1 3.0 1.1
3/4 1382 46.7 10.0 12.7 48.3 28.6 13.4 15.2 1.0
5 404 87.6 21.6 21.7 20.0 20.2 17.7 2.5
988 56.2 13.7 15.7 20.3 48.7 27.3 21.4 1.1
2898 40.5 15.9 19.0 67.2 43.1 17.5 25.6 0.9
2359 11.8 12.8 24.8 19.9 118.5 14.0 104.5 1.6
713 16.3 15.5 43.6 13.2 54.0 8.8 45.2 0.5
10 2414 33.1 16.7 21.6 27.5 87.8 29.1 58.7 2.1
II 1535 59.3 18.9 25.6 19.0 80.8 47.9 32.9 0.1
12 2766 4.6 9.1 11.2 5.3 521.9 24.0 497.9 0.7
13 1115 1.4 18.1 24.1 20.6 54.1 0.8 53.3

Note: slope figure is difference in metres between height of highest and lowest corner of square, averaged over area.

139
Jonathan Last

suggested that the topsoil assemblages contain some 9.11). It is noticeable that this general trend is mir-
elements which are later than the underlying con- rored locally within those areas consisting of more
texts (p. 132), so limited movement of material than one 10 x 10 m square- in areas 2, 3 I 4 and 7 the
downslope from the top of the eminence may have northern squares in the group have proportionately
occurred. The squares in area 314 certainly appear more Neolithic material than those to the south, so
more disturbed by later pitting. Whatever the proc- the fall-off in numbers of Classical sherds is indeed a
esses involved, the scraped square pattern again tal- consistent one.
lies with the surface collection data.
Both wheelmade and handmade sherds are Sherd weights
more frequent on the top of the eminence (area 3 I 4) Mean sherd weight for an assemblage may reflect a
than on the slope (area 2). Perhaps the greater Clas- number of processes: the intensity of trampling of an
sical activity in the former area has brought more area while the site was occupied, the degree to which
Neolithic material to the surface here. Limited an assemblage has been eroded by natural proc-
amounts of each type have then moved downslope. esses, or the extent to which movement and sorting
This hypothesis is strengthened by the more detailed of material has occurred due to animal/root action,
patterning within area 2 (p. 141). On the other hand wind, water or gravity.
area 1 has a much higher density of Classical pot- As with the sherd densities, the mean sherd
tery, implying a focus of later activity further down size for the Neolithic material relates in part to the
the northern slope, and one that may be extremely intensity of Classical activity in that part of the site.
localized since it does not appear to be reflected in Consequently, the largest Neolithic sherds are found
the surface distributions. The lack of a surface con- in those areas with the highest proportions of
centration implies this is not a slopewash deposit. Neolithic pottery: area 5 on the northern eminence
The relative proportions of handmade and and area 11 on the eastern plateau. These assem-
wheelmade pottery within the assemblages sum up blages have been less broken by later action. But
the nature of Classical activity. This confirms the area 13, with the lowest proportions of Neolithic,
focus of Classical activity on the top of the main also has a high mean weight; and the Byzantine
mound and its southern and western flanks, with an graves in areas 11 and 12 have not had the same
unexpected secondary focus in the very north of the effects on the prehistoric material, since the summit
site (area 1). Otherwise there is a fall-off in the pro- has very small sherds. In general on the main mound
portion of wheelmade sherds with increasing dis- mean sherd size increases downslope, which is the
tance north and east of the summit of the main opposite of the pattern observed for the 2 x 2 m
mound. In fact a graph of the proportion of Neolithic surface collection material. While the surface mate-
sherds against the northing value of the square shows rial probably indicates the effects of slopewash, the
the regularity of the increase, with the only excep- figures from the scraping may relate more to the
tions in areas 1, 5 (a relatively pure prehistoric as- intensity of later activity. Slight disjunctures between
semblage from the probable midden) and 11 (Fig. the two assemblages were also noted in terms of
90

80 ••
• • I•
70
60
50
• I
• I • • ••
%
40 • •
I
30 • •
20
••
10
0 • •
940 980 1020 1060 1100 1140 1180 1220 1260
northing

Figure 9.11. (.atalhoyiik East surface scraping- proportions of Neolithic sherds.

140
Surface Pottery

sherd size relative to depth and densities in the steep- aspect of the study focuses especially on the block of
est squares, implying that although the deposits are sixteen 10 x 10m squares excavated on the northern
well-sorted extensive vertical mixing may no longer eminence in 1993 and 1994 (area 2). At this relatively
be ongoing. small-scale patterns are less distinct. The density of
In addition, Neolithic sherds from the western all sherds (Fig. 9.12a) shows this apparent random-
side of the mound are smaller than those from the ness, with small clusters of high density adjacent to
east, perhaps because the density of Classical sherds groups of squares with a much lower density, rather
is greater to the west. However, the northern emi- than any clear gradation. It is, however, notable that
nence also has small sherds, which indicates that with the exception of 1030,1190, the highest densi-
erosion of the prehistoric material is not related di- ties are on the east and west sides of the area where
rectly to the extent of later activity. The wheelmade gradients are steeper. There is also a higher overall
sherds are also smaller here, so post-depositional density in the northern half of the area, and this too
processes are probably responsible. On the main is further downslope. The distributions appear there-
mound Classical sherds from the southern and west- fore to be slope-related. The three adjacent squares
ern slopes are larger than those from the north and of area 3/4 to the south on relatively flatter ground
east, not quite consistent with the Neolithic pottery, also have high densities but that area was rather
though the north/ south difference is similar. In fact more disturbed by later pitting, which could explain
there is a moderate correlation between the mean the difference.
weight of prehistoric and wheelmade sherds across In accordance with the focus of later activity,
the site as a whole (r = 0.54), which implies a fair the proportion of Neolithic material within the as-
amount of erosion and movement occurring since semblage increases towards the north (Fig. 9.12b).
the Classical occupation of the site. However, the density of Neolithic sherds in this area
As with the surface distributions, therefore, we effectively mirrors the overall totals, which is not the
can recognize two broad sets of processes in opera- case across the site as a whole. Indeed the coefficient
tion: one is the effect of differing intensities of Clas- of correlation between numbers of handmade and
sical activity on the Neolithic assemblages, the other wheelmade sherds is r = 0.77, and for the squares
is erosional processes affecting the mound as a whole south of 1190 the value rises to r = 0.93. This sup-
since the Byzantine occupation. ports the idea that post-Classical formation proc-
The direction of erosion and accumulation de- esses are responsible for much of the patterning.
pends on prevailing winds over a long period but Activity on the top of the northern eminence left
the high proportion of Classical material downslope wheelmade and handmade sherds together on the
on the southern side of the mound might indicate surface. Some of this material then eroded down-
major accumulation of material moving downslope. slope, hence the high correlation between the two
Rosen (1986) suggests the steepest face of the mound categories, with the strength of that relationship de-
would be the windward side where erosion has cut creasing further downslope.
back the tell, while soil creep would cause a shal- The hypothesis is strengthened by the detailed
lower slope to develop on the leeward side. On the results for the square 1020,1170 which was excavated
main mound at <;atalhoyi.ik the contours are indeed in 25 2 x 2m sub-squares (Fig. 9.13). The northern half
wider on the side where the bulk of the surface of the square had far fewer sherds than the south,
pottery is found, although the upper part of the fitting the pronounced difference between the squares
north slope (area 7) is also less steep. This side of the immediately north and south, and the correlation be-
mound is convex in profile while the south side is tween Neolithic and wheelmade sherd counts on this
slightly concave, which probably fits the hypoth- fine scale is still high (r = 0.79). The number of obsidian
esized direction of weathering. The steepness of the pieces is also moderately correlated with the pottery
western side of the mound and the relative lack of (r = 0.68) but clay balls and animal bone show com-
run-off deposits around the west and south sides of pletely different patterns and may therefore reflect
the mound reflect the presence of the river channel rather different formation processes.
as well as more recent tracks and drainage ditches. Mean sherd weights are also patterned rather
differently for the two assemblages. Again, for the
In close-up northern eminence as a whole there is a high degree
The formation processes in individual areas are bet- of variation between adjacent squares, but the mean
ter understood in the light of specific details con- weight of Neolithic sherds for all squares at northing
cerning their topography and archaeology. This 1150 and 1160 (area 3: 14.0 g) is greater than in those

141
Jonathan Last

0 5 10 15 20 25
sherds I cubic metre

Figure 9.12a. c;nta/lioyiik Enst surfnce scraping- northern eminence: density of n/1 sl1erds.

142
Surface Pottery

20 30 40 50 60 70 80 90
percentage

Figure 9.12b. c;atnlhoyiik East surface scraping- northern eminence: proportion of Neolithic sherds.

143
Jonathan Last

0 2 4 6 8 10 12
number

0 5 10 15 20
number

Figure 9.13. (:atalhoyiik East surface scraping- square 1020,1 170: sherd distributions.

144
Surface Pottery

at 1170 and 1180 (10.1 g), and the value declines to be the major contributary factor to the patterning
again at 1190 and 1200 (8.4 g). This supports the idea visible at a global level. At a finer scale, surface
of slope-related patterning but values for the artefact patterning may still to some extent reflect
wheelmade pottery do not show as strong a correla- underlying activity, but conclusions about this will
tion with northing, so some of the patterning may be retrospective rather than predictive, once suffi-
relate to pre-Classical formation processes or (more cient in situ material has been excavated that it can
likely, given the high correlation in counts) the Clas- be related back to the surface distributions. The sur-
sical sherds have a greater range of fabric types which face work at (atalhbyiik will probably, therefore,
are generally harder fired and therefore more resist- prove more productive as an exercise in understand-
ant to abrasion than the fairly uniform Neolithic ing mound formation processes than as a means of
pastes. Similarly, within 1020,1170 the correlation identifying any more than just the coarsest activity-
coefficient for mean weights of handmade and related patterning. It is hoped that ongoing archaeo-
wheelmade sherds is only r = 0.31. It is likely that logical and geomorphological work will in coming
abrasion processes at (atalhoyiik have been rela- seasons add to these superficial conclusions.
tively slow and more liable to affect the softer fabrics.
Given the great influence of post-depositional The Classical material
processes on the patterning of the Neolithic sherds
there is little sense of any correlation between sherd Here is not the place for a full discussion of the
numbers and aspects of the underlying archaeology, wheelmade assemblage. The general terminology of
although it is possible that those squares with high 'Classical' covers material of the Late Iron Age/Hel-
densities of pottery (1020,1170, 1020,1190, 1020,1200) lenistic, Roman, Byzantine and (probably) medieval/
are associated with larger spaces or fewer visible Seljuk periods. Despite this broad range of data the
walls (Fig. 9.12). But 1030,1190 is again an exception nature of Classical occupation and activity at (atal-
(with many walls) and it is likely, moreover, that hoyiik is not fully understood and the central
visibility of walls is itself an effect of post-depositional Anatolian typology remains generally understudied
disturbance. and unpublished. Coastal sites in Anatolia such as
Anemurium (Williams 1989) provide few clear par-
Conclusions allels. Priorities for future work on the later occupa-
The end result of the taphonomic processes described tion of the site should include the recovery of closed
above is a lateral movement of material downslope assemblages from stratified contexts and compari-
from the top of the mound. The fate of sherds and son with material from other sites in the region in-
other items brought to the surface by processes of cluding Amorium (Harrison 1990) and Sagalassos
vertical mixing and erosion will largely depend on (Waelkens et al. 1992, 91-8).
slope angle, and secondarily on the size, shape and A limited amount of permanent occupation on
density of the objects. At (atalhoyiik, many parts of the site would not be inconsistent with the evidence
the mound have sufficiently steep slopes for lateral of a much larger Classical site a few hundred metres
movement, primarily a gravity effect, to be a signifi- south and east of the site, discovered during non-
cant factor. The skirt of material seen around the systematic field-walking in 1994. The finds from this
mound, with a high proportion of small sherds, site include Late Roman red slip wares, Byzantine
largely confirms this, although it remains to be es- tiles and coarse wares with incised 'wavy line' deco-
tablished how far the prevailing direction of move- ration, as well as occasional green glazed pieces.
ment from the upper part of the main mound is Hellenistic sherds appeared to be rare, so occupa-
related to climatic factors. tion in this phase may have been focused elsewhere,
Only on the relatively flat eminences might hori- perhaps on the mound itself, which seems otherwise
zontal movement be unimportant, and even this to- to have been primarily reserved for activities such as
pography may be comparatively recent. The 'halo' quarrying clay and burying the dead. On the other
of sherds around the base of the northern eminence hand, the quantities of pottery found, particularly
perhaps indicates deeper erosion of the surface than coarse wares, imply that storage and consumption
on the main mound, to the extent that Neolithic activities also occurred on the mound.
levels have now been exposed on the eminence, wit-
nessed by the success of the surface scraping there Surface collection
(Neil Roberts pers. comm.). The distribution of the surface material has already
Post-Classical erosion of the site seems, therefore, been commented upon (p. 120), indicating a major

145
Jonathan Last

focus of wheelmade pottery on the main mound and Figure 9.14


much more limited evidence from the northern emi- 1. 191/3 Wheelmade bowl rim, flanged; diam. c. 450
nence where, consequently, the surface scraping was mm, thickness 11-15 mm; ext.light grey, painted red on
concentrated. Hellenistic slipped wares, principally flange; int. painted red, impressed twisted cord decora-
characterized by buff or orange paste and an uneven tion; fabric light pinkish brown, moderate med. mineral.
('streaky') slip applied on the wheel (Fig. 9.15:6), Unit 191, area 8; Roman.
occur across the main mound and in a secondary
concentration on the shallow slopes north and west 2. 294/1 Wheelmade bowl; rim slightly inturned and
of the northern eminence (Fig. 9.16). Roman red thickened with external bevel, diam. c. 320 mm, thick-
sigillata-type fabrics (Fig. 9.14:6) are similarly dis- ness 10 mm; base flat, pronounced junction, diam. 70
tributed. On the other hand, none of the few late mm, thickness 10 mm; vessel height 80 mm; ext. and int.
Byzantine or medieval green-glazed pieces derived pinkish brown; fabric light reddish brown, moderate
from the major concentration of wheelmade mate- mineral. Unit 294, area 1; ?Roman.
rial on the southern slope of the main mound (Fig.
9.17). The vast majority of the Classical pottery, how- 3. 24/31 Wheelmade bowl rim, slightly inturned;
ever, is coarse wares. I attempted to look for possible diam. 140 mm, thickness 4 mm; ext. slipped light reddish
functional variation across the mound by plotting brown and burnished with zigzag stripes, horizontal
attributes like ring bases, handles and the wall thick- grooves below rim; int. slipped light reddish brown and
ness of all rim sherds. But apart from the general burnished with vertical stripes; fabric light grey, sparse
increase in all types on the southern half of the main small mineral. Unit 24, area 9; Hellenistic.
mound the only evident patterning was a tendency
4. 191/2 Wheelmade bowl; rim slightly inturned, diam.
for thinner-walled vessels in the northwest part of
200 mm, thickness 7 mm; base footring, diam. 75 mm,
the site. This may be merely a reflection of the chrono-
thickness 7 mm, height c. 8 mm, width at base 7 mm; vessel
logical differences discussed below.
height 66 mm; ext. and int. light grey; fabric light grey,
sparse small mineral. Unit 191, area 8; Hellenistic/Roman.
Scraped squares
The scraped squares provide more clues to the types 5. 156/4 Wheelmade bowl rim, rather pointed; diam.
of activity evidenced. Mellaart reports Hellenistic 150 mm, thickness 5-7 mm; ext. pinkish brown, rim
pits in the area of his excavations on the west side of horizontally grooved and slipped dark reddish brown;
the main mound, and some decorated pieces pre- int. pinkish brown, rim and upper body slipped dark
sumably from these features are now in Konya Mu- reddish brown with thin horizontal bands of slip around
seum. Similar pits, other indeterminate features and base; fabric light brown, moderate brown and cream
a late wall were also observed on the top of the mineral. Unit 156, area 8; ?Roman.
northern eminence (area 3/4) but the best evidence
for actual occupation (Classical architecture and in 6. 124/sf3 Wheelmade bowl, complete; rim pointed,
situ pots (Fig. 9.15): see Chapter 7) comes from area 9 diam. 120 mm, thickness 2-4 mm; base foot-ring, diam. 45
below the summit of the main mound, while areas mm, thickness 5 mm, height of ring 3-5 mm, width at base
11, 12 and 13 revealed definite or possible graves 2 mm; vessel height 58 mm; ext. pink, painted red in band
(one Byzantine child burial was excavated in area 11 below rim; int. painted red; fabric pink, moderate small
in 1993). The largest Classical sherds, therefore, come mineral, inc. mica. Unit 124, area 10; Hellenistic/Roman.
from area 9, but denser concentrations of pottery
were found on the summit (area 12) and the slopes 7. Wheelmade body sherd with moulded decoration of
(areas 8, 10 & 13). male crouching over reclining female on a bench; body
A detailed description of the individual con- diam. 140 mm, thickness 3-7 mm; ext. and int. glossy
texts is not attempted here but the material from red slip; fabric light red, no visible inclusions. Unit 278,
unit 67 in area 10 is typical of the larger wheelmade area 6; Roman.
assemblages (cf. Fig. 9.15:10). The rim sherds from
this unit fall into a number of groups. Genuine 8. Wheelmade bowl body sherd with moulded decora-
finewares are rare, consisting of three slipped sim- tion on exterior of ithyphallic ?satyr and other figures
ple bowls, one of Hellenistic 'streaky' slip type, and below a narrow frieze; body diam. 80 mm, thickness 4-6
one late dark green-glazed bowl. 'Semi-fine' wares, mm; ext. mottled dark reddish brown to pinkish grey;
presumably of local manufacture, fall into three int. weak red; core weak red, very sparse fine mineral.
groups: a dark 'streaky' slip on buff, brown or grey Unit 191, area 8; Roman.

146
Surface Pottery

---- -

( 3

{ -I
J
7
{ 8

Figure 9.14. Whee/made rim forms from (ata /hoy iik East.

147
Jonathan Last

Figure 9.15 painted dark reddish brown above carination; int. light
1. 156/27 Wheelmade bowl rim, flat-topped and brown painted dark reddish brown in bands including
externally thickened with high, thickened shoulder; diam. stylized floral(?) decoration; fabric light greyish brown,
310 mm, thickness 10-12 mm; ext. cream with double moderate mineral. Unit 12, area 9; Hellenistic.
twisted cord impression below shoulder; int. slipped light
pinkish grey; fabric light reddish brown, moderate small/ 3. 12/7 Wheelmade jar rim with flat top and upturned
med. mineral. Unit 156, area 8; ?Roman. flange forming internal ledge for lid; diam. 190 mm,
thickness 6-8 mm; ext. dark grey with horizontal ridges
2. 12/6 Wheelmade carinated bowl body; diam. 260 and shallow grooved wavy line; int. dark grey; fabric
mm at carination, thickness 7-9 mm; ext. light brown dark grey; moderate mineral inc. mica. Unit 12, area 9;
Byzantine.
paste, generally bowls with simple or flat rims; a
reddish-brown 'streaky' slip on buff or pink paste, 4. 278/sf2 Wheelmade amphora neck, externally
generally bowls with inturned rims; a more even thickened and pointed, with loop handle; mouth diam. 55
matt red slip on pink paste, all simple bowls. Coarse mm, thickness 7-9 mm; ext. painted streaky red/reddish
wares fall into a wider range of fabric groups, gener- brown; int. light red; fabric light red, sparse mineral.
ally with sparse or moderate mixed mineral inclusions: Unit 278, area 6; ?Hellenistic.
1. Plain red surfaces on a grey paste - jars; bowls
with simple or developed rims; occasional cord- 5. 24/20 Wheelmade carinated bowl rim, inturned
impressed decoration; with flat top and upturned flange forming internal ledge
2. red or pink paste with overall or partial red slip - for lid; diam. 280 mm, thickness 6-9 mm; ext. and int.
jars; bowls with thickened or flanged rims; occa- mid brown, fire-blackened in places; ext. decorated on
sional cord-impressed decoration; shoulder with horizontal grooves framing shallow
3. red or pink paste with light-coloured surfaces grooved wavy line; fabric light grey, moderate small
and some red paint - large jars; large bowls black and white mineral. Unit 24, area 9; Byzantine.
with developed rims;
4. pink paste with cream slip - bowls with flat, 6. 603/8 Wheelmade bowl rim; diam. 140 mm,
rolled or simple rims; thickness 2-4 mm; ext. and int. 'streaky' slipped light
5. grey pastes with dark slip or burnish- mainly pinkish brown; fabric light orange brown, sparse small
jars; white mineral. Unit 603, area 12; Hellenistic.
6. unslipped brown or grey pastes, some bur-
nished -jars; bowls with simple or rolled rims, 7. 191/sf35 Wheelmade handled jar; rim everted,
occasional cord-impressed decoration. diam. 30 mm, thickness 5 mm; base flat, angular junc-
Handles occur in all the coarse ware fabrics, and are tion, diam. 28 mm, thickness 5 mm; vessel height 64
generally oval or elongated in cross-section. About mm; ext. and int. light pinkish grey; fabric orange,
20 per cent of the bases are footrings, generally but sparse small mineral; handle rectangular section 9 x 4
not always in the finer fabrics. Amphora and pedes- mm. Unit 191, area 8; ?Roman.
tal bases are rare.
Area 10 is fairly typical of the site as a whole, 8. 191/sf36 Wheelmade handled lamp(?); rim
although it does include a relatively large number of inturned with internal bevel, diam. 39 mm, thickness 4-
cord-impressed sherds (Fig. 9.15:1), as does area 11 5 mm; base flat, pronounced junction, diam. 60 mm,
nearby. The chronological significance of this fea- thickness 1 mm; ext. and int. light pinkish grey; fabric
ture is uncertain but it occurs on some flange-rimmed pink, sparse small mineral. Unit 191, area 8; ?Roman.
bowls of Roman type. It is more common on the
main mound than the northern eminence. 9. 295/13 Wheelmade jug rim with pinched spout;
It is clear that all the wheelmade assemblages diam. 40 mm, thickness 3 mm; ext. 'streaky' slipped dark
are very mixed, most including examples of every- reddish brown; fabric light orange brown, very sparse
thing from Hellenistic slipped wares to medieval mineral. Unit 295, area 1; Hellenistic.
glazes, but some chronological variation across the
site may be apparent (Table 9.8). The highest pro- 10. 67fsf30 Wheelmade base sherd, flat, pronounced
portions of 'streaky' slipped vessels occur in area 9, junction; diam. 85 mm, thickness 8 mm; ext. light grey,
while the lowest proportions occur at the north end incised geometric design; fabric mid brown. Unit 67,
of the site (area 1), on the main summit (area 12) and area 10; ?Classical.

148
Surface Pottery

6
5

-- ·- \7
I
I

~
8
7 ··-··

10
9

Figure 9.15. Other whee/made rim form s from (atalhoyiik East.

149
Jonathan Last

1300

1200

1100
.s....
g
~
.b
6

1000

900

800

900 1000 1100 1200


metres east

Figure 9.16. c;ata/hoyiik East surface collection - distribution of Hellenistic slipped wares.

150
Surface Pottery

1300

1200

.s 1100
....
g
~
s
1000

900

800

900 1000 1100 1200


metres east

Figure 9.17. Cata/hoyr:ik East surface collection - distribution of green-glazed wares.

151
Jonathan Last

on the north and east sides of the main mound (areas has more cord impressed sherds and less painted
6, 7, 10 & 11). This refines the pattern observed for decoration, and which may therefore be a little later;
the surface material. Wavy line grooved or incised and that on the northern eminence with more glazed
decoration typical of the Byzantine period (Fig. 9.15:5) wares and cream-slipped coarse wares, which is prob-
is more or less ubiquitous (though absent in areas 6 ably later again.
& 12), while glazed wares show a pronounced focus
in the northern part of the site (areas 2, 3/4 & 7) and The west mound and the Early Chalcolithic
are absent from areas 8, 9, 11 and 13, confirming the
picture from the 2 x 2 m surface collection. There Introduction
may therefore be a late but less intensive occupation The west mound is a rather different phenomenon
in the north. from <;atalhoyiik East. Smaller and lower, it has suf-
Differences between the assemblages may also fered more from ploughing. Hence when Mellaart
reflect functional variation across the site: the ratio (1965) excavated two small trial trenches in the 1960s
of jar to bowl rims varies somewhat, with the high- he found no coherent architecture, and the surface
est proportion of storage vessels in areas 8, 9 and 1, scraping in 1994 was equally fruitless in this respect
and the lowest in area 2. The large number of jars (Chapter 7). Nevertheless both Mellaart's work and
and fineware bowls in area 9 suggests a broad range the recent investigations recovered a large ceramic
of activities associated with the coherent architec- assemblage, demonstrating the ubiquity of pottery
ture in this area. A probable handled oil-lamp (Fig. in the Early Chalcolithic compared with the Neolithic.
9.15:8) also came from this area. Around this possi- In three squares some 3800 sherds were recovered,
ble occupation focus, the nearest assemblages (6, 7, about 20 per cent of the amount that came from 33
10 & 12) have fewer jars and lower numbers of squares on the east mound. Moreover, the amount
fineware and decorated vessels. The summit, for in- of Classical material on the west mound is much
stance, is characterized by simple bowls (Fig. 9.14:6) less: fewer than 7 per cent of the sherds are wheel-
and ring bases, with relatively few handles or large made, so about 39 per cent of all the prehistoric
pots; the eastern eminence (area 11) is similar. Both pottery comes from the small investigation on
these squares contain late graves. However, area 6 <;atalhoyiik West.
has some elements which may be linked to the nearby The defining element of the Early Chalcolithic
occupation, such as a well-preserved amphora neck pottery is its characteristic red-painted decoration,
and handle (Fig. 9.15:4) and a fragment of a moulded comprising a limited range of linear geometric mo-
samian-type bowl, the only other example of which tifs. 38 per cent of rim sherds found in 1994 had such
comes from area 8 (Fig. 9.14:7-8). decoration (cf. Fig. 9.18); a further 21 per cent had a
The southern slope of the main mound (area red wash. The main design elements on the painted
13) also has relatively few jar rims, but a large number pottery may be divided into three broad groups:
of handles and thick (> 10 mm) body sherds suggest parallel lines (Fig. 9.18:5), chevrons or zigzags (Fig.
many big storage vessels there. To the north, area 2 9.18:2), and lattices (Fig. 9.24:3). The designs repro-
again has an overwhelming predominance of open duced by Mellaart (1965) suggest that in general
bowl forms with few jars or handles, while area 1 zigzags are more common in the early phase (EChal
has a distinctive assemblage with many jars and jugs I) and lattices later (EChal II). Mellaart also reports
(Fig. 9.15:9) but relatively few fineware or decorated that while fabric and paint colours overlap between
pots, underlining the separateness of this part of the the two phases EChal I paint is never dark brown or
site (p. 140). black while in EChal II these shades are common.
The vast majority of the pottery recovered in 1994
Conclusions can be assigned to his EChal I. In one of the larger
More detailed understanding of the patterns dis- contexts 11 per cent of the decorated sherds had
cussed here and of the meaning of the variation brown or black paint; a few of these had appar-
between assemblages depends on further typologi- ently been burnt.
cal analysis as well as systematic investigation of As well as the introduction of painted decora-
some of the Classical features. At the moment a tion, clear differences from the Neolithic pottery of
number of different areas may be recognized: that the east mound can be seen in forms and fabrics.
focused on 9 with some early features but the assem- However, the new forms may be derived from ear-
blage becoming less distinctive in the surrounding lier prototypes: the late Neolithic S-profiled bowls
squares; that on the east of the main mound which (Fig. 9.4:3) and holemouths with incipient necks (Fig.

152
Surface Pottery

9.1:5 & 6) could have given rise to the Chalcolithic is evidenced here. Further investigation would be re-
carinated bowls (Fig. 9.18:3) and jars with everted quired to assess the significance of this, since none of
necks (Fig. 9.18:9) respectively. The generally well- the scrape squares was located on this side of the site.
fired fine buff or pink pastes are indicative of further Wheelmade fine wares are rarer than on the
technological advances in pottery manufacture. As east mound and it appears that the focus of occupa-
well as painted decoration, new surface treatments tion was later than on the other site. Hellenistic
include slips and washes as well as fine scraping of slipped wares are not common, and are restricted to
interiors, which on small fragments can resemble the top and east side of the mound. There are also a
wheel marks. Burnishing continues, although fine few green-glazed sherds, concentrated in a small
polishes comparable to the best of the Neolithic ves- area in the southeast.
sels are rare.
Scraped squares
Surface collection The three scraped assemblages largely confirm the
As at <;atalhoyiik East, the entire surface of the above picture: the western square (480,1040: see Fig.
mound was investigated by the collection of all finds 7.19) had by far the lowest proportion of wheelmade
within 2 x 2m squares at 20m intervals. The hand- pottery (1 per cent compared to about 10 per cent in
made pottery is fairly evenly distributed, with den- the other squares) which supports the distribution
sities falling off around the fringes of the site and indicated by the surface data. These sherds are also
perhaps rather fewer sherds on the north part of the very small (12.9 g on average compared to 26.6 g and
mound (Fig. 9.19). This contrasts with a focus of 30.2 gin the other squares), implying they may be
animal bone finds to the northwest. The wheelmade residual. The density of Chalcolithic material in this
pottery, although a very small assemblage, is heav- area is rather higher than in the central (580,1020)
ily concentrated on the eastern half of the site (Fig. and eastern (640,960) squares, although mean sherd
9.20), suggesting that these sherds derived ultimately weight for the Chalcolithic material also falls off
from the areas of more intensive activity on the other from south and east to north and west.
mound. This may reflect the general fall-off of mate- Without any contextual information in either
rial in the fields around a Classical site related to case the 1994 material can add little to the fairly
processes of middening and manuring. There is some
indication that mean sherd size is greatest on the top Table 9.11. c;atalhoyuk West surface scraping- Chalcolithic
of the mound and generally smaller to the north and assemblages.
west of the site.
a: Preservation
The distributions of the different painted de-
sign elements show little significant patterning. Square Sherds ')', Chal. Mean weight Mean weight Volume Chal. sherds
Chal. (g) Wheelmade (g) density
Mellaart suggests that the later (EChal II) pottery is 480,1040 1296 98.9 24.3 12.9 18.8 68.2
concentrated on the southern slope of the mound 580,1020 1519 89.7 25.5 30.2 24.5 55.6
640,960 991 90.7 31.6 26.7 19.3 46.6
(and placed his second trench to investigate this).
But the surface collection indicates dark paint is more b: Rims and bases
common on the east side of the site, with a smaller Square 7(J Plain
%Hole· <f" Jars % Carinated % Foot-rings
concentration in the west (Fig. 9.21). Pot stands with mouths bowls bowls
incised decoration are also more common in EChal 480,1040 I 1.3 5.2 60.0 23.5 19.8
580,1020 19.7 38.2 29.6 13.3 10.4
II, but these were rare on the surface; the only frag- 640,960 11.1 16.0 48.1 24.7 12.2
ment was found in the west at 380,1020.
Possible functional differentiation was consid- c: Lugs
ered in terms of a number of features. The presence Square % Lugs %Basket 7() Plain % Perforated % Crescent % Vestigial
& handles handles lugs lugs lug handles lugs
of handles and large lugs may reflect storage, since
480,1040 3.0 33.3 17.9 10.3 12.8 25.6
the bowls and open vessel forms of this period gen- 580,1020 2.9 65.0 10.0 2.5 17.5 5.0
erally lack such features. However, the only signifi- 640,960 3.0 51.9 I 1.1 11.1 14.8 11.1

cant pattern within the surface material is that basket d: Decoration


handles were generally absent from the northeast-
Square Rims ')',Straight %Zig-zag %Lattices
ern slope (Fig. 9.22), while other lug types concen- %painted %wash %burnish lines
trated in the centre and north of the mound. 480,1040 18.3 31.3 50.4 30.1 34.5 14.2
Thick-walled pots (>8 mm) were also rare in the 580,1020 42.1 5.3 52.6 20.6 50.3 18.7
640,960 61.7 21.0 17.3 25.5 50.3 15.3
northeast so perhaps a greater proportion of fineware

153
Jonathan Last

detailed typological and decorative element analysis Figure 9.18


carried out by Mellaart (1965). One unusual form 1. 860/6 Handmade bowl rim, slightly everted; diam.
(not illustrated) is a large (c. 300 mm rim diameter) 165 mm, thickness 6-7 mm; ext. slipped orange-red and
simple jar with red-painted decoration similar to painted pinkish brown, diverging lines. Unit 860, square
that on the body of a jar illustrated by Mellaart (1965, 640,960; Early Chalcolithic I.
fig. 3, no. 1), but lacking the developed neck of that
vessel. Other points of emphasis include the relative 2. 858/13 Handmade carinated bowl rim, everted;
frequency (particularly in the west) of small ('ves- diam. 130 mm, thickness 4-5 mm; ext. slipped and
tigial') lugs or knobs, usually on the carinations of painted red, parallel chevrons; int. slipped and painted
painted bowls (Fig. 9.23:3 & 4: cf. Mellaart 1965, fig. red, diagonal lines. Unit 858, square 580,1020; Early
5, no. 7). Sometimes these features occur below han- Chalcolithic I.
dles (Fig. 9.23:5) where Mellaart considers them
skeuomorphs of metal rivets. It is notable that those 3. 860/1 Handmade carinated bowl rim, slightly
handles painted with chevrons reminiscent of bas- everted; diam. 170 mm, thickness 4-6 mm; ext. slipped
ketry patterns (Fig. 9.23:6) are not associated with and painted red, parallel chevrons. Unit 860, square
such features. A further point regarding lugs is that 640,960; Early Chalcolithic I.
on the larger plain ware vessels simple round lugs
(Fig. 9.23:5) appear to be almost as common as the 4. 857/4 Handmade jar rim; thickness 6-8 mm; ext.
crescent lug handles (Fig. 9.23:1) emphasized by slipped very light brown and painted light reddish
Mellaart. brown, parallel chevrons; fabric pink with grey core.
Painted elements include a few decorative forms Unit 857, square 580,1020; Early Chalcolithic.
not previously illustrated, such as a ring of parallel
wavy lines (Fig. 9.24:4), and concentric stars (Fig. 5. 858/15 Handmade bowl rim, inturned; diam. 160
9.24:8). Simple incised decoration was found on the mm, thickness 5-6 mm; ext. slipped very light brown
underside of one small base (Fig. 9.23:9). The record- and painted mid brown, parallel lines. Unit 858, square
ing of Munsell colours allows a more precise de- 580,1020; ?Early Chalcolithic II.
scription of paint and fabric colours. The fabrics of
the painted ceramics concentrate around 5YR 6/6- 6. 856/18 Handmade jar rim; diam. 160 mm, thick-
7 I 6, although many were less red and more buff in ness 4-5 mm; ext. dark brown, incised white-filled
colour (to 5Y 8/2). Paint colours were generally red, horizontal and vertical lines of small strokes; int. dark
distributed around lOR 4/6. 'Red wash' colours were brown; fabric grey. Unit 856, square 480,1040; ?Early
more variable, possibly because the thickness of the Chalcolithic II.
application of the paint varies. There were two main
groups, one which ranged from lOR 4/6 to 5YR 7/6 7. 858/3 Handmade bowl rim with anti-splash rim;
and another smaller group around 2.5Y 8/2-4. Dark diam. 180 mm, thickness 4 mm; ext. slipped mottled
brown paint of EChal II type was rare. light brown to grey and rim painted dark red, parallel
Unusual finds include a further example of the chevrons; fabric reddish brown, sparse med. mineral.
dark grey burnished ware with white-filled incisions Unit 858, square 580,1020; Early Chalcolithic I.
which Mellaart (1965) describes (Fig. 9.18:6), and two
unparalleled fine burnished sherds decorated with 8. 855/2 Handmade carinated bowl rim, slightly
deep grooves in curvilinear motifs (Fig. 9.24:9). All everted and pointed, with 'vestigial' lug; diam. 180 mm,
these items come from the western square. thickness 3--4 mm; ext. slipped very light brown to
Although the nature of settlement on the west orange and painted dark red, parallel chevrons; int.
mound remains unknown, there are distinct differ- streaky red wash; fabric pink; lug c. 6 x 5 x 1-2 mm.
ences between the three scraped square assemblages Unit 855, area 480,1040; Early Chalcolithic I.
which may relate to a mixture of chronological and
functional factors (Table 9.9). Mean rim thickness is 9. 855/13 Handmade jar rim, slight internal bevel;
virtually the same in all three areas, although per- diam. 150 mm, thickness 5 mm; ext. and int. pink. Unit
haps fractionally greater in the centre, which has a 855, square 480,1040; Early Chalcolithic.
larger number of plain (coarse) wares. The eastern
square has most painted rim sherds; the western
square has least, but proportionally more red wash
rims. Including body sherds, this square also has

154
Surface Pottery

\ /Qf 1

l 2
3

lCD
t7 ~, "

J
,o_
,
d .
'
,.-

_:.
\ · .·.· · ·..,
'
..

(
7
~- - - - - - - - - - , - - - - -- - -- ,.,.

Figure 9.18. Chalcolithic painted rim forms from c;atalhoyiik West.

155
Jonathan Last

1100

1000

900

800

400 500 600 700


metres east

0 50 100 150 200 250 300 350


number

Figure 9.19. (:atalhdyiik West surface collection- Chalcolithic sherd numbers.

156
Surface Pottery

1100

1000

900

800

400 500 600 700


metres east

0 1 2 3 4 5 6 7 8 9 10
number

Figure 9.20. c;nta/hoyfik Wes t surface collection - wheelnwde sherd numbers.

157
Jonathan Last

1100

1000

...c::
.....

-;
....
0
c::
(fJ
~
.b
~

900

800

400 500 600 700


metres east

Figure 9.21. (ata lhoyiik West surface collection- distribution of dark painted wares.

158
Surface Pottery

1100

1000

.£j
.....

-;
0
c::
(fJ
(j)
b(j)
s
900

800

400 500 600 700


metres east

Figure 9.22. (:atnlhoyiik West surface collection- distribution of lugs and handles .

159
Jonathan Last

Figure 9.23 V-shaped, c. 35 x 12 x 2 mm. Unit 856, square


1. 858/9 Handmade jar rim, pointed, with lug handle; 480,1040; Early Chalcolithic.
diam. 270 mm, thickness 8-10 mm; ext. orange;
crescentic lug 58 x 35 x 21 mm. Unit 858, square 3. 856/1 Handmade carinated bowl body/base sherd
580,1020; Early Chalcolithic. with 'vestigial' lug; thickness 5-6 mm; ext. slipped very
light brown and painted reddish brown, parallel chev-
2. 856/21 Handmade body sherd with 'vestigial' lug; rons; int. pink with red wash; fabric light brown, sparse
thickness 7 mm; ext. and int. orange; fabric orange; lug small mineral; lug 8 x 4 x 2 mm. Unit 856, square
480,1040; Early Chalcolithic.
more straight line ornamentation and fewer zigzags
than the other squares. Whether this reflects chrono- 4. 858/23 Handmade body sherd of carinated bowl
logical differences is unclear since lattice motifs, with 'vestigial' lug; thickness 4-6 mm; ext. slipped light
which might indicate a later date, are roughly equal yellowish brown and painted dark red, chevrons; lug c. 5
in numbers across all squares. The difference may x 2 x 1.5 mm. Unit 858, square 580,1020; Early
therefore relate to variation in vessel form, or merely Chalcolithic I.
to the smaller average size of sherds in this context,
meaning more pieces have only single lines pre- 5. 858/7 Handmade strap handle with lug; elliptical
served. section 54 x 24 mm; ext. and int. slipped yellow; base
Mellaart indicates that particular decorative shows 'tenon' for joining to body; lug on ext. angle,
schemes are found on different vessel types. For in- round, 19 x 18 x 15 mm. Unit 858, square 580,1020;
stance, vertical zigzag patterns are characteristic of Chalcolithic.
everted jar rims. Despite having fewer zigzag motifs
overall, these forms are more common in the west 6. 856/15 Handmade basket handle with part of vessel
(30 per cent of painted rims compared with 20 per rim; oval section, ext. flattened, 29 x 16 mm; ext.
cent in the east and none evident in the centre). This painted red, parallel chevrons; fabric orange. Unit 856,
fits the impression from rim shapes that jars (Fig. square 480,1040; Early Chalcolithic I.
9.18:4) are most common in the west, although it is
often difficult to distinguish jar and bowl forms from 7. 858/20 Handmade body sherd of carinated bowl;
small rim sherds. Carinated bowls (Fig. 9.18:8) ap- thickness 9-12 mm; ext. slipped light orange brown and
pear roughly equal in the west and the east, but are painted dark red to dark reddish brown, parallel chev-
less common in the central square, which- in accord- rons. Unit 858, square 580,1020; Early Chalcolithic I.
ance with the larger number of plain ware sherds-
seems to be dominated by simpler forms like open 8. 855/12 Handmade base sherd, flat, angular junc-
bowls (Fig. 9.18:5) and holemouth jars (Fig. 9.23:1). tion; diam. uncertain, thickness 10 mm; ext. light orange
There are also differences in base types, with brown, painted reddish brown; fabric pink. Unit 855,
the central square having more angular base junc- square 480,1040; Early Chalcolithic.
tions (Fig. 9.23:8) and the west more rounded bases:
this may reflect the difference between open and 9. 855/8 Handmade base sherd, flat, pronounced
carinated bowl forms (cf. Mellaart 1965, figs. 4 & 5). junction; diam. 60 mm, thickness 5-6 mm; ext. very
The western square has more footrings (Fig. 9.23:11), light brown, painted pinkish brown and incised with
but the significance of this is again unclear: carinated parallel linear strokes around edge of base. Unit 855,
bowls may have either flat or ring bases. square 480,1040; Early Chalcolithic.
Lugs and handles come in a greater variety of
forms than in the Neolithic. Across the scraped 10. 860/5 Handmade base sherd, foot-ring; diam. c. 80
squares they consistently represent some 3 per cent x 55 mm, thickness 5 mm, height of ring c. 8 mm, width
of all sherds. However, there is variation between at base 3 mm; ext. slipped and painted red. Unit 860,
contexts in the proportions of different forms. The square 640,960; Early Chalcolithic.
central square, with more of the simple jar forms
associated with basket handles (Fig. 9.23:6), accord- 11. 855/5 Handmade base sherd, foot-ring; diam. 70
ingly has more of these. It also has rather more cres- mm, thickness 4 mm, height of ring 12 mm, width at
cent lug handles (Fig. 9.23:1) than the other contexts, base 4 mm; ext. light orange brown painted pinkish
which have simple round and perforated lugs. The brown; int. orange, painted pinkish brown. Unit 855,
purely decorative or skeuomorphic 'vestigial' lugs square 480,1040; Early Chalcolithic.

160
Surface Pottery

2 3

11
Figure 9.23. Chalcolithic lug forms from <:;atalhoyiik West.

161
Jonathan Last

are least frequent here and most common in the locate many of these rooms and discover the interior
west, which has more painted sherds. The feature of fixtures and decorations which led Mellaart to label
this type illustrated by Mellaart occurs on a bowl them as either 'shrines' or 'houses'. This ritual/do-
with pedestal base, so the high number of footrings mestic distinction based on the degree of elaboration
in the western square may be a related feature. of the room must, however, be called into question.
The small wheelmade assemblage from these The general similarity of the architecture of so-called
squares probably dates to the Byzantine period, al- shrines (the criteria for which are never adequately
though diagnostic pieces are lacking. Despite the defined) and houses suggests that no clear dichotomy
correlation in densities with the 2 x 2 m surface
collection, the scraped square assemblage appears Figure 9.24
rather different typologically. Coarse wares are of- 1. 855/23 Handmade body sherd, probably of
ten plain, relatively thin-walled jars in dark, mica- carinated bowl; thickness 8 mm; ext. slipped very light
ceous fabrics, presumably of a different provenance brown and painted orange/yellow, parallel lines. Unit
from the east mound pottery. There are some painted 855, square 480,1040; Early Chalcolithic I.
wares (Fig. 9.25:4) but no Hellenistic pieces. Whereas
the surface scatter of mixed date may derive from 2. 857/10 Handmade body sherd; thickness 7-9 mm;
Classical agriculture, this small assemblage indicates ext. slipped very light brown and painted light reddish
a distinctive but limited occupation, sited away from brown, parallel vertical lines; grey core. Unit 857,
the possible brick-lined graves revealed in the western- square 580,1020; Early Chalcolithic II.
most square.
3. 858/24 Handmade body sherd; thickness 7 mm; ext.
Future work slipped very light brown and painted dark reddish-
The major task for future seasons' work on the west brown, lattice. Unit 858, square 580,1020; Early
mound at C::atalhoyiik is the recovery of closed as- Chalcolithic II.
semblages from different excavated features, in or-
der to put the ceramic typology into chronological 4. 858/6 Handmade body sherd; thickness 9-10 mm;
and spatial context. Given the amount of topsoil ext. slipped light pink and painted dark brown, parallel
disturbance to which Mellaart' s soundings and the curvilinear lines; int. orange. Unit 858, square
scraped squares bear witness, fairly major trenching 580,1020; Early Chalcolithic.
will probably be necessary in order to expose undis-
turbed layers over a meaningful area. Such an in- 5. 857/7 Handmade body sherd; thickness 6-8 mm;
vestment of effort would be justifiable for many ext. slipped light reddish brown and painted red, parallel
reasons, but from the point of view of the ceramics it chevrons. Unit 857, square 580,1020; Early Chalcolithic I.
is essential in order to understand post-Neolithic
developments in pottery manufacture and use, par- 6. 858/22 Handmade body sherd; thickness 4 mm; ext.
ticularly relating to the increased consumption of slipped light pink and painted dark red, parallel chev-
ceramics and the origin of painted decoration in the rons. Unit 858, square 580,1020; ?Early Chalcolithic II.
Early Chalcolithic, as well as possible continuity be-
tween the occupation of the two mounds, which is 7. 860/10 Handmade body sherd; thickness 5 mm; ext.
more visible in some aspects of the pottery than in slipped orange and painted dark reddish brown, concen-
others. tric diamonds. Unit 860, square 640,960; Early
Chalcolithic.
Spatial analysis
8. 860/12 Handmade body sherd; thickness 6-7 mm;
Rooms and shrines ext. slipped very light brown and painted dark reddish
The second aspect of the re-analysis of Mellaart's brown, concentric 'stars'. Unit 860, square 640,960;
ceramic material, after an examination of typology, Early Chalcolithic.
considers differences between assemblages from dif-
ferent contexts. The majority of sherds from each 9. 856/3 Handmade body sherd, probably of necked
level, with the exception of Vlb, are assigned to a jar; thickness 5-7 mm; ext. burnished dark brown with
particular space, either an enclosed room or an open grooved concentric curvilinear lines; int. dark brown;
area. In combination with the plans and descriptions fabric orange with grey core. Unit 856, square 480,1040;
published in the preliminary reports it is possible to ?Chalcolithic.

162
Surface Pottery

2
1

3
I• "lnm'inm
.! l~Jm.L:.:&
.:=:··:.:···:::.··.·..··:···· .• . :.' · • .

6
5

I 7
9

Figure 9.24. Cha/co/ithic painted and incised decoration from (.atalhoyiik West. Scale 1:2.

163
Jonathan Last

of this type can be maintained. Nor is there any from different contexts in many levels may reflect
reason to presume a priori the separation of secular genuine differences in the use or meaning of these
and sacred in the Neolithic; that is probably an anach- rooms.
ronism. With these labels jettisoned we are left with
a series of formally similar architectural spaces which Levels XII-VIII
show a continuum of elaboration from the simple I will begin with the earliest layers. From levels XII
and relatively plain to the extremely decorative and up to VIII excavation was restricted to a small area,
symbolically complex (Figs. 1.3-1.7). Whether along investigated by the sounding dug in 1965. These
this continuum we can recognize qualitatively dif- rooms are generally less elaborate than those of the
ferent functions, or temporal stages in the life-cycles upper levels and Mellaart (1966) also suggests that
of structures depends upon a better contextual un- preservation was not as good lower down in the
derstanding and comparison of the room contents, mound. In level XII (Fig. 1.17) the majority of sherds
both the parietal and structural decorations and the come from the large room 29 but there are also some
assemblages of portable artefacts. Mellaart' s nomen-
clature of 'courts' for the open areas should also be Figure 9.25
questioned. While these probably do represent 1. 860/18 Handmade jar rim; diam. c. 400 mm,
unroofed spaces between structures, they should not thickness 11 mm; ext. and int. streaky red wash (int.
be assumed to be areas of activity and circulation; encrusted); fabric pinky-orange, moderate small mineral,
many seem to have been used primarily for discard principally white, calcareous grits. Unit 860, square
of refuse and lack evidence for surfaces. 640,960; Early Chalcolithic.

The nature of the evidence 2. 860/20 Handmade jar rim with remains of lug;
The aim of this analysis is to take the first step to- diam. c. 280 mm, thickness 8-9 mm; ext. burnished pale
wards a reconsideration of room use at <;atalhoyiik. brown; int. smoothed orangey-brown; fabric dark grey,
However, there are many problems with this sort of moderate small mineral, principally white, calcareous
contextual study of the areas excavated in the 1960s. grits. Unit 860, square 640,960; Early Chalcolithic.
Particularly noticeable is the fact that for most rooms
there are very few finds, and for most levels only a 3. 860/19 Handmade everted jar rim; diam. c. 300
very few of the rooms have any sherds preserved at mm, thickness 6-9 mm; ext. and int. slipped pale orange-
all. It is unclear whether this reflects a genuine ab- brown, streaky red wash around rim; fabric pink,
sence of ceramic material from the majority of the moderate small mineral, principally white, calcareous
site or merely the excavation techniques employed, grits. Unit 860, square 640,960; Early Chalcolithic.
but Mellaart (1962) mentions an average of about six
pots per house from the lower levels and more later, 4. 858/26 Wheelmade jar rim, everted; diam. c. 200
which suggests that the ceramic assemblage dis- mm, thickness 8-10 mm; ext. slipped orange and painted
cussed here is not fully representative. Collection dull purple-brown, short vertical bands on rim, horizon-
strategies and recovery rates may be presumed not tal band above wavy line(?) on shoulder; int. smoothed,
to have been very intensive, not least given the sheer orange-brown; fabric red, frequent med./large mineral
number of rooms excavated in only four seasons by inc. many white, calcareous grits. Unit 858, square
a minimal team of trained archaeologists. In addi- 580,1020; Roman/Byzantine.
tion the location of finds within the rooms is not
recorded, nor even whether they derive from fill or 5. 858/25 Wheelmade jar rim, externally thickened;
floor contexts. This question is clearly relevant to diam. 130 mm, thickness 6-7 mm; ext. and int. smoothed
how far we may interpret an assemblage as reflect- dark brown to dark grey brown; fabric grey, moderate
ing the original function of that room. In only two med.flarge mineral inc. mica. Unit 858, square
cases (VII.44 and 45) are some sherds marked 'fill'; 580,1020; Roman/Byzantine.
but does this mean that the majority of the finds,
which are not so labelled, are all from floors? In 6. 860/15 Wheelmade bowl rim, internally thickened;
addition, many of the rooms described in detail in diam. 280 mm, thickness 5-6 mm; ext. smoothed dark
the reports have no finds, while some of the contexts brown to dark greyish brown; int. smoothed mid-brown
with sherds are given only cursory descriptions or with dark brown mottles; fabric reddish brown, moderate
cannot even be located on plan. Nevertheless, de- small mineral inc. mica. Unit 860, square 640,960,
spite these caveats, variation between assemblages Roman/Byzantine.

164
Surface Pottery

{ - -~ ) 6

Figure 9.25. Handmade and whee/made forms fro m (:a talhoyiik Wes t.

165
Jonathan Last

from 25. The latter has only open rims while 29 has has somewhat smaller sherds throughout levels XII
closed forms as well; room 29 has rounded base to IX. This may represent a more trampled assem-
junctions, 25 has more angular forms. Hence there is blage here and implies an ongoing focus of activity
some indication that differences in room assemblages in this space.
can be traced back to the earliest excavated level -
without having any sense here of the meaning of this XI.29- 27 sherds: bowl/holemouth rims
difference. (rounded, flat, pointed); T4/5 (rounded)
bases; 2 potdiscs.
Ceramic finds by room: X.1- 13 sherds, including 8 from 1 vessel (form
XII.25- 21 sherds: holemouth / bowl rims (flat and unknown).
rounded); T4 (rounded) bases; 1 potdisc. X.25- 9 sherds: bowl rims.
XII.29- 12 sherds: bowl rims (flat); T1 (angular) X.28- 22 sherds: mainly bowl rims (flat,
bases. rounded); T4 bases; 1 potdisc, mean size
66.9 mm.
In the succeeding level XI (Fig. 1.17) the western part X.29- 11 sherds: holemouth/bowl rims (flat,
of the area is occupied by an open 'court' but all the rounded); Tl/2 (angular) bases; 1 potdisc,
finds come from room 29. Deposition in the open mean size 55.5 mm.
areas, of ceramics at least, does not appear to begin
until level X. While there is a single bowl rim from In level IX (Fig. 1.17) there are no large assemblages:
the earliest so-called 'shrine', X.1 (see level VIII plan), finds come from rooms 1, 8, 29, 29A and 31 as well as
most of the finds come from the area of rooms 29, 25 open area 25/28. The first two of these contain only
and 28, described as 'one large courtyard' (Mellaart dark-faced bowl rims; the remaining areas have
1966). Here the main distinction is between the west lighter wares, while holemouths are restricted to
part of the 'court', contexts 25 and 28, which have 'court' 25 and room 29A. The most complete vessel,
mainly open forms with rounded bases and large about one third of a bowl, comes from the 'large
body sherds (mean size 66.2 mm), and context 29 to storeroom' 31. But on what basis is it a storeroom? It
the east which has fewer and smaller sherds (55.5 appears considerably more complex than room 29.
mm) as well as more holemouth forms and angular And is this open form a storage vessel? The closed
bases. Hence there are evident functional or de- vessel rims occur in extra-mural areas and a possible
positional differences within the open area; and it is annex, not in the larger rooms. The distinct (though
noticeable that 29 (whether a roofed or an open space) tiny) dark-faced assemblage in rooms 1 and 8 may
be significant, because
these are called shrines by
Mellaart, both containing
bull' s- head decorations.
Room 8 also contained a
unique lug.

IX.1- 2 sherds: 1 bowl


rim.
IX.8 - 3 sherds: bowl
rims.
IX.25- 6 sherds: mainly
holemouth rims.
IX.28- 3 sherds: 1 bowl
rim.
IX.29- 7 sherds: bowl
rims; T1 base.
IX.29A- 4 sherds: 1
holemouth rim;
T1 base.
IX.31- 3 sherds: bowl
Figure 9.26. Neolithic pottery from c;atalhoyiik East. rims.

166
Surface Pottery

ln level VIII (Fig. 1.16) finds come from a number of coloured bodies, mean 40.0 mm.
buildings in the area of the 1965 sounding, discussed VIII.29- 16 sherds: mainly bowl rims (all types); T4
in some detail by Mellaart (1966). In addition there bases; 2 potdiscs, darker bodies, mean
are a few body sherds from room 45 (a 'shrine' with 45.3 mm.
animal reliefs) on the other side of the site (see level VIII.31 - 6 sherds: holemouth rims.
VII plan). For the southern area there are two clear
groups on the complexity graph: the less elaborate Levels VII-IV
rooms (3, 12, 18 & 24) generally lack potsherds al- In level VII (Fig. 1.15) a much larger excavated area
though the largest assemblage (mainly bowl rims is available, but there are few sherds. Pottery comes
and angular bases) comes from VIII.28, which is prob- from 'court' 15 and rooms 25, 43, 44, 45 and 71,
ably the room labelled on the plan as 24. With the which do not form any sort of cluster on the com-
exception of 'shrines' 1 and 8 (the former missing plexity scale. Room 45, the 'boar's head shrine', had
from the graph) the more elaborate rooms all pro- only two small body sherds from the building fill,
duced sherds, although there are some noticeable while room 44, the 'leopard shrine', contained the
differences: bowl rims and rounded bases come from smashed remains of at least one dark burnished ves-
the two least complex, 10 and 29 (called a 'meeting sel- which tallies with Mellaart's report that two
room', because of the long bench); dark holemouths dark cooking pots were found there. There is no
from 14 (with textile and net patterns on the walls) significant difference between the assemblages from
and 27 (also with net or textile motifs as well as a 43 and 44, which from their room numbers may well
leopard relief); light-coloured holemouths from 31 be related (although the precise location of 43 is
(the 'red shrine'). The most elaborate contexts have uncertain): a majority of these vessels are bowls while
primarily closed vessel forms, therefore, while the the three rim sherds from 25 and 71 are of
less elaborate rooms which have finds (10, 28 & 29) holemouths. In the absence of information on con-
contain mainly open-mouthed pots. The complex text 43 few conclusions can be drawn; VII.25 con-
group as a whole has slightly larger (45.8 : 42.8 mm) tains more closed forms than 44, which would reverse
but thinner (7.7: 9.0 mm) body sherds than the other the level VIII pattern relating to complexity, but the
rooms, perhaps indicating less trampled assemblages. latter assemblage derives mainly from the building
Room 27 is also differentiated by a number of fill. The open area yielded only a single body sherd,
fineware sherds without organic temper. Hence there but excavation in 1995 in the area of 'court' VII.15
is some indication of patterned differences between does not indicate an absence of potsherds in such
assemblages from the less and more elaborate contexts during this phase.
buildings, in terms of
both vessel types and
formation processes. Stor-
age and restricted access
seem to characterize the
most complex rooms (all
'shrines') in this level;
bowls come from 'moder-
ately complex' contexts;
while the non-complex
rooms lack ceramics.

VIII.2 - 3 sherds.
VULlO- 7 sherds: bowl
rims.
VIII.14- 1 sherd .
VIII.25 - 3 sherds.
VIII.27- 7 sherds:
holemouth rims.
VIII.28 - 28 sherds: mainly
bowl rims (flat);
T1 bases; light- Figure 9.27. Cha/colithic pottery from c;.atalhoyiik West.

167
Jonathan Last

VII.15 - 1 sherd. E.VI.71- 56 sherds: mainly holemouth rims (flat,


VII.25 - 2 sherds: holemouth rims. rounded); 1 T2 (angular) base; 3 lugs; dark
VII.43- 27 sherds: mainly bowl rims (all types); T4 bodies, mean 45.5 mm.
bases; dark bodies. E.VI.72- 8 sherds: mainly holemouth rims; 2lugs.
VII.44 - 6 sherds: 1 lug. E.VI.73- 10 sherds: holemouth and bowl rims; 1
VII.45 - 2 sherds. lug; dark bodies, mean 37.6 mm.
E.VI.74- 17 sherds (+74 of single vessel): mainly
In level VIB (Fig. 1.14) only seven rim sherds are holemouth rims (most flattened); 2lugs;
located to particular rooms. 'Storeroom' 32 had a dark bodies, mean 42.6 mm (not including
single bowl rim: this is considered to be the store of smashed vessel).
'shrine' 70 (not plotted on the complexity graph), E.VI.75- 10 sherds (+22 probably of same vessel):
which also had one bowl rim. It was this room which holemouth rims (flat, rounded); 1 T4 base;
yielded the single polychrome painted sherd dis- dark bodies, mean 44.3 mm (not including
cussed by Mellaart (1966) and now displayed in An- smashed vessel).
kara. Room 49, on the other hand, adjacent to 'shrine' F.VI.l- 10 sherds: holemouth rims (rounded, flat);
50 (with kilim paintings), has more closed forms. 1 T4 base; 4lugs, 1 basket handle.
Little else can be said about this phase since the
remaining finds have no context, although we may In level V (Fig. 1.12) the house complexity graphs
reasonably assume most of them derive from the become relevant again: there is a fourfold division of
1965 sounding. Red wares are more common among contexts in Mellaart's terms, between 'shrines' (E.3,
these unlabelled finds, which may be indicative of E.4 and F.1, which had a large painting of a red bull;
some broader spatial differences between different 28 rims mainly in F.1), 'houses' (E.7,75; F.3-5, 7, 8
areas of the site. [probably]; 36 rims), 'courts' (E, F, F.2, 5c, 6; 64 rims)
and one 'storeroom' (E.3 store; 1 rim). The first group
VIB.32 - 1 sherd. includes the three most elaborate rooms in this level
VIB.49 - 6 sherds. while the three 'houses' on the graph can be de-
VIB.SO - 1 sherd. scribed as complex (F.7), moderately complex (E.75)
or non-complex (E.7). In fact F.V.7 is more elaborate
In level VIA (Fig. 1.13) finds come from E.71-75 (in than any other 'non-shrine' in this level. While many
the southern part of the excavated area for which no of the remaining contexts cannot be located on plan
plan was published but forming part of the 'tenta- the majority of the finds clearly come from midden
tive reconstruction' in the 1965 report (Mellaart 1966, deposits in the open areas.
175)), classed as 'houses' without special decoration, There is little overall variation in the propor-
and F.1, which may be the context written up as tion of different vessel forms, with the open:closed
'shrine' E.VI.80. All three rims from this building are ratio remarkably consistent. The elaborate rooms lack
of holemouths, and lugs are particularly common; distinctive pottery: bowls are most common in the
elsewhere only the largest house, E.75 (4 rims), lacks small assemblages of E.3 and E.4, and frequent in
open forms. This is similar to the pattern observed in F.7, but F.1 has the lowest proportion of open ves-
level VIII, where the more elaborate rooms contained sels. Lugs are common in the 'houses' but less so in
more closed forms. One of the pots in F.1 comprises the 'courts' and F.l. The storeroom contained only a
six sherds; about half of the rim is preserved. Large single bowl rim (like IX.31 and E.VIb.32). There are
portions of individual vessels were also found in 71 no other consistent differences, although rarer fea-
and particularly 74, which is considered to be the tures suggest possible locations of specialized vessel
'storeroom' of 75. These rooms also have larger body types: two of the three lugs with double perforations
sherds on average, which may indicate a difference were found in F.1, and two of the three basket han-
between 'primary' and 'secondary' refuse assem- dles in the nearby room E.7. An animal head lug
blages in different contexts - although nothing is came from F.7. The body sherds show some differ-
known of the exact provenance of these finds within ences too, with more dark wares in F.S and F.6 than
the rooms. There is also some evidence that dark- in F.1-4. In terms of preservation, the F.S courtyard
coloured vessels are more common in 74 and 75 than has the largest sherds (61.7 mm average); F.1, 6 and 8
in the other rooms. have relatively well-preserved sherds (54.1-54.8 mm);
while the other spaces and rooms, especially F.3 and
E.VI.70- 3 sherds. F.S (43.2 and 45.7 mm), have smaller sherds. This

168
Surface Pottery

may reflect specific patterns of activity or discard in holemouths. Reflecting this distinction, nearly all the
this region (without location plans it is impossible to lugs come from the 'court', and while E.4 and E.S
say more) but does not reveal general differences have mainly light coloured and red vessels, the sherds
between Mellaart's structural categories. from E.11-12 and the open area are predominantly
dark wares. Footed bases are absent from the 'court-
E.V.3 - 2 sherds: bowl rims. yard' and E.11-12 but more common in E.4 and E.S.
E.V.3.St -1 sherd: bowl rim. Two rims with rare incised decoration come from
E.V.4- 1 sherd: bowl rim. E.5. Hence the assemblage from E.4 more closely re-
E.V.7- 5 sherds: T4 bases. sembles that from the least elaborate but nearby room
E.V.75- 2 sherds: holemouth rims. E.S than the more distant E.11-12. Decorated pottery is
E.V.Ct- 59 sherds: holemouth/bowl rims (most not associated with the more complex architecture.
flat); T1 bases; 3 lugs; mean 51.0 mm. Preservation appears to be less good in E.4
F.V.1 - 55 sherds: mainly holemouth rims (mean VE of 8 per cent) than in the other rooms (13
(rounded, flat), 1 decorated bowl; T1 per cent) but there are no body sherds from the
bases; 9 lugs; mean 54.1 mm. 'shrine'. The finds from area F do include body
F.V.2- 54 sherds: mainly holemouth rims sherds: among these the open area has a higher pro-
(rounded, flat); T1 bases; 4lugs, 1 basket portion of dark coloured vessels than room F.6, and
handle; mean 50.3 mm. the sherds from the latter are generally smaller (41.7
F.V.3 - 30 sherds: mainly holemouth rims (flat, mm compared to 49.8 mm), apparently reflecting
rounded); T1 and T4 bases; 5 lugs; mean rather different formation processes. Given the rela-
45.7mm. tive quantities of material, it seems to indicate that
F.V.4- 1 sherd. dumping of 'fresh' (less trampled) refuse was con-
F.V.5- 24 sherds: 1 holemouth rim; T4 bases; centrated in the open area.
dark bodies, mean 43.2 mm.
F.V.S.Ct -36 sherds: mainly holemouth rims E.IV.4- 12 sherds: mainly bowl rims (most
(rounded, flat); T1 bases; 3 lugs; dark flat); T1 and T3 (footed) bases; no lugs.
bodies, mean 61.7 mm. E.IV.5 - 10 sherds: mainly bowl rims (all types),
F.V.6- 19 sherds: mainly holemouth rims; T1 2 decorated bowls; mainly T3 bases; no
bases; no lugs; dark bodies, mean 54.7 lugs.
mm. E.IV.11-12- 21 sherds: mainly holemouth rims (flat,
F.V.7- 27 sherds: mainly bowl rims (flat); T1 rounded); T1 and T4 bases; 2 lugs;
bases; 4 lugs including 1 animal head; mean size 70.7 mm.
mean 50.5 mm. F.IV.6- 31 sherds: mainly bowl rims (flat,
F.V.8- 86 sherds: mainly holemouth rims pointed); mainly T1 bases; 1lug; mean
(rounded, flat); T1 and T4 bases; 7lugs; 41.7 mm.
mean 54.8 mm. F.IV.Ct- 181 sherds: mainly holemouth rims
F.V.Ct- 77 sherds: mainly holemouth rims (flat, (flat, rounded); mainly T1 bases; 16
rounded); mainly T1 bases; 1lug; 1 lugs; mean 49.8 mm.
potdisc, mean 51.5 mm.
Levels III and II
In level IV (Fig. 1.11) a three-way division can be In level III (Fig. 1.10) the excavated area is again
made between a 'shrine' (E.4; 10 rims), four 'houses' more restricted and the assemblage is much smaller
(E.5, 11-12, F.6; 30 rims) and a 'courtyard' in area F (20 rims with contextual information). No body sherds
(60 rims). Cross-cutting Mellaart's classification, E.4 were kept. Here the finds derive from two elaborate
and E.11 both show a relatively high level of com- rooms ('shrine' 1, with wall-paintings of a deer hunt,
plexity, while E.S and E.12 are less elaborate. Unfor- and adjacent room 7), three of moderate complexity
tunately the finds from contexts 11 and 12 were not (2, 4 & 13), and one lacking elaboration (room 12).
separated. E.4 has most open vessel forms (80 per While not recorded on the graph, the reconstruction
cent), followed by E.S (70 per cent) and F.6 (75 per drawing indicates that room 4 also contained a 'bull
cent). The latter context is presumed to be a room pillar' (Mellaart 1963, 47). Room 1 yielded a very
although it was not located on a plan (however, the small assemblage (4 rims) with open and closed
underlying F.V.6 is an open area). F.IV.Ct (69 per forms, but 7 and 4 contained predominantly bowls
cent) and rooms E.11-12 (67 per cent) have most while holemouths are more common in the less

169
Jonathan Last

complex rooms 2 and 13 (12lacked rims). Four of the complexity, the nature of that difference is itself vari-
five preserved lugs (which are always more com- able. In levels VIII and VI the most complex rooms
mon on holemouths) come from room 13. The less may have more closed forms, while in levels IV, III
elaborate contexts also contain mainly dark wares and II they have more open vessels. However, it is
while 7 has mottled colours and 1, as with 'shrine' notable that levels VIII and VI lie within a phase
E.IV.4 in the level below, has more light-coloured when holemouth forms are increasing in popularity,
wares. Room 4, in a separate block of houses, has but in levels IV to II bowls are becoming more com-
rather smaller and thinner vessels than the other mon. The 'shrine' assemblages therefore appear more
rooms. 'progressive'. An association of elaborate rooms with
innovative or unusual ceramic forms may also be
A.III.1 - 5 sherds. seen in levels IX (rare dark colours; early lug), VIII
A.III.2 - 3 sherds. (fineware sherds), VII (early lug), VIB (polychrome
A.III.4 - 4 sherds. sherd), VIA (lugs and handles), V (double lugs, ani-
A.III.7- 6 sherds: bowl rims. mal head) and II (miniature vessel). Similarly, light-
A.III.12 - 1 sherd. coloured wares increase proportionately from level
A.III.13 - 12 sherds: bowl and holemouth rims; 4lugs. V up and are more common in 'shrines' than in other
B.III.2 - 2 sherds: 1 lug. rooms during these phases. But the pattern should
not be overstated, as the occurrence of unusual forms
Finally in level II (Fig. 1.9), where wall-paintings in a non-elaborate room in level IV shows. These less
were not apparent and the ceramic assemblage is elaborate spaces themselves show a great deal of
rather different, finds come from the three most elabo- variation. 'Storerooms', for instance, generally pro-
rate rooms: A.1 (termed a shrine because of a bull vided few finds: in levels II and III these are pre-
pillar, elaborate hearth and a group of figurines), B.1 dominantly holemouths, but in the earlier levels open
and B.2, as well as their associated 'storerooms'. The forms are found. The pottery does not support un-
largest assemblage is that from B.2. Some differences equivocally this functional label.
are apparent, although there are very few sherds Variation in sherd size, which may indicate dif-
from A.1 (4 rims) which had entirely open vessel fering patterns of access and movement in contexts
forms. Holemouths are more common in the store- of different type, is less clear: in level VIII the com-
rooms (44 per cent) which fits the functional inter- plex rooms contain relatively large sherds, while
pretation (but which is not the case for equivalent those from E.IV.4 are apparently rather smaller than
rooms in the lower levels). On the other hand, the the average for the level. In levels V and IV, which
storerooms have smaller pots than the larger rooms. have the largest assemblages, the open areas contain
All spaces have a variety of surface colours although most of the ceramic refuse. It is likely that discard (of
the 'shrine' again lacks dark wares. A.1 also includes hearth and oven residues as well as pottery and
a miniature vessel and seems to have rather larger other artefacts) was concentrated in such 'dead'
sherds, although again body fragments are missing. spaces throughout the sequence at ~atalhoyiik, al-
though it appears from the preserved assemblages
A.II.1 - 8 sherds: bowl rims. that they were consistently used for deposition of
B.II.l - 18 sherds: mainly bowl rims (flattened, potsherds only in levels X, IX, V and IV. They may
rounded); all base types. also have distinctive assemblages: for instance, the
B.II.1.St - 4 sherds. predominance of holemouths in the levels IX and IV
B.ll.2 - 38 sherds: mainly bowl rims (rounded, 'courts'. However, variation between the open areas
flattened); mainly T1 bases. in level V, and within different areas of the level X
B.II.2.St - 6 sherds. 'courtyard' point to significant small-scale patterning
in depositional practice across these contexts.
Conclusions Variation between different building levels is
While there are sufficient differences between rooms marked but continuity within a single structure may
to suggest that variability in room use and formation occasionally be seen: from level XII to VIII the focus
processes can be discerned in the excavated material, for deposition of ceramics within the excavated
there is no clear distinction between assemblages area is clearly room 29, which is also marked by the
from more and less elaborate rooms that is valid for consistent occurrence of potdiscs in levels XII, XI, X
all phases. While in many levels it does seem that (when part of a larger open area) and VIII, while the
assemblage composition varies with architectural only other potdisc from these levels comes from

170
Surface Pottery

'courtyard' X.28. Issues of change and continuity, as preliminary report. Anatolian Studies XL, 205-18.
well as considerations of how discarded material Last, J., n.d. <::atalhoyiik Investigations 1995 - pottery
may relate to patterns of activity, can be further report. Unpublished archive report.
explored only within a better understanding of proc- Mellaart, J., 1962. Excavations at <::atal Hiiyiik, first pre-
liminary report, 1961. Anatolian Studies 12,41-66.
esses of abandonment or 'decommissioning' of
Mellaart, J., 1963. Excavations at <::atal Hiiyiik, second pre-
rooms. Without this information the patterns out- liminary report, 1962. Anatolian Studies 13,43-103.
lined here are difficult to explain in terms of social Mellaart, J., 1964. Excavations at <::atal Hiiyiik, third pre-
practices of ceramic use and discard, which can be liminary report, 1963. Anatolian Studies 14,39-119.
further elucidated only on the basis of more detailed Mellaart, J., 1965. <::atal Hiiyiik West. Anatolian Studies 15,
contextual recording. It may then be possible to de- 135-56.
velop a deeper understanding of the degree of Mellaart, J., 1966. Excavations at <::atal Hiiyiik, fourth pre-
routinization of activities within the context of a liminary report, 1965. Anatolian Studies 16,43-103.
domestic space that, whatever labels we apply to dif- Mellaart, J., 1970. Hacllar. Edinburgh: Edinburgh Univer-
sity Press.
ferent rooms, was clearly structured by ritual practice.
Moore. A.M.T., 1995. The inception of potting in western
Asia and its impact on economy and society, in The
Acknowledgements Emergence of Pottery: Technology and Innovation in An-
cient Societies, eds. W.K. Barnett & J.W. Hooper. Wash-
I am grateful to all on the C::atalhoytik team for so ington (DC): Smithsonian Institution Press, 39-53.
many fruitful discussions; to staff at the Konya mu- Rosen, A., 1986. Cities of Clay: the Geoarchaeology of Tells.
seum for allowing me access to the 1960s material; to Chicago (IL): University of Chicago Press.
Neil Roberts for his comments on the second section Stucki, W., 1984. Unterlagen zur Keramik des A/ten Vorderen
of this text; and to Tom Pollard for labouring so Orients 2. Zurich: E.A. Verlag.
Waelkens, M., E. Owens, A. Hasendonckx & B. Arikan,
diligently on the figures. 1992. The excavations at Sagalassos 1991. Anatolian
Studies 42, 79-98.
References Williams, C., 1989. Anemurium: the Roman and Early Byzan-
tine Pottery. Toronto: Pontifical Institute of Medi-
Harrison, R., 1990. Amorium excavations 1989: second eval Studies.

171
Chapter 10

The Knapped-Stone

James Conolly

Methodology recognition and interpretation. As such, the ideas


that shaped their production and use would have
Introduction stretched beyond problems of procedure to encom-
This section of the <::atalhoyiik lithic report presents pass concepts of the self and society.
some initial ideas for a contextual approach that In other words, the tools and waste that form the
emphasizes the interpretation of typological and tech- objects of our analyses have meaning beyond their
nological data. Although many of the objectives material boundaries: their form and patterning may
raised here cannot be met in this report because the reflect social abstractions such as domestication, gen-
type of contextual information required is absent, an der, death, ascetics and so on. The claim that a tech-
attempt will be made to apply the analytical ap- nological reading of knapped-stone industries can
proach described below to the forthcoming exca- be made independent of archaeological context
vated data. (Inizan et al. 1992, 13) may be to some degree correct,
but little information concerning socio-economic or
Towards a contextual methodology symbolic meaning can be otherwise gained. So one
Traditionally, analyses of knapped-stone assemblages possible way to discover and explore the social con-
have been approached from three perspectives: (i) text of technical actions and products is to place
typological studies, which attempt to establish mor- context as a peremptory and defining force in the
phologically defined types of tools, or interpret the methodological, analytical and interpretative proc-
results of typological classification; (ii) technological ess. A step towards this could involve questioning
studies which deal with manufacturing methods and the conventional categorization of knapped-stone
techniques and; (iii) functional studies, which in- data, using 'context' to challenge our assumptions. I
volve the study of how stone tools were used shall illustrate with three examples.
(Nishiaki 1992, 39). In isolation none of these lend The distinction between blades and bladelets is
themselves particularly well to a study of context. often defined arbitrarily. However, if the categories
However, technological analyses that follow the 'blade' and 'bladelet' are to have any real social or
methodology offered by the chafne operatoire could in technical meaning, they need to be contextually
a broad sense be called contextual, rooted as it is in grounded in some manner. Tixier's advice not to use
the French school of cultural technology, which rests his metric division of blades and bladelets devised
on the premise that technical actions are social and solely for the Epipalaeolithic of the Maghreb, but to
cultural phenomena (Inizan et al. 1992, 12; Pelegrin use context-specific criteria based on production tech-
1990). The significance of this has been succinctly niques (Tixier 1974, 7; Inizan et al. 1992, 59) is a good
outlined by Edmonds (1995, 11): example of this. The social context of the production
and use of large and blades may also suggest alter-
Like the routines that shape most of our lives to-
native classifications: functional differences between
day, the tasks undertaken by people in the past
would have provided a frame through which they
blade sizes (such as large blades produced as blanks
might have come to recognize aspects of their world for non-hafted 'stand-alone' tools and small
and themselves. The tools and waste that we re- 'bladelets' for microlithic composite tools), or social
cover would have been entangled in that world of differences (such as large blades intended for use
social practice, providing cues for this process of primarily by women, bladelets primarily by men)

173
James Conolly

are two additional examples where 'context' may In all of the above, I have ignored one impor-
aid in the definition of categories of data and make it tant aspect of the use of traditional type-categories:
less arbitrary. the value of commonly agreed terms for the sharing
The categories 'blade' and 'flake' may also be of information about the characteristics of any par-
contingent on context. At a general level, experi- ticular assemblage. In some instances this is a moot
mental research has demonstrated that blade debitage point; describing a lithic assemblage as possessing a
is a structured process requiring considerable pre- particular percentage of bladelets to blades when
planning enabling the mass production of blanks there is no reason for dividing the debitage into
from a single core. It has also been show that in blade and bladelets serves no useful purpose when
certain instances such production has social, politi- the size characteristics of all blades can be described
cal and economic associations (e.g. Clark 1987). Yet by a histogram of lengths. In other respects it is
convention dictates that any flake that has a 2:1length useful to know how many Byblos points are in an
to width ratio is called a blade (lnizan et al. 1992, 58). assemblage, or Naviform cores, or sickle-blades. Yet
Given the complexity and significant of true blade is should be recognized that the interpretative value
debitage, this cannot be a effective means of charac- of any lithic classification comes from its relation-
terization. A method of analysis based on technical ship to both a technological and social context.
and productive differences avoids such problems,
and context may also be a valuable clue to identify- Practical application
ing knapping techniques. Contextually-aware analy- Putting theory into practice is more problematic, and
ses of blades in the Aegean Early Bronze Age has, there are severe limits to what can be done with
for instance, shown that the products of some reduc- surface derived artefacts. The main application of
tion techniques only occur in particular contexts and the above ideas must therefore await the excavation
appear to have considerable social prestige and mean- phase of the renewed <;atalhOyiik project. Yet I did
ing (Carter 1994). want to begin a critical consideration of the catego-
The categorization of tools provides a final ex- ries being used and pinpoint areas where more in-
ample. Unmodified edges are often functional, so depth analysis would be profitable. The discussion
the term 'tool' itself is not particularly useful if, as is below reflects this to some degree.
the custom, the presence of retouch is used as the Fieldwork in 1993, 1994 and 1995 resulted in
defining characteristic. Clarifying this by using the two discrete samples of material obtained by differ-
term 'retouched piece' does not entirely solve the ent collection procedures. The first sample is de-
problem: Gero (1991) has questioned the distinction rived from a systematic 2 x 2 surface collection, with
made between intentionally retouched and non-re- all material being sieved through a 5 mm mesh (see
touched but utilized tools, arguing that this implic- Chapter 6). The second sample is derived from top-
itly biases our interpretations of activities towards scraping activities where all visible artefacts were
those performed by men with their assumed prefer- collected (Chapter 7). As the 2 x 2 data consisted
ence for highly modified tools (such as projectiles, primarily of small samples of non-diagnostic micro-
axes, etc.). Even so, the issue extends beyond this: debris and shatter it was recorded in less detail than
objects are made, used and discarded in a variety of the many times larger (although unsieved) top-scrape
different contexts, effectively furnishing them with a samples. Accordingly, these results place more em-
variety of different meanings. Fixed categories, such phasis on the latter.
as 'arrowheads', 'scrapers', 'blades', and so on, do
not necessarily reflect this fluidity. For instance, the Technological analysis
smallest pieces of debitage - chips and shatter - A technological analysis of all lithic artefacts was
conventionally considered as the waste by-products conducted. This involved classification by raw mate-
of core reduction or tool manufacture may be a re- rial in order provide a breakdown of the relative
quired component for the temper of mud bricks or contribution of obsidian, flint and other siliceous
mortar. Similarly, projectile points are not purely rocks to the total industry. Basic metric characteris-
functional objects used for killing animals - explor- tics (length, width, thickness) were recorded. Allar-
ing the context of their use and deposition opens up tefacts (including those that exhibited retouch or
a world of symbolic associations where their use and utilisation characteristics) were placed into one of a
meaning becomes as much ritual as functional. number of debitage categories defined by a combi-
Context becomes central to both analysis and in- nation of morphological and technological charac-
terpretation. teristics reflecting different manufacturing processes

174
Knapped-Stone Report

and/ or stages. Qualitative attributes such as scar sample. In the case of the surface sample, the tools
patterns, edge morphology and profile formed the were classified by traditional type-categories similar
basis for categorization, and categories were con- to those used previously at <::atalhoyiik and Can
tinually assessed in order to justify their existence. Hasan III (Bialor 1962; Mellaart 1964; Ataman 1989).
The term flake was used to describe debitage I found quite quickly, however, that this did not
which exhibits either a clear ventral surface, rem- work particularly well as the majority of the re-
nant striking platform, or bulb of percussion. No touched pieces could not be described by a simplis-
standardized morphology is implied by the use of tic term such as 'flake scraper' or the like. While such
the term flake and pieces of debitage which are lack- an approach provided frequencies for a type list of
ing the proximal end but can still be confidently tools that was broadly comparable to other prehis-
identified as flakes, were termed flake fragments. toric sites in the region, the system was ill equipped
The term broken flake was used to describe flakes to deal with the morphological variability of the re-
with intact proximal ends, but incomplete margins. touched tools. It was very difficult to keep the type
The term regular blade was used to refer to pieces of categories mutually exclusive, particularly when
the <::atalhOyiik debitage which exhibited regular, dealing with tools with more than one modified edge
parallel, and almost exclusively unidirectional dor- and assumed function - such as a flake with a
sal scars, and generally had parallel lateral margins. scraper-like edge and a piercing edge. Most impor-
If the blade in question was a fragment lacking ei- tant, the approach either ignored or attempted to
ther a proximal or distal end, yet still exhibits the compress the most potentially interesting aspect of
characteristics which define a regular blade, then the the non-standardised tool assemblage - the varia-
term regular blade fragment was used. Similarly, tion in tool morphology and retouch characteristics.
regular proximal blade fragment describes those In order to obtain more extensive information con-
blade fragments where only the proximal end was cerning the this variation an attribute analysis was
present, and regular distal blade fragment those conducted on a sample of retouched pieces from the
pieces with only the distal end present. If a piece of 1993 and the total sample of the 1994 and 1995
debitage deviated in some way from the defining scraped areas. The variables were adopted from
characteristics of a regular blade, such as having an Inizan et al. (1992) and comprise the description of
irregular scar pattern instead of a unidirectional pat- various retouch attributes such as location, position,
tern, or irregular lateral margins, but is still morpho- angle, etc .. Recently, Baird et al. (1995) have pub-
logically more akin to the blade group than the flake lished a comparable analytical method (The
group, it is referred to as a non-regular blade. Por- Wembach Module). 1 When at all possible, these at-
tions thereof follow the same definitions for regular tributes were recorded on all tools, including the
blades. A plunging blade or plunging blade frag- more formalized bifaces/projectiles. In this manner
ment is a blade that 'plunged' into the core on re- all modified edges could be described in a standard-
moval and thus exhibits the distal end (or portion ized manner. A functional classification of the edge
thereof) of the core from which it was struck. Debitage use was also made. Although there are obvious diffi-
which exhibit a crest produced by crossed dorsal culties in assigning functions on the basis of macro-
scarring are termed crested blades or crested blade scopic morphology alone, it would appear that gross
fragments. Very small (usually <1 cm2) and non- classifications (such as cutting edge, scraping edge,
diagnostic fragments of raw material which do not piercing edge) appear to be less problematic
exhibit the attributes of a flake are referred to as (Hurcombe 1992; Grace 1989). With the exception of
chips. Occasionally pieces under 1 cm2 are encoun- pieces esquillees2 and biface/projectiles,3 the classi-
tered which possess a clear ventral surface and/ or a fication of retouched pieces was therewith derived
butt are classified under terms used for flakes. Larger from their edge class(es). If a blank possessed two or
irregular angular fragments with no clear single ven- more different classes of working edges, then it was
tral surface and not cores, are referred to as shatter. recorded as a 'combination tool'. In sum, the objec-
tive was to establish a system that was flexible and
Analysis of retouched pieces permitted the exploration and identification of rela-
Those artefacts that exhibited (macroscopic) charac- tionships between edge characteristics, blank mor-
teristics consistent with use or intentional retouch phology, and suspected function. Ultimately I hoped
were then subjected to one of two possible analytical eventually to be able to explore these relationships
processes dependent on whether they derived from between different contexts of recovery (i.e. tool, edge
the 2 x 2 surface sample, or from the 10 x 10 top-scrape and blank variations between contexts). At this

175
James Conolly

point it is worth noting that the functional classifica- from the top-scrape sample on the east mound dis-
tion 'sickle blade' is absent despite the fact that it has play any evidence of cortex on their dorsal surface.
been mentioned as a tool form at <::atalhoyiik in the On the other hand, 26 pieces of flint from the same
past. Sickle gloss is absent on obsidian thus prevent- sample possess cortical surfaces, suggesting that the
ing the identification of any sickle elements and dis- organization of the acquisition and reduction of un-
tinguishing them from other blade cutting tools. modified flint and obsidian differed.
Consequently, the term 'knife/cutting tool' includes The first two tables provide a breakdown of
blade segments that were likely used as sickle ele- debitage categories by raw material for the east and
ments but cannot at this point be identified as such. west mounds for the surface and top-scrape collec-
tions. All soil from surface collection units was sieved
The analysis through a 5 mm mesh according good recovery of
smaller pieces of debitage. Differences in the pro-
Sample size and context portions of chips between the two sample assem-
The total sample of knapped-stone analyzed thus far blages suggest that smaller debitage was not well
from the surface of both the east and west mounds at recovered by visual pickup. In general, debitage from
<::atalhoyiik is 6133. Of this total amount,2603 pieces the two mounds is dominated by blade fragments,
are from 2 x 2 m surface collection squares, of which flake fragments and chips and shatter. Although ob-
1594 are from the east mound and 1009 are from the sidian dominates all debitage categories, the distri-
west mound. 3530 pieces were recovered from the bution of raw material suggests that flint was used
top-scrape samples, of which 3283 are from the east preferentially for flake production. A chi-square sta-
mound, and 247 are from the west mound. Thus, the tistic of the distribution of flint and obsidian for total
total sample from the east mound is 4877, and from blades, flakes and other debitage for both mounds
the west, 1256. Although the west mound data are shows a significant difference at the 1 per cent level.
briefly described, the following discussion is largely The largest contribution to the chi-square statistic is
concerned with the east mound, as that is where derived from the high number of flint flakes on the
current investigations are most closely focused. west mound. The differences observed could be at-
Where illustrations of specific debitage or tool types tributed to the local availability of flint material which
are provided, examples were chosen to depict the may not have been as suitable as obsidian for blade
range of forms possible within any category. debitage.

Raw material Cores and core-derived pieces


In both samples, a dark-grey to black translucent A range of different cores is present at <::atalhoyiik,
obsidian forms the vast majority of the raw material. and analysis is currently ongoing. To provide an
Of the 2603 pieces from controlled surface collection example of the variation, there are 36 cores and core-
contexts, 2439 are obsidian (93.7 per cent), and 159 derived pieces from the top-scrape sample, consist-
(6.1 per cent) are fine grained crypo-crystalline ing of 23 cores and core-fragments and 13 core tablets.
siliceous rock (chert or flint, although the term flint The core tablets are all from the east mound, to-
will be used). Five other pieces which look as if they gether with 17 of the cores and core fragments. Of
had been struck in a manner consistent with other the 17 cores from the east mound, there are 10 frag-
chipped stone cannot be confidently identified, al- ments of uni-directional single platform blade cores
though three are almost certainly basalt and two (9 on obsidian and 1 on flint), 1 opposed platform
appear to be quartzite. One large flake of clear, white, flake core (on obsidian) and 4 multi-directional amor-
glass-like obsidian found on the west mound does phous flake cores (1 on obsidian, 1 on quartzite, and
not correspond to any descriptions of known obsid- 2 on flint). Two further obsidian cores were too bat-
ian raw material in Anatolia. Similar proportions of tered to identify to any degree of certainty. Eight of
obsidian are found in the top-scrape sample where these 17 cores had been reworked or used as wedge
95.7 per cent is obsidian. There is some variation in or chisel-like tools, with corresponding crushing on
the proportion of obsidian between the east and west one or more edges. The 13 core tablets exhibit a total
mounds; 96.3 per cent of the east mound sample is of 6 faceted platforms and 3 plain platforms, with 4
obsidian, whereas only 89.7 per cent of the west indeterminable examples because of their fragmen-
mound sample is obsidian. It would appear that the tary nature. The 6 cores from the west mound con-
majority of the primary core shaping of obsidian sist of a single-platform uni-directional obsidian
was conducted off site as only 12 pieces of obsidian blade core, re-used as a chisel or wedge, a single-

176
Knapped-Stone Report

platform flint flake core, and 4 amorphous flint flake of waste materials have been found.
cores. Illustration of a small sample of cores and The obsidian blades from both the east and
core-derived products is provided in Figure 10.1. An west mounds are typically characterized by uni-di-
impression of other blade cores from Mellaart's ex- rectional scar patterns, straight and parallel ridges
cavations can be gained from the original report and lateral edges and trapezoidal cross-sections (Fig.
(Bialor 1962, 74), of which only five complete cores 10.2). Much less common are blades with triangular
(four from level V and one from level VIII) were cross-sections and sub-parallel lateral margins. The
reported after the first season. The cores from level V frequency of the different butts seen on both blades
are typical prismatic blade cores, and the fragment and flakes from the east mound top-scrape sample
from level VIII is probably also a fragment of blade demonstrates that the blades are characterized over-
core. whelmingly by small punctiform or linear butts with
There are only four crested-blade fragments in faceted or ground dorsal lips and small bulbar scars
the two samples, all exhibiting the characteristic (Table 10.3). Neither the punctiform nor linear butts
crossed scarring associated with this form of debitage. are isolated from the blade body. These characteris-
There are two plunging blades in the east mound tics combined with the regularity of the obsidian
sample, both from uni-directional blade cores. blades provisionally suggest their removal was from
pressure, as opposed to percussion or punch, tech-
Debitage niques (cf. Tixier 1984,58,66; Wilke & Quintero 1994,
In contrast to flint where flakes are in the majority, 41). The following table provides a breakdown of
the manufacturing of obsidian blades appears to be butt types found on blades and flakes from the east
the primary focus of obsidian production on both mound top-scrape sample.
the east and west mounds. Obsidian
blades comprise 36.5 per cent of ob- Table 10.1. Debitage categories by raw material: east and west mound 2 x 2 sample.
sidian debitage from the controlled 2 x
Debitage category Flint Obsidian Quartzite Other
2 sample from the east mound, while west east west east west east west east
flint blades contribute 10.2 per cent to Flake 3 3 18 90
Broken flakes & flake fragments 67 23 303 312
flint debitage. They are as common Regular blade 1 4
on the west mound, contributing 41.7 Regular blade fragment 1 5 326 544
Regular proximal blade fragment 2 45 7
per cent to obsidian debitage, al- Regular distal blade fragment 6 1
Non-regular blade 1
though flint blades are less common, Non-regular proximal blade fragment 2
contributing 3 per cent to flint Crested blade fragment 1
Core 1
debitage. As well as the variation seen Core fragment 1
Core tablet 2
in flint and obsidian, there is a sig- Chip 4 12 131 435
nificant discrepancy between the Shatter 22 13 69 117
Indeterminable 2 4 19
number of obsidian blade fragments
Total 100 59 905 1535 3 0
and proximal or distal blade frag-
ments. Where sufficiently detailed re-
Table 10.2. Debitage categories by raw material: east and west mound top-scrape sample.
cording permits comparison, this
appears to be a phenomenon noted Debitage category Flint Obsidian Quartzite Other
west east west east west east west east
elsewhere on Neolithic sites in
Flake 2 12 10 191 1
Anatolia and Syria such as Abu Broken flake 2 8 20 277 1
Hureyra (Nishiaki 1992; Conolly 1994) Flake fragment 9 46 41 828 3 3
Regular blade 1 2 1
and Can Hasan III (Ataman 1989). The Regular blade fragment 7 80 1028
Regular proximal blade fragment 2 26 271 2
reasons for this are unclear, although Regular distal blade fragment 4 69
it may be that the fragile distal ends Non-regular blade 1 3
Non-regular blade fragment 2 48
of blades are being mis-identified as Non-regular proximal blade fragment 1 11
Non-regular distal blade fragment 3
chips or flake fragments. An impor- Crested blade fragment 3
tant observation is that both obsidian Plunging blade 2
Core 4 2 2
and flint waste products are present Core fragment 11
Core tablet 13
in quantity, suggesting that core re- Chip 1 3 89 3
duction and tool manufacture were Shatter 4 18 21 216 5
Indeterminable 6 3 89
occurring on both the east and west total 23 108 212 3155 1 15
mounds, although no large deposits

177
James Conolly

1
3
2

4
5 6

---

Figure 10.1. Cores and core tablets - east mound: 1) obsidian blade core; 2) obsidian core tablet; 3) obsidian crested
blade fragment; 4-7) obsidian cores and core fragments; 8) flint core.
178
M ~ ~ V
l1 rn o-o ~
~ .r7Tl -- ""<~
28

9
m~ -rru ~ ~
10
noo i'- ~
11 12
~
6
~
7
~
8
-&-~
.....
30

29 I'\
::I
~
"d
31 "d
= [ -::1 1 (1)
- - . ·-
- -
0...
...... -
-
- -
o
-~ .- -
I
[J)
'I .. ·:: -
....,.
\0 0
::I
---· (1)
15 16 ---
13
14
~U -~-o
- _:~>- - - ___-
--- ---
:;;:l
(1)
"d
0
~ 33
....
....,.

~-[) ~-U ~fj R_~ 32

U ~H~ ~~ U
20
1-U-U
35

34
I

~
21 22

Figure 10.2. Blades, retouched blades, retouched flints - east mound: 1-8) obsidian blades; 13-23) retouched obsidian blades; 24-8) obsidian burins;
29) obsidian grinding/polishing tool; 30-31) obsidian drilling/piercing tools; 32) flint scraper; 33-4) obsidian scrapers; 35) obsidian chisel/wedge; 36) flint
drilling/piercing tool.
James Conolly

Table 10.3. Blade and flake 350 - - • - - east


butt types: east mound top-
scra pe sample. 300

j'·"· ---D-west

"'.". .
Butt type Blade Flake
250
Cortical 3

1 200

.
.I
Crus hed 19 175
Dihedra l 10 150
En chapea u
Faceted 41 100
Flat 24 140 ............
50
Lin ear
Punc tiform
45
156
19
53
/
Winged 1 15 0
-.o -.o -.o 00 0 N
In dete rminable 38 67 0 N
""' 00 0
......
N
...... ""'
......
00
...... 0
N
N
N ""'N N N ('() ('()
""'
('()

mm

Figure 10.3. Histogram of blade widths: east and west mound top-scrape sample.

Table 10.4. Blade dorsal scar patterning: east mound top-scrape Table 10.5. Mean and standard deviation of
sam ple. obsidian blade widths: east and west mounds
top-scra pe sample.
Dorsal scar pattern 4 5+ Indeterminabl e
Mean width Atandard deviation
Bi-directional 2
Crossed 4 East 13.10mm 3.71 mm 1373
Mu lti -direct ional 1 6 West 15.27 mm 4.09 mm 122
Uni -di rection al 195 698 65 J3
Indeterminable 436

Table 10.6. Flake dorsal scar pattern: east mound top-scrape


There are, in addition to the more regular blades, sample.
a number of non-regular blades in the assemblage
Dorsal sca r pattern 5+ Indeterminable
which are characterized by irregular cross-sections
Bi-directional 14 18 10 15
and sub-parallel margins. These seem to be a result C rossed 19 54 23 19
Multi-directional 5 28 52 133 22
of a less refined reduction strategy which does not Uni-directiona l 62 90 11 5 50 26 14
result in the standardized products seen in the regu- Ind eterminable 10 22 32 14 9 494

lar examples. Samples of this blade type are rela-


tively small, however, and it is not yet entirely clear flake fragments and broken flakes) most commonly
whether this represents a distinctly different pro- exhibit a general uni-directional scar pattern, al-
duction strategy or whether it is simply a stage in though multi-direction scarring occurs nearly as fre-
the production process of the more regular exam- quently (Table 10.6). Bi-directional and crossed scar
ples. Table 10.4 shows the distribution of the dorsal patterns, while not uncommon, are not nearly as
scar patterns on obsidian blades from the east mound prevalent. Larger flakes and flake tools preferen-
top-scrape sample. Clearly dominating the distribu- tially exhibit crossed and multi-directional scar pat-
tion is the occurrence of three uni-directional dorsal terns which are the patterns one would expect from
scars. The high number of indeterminables is a re- non-standardized debitage and amorphous cores,
sult of the difficulties of identifying both scar direc- with smaller flakes more commonly exhibiting a uni-
tion and total number of scars on fragmentary pieces. directional pattern.
There are too few examples of complete blades As with obsidian blade debitage, the mean and
to examine distributions of lengths, although histo- standard deviation of obsidian flakes from the east
grams of obsidian blade widths from both the east and and west mounds show a small but statistically in-
west mound top-scrape samples exhibit a normal significant increase in flake size between the two
distribution and therefore no clear technical basis mounds.
for distinguishing between small and large blades
by blade width alone (Fig. 10.3). The mean and stand- General characteristics of retouched pieces
ard-deviations of obsidian blade widths from the The following discussion provides an overview of
east and west mounds top-scrape sample show a the basic characteristics of the retouch and use-
slight but statistically insignificant increase in size modified pieces. A variety of different tool forms were
between the two mounds. recovered from the surface collection, all of which
Flakes are the other major debitage class at correspond to the categories of tools described by
C::atalhoyuk. On the east mound, flakes (including both Mellaart and Bialor, such as various retouched

180
Knapped-Stone Report

and worn blades and flakes, scrapers, bifaces, pro- and obsidian exploitation any further, it is perhaps
jectile points, burins, drills and piercing tools. Ex- sufficient to note the differences in the manner these
amples of the varieties of forms are illustrated in two materials were used.
Figures 10.2 to 10.6. As noted, surface collection tools
were classified in a manner similar to other Neolithic Edge attribute analysis
sites in the Near East; the distribution of these basic More detailed recording was undertaken on tools
types by raw material is given in Table 10.8. By far (and debitage) from the east and west mounds
the most frequent tools from both the east and west
mounds are retouched and utilized blades, which Table 10.7. Mean and standard deviation of obsidian flake
may have retouch on one or both lateral margins, widths: east and west mound top-scrape sample.
either directly or inversely. Mean width Standard deviation n=
Judging by the controlled surface collection sam- East 17.80mm 7.31 mm 1250
ple listed above, approximately 7 per cent of all ob- West 20.39mm 8.85mm 71

sidian and 18 per cent of all flint on the eas t moun d


exhibits signs of use or retouch, suggesting that flint Table 10.8. Tools by raw material: east and west mound 2 x 2
may have been more intensively worke d. This phe- sample.
nomenon has been noted from oth er Central East mound West mound
Anatolian early Neolithic sites, most notably Can Tool type-category obsidian flint obsidian flint

Hasan III, where Ataman (1989, 74) a ttributes the Burin


End-scraper
4
2 2
differences to either flint being conservedandmore Flake-scraper 7 I I 2
Notched blade 8 9
intensively worked than obsidian, flin t being im- Notched flake 4 5 I
ported in a finished or near-finished form, or that Piece esquillte I 7
Piercer I I 4
the debitage process of obsidian and flin t differs in a Retouched blade 37 3 68 2
Retouched flake 7 3 20 8
manner that results in greater quantities of obsidian Retouched fragment 8 2 I
waste products. I would provisionally suggest that Utilized blade 10 15
Biface I I
the first and last explanations are the mo re prevalent Biface fragment 8 I
Projectile point 4
reasons for the observed differences between the
two materials, although some flint obJ"ects such as Totals 102 10 134 13

the fine-grained dagger-like pieces


Table 10.9. Modified edge delineation by blank type: east mound top-scrape sample.
may have been imported in a finished
or near-finished state. The alluvial Edge delineation Flake Blade Shatter Core/core-derived Indeterminable/unclassified Total
Konya plain is not an area in which Burin edge 6 13 2 I 22
Concave edge 13 15 2 30
flint can be readily found (Baird pers. Convex edge 46 19 3 3 71
comm.), although Balkan-Alti (1994, Cran 3 I 4
Denticulated edge 22 91 2 2 117
37) suggests that central and south- Irregular edge 224 298 19 10 31 582
Tang 2 9 2 13
western flint sources are located near Long narrow tang 2 5 I 2 10
Ankara, the southwestern Taurus and Nosed edge I 2 3
47
Notched edge 14 32 I
possibly around Beysehir. Whether Piercing edge 17 8 I 26
Rectilinear edge 5 5 10
these areas contain tabular sources Regular edge 18 48 2 9 77
suitable for the production of large Shouldered edge
2
I 1
7
Tongued edge 5
blade cores is unknown. Large flint Unclassified 41 137 3 5 48 234
blades do occur at <::atalhoyiik, found Totals 419 685 32 19 99 1254
both in the 1993 surface sample and
from the excavations in the 1960s, and Table 10.10. Modified edge delineation by blank type: west mound top-scrape sample.
it is possible that these too were im- Edge delineation Flake Blade Shatter Core/core-derived Indeterminable/unclassified Total
ported as ready made items. The small Burin edge I 2
irregular flint flakes that are common Convex edge I 2
Denticulated edge 7 23 30
in the surface sample were likely pro- Irregular edge 19 49 2 2 73
duced on amorphous flint flake cores Notched edge
Piercing edge
I
4
5
2
from flint cobbles, some of which may Rectilinear edge I
Regular edge
have been obtained from the local Tang
area. While it is difficult at this point Totals 34 81 3 3 122
to define the differences between flint

181
11 12
rnln r~t~
~ 2

1
~-W-tl ~-~-t~j
13 14
t

\f?-V
I
()-W-8 15
tJ-~ 16
... 4
'-<
~
3
(!)
3 (f)
.......
00 t (J
N 0
~
17 2.
~
•-o• 5 18

mJf\\(:~
6 v·u · ~
20

-·-~ JS~8 &l&J 19

• ()-Q
'
7 . . . 10

Figure 10.4. Retouched/utilized pieces- east mound: 1) obsidian retouched non-regular blade; 2) obsidian scraper/scraping tool; 3) obsidian knife/cutting
tool; 4. indeterminable obsidian retouched/utilized piece; 5) obsidian scraper/scraping tool; 6) indeterminable obsidian retouched/utilized piece; 7) flint
knife/cutting tool; 8) flint drilling/piercing tool; 9) obsidian notched tool; 10) obsidian scraper/scraping tool; 11-20) obsidian pieces esquillees.
Knapped-Stone Report

top-scrape sample and the following discussion is function of the high numbers of pieces esquillees and
based primarily on these data. 35.7 per cent of the wedges and chisels. There is also a fairly high fre-
3283 lithic artefacts from the east mound and 47 per quency of alternate retouch on left and right lateral
cent of the 247 pieces from the west mound exhib- edges.
ited signs of retouch and/ or use. This is a high per- The functional classification of these retouched
centage of tools, which is likely a function of the pieces was based on an assessment of edge charac-
recovery method which missed the small, unre- teristics. If two areas of retouch perform the same
touched, chips and smaller flakes that make up a suspected function, then the artefact was classified
significant portion of a controlled sample. The 1178 by the common function. If, on the other hand, the
tools from the east mound display 1254 recorded
areas of modification, while 122 areas of modifica- Table 10.11. Edge delineation and angle types: east mound top-
tion were recorded on the 116 tools from the west scrape sample.
mound. The following two tables provide break- Edge delineation Crossed Abrupt Semi- Low Acute Indeter- Total
abrupt abrupt single minate
downs of edge delineation by blank type for the east blow
and west mounds. Sample sizes are too low to exam- Burin edge 22 22
ine the patterning in a statistical manner, although Concave edge 6 18 5 1 30
Convex edge 21 43 5 2 71
there is a prevalence of irregularly delineated edges Cran 2 2 4
Denticulated edge 7 56 50 4 117
on both blades and flakes with minor variation in all Irregular edge 3 73 338 167 1 582
edge forms between blades and flakes and the east Long-narrow tang 2 7 10
Nosed edge 1 1 1 3
and west mounds. Notched edge 10 27 5 5 47
Piercing edge 6 16 3 26
Included in the attribute analysis was the re- Rectilinear edge 3 2 4 10
cording of the delineation and angle of the modified Regular edge 21 38 17 77
Shouldered edge 1 1
edge. In all cases, semi-abrupt edges outnumber other Tang 2 8 13
Tongued edge 1 5 1 7
edge angle classes although here too there are smaller Indeterminate 15 23 16 177 234
variances in the distributions (Tables 10.11 & 10.12). Total 170 585 274 23 194 1254
For instance, on the east mound, proportions of
abrupt angles are preferentially larger on convex Table 10.12. Edge delineation and angle types: west mound top-
and regular edges, which could be directly attrib- scrape sample.
uted to functionally derived attributes, such as scrap-
Edge delineation Abrupt Semi- Low Acute Total
ing edges. A chi-square test of the distribution of abrupt single blow
abrupt and semi-abrupt angles on the east mound, Burin edge 2 2
ignoring unclassified delineations and delineations Convex edge
Denticulated edge 2
2
20 8
2
30
with a total of less than five abrupt or semi-abrupt Irregular edge 6 48 19 73
Piercing edge 6 6
examples, shows a significant difference at the 5 per Rectilinear edge 1
cent level. Regular edge
Notched edge
1
5
1
6
Further evidence of attribute patterning is sug- Tang
gested by the larger east mound sample, particularly Total 8 82 29 2 122
regarding the distribution of the location of the 1254
modified areas by blank type (Table 10.13). This dem- Table 10.13. Location of modification on blades and flake blanks:
onstrates the overwhelming prevalence of lateral east mound top-scrape sample.
edge retouch on blades and flakes, although flakes Location of modification Blade Flake Other/unclassified Total
display far more proximal and distal retouch than All edges 13 19 14 46
blades. There appears to be a slight tendency for Distal and proximal edge 12 66 33 111
Distal or proximal edge 6 23 13 42
single lateral edge retouch to be located on the left Distal edge 18 62 9 89
side on both flakes and blades. The distributions for Distal left edge
Distal right edge
7
12
6
10 1
13
23
blade and flake modification position are signifi- Proximal edge 15 14 3 32
Proximal left edge 11 6 17
cantly different at the 1 per cent level. Proximal right edge 12 7 19
A similar break-down of the location of modifi- Left and right lateral edge
Left or right lateral edge
353
18
81
41
21
18
455
77
cation shows the prevalence of direct modification Left lateral edge 79 38 117
Right lateral edge 72 27 100
on the left and right lateral edges of blanks - a Left medial edge 11 7 18
function of the high number of blades with modifi- Right medial edge
Indeterminable
8
39
4
7 37
12
83
cations in this area. Bifacial retouch occurs most com- Total 686 418 150 1254
monly on the distal and proximal edges of pieces, a

183
I

5 6

Figure 10.5. Bifaces/projectiles- east mound: 1-9) obsidian bifaces/projectiles; 10) possible flint 'dagger hilt'.

Knapped-Stone Report

functions of the two or more retouched areas dif- morphological variation exists in this tool class, rang-
fered, the implement was recorded as a combination ing from small bifacial, tanged, shouldered points to
tool. The following table provides the raw material double-tipped bifaces. No typological scheme has
types for the gross functional classification of the yet been devised, and likely will not be until such
1178 classified tools from the east mound and 116 time that a substantial sample from secure excavated
from the west mound top-scrape samples. contexts has been obtained. Drawings of a sample of
In all categories, obsidian implements outnum- bifaces/projectiles from the east mound serve toil-
ber their flint counterparts. However, flint exhibits lustrate the variation encountered (Fig. 10.5).
both a preference for tools that require more robust One additional tool class which deserves men-
edges, perhaps because of its different physical char- tion is that of the large flakes that display varying
acteristics. For instance the ratio of flint to obsidian degrees of marginal retouch, appearing in several of
drilling/ piercing and scraper I scraping tools (1 :6.3 the room contexts excavated by Mellaart in the 1960s.
and 1:11.4) from the east mound differs from the Four similar artefacts were found in a deposit on the
ratio of flint to obsidian knife/cutting tools (1:51.7). top of the northern eminence in the 1993 season (Fig.
A breakdown of all classified tools by flake and 10.6) . Although the deposit may represent a raw-
blade blanks for the east mound sample clearly dem- material cache, the objects have modified edges and
onstrates that blades are favoured for tool blanks, could be large cutting or butchery tools. The four
yet there are significant differences in the type of that were found in 1993 all have either faceted butts
blank selected for specific functions. The most nota- and/ or evidence of platform preparation and are
ble difference concerns scraper I scraping tools and fairly regular in shape suggesting some effort went
pieces esquillees, both of which have significantly more into removing a very large, regular, flake blank.
flake based forms than blade based forms. Similarly
flakes and other types of debitage (such as cores, Spatial patterning
core tablets and shatter) are also more commonly Some differences between the west and east mounds
used for piercing and drilling tools than blades. of <:;atalhoyi.ik have already been mentioned; of con-
The 36 bifaces/projectiles and fragments from cern here are differences in the distribution of arte-
the eastern mound top-scrape sample are mostly facts across the surfaces of the individual west and
fragments and thus it is often impossible to deter- east mounds.
mine original tool morphology. Some near com- On the west mound, low densities of ma terial
plete examples suggest, however, that considerable hinder identification of distribution patterning. Mean

Figure 10.6. Large retouched flakes from cache on northern eminence: 1-3) obsidian flakes from cache.

185
James Conolly

counts of artefacts by east-west transects suggest a overall densities, patterns are difficult to identify.
slight increase in density at the northern end of the For instance, the distribution of blades over the sur-
west mound, possibly relating to prehistoric surface face is very uniform, varying in density only be-
disturbance and cover by modern land-use (Fig. 10.7). tween zero and three blades per 2 x 2 unit (Fig.
There is no significant patterning in the percentage 10.10). The distribution of the smallest pieces of lithic
of blades in transects across the mound (Fig. 10.8). artefacts recovered from the surface collection- the
Turning to the surface of the east mound of small chips of obsidian - was also examined for
<::atalhoyiik, there are some minor variations in patterning. Here too there is a low density on the
density which can be correlated to distributions of southern slope of the larger eminence, possibly be-
ceramic and bone densities. For example, the south- cause of classical masking (Fig. 10.11). Generally
ern part of the mound, particularly the southern higher densities of obsidian chips are found extend-
slope of the main eminence, displays a reduced den- ing onto the lower slopes, presumably as the result
sity of flint and obsidian artefacts (Fig. 10.9). Evi- of selective downslope movement of smaller items.
dence from the ceramics suggests that this may be Because of suspected differences between the
the focus of classical occupation deposits, which thus northern and southern parts of the mound in the
mask earlier, Neolithic, deposits resulting in the lower proportion of blades and flakes, north/ south vari-
densities of knapped-stone. Lower overall densities ability was examined by the calculation of the per-
of lithic material are also seen at the northwest cor- centage of blades in the total east-west transect
ner of the mound. The distributions of various sample. These data hint at a difference as one moves
debitage categories were also examined for any spa- from south to north, showing a minor decrease in
tial variation, although because of the extremely low the percentage of blades in the northern area (Fig. 10.12).

0
0 '"'
.....
\D II
OM
o II
OM
'<!< II o§
00 II 0~ oG OLO
0 .....
0~
'<!< .....
00\
oo II
ol()
N II
l'-.5 oo5 oo5 005
N II \D II
s:: 0
..... 5 II o II
..... 5 ~5 :::5
"'5 0\
.......,
northing
Figure 10.7. Mean count of lithic artefacts in each west mound 2 x 2 unit, by northing.
80

(!)
70
60 .\ . I
.
./\ ;·'/.\__ ·-·,_ _.,_I
bO 50
....s::
ra
(!) 40 ./ ........................................
...(!) 30
u
0..
20

10
0 11!1

0~
~ ......., g oPl
oo II N
'"'
oO.:
II
0~II
\D
0~
o:::
oO\
'<!<ll'l
0 II
000
00'<1'
0 II
0~
NN
,..... II
005 "'5 "'5 0 II
..... 5 ..... .......,
s:: ..... 5 ..... 5
northing
Figure 10.8. Percentage of blades in each west mound northing.

186
Knapped-Stone Report

1300

1200

1100

1
~
s
1000

900

800

900 1000 1100 1200


metres east

0 5 10 15 20 25 30 35
number
Figure 10.9. Distribution of lithic artefacts: east mound.

187
James Conolly

900 1000 1100 1200


metres east

0 1 2 3
number

Figure 10.10. Distribution of blades: east mound.

188
Knapped-Stone Report

1300

1200

1100

1
~
e

1000

900

800

900 1000 1100 1200


metres east

0 1 2 3 4 5 6 7 8 9 10 11 12
number

Figure 10.11. Distribution of chips: east mound.

189
James Conolly

As analyses of Mellaart's data by myself and others These can be used to discuss the two major
do suggest chronological variation in the proportion sources of surface patterning; variation in the
of blades to flakes, with blades replacing flakes ap- Neolithic occupation and post-Neolithic occupation
proximately midway through the occupation (see and erosion. For the purposes of discussion of varia-
below), the pattern observed in the surface data is tion in the top-scrape sample, the mound can be
thought to relate to temporal variation in the occu- divided into thirteen separate areas:
pation of the mound. Although sample sizes are • 100 m 2 towards the base of the northern eminence
very low, this remains the clearest example of [area 1];
patterning in the 2 x 2 sample. • 1600 m 2 on the north slope of the northern emi-
In contrast to the surface sample where low nence [area 2];
overall volumes of artefacts limit detailed discussion • 300 m 2 on the top of the northern eminence, join-
of surface patterning the top-scrape sample, despite ing with area 2 [area 3];
the reduced control over recovery, affords more de- • 100 m 2 just southeast of the top of the northern
tailed discussion. There are four aspects of the top- eminence [area 4];
scrape sample which display spatial patterning • 100 m 2 on southeast slope of the northern emi-
between different areas of the mound: nence [area 5];
• differences in the density of lithic artefacts; • 100m2 towards the top of the gully between the
• differences in the percentage of blades by area; northern eminence and the main mound [area 6];
• differences in percentage of retouched pieces by • 400 m 2 on the northern slope of the main mound
area; [area 7];
• differences in the percentage of chips and shatter • 100 m 2 on the western slope of the main mound
by area. [area 8];
• 100 m 2 just north of the
summit of the main
mound [area 9];
• 100 m 2 on the eastern
slope of the main
mound [area 10];
• 100 m 2 on the eastern
plateau [area 11];
• 100 m 2 on the summit
OMII 0~ or;:; oU) 0~ o
~ ~

8 0 8 ON 0~ 00\ NOO oi'O oG o 0 8


...... of the main mound
N
oo3 003 ~II o II "" II 00 II NOOo II ~~ 0~ ..,.;:::: 00~ N II N ~""II <'"l II
o II ,..... II ,..... II ,..... II
a-3 a-3 a-3 ...... 3 . . . 3 . . . 3 . . . 3 . . . 3 ...... 3 ...... 3 ...... 3 [area 12];
northing
• 100 m 2 on the south
slope of main mound
Figure 10.12. Percentage of blades in each east mound northing. [area 13].
The density of knapped-
35 stone artefacts per cubic
380 metre has some bearing
30 on the post-Neolithic
25 processes affecting the
east mound (Fig. 10.13).
>,
.-:::: 20 The area with the highest
"'QJ~ 15 density is area 4 with over
'"0 30 pieces per cubic metre
10 .followed by areas 10 and
12, both with 13.5 pieces
5
per cubic metre. Thus the
0 steep eastern slope of the
...... C'l Cr"J -.t' t-.. ...... main mound, the area just
"' 00 0
1.1") 0\ C'l Cr"J
ro ro ro ro ro ro ro ro ro
...... ...... ...... ......
QJ
1-<
QJ
1-<
QJ
1-<
QJ
1-<
QJ QJ QJ QJ QJ ro ro ro ro to the south of the top of
1-< 1-< 1-< 1-< 1-< QJ QJ QJ QJ
ro ro ro ro ro «l ro «l «l 1-< 1-< 1-< 1-<
«l «l «l «l northern eminence and
Figure 10.13. Density of lithic artefacts by area: east mound. the area near the summit

190
Knapped-Stone Report

of the main mound have Table 10.14. Location of modification by modification position: east mound top-scrape sample.
the highest densities,
while the gentler slopes Location of modification Alternate Alternating Bifacial Direct Inverse Natural edge Burin blow Unclassified

(areas 2, 3, 5, 7, 8, 9 & All edges


Distal and proximal edge
1
2
29
93
16
7 7 2
13) and the lower slope Distal or proximal edge 24 8 4 4 2
Distal edge 5 64 18 1
of the northern emi- Distal left edge 2 5 4 2
nence (area 1) and the Distal right edge
Proximal edge
1
5
15
22
2
3
2
2
eastern plateau (area 11) Proximal left edge 12 1 3
Proximal right edge 2 7 6 3
have lower densities. Left and right lateral edge 40 5 66 303 34 1
This pattern may result Left or right lateral edge
Left Ia teral edge 1
14
10
50
87
8
18
from the exposure of Right lateral edge 2 7 73 18
Left medial edge 1 11 6
new deposits as erosion Right medial edge 7 4
proceeds from the tops Unclassified 3 8 72

of the mounds. Totals 45 262 695 133 2 17 92

One aspect of patt-


10.15. Functional classifications by raw material: east and west
erning in the lithic artefacts which is a function of Table mound top-scrape sample.
use of space in the Neolithic period concerns the
differences between the northern and southern areas East mound West mound
Gross functional classification flint obsidian other flint obsidian
of the east mound. The differences in this case are
Burin 15
largely connected with significant differences in the Carving/polishing/grinding tool 4
4
debitage, particularly in the proportions of blades, Chisel/wedge Combination tool
1
4
13
45 5
flakes and other debitage by area. For instance, the Drilling/piercing tool 6 38 4
Knife/cutting tool 10 517 78
samples collected from the northern eminence, par- Notched tool 44 4
Piece esquillie 1 166 7
ticularly on the top and north slope, display lower Biface/ projectile 2 34
proportions of blades to flakes and other debitage Scraper I scraping tool 17 80
Unclassified/indeterminable 9 171
than samples derived from the other areas of the
Totals 50 1127 109
mound (Table 10.17). A chi-square statistic estab-
lishes that the patterning is significant at the 1 per Table 10.16. Functional classifications by blank type: east mound top-
cent level. The largest contributions to the statistic scrape sample.
come from the distribution of blades where, given the
Gross functional classification Flake Blade Other Total
overall distribution, there are far fewer than expected.
Burin 2 10 3 15
As already noted in relation to the 2 x 2 Carving/polishing/grinding tool 1 3 0 4
0 9 14
patterning, it would seem reasonable to assume that Chisel/wedge Combination tool
5
23 23 3 49
such differences in debitage have chronological sig- Drilling/piercing tool 20 12 12 44
Knife/ cutting tool 107 417 4 528
nificance. A possible interpretation is that the Notched tool 14 30 0 44
Neolithic surface of at least a portion of the northern Biface/ Piece esquil/ee 96 14 57 167
projectile 1 25 10 36
eminence is earlier than the Neolithic surface of the Scraper I scraping tool 75 11 11 97
Unclassified/indeterminable 45 101 34 180
main eminence. Given that the major change in the
Totals 389 646 143 1178
replacement of flake debitage by blade debitage oc-
curs approximately at level V, this suggests that the Table 10.17. Numbers of blades, flakes and other debitage by area: east
surface of the northern eminence may not be later mound top-scrape sample.
than level VI.
Blades Flakes Other debitage Total
A further aspect of spatial patterning concerns
Northern eminence (areas 1 to 5) 672 886 287 1845
variation in the percentage of tools (Fig. 10.14) and Main mound (areas 6 to 13) 782 486 170 1438
chips and shatter (Fig. 10.15) in individual top-scrape Total 1454 1372 457 3283
sample units. Given that chips and shatter are di-
rectly associated with knapping activities, high pro-
portions of these two debitage categories may in the proportion of chips and shatter by area may
indicate deposits of knapping debris. I should em- have some archaeological significance. In particular,
phasize that no sieving was performed on these area 1, which lies towards the bottom of the slope on
samples, so absolute counts will be much lower the northern eminence contains both the highest ra-
than samples with controlled recovery. Nevertheless, tio of chips and shatter to other debitage and the
assuming consistent recovery procedures, patterning lowest proportion of tools.

191
James Conolly

58.97435897
60

50

40

30

20

10

0
area area area area area area area area area area area area area
2 3 4 5 6 7 8 9 10 11 12 13

Figure 10.14. Percentage of tools by area: east mound.


25

20
Q)
bJl
15
c"'
Q)

~ 10
Q)
0..
5

0
00 cr-. 0 N M
N M
"""
If)
"' " ..-<
..-<
..-< ..-<

"',_,
Q)
"',_,
Q)
"'~ "',_,
Q)
"',_,
Q)
"'~"' ,_,
Q) "',_,
Q)
"'~ "',_,
Q)
"',_,
Q) "'~ "'~
"' "' "' "' "' "' "' "' "' "' "' "' "'
Figure 10.15. Percentage of chips and shatter: east
mound.

10.17). The main interpretable patterning, therefore,


occurs on a macro-scale, involving variation between,
rather than within, the various sample areas.

The Konya Museum analysis

0 5 10 1 Introduction
number In the summers of 1993, 1994 and 1995, the obsidian
and flint material from <:;:atalhi:iyuk stored in the
Figure 10.16. 2 x 2 distribution of lithic artefacts in Konya Archaeological Museum was examined. The
square 1040,1170. Museum sample consists of several thousand lithic
artefacts bagged by level and/ or room number
In order to determine whether there is any together with any other information concerning
patterning within a 10 x 10 top-scrape unit or any provenance. There is no indication that sieves were
correlation between artefacts and architectural features, used by Mellaart, thus small debitage, debris and
square number 1020,1170 was excavated in 2 x 2 units. micro-tools are under-represented. Nevertheless, to-
No clear patterning emerged, however, although den- gether with larger pieces, there are considerable
sity distributions (Fig. 10.16) are roughly similar to amounts of debitage, cores and small tools. Gener-
the plots of ceramic and bone material. Similarly, ally, the samples from Mellaart's excavations have a
there was no clear patterning of lithic material within similar distribution of artefact sizes to the samples
the 1610 x 10 squares on the northern eminence (Fig. from the 10 x 10 top-scraping sample. The analysis

192
Knapped-Stone Report

0 5 10 15 20 25
fragments/m3

Figure 10.17. 10 x 10 distribution of lithic artefacts in areas 2 and 3.

193
James Conolly

70 245 559

60
50
t£·~·"" •
39

QJ
- - - - · 426
.sbOr::: 40
QJ
~
QJ
30
0...

--·---
20
77 12 4
10 •-......_ 31 • 5 • 21

0 ~.--
+-----+-----~----;-----;------r-----r-----r-----+-----+-----+----~
---...::.
II III IV v VIA VIB VII VIII IX X XI XII

Figure 10.18. Percentage of blades by level.

of the tools and modified pieces is ongoing, but I am there is a significant shift in production technology
able to offer here some preliminary results concern- at approximately level V (Fig. 10.18). While this is
ing temporal variation in debitage technology and an important observation that has significant reper-
concerning the patterning ofbifaces/projectiles, cores cussions concerning changes in the organization of
and thinning flakes in the sample. production at ~atalhoyiik, research is continuing and
I did not, however, enter into the current analy- interpretations are not yet available.
sis without the advantage of previously published There are also patterns concerning the contexts
synopses of samples of the material. Bialor (1962) of bifaces/projectiles, cores and thinning-flakes. Be-
and Mellaart (1963; 1964) are the two primary sources cause of difficulties in assigning some of the mu-
of information, and they provide some provisional seum data to contexts, combined with the fact that a
observations regarding the technological and typo- small proportion of the artefacts are not currently
logical characteristics of the industry. More recently, stored in the Konya Museum, there are occasional
in a comprehensive review of prehistoric Anatolian discrepancies between these data and the data reported
knapped-stone industries, Balkan-Atli (1994) pro- in the preliminary reports by Bialor and Mellaart.
vides a useful synopsis of the main characteristics of These problems have yet to be resolved in their en-
the knapped-stone artefacts at ~atalhOyiik. She has tirety, and correlation of the data described in the
summarized the knapped-stone assemblage from lev- preliminary reports with the museum data is ongo-
els XII to VIII based on a re-analysis of findings of ing. Nevertheless, the information collected thus far
the first three years of excavation at ~atalhoyiik and does provide some interesting insights into the dis-
describes a number of different tool forms. The over- tribution of bifaces/projectiles, cores and thinning-
all impression from both Balkan-Atli's re-analysis flakes, yet these must remain provisional until the
and the earlier reports is one of an enormous range uncertainties described above are resolved.
and variety of tool forms, from the simplest utilized The term 'biface/projectile' here refers to those
flakes to some of the most elaborately worked piece pieces of obsidian and flint which have been re-
of siliceous stone in the Near East, with no clear or touched in such a manner as to modify the original
unequivocal temporal patterning. blank to a form which resembles pieces commonly
There are suggestions from these reports, how- described as projectile-points, arrow-heads, lance-
ever, that there are temporal variations in production heads, daggers and their preforms. This does not
technology at ~atalhoyiik. Although the sample size include pieces which exhibit modification on both
is small, evidence can be found in Balkan-Atli's sum- their ventral and dorsal surfaces unless the retouch
mary of the early phases of the industry (1994, 127) is more than marginal and significantly contributes
as well as in Bialor's initial report (1962, 73) that to tool morphology. Thus, bifadal knives and other
there is a change from an earlier flake dominated modified pieces with bifacial retouch are not included
industry to a blade dominated industry. The num- in this discussion. I recognize that this is a provi-
bers of blades and flakes by level does suggest that sional and arbitrary division.

194
Knapped -Stone Report

The term core here refers specifically to all or thinning flakes appear in these two contexts and,
cores- blade, flake, amorphous or otherwise. Again, thus, there is no clear patterning.
until such tie as a detailed analysis of the distribution
of all debitage type categories has been completed, Level VIII
this larger division must suffice. The distribution of Unlike earlier levels, level VIII begins to display some
a small debitage component, thinning-flakes, is also of the patterning that emerges more definitely in
discussed. The term refers to flakes which possess a later levels at c;:atalhoyiik. Here, for instance, the
particular morphology resulting from a removal proc- structures with the fewest features, VIII.12, VIII.18,
ess associated with the 'thinning-out' of bifaces/ pro- VIII.3, and VIII.24 possess only 3 bifaces/projectiles
jectiles. Experimental knapping has indicated that between them. On the other hand, the more complex
certain morphological characteristics of flakes, largely structures, particularly VIII.29 and VIII.31, with 10
connected with the shape and angle of the butt and and 14 features respectively, contain a total of 9
lateral edges can be used to identify flakes derived bifaces I projectiles. The only core that could be placed
from biface/projectile thinning and shaping. Al- in a context within level VIII was also found in room
though the identification of these pieces is not with- VIII.29. Thinning-flakes also appear in VIII.29, as
out contention, I do think that it provides a quick well as in VIII.10, VIII.31, VIII.14, and VIII.25- again,
assessment of quantities of flakes derived from gen- generally the more complex rooms in level VIII.
eral debitage and biface/projectile thinning and
shaping. Once again, however, I do emphasize that Level VII
these categories, and thus the results derived from Only 15 of the 44 structures from level VII currently
them, are provisional; the ongoing refinement of the have available data concerning lithic artefacts. What
methodology will no doubt alter some of the data is available, however, shows some patterning con-
trends here observed. cerning biface, core, and thinning-flake distribution.
The distribution of these three classes of lithic Although the total number of bifaces/projectiles
artefact will be discussed by level, with reference to which could be assigned to level VII is low, 12 of the
buildings that appear in the assessment of complexity 276 for which contexts are available, the 12 do never-
(Figs. 1.3-1.8). Because of the ambiguous contextual theless show relatively clear clustering with the more
information for some of the lithic artefacts in the Konya complex buildings. It may be the case that some of
Museum, only 56.8 per cent of bifaces/projectiles and the patterning is spurious, having more to do with
50.6 per cent of cores can currently be assigned to the fact that data are not yet available for many of
contexts. This still represents a sample of 276 bifaces/ the level VII buildings. Nevertheless, what data there
projectiles and 40 cores, from which distribution are do suggest that bifaces/projectiles are most
patterns can be examined. There are currently 90 build- closely associated with the more complex buildings
ings for which lithic data can be examined and, as will of level VII, particularly those which have been des-
be shown, some interesting distribution patterns have ignated as 'shrines' by James Mellaart. For instance,
emerged both within and across levels. bifaces/projectiles are found in 7 structures, out of a
total of 15 for which there is available information.
Level XI Of these 7, 5 are shrines- S.VII.8, S.VII.45, S.VII.10,
Because of a lack of detailed building plans for level S.VII.35 and S.VII.21. The only core that can be as-
XII, level XI is the earliest phase for which a 'com- signed to a level VII context occurs in both a shrine
plexity index' can be calculated for any structure, and a complex room, S.VII.12. Interestingly, how-
and then only for one building, Xl.29, which exhibits ever, thinning-flakes do not cluster with the com-
11 features. Xl.29 also possess 3 bifaces/projectiles. plex buildings as bifaces/projectiles do, but
As there are no other structures available for com- nonetheless are restricted almost exclusively to
parison from level XI, little further can be said about shrines, save for room VI1.28.
contemporary distribution patterns.
Level VIB
Level IX Of the 58 structures included on the complexity scale,
The next phase with available information is level 23 have data concerning biface, core and thinning-
IX, with two structures that possess lithic data and a flake distributions. There are 78 bifaces/projectiles
complexity index, IX.29 and IX.31. These buildings which can be placed into level VIB contexts, how-
contain 7 and 11 features respectively. Three bifaces/ ever in most cases structures which possess available
projectiles occur in IX.29, and none in IX.31. No cores data concerning lithic distribution also contain

195
James Conolly

bifaces/projectiles; they occur in 19 of the 23 structures Level V


for which lithic data are available. Hence, little can There are 7 structures from level V which possess
be said regarding biface/projectile presence or ab- lithic data, out of a total of 24 which appear on the
sence patterning. Despite this, there are some nota- complexity graph. No thinning-flakes can yet be as-
ble differences in the quantities of bifaces I projectiles signed to any level V contexts, although 28 bifaces/
that appear in buildings on this level. For instance, projectiles can be assigned to level V structures. The
22 of the 78 bifaces/projectiles occur in one struc- majority of these bifaces/projectiles occur in the more
ture, shrine VIB.lO, which is the second most com- complex structures. For example, F.V.1 contains 13
plex in VIB. There are, however, exceptions as 13 bifaces/projectiles, and is the second most complex
bifaces/projectiles appear in building VIB.32, which structure on level V with 21 features. F.V.1 also con-
is far less complex - only 2 features compared to tains 4 of the 5 cores that can be assigned a context
the 30 in VIB.10. Similarly, 9 bifaces/projectiles were from this level. The structure with the second great-
found in structure VIB.17 which only has 4 features. est number of bifaces/projectiles is S.E.V.8, which
The other 34 bifaces/projectiles are distributed more possesses 9 features. Again there is an exception to
or less evenly among the remaining buildings from this emerging association of bifaces/projectiles and
this level, with no more than 5 appearing in any one complex buildings, as E.V.7 contains both one biface/
structure. There is, therefore, a definite disparity with projectile and one core despite being non-complex.
regards to quantities ofbifaces/projectiles possessed Nevertheless, the general pattern remains that the
between S.VIB.10 and the other structures from level more complex structures from level V do contain
VIB, although large numbers of bifaces/projectiles more bifaces/projectiles, and the highest frequency
were also found in less complex buildings. of cores also occurs in structures with the greatest
Unlike the bifaces/projectiles, however, the number of bifaces/projectiles.
cores from level VIB are generally exclusively re-
stricted to the more complex structures, although Level IV
once again VIB.32 does stand outside the pattern by Four structures from level IV contain lithic data, 3 of
exhibiting a single example. However, the majority which possess bifaces/projectiles in their assem-
of the cores, 6 of the 7, are found in structures with blages. No contextual information has yet been dis-
greater than 15 features. Thinning-flakes occur in covered for cores or thinning-flakes in the current
only 2 structures, VIB.13 and VIB.17, neither of which analysis, although Bialor (1962, 87-9) notes that 15
are particularly complex, although as noted, VIB.17 projectile-points and 2 cores from level IV contexts
does possess 9 bifaces/projectiles. were found in the 1961 excavation season. Only 13
bifaces/projectiles from the Konya Museum can cur-
Level VIA rently be assigned to context, highlighting the provi-
Only 9 of the 44 structures from level VIA found in sional nature of this discussion. Five of these bifaces/
the complexity index have associated lithic data. Of projectiles are found in S.E.IV.1, which is also the
these 9,5 contain bifaces/projectiles and/or cores,4 most complex structure of level IV, with 25 features.
of which were designated as shrines by Mellaart. The second most complex structure, S.E.IV.4, con-
Currently there is no information concerning thin- tains 3 bifaces/projectiles; thus the two most com-
ning-flakes from level VIA. Here the more complex plex buildings contain the majority of the bifaces/
structures do contain significantly more bifaces/ projectiles of level IV. This is, however, again an
projectiles; S.VIA.1 and S.VIA.31 contain 27 and 15 exception to this pattern, as 5 bifaces/projectiles can
bifaces/projectiles respectively, out of a total of 49 be assigned to E.IV.S, which is low on the complex-
for this level. While these two structures are not the ity scale for level IV with only 4 features.
most complex in VIA, they both exhibit over 15 fea-
tures each. However, S.VIA.7, with 18 features, has Level III
only 2 bifaces/projectiles and S.VI.30, with 12 fea- It is level III and II which provide the clearest evi-
tures, only one biface. On the other hand, VIA.49 has dence for the clustering of bifaces/projectiles and
4 bifaces/projectiles, yet only one feature. So, while cores with complex structures at <;atalhoyiik. In level
there are exceptions, the general pattern remains III, 10 of the 14 structures which are included in the
that the building contexts with the largest quantities complexity index have lithic data and a clear assess-
of bifaces/ projectiles are also the more complex. ment of biface/projectile distribution can be made.
There are also 3 cores from level VIA, 2 of which For example, 44 bifaces/projectiles can be assigned
are found in shrine VIA.1, and one in VIA.49. to structural contexts, 20 of which occur in 4 shrines-

196
Knapped-Stone Report

S.AIII.13, S.AIII.8, and S.AIII.l. Ten bifaces/projec- and B.ll.2, each containing one biface. There are no
tiles are seen also in AIII.4 which has 9 features, as other bifaces/projectiles from this level; they are thus
well as 4 cores. The shrine AIII.1 is also remarkable restricted solely to the three most complex buildings.
not only because it contains the greatest number of
features on level III and a high number of bifaces/ Summary
projectiles, but also because it contains 7 cores- the To date the lithic assemblages stored in the Konya
second highest number of cores in any structure ex- Museum from 90 structures in levels XI, IX, VIII, VII,
amined. Thus the evidence here suggests that bifaces/ VIB, VIA, V, IV, III and II have been examined for
projectiles and cores do cluster preferentially in the evidence of bifaces/projectiles, cores and thinning-
more complex buildings of level III. flakes. This is not a total sample, and once the ambi-
guities in contextual information have been resolved,
Level II additional data will become available that may alter
The pattern observed above is strengthened by the the patterning that has been observed so far. What
data from level II. Here, 6 of the 8 structures possess patterning there is, however, suggests that bifaces/
relevant data, and bifaces/projectiles are found in 4 projectiles and cores are not evenly distributed, as
structures, cores one. The shrine A.II.1 stands out, for it there is a slight tendency for bifaces/projectiles and
contains 24 bifaces/projectiles and 8 cores- the most cores to occur in the more complex buildings on any
cores in any structure examined -as well as being the particular level. It is the later levels which most clearly
most complex structure of level II, with 18 features. exhibit this patterning, although such trends are
There are no other shrines on level II, and the second hinted at by level VIII.
and third highest number of features occur in B.II.1 If the distribution of bifaces/projectiles and

Table 10.18. Counts of bifacesfprojectiles and cores and presence of thinning-flakes by structure.

Structure Bifaces/projectiles Cores Thinning-flakes? Structure Bifaces/projectiles Cores Thinning-flakes?


Xl.29 S.Vl.B.45
IX.29 S.Vl.B.61
IX.31 S.Vl.B.12 1
VIII.12 2 VI.B.25 3 1
VII1.18 S.VI.B.10 22 2
VIII.3 1 S.VI.B.8 3
VIII.29 4 yes VI.A.49 4
VIII.lO 1 yes VI.A.17
VIII.31 5 yes Vl.A.2
VIII.14 3 S.VI.A.30
VIII.25 2 yes S.VI.A.25
Vlll.27 2 yes S.VI.A.1 27 2
VII.28 yes S.VI.A.7 2
S.VII.29 S.VI.A.31 15
S.VII.31 yes S.VI.A.8
VII.33 E.V.7 1
VII.32 S.E.V.S 9
VII.2 F.V.7 3
S.VII.9 S.E.V.2 1
S.VII.44 yes S.E.V.6 1
S.VII.S S.E.V.3
S.VII.23 F.V.1 13 4
S.VII.45 E.IV.6
S.VII.lO E.IV.5 5
VII.12 S.E.IV.4 3
S.VII.35 2 S.E.IV.1 5
S.VII.14 yes A.III.3 2
S.VII.21 2 yes A.III.5 1
Vl.B.32 13 A.III.9 3
VI.B.13 yes A.III.11 1
Vl.B.27 2 S.A.IIl.13 12
VI.B.39 3 A.III.6 4
Vl.B.17 9 yes A.III.2 2
Vl.B.9 2 A.III.4 10 4
Vl.B.18 1 A.III.7
Vl.B.2 5 S.A.III.8 1
S.Vl.B.20 A.III.10 1
VI.B.3 1 S.AIII.l 7 7
S.VI.B.5 4 B.ll.3
S.VI.B.15 2 A.II.3
S.VI.B.50 2 B.ll.5
VI.B.4 1 B.II.2 1
VI.B.34 B.II.1 1
S.Vl.B.1 2 S.A.II.1 24 8
Vl.B.28
S.VI.B.44 Total 276 40

197
James Conolly

cores is not sub-divided into levels, and examined exhibit short, discontinuous and/ or partial
solely with regard to building complexity, there is bifacial retouch which does not significantly con-
also a fairly clear distribution pattern. For instance, tribute to the overall morphology of the piece in
of the 90 structures for which lithic data can cur- question.
rently be ascertained, 38 are designated by Mellaart
as 'shrines'. These 38 'shrines' contain 155 bifaces/ References
projectiles - in other words, 42.3 per cent of the
structures contain 56.2 per cent of the bifaces/pro- Adams, W. & E. Adams, 1991. Archaeological Typology and Practi-
jectiles. A similar distribution is seen for cores, as in cal Reality: a Dialectical Approach to Artifact Classification and
this instance, the 'shrines' contain 62.5 per cent of all Sorting. Cambridge: Cambridge University Press.
the cores included in this analysis. As structures Ataman, K., 1989. Lithic Analysis of Can Hasan III. Unpub-
marked as shrines are often, although not exclu- lished Ph.D. Dissertation, University of London.
Baird, D., H.G. Gebel, B. Miiller-Neuhof, K. Schmidt, G.
sively, the more complex structures on any particu-
Rollefson & M. Beile-Bohn, 1995. The Wembach Mod-
lar level, there also appears to be a .correlation ule for non-formal tool analysis. Neo-Lithics: a Newslet-
between bifaces/projectiles, cores, and structures ter of Southwest Asian Lithics Research 2, 1-4.
with high numbers of features. Balkan-Atli, N., 1994. Le Neolisisation de l'Anatolie. Paris: De
The evidence for patterning in the distribution Boccard.
of thinning-flakes is more ambiguous, although the Bialor, P., 1962. The chipped stone industry of <;atal Hiiyiik.
data from levels VIII and VII suggest that they too Anatolian Studies XII, 67-110.
may occur more frequently in the more complex Clark, J.E., 1987. Politics, prismatic blades, and Mesoamerican
buildings, and thus may be associated with the civilisation, in The Organisation of Core Technology, eds.
J.K. Johnson & C.A. Morrow. London: Westview Press,
bifaces/projectiles from which they are derived.
259-84.
However, determining whether or not the produc- Conolly, J., 1994. Obsidian Production at Abu Hureyra. Un-
tion of bifaces/projectiles occurred in situ is only in published M.A. Thesis, University College London.
part dependent on the presence of manufacturing Edmonds, M., 1995. Stone Tools and Society. London: Batsford.
debris - it would, I feel, be necessary to re-fit thin- Grace, R., 1989. Interpreting the Function of Stone Tools: the
ning-flakes to make a convincing argument. It is Quantification and Computerisation of Microwear Analy-
anticipated that this form of analysis will be intro- sis. (BAR International Series 474.) Oxford: BAR.
duced in the near future to explore this possibility. Hurcombe, L., 1992. Use Wear Analysis and Obsidian: Theory,
In any event, the continuing analysis of the other Experiments and Results. (Sheffield Archaeological
Monographs 4.) Sheffield: J.R. Collis Publications.
debitage classes and tools from these and other
Inizan, M. et al., 1992. Technology of Knapped Stone. Paris:
contexts will hopefully provide further insight on CNRS.
both this and other related issues. Gero, J., 1991. Genderlithics, in Engendering Archaeology, eds.
M. Conkey & J. Gero. Oxford: Blackwell.
Notes Mellaart, J., 1963. Excavations at <;atal Hiiyiik, first prelimi-
nary report, 1962. Anatolian Studies 13,43-103.
1. This is the product of a sub-group of the larger Mellaart, J., 1964. Excavations at <;atal Hiiyiik, second pre-
organization Southwest Asian Lithics Research, liminary report, 1963. Anatolian Studies XIV, 39-119.
founded specifically to deal with the issue of non- Nishiaki, Y., 1992. Lithic Technology of Neolithic Syria: a
Series of Analyses of Flaked Stone Assemblages from
formal tools, and I acknowledge the contribu-
Douara Cave II, Tell Damishilyya, Tell Nebi Mend,
tion informal meetings with various members of and Tell Kashkashok II. Unpublished Ph.D. disserta-
that group has made to the following discussion. tion, University College London.
2. The term 'piece esquillee' refers to small blades and Pelegrin, J., 1990. Prehistoric lithic technology. Archaeological
flakes that typically exhibit crushing and scarring Review from Cambridge 9(1), 116-25.
on opposed margins. There is debate as to whether Tixier, J., 1974. Glossary for the Description of Stone Tools. Trans.
these pieces represent small opposed platform cores M. Newcomer. (Newsletter of Lithic Technology, Spe-
or are chisel or wedge-like tools. Microwear analy- cial Publication 1.)
sis has suggested the latter (Ataman 1989, 208-10) Tixier, J., 1984. Le debitage par pression, in Prenistoire de Ia
Pierre Taillee, 2: Economie du Debitage Laminaire:
yet to differentiate these objects from examples
Technologie et Experimentation. Paris: CNRS, 57-70.
which have less equivocal chisel or wedge like Wilke, P. & L. Quintero, 1994. Naviform core-and-blade tech-
functions, the term has been retained. nology: assemblage character as determined by
3. The difficulty in distinguishing between true pro- replicative experiments, in Neolithic Chipped Stone In-
jectiles and bifaces argues for a common term. dustries of the Fertile Crescent, eds. H.G. Gebel & S.K.
Not included in this class are pieces which Kozlowski. Berlin: Ex Oriente, 33--60.
198
Chapter 11

Surface Material: Animal Bone and Worked Bone

Louise Martin & Nerissa Russell

Surface material: animal bone. ~atalhoyiik East Figure 11.1 shows how the total number of bone
Louise Martin fragments are distributed across the 298 squares sam-
pled (distribution by bone weight shows a very simi-
Animal bone material from surface collections of lar picture; hence only counts are discussed).
archaeological sites has understandably received lit- Relatively higher densities of bone are seen on the
tle attention since being a 'natural' category it can- northern eminence, on the southern part of the east-
not provide chronological or typological information. ern eminence, and on the northern and western slopes
Although surface bone material is clearly of less value of the main mound. Lower counts are seen in a halo
than pottery or chipped stone artefacts for elucidat- surrounding the northern eminence and on the south
ing patterning, the <;atalhoyiik assemblages are and southeast slopes of the main mound.
briefly described here for the possible information This patterning is extremely difficult to explain
they yield on site formation processes. An additional since numerous factors potentially affect variation.
problem with bone is that it is fragile and more Firstly, the animal bone may be predominantly
subject to weathering than the other artefact catego- Neolithic, with original variation in its spatial and
ries described, and hence more likely to have under- temporal distribution. In this case, it may either be
gone physical change since its deposition. eroding out of the mound, representing underlying
deposits, or derived from deposits now eroded away.
Methodology Alternatively, bone may be both Neolithic and later,
Bone fragments from both the 2 x 2 m surface collec- having been deposited during subsequent periods
tion and the 10 x 10 m sub-surface scraped areas of activity also, which are evidenced by the Classical
were counted, weighed and placed into size catego- period sherds covering the site. Bone might also have
ries. The following skeletal elements/zones were re- been buried or churned-up by post-Neolithic dig-
corded as identifiable: ging, either for the interring of burials or for the
• skull: occipital condyles, petrous temporal, horn extraction of brick-material from pits (both witnessed
cores, maxillary teeth (if >half tooth present); at the site). Finally, a whole sequence of post-occu-
• mandible: horizontal ramus and condyle, mandibu- pational erosional and depositional events may have
lar teeth (if >half tooth present); influenced any surface patterning. A detailed his-
• all long bones; tory of the bone material therefore appears impossi-
• scapula: articular end and neck only; ble to unravel, but comparison with the distribution
• pelvis: acetabulum only; of other finds, and consideration alongside site to-
• tarsals: astragalus, calcaneum, naviculo-cuboid; pography allows certain possibilities to be narrowed
• phalanges; down.
• vertebrae: atlas and axis only. Bone patterning does not appear explicable in
terms of erosion and deposition of this material alone:
Surface collection there is little correlation between bone density, el-
The bone from the systematic 2 x 2 m surface collec- evation (contours) or angle of slope. Also, higher bone
tion was generally extremely fragmented, with most densities are not consistently associated with slopes
pieces measuring less than 5 em in length, and the of particular orientations, which might be expected
percentage of identifiables very low (0.02 per cent). if surface run-off was a prime source of exposing

199
Louise Martin & Nerissa Russell

900 1000 1100 1200


metres east

0 50 100 150 200 250 300


number of fragments

Figure 11.1. The number of bone fragments collected from each 2 x 2 m square on the east mound.

200
Surface Material: Animal Bone and Worked Bone

homogenous underlying deposits. Clearly, factors the next most common category being the unidenti-
beyond erosion, run-off and deposition are at play. fiable medium-sized herbivores (although consider-
A striking 'negative' pattern is seen between ing the other taxa, these are likely to be sheep I goat
bone density and the number of wheelmade sherds also). The spatial distribution of these identifiables
(representing Classical period activity) (Fig. 11.2). shows no patterning, but appears to relate to the
Wheelmade sherd counts are high to the south and overall densities of bone fragments (e.g. Fig. 11.4,
southeast of the main mound, and in a halo around showing the distribution of medium-sized herbivore
the northern eminence - exactly where bone den- bones).
sity is lowest. This pattern implies that a heavy pres-
ence of Classical sherds either obscures or swamps Sub-surface scrape
bone remains (although it should be noted wheel- Finds from the sub-surface scraped areas were col-
made sherds are relatively abundant all over the lected by hand (i.e. deposits were not sieved) mean-
site). If Classical sherds are obscuring bone, it might ing that there was potentially little consistency in
be suggested that it overlies it, meaning that much of retrieval, and hence results should be treated with
the bone would be Neolithic in origin. caution.
These same areas of high Classical sherding Table 11.2 shows the animal bone data from all
and low bone density also see relatively high pro- units in each of the 10 x 10m2 scraped areas, and the
portions of bone fragments <2 em in length, i.e. the totalled results for each 10 x 10m square. Units have
smaller bone fraction (Fig. 11.3). If these areas gener- been divided into four categories, in the same way
ally see both less bone and also smaller pieces, the as for the pottery analysis: surface collection (spit 0),
high densities of Classical sherds (or high incidence topsoil removal (spit 1), second spit of cleaning (spit
of Classical period activity) may have had the effect 2), and further scraping (spit 3). In general, spit num-
of abrading or destroying the animal bone which bers increase with depth of scraping. Data are shown
appears on the surface. Another possibility is that a for the number of bone fragments collected from
high density of Classical sherds may have created a each unit, the total weight of these fragments, the
protective scattering of material over the Neolithic average weight per bone fragment, the number of
deposits, meaning that less is eroding out to be col- fragments per m 3 excavated, and the proportions of
lected. Either of these suggestions could explain the identifiables in each unit.
patterning shown. A first observation is that the average weight of
The distribution of the total number of lithics bone fragments generally increases with greater
over the 2 x 2m squares (Fig. 10.9) shows a similar depth from the surface. If weight of bone fragment is
pattern to the density of bone fragments, and might roughly equated with fragment size, then this indi-
also support the view that the majority of bone re- cates that bone is better preserved in the lower units-
mains are Neolithic. Since the lithic pieces (predomi- as are sherds -and that material in the upper units
nantly obsidian) are assumed to be of Neolithic date, is more affected by adverse preservational condi-
it is possible that lower counts also represent an tions (e.g. weathering, soil movement, trampling).
obscuring by later Classical period activity. The total densities of bone collected from each
If it is accepted that much of the animal bone is 10 x 10m square (the number of fragments per m 3
associated with the Neolithic occupation of the site, excavated) are shown spatially in Figure 11.5. Den-
then the broad-scale variations in density might ap- sity varies greatly across the site, ranging from two
pear to result from later activity. The aim then is to fragments per m 3 in square 1000,950 on the south
explain the distribution of Classical period sherds slope of the main mound, to 114 fragments per m 3 in
over the entire site, with a great concentration on the square 1040,1040 on the east slope of the main mound.
south and southeast slopes, when it appears that Even on the much flatter 1600 m 2 area scraped on the
Classical activity is more restricted (primarily to the northern eminence, great differences in bone densi-
top of the main mound, although there are also fea- ties can be seen between adjacent squares. The pot-
tures on the northern eminence). The pattern repre- tery from the scraped squares- being consistently a
sents a huge dispersal of Classical sherds across the
whole site which might cast doubt on the meaning Table 11.1. The number of animal bone fragments identified to each
of any surface patterning. taxon/category from the 2 x 2 m surface collection.
Returning to the animal bones, a total of 127
Cattle Large herbivore Sheep/goat Medium herbivore Pig/Boar Fox Total
were identified to taxon (see Table 11.1), the major-
10 5 61 43 5 127
ity belonging to sheep I goat (both identified) with

201
Louise Martin & Nerissa Russell

Fig 11:2 Number of wheelmade sherds

900 1000 1100 1200


metres east

0 25 50 75 100 125 150 175 200 225


number

Figure 11.2. The number of wheel made sherds collected from each 2 x 2 m square on the east mound.

202
Surface Material: Animal Bone and Worked Bone

1300

1200

s.... 1100
§
ifj
t;j

1000

900

800

900 1000 1100 1200


metres east

0 10 20 30 40 50 60 70 80 90 100
percentage

Figure 11.3. The percentage of bone fragments <2 em length of the total number of fragments in each 2 x 2 m square
on the east mound.
203
Louise Martin & Nerissa Russell

1300

1200

o£i 1100
g
~
s
1000

900

800

900 1000 1100 1200


metres east

0 1 2 4 5 6 7
number

Figure 11.4. The distribution of identifiable medium-sized herbivore bones (including those identified to sheep/goat)
on the east mound.
204
Surface Material: Animal Bone and Worked Bone

Table 11.2. (.atalhOyUk East scraped squares, animal bone data (ordered north to south, west to east).

Square Unit Spit No. Weight Av.gl m' Frags./ %id %id Square Unit Spit No. Weight Av.g/ m' Frags./ %id %id
frags. (g) frag. dug m3 (no.) (wt) frags. (g) frag. dug m' (no.) (wt)

294 52 286 5.5 21 30 900 19 66 3.5 5 8


295 58 389 6.7 38 38 901 148 775 5.2 30 39
296 7 64 9.1 43 63 1030,1160 total 167 841 12 13.9
297 7 23 3.3 43 57
298 16 56 3.5 25 55 903 0 21 213 10.1 38 76
1040,1240 total 140 818 5.8 20 904 421 2700 6.4 20 43
1030,1150 total 442 2913 6.6 12 36.8
817 0 44 69 1.6 7
818 428 2350 5.5 29 187 0 56 96 1.7 14 13
1020,1200 total 472 2419 5.1 6 78.7 188 1 341 925 2.7 12 29
189 2 82 194 2.4 11 18
819 0 20 18 0.9 20 39 192 3 28 88 3.1 14 42
820 1 78 525 6.7 14 35 193 3 107 370 3.5 15 32
1030,1200 total 98 543 5.5 16.3 1040,1140 total 614 1673 2.7 12.3 49.9

823 30 74 2.5 10 18 853 212 547 2.6 16 23


824 76 820 10.8 20 51 854 587 5290 9 15 28
1040,1200 total 106 894 8.4 11.8 1040,1120 total 799 5837 7.3 20 40

809 0 28 57 2 7 19 500 118 273 2.3 11 34


810 91 685 7.5 31 47 501 139 1026 7.4 22 21
1050,1200 total 119 742 6.2 7.3 16.3 502 1 487 2530 5.2 10 31
527 2 399 3550 8.9 13 21
821 0 29 179 6.2 17 16 980,1090 total 1143 7379 6.5 11 104
822 396 2670 6.7 17 37
1020,1190 total 425 2849 6.7 8.3 51.2 289 0 24 140 5.8 25 60
290 1 385 3050 7.9 20 40
280 0 38 220 5.8 21 27 291 2 88 700 8 33 56
281 1 599 4350 7.3 18 30 990,1090 total 497 3890 7.8 15.6 32
282 2 9 102 11.3 33 48
1030,1190 total 646 4672 7.2 10.9 59.2 274 0 28 107 3.8 11 25
278 1 243 1235 5.1 9 24
263 112 490 4.4 8 20 279 2 122 875 7.2 15 39
264 841 4220 5 10 31 1040,1090 total 393 2217 5.6 20.3 19.4
265 71 640 9 21 28
1040,1190 total 1024 5350 5.2 10.9 93.9 503 0 5 21 4.2 0
504 226 2860 12.7 17 35
811 0 37 138 3.7 19 38 980,1080 total 231 2881 12.5 18.8 12
812 340 1640 4.8 15 50
1050,1190 total 377 1778 4.7 6.8 55.4 505 2 8 4 0 0
506 27 167 6.2 19 54
825 12 153 12.8 50 56 601 594 4520 7.6 14 37
826 214 2080 9.7 19 48 990,1080 total 623 4695 7.5 21.8 28.6
1020,1180 total 226 2233 9.9 9.5 23.8
155 0 0 0 0 0
286 11 50 4.5 0 0 156 172 1210 7 13 38
287/288 206 2200 10.6 13 85 191 205 1520 7.4 24 47
1030,1180 total 217 2250 10.4 10.9 19.9 940,1040 total 377 2730 7.2 19.9 19

283 53 460 8.7 13 46 12 2 60 232 3.8 22 35


284 209 1340 6.4 12 37 24 2 14 75 5.4 38 28
285 0 0 34 2 8 44 5.5 50 52
1040,1180 total 262 1800 6.9 10.9 24 35 2 16 57 3.6 25 26
990,1040 total 98 408 4.2 13.2
813 0
814 239 1734 7.3 10 65 61 0 71 390 5.5 14 23
1050,1180 total 239 1734 7.3 6.3 37.9 67 1 318 1780 5.6 14 34
124 2 476 2840 6 15 38
827 6 53 8.8 0 0 269 3 2267 14,460 6.4 15 36
828-52 1,2 560 3319 5.9 17 46 1040,1040 total 3132 19,470 6.2 27.5 114
1020,1170 total 566 3372 11.8 48
195 0 16 102 6.4 19 12
708 7 21 0 0 196 1 164 823 5 10 45
709 15 137 9.1 13 21 197 2 26 206 7.9 15 32
710 190 1640 8.6 17 39 277 3 210 890 4.2 10 26
1030,1170 total 212 1798 8.5 14.8 14.3 1090,1040 total 416 2021 4.9 19 22

807 0 0 0 0 0 602 0 10 10 1 0 0
808 59 460 7.8 32 52 603 51 299 5.9 20 73
1040,1170 total 59 460 7.8 11.3 5.2 604 7 50 7.1 43 82
990,990 total 68 359 5.3 5.3 12.8
815 0 23 82 3.6 13 24
816 53 347 6.5 17 16 705 0 0 0 0 0
1050,1170 total 76 429 5.6 10.9 706 1 20 46 2.3 15 80
707 2 11 39 3.5 18 36
905 0 38 126 3.3 7 1000,950 total 31 85 2.7 20.6 1.5
906 137 1170 8.5 27 58
1020,1160 total 175 1296 7.4 12 14.5

205
Louise Martin & Nerissa Russell

940 950 960 970 980 990 1000 1010 1020 1030 1040 1050 1060 1070 1080 1090 1100
1240

1230
G
1220
1210
1200
79 16 12 16
1190
51 59 94 55
1180
24 20 24 38
1170
48 14 5 11
1160
15 14
1150
37
1140
50
1130
1120

1110
1100
1090
104 32
1080
12 29
1070
1060
1050
1040

1030
Q
1020
1010
1000
990

980
970
960
950

Figure 11.5. (:atalhOyiik East; scraped square animal bones- fragments per m3 •

mixture of Neolithic and Classical period sherds- densities would be expected if surface movement
shows that few of these deposits are in situ. How- factors were entirely responsible for the bone
ever, the great variations in bone densities in adja- patterning.
cent squares might be seen as reflecting some In short, while the bone and sherd fragmenta-
underlying (or previously overlying) spatial varia- tion analyses plus the pottery typologies suggest
tion of some sort, since perhaps a more even cline of great disturbance and mixing in the scraped deposits,

206
Surface Material: Animal Bone and Worked Bone

it does not seem that the bone patterning is entirely as being highly burnt. While accepting that all ar-
the creation of post-depositional factors. Material chaeological deposits can have complex relationships
has very likely moved from its original location, but with their associated artefacts, these results might
its current placement and patterning would not ap- suggest that the bone has been subject to some dif-
pear to be solely due to topographic and erosional ferent processes than the deposits in this square, and
factors. hence may not relate directly to the deposits or ar-
It is interesting to note that, as with the 2 x 2 m chitectural plans exposed.
squares, bone densities again do not correlate well
with those of pot (where highest densities are seen Identifiables
on the summit of the main mound). Neither do they The scraped squares have been grouped into 13 broad
correlate with the highest proportions of Neolithic areas (see Fig. 9.7), but following the pottery analy-
sherds which are seen on the northern eminence. sis, the identifiable animal bones have been analyzed
as 11 assemblages. Table 11.3a shows the number of
Test square 1020,1170 fragments identified to taxon/ category for each of
In square 1020,1170 an attempt was made to gauge these assemblages; Table 11.3b shows the same data
the relationship between the bones collected and the as percentages. Sheep I goat bones predominate in
architectural plans and deposits which were exposed. each area, and considering the counts of other taxa,
Here, material was collected in 2 x 2 m units, unlike the majority of bones in the medium-sized herbivore
the other 10 x 10m squares where finds from across category probably belong to sheep I goat also. The
the scraping were collected together (although in ratio of identifiable sheep:goat bones is approxi-
vertical spits). It was therefore possible to relate the mately 5:1, from a sample of 185 identified bones
location of the finds more closely to the deposits and (following the criteria of Boessneck 1969 and
features. Firstly, deposits were recorded as either Prummel & Frisch 1986). Cattle bones are the next
intensely burnt, generally ashy or unburnt. The per- most frequent in all areas, with the bones of equids,
centage of burnt bone fragments within each 2 x 2 m pig/boar, deer, dog/wolf, fox, small cat, hare and
unit was then calculated, with the expectation being birds being present in smaller numbers. Percentages
that high proportions of burnt bone would be found of taxa vary somewhat between areas, but without
in the units with intensely burnt deposits. fuller understanding of depositional and post-
The most intensely burnt deposits were in the depositional processes, the picture is impossible to
southeast of the square in the 2 x 2s with x,y co- interpret. Further meaning may be made of the
ordinates A,B; A,C; B,C (see Fig. 11.6). Figure 11.6 patterning and variation if these areas are excavated
shows that while there are burnt bone fragments in in future.
two of these units- A,B has 11 per cent and A,C has As would probably be expected, some bone has
14 per cent - there are higher proportions in other clearly been lost from the record in the scraped de-
units (e.g. E,E with 25 per cent burnt bones) where posits due to taphonomic factors. A comparison of
deposits are not burnt. In addition, there are no burnt counts of proximal and distal ends of humeri for
fragments in unit B,C where deposits were recorded both cattle and medium-sized herbivores and tibia
for medium-sized herbivores only (Table 11.4) shows
E 11 0 0 6 25 that in each case, the less dense proximal ends (cf.
Lyman 1994, table 7.6, 246-7) are fewer. The most
D 0 0 0 7 0 likely explanation for this is that these proximal ends
(along with other less dense bone parts) suffer more
c 14 0 13 0 0 from a suite of destructive factors including weath-
ering, abrasion and animal gnawing.
B 11 18 0 0 0
c;atalhoyiik West
A 39 11 0 11 7
Surface collection
A B c D E Figure 11.7 shows the distribution of the 1851 bone
fragments collected from the surface survey. Higher
Figure 11.6. The percentage of burnt bone fragments for bone densities are seen on the northern and eastern
each 2 x 2m unit in scraped square 1020,1170; the parts of the mound, with low densities on the south-
letters represent x,y coordinates (north faces upwards). ern and northeastern slopes. The average number of

207
Louise Martin & Nerissa Russell

1100

1000

900

800

400 500 600 700


metres east

il![illililflff~ I
I I I I I I I I I I I I I I
0 10 20 30 40 50 60 70 80 90 100
number

Figure 11.7. The number of bone fragments collected from each 2 x 2 m square on the west mound.

208
Surface Material: Animal Bone and Worked Bone

Table 11.3a. The number of identified fragments per taxon/category for each scraped square assemblage.

Area Cattle Equid Lg. herb. Red deer Fallow Deer Pig Sh./Gt M. herb. Roe Dog/Wolf Fox Cat Hare Bird Total
1 3 1 1 0 0 0 1 6 8 0 0 0 0 0 0 20
2 156 48 94 4 0 8 17 287 101 0 9 3 1 1 1 730
3,4 32 5 8 0 0 2 0 147 33 0 7 1 0 0 1 236
5 34 7 7 0 0 0 1 63 10 0 0 0 0 0 0 122
6 15 1 5 0 0 0 0 18 1 0 0 0 0 0 0 40
7 108 12 28 6 1 3 12 156 55 4 6 2 0 0 0 393
8,9 24 6 1 0 0 0 6 48 10 0 0 0 0 1 0 96
10 114 25 29 5 0 2 9 197 31 2 14 4 0 0 2 434
11 7 3 2 0 0 2 0 16 4 0 0 0 0 0 0 34
12 4 1 1 0 0 0 0 5 4 0 0 0 0 0 0 15
13 1 0 0 0 0 0 0 4 2 0 0 0 0 0 0 7

Total 498 109 176 15 1 17 46 947 259 6 36 10 1 2 4 2127

Table 11.3b. The relative proportions(%) of taxa for each scraped square assemblage.

Area Cattle Equid Lg. herb. Red deer Fallow Deer Pig Sh./Gt M. herb. Roe Dog/Wolf Fox Cat Hare Bird Total

1 15 5 5 0 0 0 5 30 40 0 0 0 0 0 0 100
2 21 7 13 0.5 0 1 2 39 14 0 1 0.5 0 0 0.5 99.5
3,4 14 2 3 0 0 1 0 62 14 0 3 0.5 0 0 0.5 100
5 28 6 6 0 0 0 1 52 8 0 0 0 0 0 0 101
6 38 2.5 12.5 0 0 0 0 45 2.5 0 0 0 0 0 0 100.5
7 27 3 7 2 0 1 3 40 14 1 2 1 0 0 0 101
8,9 25 6 1 0 0 0 6 50 10 0 0 0 0 1 0 99
10 26 6 7 1 0 0.5 2 45 7 0.5 3 1 0 0 0.5 99.5
11 21 9 6 0 0 6 0 47 12 0 0 0 0 0 0 101
12 27 7 7 0 0 0 0 33 27 0 0 0 0 0 0 101
13 14 0 0 0 0 0 0 57 29 0 0 0 0 0 0 100

Table 11.4. Counts of proximal and distal humeri for cattle and Table 11.5. The number of identifiable animal bone fragments per
medium-sized herbivores, and proximal and distal tibiae fragments taxon/category from the west mound 2 x 2 m surface collection.
for medium-sized herbivores only, from all scraped deposits.
Cattle Large Deer Sheep/ • Medium Dog/ Fox Hare Squirrel Bird Total
Bone element Cattle Medium-sized herbivore herbivore Goat herbivore Wolf

Proximal humerus 1 12 2 55 74 2 4 147


Distal humerus 10 96
Proximal tibia 9
Distal tibia 119

Table 11.6. (.atalhoyiik West scraped squares, animal bone data.

fragments collected per 2 x 2 m square is 8.7, as Square Unit No. Weight Av.g/ m' Frags./ %id %id
frags. (g) frag. dug m' (no.) (wt)
opposed to an average of 23.9 fragments per square 855 68 215 3.2 28 60
from the east mound: <;atalhoyiik East, therefore, 856 147 1198 8.1 37 59
480,1040 total 215 1413 6.6 18.8 11.4
has a higher density of animal bone fragments ap-
857 1 11 11 100 100
pearing on the surface. 858 89 933 10.5 43 64
A total of 147 fragments was identified to taxon 580,1020 total 90 944 10.5 24.5 3.7

(Table 11.5). Large herbivore bones (including cattle, 859 11 31 2.8 45 32


69 500 7.2 43 54
equids and deer) constitute 6 per cent of the 860
640,960 total 80 531 6.6 19.3 4.1
identifiables, while on the east mound they make up
12 per cent of the total. If potential taphonomic vari-
ation is discounted, the surface assemblage from the Table 11.7 presents data for the identifiables
east mound has a greater large herbivore component from each unit. As with the east mound, sheep I goat
than that from the west mound. predominate, with cattle the next most common. Sam-
ple sizes are very small, which probably explains the
Sub-surface scrape lower diversity of taxa identified. The difference in
Table 11.6 shows the data for each 10 x 10m scraped the number of identifiables between the west and
unit, and the totals for each of the three scraped east mounds does not allow proper comparison be-
squares. As for the east mound, the average weight tween the two, but a cursory scan does not appear to
of fragment increases in the lower units, perhaps bear out the suggestion of the 2 x 2 m surface
showing the upper units to have suffered more from identifiables - that there is a higher proportion of
destructive factors. large herbivore bones on the east mound than west.

209
Louise Martin & Nerissa Russell

Surface collections: worked bone The single example from the surface sieving is
Nerissa Russell & Louise Martin an indeterminate point worked on the long bone
shaft of a medium-sized herbivore. It is from unit 3
c;atalhoyiik East on the summit of the main mound.
Of the 41 pieces from the scraped squares, 22
A small collection of worked bone pieces derives are either points or what appear to be base frag-
from the surface collections from <;atalhoyiik East. ments of points; another 14 are ground astragali;
A single piece was found during the systematic and the remaining five are each single examples of
sieving of the surface deposits, and another 41 are other tools or objects (Table 11.8). These pieces are
from the 10 x 10 m sub-surface scraped areas, where of unknown date, and are likely to be from mixed
objects were retrieved by 'hand collection'. deposits.
Table 11.7. The number of identified fragments per taxon/category for each scraped square unit.

Unit Cattle Equid Lg. herb. Red deer Fallow Deer Pig Sh./Gt M.herb. Roe Dog/Wolf Fox Cat Hare Bird Total
855 2 0 3 0 0 I 0 7 6 0 0 0 0 0 0 19
856 14 0 2 0 0 I 0 28 10 0 0 0 0 0 0 55
857 I 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
858 10 2 0 0 0 I I 16 7 0 0 0 I 0 0 38
859 0 0 0 0 0 0 I 2 I 0 0 0 0 0 0 4
860 4 I 4 0 0 0 I 12 7 0 0 0 0 0 I 30
Total 31 3 9 0 0 3 3 65 31 0 0 0 1 0 1 147

Table 11.8. c;atalhOyiik East scraped squares, worked bone.

Unit Area Type Taxon Element Description


294 I ground astragalus sheep astragalus dorsal facets ground to partially flatten
295 I point- indeterminate med. herbivore long bone split?; lacks base
818 2 point standard
0
sheep/goat metatarsal - proximal split
818 2 point- indeterminate sheep/goat tibia unsplit; lacks base
280 2 point- standard sheep metatarsal -distal unsplit; unmodified base
281 2 point- tibia with ground base sheep tibia · distal unsplit; distal end ground to partially flatten
281 2 point- heavily ground metapodial sheep/goat metapodial distal0
split; base modified on all sides
281 2 point heavily ground metapodial
0
sheep/goat metapodial distal0
split? base modified on all sides
264 2 point- indeterminate med. herbivore metacarpal - shaft split; lacks base
264 2 point standard
0 med. herbivore metapodial - distal split; lacks base
264 2 point- standard sheep/goat metacarpal - proximal unsplit; unmodified base
264 2 burnisher lg. herbivore rib one end ground to rounded bevel
825 2 point - standard sheep metatarsal split; unmodified base
826 2 drilled skull fragment cattle skull · frontal fragment with three evenly spaced drilled holes
709 2 point standard
0 sheep/goat metatarsal - proximal split; unmodified base
816 2 point - standard med. herbivore metapodial split; unmodified base
904 3 point standard
0 sheep metacarpal distal
0 unsplit; unmodified base
904 3 point standard
0
sheep metatarsal- distal split; slightly modified base
853 5 ground astragalus sheep/goat astragalus dorsal facets ground to partially flatten
853 5 point - standard sheep tibia distal
0
unsplit; unmodified base
854 5 figure/spatula? lg. herbivore rib carved to represent human figure? spatula?
854 5 point · heavily ground metapodial sheep/goat metatarsal -distal split; base modified on all sides
854 5 point - standard sheep metatarsal -distal split; unmodified base
854 5 point - standard sheep/goat tibia distal
0 unsplit; unmodified base
854 5 ground astragalus sheep/goat astragalus dorsal facets ground to partially flatten
854 5 ground astragalus sheep astragalus dorsal facets ground to partially flatten
500 7 point heavily ground metapodial
0
sheep/goat metatarsal- distal split? base modified on all sides
501 7 ground astragalus sheep/goat astragalus dorsal facets ground to partially flatten
502 7 ground astragalus sheep/goat astragalus dorsal facets ground to partially flatten
502 7 ground astragalus sheep/goat astragalus dorsal facets ground to partially flatten
290 7 ground astragalus sheep/goat astragalus dorsal facets ground to partially flatten
290 7 ground astragalus sheep astragalus dorsal facets ground to partially flatten
290 7 antler plaque red deer antler split, shaped into flat piece; bitumen on edges
290 7 tool/waste? red deer antler tine tine fragment, cut at one end
291 7 ground astragalus sheep/goat astragalus dorsal facets ground to partially flatten
504 7 point - tibia with ground base sheep/goat tibia distal
0
unsplit; distal end ground to partially flatten
601 7 ground astragalus fallow deer astragalus dorsal facets ground to partially flatten
61 10 ground astragalus sheep astragalus dorsal facets ground to partially flatten
67 10 ground astragalus cattle astragalus dorsal facets ground to partially flatten
124 10 ground astragalus sheep/goat astragalus dorsal facets ground to partially flatten
269 10 point · heavily ground metapodial sheep/goat metatarsal- distal unsplit; base modified on all sides (and shaft)

Table 11.9. c;atalhOyiik West scraped squares, worked bone.

Unit Area Type Taxon Element Description


856 figurine I ornamental? lg.mammal long bone carefully carved and incised shape
856 point · tibia with ground base sheep/goat tibia . distal unsplit; distal end heavily ground to flatten
860 ground astragalus fallow deer astragalus dorsal facets ground to partially flatten

210
Surface Material: Animal Bone and Worked Bone

Points Ground astragali


As is typical of prehistoric assemblages, points are Thirteen astragali have been ground to flatten the
the most common bone tool type among the dorsal surface slightly; in one case, the ventral and
c;:atalhoyiik surface assemblage. All of the points are proximal surfaces are also very slightly ground. Most
made on medium mammal (sheep-sized) long bones. of these are sheep I goat astragali (sheep in at least two
Where taxon can be determined, all are sheep I goat, cases) (e.g. Fig. 11.8:4), but there is also one cattle and
and in at least seven cases sheep. Most (16) are made one fallow deer example. Although a few are ground
on metapodials, while five are made on tibiae and quite deeply, most bones are only slightly modified.
one on an indeterminate long bone. The articular Such ground astragali are known from many
end seems invariably to have been left on the tool to sites in both Old and New Worlds from the Upper
act as the base. Almost all have fairly slender tips Palaeolithic onward, e.g., Remouchamps (Belgium,
and thus must have been used in tasks requiring Upper Palaeolithic: Dewez 1974), Selevac (Yugosla-
relatively fine perforations. via, Middle-Late Neolithic: Russell 1990), Hacilar
Where bone is an important raw material, long (Turkey, Late Neolithic: Mellaart 1970), Troy III (Tur-
bones are often split longitudinally to achieve two or key, Early Bronze Age: Schliemann 1968 [1881]),
even four points from a single bone. The proportion Aphrodisias (Turkey, Bronze Age: Joukowsky 1986),
of splitting among the c;:atalhoyiik surface points (56 Sibri (Bronze Age, Pakistan: Russell1995b), Gordion
per cent) is substantially lower than in certain (Turkey, Classical: Sheftel1974). They are generally
Neolithic assemblages studied from the Balkans and considered to be used as 'dice' in gambling games
south Asia (e.g. Russell1990; 1995b). Mellaart (1967) ('knucklebones') or divination. This function is at-
refers to thousands of bone points from the earlier tested ethnographically in a number of regions, in-
c;:atalhoyiik excavations, so bone was clearly heavily cluding Turkey (Brewster 1960; Watson 1979).
utilized. Assuming that the surface tools are repre-
sentative of those in secure contexts, it would appear Other bone artefacts
that there was little pressure on bone as a raw mate- In addition to the points and ground astragali, a few
rial: sheep/goat metapodials were readily available other specimens of worked bone were recovered from
to those who needed them. the surface.
The c;:atalhoyiik surface bone points do not di- One large mammal (cow-sized) rib has one end
vide readily into subtypes, reflecting a relatively low ground to a rounded bevel and polish over all of its
degree of alteration of the shape of the original bone. surface, especially along one edge (Fig. 11.8:5). The
Three subtypes, however, may be distinguished on polish is not derived from contact with ceramics, but
the basis of the modification of the base (three frag- a soft substance. It is likely to have been used as a
mentary specimens lacking bases cannot be classi- burnisher for hides or similar materials. The bevel
fied): appears to postdate much of the tool's use and prob-
ably represents repair after it broke in use. Rib
Standard points burnishers are found in most prehistoric bone tool
Most points (12) fall into this category, where the assemblages, and Mellaart (1967,215) mentions such
articular end of the bone is unmodified except by tools from the 1960s' excavations at c;:atalhoyiik.
splitting where this occurs. Most are made on Two enigmatic tool fragments made on red deer
metapodials, proximal (e.g. Fig. 11.8:1) or distal (e.g. antler were found in the same unit (290 on the main
Fig. 11.8:2), but two are on unsplit tibiae. mound). One is a flat, probably deliberately split,
piece of antler beam that appears to have been shaped
Tibia points with ground bases roughly into a rounded rectangle (a piece is miss-
Two points are made on unsplit distal tibiae, where ing). Possible traces of bitumen on the edges raise
the distal articular surface has been ground to par- the possibility that it was part of a haft. The other is a
tially flatten it. piece of a large tine with traces of cutting at one end.
It may be a fragment of a tool, or waste from the cut-
Heavily ground distal metapodial points and-break method of removing tines and sectioning
Five points have been heavily abraded at the base so antlers preparatory to tool manufacture.
that the anterior, posterior, and often the distal sur- Another curious artefact is a fragment of cattle
faces of the articulation are substantially flattened skull from the frontal region with three evenly spaced
(e.g. Fig. 11.8:3). Two are split longitudinally, one is perforations drilled along a broken edge. It is possi-
unsplit, and the others are too fragmentary to tell. ble that this is a preform for a perforated bone plaque

211
Louise Martin & Nerissa Russell

em

,. - I

1 \};,-

3
- 4

,Cf~lll
'-

7

Figure 11.8. Worked bone: 1) a standard point made on the proximal end of a sheep/goat metacarpal (unit 264); 2) a
standard point made on the distal end of a sheep metatarsal (unit 825); 3) a heavily ground distal metapodial point
made on the distal end of a sheep/goat metatarsal (unit 269); 4) ground sheep/goat astragalus with dorsal facets ground
partially flat (unit 290); 5) burnisher made on rib of large herbivore (unit 264); 6) worked bone figure/spatula? made
on rib of large herbivore (unit 854); 7) part of carefully worked bone figurine or ornamental piece from (-atalhoyuk
West (unit 856)_

212
Surface Material: Animal Bone and Worked Bone

such as were found in the earlier excavations at Alternatively, these distinctions could support
<;atalhoyi.ik (Mellaart 1967), at Hacilar (Mellaart a suggestion that different areas of the site saw dif-
1970), and at Selevac (Russell 1990). Mellaart de- ferent activities. The worked bone from areas 5, 7 and
scribes some of these artefacts as archers' wrist guards 10 do not reflect a full spectrum of daily subsistence
or decorative pieces sewn onto leather belts, although activities. There are relatively few standard points
they may have served other functions, as rigid (which also explains the differences in treatment of
heddles in weaving for instance, or been ornamental bone for points from area 2/3), and hence tasks in
in different ways. which points were used may have differed from the
Finally, a carved piece of large mammal rib northern eminence. The concentration of ground
may represent a stylized human figure (Fig. 11.8:6). astragali in areas 5, 7 and 10 might suggest that they
It could conceivably be the handle of an ornate were used in divination, as in certain African socie-
spatula, although it differs in style from those recov- ties in the ethnographic present (de Heusch 1985).
ered from the earlier excavations at <;atalhoyi.ik The high degree of uniformity as well as the
(Mellaart 1967), or from the Late Neolithic and spatial concentration of the ground astragali strongly
Chalcolithic levels at Hacilar (Mellaart 1970). suggests that either they derive from a restricted
time period (given the presence of one on the West
Spatial patterning and chronology (see Fig. 9.7) mound (see below) this is likely to be Chalcolithic or
The most notable spatial patterning in the bone tool later), or else were used in a well-defined behav-
assemblage is the absence of ground astragali from ioural context that enforced uniformity through time,
the 1900 m 2 scraped area on the northern eminence as might be the case with divination. Likewise, the
(areas 2/3). With the exception of a single specimen fact that two out of three bone artefacts from
from the northernmost square scraped (area 1), all <;atalhoyi.ik West are typologically identical to mul-
the ground astragali from <;atalhoyi.ik East come from tiple specimens from the east mound indicates ei-
further south (areas 5, 7 & 10). Points, on the other ther persistence of these types through considerable
hand, while found in all areas where worked bone lengths of time (implying substantial continuity of
was present, are common in areas 2/3 on the north- occupation or at least population between the two
ern eminence. The ground tibia and metapodial mounds) or the presence of a Chalcolithic or later
points are quite evenly distributed across the scraped occupation in the surface levels of both mounds.
areas where worked bone was found, but it interest-
ing to note that - if this distinction between areas <;atalhoyiik West
2/3 and the others is maintained- then three-quar-
ters (9 of 12) of the standard points derive from the Three bone artefacts were recovered from surface
former. Moreover, splitting occurs in 10 of the 14 work on the west mound (Table 11.9). There is one
points in areas 2/3, but a lower 4 of 8 points in the tibia point with a ground base similar to those found
other areas combined. An additional observation is on the east mound. Similar points are also character-
that points from areas 5, 7 and 10 seem to divide istic of the Chalcolithic component of the Mehrgarh
more clearly into either expedient or carefully worked bone tool assemblage in Pakistan (Russell 1995b),
categories (two expedient, five carefully worked) than but this may be coincidental. There is also a fallow
those from area 2, although the very small sample deer astragalus ground on the dorsal side, again
size should inspire caution. resembling examples from the east mound. Finally,
These spatial distinctions could be either func- a fragment of a carefully carved and incised piece of
tional or temporal. Bone tools are notoriously poor large mammal long bone (Fig. 11.8:7) may be part of
chronological markers, and the types of points and a figurine or some ornamental piece.
ground astragali found here have a very wide distri-
bution in time and space. One interpretation might References
be that the treatment of bone as a raw material seen
in areas 2/3, with its higher degree of splitting, and
absence of clear categories of expedient and care- Boessneck, J., 1969. Osteological differences between sheep
(Ovies aries Linne) and goat (Capra hircus Linne), in
fully worked points, is typically Neolithic; the lower
Science in Archaeology, eds. D.R. Brothwell & E.S.
incidence of splitting and clear separation of expedi- Higgs. 2nd edition. London: Thames & Hudson,
ent and carefully finished points in the areas to the 331-58.
south is a pattern usually found in later periods Brewster, P., 1960. A sampling of games from Turkey. East
(Russell 1990; 1995a,b). and West 11, 15-23.

213
Louise Martin & Nerissa Russell

de Heusch, L., 1985. Sacrifice in Africa: a Structuralist Ap- Krstic. Los Angeles (CA): UCLA Institute of Ar-
proach. Trans. Linda O'Brien & Alice Morton. chaeology, 521-48.
Bloomington (IN): Indiana University Press. Russell, N., 1995a. Neolithic Relations of Production in
Dewez, M.C., 1974. New hypotheses concerning two en- Southeast Europe and South Asia: Insights from the
graved bones from La Grotte de Remouchamps, Bel- Bone Industry. Unpublished paper presented at the
gium. World Archaeology 5, 337-45. meetings of the Archaeological Institute of America,
Joukowsky, M.S., 1986. Prehistoric Aphrodisias: an Account San Diego.
of the Excavations and Artifact Studies, vol."I: Excava- Russell, N., 1995b. The bone tool industry from Mehrgarh
tions and Artifact Studies. (Archaeologia Transatlantica and Sibri, in Mehrgarh: Field Reports 1974-1985, From
3.) Providence (RI): Brown University, Center for Neolithic Times to the Indus Civilization, eds. C. Jarrige,
Old World Archaeology and Art. J.-F. Jarrige, R.H. Meadow & G. Quivron. Karachi:
Lyman, R., 1994. Vertebrate Taphonomy. Cambridge: Cam- Department of Culture & Tourism, Government of
bridge University Press. Sindh, 583-613.
Mellaart, J., 1967. Catal Hiiyiik: a Neolithic Town in Anatolia. Schliemann, H., 1968 [1881]. Ilios: the City and Country of
London: Thames & Hudson. the Trojans. New York (NY): Benjamin Blom.
Mellaart, J., 1970. Excavations at Hacilar. Edinburgh: Edin- Sheftel, P.A., 1974. The Ivory, Bone and Shell Objects
burgh University Press. from Gordion from the Campaigns of 1950 through
Prummel, W. & H.-J. Frisch, 1986. A guide for the distinc- 1973. Unpublished Ph.D. thesis, University of Penn-
tion of species, sex and body side in bones of sheep sylvania.
and goat. Journal of Archaeological Science 13, 567-77. Watson, P.J., 1979. Archaeological Ethnography in Western
Russell, N., 1990. The Selevac bone tools, in Selevac: a Iran. (Viking Fund Publications in Anthropology
Neolithic Village in Yugoslavia, eds. R. Tringham & D. 57.) Tucson (AZ): University of Arizona Press.

214
Chapter 12

Figurines, Clay Balls, Small Finds and Burials

Naomi Hamilton

The figurines from Mellaart's excavations four seasons of work during the 1960s. A full list of
these is still not available, as it is likely that a number
The aim of this section is not to offer definitive inter- of others are concealed in boxes and bags of bone,
pretations, nor refute those which are current, of the stone and clay artefacts in the store room of Konya
figurines found during Mellaart's excavations. Some Museum. So far I have been able to track down 254
of the material is extremely well known, and this figurines and fragments, some of which are known
territory has been well trodden before me. My inten- only from records, others of which there is no written
tion is to assess the data in a more complete form record. I have been able to examine 181 in some detail,
than has been done previously, and put forward a although a number were behind glass, and a further 47
number of issues and ideas which the figurines sug- only briefly. The majority of those I have not recorded
gest to me. When dealing with material from old in detail are fragmentary animal and humanoid figu-
excavations at which one was not present, a wide rines. Unless prefaced by CHC, record numbers given
range of problems can arise in the areas of documen- in this section all refer to new numbers given by
tation and context. These difficulties are discussed myself for analytical purposes only. These were nec-
briefly below, and will naturally affect the amount essary partly because some figurines had no record
of information that can be drawn from the data, and numbers, and because others shared a group number.
the weight of inference and interpretation that they The data base contains details of the original excava-
can bear. Nevertheless, the material is very rich, and tion numbers and museum inventory numbers for
presents interesting challenges. each figurine, where they exist. Mellaart also men-
tions three of which I have found no trace and suspect
The data may be publication errors - in EVI:5, EVI:23 and
There are several broad types of figurines which are ?VII:22. They have not been included in the data base.
described for simplicity as human, schematic, hu- The information available about the find con-
manoid and animal (Fig. 12.1). However, the term texts is extremely varied- precise locations are given
'schematic' is not used in a strict typological sense, for only 5, fairly precise locations for a further 24,
but rather to stay in touch with Mellaart's usage (see and a couple have contextual information without
Typology below). The term humanoid has been used location. A greater number can be assigned to par-
for one group of schematic forms, a number of which ticular buildings, but these are almost exclusively
have been inventoried (incorrectly in my view) as the human and schematic figurines. The majority of
animals. Although Mellaart does not use the term the others have either just a level, or no record at all
humanoid, I use it when referring to those he la- of their context. This information is presented in
belled ex-voto, to avoid confusion with large-scale Tables 12.1 and 12.2.
schematic figures. In addition there are a number of Most figurines are made of baked or sun-dried
multiple humans, humans with animals, unclear, and clay. However, schematic figures are all of stone, as
unfinished figurines. are the majority of human ones up to and including
Mellaart published the majority of the human level VI. Humanoid and animal figurines are all of
and schematic figurines in varying detail, but these clay with four exceptions. The materials used for sche-
are only a minority of the figurines found during the matic and human figurines are shown in Table 12.3.

215
Naomi Hamilton

545
544

498

532

527
513

Figure 12.1. Examples of Neolithic figurine types from the 1960s' excavations at (:atalhoyiik East; 544, 545) animals;
532) human; 498) humanoid; 527, 513) schematic human . (Reproduced at 75% of original size.)

216
Figurines, Clay Balls, Small Finds and Burials

Table 12.1. Figurines by building. Table 12.2. Schematic and human figurines by level.
rubbish disposal; and
pits cut into the depos-
All:!= 9
Alll:l =2
Level II III IV v VI VII VIII X Unstrat. its in courts. Five ani-
AIII:2 =1 Schematic 0 0 1 0 6 0 0 0 2 or 3 mal figures and one
AV!:l =1 Semi-human 0 0 0 0 4 1 0 0 0
Human 13 5 3 1 22 2 2 or 3 1 8 humanoid are recorded
em= 1
Unclear 2 0 0 0 1 1 0 0 0
to come from outside
E!V:l =1
E!V:4 =2 the east wall of EVI:10,
EV :8 =1 (animal)
EV!:JO =14 Table 12.3. Schematic and human figurines by material and level. which means they were
EV!:25 =4 (3 human, 1 bird) also outside the west
EV!A:30 =1
EV!A:31 =1 (animal)
Level II III IV v VI VII VIII X XII Unstrat. walls of one or more of
EV!:44 =9 (1 animal)
EV!:45 =1
Clay
Stone
11
4
2
3
2
2
1
0
9
25
1
3
2 or 3
0
1
0
0
0
7
4
the three ad joining
EVII:21 =1 buildings (EVI:25, 26 &
EVII:24 =1
EVII:28 =1 27) as the walls abut.
EV!!:45 =1 (animal)
EV!ll:l =1
These outside spaces betwe en walls have been found
EV!ll:25 =1 (humanoid) during our surface work t o be full of a variety of
EV!ll:28 =1 (humanoid)
EX:28 =1 artefacts and probable refus e. The deposition of figu-
EX!l:29 =2 (1 animal, 1 humanoid)
rines in these spaces suggests deliberation, although
this is not clear with other types of material. On the
The issues other hand, they could have been thrown away along
Numbers in context with the rubbish. A seventh figurine, a pebble with
254 is a considerable number of figurines to come incised eyes and mouth, is recorded to have come
from a fraction of one site, and indeed it is a substan- from between the west wall of EVI:10 and its adja-
tial group, although prehistoric Anatolian sites gen- cent storeroom. It is not clear whether this last space
erally do appear to be figurine-rich. Mellaart excavated was genuinely - or conceptually - external, or
more than 200 buildings, and in only 21 are there whether it was regarded as part of building EVI:10
records of figurines. He also excavated a number of as it is within one overall complex. No other figu-
open spaces or courts, some of which contained hu- rines are recorded specifically as coming from spaces
manoid and animal figurines although context is between walls, but Mellaart does state that animal
lacking. Of a total 254 known figurines and frag- and sometimes humanoid clay figures were stuck
ments, 120 are representations of animals (of which into the walls of most of the level VIA shrines in area
two are birds). The figurines which were found in E (1967, 78). Other external spaces have less precise
buildings are largely human and schematic - in- records still- Mellaart reports (1967, 78 & 182) that
deed all but 2 schematic figures come from build- animal figurines with wounds or deliberate damage
ings - but the numbers are small. Out of the 58 were found in pits near buildings VIB:12 and IV:4;
figurines known to have been found in buildings (of and 'a rich deposit of crude clay votive figurines of
which four are animals, and two are unclear), 32 wild animals, ritually broken, crushed or wounded
come from only three buildings (AII:1-9; EVIA:10- with arrow or spear points, was found in a pit in
14; EVI:44-9), while 14 buildings had one each. The building level VIII', immediately west of building
numbers must therefore be broken down by context VII:1 (1962, 51). Apart from these, one figurine has a
before any assessments or interpretations can be 'general level VII' context, one came from a court in
made. It is also clear from these numbers that figu- level V and another came from Court X:28. Current
rines are rare finds in buildings, and more common fieldwork has recovered half a dozen humanoid and
finds in external spaces. animal figurines from a courtyard dating probably
to level VII, and it is likely that many of those with-
Context by numbers out recorded context also came from such open spaces.
It is clear from the limited data we have relating to Internal spaces are recorded more precisely, and
context that the number of figurines varies enor- again there is more than one type of space. Almost
mously among buildings and between buildings and all the figurines found in buildings are human or
external spaces. It is unfortunate that the data are schematic. Storerooms account for several - one in
not more precise, but there are several types of open EIV:4 (577), two in AII:1 (639, 711), one in EVI:10 (591);
spaces in which figurines are found - spaces be- one in A VI:1 (569); while number 635 came from the
tween the walls of adjacent buildings; open areas 'narrow room' of EVI:45, and 585 from the anteroom
referred to by Mellaart as courts, which are used for of EVII:21. The differences between storeroom,

217
Naomi Hamilton

narrow room and anteroom are not explicit. For those this building; added to the fact that it contained the
figurines found within the main rooms of buildings, largest number of figurines found in a single build-
distinct spaces become apparent- seven were found ing, does point to deliberate deposition. On the other
in AII:1 'scattered around the hearth', six were found hand, the discovery of such a large group of figu-
in EVIA:44 on the floor between the east and west rines within the building may have led to more care-
platforms in a space adjacent to the screen separat- ful recording of those found when dismantling the
ing the kitchen area from the main room, but 583 was walls, while others found around 'less important'
found in EVIA:44 on the east platform, and 586 came buildings were given less attention, and their con-
from the east central platform of EVI:10. Spaces may textual information was lost.
also be related to room fittings and furnishings - The presence of 'some intact weapons (lance-
thus 586 was below a niche, above which was a large heads) and numerous clay balls (sling ammunition)'
sculpted cattle head; 583 was below two sculpted in the pits near EVIB:12 and EVI:14, in which figu-
leopards; four figurines plus four concretions were rines portraying wild animals were found (1967, 78),
found in EVI:25 near a 'coarse flat plaster relief' assists Mellaart in viewing them all as part of a hunt-
which was not further described or illustrated; and ing ritual, at the end of which all these items were
figurine 639, from a storeroom attached to AII:1, was deliberately deposited in the pits. The interpretation
found in a grainbin, which may also be the case with of clay balls is pursued further below.
577 (reports in Mellaart 1962, 1963, and 1967 are The context of deposition within buildings may
contradictory both as to its building context and its have been affected by factors beyond control, in par-
precise finds pot). Mellaart also mentions a stone figu- ticular the spectacular fires described by Mellaart. It
rine from a granary in EVI:5 of which I have found is to this that he ascribes the recovery of a number a
no trace, and 585 was in a basket. Several figurines figurines. Buildings EVIA:10, EVIA:14, EVIA:31 and
also have a context of association- 585 was not just AVI:1 (later known as EVI:61) were heavily burnt so
in a basket in an anteroom, but it shared the basket that they could not be emptied, extra fill being added
with a stone pestle and bone scoop; 583 was not just to collapsed walls. Thus they were 'found intact with
on a platform below leopards, but lay among grain their contents' while other buildings (AIII:1, AIII:8,
and crucifer seeds; 589, which shares a confusion of EIV:1, EVI:7 and EVI:8) were burnt, cleaned,
building context with 577 but I believe came from replastered, and then filled - they 'had been more
EIV:4, was found in a deposit of peas. Where the or less cleaned out' (1963, 48). Although many so-
peas were is not described - current work shows called shrines were burnt, only a few contained figu-
that they could have been in a number of places rines. Of those mentioned as being too burnt to
including platform, floor, hearth, and probably store- empty, EVIA:10 had a fine collection of figurines,
room. including a single clay one in the storeroom along-
It is clear from this breakdown of find context side some flint and obsidian projectile points and
by number that figurines are found in a wide variety daggers, one in its leather sheath; EVIA:14 had none;
of spaces and associations, although it must be rec- AVl:1 had a most unusual painted clay figurine in a
ognized that the find context may bear little or no storeroom which it shared with a large wooden plat-
relation to use before deposition. ter, two stone maceheads with handles in situ, three
horn cores, two circular baskets, a number of obsid-
Contexts of deposition ian and flint weapons and some polishing stones, all
The range of find context described above suggests lying on marsh grass matting; and EVIA:31 had a
that the context of deposition - the reason, method, single animal figure. Of those buildings which were
and level of deliberation guiding deposition- would cleaned out before being filled in, AIII:1 contained
have been similarly varied. As mentioned already, two figurines, and a number of flat stone pendants;
the presence of humanoid and animal figures ·be- EIV:1 contained one figurine; EVI:8 was cleaned out
tween walls could be regarded either as deliberate - but 'a group of objects that lay on the floor of the
as Mellaart saw it (he related them to the building he doorway was carefully left in situ' (1963, 61); and
regarded as a shrine, EVIA:10, although presumably AIII:8 and EVI:7 had no figurines or reported finds.
he could equally have related them to the buildings AII:1 is reported to have been destroyed by fire, but
providing the other boundary wall) - or as the by- was close to the surface of the mound so that little
product of the disposal of refuse into empty spaces survived of its walls, and it is not clear whether or
in a confined area. In this particular case, the number not it had been replastered and then filled in. It
of figurines recovered from the spaces surrounding contained a large collection of figurines in three

218
Figurines, Clay Balls, Small Finds and Burials

different rooms, four 'stamp seals', a dozen pots, at but this still involves disposing of one complete and
least seven small deposits of grain, much obsidian, serviceable figurine even when, as we know, the
some chert and flint, and several hundred palettes, others could have been mended. It also involves
pounders, querns and polishers. AIII:2, next door to depositing or throwing away eight other complete
AIII:1, is not reported as a shrine: it contained one figurines, of various styles ranging from very sche-
clay figurine, and a large number of stone tools as matic to highly complex, all but one of them made of
well as raw material. Mellaart suggests it could have stone. Similarly the nine figurines in AII:1 included
been a stoneworker's shop (1962, 55). These data, one complete, one fragmentary, six missing the head
chosen largely from Mellaart' s list of burned 'shrines' only and one missing the head and one shoulder.
rather than being comprehensive, suggest that the Although the clay ones could have been damaged
discovery of figurines in buildings is not solely de- by the fire and building closure, the broken stone
pendent on them being unobtainable due to fire dam- figure was almost certainly in that condition at depo-
age, nor to being within a 'shrine' in the first place. sition. The alternative, then, is that these figurines
Furthermore, the position of the figurines within the really were actively in use when the fire broke out,
buildings is varied. If we look at the hypothesis that and that broken and fragmentary figurines remained
the fires were not accidental but were part of a sys- in use. This would support that idea that they were
tem of deliberate closure (Chapter 19), the position sacred objects of some sort. Building EVI:44 con-
of the figurines has other implications, particularly tained nine figurines, only one of which was broken
those grouped together as though in use. (see below). The positions of the three large groups,
in buildings AII:1, EVIA:10 and EVI:44, on the floor
Condition at deposition in areas where they were likely to be in the way of
A considerable number of figurines are damaged, the occupants, makes it possible that, as Mellaart
and this applies also to those found in buildings. suggested, certain buildings were 'shrines' and were
This is relevant to the context of deposition, whether not occupied all year round - except by figurines.
or not the fires are regarded as accidental. All defi- Nevertheless, figurines were not found in all the
nitely schematic figurines are whole, but the major- burned 'shrines', and their purpose when they are
ity of human, humanoid and animal figurines are found is less easy to determine due to the variety of
damaged in some way. Starting with the human findspots within, for instance, building EVI:10 which
ones, it is clear that figurines could be broken and include a platform, a storeroom, between the walls,
remain in use. We have one clear example of a and on the floor. The findspots of the figurines, and
mended figurine, with holes drilled through for fix- their condition, do therefore pose some difficulties
ing (633). When found, it was incomplete, suggest- to anyone trying to offer a single generalized inter-
ing that it had been broken again and still remained pretation of the use of figurines, or of the context of
in use. It is possible, however, that it had been taken their deposition.
out of use before being deposited in a building due
for closure. This particular figure formed part of a Fragmentation
distinct group of four, all related to leopards, and Of the 181 figurines examined in detail48 are essen-
this, together with the complexity of the design, may tially complete (a few were found broken in situ but
account for the mending. Two others of the group of with all parts present). Of these, six are truly human,
four were missing their heads, and two more found 17 are 'schematic', 13 are humanoid, nine are ani-
in the same building (EVIA:10) were broken, one mals, and three show a human with an animal. Some
consisting only of the upper body and head. It could information on fragmentation is shown in Table 12.4.
be suggested, therefore, that these figurines were The implications of these numbers are not entirely
old and were no longer in use but, having a sacred clear. The low number of complete animal figures
character, had to be disposed of carefully. Thus they may be a result of deliberate mutilation during a
were placed together in an old building which was hunting ritual, as Mellaart suggested, or a result of
deliberately set on fire, and they were left in the their deposition in outdoor areas, or their place of
ruins. Some such explanation could apply to some of deposition may be a result of their damage. Up to
the buildings and figurines, but not all. For instance, three complete animal figures were found between
one figure in the group of four containing the mended the walls of EVIA:10 and one of its neighbours. Their
item was complete. Here one could suggest that since preservation may be due to intentional deposition or
these four formed a special group, they were disposed an accidental effect of deposition in a relatively pro-
of together when all but one were old and damaged, tected space. 50 per cent of complete figurines were

219
Naomi Hamilton

a head was otherwise almost identical to four others,


Table 12.4. Mellaart' s figurines by fragmentation and context.
and was one of the group of seven by the hearth. The
Building Court Level Unstratified head had in fact broken off, and was found in the
Hullwll/!:ichcmatic building; the other heads were not recovered. This
Complete 23 I 2
Broken bodies 19 12 6 may be accident - all the headless figures, and the
Head only 4 I 2 one whose head was found, are made of clay, and
Huma11oid
Complete 6 7
perhaps the heads disintegrated after the collapse or
Minor damage 7 4 infilling of the building; they may simply not have
Major damage 5 I
Head only 2 been found by the excavators; they may have been
Auimal destroyed in the same process that destroyed the
Complete
Minor damage II 6
4 building, and the figurines then either just left there
Major damage 15 2 if it was one of the buildings which was not cleared
Head/horn only 9 16
out, or placed there headless before intentional fill-
Note: No information is currently available concerning fragmentation of a further
60 figurines. The majority of these are animal, and almost all of them have level
ing-in; or they could have been in use in a fragmen-
information only. A small group consisting of unfinished/unidentifiable frag- tary, headless condition. This situation is echoed
ments has also not been included.
throughout the site - the carefully sculpted and
Damage levels have been assessed on the basis that minor damage consists of
missing extremities such as horn tips, ends of limbs, headdresses, as well as
painted figurine from the storeroom of AVI:1, the
similar levels of damage to other parts of the body. Major damage includes the only clay figurine of this type from a level VI build-
loss of the head, loss of entire limbs, and more severe damage.
ing (which normally contain stone representations
of humans) had no head; similarly the painted figu-
found inside buildings, which may be regarded as rine from EIV:4 is headless. On the other hand, EVI:44
safe and protective areas; on the other hand, at least contained a stone head with no body (643) - the
29 figurines found in buildings were broken, and of only broken figurine in the building (Mellaart 1964,
the other two animal and one humanoid figures de- fig. 31b, pl. XVIc). Mellaart suggests that its proxim-
posited between the walls of EVIA:10 and its neigh- ity to a Hellenistic pit may explain the loss of the
bours, two were broken, although the third may body, but this may not be so- the head has a dowel
have been complete. The poor survival rate of hu- hole in its base, for attachment to a body made sepa-
man as opposed to schematic figurines is rather sur- rately or simply for mounting it on a stick. As seen
prising, considering the level of work required to above, fragmentary figurines appear to have re-
create them. It may simply be that properly sculpted mained in use alongside complete ones, and this
figures are more vulnerable to damage, even if they head may well be a remnant of a figurine which was
are in a protected environment, or that because of still useful alone. 643 is not cut regularly along the
the effort required to make them, they were kept base - it appears to have been broken mid-neck,
even when broken. It cannot just be due to the ease and the hole may have been made at a later date,
of making schematic figures by adding a few fea- either in order to mend it when it broke from its
tures to a pebble or concretion, as overall the number original body, or to use it in some other way. Cer-
of schematic figures is small- parts of 53 human tainly it would appear that the head of a figurine
figures are known, compared to 17 schematic ones. was worth saving, yet at the same time many head-
The use of stone for all schematic figurines may help less figurines also seem to have been saved. A rather
explain this, as 23 complete figurines are made of similar head was found in the surface survey, also
stone, yet humanoid figures are all clay and have broken off at the neck, but this has no dowel hole.
similar survival rates. The absence of heads could well be related to other
The most commonly missing body part is the features of the site - the deliberate defacement of
head. This affects all types of figurines except sche- the heads of large-scale sculptures, the presence of
matic ones, which are all apparently complete (there skulls on platforms in two buildings, and heads and
is one possible exception). 17 clay and 6 stone hu- headless bodies depicted on wall-paintings. Several
man figurines are missing the head only or head and other figurine heads are known, offering a wide va-
one shoulder. While the neck is clearly a vulnerable riety of images far beyond the round face with circu-
area, along with arms and legs, this may not be the lar crown of hair used in the reconstruction of the
whole answer to the frequent absence of heads. Of famous birth figure: One stone figurine head (535:
the 9 figurines found apparently in situ in building see Mellaart 1962, pl. IXd) from building CIII ap-
AII:1 at least seven had no head (information is not pears to have been deliberately defaced. It was care-
available about one figure). The one which did have fully sculpted to show ears, chin, hairline and the

220
Figurines, Clay Balls, Small Finds and Burials

join of head and neck, and has two holes drilled in according to the occasion, depending on the circum-
the top for the attachment of hair or a headdress of stances of their use (see below Interpretations- Wom-
some sort, seen frequently on the humanoid figures. en's rites or women's rights). Alternatively, extra heads
(A similar hole is drilled in the head of a complete may have been stuck on broken figurines with sub-
alabaster figurine from building EIV:4.) The only stances which have left no trace, or were not looked
facial feature is a faint trace of one eye, the rest for. Unlike the shortage of heads for human figu-
seems to have been chipped off. The base of the head rines, a considerable number of heads of humanoid
is roughly cut but it can stand unaided. There is no figures has been found. These are also made of clay,
hole in the base for attaching it to a body. A rather which should give them a similar chance of survival
different head from level II (502) is made of clay. The other things being equal. Possibly we will be fortu-
upper part is almost conical but the back is flattened. nate enough to find some more figurines during the
It has a large modelled nose with nostrils shown, current excavations which will help us to answer
and squarish eyes of obsidian, only one of which these questions. At present, they must at least be
survives. In the base is a small hole, which could be borne in mind. Fragmentation of animal figures is
for attaching it to a body but is very small to take addressed below.
such weight. A very strange head from EX:28 (631)
seems never to have been attached to a body. Made Typology
of a ball of clay, - and similar in size to the clay Mellaart referred to human, schematic and animal
balls- in profile it has a straight nose, incised mouth figurines. Some of the schematic, and all the animal
and eye, and curve for an ear. The front view, how- figures were called ex-voto, implying function rather
ever, offers small wide-set slit eyes, a gash for a than form. None of these descriptions is unproblem-
mouth, and two large nostrils. Overall it has a pig- atic. The typology used in the data base is concerned
like look, but is also reminiscent of the 'dead man's with fragmentation, based on loose types, and is not
head' from the wallpainting in EIV:l. Another strik- a typology based solely on form or function. The
ing clay head (525) has a rather venomous expres- most striking aspect of the figurines from the site is
sion, due to the 45 degree slant of the incised oblong their enormous variety. Within that variety there are
eyes and upward tilt of the flattened face- similar certain clear types, and others which stand free of
in many ways to the stylized heads seen on Chal- easy classification.
colithic figurines in Cyprus. It is broken through the The term schematic has been used very loosely
thick cylindrical neck. Unfortunately it is unstratified. in this chapter to designate a group of stone figu-
Perhaps the most naturalistic head of all is in mini- rines which suggest the human body yet range from
ature- only 15.6 mm high including the remains of pebbles with incised eyes or mouth, to fairly de-
the neck where it broke from the body (478). This is tailed figurines which nevertheless are not naturalis-
also of clay, and is unstratified. Delicate modelling tic in form. Because of the enormous variety involved,
of the chin, eyes, eyebrows, nose and ears have cre- I have chosen not to define them more closely, and
ated a most unusual and striking image, although it see instead a continuum from concretions with hu-
is fairly roughly made due to its size. In some areas man heads carved on them to the finely worked
it is polished, possibly from handling. It must have human figures shown with leopards. In some ways,
been joined to a very small body, of the size nor- the schematic/human classification is of less use than
mally used for humanoid figures. a division into those of stone and those of clay. When
The variety of imagery, and degree of attention viewed in this way, the continuum in the design of
to detail, in this small sample of heads detached stone figurines shows clearly, although some figures
from bodies, suggests that heads and faces were not are hard to place - for instance, some with little
regarded as sketchy and dispensible items used sim- naturalistic human form have more detail of cloth-
ply to finish off a figure. Rather they were used to ing or facial features, such as 576, than some more
portray a range of emotions, attitudes or states of obviously human ones. The majority of stone figu-
being, and were treated with great care in some rines come from level VI - 28 out of 45. Of the
instances at least. This must be relevant to our inter- remainder, level VII has 4, level IV has 2, level III has
pretation of headless figurines. Several clay figures 4 and level II has 5, with two unstratified. With the
have holes for the attachment of heads - two of exception of an animal head of pumice in EVI:44 and
them from AII:l.lt is possible that this was not just a a bird of prey (vulture?) in EVI:25 they all represent
precaution against them breaking off if made in one humans, sometimes in association with an animal.
piece; it may be that heads on figurines were changed Most are standing, a few are seated, sometimes on

221
Naomi Hamilton

an animal. None are known from earlier than level leopard-skin top and fringed skirt. Both these hu-
VII. Mellaart pointed out that it would be wrong to man forms seem to exude confidence. The seated
see the concretions and schematic figures as earlier ones are relaxed and comfortable, the standing ones
than the complex ones, as they occur in the same strong, even commanding. Other types exist along-
buildings, along with unmodified concretions and side, in a range of attitudes, independent of the main-
stalagmites. However, the skill required for the more stream, from level VIII.
detailed ones suggests that stone carving was not The humanoid figures have sometimes been
new in level VII, and the wide range of high-quality regarded as animals or ducks, as a number have few
stone artefacts from the site proves this. The quality features to relate them to the human form. However,
of stone figurines after level VI seems far lower. there is again a huge range of representations, and
Why there should have been a floruit of stone figu- many show clear human aspects. They occur from
rines in an extraordinary range of styles at one time level VII onwards, with a possible one also in level
is not clear, although of course level VI lasted for XII. In general terms they are small clay figures
perhaps a century. The discovery of many of them in (mainly 35-45 mm high, but a few smaller and the
just two buildings could suggest specialization in largest 60 mm), with roughly conical bases, long
stone manufacture by the occupants of the build- necks and schematic heads with pinched noses. Some
ings, or that the apparent concentration in certain have a divided base at the front, representing legs or
levels is an accident of recovery. Alternatively, they possibly arms, which sometimes protrude, but undi-
could have been disposed of deliberately as part of a vided ones are more common. Seen from above, the
change taking place in society or the 'household'. head is generally triangular. A marked feature is the
The images portrayed are discussed below under headgear worn - headresses or scarves were cre-
Sex and gender. ated by extending the clay at the top of the head,
Once separated from the stone figures, the clay flattening it and folding it down behind (e.g. 521);
ones can also be seen more clearly, and there is little others seem to have pointed hats, or flat caps- one
hesitation is splitting them into three groups - hu- has three incisions across the top as though showing
man, humanoid and animal. A number seem to cross sewn fabric (516); one is possibly wearing a cape
these boundaries, but it is only a handful. The hu- depicted in black paint (537). Several humanoid fig-
man figures tend to have two basic designs. The first ures are rather different, with facial features, or
is the 'fat woman' sitting with legs crossed or tucked greater attention to detail. The earliest ones are more
underneath, hands on knees or breasts, with fingers human- 557 shows the curve of the stomach on the
carefully delineated and a lot of attention to detail. seated figure, and 533 sits uneasily in the humanoid
This type is common in the later levels, occurring category, having breasts (one pinched, one appli-
first in level V (593). Five possible precusors in level que), stub arms (one missing), an elaborate head
VI are all atypical: one (569) has its legs stretched out arrangement with multiple piercings around the side,
in front, and is also painted with a meander design; possibly for the attachment of hair or fabric but also
another also has traces of red paint (520) and resem- strongly reminiscent of the much later multi-
bles the humanoid figures in form, as does 515, which earringed bronze age figurines in Cyprus, and in
is however on a massive scale; 514 shows extraordi- addition it has streaks of red paint on the head and
nary detail of hands and feet, with knees drawn up coming from the nose. Both 557 and 533 are from
in front, similar to the pose of 541 which is a com- level VIII. During the later levels the figures become
posite figure made of rolls of clay stuck around a more schematic, and the last stratified ones are two
core. The general impression is that the style was from level V, one again with breasts (477), one with
being developed during level VI, at the same time paint on the head (497), both with slim conical undi-
that the high quality stone figurines reach their peak. vided bases. This type of figurine seems therefore to
By level II it had become standardized, although have a specific time span just as the stone ones and
other forms still occurred alongside. The second hu- 'fat woman' ones do.
man type is a standing figure with divided legs and The animal figures, with four exceptions, are
arms by the side or near the breast, which occurs all of clay. These four exceptions are a stone bird of
from level IV onwards. It seems to derive more from prey (vulture?), headless (599), a bone vulture beak
the stone inheritance of the level VI figures, and (392), a bone animal, uncertain (788), and a pumice
stone versions do occur, but neither stone nor clay animal head, uncertain (712). Three of these were
ones attempt the elaboration of the earlier ones found in buildings, the other has no context. The
except for one figurine in level II, clothed in a clay figures offer great variety of size, stance and

222
Figurines, Clay Balls, Small Finds and Burials

surface treatment. They represent a range of ani- The animal figurines from level VI were inven-
mals, the most common of which appear to be cattle toried in two large groups, number CHC319 (18 fig-
(8 clear, 21 possible) and boar (12 clear, 7 possible). ures) and CHC322 (19 figures). I have been able to
Other identifiable types are dog (8 probable), leop- examine only a small number of these, which makes
ard/ feline (3), goat (or possibly deer?) (3), and possi- any inferences unreliable, but nevertheless some pat-
bly bear (3?) and sheep (1). There are also nine horns, terns do emerge which will bear further investiga-
probably from cattle. In addition to this a considerable tion. Mellaart reported that large groups of animal
number are impossible to identify, either because figurines were found in two pits, outside EVI:12 and
they are of no distinct type other than quadruped, or EVI:14 (see above), which he interpreted as having
because defining features such as the head are miss- been used in a hunting ritual. Of the seven CHC319
ing. Size ranges from approximately 50 mm head to figures examined, six are lying down. They repre-
tail when complete to 150 mm, although due to the sent two goats, three cattle, and an unidentifiable
damaged state of most figures precise measurements headless quadruped. This last has been stabbed, the
are not possible; the stances depicted are lying, sit- only one in the group which has been. The seventh
ting, standing and possibly running; surface treat- figure is standing, and may be a boar. All these
ment may include smoothing, polishing, stabbing, figurines are small, up to about 55 mm long. Two
and indication of bristles or spots. Interestingly, the may be complete, four have minor damage, and one
species of animal represented does not seem to con- is only a torso. The nine CHC322 figures examined
trol exclusively the method of represention, although are very different - five of them are large or very
there are patterns- all three goats are shown lying large, none has been stabbed, and all are damaged -
down, and none has been stabbed. The animal shown four are heads, two are just horns. The animals rep-
most frequently to have been stabbed is the boar (11 resented are four cattle (plus two horns), one boar,
instances), and these are also the most likely to be and two unidentified- one of which may be a goat,
whole or nearly so, but boar which have not been deer or possibly boar, and is lying down. The differ-
stabbed, or which are fragmentary, occur also. Cattle ences in stance, fragmentation and species dominat-
show the widest variety - although most are stand- ing these two groups suggests that they may have
ing, at least two are lying down, some are stabbed and been used in different ways, and should not be
some are not, and the degree of fragmentation ranges lumped together. It is possible that the figurines
from complete to just a horn. should not be divided in this way - Mellaart also
Animal figurines were found almost exclusively reports animal and humanoid figurines from pits
in levels VI and VIII. Besides these, there is one from outside buildings EVIA/B:61 and EVIB:23, and sug-
level XII, three from level VII, and two from level V. gests another occurrence outside EIV:4 (1967, 102-3)
Certain typological aspects can be approached bet- although no artefacts can be tied to this last group.
ter by looking at them by level. 26 figures are known He himself suggests a division between the largely
from level VIII, 21 of which I have been able to sheep, goat and cattle figurines which he says were
study. A group of 19, of which I have seen 15, were found between the walls of buildings, and those
inventoried under a single number (CHC318) and representing mainly wild animals such as boar, cat-
are likely to be those mentioned by Mellaart as com- tle and ?deer which have been wounded (1963, 78).
ing from a level VIII pit near building VII:l. These
represent two whole ?boar, one stabbed, one not; Types in context
one whole ?bear, one whole stylized cattle head; I have already mentioned the enormous range of
three damaged hornless quadrupeds (dog or boar); depictions of humans- and animals- at <::atalhOyiik.
two rear ends, one stabbed, one not; and six varied Within this range I have discussed broad types. How-
body fragments. Three were stabbed, three had a ever, there are certain groups which are linked not
possible stab mark, and one had incisions on the just by style but also by context, and these require
neck. Looking at the level VIII figurines overall, they further investigation. Similarities in the stance of
tend to be large, although a few medium-sized ones some animals found in pits has been mentioned
occur; five were complete, the remainder broken to above. Here I would like to look at the figurines
various degrees; five were stabbed, four had a possi- from certain buildings.
ble stab mark, one had incisions and nine had no Building AII:1 contained nine figurines, eight
sign of stabbing. This suggests that the figures may of which are known. The ninth figure is described as
have been used in different ways before their depo- being a broken standing figurine made of stone, but
sition in the same area. has not been illustrated or located. Of the eight

223
Naomi Hamilton

figures known about, all are made of clay, and five non-feline animal, one is a bust, one is a human head
are of one type - the 'fat woman' style of seated and one is an unknown animal head.
figures with large breasts. The three other figurines Building EVI:25 contained four figurines, all
are all different - one is a fat standing figure with quite varied- a person seated on a stool (570), the
large breasts, one is a fat woman with large breasts lower part of possibly a schematic figure (505), a
who is (perhaps giving birth) seated on a throne of kneeling human (574), and a probable vulture (599).
felines, and the last is a standing figure with small It is notable that the vulture did not occur in a build-
breasts, divided legs, and wearing a skirt and leop- ing in which vultures were depicted in other ways.
ard-skin top. It has a hole for the attachment of a Three buildings had wall-paintings depicting vul-
head. The five forming the group were found to- tures, and vulture skulls were moulded into a relief
gether, and are remarkably similar. Their differences in another, but these did not contain vulture figurines.
are in the position of the hands and feet. The two A carved vulture head was found in another building
smallest have both hands on their legs, which are (EVI:70) but this did not contain vulture paintings.
crossed, and one has a hole for attaching a head; two
have one hand on one breast and one on one leg, Interpretations
with their legs tucked underneath to one side; the Three interpretations have been offered for the figu-
largest, which is the only one with a head, has a rines, each one applying to a different group and all
hand on each breast and legs tucked underneath. offered by Mellaart. The human and schematic fig-
They are well made and give the impression of be- ures are seen as deities, with an emphasis on fertility
ing a deliberate group, and could have been made by seen in fat females and an abundance of bull im-
one person. On the other hand, four other figures of the agery. This interpretation has been widely accepted.
same style are known from other contexts, the most The humanoid and animal figures are ex-voto images,
similar being 648- also from level II and 593- from a the humanoid ones having perhaps a protective use,
court in level V; the other two are 590 - a stone ver- the animal ones used in hunting magic. As the hu-
sion from AIII:l; and 589- a clay one with painted man figures are the only ones generally known about,
leopard-skin garment from EIV:4. their interpretation is the one which dominates ideas
Building EVIA:lO contained a large number of about ~atalhoyiik.
figurines, including a group of four related to leop-
ards. Two show a human standing behind a leopard, Function
one of them wearing a leopard-skin scarf; one shows As noted at the beginning, I am not concerned here
a human sitting on or riding a leopard; the fourth with defining functions clearly, but with raising ques-
shows a person wearing a leopard-skin hat. The first tions. However, it seems possible that the function
two are stylistically very close, although one person of figurines varies both according to the type of fig-
has breasts and the other does not, but similarities ure, and over time. The possible functions of human
with the other two are not profound, depending and schematic figurines are discussed below (Sex and
mainly on the presence of spots presumed to be gender, Women's rites or women's rights). First I will
leopards'. Nevertheless, when this is combined with look briefly at the humanoid and animal figures.
their find context, they do make a cohesive group. There are two easy options for the humanoid
Whether or not they all date from the same time, or figurines - one is that they are votive objects, the
were made by the same person, is less certain. Four other that they are toys. The toy interpretation de-
buildings were found with reliefs of leopards pends largely on the denial of their human traits, for
moulded onto the walls, one had a feline head cut they would not be viewed as dolls - they are too
into the plaster, and several others had wall-paint- small- but as gaming pieces or counters. It is possi-
ings showing people wearing leopard skins, yet the ble that, taking into account their clear human as-
leopard-related figurines were not found in associa- pects, they could fit this category. Their rough and
tion with any of these - instead the building con- ready manufacture, the frequent damage, the dis-
tained modelled cattle and sheep heads, although posal in rubbish areas, are all consistent with this
one of Mellaart' s reconstructions does suggest a fe- idea, but a context of use which could support this
line head on the huge human relief on the wall of view is lacking so far. It will be interesting to see
EVIA:lO. One of the 'leopard shrines' containing leop- what the 'chessman' figures reported from Can Hasan
ards in relief (EVI:44) did contain figurines, but they look like (French 1968, 48 & 52), and what contexts
did not show leopards - four are schematic, one is a are available for them. Mellaart's reports of humanoid
standing figure, one a bearded figure seated on a figures in the walls of buildings suggest that a votive or

224
Figurines, Clay Balls, Small Finds and Burials

protective interpretation is most likely to be correct. women. Female bodies and women are rarely, if
Animal figurines may have had more than one ever, considered separately, although there is by now
function. While some were clearly stabbed, and oth- ample literature to document the social construction
ers may have been deliberately mutilated, which con- of gender as sex-based 'complementary' roles, and a
forms to the hunting ritual interpretation, the growing body of research on the social construction
presence in the same contexts of whole figures which of sex itself (see Hamilton 1994 & in press for brief
have not been stabbed suggests they may have been examples and references).
used in another way. The difference in size and stance The definers of sex have been much argued
may also be relevant. Several animals are shown over in studies of Cypriot and certain Near Eastern
lying down - three goats and at least five cattle. figurines (Hamilton in press; Morris 1985; Ucko 1969)
These have not been stabbed, so they may not repre- but little attention has been paid so far to those in
sent dead animals, but rather be related to other Anatolia. A handful of figurines at <;atalhoyiik have
activities such as domestication. These are all from been interpreted as male (Mellaart 1963, 83-90; 1964,
level VI with the exception of two unstratified cattle, 75-81). Some of these have beards, and are seated on
and animal figurines have not been found in strati- animals regarded as bulls (e.g. 282, 632, 592), al-
fied contexts in later levels. Further investigation though the absence of horns (a major symbol at the
with a more complete data set may allow some clearer site) or any clear defining characteristics, when com-
insights into differences over time. pared to the careful drilling of spots on leopards,
makes them just as likely to be sheep. Several figu-
Fertility rines regarded as male have no claim to the ascrip-
A striking aspect of the human figurine assemblage- tion other than the absence of breasts - 580, the
and indeed the assemblage as a whole - is the lack headless 'youth' riding a leopard; 630, a cloaked
of symbols usually associated with fertility in dis- standing figure; 638, a seated figure with leopard-
cussions of figurines. Phallic imagery is extremely skin hat and armlets; 570, seated on a stool; 643, a
rare; there is only one possible birth scene; there are head very similar to those on the twin 'double-
only two possible babies (including the one in the breasted' figure 572. These all come from two build-
birth scene); and although many of the human fig- ings, EVIA:10 and EVI:44. One is known from level
ures are fat women, there is no clear reason to view VII (585) - beardless, with small breasts, the back
this as pregnancy rather than the mature female fig- view of the torso shows broad shoulders, and it is
ure much in evidence in central Anatolia today. As riding a hornless animal. Although I am sceptical
mentioned above, there are some contextual reasons about some of this sexing, and the animals deemed
for associating a few figurines with crops, but abun- suitable to accompany men, there are also some figu-
dance of food and human fertility are not precisely rines generally regarded as female which have no
the same. There are also some relief figures in build- clear claim to the name - breastless and beardless,
ings termed shrines which were viewed by Mellaart such as one standing behind a leopard in EVIA:10,
as giving birth, as they were placed above animal or indeed the majority of stone figures and the ma-
heads. This imagery, whatever its meaning, is obvi- jority of figures pre-dating level VI. The figurines
ously rather different to women having babies. with large heavy breasts start in level VI and domi-
Mellaart thought the goddess was shown giving birth nate in the later levels. Overall, level VI seems to be a
to a son - as bull or ram - displaying a common time of change where figurines are concerned- in
binary and complementary attitude to sex and gen- terms of materials, styles, and types of representa-
der (discussed below). His suggestion that an aurochs tion. After level VI male-defined figures have not
bull or a large ram is a more impressive symbol of been found, nor any humans riding animals, and
male fertility than men themselves (1967, 181) may although breastless figures do continue the domi-
well be true, but does not explain why a woman nant form is demonstrably female. The sexless hu-
should be impressive enough as herself. The dispar- manoid figures, and the animal figurines (which
ity between male and female images both in relief never show sex) also cease soon afterwards. The
sculpture and in figurines needs to be examined, multiplicity of overtly female representations com-
and is discussed below. pared to demonstrably male ones - even when pro-
portions are adjusted - suggests a growing focus
Sex and gender on the female which was less apparent in the figu-
There is a widespread belief that the majority of rines from the earlier levels. There are other reasons
prehistoric figurines represent female bodies/ for viewing level VI as something of a watershed at

225
Naomi Hamilton

<:;atalhOyiik- the change in pottery technology and new emphasis on femaleness in figurines after level
knapped-stone tools. It is probable that these were VI. The production of clay figurines depicting ample
accompanied by altered social dynamics, which may women with large breasts, in which not 'fertility' but
have been reflected in the altered representation of femaleness was emphasized, along with the absence
humans -and animals. of maleness, might suggest an increasing concern
The dearth of overt representations of males with women's roles. Rather than taking the 'accept-
has been assumed to have been balanced by a sym- able' view that women tamely embraced a life of
bolic representation of males in the guise of cattle endless childbearing, the figurines may also be dem-
and sheep heads, always referred to as bull and ram onstrating their challenge to fulfil other roles in soci-
heads even though some of the horns are known to ety - perhaps in debate over lineage, access to
come from females, and sex is rarely depicted on power, etc. (see Chapter 19). The mature bodies de-
animals in paintings or reliefs. The interpretation of picted in the most stylistically unified group accord
the large relief figures with animal heads beneath with the decision-making role played in some socie-
them as goddesses giving birth to sons fits beauti- ties by the longer-lived experienced members. The
fully with Western attitudes to virility and its appro- relaxed, confident, even commanding pose of many
priate symbols, but less well with the evidence. Most of these figures is more suggestive of the matrons,
of the relief figures have no overt sex, and although the elders, than of young women's rites of passage.
this does not mean they are not female, as sex may be
signified in ways other than those widely understood Shrines and non-shrines
in the West today, nevertheless it should not be as- Some buildings have large numbers of figurines in
sumed without question. The identification of the them- EVIA:10 (14 + 3 between walls), EVI:44 (9),
animals with males is more problematic. If we look AII:1 (9). These are regarded by Mellaart as shrines
not from a viewpoint in which maleness is privi- on other grounds, such as elaborate decoration -
leged and virility highly prized, but from the wall paintings or reliefs (EVIA:10 and EVI:44)- or
economy of <:;atalhOyiik, in which cattle meat played unusual size and complex architecture (AII:1). A few
a significant part, the presence of cattle symbolism is more have more than one figurine - EVI:25 (4),
not strange. If we accept that the relief figures are EVII:24 (2), EIV:4 (2), AIII:l (2) and EXII:29 (2). Of
giving birth to animals, this is then simply the crea- these, only AIII:1 is consistently regarded by Mellaart
tion of the economic requirements of humans. as a shrine, while for EIV:4 and EVI:25 reports differ.
Of the buildings containing one figurine, AIII:2,
Women's rites or women's rights? AIV:1, EVI:28, EVII:28 and EVIII:28 were viewed as
The presence of predominantly female figurines - houses, AVI:1, EVI:31(animal figurine), EVI:45,
or a widespread belief that they are female - has EVII:21, EVII:45, EVIII:1 and EVIII:25 were called
led to interpretations of their meaning and function shrines, and EV:8 was uncertain. The majority of
which generally relate to biological roles and either buildings in which figurines were found -albeit in
elevate or downgrade women through that medium, small numbers - were not regarded as shrines, nor
according to the viewpoint of the scholar concerned. were figurines found in the majority of buildings
Such interpretations have been 'acceptable' due to which were designated shrines. All the buildings with
the dominant Western view of women as 'natural' the largest numbers of figurines are amongst the
birthers and 'natural' mothers, which is part of the most elaborate in the complexity graphs (Figs. 1.1-1.5).
naturalization of the restriction of women to the do- Mellaart pointed out that elaborate buildings
mestic sphere. The elevation of the role of birther are often built above earlier elaborate buildings, al-
and mother to a divine one, as in the mother god- though this is not always the case (see Chapter 19 for
dess interpretation, does not change or challenge quantitative data). Whether or not these are shrines,
this underlying belief. it would not be entirely surprising if the form of
While figurines may well have played a role in decoration used in one building occurred in the suc-
women's rites of passage, such as those of birth and ceeding one, since it is likely - given the settlement
motherhood - and as noted above, the possibility layout and lack of fresh building areas- that the
exists that detachable heads were used to portray new buildings were built by the occupants of the old
different ages, emotions, rites etc. - it is equally ones. It may be for the same reason that figurines are
feasible that they were concerned with sex-based sometimes found in consecutive buildings, or in the
gender roles and the consequent social conflict this same vertical column of building but skipping a level
doubtless created. This may be supported by the or two- EVIA:31 is below EV:8; EVII:45 is below

226
Figurines, Clay Balls, Small Finds and Burials

EVIB:45, EVIII:25 is below EVI:25, with level VII miss- outstanding features are the range of find contexts,
ing out on figurines, while AVI:1 is below AIII:1, the changes in the assemblage over time, the differ-
jumping a huge time-span. However, these build- ent time span for the different groups, and the links
ings are not always similar - EVIII:25 was called a these aspects must have, in some way, to the other
shrine by Mellaart, and had one figurine, while EVI:25 changes at the site. Other aspects of figurines could
was initially seen as a house, although it contained give us information both about themselves and the
four figurines. The designation of EV:8 as a shrine culture from which they come: use-wear analysis
was left uncertain. may shed light on their function, and preliminary
work has shown that this could yield results, as ar-
Totemism eas of polish suggest handling, but newly-excavated
One possibility which could be investigated further artefacts are most likely to offer relevant informa-
is a totemic base to some of the figurative represen- tion; a study of which figures can stand unaided is
tations at C::atalhoyiik. Attention has been drawn also helpful in determining function; an investiga-
above to the dyssynchrony between representations tion of fingerprints on many of the humanoid fig-
on wall-paintings and in reliefs and those on the ures could be revealing, but is complex to carry out;
figurines - in particular the leopards and vultures. and the iconography offers assistance with clothing,
It is feasible that a totemic clan structure existed an area generally difficult to examine. Some of these
which utilized certain animal symbols- the leop- directions must wait for the future, but the theoreti-
ard and vulture have already been mentioned, but cal debate has been underway for many years. My
deer and boar could be added to the list. The pres- challenges to orthodoxy are not offered as proven,
ence of leopard-based figurines in a building with- but as matters to consider. Further discoveries dur-
out other leopard symbolism could reflect a different ing the current excavations may be able to enlighten
usage of the same symbol by members of the same us in some of these areas, while others are likely to
totemic group. Aspects other than figurines and remain bones of theoretical contention.
building 'decor' could have been involved - cos-
tume in particular. Thus people shown on paintings The figurines from the surface
in leopard-skin clothing could be demonstrating their
clan affiliation, and those shown in plain red skins Details of the figurines found during the surface
could perhaps be wearing deer skin. From this view- investigations in 1993-95 can be found in the discus-
point, the variety of depictions of human skin and sion of small finds below. Here I am concerned with
clothing, grouped in the paintings, taken in conjunc- their presence on the surface of the mound.
tion with the lack of killing of animals which was Both human and animal figurines were found
noted by Mellaart, could lead to an interpretation of during the survey in small numbers - 20 definite
the scenes as contests between totemic groups. The figurine fragments, of which 17 come from the east
importance of cattle in some of the scenes, and in mound survey (5 human, 3 humanoid, 9 animal);
buildings which also have depictions of other ani- one from Mellaart' s sections (animal), and two from
mals, could suggest that cattle were the totem of the the west mound (1 animal, 1 uncertain); and a fur-
whole society, and that inter-community contests ther nine possible fragments come from the east
were played out against a backdrop of cattle im- mound (3 human, 2 humanoid, 2 animal, 2 indeter-
agery. The presence in some graves of items made of minate), one of which is from the sections. Only one
antler, and deer teeth necklaces, could also be inves- is of doubtful date, the rest being clearly Neolithic.
tigated within this framework, although at present Details can be found in Table 12.5. Although this is
the data are poor. too small a group of artefacts from which to draw
statistical conclusions, it is nevertheless a surpris-
Conclusion ingly large number when compared with the num-
I have not aimed to reach definitive conclusions about bers found by Mellaart in large-scale excavations -
the figurines from Mellaart's excavations, but it is around 250 figures, of which about 130 are human or
clear from the afore-going that the figurines from humanoid. We have perhaps a slightly larger pro-
C::atalhOyiik do not fall into a single category, and portion of small fragments than Mellaart, which is to
did not all have the same function. There is scope for be expected of purely surface material, but the re-
a great deal more work in this area- this discussion covery methods used were the same in the case of
has merely brushed the surface, as a re-assess- the scrape squares- hand-picked by workmen from
ment of the evidence, not a re-interpretation. The shovelled soil.

227
Naomi Hamilton

With such limited data, and general find spots, higher. Of the 44 definitely deposited in buildings,
it is not possible to carry out detailed quantitative 23 were broken or fragmentary, sometimes consist-
analyses of the type used for the knapped stone, ing solely of the head. Humanoid and animal figu-
ceramics and bone. However, there are some points, rines were very rarely deposited in buildings, and
largely based on qualitative information, which may are reported to be found in pits and courts. Of the 47
be pertinent to the interpretation of the distribution humanoid figurines with records, 15 are complete
of those materials. and a further 11 have only minor damage. Of the 76
animal figurines with records, only 8 are complete
Distribution with a further 18 having minor damage to horns and
20 of the total 29 possible figurines came from the legs. Therefore over 50 per cent of humanoid figures
large scraped area on the upper north slope and are in a good state of completeness, a far greater
northern-most edge of the top of the northern emi- percentage than the human figures, yet they were
nence, which could be assumed to have been subject apparently discarded in pits and rubbish dumps
to considerable erosion. The figurines were discov- while the human ones were deposited in buildings,
ered in sub-surface units giving good Neolithic ar- sometimes in a broken or fragmentary condition.
chitectural contexts, and not in the surface units, The animal figurines show similar damage rates to
suggesting that the figurines found may be basically those for human ones, despite very different de-
in situ, and may not have been subject to a great deal positional contexts. These data agree well with the
of movement. The absence of a single recognizable surface material- the humanoid figures are in the
figurine fragment in the 2 x 2 m survey of the east best state of preservation, the animals a mixture of
mound - in which artefacts were retrieved by good and very fragmentary, and the human figures
seiving in a 5 mm mesh - must be relevant to the include nearly complete, head only, and badly dam-
question of erosion and movement of artefacts, and aged fragments. The close find spots of two almost
supports the suggestion that many of the figurines identical figurines in very different states of frag-
which were recovered had not travelled far from the mentation may therefore be a result of deliberate
place of their original deposition. depositional practices in which broken human figu-
This contention is supported also by the con- rines are deposited in buildings. It is also possible,
text of two particular figurines - from unit 821 however, that the severely damaged fragment had
(CH94:32) and unit 282 (CH93:30). These two arte- been discarded in the adjacent open area/ street.
facts are almost identical. They were retrieved from
adjacent 10 x 10 squares. They are the only figurines Dating
of their type, and in my opinion it is possible that The human head CH94:31 is the only stone figurine
they were made by the same people for use in the fragment found on the surface. It came from the
same building. Deposits of lime on the surface of southeast quarter of sub-surface unit 820, overlying
both also suggest they lay in similar soils. a Neolithic building. Stone figurines are most com-
mon in the level VI material from Mellaart' s excava-
Fragmentation tions - a total of 43 stone figurines was found,
The majority of figurines from the survey are bro- ranging from levels VII to II, of which 26 were from
ken. In the light of my suggestion that many have level VI and four were unstratified. None were found
not travelled far from their place of original deposi- in level V. A head rather similar to this one, also of
tion, it is necessary to ask how much this is a prod- stone, was found in building VI:44, the 'leopard
uct of erosion and post-depositional damage. In other shrine'. It is considerably larger than ours, being 54
words, would broken figurines have been deposited mm high, and it has a dowel hole in the base for
in buildings? Here, information derived from the attaching it to a figurine, whereas ours was clearly
figurines found in Mellaart's excavations can shed made all in one piece and has broken off above the
light on depositional practices. The data is discussed shoulder. Much more similar is a large clay head
in detail above, and it is clear that different types of found in Mellaart's backfill in 1995, unfortunately
figurines were deposited in different types of con- unstratified. The humanoid figurines can probably
text and in very different states of completeness. be dated reasonably by comparison with Mellaart's
Gross counts in Table 12.4 (p. 222) show that well material. Two figures with conical bases surmounted
over 50 per cent of human and schematic figurines by fairly featureless heads were found in level V
were deposited in buildings, and as the remainder (477 & 497), and compare well with CH95:39 al-
have no context the percentage could have been though this is a little smaller, and one of Mellaart's

228
Figurines, Clay Balls, Small Finds and Burials

had breasts. This style does not appear in other levels. missing its head; the other a horn or arm. There is no
CH94:34 and CH94:1 resemble many figures from comparative material from Mellaart's excavations and
levels VII and VI. The two figurines discussed ear- with only two pieces and no sub-surface investigation,
lier (nos. CH94:32 & CH93:30) as well as CH93:4 nothing can be said concerning their position on the
have no clear parallels among the excavated mate- mound.
rial. The fragment impressed with small circles
(CH93:31) may however be compared to four stone Conclusion
figurines from building VI:10 decorated with similar Although the dating evidence is scanty, relying on a
circles which clearly represent leopard spots- in restricted sample, it does suggest overall that the
three cases the spots are on animals, and in two figurines found on the northern eminence belong to
cases on clothing. A clay figurine from building AII:1 level VI or V, which is entirely in agreement with the
is also wearing clothing on her upper body deco- ceramic and knapped-stone evidence not just for the
rated with similar spots. This may provide a better surface material but also from the 1995 excavations.
date for this fragment, both in terms of its material The lack of figurine material washed down to the lower
(clay is rare for level VI human figures but common slopes, when so much is available at the top, does
in level II) and its position on the mound. CH93:31 is suggest either very restricted movement of this arte-
the only clearly human figurine to be found on the fact class or its rapid weathering and deterioration.
southern eminence, and came from close to Mellaart' s
old trenches. It was found in unit 24, a lower spit at Clay balls from the surface survey
least 7 em below the surface and possily twice as (with Mehmet Uluceviz)
deep, in the southwestern quarter of a square not far
from Mellaart's Area A in which AII:1 is situated. Clay balls occur at C::atalhOyiik in two broad size ranges
The evidence of the animal figurines is also -small, measuring roughly 1-2 em in diameter; and
largely in agreement with the dates suggested by the large, measuring in general 6-8 em in diameter, oc-
human and humanoid examples. By far the majority casionally as large as 11 em or in two cases appar-
of animal figures found by Mellaart come either from ently 17 em. Here I shall discuss the large clay balls.
level VI or level VIII. Only five are known from Fragments of clay balls have been found scat-
other levels - one from level XII, two from level VII tered across the surface of the east mound- this at
and two from level V. CH94:33 and CH95:42 both least is the impression of those working on the survey,
fit happily within the level VI
reportoire. The animal head pot Table 12.5. Figurine fragments from the surface survey.
lug CH94:3 also has parallels. Four
similar lugs were found by Unit Square Unit type Figurine type Part present Material Notes

Mellaart and are in the Konya Mu- Definite figurine fragments


818 1020,1200 Sub-surface Humanoid Leg missing Clay Lime?
seum. Two were from level V, the 821 1020,1190 Surface Human Body Clay Lime deposit
282 1030,1190 Scrape Human Lower body Clay Lime deposit
other two unstratified. The item 820 1030,1200 Sub-surface Human Head Limestone Painted
which has no clear parallels is 816 1050,1170 Sub-surface Humanoid Complete Clay Lime
901 1030,1160 Sub-surface Humanoid Complete Clay
CH94:2. It is larger and more care- 820 1030,1200 Sub-surface Animal Complete Clay Stabbed
901 1030,1160 Sub-surface Animal Rear end Clay Stabbed
fully modelled than most, and 824 1040,1200 Sub-surface Animal Head Clay
seems to be either an equid or ovi- 824 1040,1200 Sub-surface Animal Horn Clay
816 1050,1170 Sub-surface Animal Head Clay Pot lug
caprid, neither of which occurs in 826 1020,1180 Sub-surface Animal Horn Clay
822 1020,1190 Sub-surface Animal? Horn? Clay
this size in the excavated material. 822 1020,1190 Sub-surface Animal Horn Clay
It is the most likely candidate for a 279 1040,1090 Lower spit Human Body Clay Incised triangle
24 990,1040 Sub-surface Human Upper body Clay Impressed
post-Neolithic date apart from the 295 1040,1240 Scrape Animal Horn Clay Incised? Lime?
565 Section 6 Animal Neck Clay
possible horn CH94:12 from the 802 460,840 2x2 Animal Body Clay
west mound (not mentioned in Ta- 648 520,940 2x2 Arm? Horn Clay

ble 12.5 but can be found in the Possible figurine fragments


710 1030,1170 Scrape Humanoid Lower body Clay
catalogue). However, its context is 288 1030,1180 Scrape Human Upper body Clay Impressed
a stongly Neolithic sub-surface unit. 841 1020,1170 Sub-surface Humanoid? Lower body Clay
264 1040,1190 Sub-surface Human? Upper body Clay Impressed
Only two certain figurine 196 1090,1040 Sub-surface Human? Lower body Clay Incised
818 1020,1200 Sub-surface Limb? Horn? Clay
fragments were found on the west Section 3 ? Limb? Horn? Clay
mound, both in the 2 x 2 survey. 706
269
1000,950
1040,1040
Sub-surface
Lower spit
Animal?
Animal?
Horn?
Horn?
Clay
Clay
Incised

One is the body of a quadruped,

229
Naomi Hamilton

by whom clay balls have come to be regarded as surface detail. The surface is generally only roughly
ubiquitous. A closer look at their distribution sug- smoothed, although some are well-smoothed and
gests that this is not quite true. The numbers found one is recorded as almost polished. However, sev-
vary enormously across the site and from unit to eral examples have been 'decorated': a few have a
unit, as will be seen from Table 12.6. dip or dimple, occurring either singly or in threes
Clay balls are generally approximately spheri- arranged as a triangle fairly close together; others
cal in shape as their name suggests, but as all but one bear scratch or chisel marks. However, deliberate
example from the surface are fragmentary (the one surface treatment is very rare (19 examples among
that is complete is broken into two joining pieces), it 361 fragments) and some of these marks may be
is not possible to say with certainty that they were accidental. At least three are probably the imprint of
all spherical although a few complete, spherical ex- matting or textile, a feature found on perhaps a dozen
amples have been found in excavation. Indeed, as balls and fragments found in Mellaart's area in 1995.
mentioned below, a number have one suspiciously
flat surface. All the fragments are burnt, although Distribution
the degree of burning varies considerably. They are The distribution of clay ball fragments on the east
usually quite carelessly made, with little attention to mound is shown by Table 12.6 to be restricted largely
to the large area scraped on top of the north-
Table 12.6. Distribution of clay ball fragments on the surface. ern eminence. No data are available for the
1993 10 x 10 survey. However, all but four
East mound 10 x 10 tmits
Unit Grid No. of balls Level Decoration/Notes
of the fragments found in the 10 x 10 sur-
500 980,1090 Surface fingerprint
vey in 1994 and 1995 derive from this area.
504 990,1080 Sub-surface Of the 5 fragments found in the 2 x 2 sur-
602 990,990 Surface
603 990,990 Sub-surface vey, four also came from this area. Of the
710
808
1030,1170
1040,1170
14
9
Scrape
Sub-surface 1 complete (broken)
remaining five fragments, two come from
809 1050,1200 2 Surface the top of the southern eminence, close to
810 1050,1200 Sub-surface 1 crescent incision
811 1050,1190 Surface Mella art's trench, two from the 20 x 20
812
814
1050,1190
1050,1180
17
37
Sub-surface
Sub-surface
1 wavy lines
3 dots in triangle; 1 circle
scrape on the northern slope of the south-
815 1050,1170 Surface ern eminence, and one from the north end
816 1050,1170 11 Sub-surface
817 1020,1200 Surface of Mellaart's long spoil heap. 43 fragments
818
819
1020,1200
1030,1200
47 + 1 mini
2
Sub-surface
Surface
1 'insect' impression
very small
were found on the surface, in 14 units; the
820 1030,1200 11 Sub-surface 3 small holes in triangle vast majority come from sub-surface units
821 ]()20,1190 Surface
822 1020,1190 53 Sub-surface lime; 1 nail impression; 1 2 lines on the northern eminence overlying good
823
824
1040,1200
1040,1200 26
Surface
Sub-surface
very small
1 semi-circle
Neolithic architecture. They may not, there-
825 1020,1180 8 Surface 1 cut fore, be the ubiquitous artefacts we have
826 1020,1180 14 Sub-surface
827 1020,1170 Surface come to believe them to be.
828-52
854
1020,1170
1045,1125
47
2
Sub-surface
Sub-surface white semi-circle round outside
Due to the methods employed in the
901 Hl30,1160 8 Sub-surface 1 incision +dimple survey, it is not easy to establish precisely
904 1030,1150 7 Sub-surface eye+ ear?
the type of context in which clay balls were
2 x 2 units in scraped area 1020,1170 on northern eminence
Unit Grid No. of balls Decoration/Notes
found. However, we do have some indica-
829 1020,1170
tions. The 10 x 10 area scraped at 1020,1170
830 ]()20,1170 was divided into 25 sub-surface 2 x 2 units
831 1020,1170 3 + 1 mini
832 1020,1170 2 (numbers 828-52). Clay ball fragments oc-
833
834
1020,1170
1020,1170
2
2
include 1 x 11 em
include 1 x 17 em
curred in 21 of these units. According to
835 1020,1170 4 include 1 x 17 em the plan, these units lay almost exclusively
836 1020,1170 3
837 1020,1170 2 1 groove/piercing? over internal areas apparently covering
838
839
1020,1170
1020,1170
1 + large mini
1
1 straight incision parts of 4 buildings. The only external ar-
840 1020,1170 cut marks? eas are in the extreme northeast corner,
841 1020,1170
843 1020,1170 which just clips the street, and possibly the
844
845
1020,1170
1020,1170
extreme southwest corner which has a sin-
846 1020,1170 1 short incision gle wall running across it and may there-
847 1020,1170
849 1020,1170 4 1 chisel mark, 1 'tear' fore adjoin an open area- no investigation
850
852
1020,1170
1020,1170
was made beyond this square to the south
or west. This is one of the four units which

230
Figurines, Clay Balls, Small Finds and Burials

did not contain clay balls, but since the major part of The west mound
it lies within a building anyway, as do the other 17 fragments of clay balls were recovered from the
three units without clay balls, this does not help us west mound. Of these, six came from the 2 x 2 sur-
to narrow down context. However, in other parts of vey, three came from surface 10 x 10 units, and eight
the northern scraped area the situation appears to be from sub-surface units. No further information is yet
different- unit 818 is recorded as having clay balls available, so no comparisons can be made with those
largely in the western half, which appears to be a from the east mound.
court with dumps. Clay balls are also recorded at the
western edge of unit 814. A note written by the Dating
archaeologist scraping square 1045,1125, down on Following the conclusion above that the clay balls
the southeast slope of the northern eminence, re- are approximately in situ, they presumably date from
ports that clay balls are not found there, in what the same time as other material from the top of the
appears to be a large rubbish dump, but do occur in northern eminence. According to the ceramic,
houses. Since the scraping did not involve the exca- knapped-stone and figurine typology derived from
vation of closed contexts, the source of all the clay Mellaart's excavations, this is level V or VI. This may
balls recovered by this method must remain in doubt. actually be the final fling of clay balls. A quantity of
However, clay balls have been found in considerable clay balls is stored in Konya Museum, and although
numbers in an open area excavated in 1995 in to date these have not been counted or studied in
Mellaart's old trench, several of them complete. They any way, a list of their contexts was made. It is
are common in dumps and room fill in the sections possible that these all come from the 1965 season,
in Mellaart' s trench. Mellaart reported that they are although this is not clear. Nevertheless, the list is
found on rooftops, but it is clear from our data that informative- clay balls seem to be common both in
they are found in several types of context, though buildings and in outside areas in all levels from XII
not invariably in each example of each context. to VII except XL However, there are only three bags
of clay balls from level VI, although many buildings
Fragmentation and distribution of that level were excavated. Taken in conjunction
So far, no detailed records have been made concern- with Mellaart's reports that they are common in the
ing the fragmentation of clay balls. Only the most lowest levels, and the lack, or extreme scarcity, of
basic information is available, but this does tell us clay balls from the 2 x 2 survey of other areas of the
something - every ball was broken, and those from site which are known to have later occupation (the
two surface units (818 and 823), comprising 8 frag- eastern and southern eminences, and the west
ments in all, are described as 'very small'. Although mound) it may be that the date range for clay balls is
some of this fragmentation may be a result of post- restricted to the earlier phases of the site. The data
depositional factors and proximity to the surface, it may be skewed by the concentration on scraping the
does appear that clay balls were normally broken in top of the northern eminence, but this area accounted
antiquity. It may be that their weight has restricted for 19 (for 15 of which data are available) out of a
their movement, but details of the weights are not total of 35 scrape squares (23 of which have data
available, and the 'very small' fragments have not available), 32 on the east mound, 3 on the west
strayed far if at all, although they are on the north- mound, a proportion which bears no relation to the
ern slope and may therefore have moved slightly overwhelming proportion of clay balls found there.
downwards. However, considerations of weight and
size would apply also to ceramics, stone and bone, Function
and these are found all over the surface of the mound. Little attention has been paid to clay balls, and this
The gully between the southern and northern includes their purpose or function. Mellaart reported
eminences is noticeably full of large Classical sherds that they were often found on roofs, and suggested
in a deep deposit of water-lain soil. It would appear, they could have been used as slingstones when
therefore, that the distribution of clay balls is not a wielded from rooftops or put in leather bags on
random result of erosion and post-depositional proc- sticks to create a 'cheap' version of a macehead (Mellaart
esses, but that the clay balls are actually approxi- 1966, 188), but it would probably be simpler to make
mately in situ. This is exactly the situation regarding clay maceheads. Similar items have occasionally been
the figurines from the surface, which also derive reported from other Neolithic sites in the region, but
almost exclusively from this part of the site in sub- not in the quantitities that occur at ~atalhoyiik, and
surface units. there is no mention so far of any at Can Hasan, the

231
Naomi Hamilton

other major Chalcolithic site in the area. Furthermore, some items of burnt clay are likely
In attempting to understand the function of clay to have been incorrectly classified as clay balls, as
balls, the most important features appear to be the their fragmentary condition makes their real shape
burning, the fragmentation, their discard in dumps, unclear. They probably belong with a number of
room fill, and open areas, and their presence appar- baked clay objects of uncertain shape which have yet
ently in all the lower levels, but not necessarily in the to be studied but are thought to be parts of pot
upper ones. Another important point is that a number stands. Frequently they are distinguishable from clay
of fragments classified initially as clay balls actually balls only be having one flattened surface, but some
have one or more flattened surfaces, or are not genu- have squarish or rectangular bases curving towards
inely spherical (see below). the top, and one from Mellaart's backfill looks like
My suggestion is that clay balls functioned pri- an egg, broken at the base and with incised decora-
marily - though not necessarily exclusively - as tion around the top. These appear to be a develop-
simple pot stands. While it may initially seem strange ment from clay balls into recognizable pot stands.
that roughly spherical objects would be used for The two fragments recorded as having diameters of
balancing a pot on, it makes sense on several levels. 17 em may belong to this group, as they are very
1) they are simple to make; 2) they could be rolled in large for clay balls. Mellaart found several examples
hot ashes with ease; 3) once in position in the ash of Neolithic pot stands, all of undecorated baked
they should be stable; 4) round-based pots would sit clay. Three are held in Konya museum- one is a
more easily on three or four spheres than on flat- plain trapezoid vertical block, narrowing at the top
topped objects. This function would account for the (Inv. no. 1984.14.67); the other two are surmounted
burning, the fragmentation - the result of firing by two 'horns' - one is complete and measures
such a solid lump of clay - and their common dis- 102H,58W,48D (Inv. no. 1984.14.68), the other, miss-
card as easily replaceable items. The occasional'deco- ing the horn tips, is much smaller- 48H, 36W, 300
ration' in the form of dips in the surface would have (Inv. no. 1984.14.72). Three others are in Ankara mu-
facilitated adjusting their position in the fire, provid- seum, and a further three are in Aksehir. Mellaart
ing a grip for sticks. Such a function could also ex- reports that the earliest pot stands on the east mound
plain the matting imprints, which suggest that they come from level III, presumably those mentioned
were made indoors rather than at a clay source, and above. In contrast, large incised pot stands abound
therefore probably destined for use in the settlement on the Chalcolithic west mound, and also at
rather than outside it. Presumably they were placed Chalcolithic Can Hasan where they are very similar
on mats to dry, probably by the hearth, where they (Figs. 12.7:1 & 12.7:2). They frequently have holes
would be at hand when required. Their reported cut in the lower part, assumed to be for the insertion
presence on roofs, of which no evidence is available of sticks to aid moving them around and/ or for
from the surface survey (the 40 x 40 scrape area cooking on, and to reduce the fire damage likely to
appears to be coming down onto floors), is not in- result if they were made of solid blocks of clay .
consistent with a use as pot stands. Roofs must have If pot stands do not occur at <::atalhoyiik East
been used as living and working spaces much as until level III, what were the inhabitants using in-
they are now, and although clay balls would have stead? While Neolithic pottery is never abundant on
been less stable on a smooth surface than in ashes, the east mound, it is present in every level, and the
they may have been used nevertheless to support sherds from level XII do not appear to be the earliest
pots in use on the roof. We are not certain what the attempts at pot-making. Although clay balls seem to
upper surface of the roofs consisted of, whether mat- occur frequently on the top of the northern emi-
ting would have been a normal feature outdoors, nence in level VI or V, a number of these are cer-
whether there was a superstructure sheltering the tainly not entirely spherical, and other evidence
doorways and shafts (Mellaart (1962, 46) felt there suggests clay balls may cease to be common after
must have been as the shafts had unweathered plas- this time. Perhaps the apparent balls with one flat-
ter on their walls), nor whether there were ledges or tened surface represent the transition from balls to
low walls delimiting them. Indeed, broken clay balls what we recognize as pot stands. Certainly the clay
would be of use in propping up round-based pots balls need to be looked at more closely in the light of
on any smooth surface such as a floor, platform, the information discussed above. At this stage it is
shelf or possibly roof, with fragments being used as not possible to prove that clay balls were pot stands,
wedges, the broken part gripping the surface on but as excavation proceeds it will be interesting to
which it was placed. discover whether or not fragmented clay balls fade

232
Figurines, Clay Balls, Small Finds and Burials

out as fire-proof pot stands arrive on the scene. It is contains only descriptions and comparative infor-
also possible that the frequent cracking and fragmenta- mation. Because of the difficulties in giving stand-
tion of the clay balls is the result of their additional ardized descriptions of the dimensions of objects of
use as pot boilers - the frequent heating and immers- such varied shape, measurements for figurines are
ing in water leading to cracking. Experimental work is given asH (height), W (side to side), D (front to back)
needed to evaluate the feasibility of this proposal. regardless of the normal applicability of the terms.

The small finds 1) Human figurine head. Inv. no. CH94:31, Fig. 12.2:1. Unit 820,
sub-surface, 1030,1200. Beige limestone, 22H, 28W, 190. Munsell
Bone, ceramic and knapped-stone small finds have 10YR8/2 white (back), 10YR7 /3 very pale brown (top, core),
5Y5/1 gray (paint patches). Weight 15.54 g. Stylized head broken
been dealt with separately in this volume. Here I from body at neck/shoulder junction. Flat top, large hooked
shall present a number of small finds from the sur- nose, projections for ears, incised rectangular slashes gouged out
face investigations which are not covered by special- for eyes, a similar incision immediately below the nose may
ist reports. All measurements are in millimetres. The represent a mouth. The face appears to be tilted slightly up-
wards. There are traces of dark brown paint on the face (by nose,
notes below on the distribution of small finds ex- ear, chin), and on top of the head. The whole head is smooth all
clude bone, ceramic and knapped-stone artefacts. over, which may be the result of constant handling which could
It is noticeable that very few small finds come account also for the worn paint; however, paint is likely to be
from the surfaces of the mounds, particularly the unstable on this surface. This head is similar in style to a large
clay head found in Mellaart's backfill in the 1995 season, and also
east mound. Only 5 were found in the 2 x 2 collec- resembles one made of alabaster from Mellaart's excavations
tion on the east mound, with a further 9 coming (Mellaart 1964, fig. 31b, pl. XVIc).
from the surface of the 10 x 10 scrape squares and 1
from the spoil of mixed surface and sub-surface units. 2) Standing human female figurine. Inv. no. CH94:32, Fig. 12.2:2.
Unit 821, surface, 1020,1190. Pale grey baked clay, fabric fine
The majority came from sub-surface units- 32 from with occasional fine mica and black mineral inclusions. 45.5H,
simple sub-surface units, 48 from combined sub-sur- 34W, 330. Munsell 5Y6/1 grey /light grey. Weight 38.49 g. This
face and scrape units, 11 from simple scrape units freestanding figurine is missing its head and one arm, and the
and 1 from a lower sub-surface unit. On the west breasts were shattered by a shovel. The base is roughly cylindri-
cal, widening at the middle to form a large stomach and large
mound the surface has been altered by the recent protruding, rounded bottom. Above the waist it narrows sharply
and continuing use of the mound for the winter into an eliptical upper body with stub arms. Most of the surface
storage of chaff. This involves digging a broad, shal- of the figure is smooth as though well handled. There are dark
low hollow into the surface of the mound, placing grey patches on the front, and the back, sides and base have a
whitish deposit, probably lime.
piles of chaff inside, and covering them with soil
and vegetation obtained from the hollow and its 3) Fragment of human standing figurine. Inv. no. CH93:30. Unit
surrounding area. As a result, much sub-surface soil 282, sub-surface, 1030,1190. Pale grey very hard baked clay, fab-
is now on the surface of the mound, and this may be ric fine with occasional fine mica and black mineral inclusions.
26.6H, 30W, 230. Munsell5Y6/1-4/1 (core), base 5Y4/1. Weight
reflected in the finds survey. A total of 23 small finds 13.70 g. Lower part of freestanding figure with cylindrical base,
was collected, of which 2 are from the western edge identical in style and fabric to no. 2 above, with the front missing.
of the mound outside the survey area, 14 were found Smooth outer surface, covered with whitish deposit in lower
in the 2 x 2 survey, and 7 came from the 10 x 10 region, probably lime.
These two figures are practically identical, and were found in
scrape squares - 1 from a surface unit, 2 sub-sur- adjacent squares on the northern eminence, overlying Neolithic
face, and 4 mixed surface/sub-surface. Although 16 architecture (see discussion above).
of the 23 finds appear to come from the surface, in
contrast to the situation on the east mound, this is 4) Standing human female figurine. Inv. no. CH93:4, Fig. 12.2:3.
Unit 279, scrape, square 1040,1090. Grey baked clay. 30H, 13W,
probably a distortion created by disturbance. 150. Munsell 7.5YR7 /2 pinkish gray (exterior), N4/0 dark gray
(interior). Weight 6.33 g. Missing head and left arm/upper body
Small finds from the east mound and chipped on the lower front. Roughly cylindrical base bulg-
As all these small finds are unstratified, help in iden- ing a little to portray hips and slightly protruding bottom. The
upper body remains thick, showing either a large drooping stom-
tifying and dating them can be sought from the pre- ach or breasts, with the surviving short horizontal arm shown to
viously excavated material at this site, and that of be bent forward with hand touching breast. Two deeply incised
other sites in the area of similar date. 'lines depict a pubic triangle. There is no close parallel for this
figurine among the surface finds nor among Mellaart's exca-
vated material.
Figurines
A general discusison of the context of these figurines 5) Fragment of human figurine with impressed decoration. Inv.
can be found earlier in this chapter. This catalogue no. CH93:31, Fig. 12.2:4. Unit 24, sub-surface, square 990,1040.

233
Naomi Hamilton

Q .
.
.

-
.

~~:
fl '/ .

:~
'i. -
2 .,. ·-_
.

~ ·~;~: :. -
D .
-
5

1
-- --
~ I

~
Vj
I
'
~ 1
. '
- ~ . '
~

? ' > :~ I •'I

6
--

-- 8

\>YJ I
- 9

Figure 12 .2 . F'lgunnes
·
~10 \1
from (ata /h oyiik East.. Ford escnptiOns
. . see text.

234
Figurines, Clay Balls, Small Finds and Burials

Fire blacke ned clay. 38H , 31 W, 170. Munsell 7.5Y R5 / 2 bro w n to level VI, poss ibl y level VII. Humanoid fi gures occur onl y be-
N3 / very d a rk g ray. We ig ht 14.72 g. Badly d a maged fragment of tween levels VIII and V.
th e upper pa rt of a hum a n fi gure, mi ss ing th e hea d . Th e a rm s
seem to be dep icted ha ng ing by th e sid e o f th e torso, the shoul- 8) Hum a noid fi g urine. ln v. no. CH95:39. Unit 901 , sub-surface,
ders slopin g sha rpl y. Th e fro nt has circular d epress ions in verti- 1030,11 60. Pa le beige/grey baked clay. 17H, lOW, 12.50. Com-
ca l lines, about 1 mm d eep . Sex is not indi ca ted. The fi gure gives pl ete. Sma ll fi gure w ith ovoid base rising with gentl y inwa rd-
th e impressio n of being cl oa ked. There is no para ll el a m ong slopin g sid es to a neck / hea d fea tureless except for a slig ht nose.
Mellaa rt's excava ted ma terial, altho ugh cl oa ked fi g urines were Thi s does no t have di vid ed 'legs' . Roughl y smoothed su rfa ce.
fou nd. A clay frag men t w ith similar d ecora tio n was found in th e Slig ht chip on base. A simila r fi gure was found by Mella a rt in
su r face survey a nd may be part of a fi gurin e (no. 14). level V (Konya Mu s. 1984.14.217) while a second , also from level
V (1 984.14.21 6) has th e sa m e d esig n with the additi on o f breasts.
6) Hum a noi d figurin e. ln v. no. CH94:34, Fig . 12.2:5 . Unit 816,
su b-surface, 1050,1 170 . Da rk brown ba ked clay . 20H , 11 .3W, 170 9) Animal figurine. ln v . no. CH94:33, Fig . 12.2:7. Unit 820, sub-
Mun sell 7.5YR3 / 0 very d a rk g ray, 2.5Y8/2 w hite (base). Weig ht surface, 1030,1200. Lig ht brown baked cla y. 27H , 21W, 360. Munsell
2.64 g. Co mpl e te. Frees tanding fi g ure w ith ovo id fl a t base mad e 10YR6/3 pa le brown . Weig ht 13.26 g. Almos t complete - one
of two 'a rm s' or 'legs' pinched toge th er, sloping up wa rd s a t a 45 front leg broken. Sta nding qu adruped which is genera ll y bov ine
deg ree a ng le to a thi ck neck surmo unted by po inted face w ith in a ppea rance but lacks horns, possibl y therefore some other
round ed top crea tin g a bird -like hea d. A sli g ht hunch w here the anim a l. Th e fi g ure was w ell mad e, with a ttentio n to d eta il -
'legs' jo in th e neck g ives a suggesti on o f a nn s o r bent kn ees. eyes, ea rs a nd nose are d epicted , w hi ch m akes the lack o f horns
C ru dely mild e. A s li g h t polis h around the back a nd s id es su g- releva nt. There il re two possible s tilb marks, on e on the rea r, one
ges ts hand li ng. A comparison w ith simila r fi gures from Mell aa rt's und ernea th. These a re generall y found on m od els of ca ttle a nd
excava ti o ns sugges ts this d a tes to leve l VI, possibl y leve l VII . boar. Th e surface is fa irl y rough, smoother on the le ft sid e.
Hum a noid figures occur onl y between leve ls VIJI a nd V.
10) Fragm entary a nim a l fi g urine. In v. no . CH 95:42. Unit 901 ,
7) H uma no id fi g urine. ln v. n o. CH94:1, Fig. 12.2:6. Unit 818, sub- sub-surface, 1030,11 60. G rey-brown lightly baked cla y w ith pur-
surfilce, 1020,1200. Light b rown ba ked clay . 31H, 10.5W, 200. pl e a nd ru s ty tinges. 38H , 27W , 300. Badly d a ma ged fra g ment
Mun se ll 10YR6/ 1 li g ht grey I grey (surfa ce), 10YR6/ 4 li g ht yel- d epi ctin g th e rea r end o f a s ta nding qu a druped w ith a short
low ish b row n (core). We ig ht 3.94 g. Broken hum a no id fi g ure o f stumpy ta il. Th e fi gure has mu ltipl e d eep stab ma rks whi ch may
simila r ty pe to no. 6 but wea rin g a headdress or sca rf w hich have co ntributed to th e heavy dama ge it ha s suffered .
protrud es a t th e back o f th e hea d , slopin g d ow nwa rd s. One
'a rm / leg' is mi ss in g. The remaining one has a fl a t base curv ing 11) H ead o f anim a l fi gurin e. ln v. no. CH94:2, Fig. 12.2:9. Unit 824,
s lightl y u pward s a t its point. This is surmounted by a long thi ck sub-surfa ce, 1040,1200. Beige ba ked cla y with black sand inclu-
nec k lea nin g s li g htl y fo rwa rd s, w ith hea ddress form ed a ll in one sions. 28.5H , 37W, 58.50 . Munsell 2.5Y6/ 2 li g ht bro wni s h grey,
by p in chin g th e clay a t th e top o f th e neck upwa rd s th en bendin g 2.5Y5/5 g rey ish brow n (ex te rior), 2.5Y4 / 0 dark grey (interi or).
it d ow n. No face is s how n. Roug hl y mad e. A co mpa rison w ith We ig ht 40.11 g . Ca refull y mod elled hea d o f proba bl y ho rned
s imilar figures from Mell aa rt' s excavatio ns sugges ts thi s d a tes to animal, bro ken off th e bod y a t the top of th e neck, a nd miss in g

Figure 12.3. Human figurine from (ata/hoyiik East. Figure 12.4. Animal figurine from (atalhdyiik East.

235
Naomi Hamilton

one horn. The shape of the head is suggestive of an equid, and 25H. Munsell10R3/ 4 dusky red, 10R5/ 4 weak red, 5YR8/2 pink-
the surviving horn could possibly be an ear. Equid has been ish white veins. Weight 40.40 g. The shape of this dish is unclear
found in great quantity at the contemporary site of Pmarba~1, not as the fragment has one almost straight edge followed by curv-
far distant at the base of Karadag. Alternatively this could be a ing ones- it could be sub-rectangular, oval, or a poorly shaped
sheep or goat, rather than cow /bull. Despite the careful model- circle. The sides are curved, the rim almost flat on top. A long
ling, no eyes or mouth are marked. It is unusually large, al- foot survives, original length unknown. Its oblong shape sug-
though a few figurines of this size were found in Mellaart's gests the dish is not round.
excavations.
20) Fragment of a bowl. Inv. no. CH93:36. Unit 269, lower spit,
12) Animal head lug. Inv. no. CH94:3, Fig. 12.2:10. Unit 816, sub- square 1040,1040. Creamy/white marble. 65L, 39W, 11TH.
surface, 1050,1170. Pinky brown fired clay. 27H, 26W, 29D. Munsell2.5Y8/2 white. Weight 30.27 g. This rather thick triangu-
Munsell 5YR7 /4 pink (exterior), 7.5YR7 /4 pink (interior). Care- lar body fragment has no rim or base, and comes from a large
fully modelled bovine head with eyes depicted, horns broken off. deep bowl.
Traces of a light burnish to surface. A similar lug was found by
Mellart in level V and is in the Konya museum (lnv. no. 1985-14- 21) Fragment of a bowl. Inv. no. CH93:37. Unit 277, lower spit
113). plus scrape, square 1090,1040. White alabaster or marble. 44L,
27W, 4TH. Munsell 2.5Y8/2 white. Weight 7.23 g. This is a small
13) Fragment of animal head. Inv. no. CH94:4. Unit 565, section 6. triangular rim sherd from a thin-walled open bowl form with a
Dark grey baked clay. 36H, 25W, 17D. Munsell N4/- dark grey. diameter of c. 20 em.
Weight 8.93 g. Badly broken neck and head of animal, consisting
of neck broken off at shoulder, plus head with one ear but no Discussion:
face. Possibly an equid? Smoothed surface. The first three of these vessels appear to date from the end of the
Neolithic or the early Chalcolithic. Several similar examples were
14) Fragment of human figurine? Inv. no. CH93:32. Unit 288, found by Mellaart, although stone vessels are rare. He comments
scrape, 1030,1180. Beige baked clay. SOH, 38W, 19D. Munsell (1964, 84-5) that fewer than a dozen stone vessels had been
10YR7 /2light gray. Weight 27.54 g. This irregular fragment may found in three seasons of work, but these included a thin-walled
be part of the torso of a figurine. Its smoothed surface is deco- marble bowl on crescent feet from level IV, four spouted dishes
rated with a horizontal row of impressed circles 1.5 mm deep, in fine veined red limestone from various shrines in level VIA,
reminiscent of those on no. 5. and a white marble spouted dish pierced either side of the spout,
from a grave in VIA:29. He illustrates (1962, 53) a square mini-
15) Humanoid figurine fragment? Inv. no. CH94:8. Unit 710, ature table/footed dish of clay found in AIII:4, adding that stone
scrape, 1030,1170. Dark brown baked clay. 18H, 21.5Dia. Munsell dishes in white veined marble with four feet occurred similar to
10YR3/2 very dark grayish brown. Weight 6.53 g. Conical clay ceramic versions. In addition to these examples from the upper
object with its top missing. This may have been similar to no. 8. levels of the east mound, two corner fragments of footed square
stone vessels were found during Mellaart's excavations on the
16) Horn-shaped fragments of baked clay were found in units west mound (1965, 148-51, fig. 11.18, 19), both unfortunately
822 (x2), 824, 826, 295, and 269. Details will be available on the unstratified. One has incised, the other relief, geometric decora-
data base. tion. These two pieces provide the closest parallels to our finds
17 and 18, while 19 is closer to the examples from the east mound.
A number of baked clay fragments which could belong to figu- Stone bowls are also reported from various Chalcolithic levels at
rines but cannot easily be understood are mentioned at the end Can Hasan (French 1962; 1963; 1964; 1967; 1968). None has yet
of the section on clay objects, below. been published, so comparisons cannot be made. Miniature footed
dishes and tables in clay have also been found (see below no. 39)
Stone vessels Nos. 20 and 21 seem as yet to have no parallels within the
17) Fragment of a footed rectilinear dish/miniature table. Inv. no. excavated, published material from the site. No. 20 was found in
CH93:33, Fig. 12.5:1. Section 3. Volcanic stone, beige with linear the same unit as no. 19, and the dateable finds from this unit
black shiny inclusions. 45 x 50, 25H. Munsell 10YR8/3 very pale were mainly Neolithic, but the southern part of this square had
brown. Weight 28.73 g. This is a corner of a shallow, probably considerable remains of the Classical period. Unit 277 from which
square, vessel with a flattish base and sloping sides - one side no. 21 came lay below a child's grave of the early Byzantine
sloping inwards, the other outwards. A curved rim, slightly flat- period, in a part of the site (the eastern skirt) which appears to
tened on top, survives on two sides. One foot is present, pendant have very late Neolithic occupation. Several similar fragments
from a raised 'strut' which appears to run across the base on one were found in Mellaart's backfill in 1995, but although this spoil
side of the dish. came generally from Neolithic strata, some parts of his excavated
area were damaged by Hellenistic pits. Therefore, although these
18) Fragment of rectilinear shallow dish with incised decoration. pieces are probably prehistoric, they may also belong to the later
Inv. no. CH93:34, Fig. 12.5:2. Unit 278, sub-surface, square periods on the mound. The nearest prehistoric parallels appear
1040,1090. Pale brown stone. 32L, 22W, 28H. Munsell 10YR7 /3 at Hacilar in most levels (Mellaart 1970, 149-52), including
very pale brown. Weight 16.95 g. This is a corner of a shallow, aceramic (2 fragments), but particularly the late Neolithic level
probably square, vessel with a flattish, thick base and slightly VI (21 bowls and fragments), and the early Chalcolithic levels liB
convex, slightly outward-flaring sides, rim sloping inwards. The and IIA (24 bowls and fragments). All but 2 of these are white
exterior of one side has faint incised decoration in a geometric marble. The majority are small footed bowls, and Mellaart com-
design of diagonal lines. ments on the frequency of mending holes, suggesting that al-
though not rare, such items were not easy to make or obtain. We
19) Fragment of footed sub-rectangular dish. Inv. no. CH93:35, await the publication of Can Hasan to discover whether the Early
Fig. 12.5:3. Unit 269, lower spit, square 1040,1040. Veined pink, Chalcolithic stone bowls there were of marble as many Middle
red and white conglomerate with white inclusions. 69L, 43W, Chalcolithic ones were.

236
Figurines, Clay Balls, Small Finds and Burials

0
.
. ·
:
: ..

----
.,
.,
3
4

r ."
I
\
QBJ I

8 5 0-0 0 6

F 4

[ 8

Figure 12.5. Stone artefacts from c;atalhoyiik East. For descriptions see text.

237
Naomi Hamilton

Ground-stone tools with at least four triangles cut into its upper surface to a depth of
22) Macehead. Inv. no. CH95:41, Fig. 12.5:4. Unit 906, sub-sur- 1-1.5 mm, and a groove with semi-circular profile cut into one of
face/scrape, square 1020,1160. Orange-brown slightly mottled its long sides. One end is broken off. Its purpose is unknown -
polished stone. 66Dia, 52H. Complete. Neolithic. This is a the triangles could be used as moulds, but the groove is more
squashed sphere in shape with flattened upper and lower sur- difficult to explain. The context is strongly Neolithic, but there
faces. The central perforation narrows at the middle. The surface are no parallels for this artefact.
has been scratched in places. Over about a third of its surface the
stone alters, becoming a pale orangey-cream colour and, although 28) Stone palette fragment? Inv. no. CH94:17. Unit 569, section 8.
well polished, it is not shiny, only smooth. Typologically this Reddish-grey sandstone. 51Dia?, 11H. Weight 21.25 g. Part of an
macehead is unusual in having flattened surfaces, a feature found irregularly shaped stone disc with parallel upper and lower sur-
on only four of the 29 maceheads found by Mellaart, two in level faces, one smooth, the other rough. Probably part of a palette of
VI, two unstratified. Two more have one surface flattened, both the type Mellaart found, used for grinding pigment.
from level VI. Overall, 14 were found in level VI, two in level V,
one in level VIII where it was found in an unusual burial; and 29) Incised stone fragment. Inv. no. CH93:46. Unit 287, sub-sur-
one in level !I, with 11 unstratified. face, 1030,1180. 53H, 43W, 44TH. Munsell 5B4/1 dark bluish
gray. Weight 127.53 g. Badly damaged worked stone with prob-
23) Celt. Inv. no. CH93:6, Fig 12.5:5. Unit 264, sub-surface, square ably sub-rectangular base which was polished but is very chipped.
1040,1190. Mid grey-green polished stone. 44L, 25W, lOTH. All vertical sides and the top are broken, but one face has a
Munsell SGS/2 greyish green. Weight 21.23 g. This celt has gen- substantial horizontal depression with a semi-circular profile
tly curving parallel edges narrowing towards the butt, and a running right across it. This depression is highly polished, and
fairly flat well-ground blade. In profile it tapers smoothly to a incised with irregular criss-crossed lines. Unit 287 is strongly
fine blade, narrowing slightly at the butt. Broken at blade end, Neolithic.
and some damage on the butt.
Details of other fragments of ground stone will be available on
24) Celt. Inv. no. CH94:37. Unit 854, sub-surface/scrape, square the data base.
1040,1120. Dark green stone with black and light green speckles.
43L, 25.5W, 11TH. Weight 20.26 g. Polished celt similar to no. 23 Beads
but with narrower butt and curved blade. In profile it narrows 30) Clay bead. Inv. no. CH93:47. Unit 95, surface, 2 x 2. Dark
considerably towards the butt as well as the blade, and is un- brown unbaked clay. 18L, 6Dia. Munsell 7.5YR4/2 dark brown.
evenly shaped. Broken at blade end. Four fragments of blades of Weight 1.05 g. Cylindrical bead perforated longitudinally.
similar celts and axes were found in units 269 (lnv. no. CH93:38), Neolithic. A similar bead was found in unit 905 (inv. no. CH95:45).
818 (inv. no. CH94:5), 290 (inv. no. CH93:39) and section 3 (inv. Such beads occur among the grave goods found in Mellaart's
no. CH93:40). Three are of dark green/black stone, the fourth of excavations, although they are not very common.
grey-green stone. Details will be available on the data base.
31) Clay bead. Inv. no. CH93:48. Unit 265, scrape, square 1040,1190.
25) Celt fragment. lnv. no. CH93:41, Fig. 12.5:6. Unit 61, surface, Orange-beige unbaked clay. 18.5Dia, 20H. Munsell7.5YR7 I 4 pink.
square 1040,1040. Mid-dark green stone with grey streaks. 19L, Weight 3.79 g. Roughly spherical bead with slightly flattened top
ISW, 8.5TH. Munsell5GY4/1 dark greenish grey. Weight 5.15 g. and bottom, pierced vertically. Neolithic. Similar beads were
Highly polished fragment of a miniature celt. Blade missing. found in units 269 (x2), 281, 283 (x3), 288, 820 and section 3. They
Several very similar complete examples were found during are paralleled by beads found by Mellaart among grave goods,
Mellaart's excavations. A fragment of a similar miniature celt although they are not very common.
was found on the surface, Inv. no. CH93:42, made of the same
stone type. Details can be found on the data base. 32) Clay bead. Inv. no. CH93:49, Fig. 12.6:1. Unit 284, sub-surface,
square 1040,1180. Dark grey unbaked clay. 23L, 9Dia. Munsell
The majority of these celts come from strong Neolithic contexts N4/0 dark gray. Weight 1.82 g. Elongated biconical bead pierced
and are within the broad range of celts and axes found during longitudinally, chipped at both ends. Neolithic. A similar bead
Mellaart's excavations. was found in unit 264, and parallels are found in the grave goods
Two fragments of axes with strongly angled facets at the butt from Mellaart's excavations, although they are not common.
end were found in units 156 (lnv. no. CH93:43) and 808 (lnv. no.
CH94:27), both of dark green-grey stone. Details will be available 33) Shell bead. Inv. no. CH93:50. Unit 269, lower spit, square
on the data base. 1040,1040. Off-white dentalium shell. IlL, 3.5Dia. Munsell5YR8/
1 white. Weight 0.25 g. Short slim cylindrical section of dentalium
26) Pounder. Inv. no. CH 93:5, Fig. 12.5:7. Unit 269, lower spit, shell. Part of a similar but much wider and shorter bead was
square 1040,1040. Dark green/black stone. 109L, 54W, 51TH. found in unit 506 (inv. no. CH94:61), and there are many in the
Munsell5GY3/1 dark greenish grey. Complete roughly cylindri- grave goods from Mellaart's excavations.
cal pounder narrowing at one end. A chip had broken off the
broader end and the scar had been reworked in antiquity. This is 34) Stone bead. Inv. no. CH95:36. Unit 900, surface, square
the only complete grinding tool found during the surface investi- 1030,1160. Off-white polished stone. 9.25Dia, 1.25TH, 3.65 hole
gations. The date of this object is not certain- although unit 269 dia .. A disc-shaped bead with flat surfaces and straight sides
contained largely Neolithic material, artefacts of the Classical showing sharp angle between the two. This shape is typical of
period have also been found in this area. beads found in huge numbers among grave goods during
Mellaart's excavations, although this one is very large- the vast
Miscellaneous ground stone majority do not exceed 5 mm dia., although some are of this size.
27) Mould fragment? lnv. no. CH93:44, Fig. 12.5:8. Unit 281, sub-
surface, square 1030,1190. Mid-grey stone with dark and light 35) Stone bead. lnv. no. CH94:6, Fig. 12.6:2. Unit 812, sub-sur-
grey mottling. 30L, 9W, 7.5H. Munsell 5B4/1 dark blue grey. face/scrape, square 1050,1190. Beige polished stone. 11L, 6W,
Weight 7.47 g. This is part of a rectangular piece of ground stone I .5TH. Munsell IOYR7 /3, very pale brown. Weight 0.212 g.

238
Figurines, Clay Balls, Small Finds and Burials

-Qo 2
-
f)
.

@
n_rn
'
u v
9 I

.~
'
0
3

~ I

g~g
I

Figure 12.6. Miscellaneous small finds from c;atalhoyiik East. For descriptions see text.

239
Naomi Hamilton

Eliptical thin flat bead pierced centrally. This is an unusual shape children's graves in level II, which appears to be Early Chalcolithic.
for a Neolithic bead. One has four feet, and is incised with a geometric design similar
to that on find no. 18 above. It was accompanied by a clay spoon
36) Stone bead. Inv. no. CH93:51, Fig.12.6:3. Section 3. Turquoise. (Silistreli 1984, 33, fig. 12). The other has no feet, and no decora-
9L, 6.5W, 4.5TH. Weight 0.48 g. Eliptical both longways and in tion, but a small lip or groove on the rim on one side (Silistreli
section, pierced longitudinally. Turquoise is a long-distance im- 1990, 97, fig. 8). The best Neolithic parallels for this table occur at
port, its nearest known sources in Iran and Sinai. Turquoise Hoyii<;ek in the Burdur region (Duru 1991; 1993). Two miniature
beads are rare among the grave goods excavated by Mellaart, but tables, rather larger than ours, were found in Level II, dated to
generally occur in unusual shapes, often more complex than this the late Neolithic period. One has a rectangular upper surface
one. Fragments of two more turquoise beads were found in units with four tall tapering legs adorned with animal heads in relief,
156 (Inv. no. CH93:61) and 191 (Inv. no. CH93:60). Details can be and was found in a silo; the other has a cut-away stepped design,
found in the data base. with four short legs clustered under the centre, and was found
on the floor of the work area outside the shrine. Both are viewed
Clay objects as offerings. The early Chalco lithic levels of nearby Kuru<;ay also
37) Sub-oval stamp with geometric meander design. Inv. no. have miniature clay tables (Duru 1994, 63-4, figs. 203-4). A square
CH94:30, Fig. 12.6:4. Unit 601, lower spit, square 990,1080. Brown table with four thick legs and shallow top has an unclear paint-
baked clay with black sand inclusions. 28L, 22W, 17H. Munsell ing on the top, suggestive of two opposing hull's heads and a
10YR6/2 light brownish gray. Weight 7.80 g. Flat face with similar decoration in relief on the side; another has a three-
rounded upper pulled up into a central knob. Broken at one end, fingered hand painted on its upper surface; and a number of
about 70 per cent of the seal face survives, with the design incised to fragments were found. Mellaart offers comparisons with items
a maximum depth of c. 3 mm. Original length probably c. 40. from Demircihoyuk, further north in the Eski~ehir region. Al-
though distant from <;:atalhoyuk, recent survey work in the
38) Rectangular stamp with symmetrical geometric design. Inv. Eski~ehir region suggests that it was in contact with Konya plain
no. CH95:37, Fig. 12.6:5. Unit 902, spoil from 1994 40 x 40 on cultures from the end of the Early Chalcolithic period at the very
north eminence. Mid-grey-brown baked clay. 27L, 24W, 20H. latest (Efe 1990, 39), and Duru considers it possible that the
Flat face with pyramidal back pulled up into knob. Both ends are Demircihoyuk tables are Late Neolithic (1993, 132).
broken, about 85 per cent of the decorated face survives, with a
design of opposed chevrons incised to a depth of c. 3 mm. Origi- 40) Fragment of horned pot stand/portable hearth. Inv. no.
nallength probably c. 30. The knob has signs of burning, and is CH94:7. Unit 504, sub-surface, square 980,1080. Pale brown baked
chipped on top. clay with sand and mica inclusions. 111 L, 64W, 133H. Munsell
10YR7 I 4 very pale brown. Weight 680 g. This is one end of what
Both these stamps are similar in style to the twenty-three exam- was probably a crescent-shaped pot stand with vertical ends and
ples excavated and published as stamp seals by Mellaart (1963, a centrally placed third support. The surviving end has a stubby
figs. 40 & 41). The design on no. 37 resembles several of them, horn-type projection for supporting a pot, and is decorated on
particularly those from levels II and III, whereas no. 38 is rather the outer edge at the base of the horn with a human ear. Others
different and stylistically might be considered a little earlier as were found on the west mound (see below).
the symmetrical designs tend to occur in levels VI and IV. Both
stamps are fairly small compared to Mellaart's examples, al- 41) Fragmentary model human face. Inv. no. CH93:1, Fig. 12.6:7.
though as no scale nor dimensions are given in the illustrations, Unit 124, scrape, square 1040,1040. Grey ceramic with red paint.
and only nine of them have been located in the museums, com- 57H, 31W. Munsell10YR6/1 gray, with 10R4/8 red paint. Weight
parisons are difficult. 31.7 g. Broken at the top and one side, around 60 per cent of the
Perhaps it is time to change the name of these items from original width survives; the original height is uncertain. Hollow
seals to stamps, as there is no evidence that they were used for back. This is a detailed, slightly stylized, modelled human face
sealing, not a single fragment of a sealing showing one of these with straight nose, unsmiling mouth, heavily lidded eye, and
designs having been found. Mellaart suggested that they could detailed ear. A strong line runs from the outer edge of the nose to
have been used to stamp textiles, or to stamp designs on human the jaw. Patches of red paint survive on the cheek, nose, chin and
bodies. No sign of dye was found on them by Mellaart, but it is brow, and may originally have covered the whole face. The date
not clear whether chemical analyses have been made. is unknown, no parallels have been found. Clearly it is not
Neolithic, and no finds are known from the site between the
39) Miniature clay table. Inv. no. CH93:2, Fig. 12.6:6. Unit 269, Chalcolithic and Hellenistic periods. It does not appear to be
lower spit, square 1040,1040. Grey unbaked clay. 30L, 30W, 20H. modern.
Munsell N6/ gray. Weight 21.98 g. Handmade with fingernail
marks visible. Complete. This is an irregular square table on two 42) Fragmentary loom weight. Inv. no. CH94:35. Unit 603, sub-
bar legs running parallel to each other, and vertical sides sur- surface, square 990,990. Brown baked clay. 63H, 45W, 35D.
rounding the flat upper surface. In the centre of the upper sur- Munsell 5YR6/ 4 light reddish brown. Weight 101.43 g. Trap-
face is a hemispherical clay blob. ezoid in shape, with flat sides and base, narrowing towards the
This is possibly a model offering table, complete with offer- top both from the side and front, it was pierced near the top with
ing in situ. Nothing quite like it is known from any site, but a substantial hole. This is probably a loom weight, although it
miniature clay tables were found by Mellaart in the later levels of could be a small andiron. Unit 603 was part of a late cemetery,
the site (1962, 55, fig. 9 /3). This one is particularly small, and has probably of Byzantine date. The top is broken off, but half of the
unusual legs - similar square examples either have four inde- piercing remains.
pendent feet, or are dishes without feet. Unit 269 had mainly
Neolithic material, and there can be little doubt that this is a 43) Fragmentary loom weight or sheep tag. Inv. no. CH93:52.
prehistoric artefact. Unit 191, scrape, square 940,1040. Grey baked clay. 90L, 48W,
The nearest site from which similar objects are reported is 16TH. Munsell 10YR7 /1 light grey (exterior), 10R6/ 4 pale red
Ko~k Hoyuk, where two square baked clay vessels were found in (core). This is slightly more than half of a circular, flat object,

240
Figurines, Clay Balls, Small Finds and Burials

pierced off-centre and incised on one side just below the piercing 10YR7 /1 light grey. Weight 181.34 g. Probably the base of either
with a horizontal line with three surviving pendant lines. Sug- a figurine or pot stand, this badly damaged object has sub-rec-
gested to be a sheep tag because of its inscription, it could also be tangular base rising vertically at the sides and sloping slightly
a loomweight. This unit lay about 25 em below the surface, and inwards on the incised face. This face appears to be the front, as
contained some late material. This object probably belongs to the the sides then curve at around the level of the decoration, giving
Classical period. the impression of hips and waist. The decoration consists of a
lower line slightly off the horizontal, and above it two further
44) Fragmentary shaft straightener/polisher. Inv. no. CH94:9. lines, one parallel to the first, the other raised at one end to form
Unit 565, section 6. Light brown baked clay with sand and mica a triangle. Diagonal incised lines fill this triangle. Unit 196 is on
inclusions. 74L, 22W, 44TH. Munsell 7.5YR6/4 light brown. the low eastern eminence and overlay or contained a Byzantine
Weight 89.88 g. Ovoid in shape, broken on one side of the central child burial. It bears some resemblance to a few figurines from
slot, with a thick base. This is similar to other shaft straighteners Mellaart's excavation, but may well be something else. Its date is
found by Mellaart, and two which have been found in his backfill, uncertain.
one of them also made of clay, the other of stone which is more
usual. 52) Fragment of impressed clay. Inv. no. CH93:55. Unit 264, sub-
surface, 1040,1190. Pale beige baked clay. 32H, 33W, 18TH.
45) Star. Inv. no. CH94:10. Unit 814, sub-surface/scrape, square Munsell 10YR8/2 white. Weight 18.89 g. Badly damaged frag-
1050/1180. Beige baked clay with sparse small white mineral ment decorated with thumbnail impressions in a broad row.
inclusions. 39 x 36, 11TH. Munsell2.5R6/ 6light red. This is a flat, Possibly part of a figurine or pot stand. Unit 264 overlies good
six-pointed star of no clear purpose. It is not pierced. This unit Neolithic architecture.
produced largely Neolithic material, and this is probably
Neolithic. Miscellaneous
53) Bone ring. Inv. no. CH93:56. Section 2. Shaped and polished
46) Circular clay token? Inv. no. CH94:11. Unit 565, section 6. bone, creamy brown. 24 Dia. ext., 18 Dia. int., 5.5W. Munsell
Beige baked clay. 17.5Dia, 7H. Munsell10YR6/2light brownish 10YR8/3 very pale brown. Half a ring, apparently originally
gray. Weight 2.22 g. Small clay disc with slightly concave faces circular in shape. Similar rings were found during Mellaart's
and convex sides. A similar object was found in Mellaart' s backfill excavations as grave goods, and frequently they are irregular in
in 1995. No other parallels are known. shape.

47) Fragment of weight or spindle whorl? Inv. no. CH93:53. Unit 54) Plaster cone. Inv. no. CH94:14. Unit 816, sub-surface, square
156, sub-surface, square 940,1040. Grey baked clay. 31Dia, 25H. 1050,1170. Creamy white roughly smoothed plaster. 33H, 24D
Munsell 10YR7 /2 light grey. Weight 13.33 g. Fragment of circu- max. Munsell 10YR8/2 very pale brown. Weight 11.66 g. Solid
lar, centrally-pierced object with vertical sides and rounded junc- cone with roughly flattened top. This object has no obvious pur-
tion between sides and top/bottom. Original diameter c. 35, just pose but could also be part of a humanoid figurine. The unit is
under 50 per cent survives. Probably Classical. strongly Neolithic.

48) Mini clay balls. A number of small clay balls have been 55) Mosaic fragment. Inv. no. CH94:15. Unit 291, scrape square
found, ranging in size from 5-21 mm diameter. There is no rea- 990,1090. Grey-blue and white stone tesserae. 86L, 71W 16TH.
son to suppose they all have the same function, which could Weight 81.70 g. This chunk of mosaic has a later plaster layer
include counters, tokens, toys, gaming pieces etc .. They are all covering the tesserae. Classical.
roughly made of lightly baked clay, and they come from Neolithic
contexts. At least 9 have been found on the east mound, all but 56) Stone pendant? lnv. no. CH94:16. Unit 527, lower spit/scrape,
one in sub-surface units overlying Neolithic architecture on the square 980,1090. Black metallic stone, probably a single crystal.
northern eminence. Details of these will be made available on the 23L, 16W, 3TH. Weight 2.52 g. Irregular sub-rectangular flat
data base. The largest of these has a flattened surface, and may shiny stone with a hole drilled through in one corner and natural
have a different purpose- several similar objects were found by depression in the adjacent corner, giving the impression of eyes.
Mellaart in houses. However, there may have been no intention to make a face on
this object, which appears to be a pendant. Similar pendants of
49) Horn-shaped decorated object. Inv. no. CH94:12. Unit 706, mica and slate occur among the grave goods excavated by
sub-surface, 1000,950. Fine light grey baked clay with sparse Mellaart, none of them engraved with faces or designs.
small mineral inclusions. 43L, 26W, 17TH. Munsell10YR7 /1light
gray. Curved into a semi-circle, broken at both ends, this object 57) Metal. Coin. Inv. no. CH93:57. Unit 294, sub-surface, square
has broad incised lines radiating outwards on one side. The other 1040,1240. A coin was found in this heavily Roman unit. No
side is undecorated. Both faces have traces of an abraded dark, details could be made out. Another coin was found in unit 188
glossy slip or paint. This could be part of an animal figurine, or a (lnv. no. CH93:45). Details will appear in the data base. Several
handle. Date unknown. iron nails and a coffin fitting were also found, probably all deriv-
ing from Byzantine burials on the mound.
50) Fragment of pot stand? Inv. no. CH94:13. Unit 707, lower
spit/scrape, square 1000,950. Red-brown baked clay with fre- 58) Glass. 7 fragments of ancient glass have been recovered from
quent gritty inclusions, surface burnished. Dimensions 34H, 32W, the surface of the mound. Of these, six are parts of bracelets units
20TH. Munsell2.5R6/6light red surface, 2.5YR6/ 4 colour fabric, 248 (CH93:58), 244 (CH93:59), 672 (CH94:45), 701 (CH94:58), and
5R7 /2 pinkish gray core. This object is unclear. Described as a 706 (CH94:65; CH94:44), one is a flask or jug handle unit 710
figurine or vessel fragment, it may be the upper part of a pot stand. (CH94:57) and one is body fragment from a vessel unit 701
(CH94:59). All the bracelet fragments are of blue glass, the han-
51) Incised baked clay object. Inv. no. CH93:54. Unit 196, sub- dle is green, the vessel fragment is semi-transparent white. De-
surface, 1090,1040. Beige baked clay. 62H, 56W, 53TH. Munsell tails of these can be found in the data base.

241
Naomi Hamilton

The west mound shell. 24L, 9Dia. Pierced through one side and its broader end.
59) Fragment of anthropomorphized pot stand/portable hearth.
Inv. no. CH94:36, Fig. 12.7:1. West edge of west mound, surface. 66) Stone bead. Inv. no. CH94:24. Unit 855, surface, square
Light grey baked clay with black sand inclusions. 120H, 113W, 480,1040. Light blue stone. 3.5Dia, 1.5TH. Weight 0.04 g. Tiny
88TH. Munsell 5Y6/2 light olive grey-5Y6/3 pale olive. This is circular disc bead, pierced centrally. In shape and size, this is
part of the vertical end of what was probably a crescent-shaped typical of the beads found in huge numbers among grave goods
pot stand. It has one flat surface incised with parallel horizontal during Mellaart's excavations.
lines, and a rounded surface also incised with parallel horizontal
lines and surmounted at the lower, broken, edge by a projection Similar but rather larger beads were found in units 608 (CH94:62
resembling a nose. Flat, vertical multiply pierced projections pinkish white), 744 (CH94:63 beige) and unstrat. (CH94:64 creamy
down the side of the object, curving at top and bottom towards white). Details will be on the data base.
the side with the nose, look like ears, adding to its anthropomor-
phic aspect. 67) Stone bead. Inv. no. CH94:25. Unit 749, surface 2 x 2. Off-
white marble. 24Dia, 8.5TH. Munsell 10YR7 /2. Weight 1.36 g.
60) Fragment of pot stand with incised geometric decoration. Squashed sphere in shape, centrally pierced.
Inv. no. CH94:18, Fig. 12.7:2. West edge of west mound, surface.
Baked clay. 64H, 48W, 38TH. Munsell10YR7 /2light gray. Weight 68) Clay bead fragment. Inv. no. CH94:26. Unit 858, sub-surface/
101.18 g. This is the upper part of probably a free-standing pot scrape, square 580,1020. Beige/grey clay. 15.5Dia, 13H. Munsell
support. It is an oblong upright with a slight inward projection at 10YR5/1. Weight 2.93 g. Just over half of this bead survives. In
the top providing a flat surface for a vessel. It has inscribed shape it is a squashed sphere, with a central vertical perforation.
decoration on four surfaces - two parallel zigzags on the front/
inner face; chevrons on the sides - one side pointing up, the 69) Glass. Four fragments of bracelets of blue glass were found in
other down - and two irregular zigzags on the upper face. The units 859 and 860, square 640,960. Details will be made available
upright widens slightly towards the bottom, and may well have on the data base. They probably come from Classical period
widened considerably at the base. graves in the area.

Discussion 70) Mini clay balls have been found in at least two units- 785
A fragment almost identical to no. 59 was found by Mellaart and and 793. Details will appear in the data base.
is currently held at the Konya Museum (Inv. no. 79-85-65). It is so
similar in all respects that, although it has not yet been possible
to look at the two pieces together, they may well belong to the
The burial data from Mellaart's excavations
same artefact. Similar pot stands from the site, bearing the same
parallel lines on one surface, suggest that we have here the upper There are three sets of data to consider in examining
tip which sat at roughly a 40 degree angle on the top of an the burials - the skeletons, the grave goods, and the
upright, thus offering a sloping yet anti-slip surface for a pot
with an angled lower body. They are dated to the early
published information. These three groups offer dif-
Chalcolithic II period. A number of pot stands similar to no. 60 ferent challenges, and it must be said from the start
were found by Mellaart on the west mound (1965, 151, fig. 10). that all of them are extremely difficult to work with.
Many have holes cut in the lower part, to disperse the heat and In addition to discussing each data set in some de-
prevent fire damage, and it is quite possible that this fragment
was of that style. Typologically it appears to be closest to those
tail, I shall suggest a range of questions that could
dated to the Early Chalcolithic I period. potentially be asked of the data, and later consider
how far we can attempt to answer these questions
61) Animal figurine. Inv. no. CH94:19. Fig. 12.2:8. Unit 802, sur- with the information available.
face 2 x 2. Mid-brown baked clay. 16H, 14W, 26L. Munsell10YR5/
2 greyish brown. Weight 4.39 g. Standing quadruped, head miss- Throughout the preliminary reports, and in his
ing. In the absence of the head it cannot be confirmed that this is 1967 book about the site, Mellaart attempted to as-
a cow /bull, but it has a bovine appearance. sess and interpret the information he was present-
ing, rather than leave it in a raw form, even though
62) Fragment of figurine. Inv. no. CH94:20. Unit 648, surface 2 x
2. Baked clay. 19L, 7Dia. Weight 1.01 g. Arm or horn of figurine. he was working inevitably with partial data. A
number of conclusions were drawn regarding social
63) Fragment of spindle whorl. Inv. no. CH94:21. Unit 858, sub- organization, gender status and rank based on the
surface/scrape, square 580,1020. Colour-baked clay with sandy burials. These included burial in different areas and
inclusions. 34Dia, 20H. Munsell 10YR5/2 colour. Weight 8.65 g.
This is a quarter of a probably biconical spindle whorl. This with different grave goods according to age and sex,
fragment is half of one conical section, broken vertically through and a distinction between shrines and houses. The
the central piercing. The surface is undecorated. Not prehistoric. ideas are well known, and have recently been the
64) Fragment of celt. Inv. no. CH94:22. Unit 628, surface, 2 x 2.
focus of some debate (see Chapter 1). In the follow-
Dark red polished stone, possibly carnelian. 17L, 9.5W, 4.5TH. ing pages I will assess the data on which these
Weight 0.89 g. This is a small fragment of one side of a celt at the theories have been based, as well as looking more
blade end. Highly polished, the stone is very unusual. It is possi- widely at the economic information contained re-
ble that this is part of a pendant. No other celts or axes of this
stone have been found.
garding trade and technology, and examining
changes over time. Potentially, the unusual stratigraphy
65) Shell bead. Inv. no. CH94:23. Unit 760, surface 2 x 2. Whelk-type at C::atalhOyiik - with buildings placed one above

242
Figurines, Clay Balls, Small Finds and Burials

Figure 12.7. Decorated pot stands from c;atalhoyiik West. For descriptions see text.

243
Naomi Hamilton

the other - should offer a chance to look at vertical these were not included in the published data.
correspondences of burial habits, distribution and Ferembach apparently examined the same skel-
grave goods, and the quality of the data in this re- eton collection as Angel, although she worked only
spect will be examined. The grave goods, with all on the adult ones, but it is clear from her records that
their problems (see below) constitute one of the ma- there were some disparities. She has a greater number
jor artefact groups from the site and comprise a broad of skeletons from some buildings than Angel did,
range of types and materials which permit us to fewer from others, and the identifications do not
approach a variety of issues through the same data always match up. It has proved impossible to match
set. their records completely, so in addition to poor skel-
etal survival, we have a further reduced set of data
The data and their problems to work from. The importance of trying to match
The skeletons their data lies partly in the different sexing of a small
The skeletons are probably the best studied and pub- number of adults by the two specialists, and also in
lished data set from C::atalhoyi.ik. This does not mean trying to discover how many additional individuals
they are unproblematic. They were studied in the were identified altogether- it would appear that
1960s by both Lawrence Angel and Denise Ferembach, Ferembach divided again some of the material
each of whom published some results (Angel 1971; viewed as a single individual by Angel. Unfortu-
Ferembach 1972; 1982). Angel reported that the ma- nately, Ferembach's records rarely identify the skel-
terial was in poor condition and some disarray, with etons by the CH numbers used by Angel. As a result
labels missing, and that many skeletons never only 86 skeleton records can be matched up, of which
reached him- Mellaart mentions the excavation of 10 or so skeletons are sexed differently by each of
approximately 400 skeletons during the three sea- them. Ferembach has a maximum total of 282 adults,
sons 1961-63, and at least 80 more in 1965, yet Angel which added to the 89 juveniles found by both her
received only 297. Angel comments on the huge loss and Angel makes a minimum of 356 individuals.
that has obviously occurred, almost 40 per cent. This This is 78 more than the number Angel originally
means that any attempt to construct statistics is likely received, which - working on the same 40 per cent
to be inaccurate. From Angel's records it appears loss - suggests an original excavated assemblage
that the skeletons arrived as a group of 275 individu- approaching 600 (but see Fragmentation p. 257).
als, and that he sorted them further to arrive at the To add to the difficulties, both Angel and
figure of 297 (published as 294). Of those skeletons Ferembach seem to have had incomplete knowledge
which Angel did receive, 27 had no labels, while of the recording system used on site by Mellaart.
none arrived from building VIA/B:10, which was Thus some skeletons have bizarre building num-
supposed to have the largest group in the entire site, bers, whose original can only be guessed at. In addi-
consisting of 32 skeletons. Angel suggests these may tion, some mistakes have clearly been made at some
be the 'no label' group, although they could also point in the writing of labels or notes, so that skel-
have come from a mixture of other buildings said to etons are assigned to buildings which were never
have burials but to which no skeletons were assigned. excavated, or identified differently by the two spe-
In addition, a number of skeletons reported by cialists. Altogether the skeleton data from both spe-
Mellaart as noteworthy due to special treatment were cialists identifies burials deriving from 38 buildings,
not among those received. a small fraction of the 200+ buildings excavated by
Angel aged and sexed all the individuals he Mellaart.
identified, and published the results (Angel 1971).
He also gave parity figures based on examination of The grave goods
the pubic symphysis of female skeletons. All these A collection of some 300-450 grave goods is held in
data are now in question. The sexing of children and the museums in Ankara and Konya. The precise
childbirth estimates have been discredited, and even number is unknown- the figure given is an edu-
the aging of adults is now under scrutiny, as the cated guess. The reason for this is the poor level of
work at Spitalfields, London suggests that many have documentation. A total of 13 artefacts is marked in
been under- or over-aged (Molleson & Cox 1993, the Konya Museum inventory as coming from buri-
167-79). Angel was well aware that the sexing of als; a rather larger number is labelled on the item,
children was problematic, and in his notes indicate while information gleaned from the publications ar-
his level of certainty with a range of from one to rives at a total of 110 definite grave goods. In this
three question marks, with or without brackets, but situation, records were made of all inventoried

244
Figurines, Clay Balls, Small Finds and Burials

artefacts known to be almost certainly from burials, may consist simply of 'male' or 'female', and a fur-
whether or not contextual information was available ther 14 are identified by which platform they carne
(such as beads, buckles, wood, textile, spatulae etc.), from. A number of these close identifications belong
and all those with a building context which were to the same skeleton, reducing further the number
potentially from burials (bone tools, ground stone, of individuals to whom grave goods can be as-
knapped stone, etc.). As this selection was made signed.
only from inventoried items, it is quite possible that
other grave goods will be found among the bulk The published information
finds which have not yet been examined- princi- Mellaart published a considerable amount of infor-
pally bone tools and ground stone, substantial quan- mation about burial practices in his reports, and the
tities of which exist. The majority of knapped stone main themes identified by him are clear. Burial was
has been examined, and did contain a number of intramural, beneath the platforms with which all
contexted grave goods in bags of mixed materials in buildings except stores and entry shafts were provided;
which obsidian was dominant, and this situation it was generally, if not always, secondary inhuma-
may occur with the bone and stone. The reasons tion following excarnation to varying degrees; it was
behind the choice of items to be inventoried is not normally communal; and this communality was usu-
known - decisions may have been based on the ally diachronic. The first of these should make the
condition of the artefacts, or on the existence of good <::atalhoyiik burials an invaluable source of informa-
context, or their recovery from buildings, or burials, tion; the others diminish this promising start.
or a mixture of them all. However, we do know that With the knowledge that burials were second-
in general few artefacts were found on floors, al- ary, communal, and were added to over time, it is no
though exceptions are reported, and that the major- longer surprising that few grave goods have close
ity of finds probably come from burials and outside contexts. A number of photographs show tumbled
areas known as courts. Working on the assumption masses of skeletons mixed inextricably together, and
that burial goods have a better chance of survival the problems of assigning grave goods to individu-
than artefacts discarded in open areas, and may have als were insurmountable. Indeed, perhaps that is as
been deposited in a better condition than those dis- it should have been, for if the skeletons were mixed,
carded in open areas, added to the knowledge that why should we assume that artefacts found in the
beads were the most common grave goods and were grave were not the communal property of them all?
apparently all inventoried, it is likely that the major- However, some skeletons were buried separately, a
ity of grave goods have been recorded. A number of number of them with grave goods, and it is largely
items known to be from non-burial contexts were on the basis of these that a division of artefacts ac-
also recorded deliberately, in order to compare them cording to sex/ gender was offered by Mellaart. It
with grave goods. should be noted that detailed skeletal analyses were
Overall the number of grave goods can be in- not available to Mellaart at the time.
creased with considerable certainty to 300, by the Mellaart tells us (1963, 46-50) that no burials
addition of artefacts which are mentioned as always were found in three buildings in 1962 -All: 1, AIII:8
or generally being found in burials, or which are (directly below AII:l) both of which he regards as
extremely unlikely to have been recovered from other shrines, and AV:l. He does not give similar informa-
contexts - for instance, necklaces, and wooden tion in other reports, but does say that burials occur
bowls. However, a large number of artefacts remain in all houses and most shrines.
in the uncertain category- notably the bone points,
celts, knapped stone, and rnaceheads, all of which The artefacts
are known to occur in burials. As far as is known, The objects found in burials, according both to the
figurines and pottery were never found in burials. items themselves and to Mellaart's reports, are items
Given the level of documentation, it is not surprising of personal adornment or aggrandisernent (beads,
that of those 300 artefacts which can definitely or pendants, wristguards, buckles, toggles, pins,
almost certainly be treated as grave goods, many rnaceheads), obsidian and flint tools (projectile points,
have a poor context. 54 have no information at all, daggers, knives), bone tools (borers, needles,
and 70 have a level only, leaving 176 well-contexted spatulae, spoons, ointment sticks), ground-stone tools
artefacts. Of these, only a small proportion can be (adzes, celts, grinders), textile, vessels (wood, bone,
assigned a tighter context - 39 have information stone, basket), paint or pigment, 'stamp seals' and
about which skeleton they were found with, but this mirrors.

245
Naomi Hamilton

Beads than average numbers occurring with some individu-


By far the most common artefacts are beads, which als. The beads themselves are extremely interesting,
occur in large numbers - over 20,000 have been from the point of technology, materials, and design.
found, in groups of 3 to 2000. Beads were generally Almost all the materials used for beads are im-
found in the form of necklaces or bracelets on skel- ports, some from distant areas. Mellaart lists the
etons, although sometimes they were scattered over materials as follows, commenting that most come
a skeleton and may originally have been attached to from areas involving at least a few days travel (1964,
fabric. Mellaart reports that on occasion it was possi- 97): copper (nearest source near Bozkir), lead (ga-
ble to see the order in which the beads were threaded, lena is available at the Cilician Gates), white paste,
sometimes forming multi-strand necklaces of up to 8 chalk, red ochre, baked clay, lignite, slate, boar tusk,
strands; however, when this was not possible, re- bone, schist, mica, animal teeth, calcite, alabaster,
stringing has been left to the taste of archaeologists carnelian, obsidian, apatite, serpentine, limestone,
or museum staff, and it is therefore quite likely that dentalium, cockle and whelk shell, freshwater shells,
any division into necklaces or bracelets simply re- cowrie shell, and mother of pearl. A look at the
flects their present form. For instance, in some cases occurrence of different materials over time might
I have examined two or three threaded groups of give us an insight into travel or trade patterns,
beads which were inventoried under a single number, changes in economy or ecology, as well as social and
and I have no way of finding out on what basis these symbolic patterns. We are not yet in a position to say
decisions were taken. Beads given number CHC712 whether the population of <::atalhOyiik left the settle-
are strung in three groups, one of which has a build- ment in search of all their raw materials, either for
ing number attached- can it be assumed that the their own use or to trade, or whether they obtained
other two groups can also be assigned to that build- them from outsiders, but one thing is clear - com-
ing?; one of the necklaces reported to have been monly available materials such as clay (used in
found with one of the atypical burials in EVIII:31, architecture, figurines, pottery etc.), or bone (a hi-
numbered CHC672, is strung in two groups, each product of food, used for a range of tools) were not
consisting of the same three colours and the same widely used for beads, although they do occur. This
type of beads, while another necklace from this build- information could lead into an investigation of rank
ing, numbered CHC669, is currently threaded in two and status. Before looking at the beads in detail, it
groups, one with no context and the other thought to should be made clear that Mellaart does not say who
belong to EVIII:25. If this is the situation with the identified the materials for him and how certain these
best-published grave goods from the site, described identifications are. In addition, materials are rarely
in detail in the report (Mellaart 1966, 182), any analy- mentioned in detail on inventories, and there has
sis of the occurrence of bead ornaments becomes been no recent specialist identification.
much more difficult than expected. Altogether I have The majority of beads are made of various types
examined 119 groups of beads, recorded on the data of limestone, but dentalium shell and deer teeth were
base, but this may not be an accurate representation also popular. Table 12.7 shows the quantity of each
of how they were found as 15 of them share only 7 material over time, and Table 12.8 shows the number
CHC numbers. Of the total 119, 32 have no informa- of bead groups containing each material. From these
tion physically attached to them concerning context, some changes in material can be seen- for instance,
43 have information about level, and 44 have clear dentalium shells seem to have their heyday in level
building contexts - some of which may be inaccu- VI with 10 examples, 4 in VII, 3 in VIII and 2 in IX,
rate, as shown above. Although a number could be but the majority of beads actually come from a sin-
assigned to buildings with some certainly via shared gle level VII necklace (c. 780 beads), and level IV has
excavation/inventory numbers, and some with no marginally more beads than level VI, though on fewer
context could be assigned to levels in this way, there necklaces. Later levels have 1 from V, 4 from IV and
is clearly a definite risk that mistakes will result 1 from III. As the shells come from the Mediterra-
from such an approach. nean, this suggests long-distance travel or contacts
Despite the poor contexts, we can work with around the time of levels VI and VII, which also
the beads in some ways. In a few instances we can appears to be the most flourishing period of the
assign them to a particular skeleton, or associate settlement. If we look at other materials from the
them with other grave goods. We can also calculate Mediterranean, we see that whelk also occurs mainly
the approximate occurrence of bead ornaments as in level VI, with 3 definite occurrences, one which is
grave goods, and assess the implications of other either VI or V, and only one other stratified example,

246
Figurines, Clay Balls, Small Finds and Burials

in level IV; cowrie occurs once in level VII, and the necklaces from level VIII, often in multiple col-
cardium in level IV. The majority of these shells are ours chosen generally from white, terracotta red,
not found on the shoreline, but in shallow water, so and black, although beige, dull yellow, orange and
are unlikely to have been collected without premedi-
tation. How much shell was required has not yet Table 12.7. Beads: number of beads made of each material, by level.
been assessed. Lead or lead ore beads are most com-
mon in level VI, although the earliest known are in Material II III IV v VI VII VIII IX Unstrat. Total

level IX and the latest level II- galena is common in Stone 196 166 145 137 5719 2410 5671 1254+ 2763 18,461
Agate? - - 12 15 49 - - - I 77
the mountains between the Konya plain and the Apatite? - - 2 2 I - - - 4 9
Calcite? - - - - - - - - 5 5
Mediterranean. Mother-of-pearl pendants, which are Jasper? - - - 5 14 - - - 3 23
reported from burials of babies, occur in small num- Marble - - 77 - - - 425 5 2 509
Pebble - - - - I - - - - 1
bers from levels VIII to V, but this may derive from Rock crystal - - - - I - - - 3 4
Serpentine - - - - 8 - - - - 8
fresh water mussels rather than oyster. The concen- Turquoise 17 - 60 21 6 - 6 - 26 136
tration in level VI of materials from the Mediterra- Shell I - 3 - 3 13 I - 6 26
- 167 781+ 73 6+ 1316 2594
nean area or en route to the coast does suggest regular Den taliurn
Whelk -
51
-
181
3
70
21 22 - - - - 46
contact at that time, but the numbers involved are Bone - - I I 91 - I 47 - 140
too small to say more than that. Given this contact, it Deer tooth I I 18 I 17 25 50 6 20 139
may be less surprising to discover than obsidian Copper - - I - - I 6 I 5 14
from the same region as that used at <;atalhoyiik Lead - I - - 36 - 8 ? 3 48+

was reaching Cyprus in small quantities at some Clay - - - - - - - ? 17 17+


Plaster - - - I - - - - 6 7
point in the early seventh millennium. Although the Red ochre - - - - 1 - - - I 2

data could be skewed by the greater area excavated Coal - - 4 - - - - - - 4


Mica - - - - - - - - 6 6
in level VI compared to the earliest and latest levels, Obsidian - - - 10 I - - - - 11
this is not apparent when considering deer teeth, Shale - - - - 3 - - - - 3

which are more common in levels VIII and VII than


Note: This table does not contain details of every bead~ as a number were too broken to
in later periods, with nine of the 18 occurrences dat- count, or were inaccessible, but includes probably 99 per cent of those I saw. Some were
counted from photographs or through glass. The addition of + to a number means that
ing to these two levels. The earliest occurrence is in an unspecified number of additional beads of this material are known. The term 'stone'
level IX and the latest in level II. This temporal dis- used for the first group should normally be regarded as implying limestone, but this
may not be correct in every case. As stated in the text, some other stone identifications
tribution could be tied in with changes in habitat, or are uncertain.
changes in hunting practices generally, or a reduc-
tion in meaning assigned to deer teeth as beads. The Table 12.8. Beads: number of bead groups containing each material, by
source of the green and blue apatite used particu- level.
larly for beads of unusual design is not known. Some Material II III IV v VI VII VIII IX Unstrat. Total
of the beads may actually be of turquoise, which has Stone 2 2 5 2 27 8 16 5 20 87
sources in Sinai and Iran. Most of these beads occur Agate? - - 2 I? 2 - - - I 6
Apatite? - - 2 I 1 - - - 1 5
in the middle levels of the settlement, clustering in Calcite? - - - - - - - - I 1
jasper? - - I I - - - - I 3
VI, V and particularly IV, although three beads were Marble - - I - - - 2 I I 5
Pebble
found on one necklace of level VIII. A few beads Rock crystal - - - - - I - - - - 1
- - - I - - - I 2
may be made of copper ore, which was used to make Serpentine - - - - I - - - - 1
Turquoise 1 - 5 2 2 - 1 - 4 15
pigment. Copper is known from levels IX, VIII, VII
Shell I - 3 - 1 I 1 - 4 11
and IV, formed into beads or coverings for what Dentalium - I 4 I 9 4 3 2 14 37
Mellaart believed were weights on string skirts. Their Whelk - - I I 3 - - - - 5

poor state makes examination difficult. Bone - - I I 3 - I - 7 13


Deer tooth I I 2 I 2 4 5 1 4 21
Most beads are made of polished limestone in Copper - - I - - I I I 2 6
various colours. Technically they are of very high Lead - I - - 4 - I I I 8

quality. Around half the total are tiny flat discs with Clay - - - - - - - I 3 4
Plaster - - - - 1 - - - 1 2
polished outer edges sliced from cylindrical preforms Red ochre - - - - I - - - I 2
which often bulge slightly in the middle. In diameter Coal - - I - - - - - - 1
they are commonly as small as 2.5 mm, sometimes Obsidian Mica - - - - - - - - I 1
- - - I I - - - - 2
as large as 5 mm, and range in thickness from 3 mm Shale - - - - I - - - - 1

to literally paper thin, with sizes of less than 1 mm far


Note: 'Stone' should normally be regarded as implying limestone, but this may not be
from unusual. These beads (type 1 in my typology) correct in every case. As stated in the text, some other stone identifications are uncertain.
were used in large numbers- up to 1657 on one of

247
Naomi Hamilton

turquoise also occur. When viewed through a glass Mother-of pearl pendants are reported to come
case in a museum the small size of many of the largely from burials of babies (1964, 95). One is men-
beads is not readily understood, but Mellaart states tioned on a necklace in the unusual adult burial in
that modern steel needles cannot pass through the EVIII:31 (1966, 182), but the inventory records this as
piercing (1967, 211). Larger versions of the same corning from the child's grave in the same building.
type also occur (type 21) but are not common, and The pendants are roughly triangular and almost al-
often appear as loners in the midst of a string of ways double pierced in one corner, giving them the
another type. Perhaps originally they were not on look of a modern child's version of a ghost! Other
necklaces, but tied onto clothes or used in some other pendants have also been found, in a variety of shapes
way. Stone beads are most common in the early and materials- an animal head in dark green-black
levels, reducing drastically after level VI. Although stone with inlaid green eyes (369), a miniature celt in
the total number in level VI is similar to that in level red-brown stone (93), an oval pendant of coal or
VIII, and they occur on twice as many necklaces, in wood (118), a small perforated bone spoon (318)
comparison to the number of burials excavated in which is probably a pendant, and a number of oth-
level VI they are far less popular. ers in clay, rock crystal, shell and stone.
While type 1 and 21 beads comprise the major- No work has yet been done on how these beads
ity, a wide range of other forms occurs in small were produced, but clearly considerable technical
numbers. Probably the next most common is a skill was required - these are not odd pebbles or
straight-sided cylinder pierced longitudinally, with bones with fortuitous holes in them being put on a
a diameter of 3-5 rnrn and some 5-10 rnrn in length. piece of string, but highly standardized and attrac-
They occur in a variety of exotic materials, often tive artefacts. Although beads are the most common
highly polished and beautifully coloured marbled form of grave goods, and type 1 beads the most
stones which are shown off more effectively this frequent, they should not be regarded therefore as
way than if sliced up, as well as rock crystal, marble, 'cheap', 'simple' or 'low status' items. Mellaart com-
lead, copper, plus some dentaliun. The colours are ments that no tools have yet been identified as bead-
somewhat unusual compared to those found among making equipment (1964, 105).
type 1s, with blue-green and greys, orange-red and
copper, as well as black and white. Ground-stone tools
The turquoise stones attracted noticeably more Mellaart reports that greenstone axes and celts were
interesting design than most. Although a few disc- found in women's burials (1964, 95). Only two have
shaped beads are found, the majority are in fancy been identified as having burial contexts, as has one
styles such as lozenges, 'double axes', and the fa- pounder. Of the remainder, at least four carne from
mous birds or stylized goddesses published by storerooms, and one from a courtyard. They tend to
Mellaart (1963, pl. XXVII). Turquoise beads usually be small, and some could be classed as miniature,
occur in clusters, sometimes without any other col- while large ones are rare. Almost all are made of a
ours, and rarely with the standard type 1 stone beads. dark green stone, although some are a light grey-
They also display a range of unusual piercings, such green, and the size of stone nodule available seems
as double piercings invisible from one side. Some of to dictate the size of the product, as many bear traces
these may have been used as spacer beads or pen- of cortex. The two recorded as corning from burials
dants. Certainly they stand out from the crowd. are both miniatures, the slightly larger of the two
Dentaliurn shell is nearly always sliced into being 53.8 rnrn long, although much smaller ones are
discs, falling largely into the type 1 and 21 varieties. known. A number of others must have come from
This is quite different from Cyprus, where sites of burials, but information is lacking. No patterns can
the same date have produced necklaces of multiple be seen between those found in burials, store and
strands of whole dentaliurn shells - a well-known courtyards, but the numbers involved are very small
example comes from the acerarnic Neolithic site of and a larger sample might suggest differences - as
Khirokitia (Dikaios 1953, 303-4, pl. XCIX) - and could use-wear analysis. Maceheads are reported
from earlier Natufian sites in the Levant. Complete from male burials (1964, 94), but only four can now
dentaliurn shells are rare at ~atalhoyiik, perhaps be associated with buildings. Mellaart also mentions
because necklaces of type 1 stone beads had become them as offerings in shrines. I have seen 29 alto-
an established style before dentaliurn is first attested gether, a number of which- all from level VI -
in level IX, where they occur in a fully-developed appear to be unfinished, having clear shaping marks
type indistinguishable from later type 1 beads. which have not yet been polished away. It is

248
Figurines, Clay Balls, Small Finds and Burials

possible that rather than being offerings, they were does not appear to have been finding its way into
being made in a building. Unfortunately context is graves. Perhaps the greater deposition of obsidian
lacking, so this idea cannot be explored at present. and flint in burials in level VI is a result of changes in
Maceheads cluster in level VI although two are society, of which other evidence is available- there
known from level V and one from level II. Only one is a change in knapped-stone tool types, in pottery
is known from earlier, from a burial in EVIII:31, of technology, in figurine typology and material. Small
which much more will be said later. Two of those numbers of blades and scrapers are known from
from level VI were found in one burial, said to be burials across the levels, and according to Mellaart' s
male, in building EVI:20. Mellaart also reported find- sexing of the skeletons, they occur with both sexes,
ing two maceheads in the storeroom of AVI:l, one whereas he says projectile points and elaborate dag-
with a bone handle and one with a stick (1963, 52). gers normally occur with males (1964, 94). Unfortu-
The majority of maceheads have very narrow perfo- nately, with so few artefacts having strong contexts,
rations, as small as 7.6 mm, and rarely greater than this is yet another matter which is difficult to evalu-
13.5 mm. Few sticks or bones slim enough to fit ate with the old data, but exceptions will be dis-
through these perforations would be strong enough cussed below.
to support their weight, and I suspect many of them
were slung on ropes. This would affect how they Bone tools
were viewed, for a stone dangling on a string offers A number of modified animal long bones have been
rather different visions of authority from one held found, mainly worked into points plus a few chisels
aloft! Some experiments may be in order. Obsidian (cf. Chapter 11). None of these have a clear burial
mirrors are also known as grave goods- Mellaart context. However, a range of needles, pins and
reports ten, all apparently buried with female skel- spatulae have also been found which do seem to
etons and all in shrines, although he later says that come largely, if not entirely, from burials. Several
two occur in houses (1967, 79). Of the ten I have were illustratrated by Mellaart (1964, fig. 42/ 43).
seen, two of which are broken, three have no con- These were interpreted by him variously as bodkins
text, one is from level IV, three from level V and for basketry, clothing fasteners and ointment sticks.
three from level VI; only eight are shown on Of the seven needles and bodkins seen, three are
Mellaart's chart (1967, 81), four date to level IV, two recorded from burials; of the thirteen spatulate ob-
to level V, and two to level VI. The technology for jects, five have burial contexts; and of three miscella-
polishing mirrors is still not understood. They ap- neous pins, forks etc., one has a context. Spoons and
pear to be a rather late addition to material culture, spatulae are reported from burials of women with
first occurring in level VI, but this may prove in babies (1964, 103) - of the six seen, two have con-
future to be incorrect. texts. Another item called a toggle by Mellaart was
interpreted as a fastener for a cloak (1964, 100). Eleven
Knapped-stone tools are known, of which six have burial contexts. He
Mellaart mentions obsidian and flint daggers, knives reports that they were found with male skeletons,
and projectile points occurring in graves, mainly in normally behind the shoulder. These come in a range
association with male skeletons but sometimes with of forms and sizes, generally made of antler but
female ones. There are few with good contexts, but sometimes of bone, and could also be seen as weav-
of those the earliest occurrence is in level VIII, a ing shuttles. At present, the purpose of the ones at
single flint blade; two obsidian projectile points in <;atalhoyiik cannot be ascertained. Belt hooks and
level VII; 20 in level VI- a mixture of blades, scrap- buckle plates, and bone wristguards, are also regarded
ers and projectile points (eight of which were found by Mellaart as male items. They are rare - three
in a group by the leg of a skeleton), a mixture also of wristguards, and 14 parts of buckles were found-
obsidian, flint and chert; one obsidian fragment in eleven of them with burial contexts, including two
level V; four obsidian tools in one building in level sets of both hook and plate. In the future use wear
IV; and two flint items in level III. Many others may analysis could potentially answer some questions
have been found in burials but no record is avail- about what some of these items really were, and
able. The huge preponderance in level VI is prob- how they were used.
ably more than a result of the greater number of
burials excavated, although it is affected by the cache Wood, basketry and textile
of projectile points in a single burial. Obsidian is Several burnt buildings in level VI had burials con-
found in large quantities in the lower levels, but taining carbonized wooden vessels and quantities of

249
Naomi Hamilton

textile - in which the skeletons had often been and by ourselves. It differs from some types of social
wrapped. Remains of at least 25 wooden vessels have status in having a biological base, and is justified on
been examined, with contexts in AVI:1, EVI:1, EVI:5, those grounds. However, the development of gen-
EVI:10, and EVI:25. This is probably not the com- der as a social construct is dependent first on the
plete record - Mellaart reports that a minimum of development of a concept of sex not just as a differ-
20 wooden vessels was found in EVI:10 alone (1964, ence, but as the difference which matters.
86). They include bowls, cups and boxes with lids in In problematizing sex and gender, however,
various shapes. Baskets are also reported from buri- we remove many of the hooks we use to hang our
als, often with the skeleton placed inside. However, data on. If we abandon the division, do the grave
such items have not survived. Textile was recovered goods have a basis for patterning? Is much of the
mainly from burials in buildings EVI:5 and EVI:25, meaning of grave goods dependent on the sex-basis
which according to Mellaart provided over 100 frag- that we have given? If we abandon the concept, do
ments (1963, 101). I have seen some 100 pieces of we also throw out the tool for overturning it? -
textile, many very small and badly damaged, with surely contravening the regulations is the surest way
contexts not only from these buildings but also from to prove them wrong, but is dependent on their
EVI:1, EVI:3 and EVI:8. Doubt remains about which prior existence. Because of the difficulties expressed
raw material was used (Burnham 1965; Ryder 1965), here, I shall initially treat the material as though I
but the fineness of some of the weaving is very clear. accept the division, and work through other issues
A number of fragments are attached to or wrapped rather than attempting to address sex and gender
around pieces of human bone, frequently mandi- separately. I shall then assess the data with reference
bles, and many have skull contexts on their labels. to current theories of sex and gender at <;atalhoyiik,
Mellaart reported finding textile in the brain cavity to investigate the validity of both the theories and
of one skull (1964, 93), and the wrapping of indi- the concepts themselves.
vidual limbs found in the unusual burnt burial of
EVI:5 (see below) in textile after initial burning (1963, The skeletons
99-101). Ferembach found 135 female and 106 male adults, as
well as 11 uncertain, 7 with female and uncertain
Sex and gender elements and 8 with male and uncertain elements,
The information provided by Mellaart focused largely totalling 267-82 adults. Angel found 131 female and
on a division of people into two sexes which were 78 male adults, plus 2 female?, 3female??, 8 male?
treated differently in a range of ways - in place of and 1 male??, making a total of 223 adults. Either
burial, treatment of the body, and grave goods. This way it is clear that female skeletons outnumber male
aspect of his work has been a focus of considerable ones among adults, although the proportions are
debate regarding the roles of men and women, par- considerably more equal in Ferembach' s data than
ticularly with respect to social power and hierarchy Angel's. The question I now want to ask is, did the
(see Chapters 1 & 19). However, before the answers people of <;atalhoyiik recognize this sex division
can be forthcoming some serious questions have to themselves; if so, did they codify it in some way
be asked. recognizable in the archaeological record; and if they
did, can it tell us anything about the state of gender
The problem at <;atalhoyiik?
Sex is regarded by archaeologists as a natural, given In his first report Mellaart (1962, 51-2) said that
attribute of humans, and the division of humans into people were generally buried under the northeast
two sexes, usually with associated sex-based gender platform of the building, that grave goods were rare
roles, has rarely been questioned. While I am not and that the use of red ochre on skeletons was not
disputing that in various biological formats sex does attested. In subsequent reports this situation was
exist, what is clear is that even in biology there is a changed. Mellaart began to say that males were bur-
continuum, which makes the sexing of skeletons un- ied under the northeast platform and females and
certain, and that outside biology the concept of sex children under the main (that is, east centre) plat-
may not have existed in the past in the way it does form, that grave goods- while not present in all
now. A person whose skeleton we view as female graves or in large numbers -were not rare, and that
may not have had a concept of femaleness, or placed red ochre was applied to a small number of female
itself within that category. Gender is a social status skeletons. In addition, people were sometimes bur-
based on perceived sex - perceived, that is, by others ied under other platforms, children were sometimes

250
Figurines, Clay Balls, Small Finds and Burials

buried under the floor, blue and green pigment was to Angel, it is female; it cannot be identified with any
applied to some skeletons regardless of sex (and one of Ferembach's skeletons. Those from the southwest
case of grey pigment is attested, sex not mentioned). platform are also only identified by Angel, and con-
In addition a range of complex information was sist of one female, one possible male and three juve-
given- that burials in rooms called shrines tended niles - a child aged 10 (long bones only), a child
to be 'richer' than those in houses, that red ochre aged 5, and a baby of around 15 months. The skel-
burials occurred in shrines, that blue and green buri- eton from the storeroom is male.
als tended to be 'richer' than others, that red ochre Ferembach's data seem to give positional infor-
burials were sometimes fairly poor, that the number mation on ten skeletons, regarded by Angel as six
of burials per building depended on the size of the individuals. These skeletons have the letters N or
building, that it depended on whether or not it was a NE added to their building number, and those which
shrine, that burials below level VI tended to be re- have been subdivided also have their CH number.
spected while those in level VI and above were Those coming from the north platform (N) are four
pushed aside into bone piles, that women were bur- females according to Angel, or seven females and
ied with necklaces and men never had more than a one male according to Ferembach. All but one of
few beads, that men were buried with maceheads them are identified by Angel as Group Y, the odd
and daggers, that women had make-up and oint- one out being Group AI. The Group AI skeleton
ment sticks and tools for feeding babies with, that had been divided by Angel into an adult female and
males were never buried with children and tended an adolescent which he sexed as male, and Ferembach
to be single burials. Finally, details of a few notable then subdivided it into two females and one male.
burials showed that there was considerable varia- Five other skeletons from this building are labelled
tion from any clear norm. Group Y, and it would seem reasonable to suppose
The skeletons received for analysis derive from that they came from the same platform, hence the
all four seasons of excavation, although the majority grouping. This extra group consists of two males,
date to 1963. Area E is the main area represented, two females and two juveniles- a child aged around
and level VI dominates heavily- of those skeletons seven and a newborn baby.
with levels recorded 235 come from level VI, the The skeletons identified by Ferembach as NE -
remainder being shared rather unequally as follows: presumably northeast platform - are viewed by
level 1:0, 11:8, III:lO+, IV:20; V:21+, VII:57+, VIII:ll, Angel as two males, and by Ferembach as four males.
IX:l, X:O. XI:4, XII:O. Hence any analysis must take They belong to Group A, to which a further two
into account that the data are strongly skewed, and skeletons belong, one a juvenile aged about five, one
that statistics are likely to present the situation ac- a baby aged 12 months+. Two more are labelled
cording to level VI rather than the site as a whole. Group A(?)- a baby of about six months and a mid-
Few skeletons have recorded information regarding adolescent separated by Angel from one of the adults.
their position within the building, but 20 do appear One further skeleton is labelled Group C, and if
to have these details. the grouping has a straightforward meaning we
ought to be able to assign it with the other Group C
Burial positions within buildings skeleton to the southwest corner. This one is female
Building EVII:31 contained 46 individuals according and was treated with red ochre, which makes its
to Angel, and 43 adults according to Ferembach. position within the building of particular interest.
Angel's data break down as 16 female, 14 male, 2 One other skeleton from EVI1:31 is mentioned by
possible male and 14 sub-adults (ranging in age from Angel as being treated with red ochre, this time a
adolescent to baby) giving only 32 adults against male aged around 29. However, this one is problem-
Ferembach's 43, which break down as 15 female and atic: Angel labels it 75CH, and also identifies it as
26 male skeletons with 2 indeterminate. number 20, a number it shares with a female (76CH)
Angel's data give positional information for 7 and a baby (77CH). Ferembach identifies two male
skeletons, one of which also bears a group number burials as 75CH, one in EVII:31 with no identifier,
which may enable us to place a further skeleton in and one in EVIII:31. This latter is identified by Angel
that area. The positions given are southwest corner as a male aged 21+, with number 51CH.
(one skeleton, Group C), southwest platform (five Three other groups of skeletons are mentioned
skeletons), and storeroom (one skeleton). The skel- in this building- Group B (two females, separated
eton in the southwest corner may have come from by Angel from one original), Group W (one female),
below the platform but this is not stated. According and Group Z, (one male). It seems likely that more

251
Naomi Hamilton

than one individual was identified by the excavators pared to some other parts of the buildings, and when
in each group, in order for it to be called a group, but it is decorated, it is nearly always with paintings or
on available data no others can be added at present. reliefs running along the north or east centre walls
According to the plans, building EVII:31 had as well. Mellaart says quite clearly (1964, 93) that the
platforms in the north, northeast, east-centre, south- main platform is the east-central one, based on the
west and probably southeast. The northeast plat- level of elaboration of reliefs and paintings, and it is
form has a crawl-hole adjacent to it, leading to the under this platform that he places the female and
storeroom along the north end. Symbolic assessments child burials. Unfortunately I have no documenta-
of the buildings at <;atalhOyiik, based on Mellaart's tion concerning the burials under this platform in
publications, have suggested that the north end is building EVII:31, but obvious candidates are Group
male (Hodder 1990, 10), and the south end female. B (two female, one male), Group W (one female),
According to the available skeletal evidence, the situ- and Group Z (one male). A further seven males, one
ation is rather more complicated. Although males possible male, five females and four juveniles (An-
may occupy the northeast platform, not only do they gel), or ten females, 21 males and two indeterminate
share it with a number of juveniles (sex unknown), (Ferembach) need to be placed. The recurrent pres-
but the large north platform adjacent is heavily ap- ence of bull figures on the north wall, noted by
propriated by adult females. The southwest area also Mellaart (1967, 104), has been used by Hodder to
has a high density of females, with one probable support the association of the north end of the build-
male, and one male is in the storeroom - an excep- ing with males - through the belief that the bull
tional place for any burial. Hodder regards the north (which is, in fact, generally shown sexless) is a male
end as symbolically more elaborated, and as an in- symbol. The presence of predominantly female skel-
ner area (Hodder 1990, 9-10). Both of these concepts etons under the north platform challenges this inter-
must be questioned. Firstly, an aspect of architecture pretation. As it happens, EVII:31 is one of the
which seems to have gone unnoticed by commenta- buildings represented on Mellaart' s chart of wall
tors, though mentioned by Mellaart (1962, 46; 1963, paintings and reliefs. It is the west wall which is
56), is that many buildings have an entry shaft in most heavily elaborated, with cattle heads and a
addition to the ladder at the southern end. This shaft relief 'goddess', the north wall is blank and contains
can be in any area outside the main room, but in two crawl-holes (one to the storeroom, one to the
practice is frequently next to the storeroom, from entry shaft), the east wall was destroyed apart from
which it is divided by a wall. Entry to shafts and the base of a red panel above the central platform
storerooms from the main room is effected through and a relief 'goddess' and pair of breasts at the south
a crawl-hole. Mellaart comments that in level VII end, and the western half of the south wall most
and earlier the shaft tends to take the form of a unusually has another relief 'goddess'. In this par-
corridor along one end of the building (1964, 50), but ticular building then, rather exceptionally, the south-
in level VI and later the use of a dividing wall creates west corner is the most elaborated and the north
two separate spaces, a small squarish shaft and larger wall and northeast corner the least, with unanswer-
rectangular store. The shaft is frequently at the north able questions over the east wall. If the position of
end, and may have been intended to offer an alterna- the skeletons is plotted onto this 'symbolic' map, the
tive method of entry when a smokey hearth makes elaborate southwest corner contains mainly female
the southern entrance uncomfortable to use. Al- and juvenile skeletons, including one treated with
though the north is often used for the shaft (for red ochre; the plain north central area is also domi-
instance VI:1 and VII4- northeast; VI:28 and 31 -north- nated by females; the plain northeast corner has males
west, and VI:7 and 14 sharing a northern shaft) - and probably juveniles. This clearly contests both
even sometimes taking the space used for north-east the division of the building into sex-segregated zones,
platforms! (e.g. VI:44, VIB:l) , the south seems to be and the association of males with the more elaborate
just as good (for instance VII:10, VI:8, AII:1 - south- areas.
west; VII:22- southeast). The structural and symbolic A small amount of information regarding con-
inner nature of the north end of the building can no text is available from the skeletal data concerning
longer be maintained. Turning to the elaboration of three other buildings: EVI:8, in which Skeleton A,
the north end in general and northeast corner in sexed female by Angel, came from the ante room
particular, a glance at Mellaart's chart (1967, 102-3)- (which was to the south of the main room on the
which does not include all buildings- shows that eastern side). According to Mellaart's chart, in both
the northeast corner is not heavily decorated com- VIA and VIB (it is not known from which level the

252
Figurines, Clay Balls, Small Finds and Burials

skeleton derives) the whole of the west, north and female, age illegible but possibly 'youth'. Ferembach
east walls were decorated including the southern did not record these skeletons, although they may
end of both east and west walls, and the south wall appear without their building number. There is no
was blank. EVI:1 had a male lower jaw from the trace in the skeletal record of the children or the red
entry shaft in the northeast corner - possibly ac- ochre skull.
companied by three others bracketted by Angel with There could be a good reason for the disparity
the male jaw for unclear reasons - all female in between the burial record in EVI:5 described by
Angel's view, one female and one uncertain in Mellaart and the general rules he sets out - the
Ferembach' s notes. According to Mellaart' s chart, bodies had not been excarnated, but had been burnt
VIB:1 (VIA:1 is not shown) has only simple decora- prior to burial and wrapped in textile after this burn-
tion in the form of red panels which occur on the ing and before the fire which destroyed the building.
eastern half of the north wall, the central and south- This could be seen as a 'foreign' burial rite. On the
ern sections of the east wall, and the eastern end of other hand, the grave goods are no different to any-
the south wall; the west wall was destroyed. No thing else on the site, so these are unlikely to be
decoration occurs on the northern end of the east foreigners. Mellaart stated (1963, 98) that in no sin-
wall, which contains a crawlhole. Finally EVII:45 has gle case was there evidence of a hole made hastily in
7 skeletons listed, of which two are assigned to the a platform for burials, and proposed instead that
west platform (one female, one male according to burials took place at a set time, perhaps coinciding
Angel, one female and one uncertain according to with a festival and probably with the annual re-
Ferembach) and one to the central platform (a baby plastering of the buildings. Some evidence for this
of about 6 months). According to Mellaart's chart, lies in the different degree of excarnation shown
the central section of the north and east walls and among the skeletons, some being disarticulated whilst
the southern end of the east wall were decorated, others seem to have had ligaments intact. This could
while the west wall was destroyed. The walls sur- be the reason for an unusual burial, if a group of
rounding the northeast corner were blank. It should people died just at burying time. By burning them,
also be noted that published reports say that burials the worst problems of putrefaction could be over-
did not occur in anteroom, storerooms or shafts (1964, come, and by putting them all in one building,
92). Angel's notes show that they occurred in all whether or not they all lived there, any problems
these, although they should probably be considered would be restricted to a small area which could be
exceptional. closed if necessary. Why they should choose to put a
large number of bodies under the small 'male' plat-
The northeast platform form is more difficult - unless there was a strong
Mellaart reported a burnt burial from building EVI:5 taboo concerning burial areas for men, restricting
which contained the remains of 6-8 individuals, 2 or them to one part of the building, which did not
3 of whom were children under the age of ten (1963, apply so strongly to women or children. As seen
99-101). The building is recorded as a 'less distin- from EVII:31, males do seem to be buried under the
guished house' which nevertheless had rich grave northeast platform, but not exclusively there, and
goods with burials (1964, 94). The burial reported they do share it with juveniles.
was made under the northeast platform. In the midst In 1961 Mellaart reported that people, not males,
of the most 'characteristic' level of the site, with were regularly buried under the northeast platform.
fairly standardized architecture, elaborate symbol- This suggests that in the first season of excavation
ism etc., is a burial of children under the 'male' few burials were found under other platforms.
platform. Artefacts found in the burial included a Mellaart never published a list of which buildings
belt buckle, seen as a male item, which, if true, sug- were excavated each year, but a good idea can be
gests the breaking of another rule - the burial of obtained from the plans, reports, and inventory lists.
children with a man. One of the skulls was treated As far as can be ascertained buildings BI:4, 5, 6, 7;
with red ochre, which contravenes several other BII:1, 2, 3; AIII:l, 2, 3, 4, 5, 6, 7; EIV:1, 2, 3, 4, 5, 6, 7, 8,
'rules'- that red ochre was applied to females only, 9, 10, 11, 12; EV:1, 2, 3, 4, 5, 6, 7; EVI:1, 2, 3, 4, 5, 6, 7
or else that females are not buried under this plat- and EVII:1 and what later became 28 and 29, were
form; that red ochre burials occur in shrines only; excavated in 1961. Of these, no platforms are shown
and that red ochre burials tend to be poor in terms of on the plans of level I; BII:3 has none, BII:2 has two,
grave goods. Two skeletons from EVI:5 were seen by northeast and east centre, BII:1 has three- northeast,
Angel, who recorded them as a male aged 40 and a east-centre, and northwest; in level III platforms do

253
Naomi Hamilton

not abound- AIII:3, 5, 6 and 7 have none, AIII:2 has single-sex burials: BII:1 (two females), BII:? -
one in the northeast, AIII:4 has an east central one whether more of BII:1, or BII:3 - which definitely
with burials under and another in the south-west, had burials (five females and one possible female);
AIII:1 has three, northeast, centre-east and west/ FV:75 (two males [Angel], or five males [Ferembach,
centre north; in level IV buildings 3, 5 and 9 have no who mis-labels them EV:75]), EVIII:25 (one female,
platforms, EIV:2 has one in the southeast, EIV:6 has one juvenile), and EIX:1 (one female). Furthermore,
one in the northeast, EIV:10 has one east centre an- despite the preponderance of females overall, in sev-
notated as having numerous burials underneath, eral buildings males are in a majority: AVI:1 (four
EIV:12 has one east centre/north, EIV:7 and 8 have males, three females), EVI:20 (four males, three fe-
two, in the northeast and east centre, EIV:4 and 11 males [Angel], eight males, three females, one inde-
have three, northeast, east centre and southwest, and terminate [Ferembach]), EVI:29 (three males, five
EIV:1 has five or more, leaving only the south-centre females [Angel], four males, three females
and part of the north centre definitely platform free. [Ferembach]), EVI:34 (five males, six females [An-
In level V buildings seem rather more standardized, gel], 11 males, seven females, three indeterminate
with all but storerooms having platforms, always a [Ferembach]), EVII:31 (14 males, 16 females, one pos-
northeast platform and the others generally ranged sible male [Angel], 26 males, 12 females [Ferembach]),
on either side of it, with only one southwest plat- EVIII:31 one male [Angel], four males, one female
form occurring, in building 2, and building 4 being [Ferembach]) and an uncontexted group called G by
unusual in having only two platforms, in the north/ Ferembach (six males, five females). Even bearing in
centre-east and north/ centre-west. Levels VI and VII mind that the great majority of skeletons examined
show more variation again, with a single north/ are from level VI, it is still interesting that all the
centre east platform in level VI:1 (burials below) and groups with known context with high numbers of
2, burials below a possibly centre east platform in males come from levels VI to VIII, mainly level VI,
building EVI:3, two platforms in EVI:4 at north cen- and the later levels seem to have greater numbers of
tre and southwest, and two also in EVI:7 in the north- females. Why this should be is not clear - it could
east and east centre, both with burials; and in level relate to economic changes, trade patterns, health
VII platforms against most wall space in EVII:1, a and disease, emigration, falling birth rates, changes
northeast one in EVII:29 and northeast and centre in diet etc .. It does appear that age at death is slightly
east in EVII:28, both with burials underneath. lower for adults during level VI, and the majority of
These rather complex data suggest that the juvenile skeletons are also from levels VI and VII.
patterning offered by Mellaart was only a rough This may be tied in with other aspects such as the
guide in the first place, and has had too much built close packing of houses, economic and technical
upon it. If the northeast corner is the one most likely changes, and social changes which seem to take place
to have a platform, it is not surprising that it is also at that time. The preponderance of female skeletons
most likely to contain burials. It is clear from the in the later levels may well tie in with the new and
limited information available in the publications that growing emphasis on the depiction of mature women
burials occurred in other parts of the buildings in the figurines and the absence of clear representa-
throughout the sequence. tions of males after level VI. Whether it also reflects
a reduced male population living at the settlement is
Male and female: numbers in context not known. Although it has been suggested (Angel
The 'excess' of adult female skeletons at C:::atalhoyiik 1971) that population crash may have been involved
has already been mentioned - although as Angel in the abandonment of the site, there is no evidence
comments, it probably actually represents a dearth yet that it was really abandoned, as the lowest levels
of adult males (1971, 79). So far, no basic counts of the western mound may prove to follow immedi-
seem to have been made of the number of skeletons ately on from the upper levels of the east mound. We
of each sex per building, and although we must must await its excavation to discover whether the
always bear in mind the incomplete nature of the settlement simply shifted, or ceased.
data, and the lack of certainty as to whether those The lack of males in Angel's data between the
groups which did survive for analysis represent prac- ages of 13+ and 20, compared to the 17 females in the
tically the entire assemblage from each building or same age range, may be the result partly of the greater
only a fraction of it, sex counts can be illuminating. ease of sexing females after puberty. It is also likely
A number of buildings appear from the surviving to be related to cause of death, although suggestions
data, some of which is apparently complete, to have concerning the presence in the settlement of differ-

254
Figurines, Clay Balls, Small Finds and Burials

ent groups at different ages should also be taken into A further 11 females and 14 males (Angel), or
account. Adult age at death may be caused by a 35 females and 23 males and a number of indetermi-
number of factors. We see a broad range of ages for nate (Ferembach) and several juveniles are unstratified.
both sexes, but male burials are at their highest in Ferembach only identified a small number of juve-
their thirties, while female deaths peak rather niles, and they have been added to Angel's data
younger, in their twenties. This may be the result of when it is clear there is no duplication. The most
differential depositional practices. Another possible striking differences between the data sets are levels
explanation is that female deaths are related to child- V and VII, in which Ferembach has a majority of
birth, and this may well be true, but a second expla- males against Angel's majority of females.
nation may underlie the data. Weak male babies
tend to die early, while females have better survival The juvenile data
rates. The stress of childbirth may be the event for It is not only male and female adult skeletons that
females which shortens life-span. If so, this ought to refuse to comply with neat normative patterns. The
be seen as equalizing the sexes, rather than creating juveniles must also be considered. Altogether 90 ju-
a net drop in the number of females. Death in battle veniles were identified, with ages ranging from 0 to
or while trading is often posited as the male equiva- 15, and all ages have fairly equal representation (Ta-
lent, but despite sundry comments about sling stones ble 12.10). No age is missing, and although a few
and head wounds, the evidence for violent deaths is seem to have slightly higher numbers than others
minimal. The occasional head-wound or parry frac- this is partly due to the inexactness of age measures,
ture - regarded by Angel as suggestive of fighting and to the imprecise ages given for a minority of
(1971, 91)- could also be a result of the system of
roof entry, at least in some cases. The deaths of males Table 12.10. Juveniles by age. (Based on Angel with some
in their thirties would therefore be viewed as natural additions from Ferembach.)
deaths, and a substantial number of females also
Age at death Totals by 1-year intervals Total by 3 year age-groups
survived into their thirties and beyond, the sexes
0=3 0-1 = 6 0-3: 21
being roughly equally represented with 36 females 0.5: 2 1-2: 12 3-6: 12
<1 : 1 2-3:5 6-9: 18
and 37 males in their thirties, 18 females and 15 1+ = 5 3-4:6 9-12 = 7
males in their forties, one female and two males 1.25 = 2 4-5:5 12-15: 29
<1.5 = 5 5-6:5
around 50 and two females and one male getting to 2=3 6-7:5
2.5:2 7-8:8
60. (All ages taken from Angel's data alone.) A re- 3=2 8-9:5
aging of the skeletons might well overturn these 4=4 9-10: 1
4+: 1 10-11:2
data (Molleson & Cox 1993). 5:5 11-12=4
6:4 12-13:9
The sex ratio of skeletons varies considerably 6+: 1 13-14:6
over time, and although this may be purely a result 7=6 14-15 = 6
7+: 2 15 = 3
of the unequal excavation of different levels and 8=2 plus 5 imprecise
8+: 3
minimal data available from some of them, it is worth 9:1
considering. Unfortunately, adult (aged over 15) data 10: 1
10+: 1
from Angel's and Ferembach's analyses differ some- 11 = 2
what. The results are shown in Table 12.9 (numbers 11+=2
12: 3 Note: Some of the + ages were very im-
of juveniles up to and including age 15 are added for 12+: 6 precise (8-10, 10-13 etc.), and these start
around age 7, making some imbalance
13:3
reference). 13+=3 in the older ages.
14:5
14+ =I
15 = 3
adolescent = 4
Table 12.9. The sex of adult skeletons in the different levels mid-adolescent= 1
according to Angel and Ferembach.

Level Angel Fe rem bach Juveniles by age and sex: Angel's level of certainty (total 44).
II 7 female 0 male 1 juvenile 3 female 0 male 5 = M??, M??, M? 12 = M??
1+=M??,F
Ill 3 female 3 male 3 juvenile 0 female 0 male 1-1.25 = M???, F?? 6 =M??, F?? 12+ = M?, M?
IV 12 female 6 male 3 juvenile 13 female 6male 1.25 = (M) 7 = M??, M??, M??, M?, F? 13 = M?, M??
v 11 female 5 male 1 juvenile 11 female 17male
1.5 = M?? 8 = F??, F?? 13+ = F??
VI 54 female 29 male 52 juvenile 49 female 30 male 8+ = (M), M?, M 14 = F??
2 = M??
VII 25 female 19 male 18 juvenile 27 female 32 male 15 = F, F
2.5 = M?, M 9 = M??
VIII 3 female 2 male 4 juvenile 3 female 4 male 10 = (M)? adolescent = M
3= M
IX 1 female 1 male 0 juvenile 1 female 0 male 4 = M??, M??, (M) 10+ =(F), M??
XI 1 female 0 male 3 juvenile 0 female lmale
4+ = M? 11+=M

255
Naomi Hamilton

Table 12.11. Juveniles


individuals (e.g. 8-10, or 11+). juveniles were regarded as male (33 males, 11 fe-
by building. The number of juveniles per males). Angel may have been correct, and this would
building varies widely, from certainly help explain the low number of male adults.
AIII:I = 3
EIV:II = I none in some, to 14 in EVII:31 On the other hand, it may be that juveniles at
EIV:I2 = 2
FV:I =I
(Table 12.11). Of the 38 build- C::atalhoyuk displayed an unusual robustness of phy-
AVI:I =I ings with skeletons attributed sique, reflecting a general robustness which could
EV1:2 = 2
EVI:I = 5 to them, 25 have the remains also partly account for a low rate of child mortality.
EV1:7 = 10
EVI:S = 2
of individuals aged 15 or Of the 17 infants and babies under the age of five, 15
EVI:II = 3 under using Angel's esti- were sexed male; if we look just at the nine babies
EVI:20 = 2
EVI:29 = 2 mates, with additions from under two and a half, seven were sexed male. While
EVI:32 = 6
EVI:34 = 12
Ferembach, who aged but did this suggests a greater death rate of male babies, the
EVI:35 = 2 not sex a number of juveniles. extent of the sex difference is rather startling. One
EV1:44 = 3
EVI:50 = 2 This means that for unknown explanation would be that Angel was deceived by
EVII:29 =I
EVII:31 = 14
reasons 13 buildings - or one the robustness of the population into incorrectly
EVII:35 =I third of those represented in sexing them male. If the sexing is accepted, social
EVII:45 = 2
EVIII:IO = I the skeletal data- had there- selection for burial could also be implied, but the
EVIII:25 =I
EVIII:31 =I
mains of adults only. The pro- low rate of juvenile females when compared to the
EXI:35 (25) = 3 portion of juveniles to adults high proportion of adult females receiving burial in
also varies, from a single baby the same places is perplexing. The extremely low
with two or three adults in EVII:29, to one juvenile number of females could suggest a female-preferred
with eight adults in EIV:11, to six juveniles with one culture- one of Angel's explanations for the high
adult and one sub-adult in EVI:32, to 12 juveniles number of adult females found in buildings. Statis-
with 11 (Angel) or 14 (Ferembach) adults in EVI:34. tics show that in male-preferred societies, while
It is difficult to interpret these figures without any young male babies still die at a higher rate than
knowledge regarding biological relationships or female ones, once past the age of 12 months this is
cause of death. If we assume that the dead were reversed as female toddlers die of neglect (Janssen-
buried where they had lived, this suggests very var- Jureit 1992, 72; Morgan 1984, 297, 427, 457, 460, 638;
ied child mortality rates in different living units. It Venkatramani 1992, 125). Another explanation would
has been argued (Forbes, pers. comm.) that the burial be male infanticide, regarded in some cases as a
record cannot be complete due to the over-represen- reasonable explanation for surplus female deaths or
tation of adult females and under-representation of a shortage of female adults (Ucko 1969). Such an
babies. However, there is no question that all ages explanation could also suggest a female-preferred
are represented at C::atalhoyuk, and that any social culture (the reverse of the Yanomamo), or could re-
selection was thus not based purely on age. flect a struggle within society over sex-based power
Angel sexed 44 of the 77 juvenile skeletons he or the development of gender roles. The removal of
examined. Although the sexing of juveniles is inex- male babies could be an effective weapon for women
act, the data may not be totally redundant. Angel whose social power was being eroded, both in op-
stated that the sex imbalance seen in adult skeletons posing an ideology of women as mothers and carers
was reversed among juveniles, and suggested that of males, and in creating their own majority.
the one explains the other. In fact, looking at the Looking at the data by level is complicated by
complete assemblage rather than level by level, adult the small numbers involved. Level VI, which gives
females never outnumbered males by 2:1, yet among the greatest range of context with data from 13 build-
the juveniles Angel's sexing shows males outnum- ings, shows a high level of juvenile mortality with 52
bering females by 3:1. Although these 33 juvenile juveniles to 84 adults. Of these, 13 are infants under
males could still not make up the entire 'deficit', it three, only three of them being less than a year old.
would make some difference, particularly if we as- Similar numbers of skeletons are attested within each
sume that the juvenile assemblage has suffered roughly three-year age group from 0-15, and the
greater loss due to fragility or alternative methods of fairly equal spacing of deaths across these ages, with
disposal. the exception of the very low number of young ba-
Angel sexed skeletons partly on robustness, and bies, does not suggest a single cause of death such as
this would have been his main criterion for juve- male infanticide. The low number of neonatal deaths
niles, other attributes being less developed. It is could be seen as an indicator of different treatment,
largely on this basis, therefore, that the majority of but the presence of six neonate skeletons across two

256
Figurines, Clay Balls, Small Finds and Burials

levels (VI & VII) in the restricted assemblage re- categories, and problems during and after excava-
ceived by Angel shows that if such different treat- tion, it cannot all be, as demonstrated by the particu-
ment did occur, it was not a fixed rule. Although we lar bones present. This shows a level of fragmentation
know from Mellaart of a neonate buried in a brick not previously suggested, and causes major prob-
in the wall of EVI: 14 (1963, 75) - a skeleton not seen lems with counting individuals. Furthermore, there
by Angel and therefore additional to the six he re- seems to have been some confusion of identities,
corded- it had apparently been given similar treat- either deliberate or accidental, during or after
ment in terms of excarnation, application of red ochre, burial - 45 individuals identified by Ferembach
wrapping in textile, and provision of grave goods. were separated from skeletons previously thought
Its presence in a brick cannot therefore be seen as the to represent one person; 26 of these were mixed
hurried disposal of rubbish, but could indicate a male and female bones, while a further 15 were mixed
death at the wrong time of year for normal burial, a with bones of indeterminate sex, the remainder be-
wish to preserve delicate bones from long exposure ing duplicate bones of the same sex. These occur in
while awaiting burial, a foundation deposit, or a levels V, VI and VII, perhaps III, and a number of
desire to provide a baby-sized equivalent of a mud- burials without context. If Mellaart was right, and
brick platform which would still ensure that the bodies were exposed on platforms- presumably on
burial took place within a building, and a flexibility the textile in which they were wrapped for burial-
of custom which made it possible to seize opportu- it is possible that bones which fell from the platform
nity (brick-making and house-building) when of- were added at random to any skeleton nearby, thus
fered. It is possible that more neonates would be inflating our estimates of individual skeletons; it is
discovered if all bricks were broken, and it is almost also possible that skeletons were deliberately mixed
certain that we have lost a disproportionate number and then shared out among various buildings or
of infant skeletons during excarnation, burial, exca- platforms within a building. This could imply that
vation and cleaning due to their fragility. complete skeletons are intact because there happened
to be no others available to mix them with - an
Fragmentation accident of death. Such mixing could relate to kin-
One aspect which is quite clear from Ferembach's ship or non-blood relationships, in order to share a
data, and to some extent also from Angel's, is the skeleton amongst significant others, or could have
high degree of fragmentation of the skeletons - by ritual implications concerning the mixing of clans,
which I mean separation of body parts from each social groups, both sexes etc. in death.
other, not damage to individual bones. Mellaart was We do not know whether the dead had lived in
clear that the burials were secondary, occurring after the building they were buried in. Mellaart reported
excarnation, and suggested some sort of mortuary that no evidence was found of holes being made in
structure where dessication, vultures or insect ac- platforms (1963, 98), and rather than assume that all
tion reduced the corpses to bones and ligaments buildings in which a death had occurred were dis-
(1963, 98; 1964, 92). This has been questioned by turbed for a single burial, it is possible that each
Todd (1976, 67), yet the skeletal data show even year, for instance, all the skeletons were gathered up
greater disturbance of the bones than Mellaart re- and given burial in a few selected buildings. This
ported. Mellaart said that complete/intact burials could account not just for the mass burnt burial in
were rare, athough the majority of skeletons dis- EV1:5, but also for the bone piles common in level
played sufficient articulation to suggest that some VI. Some levels may have lasted for at least a cen-
ligaments survived at the time of burial. According tury, but this alone does not explain why according
to Ferembach's data, complete or almost complete to Mellaart some buildings contained up to 32 indi-
skeletons are extremely rare - only nine approach viduals (probably a considerable under-estimate in
such a state, five female and four male. Ferembach view of Ferembach's data) buried in three or four
used 20 bone categories in her tables, omitting du- layers, while others contained only a couple of skel-
plicates for left and right sides. Of the maximum 282 etons. Mellaart suggested that burial in 'shrines' was
adults and 35 juveniles she identified, only one skel- more popular than burial in houses, to account for
eton had examples of all20 categories, while 88 adults this discrepancy (1967, 206), but his description of
and 24 juveniles were identified from a single bone layers of burials also suggests several episodes of
category. (This does not mean only a single bone group burial rather than a regular trickle of deaths
was present, however.) While some of this is cer- as the years passed. Occasional mass burials could
tainly due to variable survival of different bone also explain the use of several platforms, one after

257
Naomi Hamilton

another. The selection of buildings for burials could accompanied by two necklaces.lt is not clear whether
be purely practical, depending on which ones had the palette and bone fish-hook from this building
experienced a death among the occupants, which were found in the burial, although palettes were
ones had a suitable platform, which ones needed reported by Mellaart from female burials.
repairs anyway or were due to go out of use, which A few burials have strong associations with
clan, totem or 'religious' symbol was favoured by grave goods. These are described in the reports and
the dead individual and occurred in the building are generally individual burials. Thus the burial of a
etc .. This last suggestion could explain the higher num- child in EIV:8 was accompanied by a bracelet of
bers of skeletons in some heavily ornamented build- turquoise-coloured bird or goddess beads, a bone
ings, regarded by Mellaart as shrines. Alternatively, pin and a fine obsidian blade. Red ochre was applied
we could be seeing the result of epidemic diseases. to the skull and entire upper body (1964, 93). This
However, there is some evidence of more individual burial was unusual in that the body was in an ex-
burial, and the presence of mainly male adult skel- tended position with its head to the wall and feet to
etons (accompanied by juveniles) beneath the north- the centre of the room- though this may have less
east platform of one of the buildings which could be ritual than practical meaning, as a child might be
investigated through the skeletal data does suggest short enough to fit within the platform space in an
that mass burial was not always carried out, and that extended position, whereas adults are too large and
some information about the individual was retained have to lie in a contracted position. The bracelet,
to the point of burial. The identity of skeletons could bone pin and red ochre are regarded by Mellaart as
have been retained through the textile on which they female, the blade should normally be male and adult.
were exposed for excarnation, or a similar method. A skeleton in EVI:7 had a dentalium-bead necklace,
A scattering of female bones among mainly male a white marble armlet, and a red-painted basket
ones or vice versa could represent, as suggested (1964, 95). It was sexed by Mellaart as female. Arm-
above, some form of ritual or deliberate mixing, or lets are shown on several figurines, some clearly
accidental addition during excarnation. Overall, how- female but others generally regarded as male. In
ever, individual burial seems to have been rare. EVI:14 the baby in the brick was treated with red
ochre and accompanied by an obsidian chip and a
Adding the artefacts piece of shell, while a child was buried with eleven
The excarnation and fragmentation of skeletons also bone rings on its fingers (1964, 95). In EVIB:20 a
raise issues concerning grave goods. The common skeleton said to be female had an obsidian mirror, a
occurrence of mass burials and disarticulated skel- dentalium bead necklace, a round basket and an
etons interferes with simple identifications of many oval basket containing 'rouge'. The skull had been
grave goods, or associations with individual skel- treated with cinnabar; another skeleton, said to be
etons. Mellaart mentions necklaces around necks, male, had a bag of 8 obsidian projectile points beside
rings on fingers, belt hooks at the hips, even a string its leg. EVI:25 contained only two skeletons - a
skirt, and suggests that many skeletons were clothed young adult said to be female and a child (1963, 101).
when buried. This is surprising if excarnation was Although the burials are reported to be among those
taking place. It is possible that some items were in shrines which border on poverty (this building
placed on the body before excarnation, and others was earlier regarded as a house, not a shrine) the
were added to the grave during burial. If bodies adult was buried in a string skirt weighted down
were exposed on textiles, it would be a simple mat- with copper tubes, and accompanied by a necklace
ter to carry the bones back to the settlement for and bracelets, an oval wooden bowl and another
burial with clothes and ornaments in situ, but it is wooden object as well as much textile. The burnt
unlikely that they would be undisturbed. Mellaart burial in EVI:5 contained wooden vessels, fruit, flow-
regards some antler and bone items as fasteners for ers, a mica pendant, a polished bone tool perforated
cloaks for males, as they were found sometimes be- at one end, a two-part belt buckle, textile, fur and a
hind the shoulder, while females sometimes had bone wooden peg with copper on it, all from a mass burial
pins at the shoulder. Few grave goods have clear (1963, 99-101); a needle, pendant and flint tool are
skeleton contexts, although a burial from EVI:20 sexed also known from this building. Since these skeletons
male by Mellaart and also by Angel and Ferembach were not identified by Angel or Ferembach, it is not
was accompanied by a buckle, an item Mellaart claims possible to test the sexing given by Mellaart against
is male, and only one skeleton is reported from EIX:1, the grave goods described, in order to investigate
sexed female by Mellaart, Angel and Ferembach and the gender aspects.

258
Figurines, Clay Balls, Small Finds and Burials

EVIA:29 is described by Mellaart as a less dis- Juveniles tend to be found buried with adults.
tinguished house which contained rich grave goods Mellaart reported that babies and infants were bur-
(1964, 94). There were several skeletons, ten of which ied with women (1964, 93), but as in the above case
were seen by Angel. He sexed them as three males, the sexing may need to be questioned. He also re-
aged 30??, 35 and 47+, five females aged 18, 27, 28, ported that certain grave goods sometimes accom-
30- and 30, plus two juveniles aged 3 and 7. Unfortu- panied such burials, such as spoons and spatulae
nately there is no method of identifying one of these which he thought were for feeding the infant with,
skeletons with the one Mellaart describes as a male and shell or mother-of-pearl pendants (1964, 103).
accompanied by a white marble bowl, a fine flint dag-
ger with bone handle in the shape of a snake, a bone Pigment
or horn scoop, green pigment and ointment sticks. Mellaart reported the use of red ochre on 21 skel-
EVIII:1 had a burial described by Mellaart as etons, blue pigment on 10, green pigment on three,
consisting of a woman with a child on top of her, grey on one and cinnabar on several. He did not
with red ochre applied to the skeleton, and accom- attempt to explain it, but did try to give it associa-
panied by fresh water mussel shells filled with red tions. Red ochre was applied almost entirely, and
ochre, an adze, a flint dagger with a chalk pommel, a possibly solely, to female skeletons, and these al-
spoon and a spatula (1967, 207). These last two, he most always occurred in 'shrines'. Mellaart is also
says, generally occur when a child is buried with a clear that red ochre burials, although possibly high
woman (1967, 209). Red ochre is regarded as female, status, are not rich- they are generally accompa-
and so are adzes. Daggers are supposed to be male. nied only by a necklace. Red ochre also occurred in
This particular dagger has not yet been identified graves in other forms- in baskets or shells as 'rouge',
among the artefacts. or in one case just lumps of it in a grave with some
EVIII:31 contained the most complicated buri- blue pigment but not on the bones (EVII:1). I have a
als in some ways. They were reported by Mellaart in total of 26 occurrences of associations between skel-
great detail (1966, 180-83), and consisted of an un- etons and red ochre culled from the records. By far
specified number of burials without grave goods the most are in levels VII (six or seven) and VIB (six)
below a white floor, and two strange burials under but with the exception of VIA, which may have suf-
painted platforms- a young child ('little girl') in a fered in the records from the adjustments between
basket, the skeleton partly stained with cinnabar, VI, VIA, VIB and VII occasioned by changes in the
accompanied by two bracelets and two necklaces - understanding of the stratigraphy, the number of
one with a mother-of-pearl pendant and many deer instances seems to correlate quite well with the
teeth; and a young adult sexed in situ as female, number of burials excavated per level, although none
disarticulated yet in a vertical sitting position, with is known from level II or from earlier than level IX.
red ochre covering the body and applied in stripes The skeletal data do not give a great deal of informa-
to the skull and neck, wearing three long necklaces tion, as only five skeletons with red ochre were ex-
including a mother of pearl pendant, with two bone amined, of which three were female and two male. It
rings and a macehead near the body, as well as the does seem to occur in certain buildings - there were
long bones and skulls of many mice and a single two in EVII:31 (one male aged 29, one female aged
shrew. This burial is of interest first because of the 42), two or three in EVII:10 (one female labelled as
unusual method of burial and the unusual surround- having red ochre but none visible, aged 34, one fe-
ing building, next because of the presence of a male with no label but ochre visible, aged 37, one
macehead with a woman- regarded by Mellaart as female labelled and with ochre visible, aged 39), and
a symbol of authority and a male attribute - and two in EVI:14 (one of them a neonate). In addition, in
finally by Angel's sexing of the skeleton as male. some cases they occur in buildings above one an-
This apparently male skeleton was literally covered other - in EIV:14 and EV:17; EVIA:5 and EVIB:5;
with necklaces of a type apparently only found with EVIB:12 and EVII:12; EVII:1, EVIII:1 and EIX:1; and
female skeletons - not the few beads Mellaart says with a pause during levels V and VI, EIV:8, EVII:31
occur with males, not just one long and elaborate and EVIII:31.
necklace, but three. A bodkin and a bone handle/ Green and blue pigment are said to occur on
comb are also annotated as coming from this burial. skeletons of both sexes, and with rich burials (1964,
Another bone item, possibly a palette, comes from 94). Green pigment was applied as eyebrows to a
this building and although it does not have a burial skull sexed female in EVIB:20 and occurs on a skel-
context it may well belong to the same group. eton sexed male in EVIA:29 (1967, 208). I have also

259
Naomi Hamilton

found traces of green pigment on a long dentalium the building; whether the burial is individual or
necklace from level VII (171) and one necklace from mixed; if mixed, with whom; articulated or not; in-
AVI:1 (208). Blue pigment was applied to the skull volves pigment; is in the earth, on a mat, in textile, in
and neck. Traces of blue pigment survive on a neck- a basket; is contracted, extended or sitting; is accom-
lace from EVI:10 burialS and on another from EVI:44. panied by grave goods; if so which grave goods, etc ..
There is no record of which building the grey pig- It is not possible at present to address many of these
ment occurred in, but it was in level V. Cinnabar easily, and some have been discussed already.
was found on the child in EVIII:31, on a skeleton Grave goods may or may not indicate status. A
sexed female in EVIB:20, accompanied by a mirror, major problem is whose status they represent- that
necklace and rouge (1964, 95), and on the skull of of the dead person, the person responsible for bury-
another in EIX:1 whose body was covered in red ing them, the person who lives in the house they are
ochre, accompanied by necklaces and some copper buried in, the 'family', the community etc .. What is
and lead beads. clear at <::atalhoyiik is that while nothing was obvi-
Pigment could have been applied to skeletons ously made purely for burial, many grave goods
for a range of reasons. They could relate to the status required considerable investment of effort when pro-
of the dead person in terms of their occupation (wall- duced. Maceheads are regarded as high status sym-
painter?), their role within the 'family' or 'society', bols, symbols of authority, and Mellaart reports that
their age, sex or gender, or could refer to an outside they were found with male skeletons although few
factor such as cause of death. The presence of 'rich' can now be assigned to burials. Necklaces of hun-
grave goods need not imply that they themselves dreds of tiny stone beads are the most common grave
were wealthy or had high status (contra Wason 1994), goods and occur apparently with female skeletons,
as we do not know who the artefacts belonged to although there are clearly exceptions. They are not
before burial, why they were put in a grave, nor regarded as high status, although Mellaart makes it
what was regarded as wealth. At present there is no clear that the majority of burials have no grave goods
indication from the records nor the skeletal data that (1962, 52) and therefore a necklace could be regarded
the use of pigment was restricted to a particular sex as a step up from none. However, the effort required
or age group, and although it may cluster in certain to produce a necklace was probably far greater than
buildings, these often also contain many non- that involved in making a macehead. The materials
pigmented burials. It is possible that future work used for necklaces are largely imported and include
can investigate any links with cause of death, and exotica such as Mediterranean shells, deer teeth, lead,
any particular accompanying grave goods that could copper, rock crystal, and stones of many colours
suggest a pattern. carved into interesting shapes. Maceheads were made
Mellaart suggested that red ochre burials were generally of limestone, sometimes with unusual
often accompanied by obsidian mirrors, and accord- mottling but sometimes plain. If the level of status
ing to my data this occurred in three, or possibly depends on the level of investment in work and
four cases (the burial in EVIB:20 is not certain- the materials, many of the necklaces would outweigh
skull is said to be treated with cinnabar, but a red the maceheads very easily. Mirrors were regarded
ochre burial is also recorded), while five burials with by Mellaart as high status items for women, and he
mirrors have no record of ochre being used. Of the reports that they occur mainly in 'shrines'. They
three definite burials with red ochre and mirrors, would have required a great deal of work- it is still
one took place in a building regarded by Mellaart as not known how they were polished without being
a house rather than a shrine (EV:17) as did at least scratched, and of course they are made from im-
one other red ochre burial (EIV:12). ported obsidian. Bone belt-buckles were high-status
objects for males, suggested to be used for fastening
Wealth, status and hierarchy leopard-skin clothing on priests and occurring in
The question of wealth and status is difficult to ap- 'shrines'. It is clear that they did not occur solely in
proach, as the meanings of these words are value- 'shrines', and evidence from the figurines shows that
laden. Indicators of status can be very variable, and leopard skins were not regarded as male rather than
before we can recognize them we need to have some female attire. The buckles are made of easily obtain-
idea of which statuses could be involved. Status in- able and worked material (bone) and are not very
dicators in the burial record could include the type elaborate. Their status seems to be based on their
of burial/ disposal; place of burial (within or outside rarity alone. The suggested female equivalent of belt-
a building); which building is chosen; which part of buckles, bone pins, is not considered high status,

260
Figurines, Clay Balls, Small Finds and Burials

although at least one is decorated. Elaborate obsid- spatula (AIII:l), necklaces, armlets, mirror, celt, pro-
ian and flint blades and projectile points are seen as jectile points, scraper and pendant (AIV:1- no indi-
occurring with males, and regarded as items requir- vidual attribution of grave goods), wooden vessels,
ing specialist craftspeople, while polished stone adzes fruit, flowers, a pendant, a perforated bone tool, a
occur with females and pass without comment. The buckle, textile and a copper-covered wooden ?peg
investment of time is far greater for a polished than a (EVI:5, the mass burnt burial). This one seems to me
chipped stone tool, and the materials for both were to differ little from the impoverished burial in EVI:25,
imported. particularly when considering that only two skel-
Mellaart describes some buildings as containing etons were found in EVI:25 and 6-8 skeletons were
rich burials or burials bordering on poverty (1967, found under a single platform in EVI:5 - no infor-
207). Returning for a moment to pigmentation, and mation is available as to whether others were found
remembering that he argued that red ochre burials under other platforms. EVI:29 was a 'less distin-
were not rich but may denote high status, it is inter- guished house' (1964, 94) but had rich grave goods,
esting to look at his assessments of wealth. Of those said to include obsidian tools and weapons although
burials described as rich, three had red ochre burials they have not been traced (a single projectile point is
(AIII:l, EVI:5, EVII:12), while those bordering on known from the building), two necklaces (which in-
poverty included two with red ochre on the skeleton clude 18 beads probably made of lead ore) and a
(EVI:12, EVI:14) and one burial with lumps of red mother-of-pearl pendant which I have seen, and a
ochre as a grave good (EVII:1). Red ochre survives collection mentioned above said to be buried with a
on a necklace of deer teeth and copper from level male. EVII:12 is a big building with wealthy grave
VIII, and the two red ochre burials in each of EVIII:31 goods, said to include obsidian weapons and tools,
and EVIII:25 had long elaborate necklaces. The buri- and a boar's tusk collar (in Ankara Museum). I have
als in EVII:21 were described as simple and include also found a spatula and buckle belonging to this
a skull treated with red ochre and with cowrie shells building and doubtless deriving from the burials.
placed in the eye sockets (1966, 183). Broadening the With data of this nature, assessments of rank and
view to include other grave goods, building EVIA:25, status are bound to fail.
known variously as a house or a shrine, contained A brief comparison can be made with what is
only two burials- an adult and child. Said to bor- known of the burials at two very important sites in
der on poverty, this burnt burial included two the broad neighbourhood of <;atalhoyiik. A~1kh
wooden artefacts, a necklace, bracelets of stone and Hoyiik) is an aceramic site, somewhat earlier than
coal, textile, much thread and a string skirt with the <;atalhoyiik but with many similarities which prof-
ends weighted with copper tubes. Considering that fer it as an ancestor of sorts. Intramural burials be-
the majority of burials are recorded as having no low the floors of dwellings are known in small
grave goods, this one seems rather rich to me, par- numbers - four males, five females, two infants
ticularly in the use of copper - a very rare material, (Esin 1991). They are generally single, although buri-
and the presence of more than one string of beads. als of infants with adults are known, as is one double
Wood and textile may be considered to have been burial of adults. Grave gifts are said to be rare, but
lost from most burials, as these were preserved in beads appear not to be regarded as grave goods,
this case by the severity of a fire which destroyed the rather they are personal items. The beads themselves
building. The burials in EVI:14 are also said to bor- show strong similarities with <;atalhOyiik, and both
der on poverty (1967, 207), although the baby in the deer teeth and copper have now been found on neck-
brick had a piece of shell and an obsidian chip, and a laces in graves (Esin 1995). Other grave goods con-
child had eleven bone rings, while a needle, a fas- sist of a stone axe and a sharpener, items which-
tener and a stone armlet have not been attributed to along with beads - are rare among the artefacts
individual skeletons. Simple burials are mentioned assemblage. At Ko~k Hoyiik (Silistreli 1984; 1990),
in EVI:10, which contained at least 20 wooden ves- somewhat later than <;atalhoyiik, at least 17 burials
sels including boxes with lids, and several baskets. were excavated, all but two juveniles. Grave goods
The simple burial in EVII:21 has been mentioned consist of shallow ceramic bowls, sometimes with
above. EVI:34, a 'large well appointed house' also spoons, weapons and tools made of bone and obsid-
had 'no signs of luxury' in the grave goods (1964, ian, and beads made of various stones. One burial of
94), which consisted of a number of baskets, two sets a child (said to be female) had a clay figurine, a stone
of beads and two projectile points. Rich burials, on miniature 'idol' and a bone stamp with geometric
the other hand, may contain only necklaces and a curvilinear motifs. On the other hand adult skulls

261
Naomi Hamilton

are represented, with plastered features similar to crossing, but perhaps rather than suggest they were
those known from Jericho. One, treated with red bending their gender, we should consider whether
ochre after plastering, was found with a pair of black we have not been bending the evidence to fit our
stone eyes (unfortunately not in situ but believed by own preconceptions.
the excavator to come from the eye sockets). Both In attempting to separate grave goods accord-
these sites show rather different burial customs from ing to sex we are bound to run into difficulties due
<:;atalhoyiik - a dearth of burials, or emphasis on to the mixing and fragmentation of burials. We know
juvenile burials within the settlement, the presence from the skeletal record that mixed sex skeletons
of apparently adult grave goods with children, the were buried together. In such circumstances, not only
presence in graves of items not found at <:;atal- need we question whether grave goods belonged to
hoyiik - yet there are similarities such as intramu- individuals, but also whether the sex of the indi-
ral burial, beads in graves, use of red ochre. viduals was relevant at all. Burial practices at
<:;atalhoyiik even has one skull with cowrie shells in <:;atalhoyiik seem on one level to be quite fixed and
the eyes, although it was not plastered, and bone conservative, yet in others they show flexibility -
items very like the belt buckles at <:;atalhOyiik are the range of position, wrapping, pigment, grave
found at A~xkh Hoyiik, but are not mentioned as goods etc., and the great variation in numbers of
grave goods. Given the variation, we cannot look to burials per building. Some of it may reflect expedi-
other sites for assistance, but can be aware of the ency, some design. Most of these variations have
range of artefacts and burial rites in use in the area. been viewed as reflections of the status of the dead,
but I remain unconvinced by the data at present.
Conclusion Difference need not mean structural inequality
As noted at the outset, there are serious problems (Wason 1994). Ranking by age, achieved status, so-
with the data. Nevertheless, I hope that in the afore- cial roles based on skill and knowledge etc. do not
going I have demonstrated that some information necessarily contradict an egalitarian ethos.
can be teased from them. The sex segregation of the This examination of the evidence has by no
burials has been demonstrated, in those few groups means exhausted all avenues, but it is clear to me
of skeletons with sufficient information, to be impre- that although more statistics could be examined, and
cise. There may be an element of segregation in the labels could be checked again, the information we
burial of males under the northeast platform but this will gain is limited. Sufficient has been found to
must be tested further, as few examples are avail- show that burial practices at <:;atalhoyiik were not as
able, and they were certainly accompanied by juve- simple as was thought. In order to progress, we now
niles. Males were definitely not only buried under need new data, which must be collected with the
the northeast platform, and its symbolic elaboration questions I have raised in mind.
is dubious. The sex/ gender base of grave goods must
certainly be scrutinized further before it can be ac- References
cepted - there has been little opportunity to evaluate
published records against the skeletons themselves. Angel, L., 1971. Early Neolithic skeletons from <::atal
It may be that what needs to be questioned is not the Hiiyiik: demography and pathology. Anatolian Stud-
sex of the skeleton, but the gender of its owner. If ies 21, 77-98.
Burnham, H.B., 1965. <::atal Hiiyiik - the textiles and
necklaces were worn exclusively by women, then
twined fabrics. Anatolian Studies 15, 169-74.
the male skeleton in EVIII:31 may have been owned Dikaios, P., 1953. Khirokitia. Oxford: Oxford University
by a woman, yet if maceheads are really male items, Press.
that woman was also somehow a man - a gender- Duru, R., 1991. Hoyiicek Kazxlan 1990, in K.azz Sonur;larz
crosser. Gender-crossers can only exist where gen- Toplantzsz XIII. Ankara: Kiiltiir ve Turizm Bakanhg1
der exists, and perhaps the entire edifice has been Eski Eserler Miizeler Miidiirliigu, 155-61.
built on modern assumptions - perhaps, after all, Duru, R., 1993. Zwei neolithische Miniaturische aus
the skeleton in EVIII:31 represents a person, who Hoyiicek. Istanbuler Mitteilungen 43, 129-32.
wore necklaces and had a macehead, and had no- Duru, R., 1994. Kurur;ay Hoyiigii I. Results of the Excavations
1978-1988: the Neolithic and Early Chalcolithic periods.
tions of men and women very different from our
Ankara: Tiirk Tarih Kurumu.
own. The female skeleton in EVIII:l with the dagger, Efe, T., 1990. Three early sites in the vicinity of Eski~ehir:
the male skeleton in EVIA:29 with the ointment sticks, Asmainler, Kanhta~ and Kes Kaya. Anatolica XVI
the male skulls with red ochre, males buried under 1989-1990,30-50.
female platforms etc. - all are candidates for gender- Esin, U., 1991. A(;)Ikh Hoyiik kaz1s1 1990, in K.azz Sonur;larz

262
Figurines, Clay Balls, Small Finds and Burials

Toplantzsz XIII. Ankara: Kiiltiir ve Turizm Bakanhg1 London: Thames & Hudson.
Eski Eserler Miizeler Miidiirliigii, 131-53. Mellaart, J., 1970. Haczlar. Edinburgh: Edinburgh Univer-
Esin, U., 1995. Early copper metallurgy at the pre-pottery sity Press.
site of A~1hk, in Readings in Prehistory: Studies Pre- Molleson, T. & M. Cox, 1993. The Spitalfields Project, vol. 2:
sented to Halet Cambel. Istanbul: Graphis Yaymlan, the Anthropology. (Council for British Archaeology
61-78. Research Report 86.) York: CBA.
Ferembach, D., 1972. Les hommes du gisement neolithique Morgan, R. (ed.), 1984. Sisterhood is Global. Harmondsworth:
de <:;:atal Hiiyiik, in VII Tiirk Tarih Kongresi (1970). Penguin.
Ankara, 15-21. Morris, D., 1985. The Art of Ancient Cyprus. London:
Ferembach, D., 1982. Mesures et indices des squelettes humains Phaidon.
neolithiques de (:atal Hiiyiik (Turquie). Paris: CNRS. Radford, J. & D. Russell (eds. ), 1992. Femicide: the Politics
French, D., 1962. Excavations at Can Hasan. Anatolian Stud- of Woman Killing. Buckingham: Open University
ies 12, 27-40. Press.
French, D., 1963. Excavations at Can Hasan. Second pre- Ryder, M.L., 1965. Report of textiles from <:;:atal Hiiyiik.
liminary report. Anatolian Studies 13, 29-42. Anatolian Studies 15, 175-6.
French, D., 1964. Excavations at Can Hasan. Third pre- Silistreli, U., 1984. Ko~k Hoyiigii 1983, in Kazz Sonuc;lan
liminary report. Anatolian Studies 14, 125-38. Toplantzsz VI. Ankara: Kiiltiir ve Turizm Bakanhg1
French, D., 1967. Excavations at Can Hasan. Sixth prelimi- Eski Eserler Miizeler Miidiir!iigii, 31-6.
nary report. Anatolian Studies 17, 165-78. Silistreli, U., 1986. Ko~k Hoyiik 1985, in Kazz Sonuc;lan
French, D., 1968. Excavations at Can Hasan. Seventh pre- Toplantzsz VIII. Ankara: Kiiltiir ve Turizm Bakanhg1
liminary report. Anatolian Studies 18, 45-54. Eski Eserier Miizeler Miidiirliigii, 173-9.
Hamilton, N., 1994. A fresh look at the 'seated gentleman' Silistreli, U., 1988. Ko~k Hoyiik 1987, in Kazz Sonuc;lan
in the Pierides Foundation Museum, Republic of Toplantzsz X. Ankara: Kiiitiir ve Turizm Bakanhg1
Cyprus. Cambridge Archaeological Journal4(2), 302-12. Eski Eserier Miizeler Miidiir!iigii, 61-6.
Hamilton, N., in press. Ungendering archaeology: concepts Silistreli, U., 1989. Ko~k Hoyiik Kaz1s1 1988, in Kazz Sonw;lan
of sex and gender in figurine studies, in Gender and Toplantzsz XI. Ankara: Kiiitiir ve Turizm Bakanhg1
Material Culture: Representations from Prehistory to the Eski Eserler Miizeler Miidiir!iigii, 91-7.
Present, eds. M. Donald & L. Hurcombe. London: Silistreli U., 1990. Ko~k Hoyiik Kaz1s1 1989, in Kazz Sonuc;lan
Macmillan. Toplantzsz XII. Ankara: Kiiitiir ve Turizm Bakanhg1
Hodder, 1., 1990. The Domestication of Europe Oxford: Eski Eserler Miizeler Miidiirliigii, 95-104.
Blackwell. Todd, 1., 1976. (:atal Hiiyiik in Perspective. New York (NY):
Janssen-Jureit, M., 1992. Female genocide, in Radford & Cummings.
Russell (eds.), 87-94. Ucko, P., 1968. Anthropomorphic Figurines of Predynastic
Mellaart, J., 1962. Excavations at <:;:atal Hiiyiik, first pre- Egypt and Neolithic Crete with Comparative Material
liminary report, 1961. Anatolian Studies 12,41-65. from the Prehistoric Near East and Mainland Greece.
Mellaart, J., 1963. Excavations at <:;:atal Hiiyiik, second London: A. Szmidla.
preliminary report, 1962. Anatolian Studies 13, 43- Ucko, P., 1969. Ethnography and archaeological interpre-
103. tation of funerary remains. World Archaeology 1, 262-
Mellaart, J., 1964. Excavations at <:;:atal Hiiyiik, third pre- 80.
liminary report, 1963. Anatolian Studies 14,39-119. Venkatramani, S.H., 1992. Female infanticide: born to die,
Mellaart, J., 1966. Excavations at <:;:atal Hiiyiik, fourth pre- in Radford & Russell (eds.), 125-32.
liminary report, 1965. Anatolian Studies 16,15-191. Wason, P., 1994. The Archaeology of Rank. Cambridge: Cam-
Mellaart, J., 1967. (:atal Hiiyiik: a Neolithic Town in Anatolia. bridge University Press.

263
Chapter 13

Trace Element Analyses of Bones and Teeth from


~atalhoyiik

Theya Molleson & Peter Andrews

Introduction Material and methods

The re-opening of excavations at <;:atalhoyiik has Samples of human bone, permanent and deciduous
provided the occasion for a re-examination of the teeth as well as soil and an animal bone were
human skeletal material from the part of the site that analyzed by induced coupled plasma spectrometry
was excavated between 1962 and 1965 by James (ICP) for 22 elements. For this pilot study most parts
Mellaart (Mellaart 1962; 1963; 1964; 1966). This ex- and levels of the site were examined. Several sam-
tends the studies carried out by Denise Ferembach ples were taken from a single individual in an at-
(1969; 1970; 1973) and Larry Angel (1971). We are tempt to document variation in preservation and in
most grateful to Professor Berna Alpagut for ena- diet. A list of material analyzed is given in Table
bling us to carry out these studies. 13.1.
The new studies of the mandibles and denti- The elements studied provide information as to
tions are being undertaken by a team of researchers both diet and to contamination or post-burial altera-
based in the Natural History Museum, London. They tion. A rough guide to contamination, i.e. inclusion
include a detailed recording and radiography of the of extraneous material, of the human samples is in-
dental morphology and oral health. A preliminary dicated by the presence of elements that are present
study of the dental microwear features using scan- in soils and groundwaters. These include silica, tita-
ning electron microscopy has also been undertaken nium, and aluminium. Apart from one of the soil
(Molleson et al. 1996). The analysis revealed a large samples, one bone from K10.62, and the pig bone
pit size, which suggests that large objects were being EVI.34i, had Si, Ti, and Al concentrations signifi-
processed, while the paucity of scratches suggests cantly above the range of the remaining 21 bones
that there was little need to masticate or grind the and teeth. The tooth from K10.62 did not appear to
food. The main food type was probably not cereal be contaminated. Iron and manganese were also high
and certainly not boiled cereal. Taken together with in these samples. Calcium and phosphorus, the main
the general low degree of abrasion on the molar constituents of skeletal mineral, were reduced in the
teeth, the anterior faceting and large molar pits, sug- animal bone, implying alteration replacement. It is
gests that the most likely types of food being eaten unfortunate that the only animal bone available, a
were tubers and pulses. In order to test this further, pig humerus, showed this alteration, since as a con-
we have initiated an investigation into the trace ele- sequence no dietary interpretation of the data using
ments and stable isotopes present in the bones and a comparison of the pig with human data could be
teeth of the <;:atalhoyiik human population, and by made with confidence. It is understood that most of
this means it should be possible to gain further the inhumations had been secondary burials placed
insights into the diet of these Neolithic people. This under the floors of the rooms after the body had
analysis will have the further object of measuring been essentially defleshed and dried. Further, being
the extent of chemical alteration through burial in within the building structure the skeletons would
the ground. not have been subject to percolating waters. Both

265
Theya Molleson & Peter Andrews

Table 13.1. c;atalhoyiik bones and teeth analyzed by ICP. are potassium, manganese and magnesium, whereas
sulphur, zinc and iron are found in both plant and
High trace elements Mg!S Indices animal products, but in differing proportions.
EIV 85 Soil Si. Ti, AI, Pb, Mn, Fe 58.9
EIV92 Soil Si 48.2
Concentrations of these elements are particu-
EIV3 Teeth Na,K 45 larly difficult to interpret since they are essential
EIVHI Mandible bone 50
EIVHI M2roots Pb 66.7 elements and therefore always present; they are also
EIV92 H M3 roots 40 highly mobile and readily taken up by buried bone.
EVI lh 37 de root Pb 25.6
EVI lh 37 Ml roots Cu,Pb 37.5 A Pearson correlation matrix was produced after
EVI lh 37 Maxilla bone V,Pb 40
EVI 7h,i M3 roots Pb 87.5 removal of samples with anomalously high concen-
EVI 7h.i Canine germ Ba 88 trations of copper, arsenic and iron. These pairwise
EVIId Mandible bone As,Pb 48.7
EVIl 8c72 Maxilla bone Ti 45.5 correlations, were examined in conjunction with the
EVIl 8c72 M2germ 81.8
EVIl 8c72 dm2 roots 37.8 distribution in bivariate plots to elucidate relative
EVIl lOb Mandible bone As, V, Pb, Hg 23.9
EVII31 Mandible bone V,Pb 33.7
dietary input.
EVIl 35 Mandible bone Pb 54
EVIl 45 9a Bone Pb 50
EVIl 45 9a M2germ Pb 75.9 Trace element analysis
EIX 17b Red stain Ba, Pb, Hg 64
K1062 Mandible bone Si, Ti, AI, Pb 59
K1062 M2 roots 61.7 Soil strontium levels, 520-645 ppm, were found to
EVI34i Pig bone Si, Ti, AI, Sr, Ba, Fe, Mg
be within the range for the bone and tooth samples.
Strontium was not correlated with the soil-borne
these factors would have contributed substantially elements but it was found to be correlated with all
to creating an inert environment for the bones and the dietary elements. Soil manganese levels, 320-730
original trace element content has remained unal- ppm, were higher than the levels found in bone or
tered in many cases. tooth samples except for one contaminated bone,
The numbers found on the specimens were used K10.62.
in all cases, although unfortunately the meaning of Strontium, potassium and magnesium, plant
this numbering is not always very clear. The prefix E indicators, were correlated at the following levels:
and K refer to areas of Mellaart' s excavation (Mellaart Sr-K, r = 0.59; Sr-Mg, r = 0.43; K-Mg, r = 0.6. The
1962-1966), and the Roman numerals IV to IX refer strontium correlation with magnesium (r = 0.43) sug-
to successive levels. The meaning of the number 10 gests plant leaf as a primary source for these ele-
in KlO is not understood at present, so we do not ments rather than cereal. The lowest values for
know what level this specimen came from. The num- strontium were in samples from EVIL These sam-
bers that follow the letters and roman numerals in ples also had low concentrations of magnesium,
Table 13.1 in some cases designate house numbers as which is found in leaf vegetables, and higher than
allocated by Mellaart (1962-1966), and these are average concentrations of zinc, an essential element
sometimes given a letter prefix H; in other cases, most readily available in meat, fish and legumes.
particularly the numbers 62, 85 and 92, the numbers The indications then are that the individuals recov-
are probably specimen numbers of unknown origin, ered from EVIl had more meat in their diet than
and the lower case letters after a number are prob- those from other parts of the site. It may be signifi-
ably secondary categories for that number. This un- cant that some individuals from this level in E had
satisfactory numbering system may be clarified with also been treated with cinnabar, lead and arsenic
further work on the old collections. either in life or after death.
The technique used to infer dietary levels of Potassium was strongly correlated with sodium
plant and animal products in the <;atalhOyiik re- (r = 0.84). Good sources for this are fruit, vegetables,
mains was twofold: by examination of correlations milk and cereals. Sodium levels are much higher
between pairs of elements that have a similar chem- than those for potassium, which is unusual and pre-
istry and by examination of the correlations between sumably reflects high ambient levels in the soil. So-
elements that occur in different foodstuffs. Thus dium is also correlated with magnesium (r = 0.79)
strontium and barium levels are correlated in plants but not with the soil elements.
and can be expected to show a relationship higher Zinc shows correlations with soil elements and
up the food chain in herbivores, omnivores and car- cannot be a reliable indicator of dietary intake. In
nivores, although the absolute values will be pro- general, however, zinc levels were low (174+ I -44
gressively decreased. In practice barium proved not ppm compared to 250 ppm in the reference stand-
to be useful. Other elements abundant in plant foods ard, analyzed with the same equipment). This is lower

266
Trace Element Analyses of Bones and Teeth

than in the soil samples and was correlated (r:0.54) magnesium and strontium concentrations generally
with strontium, suggesting a plant source rather than recorded for bone samples can be taken to reflect
animal for this element; nor is it strongly correlated higher reliance on plant foods in adults.
with iron (r:0.29). A rich source for zinc would be the Further work will explore the relationship be-
pulses, peas and beans. Food offerings in the graves tween the trace elements within the bones and teeth
comprise these legumes and meat rather than cere- and microwear on the teeth. Initial studies of the
als. Zinc was high relative to magnesium in several dental microwear has indicated a large particle size
bone sa mples from EVIl suggesting a higher meat in the food (Fig. 13.1) but without a strongly abra-
component in the diet here. sive component in the diet (Molleson et al. 1996).
Even after removal of outliers, iron shows high- There is not a wide variety of evidence for the exist-
es t correlations with soil elements and cannot be a ence in <:;atalhoyuk of techniques for the preparation
reliable indicator of dietary intake in this material. of grains, and the microwear evidence suggests that
Sulphur is most highly correlated with stron- cereal grains, which are hard and abrasive, were not
tium and potassium (r:0.64, 0.55) and also with mag- being widely consumed by the humans. Further, the
nesium (r:0.38) and may be a pointer to the relative similarity of the crazing of the enamel of some of the
importance of pulses in the diet. teeth to those published by Cutress (Hoffman et al.
As an experiment, ratios of the values of these 1989), which describes the damage done to the enamel
elements were compared to data for groups of pulses, by Polynesians who use their front teeth to peel
roots, cereal and meat foods, taken from the litera- coconuts, suggests that roots or tubers may have
ture (Documenta Geigy 1962). The rationale for this been prepared in this way (Fig. 13.2).
is that although there will be fractionation during Direct comparison of the human bone and tooth
metabolism of these elements, the proportion should samples with trace elements from Abu Hureyra, a
remain more or less the same, and large differences contemporary Neolithic site in Syria, reveal broadly
in proportion might be reflected in the measured similar values for Sr and Ba. At Abu Hureyra, how-
bone or tooth content. Thus suckling infants will ever, the levels for these elements in the human bones
have a meat signal (i.e. their main food source is are considerably less than in the animal bones. As at
animal); while adults might have meat, plant or an Abu Hureyra the manganese values are similar in
intermediate signal depending on the predominant human and animal bone although the absolute val-
nutritional component of the diet. This reasoning is ues are higher at C::atalhoyuk. The zinc and magne-
still very crude, but was thought worth trying. sium values are also similar to those from Abu
Magnesium:sulphur ratios are highest in fruits, Hureyra.
nuts and cereals (1.40, 1.04, 0.69), lowest in meat fish Some samples from <:;atalhoyuk showed unex-
and fats (0.11, 0.11, 0.10) and intermediate in leaf, pectedly high values of heavy metals. One sample
pulses and roots (0.55, 0.45, 0.39) . The ratios for the (EIX.17b) with red staining contained mercury (as
C::atalhoyuk bones and teeth fall mostly within the
range of plant foods especially pulses (0.40-0.50).
But the low Mg/S ratio of EVI lh,i, the root of the
d eciduous canine formed in infancy, is consistent
with an animal (i.e. suckling) diet. A deciduous mo-
lar and an Ml, both have low Mg /S ratios (0.37) that
would be consistent with a mixed diet, whereas other
teeth (M2) that formed when the individual was
young though weaned have higher ratios (0.61-0.81)
which could imply a cereal component in the diet.
Late forming teeth (M3) from EVI have high ratios
(0.87). The mandible from EVIl lOb appears to have
a predominantly meat diet (Mg/S = 0.23). These
results are sufficiently encouraging to warrant fur-
ther work. Figure 13.1. The occlusal surface of an adult mandible,
Comparison of levels of magnesium in teeth EVllg, showing cracking of the enamel of the anterior
suggests a lower plant component in the diet of young teeth . The teeth are also somewhat procumbent, suggest-
children. They would be either suckling or possibly ing that they have been used for pulling or stripping
rec eiving milk from domestic animals. Higher objects.

267
Theya Molleson & Peter Andrews

Figure 13.2. A) The occlusal surface of the mandible of a


4-5 year old individual, EVIII 25, with two second and
one first deciduous molar in place. B) SEM micrograph
showing the large pits and strongly directional striations
on the crushing facet of the first deciduous molar on the
mandible. This pattern was probably produced by the
pulling of fibrous material along the tooth surface.
Magnification x180.

well as cobalt and high barium) suggesting the use High levels of arsenic and lead were recorded
of cinnabar (mercuric sulphide) although sulphur in two samples from EVIl (EVII.d, 1500, 30 ppm; and
was very low. Cinnabar was recorded as being ap- EVII.lOb, 380 ppm, 13.5 per cent). The mandible
plied to the face of several skulls (Mellaart 1967). It is EVIl.lOb had advanced periodontal disease, but the
a traditional medicine used to treat skin conditions, level of lead, 13.5 per cent, has clearly been incorpo-
especially fungal, such as ringworm, and to ward off rated postmortem (calcium and phosphorus were
ectoparasites. low). Low levels of lead (10-55 ppm) were recorded

268
Trace Element Analyses of Bones and Teeth

in 11 of the human samples and could represent arsenic eating to increase their weight, strength and
lifetime intake. It occurs in both bone and tooth sam- appetite. This habit is attributed to the Styrians of
ples and in both adults and juveniles. In the perma- Austria (Encyclopaedia Britannica 1911). It may not be
nent tooth of one individual, EVI 1h 37, lead (10 too far fetched to associate the physique of the god-
ppm) is associated with a very high level of copper dess statuettes with the ingestion of arsenic.
(125 ppm). Other samples from this individual have This multi-element examination of the human
elevated lead (15-55 ppm) but copper is within the remains from ~atalhoyiik has enabled the identifica-
normal range for the site. tion of altered samples, which cannot reasonably be
The presence of arsenic, both in low quantities used to interpret diet. Contaminated specimens can,
generally and occasionally at highly elevated levels, with caution, be used to infer diet since the addition
is intriguing. All except two of the bone and tooth of grains of say feldspar, high in potassium, should
samples contained arsenic and it was present in the not alter the relative values of dietary indicators such
soil samples. Excluding the two outliers, the arsenic as strontium and magnesium. Tooth samples are
levels are correlated with sulphide (r = 0.48) and less likely to show alteration or contamination. The
with strontium (r = 0.5) which could indicate that study has also identified the use of metals and
the element, already present in the soil, was being pointed to the possibility that the people of
ingested with plant foods. It is not strongly corre- ~atalhoyiik were experimenting with the medicinal
lated, however, with the other plant indicators po- properties of certain elements, perhaps the begin-
tassium or magnesium (r = 0.26 and 0.13 respectively) nings of inductive science.
and recent studies have shown that plant uptake is
not the major pathway of such metals as arsenic or Conclusions
lead. On the contrary 9-80 per cent of lead and 34-90
per cent of arsenic uptake in cattle is due to ingestion Samples of human bone, permanent and deciduous
of soil (Thornton 1988). Arsenic occurs naturally as teeth, soil and one animal bone have been analyzed
the sulphide, in copper ores, and as an impurity in by induced coupled plasma spectrometry (ICP) for
zinc with lead. It is normally present in soils in the 22 elements. Several parts and levels of the 1960s
range <5-40 ppm and at ~atalhoyiik levels of 20 and excavations at ~atalhoyiik were examined. There-
45 ppm were measured. Arsenic can be alloyed with sults give tentative support to an earlier analysis on
copper to make bronze. The alloying of copper with the microwear of the teeth that a non-abrasive diet
arsenic gives a harder, less brittle material and im- was being eaten by the individuals tested from
proves the workability of the object by allowing re- ~atalhoyiik, and despite the evidence for grain prepa-
peated hammering. Mellaart records beads of copper ration at the site the most likely diet was one of
and lead worn by the women as early as level IX pulses and tubers. Some elements may have had
(Mellaart 1967). Mellaart also records the presence of other uses, such as for medicine or cosmetics.
a green paint (Mellaart 1967, 209). Scheele's green is
a compound of arsenic and copper. The oxide gives Acknowledgements
a white colour to copper and has been widely used
as a cosmetic. Combined with potash or soda it is We are grateful to the General Directorate of Monu-
used as an insecticide. ments and Museums for permission to study the
There is also the possibility that arsenic was material from ~atalhoyiik, and to Professor Berna
being used medicinally by the people of ~atalhoyiik. Alpagut for allowing access to the mandibular mate-
Externally it has been used to clear the skin in cer- rial. The ICP analyses were performed in the Depart-
tain skin diseases. Taken internally arsenic is a cu- ment of Mineralogy of the Natural History Museum
mulative poison but in minute quantities it is a gastric by Vic Din, whose help and advice is gratefully ac-
stimulant. In certain forms of anaemia it increases knowledged.
the number of red blood corpuscles and also their
haemoglobin content and has been used to treat anae-
mia following malaria. Angel (1964) observed a high References
frequency of cribra orbitalia in the skulls from ~atal­
hoyiik and proposed that this was the consequence Angel, L., 1964. Osteoporosis: thalassaemia. American Jour-
of a high incidence of malaria in the population. nal of Physical Anthropology 22, 369-73.
Arsenic also affects the general metabolism, so Angel, L., 1971. Early Neolithic from <;:atalhoyiik: demog-
that occasionally individuals acquire the habit of raphy and pathology. Anatolian Studies 21, 77-98.

269
Theya Molleson & Peter Andrews

Ferembach, D. 1969. Apen;u sur le peuplement de preliminary report, 1962. Anatolian Studies 13, 43-
l' Anatolie et du moyen-orient au Neolithique. Tiirk 103.
Tarih Kurumu Belleten 33, 137-43. Mellaart, J., 1964. Excavations at <;atal Hiiyiik, third pre-
Ferembach, D., 1970. Les hommes du gisement neolithique liminary report, 1963. Anatolian Studies 14,39-119.
de <;atalhoyiik. Bulletin de la Societe d' Anthropologie Mellaart, J., 1966. Excavations at <;atal Hiiyiik, fourth pre-
du Sud-Ouest 6, 7-14. liminary report, 1965. Anatolian Studies 16, 165-91.
Ferembach, D., 1973. L'evolution humaine au proche-ori- Mellaart, J., 1967. c;atalhoyiik: a Neolithic Town in Anatolia.
ent. Palaeorient 1, 213-21. London: Thames & Hudson.
Hoffman, M.P., T.W. Cutress & M.C. Crooks, 1989. Some Molleson, T.I, B. Boz, K. Nudd & B. Alpagut, 1996. Dietary
epidemiological and scanning electron microscopic indications in the dentitions from <;atalhoyiik, in Arkeo-
features of crazing of the dental enamel of Poly- metri sonuqlarr toplantzsz XI. Ankara: T.C. Kiiltiir bak-
nesians. New Zealand Dental Journal 85, 86-90. anh~ Antilar ve Miizeler Genel Miidiirliigu, 141-50.
Mellaart, J., 1962. Excavations at <;atal Hiiyiik, first pre- Thornton, I., 1988. Soil features and human health, in Trace
liminary report, 1961. Anatolian Studies 12,41-65. Elements in Environmental History, eds. G. Grupe &
Mellaart, J., 1963. Excavations at <;atal Hiiyiik, second B. Herrmann. Berlin: Springer-Verlag, 135-44.

270
Chapter 14

Exploring the 1960s' Surface: the Stratigraphy of


~atalhoyiik

Wendy Matthews & Shahina Farid

In this chapter we discuss firstly the nature and a large 20 x 20m sounding in Area E with Mellaart's
conditions of the sections from the 1960s' excava- excavations, in order to ensure continuity between
tions in the southwest of the site, and our approaches the new work and the 1960s' records. The aim in
in recording them in the field. We then consider the these new excavations is to investigate the earlier
nature of the overall stratigraphic problems and spe- levels of the site, by taking advantage of the expo-
cific correlations with Mellaart's excavations and ob- sures already opened by Mellaart, including the area·
servations, before discussing the sequencing and of his 1963 deep sounding.
replacement of buildings and open areas, and the
site structure thematically. The nature and condition of the sections, and
In the first three seasons of renewed investiga- recording methods
tions at (atalhoyiik a total length of c. 160m of field
sections from the 1960s' excavations, has been Mellaart's excavations
cleaned, photographed, drawn at 1:20 and described, Mellaart excavated at (atalhoyiik for four seasons in
in Areas A and E of Mellaart's excavations in the 1961, 1962, 1963 and 1965. He focused his excava-
southwestern side of the east mound (Fig. 14.1). These tions on the southwestern side of the east mound,
sections have been correlated with plans from where he had first observed plastered walls eroding
Mellaart' s levels XII-III. They provide a con joinable out of the surface of the mound during survey. The
section through 11.04 m of the mound, from 1004.38- results from each season were published annually in
1015.42 m a.s.l., 2.08-13.12 m above the level of the preliminary reports in Anatolian Studies, and in an
plain and 4.88 m below the summit of the mound overview of the site in 1967 (Mellaart 1962; 1963;
which is at 1020.3 m a.s.l.. Selected areas of these 1964; 1966; 1967).
sections were in addition drawn at 1:5, described in By the end of excavations in 1965 Mellaart had
further detail and sampled for micromorphological, excavated an area with maximum dimensions of c.
chemical, conservation and pollen analyses (Chap- 80 m east-west and north-south, in the southwest-
ter 15). ern area of the site. During the course of his excava-
There are three principal objectives in studying tions he identified twelve stratigraphic building levels
the sections from the 1960s. The first is to record the from 2.12-14.5 m above the level of the plain (Figs.
sections before further erosion, to ensure new infor- 1.8-1.11). In a deep sounding in 1963 he encountered
mation is recorded. Once recorded the aim is to ar- a series of occupation horizons which extended five
rest erosion by a programme which will include metres below the current level of the plain, in a
backfilling and stepped excavations. The second ob- narrow trench below level X Shrine 8 which he esti-
jective is to study spatial and chronological variation mated was already one metre below the level of the
in the architecture, occupation sequences, and struc- plain (Mellaart 1964, 73; Chapter 2 this volume). Ref-
ture of the site, and to become familiar with the erences to Mellaart' s levels are written in Roman
nature of the deposits and potential archaeological numerals. References to individual buildings are writ-
information and contexts, before renewed excava- ten in Arabic numerals. These references may be ab-
tions. The third is to correlate the new excavations in . brevia ted to read for example 'Shrine X.8' following

271
Wendy Matthews & Shahina Farid

960,96o+-

-------- top of wall/ section


~----' bottom of wall/ section

numbers refer to section drawings


------
0 10m

Figure 14.1. Plan of sections recorded in 1993-95 in Mellaart's Areas A and Eat the southwest of the site.

the format used by Mellaart, denoting in this case had been backfilled during Mellaart's excavations in
'level X Shrine 8'. 1965, for health and safety reasons. During the last
three decades the upper reaches of the sections and
Section conditions 1993-95 walls have been subject to erosion. The bases of each
When we resumed investigations at <:;:atalhoyuk in section or wall lie buried below steeply sloping banks
September 1993, 28 years later, the area had altered of colluvial debris. In Area E the steep eastern sec-
in appearance due to erosion and backfilling. The tions towards the centre of the mound have eroded
deep sounding of 1963 was no longer visible as it back as much as 1-1.8 m , depositing a bank of eroded

272
1960s' Surface: the Stratigraphy of C::atalhoyiik

debris which is up to 2.28 m thick where Mellaart excavations were still detectable in the face of some
excavated down to level XII. The main eastern sec- sections such as section 6, where deposits had been
tion of this area, section 7, currently exposes 6.27 m buried by eroded material.
of stratigraphy from the banks of collapsed debris at The plan of level XII.25 was re-exposed by care-
its base, 1006.63 m a.s.l., to the top of its grassy ful cleaning (Fig. 14.21). Only 25 em of eroded mate-
surface at 1012.90 m a.s.l.. rial had collapsed onto the floors of level XII.25 in
In many instances Mellaart excavated up to the front of section 6. This collapsed debris was removed
inside wall face of each building unit, leaving walls with care by shovel and trowel, down to the first
standing along section faces. Many of these walls layers of water-laid rainwash c. 1 em thick which
have since eroded and fallen away from the section had accumulated after excavations ceased in 1965.
face, exposing initially the walls and then occasion- These water-laid layers were removed to expose the
ally the internal room fills of unexcavated buildings extant surface.
behind. Walls are much better preserved at the base The northern half of section 7 was not cleaned
of sections, protected behind the banks of eroded due to its great height and potential instability. At
materials. Fragments of white wall plaster still ad- the end of each season earth was banked back up
here to some of these basal wall faces. Traces of against the base of the sections in order to minimize
small fragmentary red wall painting were uncov- further erosion. In 1995 a programme of infilling
ered on a patch of plaster on the western wall of recorded areas was started.
Shrine 1 (Fig. 14.10). The painted designs comprise
two small geometric and curvilinear motifs. These Section recording
fragments were conserved and lifted in 1995. Once cleaned, the sections were photographed, both
Room fills and ashy deposits in middens gener- for colour slide and black and white records, and
ally are more susceptible to erosion than mud brick drawn at 1:20 (Figs. 14.3-14.20, 14.22). All of the
walls. However, where cracks have developed be- mud bricks in the walls were drawn, using a vertical
hind the face of a mud brick wall, whole segments of gridded frame where possible, in order to study wall
wall have clearly fallen away from the section face. construction. Each stratigraphic unit within build-
Much of the 1960s' excavation area was cov- ings and open areas, and the bricks, mortars, plas-
ered by vegetation. In many instances this vegeta- ters and packing between walls, were described
tion, though destructive as far as the roots penetrate, recording: unit number, moisture, Munsell Colour
provides a protective and binding mantel stabilizing Reference, structure, texture; inclusion type, size and
the banks of colluvial debris at the base of the sections. abundance; interpretation and any additional com-
ments.
Section cleaning Floor sequences, fire installations and fixtures
Each section was photographed to record the condi- within these sections were photographed, drawn at
tion of the section face prior to cleaning. Any vegeta- 1:5 (Figs. 15.1-15.8 & 15.11), and described in greater
tion on the tops of the sections outside the 20 x 20 m detail in order to record the microstratigraphy of
area of new excavations was trimmed to reveal the these activity areas. All descriptions used standard-
underlying section face. Vegetation within the 20 x ized procedures and terminology (Limbrey 1975;
20 m area was removed. The banks of debris at the Hodgson 1976; Courty et al. 1989), and are recorded
base of sections were removed by shovel in a series and stored in the C::atalhoyiik archive. These records
of platforms to provide access to the upper reaches will be entered into the site data base in the near
of the sections for cleaning and recording. future, enabling computer-aided analysis of specific
The thin skim of silty wash which adhered to research questions.
the section faces was removed by trowel, artist's Although the northern half of section 7 was not
palette knife and plasterer's leaf blade to expose a cleaned due to its great height of 6.27 m, many indi-
freshly cut section face. In 1993 and 1994 the banks vidual mud bricks, walls, plaster faces, floors and
of debris were only removed to the top of any sur- pits were distinct enough to enable drawing at 1:20
viving wall plaster. No wall plaster was uncovered and photography. The outline of the top and bottom
during the first two seasons and remained protected of the section, and a datum line were measured us-
in its current buried environment. In 1995, in the 20 ing an EDM. Ranging poles, two metres in length
x 20 m excavation area, any surviving patches of with 50 em markings, were set both vertically and
wall plaster were exposed down to floor level, and horizontally along the base of the section to enable
conserved by C. Silver. Pick marks from Mellaart's details to be measured accurately by eye and drawn

273
Wendy Matthews & Shahina Farid

from a distance. Overall stratigraphic problems and correlations


Visibility during recording of all sections var-
ied according to the angle of the light. Visibility was Stratigraphic problems and procedures in
bad when the sunlight was parallel to the section, correlations
but considerably improved as the sun moved to We encountered a number of stratigraphic problems
throw the section into either full sunlight or shade. in interpreting the depositional sequences in the sec-
The sections were at their freshest and most visible tions and correlating them with Mellaart' s levels.
after cleaning. They were less clear when they tended These problems related on the one hand to inherent
to dry out. All Munsell colour readings were taken difficulties in trying to correlate the two different
in the shade. sets of data available, and on the other to the com-
The three-dimensional location of all sections plex nature of the settlement and stratigraphy. The
was recorded and planned (Fig. 14.1). The buildings stratigraphic evidence from both sets of data were
within the 20 x 20 m area, and the architecture and examined independently, before attempting to cor-
exposed features of building XII.25 were also planned relate them.
(Figs. 14:2 & 14:21). Much of Mellaart's information was retrieved
during three-dimensional excavations. Much of our
Sampling data from 1993-95 relates to two-dimensional sur-
The Directorate-General of Monuments and Muse- face data whether obtained from analysis of features
ums in Turkey kindly permitted export of samples in horizontal plan, or from vertical sections. Investi-
of deposits for micromorphological, organic, inor- gative cross-sections through walls, floors and features
ganic, pollen and conservation analyses. The ana- were carried out in 1993-95, but were often restricted
lytical results are discussed in Chapter 15. in size. Analysis of stratigraphic relationships both
r ..,
VII.l9

S.VIII.l4

=======~

Courtyard
VII/VIII.lS N

~
VII
11 m

0
S.X.l S.X.8 S.VIII.lO

.................

-------------~ Antechamber

VIB.3 VIB.28
1
~~======~~~-
I
I
I
I
I

Figure 14.2. Plan of upstanding walls within the 20 x 20m area of excavations renewed in 1995.
274
1960s' Surface: the Stratigraphy of <;atalhoyiik

in plan and in cross-sections will be possible during of heterogeneous constructional packing. Mellaart
the course of renewed excavations at the site, when identified his building levels on the basis of area
we shall be dealing with three-dimensionally re- exposure, brick sizes, materials and colours (Mellaart
corded evidence. A catalogue with a summary of the 1966, 176).
information recorded during the 1995 season is pre- The problem in analysis of single building units
sented in Tables 14.3-14.4. in the sections, is to establish how changes in mud
bricks and mortar relate to changes in floor levels,
Nature of the settlement mound many of which had been removed during the course
Like all tell sites, the mound at <;atalhoyiik com- of excavation or by subsequent erosion. Does a
prises a complex sequence of buildings and occupa- change in mud bricks or mortar represent a change
tion debris which were constructed or deposited over in building level, or perhaps just a change in the
a period of time, and have been subject to anthropo- batch of mud bricks during construction of one wall?
genic and natural modifications and post-deposition- Where floors survive, they often occur at the same
al alterations (Butzer 1982; Rosen 1986; Schiffer 1987). level as the base of the walls, or at the level of only
The structure of the mound at <;atalhoyiik is one to two mud bricks up from the base of the wall
similar to that of the earlier Aceramic Neolithic site (Table 14.1). Some walls have foundations up to 60-
of A;akh Hoyiik, in the foothills on the eastern edge 80 em deep, particularly those which were built on
of the Konya plain, excavated by Professor Ufuk the downward slope of the mound, on top of ashy
Esin (Esin 1991). At both of these sites buildings deposits. It is apparent, therefore, that many of the
were often constructed next to one another, sepa- changes in mud bricks and mortar do correspond
rated by distances of only c. 2-35 em, which were with changes in floor and building level.
often infilled with packing. Entrance into buildings The danger lies in assuming that these visible
was from the roof by ladder (Mellaart 1962, 46; Esin changes in mud brick and mortar always represent
1991, 130). New buildings were often constructed on changes in building levels, or the relative heights of
the same alignment as previous underlying build- floor levels. There are clear exceptions.
ings, directly on top of earlier walls (Mellaart 1967, Some walls comprise several types of mud brick
67; Esin 1991, 137). At <;atalhOyiik, although there is which are intermixed and were almost certainly laid
considerable continuity in both the ground plan and at the same time during the construction of a single
complexity of buildings (Chapter 19), there are modi- wall. There is some evidence, moreover, to suggest
fications such as the addition of an antechamber, or that the top of an earlier wall is, on occasion, used as
larger-scale changes and fluctuations in the settle- the upstanding base for a later wall. In section 27, it
ment where buildings become open areas and vice is apparent that a later wall was constructed as a
versa. The buildings at A;;1kh Hoyiik appear to have slight overhang on top of an earlier wall (Fig. 14.9:0).
been constructed in groups with either radial or con- Both of these walls were rendered with a continuous
centric building arrangements, sharing some simi- layer of plaster, suggesting the uppermost remains
larities with elements of the settlement pattern on
the east mound at <;atalhoyiik (Esin 1991, 127-9 & Table 14.1. Height of the first floor in each building above base of walls.
144). At A;;1kh Hoyiik a larger more complex struc-
ture is separated from the rest of the settlement by a Section Building Wall Height of floors above walls
em no. mud bricks below in wall
substantial stone wall. No substantial walls dividing
V /VI, northern 5 12
different areas of the settlement have yet been un- v w II
lll.6 5 86 indistinct, at mound surface,
covered at <;atalhoyiik, although the eastern settle- built on top of ashy deposits
ment mound does comprise several distinct eminences. VIA/V, southern N >8 I
central party 14 1-2
s 60 7, on top of ashy deposits
4 VIA.46, E of 5 14 1.5
Stratigraphic problems 6 X/Vlll.25/15 Court 5 18 ?2
We were faced with two principal stratigraphic prob- 6 Vlll.25 E 5 0.5
7 Vlll.29 5 7 I
lems in interpreting the data from the field sections. 7 VII.31, antechamber N 0 0
7 Vll.31 central party 0 0
The first lay in interpreting the significance of changes 7 VII.31, E of w 12 1-2
within single building units, the second in relating 7 Vll.29, E of w 0 0
7 VIB.31 N inserted 25 2
changes in one building to the history of surround- 7 VIA.32 N 3 0.2
ing buildings. Each building was often built on top s 58 4, on top of ashy deposits
v N 0 0
of earlier walls within its own ground plan, and 5 47--60 5-7, on top of ashy deposits
37 Vll.8 w c. 18
separated from adjacent buildings by only 2-10 em

275
Wendy Matthews & Shahina Farid

of the earlier wall were in use at the same time as the and reconstruction. These and other specific rela-
base of the later wall. Mellaart observed a similar tionships are discussed for each building and area of
overhang in Shrine Vl.lO, where 'the lower 2 feet are the site, and under thematic headings below. Some
part of the original structure' (Mellaart 1967, pl. 9). stratigraphic relationships and complexities, how-
In other instances, the floors of some buildings ever, remain unresolved until future excavation.
were raised substantially within pre-existing walls.
The floors in the building east of VIB.31 were raised Procedures in correlations
1.4 m above those of level VII, in section 7 (Fig. 14.6). Despite erosion since the 1960s, many of the walls of
The floors in Shrine VIA.10 were raised 60 em above the last buildings excavated by Mellaart are visible
those of VIB (Mellaart 1964, 47). as upstanding features (see Fig. 5.1). It has been
The second principal stratigraphic problem was possible to correlate extant walls and features both
to establish how changes in the materials, construc- in plan and in the sections, with Mellaart' s plans and
tion, floor levels, use, modification and destruction transparent overlays, by analysis of settlement lay-
and levelling of one building relate to changes in the out, building and room sizes, and location of walls
history of other buildings. and features.
Most of the buildings at C::atalhoyiik were built Plans from both the preliminary reports and
next to an adjacent building, separated by a layer of from Mellaart (1967) were reproduced at uniform
deposit 2-10 em thick, between the walls. Most of scales on transparencies and overlain in order to
these deposits between walls appear to have been reconstruct the series of building levels excavated by
laid as constructional packing, which has massive Mellaart. These overlays were used in the field to
bedding and was laid during a single construction compare to the extant upstanding walls in plan, and
episode. A number of secure stratigraphic relation- to walls and features in the sections. Mellaart changed
ships between buildings have been observed during his identifications of different building levels during
the section recording operations. excavation (e.g. Mellaart 1964, 42; 1966, 166). These
1. The walls of buildings east of VII.29 and 31 had changes were due in part to the continued use of
clearly been built on top of a levelled area of some buildings whilst others were levelled and com-
earlier buildings from level VIII, visible in sec- pletely rebuilt.
tion 7 (Fig. 14.6). This levelling coincides with a The surveyed height of many of Mellaart' s floors
slight alteration in the relative plans of Shrines within buildings is recorded on the plans in the
29 and 31, and the sharing of party walls indicat- Anatolian Studies preliminary reports, as metres above
ing that at this level they are clearly contempora- plain level. Mellaart's plain level was designated as
neous. being 18m below the summit of the mound (Mellaart
2. One wall had evidently been cut into deposits 1967, fig. 3). The current absolute height of the top of
which abutted a pre-existing wall. The building the mound is 1020.3 m a.s.l.. Mellaart's plain level
east of VII.32 appears to have been particularly therefore lies 18m below this, at or close to 1002.3 m
long-lived. Ashy deposits which abutted its north a.s.l.. If we add his heights in metres above plain
wall for a short time, have been cut by the south- level to the figure of 1002.3 we can correlate his
ern wall of a later phase of a building, east of levels with the absolute heights surveyed in 1993-95
VII.31, section 7 (Fig. 14.6). (Table 14.2). The close accuracy of these calculated
3. The eastern wall of Shrine VII.10 had been cut figures for Mellaart' s plain level can be corroborated
into the western wall of Shrine 27 level X/VIII, by examining the absolute height of the floors/
removing the western half of mud bricks in the subfloors in XII.25 measured in 1994, of 1004.48 m
surviving wall, which slumped at some point in a.s.l., and comparing it to Mellaart' s height for XII.25
antiquity, evident in section 6 (Fig. 14.20) and as 2.12 m above plain level. If we subtract 2.12 from
during cleaning of section 8 (Fig. 14.16). 1004.48, the calculated height for Mellaart's level of
4. Many walls appear to abut uniform packing sug- the plain is 1002.36, a discrepancy of only 6 em be-
gesting they are contemporaneous, and share tween two different methods for calculating the cor-
similarities in materials and sequences of mate- related absolute height for his level of the plain.
rials to adjacent buildings, as in Shrines 27, 10, 8, There is a strong correlation in the heights re-
and 1levels VIII-VII (pp. 300-301). corded by Mellaart, and absolute readings taken in
It has been possible, therefore, to establish the 1993-95 of floors and architectural features visible
relationships between many buildings, and to trace today both in plan on the ground and in the surviv-
phases of both organic development, and levelling ing sections. Discrepancies between floor levels are

276
1960s' Surface: the Stratigraphy of <::atalhoyiik

Table 14.2. Correlation of the absolute readings taken in 1993-95 of floors and architectural features visible today in the surviving sections
and the heights recorded by Mellaart on plans in Anatolian Studies 1962, 1963, 1964 and 1966.
Context Section Absolute height Mellaart HI a.p.l. 1960s' plain Calculated level Discrepancies m
Summit 1020.3 18 1002.3 1020.3
Plain level 0 1002.3 1002.3
111.01 12.75 1002.3 1015.05
111.02 12.95 1002.3 1015.25
11/.06 1014.647 12.3 1002.3 1014.6 0.05
111.13 13.4 1002.3 1015.7
IV.04 11.5 1002.3 1013.8
V.61 1013.669
VIA/V, south 1012.837
VIA.09 1011.507
VIA.29 9.27 1002.3 1011.57
VIA.30 Grass 1012.63 10.36 1002.3 1012.66 -0.03
VIA.31 9.15 1002.3 1011.45
VIA.31 annex 7 1011.05 8.92 1002.3 1011.22 -0.17
VIA..l2 7 1010.6 6.78 1002.3 1009.08 1.52
VIA.33 Grass 1012.11 10.23 1002.3 1012.53
VIA.44 3 1011.167 9.16 1002.3 1011.46 -0.29
VIA.46 <3 1012.547 10.16 1002.3 1012.46 0.09
VIA/8.46, N of 4 1012
VIA.47 4 1011.265 9.12 1002.3 1011.42 -0.16
VIA.61 annex 2 9.95 1002.3 1012.25
VIA.61 shrine 9.9 1002.3 1012.2
VIB.10 5.93 1002.3 1008.23
VIB.25 7.02 1002.3 1009.32
VIB.29 6.21 1002.3 1008.51
V/8.29, £of 7 1009.51
VIB.31 5.64 1002.3 1007.94
VIBJ1 annex 7 1009.74 6.29 1002.3 1008.59 1.15
VIB.31, E of 7 1009.79
V/8.32 7 1010.6 6.78 1002.3 1009.08 1.52
VIB.44 9.16 1002.3 1011.46
VIB.45 8.21 1002.3 1010.51
V/8.47 4 1011.265 9.12 1002.3 1011.42 -0.16
VIB.61 <1 1012.415 9.9 1002.3 1012.2 0.22
VIB.61 annex <2 1012.487 9.95 1002.3 1012.25 0.24
Vl/.01 27 1005.76 3.13 1002.3 1005.43 0.33
VI/.02 33 1006.53 3.87 1002.3 1006.17 0.36
VII.04 <10 1006.27 3.41 1002.3 1005.71 0.56
VII.05 <11 1006 3.32 1002.3 1005.62 0.38
VII.06 <19 1006.35 3.52 1002.3 1005.82 0.53
Vl/.08 37 1006.18 3.67 1002.3 1005.97 0.21
Vll.10 8 1005.61 3.28 1002.3 1005.58 0.03
Vll.10, E of 12 1007.2
Vll.12 30 1006.35 3.55 1002.3 1005.85 0.5
Vll.16 4.43 1002.3 1006.73
Vll.27 7 1006.78
Vl/.29, E of 7 1008.72
V/1.31 7 1008.73
Vll.31, Eof 7 1008.31
V/1.32, £of 7 1008.985
Vl/.32, 5 of 9 1008.53
Vll.44 <3 1010.627
Vl1.45 <4 1009.93
VIII.01 <40 1005.1 2.55 1002.3 1004.85 0.25
Vlii.08 <37 1004.95 2.64 1002.3 1004.94 0.01
VIII.JO <8 1005.05 2.33 1002.3 1004.63 0.42
Vlll.12 court <30 1006.35 2.74 1002.3 1005.04 1.31
V/l/.25 6 1007.09 4.78 1002.3 1007.08 0.01
Vlll.27 18 1006.02 3.38 1002.3 1005.68 0.34
V/l/.29 7 1007.69 5.34 1002.3 1007.64 0.05
Vlll.31, Eof <7 1006.96 4.16 1002.3 1006.46 0.50
IX.29 4.43 1002.3 1006.73
IX.29a <7 1006.66 4.24 1002.3 1006.54 0.12
IX.31 2.9 1002.3 1005.2
IX.31, Eof <7 1006.77
X.08 -1 1002.3 1001.3
X/VIII.27 6 1005.85
X/V/l/.25/15 C 6 1005.82
X/.25C 6 1005.58 2.8 1002.3 1005.1 0.48
Xl.29 3 1002.3 1005.3
Xl.29a 2.72 1002.3 1005.02
X/1.25 6 1004.48 2.12 1002.3 1004.42 0.06
Xl1.29 2.27 1002.3 1004.57
Xll.29a 2.47 1002.3 1004.77
I 963 sounding -5 1002.3 997.3

Key
Context: Building levels and numbers. Italicized context references indicate floors for which absolute heights were recorded in 1993-95. Contexts in plain text refers
to wall bases for which absolute heights were recorded in I 993-95.
Section: 1993-95 field section numbers (Plan 14:1). <#refers to contexts where the floors lie just below the base of the 1993-95 sections. The absolute heights in
these cases refer to the top of debris at the base of the 1993-95 sections.
Absolute heights: Heights above sea level recorded in I 993-95.
Mellaart Ht a.p.l.: Heights of floors from 1%0s' excavations written on plans published in Anatolian Studies, above plain level, which was 18m below summit of mound.
1960's plain: Absolute height of I%0s' plain, based on height of summit 1993-95 1002.3-18 metres.
Calculated level: = Mellaart's height above plain level+ 1002.3 m a.s.l..
Discrepancies: =Absolute heights a.s.l. recorded in 1993-95 minus calculated heights from the 1960s. Where section numbers are prefixed by'<', the heights refer to
contexts where the floors lie buried just below the base of the section, and the absolute heights in these cases refer to the top of debris at the base of the
1993-95 sections, accounting for the apparent discrepancies.
Wendy Matthews & Shahina Farid

only in the order of several centimetres to tens of heights with the levels excavated in the 1960s (Table
centimetres (Table 14.2). As already observed, the 14.2), suggests that level III buildings have largely
current recorded height of the extant surface in XII.25 been eroded from the tops of the extant sections in
is only 6 em higher than the translated absolute height this area. Level III Shrine 1 had elaborate wall paint-
recorded by Mellaart. The current height of the ex- ings which depicted scenes of men dressed in leop-
tant floors in Shrine VIII.25 is only 1 em higher than ard-skin loin cloths hunting a large bull and deer
the calculated height of floors recorded by Mellaart. (Mellaart 1962, 62-5, pls. XIV-XVIII). The walls of
Some discrepancies are to be expected given varia- unexcavated buildings which have been exposed by
tions in the height of floors due to the presence of erosion of the section faces, probably correlate with
platforms, features, and slope/subsidence, for ex- levels V-IV, although no plans of buildings from
ample. Mellaart also observed considerable differ- level IV have been published for this area.
ences in the height of floors in different rooms within
some buildings. The floors in the main room of Shrine Area A-E, sections 3, 4 & 5 (Figs. 14.4-14.5 & 15.7-15.8)
VIB.31, for example, were 65 em higher than those in The lowermost excavated building in this area is
the adjoining antechamber (Mellaart 1964, fig. 2). Shrine VII.44, where a spectacular relief of two plas-
There are two considerable discrepancies. The tered and painted leopards was uncovered, con-
first is in the height of Shrine VIB.31 antechamber, served and lifted in 1965 (Mellaart 1966, 176-8, pls.
section 7 (Fig. 14.6), which in our correlations lies XXXV-XL). The walls of this shrine, and its ante-
1.15 m above the height recorded by Mellaart. This chamber which was only excavated to level VIB room
discrepancy could be due to the fact that Mellaart re- 47, are illustrated in plate XXXVa (Mellaart 1966)
attributed some buildings from level VIB to level and are the same as those which were uncovered in
VII, after further excavation (Mellaart 1966, 166). In 1993 and re-photographed. The uppermost wall in
section 7, it is evident that the floors for Shrine VIB.31 this section is a fragment of building III.6. The
are associated with an inserted wall which may cor- unexcavated buildings in the south of this section
respond to modifications recorded by Mellaart. The probably correlate with levels VIA/Von the basis of
current height of our correlation for Shrine VII.31 stratigraphic relationships between sections 3 and 7,
antechamber is 1008.73 m a.s.l., which is only 14 em and absolute heights. The excavated building from
higher than the calculated height recorded by level IV to the west, Shrine 4, would have been at a
Mellaart for VIB.31. The second discrepancy is in the calculated height of 1013.8, which corresponds with
height of VIB-A.32 which is 1.52 m higher than re- the very top of this section, one metre above the
corded by Mellaart, and currently remains an floors exposed in one of these unexcavated build-
anomaly. ings. These buildings would have lain to the east of
Other floor levels still lie buried below banks of level VIA 'court', or V building 9.
eroded material. A three-dimensional computer gen- There are two types of intrusive Classical fea-
erated layout of Mellaart' s plans is being configured ture in section 3, bell-shaped pits, and a large shaft
by researchers in the Staatliche Hochschule fur which is 7 m long and undercuts more than two
Gestalung Karlsruhe. Analysis of Harris matrices of metres of intact Neolithic stratigraphy. Both contain
the site structure should facilitate the construction classical pottery sherds. Dr I. Todd has suggested
and illustration of independent sequences and cor- the large shafts may have been dug in search of the
relations with other sequences, particularly when sandy materials employed in the construction of the
combined with evidence from renewed excavations. Neolithic mud-brick walls (Todd 1976).
The floors exposed in the north of section 4 are
Specific correlations probably from an unexcavated building/room to the
A brief summary of the key points on which correla- north of VIB/ A.46.
tions have been based for the major buildings and
features in each section area is presented here. Area E, section 6 (Fig. 14.20, 15.5-15.6 & 15.11)
Section 6 has been securely correlated with build-
Area A, field sections 1 & 2 (Fig. 14.3 & 15.1) ings and open areas of Mellaart's levels XII-VIII.25.
The lowermost building recorded in these sections The walls and features of XII.25 (Fig. 14.21) correlate
correlates with Mellaart's plans of level VIA-VIB closely with Mellaart's plan (Mellaart 1966, fig. 3).
Shrine 61, which, when excavated in the 1960s, had a The small circular hole from geomorphological
bench with bulls horns set along its length (Mellaart augering in 1993 cuts through the extant sub floor
1963, 50-52, pl. VI). Comparative analysis of absolute surface of level XII at 1004.46 a.s.l.. The surrounding

278
Key (for all figures in Chapters 14 and 15)
I top of section
I
.. paint
thin-section sample animal hole w hite wa ll plaster (Shrine 1, Fig. 14.9 only)
I I
l__ j
bottom of section
® I i/\l D
I
I
W 1N

1015
1015

.......
1014 \0
1014
0'>
0
(f)'

(fJ
1013 1013 c
.....
~
n
Scale in metres ~
g.
(!)

~
.....
t:j Figure 14.3. Area A, levels VIA/B-Ill. Sections 1-2 , including Shrine VIB-A.61, large open areas, and fire-installations (m icrostratigraphy Fig. 15.1). tlJ
.....
\0 oq"
.....
tlJ
'lj
;:r
'<
0
.....,
lOIS 1015
.n
tlJ
8"
s=
Q:
1014 1014 '<
C:
:A"

1013 1013

1012 1012

lOll Vll.S44 ~ Vll.Court42 1011

Scale in metres
==
Figure 14.4. Area A-E, levels VII-III. Section 3, including Shrine VII- VIA.44, large open areas, and a complete section through an unexcava ted building
level VIA-V (microstratigraphy Fig. 15.7).
Wendy Matthews & Shahina Farid

buildings, Shrines 27, 10, and 14, and Court 15, were section 7 represents the area where Mellaart exca-
not excavated by Mellaart. below level VIII and still vated VIA.30 and 33. Calculations suggest that these
have remnants of walls which can be clearly traced floors would have lain above the top of the section.
on the ground and correlate with his plan of level The earliest architecture now visible is the walls
VIII (Mellaart 1966, fig. 7). but not the floors of IX.29, which have very dark
The uppermost level in section 6 comprises a nar- brown bricks. In Mellaart's levels IX-VIII building
row strip of floors, 36 em wide along the entire south- 29 floors were 1.53 and 1.18 metres higher respec-
ern wall face of Shrine VIII.25, which was left by tively than those of building 31 to the south. The
Mellaart as he excavated down into levels IX-XII. double wall in these levels in section 7 (Fig. 14.6)
There is a multiple sequence of at least 10 phases of separates buildings IX.29 and VIII.29 from an
floors and periodic modifications which have been unexcavated building to the south, east of building
sampled for micromorphological analysis (Chapter 15). VIII.31. There was a major phase of rebuilding across
According to Mellaart, Shrines VIII.27 and this area of the site in level VII, discussed below.
VIII.25, were built at the junction of an upper (east- A large building has been exposed by erosion
ern) and lower (western) terrace (Mellaart 1966, 180). to the east of VII.32, which is more than 9 m in
The features in the western face of section 6 there- length from north to south. The outer wall of this
fore, could potentially relate either to the upper ter- building has partly eroded to expose well plastered
race to the east or to the lower terrace to the west as floors, and room fill which comprises steeply slop-
identified by Mellaart. ing layers of burnt and unburnt building materials.
There are four apparently secure correlations The floors were drawn at 1:5 and sampled (Fig. 15.4).
of features in the west of section 6 with Mellaart' s During cleaning of the southern sector of sec-
levels: tion 7, we uncovered the edge of a platform and
1. the wall at the base of the section is the western floors in an unexcavated building to the south of
wall of XII.25; building VII.32, which was traced 3.24 m west from
2. the overlying ashy deposits correlate with the face of section 7, where it petered out due to
Mellaart's level XI Court 15; surface erosion. This building like the well preserved
3-4. the southern walls in this section comprise the building in section 7, east of VII.32, had been heavily
northeastern corners of VIII.10, and VII.10 on burnt, dark reddish brown in colour. Multiple layers
the basis of evidence in section 8 (see below). of red paint were visible in the white wall plaster
There are, however, stratigraphic complexities which had collapsed on to the floors.
within this section which are discussed at the end of Mellaart stated that 'upon partial removal of ...
this section. houses the stratigraphy showed that Shrines E.VIB.45,
E.VIB.31 and House E.VIB.29 did not belong to level
Area E, section 7 (Fig. 14.6 & 15.4) VIB, but to level VII' (Mellaart 1966, 166). The conti-
It has been possible to correlate much of this section nuity in the use, in level VIB, of buildings which had
with levels and buildings identified by Mellaart, by been constructed in level VII is clearly visible in
analysis of the configuration of his plans and the section 7. In level VIB the floor levels were raised
heights above plain level (Table 14.2), and study of a 0.95-1.1 m, and internal walls were constructed
photograph of this section taken in 1965 (Mellaart against the face of the party walls from level VII in
1966, pl. XXXIa). Section 7 principally records a the building east of VIB.29, and in VIB.31. The build-
north-south cross-section through three linear ar- ing east of VIB.31, by contrast, was completely re-
rangements of buildings in this area of the mound built on top of cut down walls from level VII (Fig.
which are on a slightly radiating east-west align- 14.6). The southern sector of section 7 was an
ment. The easternmost buildings excavated by unroofed area in VIB. Layers of ashy midden accu-
Mellaart in each of these lines are buildings 29, 31 mulated on top of the burnt building, perhaps in an
and 32 respectively. Mellaart was not able to exca- extensive pit.
vate the southeastern corner of building 29 nor the The surviving strip of floors in the northeastern
northeastern corner of building 31 in levels VIII-VIB corner of a building are almost certainly those of
because they extended obliquely into the section face, VIA.32, although they represent one of the few cur-
and are still visible today. rently inexplicable anomalies in correlating build-
Level XII.29 floors currently lie below more than ings with Mellaart' s plans and heights, lying slightly
two metres of collapsed debris at the base of section further south than the plans suggest, and 1.52 m
7. The depression in the top of the northern half of higher. The outer face of an unexcavated building to

280
N s 1 w E

1013 1013

1012 1012

I
,_ - '- - __j

1011 1011

......
\0
0'-
1010 1010 0
[Jl-

Cfl
Scale in metres >:::
Section 4 Section 5 .....
OT'
n
!'!
Figure 14.5. Area A-E, levels VII-VIA. Sections 4-5, including Shrines VII.45 and VIB-A.47, and an unexcavated building east of VIB-A.46 s:
f1)
(microstratigraphy Fig. 15.8). $!l
N .....
Ill
00 ,..,.
...... o'Q"
.....
Ill
'"d
::r
'<
N 0
...,..,
.n
Ill
6i
s::
Q:
'<
>::::
:;-;-'

1010

1009

""'
Scale in metres

Figure 14.6. Area E,levels IX-V. Section 7, including Shrines 29 and 31, and an unexcavated building east of VII.32 (microstratigraphy Fig. 15.4).
Wendy Matthews & Shahina Farid

the east has been exposed by erosion. The internal been combined into two major composite sections
southwest corner of this building has eroded, expos- through Area E (Figs. 14.8-14.9), in order to illustrate
ing lines of the interior white wall plaster. These different wall phases, and to show the stratigraphic
buildings are surrounded by ashy deposits, like build- relationship between the detailed 1:20 section draw-
ings VIA.30 and 31. Mellaart recorded that although ings, and walls only shown in outline. Summaries of
Shrine VIB.31 was one of the best preserved build- the section information recorded in 1995 are pre-
ings standing to a height of 6 feet, its north and the sented in Tables 14.3-4 and refer to Figures 14.8-14.9.
greater part of its eastern wall had been destroyed Two small and one larger unexcavated build-
right down to the lower red dado on the platform ings were visible to the south of Shrines 27 and 10, in
(Mellaart 1964, 45). the upper part of the southern sections, each sepa-
The latest surviving level now visible in section rated by double walls (Figs. 14.9, 14.17 & 14.22). All
7 is level V in the south of the section. The floors and of these had been extensively burnt. A series of fine
walls of the building east of V.7 and the lane be- floors was visible to the south of Shrine 10, at a high
tween this building and F.V.1 have been cut by the level.
section. Excavations in F.V.1 uncovered spectacular The eroded outlines of Shrines 27 and 10 could
wall paintings with hunting scenes (Mellaart 1966, be clearly traced on the ground and corresponded to
184-94). This building is not present in the section the measurements on Mellaart's plan for level VIII,
further to the south, because the open area which the lowest level excavated in this area (Mellaart 1966,
continues to the east and south was traced by exca- fig. 7). The floors of Shrine VIII.10 are recorded as
vation. Mellaart excavated below level V in Area F being 2.45 m lower than those of Shrine VIII.25, and
to levels VIA and VIB, including Shrine 80 which 1.05 m lower than those of Shrine VIII.27 (Mellaart
lies below F.V.l. There are no published plans of 1966, fig. 7). Similar differences in height have been
these buildings in this southernmost quarter only recorded firstly in the current absolute heights of
three-dimensional reconstructions (Mellaart 1966, Shrine VIII.25 floors and the now buried floor of
figs. 5 & 6). Shrine VIII.10, estimated as a difference of 2.46 m on
the basis of calculations in Table 14.2, and secondly
Area E, sections 8 & 10-40 (Figs. 14.8-14.18 & 14.22) in the change in brick and mortar types in Shrines 27
We have been able also to correlate upstanding walls and 10 from dark bricks in level VIII to orange brown
further west in the 20 x 20 m area being cleared for bricks in level VII, which currently shows a 1.34 m
excavation in Area E with Mellaart's building plans drop in height from the top of wall DO to the top of
and floor heights (Fig. 14.2) . These sections have wall Z (Fig. 14.9).
Mellaart did not ex-
cavate below level VIII in
Shrine 10, nor level VII in
the antechamber (Mellaart
1966, 178). According to
Mellaart, level VIII.10 was
rather plain with little
decoration, unlike the later
buildings. Shrine 10 in lev-
els VIB-VIA in particular
was one of the best pre-
served and most complex
buildings at (a talhoyi.ik
and is reconstructed in
the Ankara Museum of
Anatolian Civilizations,
where it has stacked bulls
heads and goddesses
(Mellaart 1966, fig. 9, for
level VII see fig. 17). In sec-
tion 8 there are only two
Figure 14.7. Overall view of section 7. extant building levels, the

282
J'
1007 iA 1007

~
1006
__ F · ... l :
1006
i\ - - - - - - - - - - -E- ......
\,
~
1005 ~--~----~- y J Q ~~--------J 1005
;__ _________ _..,\,__ _____ __
Section20 Scale in metres Se-ction36 SectionS Section6
'house6' 'shrines1. 8' 'shrineJO' 'budding25'
Fi gure 14.6 Figure 14.20 r,.w,

Figure 14.8. Composite of west-east sections throu:;;h Area E, south facin:;;. Letters refer to descriptions in Table 14.3.

N E w
\ EE w s
1008 .......
1008 \0
! \ 0'\
0
[Jl'
1007
1007 ~ ' cc
~- s _,/ (fJ
E _:L~ DO
;·:·.:·: ...
DO
AA
~ L~ W~woll [:]
c.....
\ ~ '
. . :.-==r:::=·-1...)"'""'~'
\_________ .i... !;'
1006 n
1006 ----- H . i I o r' A
[ F i "~"mplo"""'""' \ ( . ~
\_ .__ j .....
~ 1()()5
::r
(J)
----- ---
Section 18 Section 39 Section 12 Scale in metres Section 37 Section 27 Section 14 ~
N
00
- 'shrine 27' east facing -'shrine 27' 'shrine 10' 'shrine S' 'shrine 1' 'houseS' .....
Q.l
.....
Figure 14.22 Figure 14.17
(.;J
oo·.....
Q.l
Figure 14.9. Composite of east-west sections through Area E, north facing. Letters refer to descriptions in Table 14.4. '"0
::r
'<
0
......
.()
House 2 - level VII Q.l
EU
s::
Q:
5 5
1008 1008 '<
C:
:>;"

1007 1007

scale in metres
Section 32 Section 33 Section 34
east facing south facing west facing

Figure 14.11. Hou se 2, level VII. Sections 32-4.


Wendy Matthews & Shahina Farid

Table 14.3. Summary of section information recorded in 1995 (letters in column refer to Fig. 14.8).

No. Building Type Wall Level Bricks Mortar Plaster Colour Notes
A House 6 wall N VII clay pale brown
silty clay light grey /light brownish
with fine grey
sand
silty clay light brownish grey
silty clay light grey /pale brown
B House 6 fill VII Unexcavated fills of House 6.
c House 2 wall s VII clay pale brown
fine sand/ light grey /pale brown Back view of S wall of House 2, Level VII.
silty clay
D House 2/ fill/ N VII Layered fill/packing between walls 'C'
Shrine 1 packing and '1'.
E Shrine I wall N VII clay with pale brown/light
hint of sand yellowish brown
clay light yellowish brown
F Shrine I wall N Vlll silty clay light greyish brown/ Note change in composition.
greyish brown
clayey lime white
G Shrine 1/8 wall E/W Vll clay with pale brown/light Southbound divide
hint of sand yellowish brown between Shrines I and 8.
clay light yellowish brown Appears to be party wall.
H Shrines 1/8 wall E/W Vlll silty clay light greyish brown/ Ditto
greyish brown
silty clay pale brown
Shrine 8 wall s Vll clay with pale brown
hint of sand
clay light yellowish brown
Shrine 8 wall s Vlll silty clay light greyish brown/ Note colour change in levels VII and VIII.
greyish brown
clayey lime white
silty clay pale brown
clay with white
hint of silt
K House 16 wall E Vll clay light greyish brown/ Therefore same as 'C'.
brown
clay with light yellowish brown
hint of silt
L Court 15 wall w Vll silty clay brown/yellowish brown
silty clay light yellowish grey I
brown
M Shrine 8 wall N VI or VII? silty clay greyish brown/brown Same as 'I' although brick type
clay with light yellowish brown same as ·o·.
hint of silt
N Shrine 8 wall E VII? clay with brown Rebuild or change of brick and
hint of silt mortar batch.
gritty clay yellowish brown
0 Shrine 8 wall E VII? silty clay greyish brown/brown Ditto
clay with light yellowish brown
hint of silt
p Shrine 8 wall E VII silty clay pale brown Ditto
with hint of
sand
clay light yellowish brown
lime light grey
Q Shrine 8 wall E VIII clay with light yellowish brown Southward return therefore same as T.
hint of silt
silty clay light greyish brown/
greyish brown
lime white
lime light grey
R Shrine 10 wall W/E Vll silty clay pale brown/light Possible party wall, not fully exposed.
antechamber yellowish brown
silty clay light yellowish grey +
white

284
1960s' Surface: the Stratigraphy of ~atalhoyiik

Table 14.4. Summary of section information recorded in 1995 (letters in column refer to Fig. 14.9).

No. Building Type Wall Level Bricks Mortar Plaster Colour Noles
A House 5 wall s VII sandy clay dark brown
loam
silty clay dark brown+
dark yellow-brown
silty clay light grey /light brownish Comprised several layers.
grey
B Shrine 1 wall s VII? clay yellowish brown Wall overhangs 'D'. Either new phase of
silty clay pale brown wall or architectural feature.

c Shrine 1 packing/ s VII? Fine layers representing packing


floors between 2 phases of wall or remnants of
floor overlain by overhanging wall '8'.

D Shrine 1 wall s VII? lime white Comprised several layers, obscurred


brickwork and rendered ' C' and
B' in a continuous layer.
E Shrine 1 wall s VII? clay silty brown Possibly same as 'D'.
loam
clay brown

F Shrine 1 wall s VII clay grayish brown/brown Same as 'D'.


silty clay yellowish brown
G Shrine 1 floors VII? Several laminated plaster floors with mud
packing surviving in SE corner of Shrine 1
abutting plastered walls E and F.
H Shrines 1, S walls E+W VII Double wall, northbound,
dividing Shrines 1
and S. Not fully exposed.
ShrineS floors/ VII? Fine layers of plaster and mud packing
feature? representing a feature or floors surviving
in SW corner of Shrine 8.

ShrineS wall s VI or VII? clay with pale brown/yellowish No clear bricks visible. Possibly a Level
hint of fine brown VI wall or change of brick batch.
sand
silty clay brown
loam

K ShrineS wall s VII clay dark brown Contains blocking 'L' and comprises two
clay with pale brown/yellowish types of brick and mortar. Some similar
fine sand brown to those in blocking.
clayey lime white
with fine
sand
silty clay light yellowish brown
L ShrineS blocked s VI? clay dark brown Level VII crawlhole possibly blocked in
crawl hole clay with pale brown/yellowish level VI.
fine sand brown
silty clay yellowish brown
M ShrineS wall s VII clay dark brown Same as 'K' but change in mortar
silty clay light yellowish brown composition

N ShrineS wall s VII lime white Either antechamber back wall or cut
out feature.

0 ShrineS wall E VI? silty clay pale brown


antechamber with fine
sand
clay pale brown + yellowish
brown
p ShrineS wall E VII silty clay pale brown/brown Same as 'P' with change of brick and
with fine mortar batch or different phase.
sand
silty clay pale brown /brown

285
Wendy Matthews & Shahina Farid

Table 14.4. (continued)

No. Building Type Wall Level Bricks Mortar Plaster Colour Notes
Q ShrineS wall E VII clay dark brown
clayey lime white
with gritty
sand
clayey lime white Comprises several layers.

R Shrine 8 wall E VII clay dark brown Northbound return of 'K' and 'M'.
silty clay light yellowish brown
s House28 fill VII Unexcavated Level VII backfill compris-
ing burnt building debris and collapsed
burnt wall plaster.
T House 28 walls E+W VII sandy loam dark brown Unexcavated double wall of Level VII
silty clay brown house, S of Shrine 10.
clay silty pale brown
silty clay pale brown

u House 28 wall N VII silty clay greyish brown Back view of unexcavated Level VII
silty clay greyish brown house to S of Shrine 10.
v House 28 packing/ VII Mixed silty fill/ packing between
fill unexcavated walls of House 28.

w House 28 wall N VII clayey silt greyish brown Back view of unexcavated Level VII ,
clayey silt pale brown/very pale house 28 to S of Shrine 10.
brown

X Shrine10 wall s VII silty clay pale brown/light yellow-


ish brown
silty clay light yellowish grey
and white
y Shrine 10 wall w VII silty clay pale brown Possibly wall with plastered W
clay pale brown face or small antechamber to
W of Shrine 10.
z Shrine 10 wall s Vlll silty clay light brownish grey I Note clear change with overlying wall.
greyish brown
clay with brown/ dark brown
hint of silt
silty clay grey
AA House 28 ? wall E VII clay with light brownish grey I Back view of unexcavated Level VII
hint of sand greyish brown house to W of Shrine 27.
clay light brownish grey /pale
brown

BB House 28? I fill/ VII Mixed silty fill/packing between


Shrine 27 packing double walls.
cc Shrine 27 wall w VII silty clay light yellowish brown Note clear change between levels VII and
silty clay light brownish grey Vlll.
loam

DD Shrine 27 wall w Vlll silty clay greyish brown/ dark Note terrace drop to level Vlll in Shrine 10
greyish brown wall'Z'.
silty clay light brownish grey I
with hint of greyish brown
fine sand
lime white/light grey
EE backfill VII Unexcavated backfill to S of Shrine 27.
FF wall N VII clay pale brown/light Ditto
yellowish brown
clay very pale brown/pale
brown

GG wall E VII clay pale brown/light Unexcavated southbound wall to


yellowish brown undefined building S of Shrine 27.
clay very pale brown/pale
brown

HH fill/ VII Mixed fill/packing between


packing double walls 'GG' and 'II'.
II wall w VII sandy clay pale brown/ light Unexcavated southbound wall to
yellowish brown undefined building S of Shrine 27.
gritty clay light yellowish brown/
brown

JJ Shrine 27 /?fill/ Vll Fill/ packing in cavaity between double


packing walls 'CC' and 'FF' comprised several
deposits on very distinct tip lines.
1960s' Surface: the Stratigraphy of C::atalhoyi.ik

Shrine 1
s N

1007 1007

1006 1006

Section 40
east facing

em em

Figure 14.10. Shrine 1, levels VIII-VII. Section 40.

lowest of which is level VIII.lO, the uppermost level VII.2 is visible in the composite section, north of
VII.10. Shrines 8 and 1 (Fig. 14.8). The height of the floors in
A change in brick colour from grey to pale House 2, correlate quite closely with Mellaart's (Ta-
brown in levels VIII to VII has also been observed ble 14.2). These floors both in the field and in
further west in Shrines 8 and Shrine 1. The absolute Mellaart's plans are in the order of 20-70 em higher
heights of fragments of plaster floors in the corners than those in Shrines 8 and 1, and according to
of these buildings are 21 and 33 em higher respec- Mella art's plan occur on one of a series of north to
tively than the heights of floors in the centre of the south terraces (Mellaart 1964, fig . 11).
rooms recorded by Mellaart (Table 14.2), with the Houses VII.4, 5, 6 and 12 lie on the western
floors in Shrine 8 being 42 em higher than those in edge of the 20 x 20 m area of new excavations with
Shrine 1, similar to the 54 em difference in height floors 20-40 em higher than those in Shrine 1 (Figs.
recorded by Mellaart. The walls separating these 14.12-14.14 & 14.18).
two shrines have eroded. Mellaart excavated both of
these shrines down to level X floors, which lie an Stratigraphic complexities
estimated 3.64 m below remaining backfill at the end The correlations of extant architectural remains with
of the 1995 season. The floors of Shrine X.8 lay at Mellaart's building plans and floor heights have been
approximately 1 m below the level of the plain consistently close across the whole area of his exca-
(Mellaart 1964, 73) at c. 1001.3. a.s.l .. Mellaart states vations. We have, however, encountered several
that the bottom of the sounding was a further 4 m stratigraphic complexities which remain tantalizingly
below this level, suggesting that it lies below ap- unresolved, until future excavation in this area.
proximately 7.64 m of extant backfill. The first stratigraphic complexity encountered
The back view of the north facing wall to House is at the juncture between Shrines 25 and 27, and

287
Wendy Matthews & Shahina Farid

House 4 - level VII


w EN w

1007 1007

scale in metres
Section 16 Section 10 Section 15
south facing west facing north facing

Figure 14.12. House 4, level VII. Sections 10,15 and 16.


House 5 - level VII
w EN w
West wall of shrine 1
1007 1007

Section 13 Section 11 Section 14


south facing west facing north facing

Figure 14.13. Hou se 5, level VII. Sections 11 , 13 and 14.

House 6 - level VII


W E

1007 ~
I
~ I

Lc> ·· . .£? ··: :'" l


I I
I I
I ~I
I ~W I
I
I
W I
I
'-·--·--"
scale.
in metres

Section 20 Section 19 Section 21


south facing west facing north facing
Figure 14.14. Hou se 6, level VII. Section s 19- 21 .
Court 15 and Shrine 10. This juncture was identified need to be viewed in plan and other selected sections
by Mellaart as a terrace (Mellaart 1966, 180). The during excavation. The unresolved questions are:
stratigraphic evidence available is incomplete because 1. How do the horizontal interbedded layers of ?sta-
the surviving sections are at the junction of the offset ble deposits X/ VIII.25 / 15 in the west of section 6,
corners of several buildings and open areas, which discussed in Chapter 15, relate to the surrounding

288
1960s' Surface: the Stratigraphy of <::atalh6yiik

buildings? the preceding part of this report, we wish now to


The relative stratigraphy in the western section turn to the evidence available for the sequencing
(Fig. 14.20) suggests they could be attributed ei- and replacement of buildings and open areas.
ther to: Stratigraphic analysis of the type and size of
i) level X, on the basis that the lowermost wall mud brick and mortar of adjacent buildings may
can be securely correlated with building provide information on:
XII.25, and the underlying ashy deposits can 1. access to raw materials and thereby land or, al-
be correlated with level XI Court, and Mellaart ternatively, finished structural products;
only encountered a 'court' in level X. 2. the chronology of construction;
ii) level VIII, if they abut the Shrine VIII.lO wall. 3. the collaborative or organic nature of building
The ?stable deposits abut an irregular dark construction and associations between buildings.
grey brick like feature, probably a wall face,
in the south of the section. The relationship of Construction materials
this dark grey material to the dark grey wall Samples of mud brick, mortar and wall plaster have
of VIII.lO, is not at all clear in this small sec- been analyzed in thin section in order to study their
tion. They are either contemporary, or there composition and structure, and evidence for source
is a small possibility that the VIII.lO wall is materials and manufacturing techniques (Chapter 15).
cut into the remains of an irregular wall from The walls at <::atalhoyiik are constructed from
level X which is associated with the stable large mud bricks which in the sections studied may
deposits. range in length from 76-125 em, in width from 22-33
2. Similarly, the south facing wall below Mellaart's em, and in height from 5-15 em, and were laid in
Shrine VIII.25, initially considered as a new re- walls a single brick width thick. The bricks would
cently exposed wall in level X, is in fact on the have required considerable care in handling, and
same absolute level as the western wall in Shrine perhaps some form of rigid support to transport
VIII.27. On Mellaart's plan for level VIII there is them from the place of manufacture to the construc-
a 1.4 m drop in height from floors in Shrine tion site. It has been suggested that the bricks may
VIII.25 down to floors in Shrine VIII.27 to the have been moulded in situ on the walls, but this
south, which we have identified but have been requires experimental testing.
unable to verify on the basis of the stratigraphic The bricks range in colour and texture from:
evidence currently available in the field sections. silty clay 5YR 7 I 1 light grey
There is moreover a complicated three wall clay loam 10YR 6/4light yellowish brown
boundary between Shrines 25 and 27 (Mellaart sandy clay 10YR 5/4 yellowish brown
1966, fig. 7). fine sandy loam 2.5Y 5/2 greyish brown
3. Multiple and complex wall boundaries exist fur- sandy loam lOYR 4/4 dark yellowish brown
ther west in the 20 x 20 m area being cleared for Many of the bricks were made from natural sediments
excavation. The return walls between Shrine 8 exploited from the surrounding region, and stabi-
and Shrine 10 antechamber comprise a complex lized with vegetal remains represented by linear and
series of four different changes in mud brick and curvilinear voids in the mud-brick matrix. Other
mortar. In one area, continuous wall plaster ad- bricks were made from sediments mixed with het-
heres to three of the wall types indicating that if erogeneous anthropogenic debris, similar to many
they do represent different walls from different of the mortars. The bricks were laid with alternating
building levels as opposed to different uses of layers of mortar, and generally overlapped one an-
material in one construction phase, then at some other for structural stability. The mortar in some
stage the middle wall was cut to a stub, and all instances was at least as thick as the mud bricks.
three used as one contemporary wall face. The The mortar is often greyish brown (1 OYR 5 /2) -
white plaster face of a west facing wall in Shrine very dark greyish brown (lOYR 3/2) in colour and
27 may relate to the unexcavated antechamber. has many anthropogenic inclusions and charred frag-
ments suggesting it was excavated from surround-
The sequencing and replacement of buildings and ing occupation debris, probably from open areas,
spaces with some addition of sediments. Similar types of
mortar were also used in the third millennium BC in
Having considered the nature of overall stratigraphic Tell Abu Salabikh in central southern Iraq. Other
problems and the details of specific correlations in types of mortar were made from natural sediments

289
Wendy Matthews & Shahina Farid

ranging in colour and texture from brown loam to applications of plaster 1-3 em thick. At least three
white marl. The combinations of mud brick and mor- layers of plaster were applied in the eastern fire-
tar are manifold, and await computer aided analysis installation, twelve in the western fire-installation.
in the near future. The fire bowls were cleaned out before the applica-
The walls were plastered with couplets of firstly, tion of each layer of plaster. In thin section only
a pale brown preparation plaster and secondly, a isolated aggregates of fuel residues were visible be-
finishing coat of white plaster. 160 layers of plaster tween each layer of plaster, less than 2 mm thick.
comprising 80 couplets have been identified in one Each layer of plaster was rubified by the heat at the
fragment of collapsed wall plaster 3 em thick in thin base of the fire.
section (Chapter 15). The north facing wall face in A shallow-rimmed basin in the northwestern
Shrine VIII.8 antechamber was well preserved when corner of House VII.2 was rendered with layers of
exposed in 1995, section 22 (Fig. 14.15). A series of white plaster, and is currently of unknown function .
narrow vertical slots in the wall <1 em wide filled The bench in section 3 was constructed from a
with white plaster were clearly visible. The slots variety of mud brick sizes, and was covered in a
may have been deliberately formed in order to pro- layer of plaster up to 1.3 em thick. The platform
vide a secure surface into which a preparation plas- make-up in section 9 includes aggregates of multiple
ter could be keyed, before subsequent application of layers of white plaster visible in thin section.
successive layers of plaster during the use of the A number of moulded features such as horn
building. fittings were visible in the sections, particularly in
The lumps of pise and mud brick encountered section 3, associated with a Bas jaw and 3 horn cores.
during section cleaning require further study, before These were manufactured from a distinctive reddish
we can comment on other materials such as roofing. sediment which was heavily stabilized with added
Mellaart records that the roof of Shrine Vl.lO ap- vegetal remains represented by voids in the sedi-
pears to have been 'c. 0.3 m thick, flat and made of mentary matrix.
layers of mud and reeds laid on small beams' (1963,
70 & fig. 14). Wall construction
Timbers served a range of structural roles in All of the walls at <:;atalhi:iyi.ik are constructed from
the buildings at <:;atalhi:iyi.ik (Mellaart 1963, 60). Large single widths of mud bricks each up to one metre or
charred juniper timbers were exposed in the burnt more in length. Most buildings were constructed
fill of the building to the east of VII.32, section 7. with their own walls, and occur as a double wall
These were sampled for dendrochronology (Chap- construction, next to walls from other buildings. The
ter 17). load bearing outer walls of buildings are often sepa-
Few fixtures or plastered features were exposed rated from walls of adjacent buildings by a thin layer
during the section-cleaning operations. The two fire- of constructional packing, 2-10 em thick which often
installations from section 2 were constructed with a includes lumps of assorted hard-core. Dr J. Coulton
mud brick base and sides, and lined with successive has suggested that the characteristic abutting or

Shrine 8 antechamber - level VII


E

1007 1007

scale in metres
Section 22
north facing
Figure 14.15. Shrine 8 antechamber, level VII. Section 22.

290
s
1007 1007

1006 1006

1005 1005
.......
\0
Scale in metres 0\
0
- ~ 1006 [f)'

[J)
~
.....
.......
ilJ
n
~ ~
w EIN siE w ,...,.
::;-'
l1l

Figure 14.16. ?House/building and Shrine 10, levels VIII-VII. Section 8. ~


.....
N ilJ
,...,.
\0
....... ciQ"
.....
ShrinelO South of shrine 10, level VII ilJ

N 5 E w I 15-
'<:
0
.......
---
1007 1007 I ~s::
o:
'<:
~:
~

1006 1006

uI •
I :
I ' '
\ L_-: ~ ·
scale in metres
Section 17 Section 12
west facing north facing

Figure 14.17. Shrine 10, levels VIII-VII. Sections 12 and 17.


House 12 - level VII/VII
w EN s w EN s

1007 1007

---
------------- .------ -
scale in metres
Section 28 Section 29 Section 30 Section 31
south facing west facing south facing west facing

~
Figure 14.18. Hou se 12 , level VII. Sections 28-31. ro
:::1
0..
'<
$::
t:ll
....,_
....,_
::;'
ro
N ~
\0 (fl
N
~
[J)
::;'
t:ll
N I I a:
:::1
1007 1007 l t:ll
'Tj
t:ll
::::1.
0..

I
1006 1006

1005 1005

Scale in metres

Figure 14.20. House, Court and Shrine 25, levels XII-VIII (m icrostratigraphy Figs. 15.5-6) and Court X/VIII.2 5/15 ?stable deposits (micros tratigraphy Fig.
15.11) . Section 6.
1960s' Surface: the Stratigraphy of <;atalhoyuk

double wa lls at <;atalhoyuk and ea rlier sites such as common comprises red eposited anthropogenic de-
A;;1kh Hoyuk may have been critical design elements bris with heterogeneous aggrega tes. These deposits
in sharing or supporting the weight of the roofs, on are massively bedded and have unoriented, ran-
th e basis of ethnographic parallels (Coulton pers. domly distributed inclusions suggesting they were
comm.). The use of double walls may have been deposited as a single depositional unit. A second
employed, in addition, for insulation from the cold, more unusual series of deposits was observed in the
fire hazard s or noise for exa mple, or as rigorous or packing between Shrine VII.27 and an unexcavated
symbolic boundary markers. building to the south (Fig. 14.22). These deposits
Three major types of packing between the walls comprise parallel diagonal wedges of pise-like ma-
have been observed. The first and by far the most terial with a sharp edge which appears to have been
pressed up against an ob-
ject such as a wooden plank.
The deposits between these
diagonal wedges are uncon-
solidated with some
banded bedding similar to
that observed on modern
spoil heaps. It is likely that
these deposits were shov-
elled down between th e
walls, after the pise like
material had hardened, and
the object such as a plank
had been removed. Large
mammal scapulae have
been found in packing be-
tween walls, and are likely
to have been used as shov-
els, and/ or possibly as
ritual deposits.
The third type of de-
Figure 14.19. Sections 28 to 36 and buildings VIII:12, VIJ:2 and Vll:16 in the area posit between walls occurs
excavated by Mellaart. between building VII .32

,..--- .....
---.. ...... , I / '\
\ I' II
N I I I
I ~ ,' 0
I I I
I \ I
\ /
I .............. _...,/
I
I
I
I
I
r---L-- .... ,
I \
I I
I I
0 I I
I I
I I
I

Figure 14.21. Plan of House 25, level XII floor, showing location of sou thern platform and hearths, eastern bench, the
eroded base of a fire installation in the north of the room, and the 1993 geomorphological auger hole (G) .

293
Wendy Matthews & Shahina Farid

and the building east of VII.31 which are further of four layers of burials below it (Mellaart 1966, 166).
apart than usual, at 20 em. The upper layers between As more dates from radiocarbon, optically stimu-
these walls comprise heterogeneous layers of ashy lated luminescence and dendrochronological analy-
refuse and ?wind and water laid deposits, which ses become available, we will be able to document
may have accumulated over a period of time. The the longevity of buildings and the rate of structural
southern wall of the building east of Vl1.32, appears changes.
to have cut through these deposits, suggesting that At least nine phases of modification of fire-
they did not accumulate between walls, but rather in installations and perhaps other structures have been
an open area up against VII.32. The overlying build- identified in Shrine VIII.25, represented by sequences
ing VIB-VIA.32, excavated by Mellaart, was also of periodic levelling debris, rebuilt fire-installation
separated from the surrounding buildings, by in this bases, and successively raised platforms (ten phases)
case, distances in the order of 60 em. (Fig. 15.5).
There were no double walls in levels XII and The southern wall of the building to the east of
XI, according to Mellaart, but instead there were VII.31 was cut down and rebuilt, and the floor surface
party walls occasionally supported by buttresses raised 1.4 m, without any associated modifications
(Mellaart 1966, 168). Partition walls within buildings to surrounding buildings at this time (Fig. 14.6).
are a single brick width thick, and may be thinner Two blocked entrances have been identified, a
than the standard brick sizes. blocked crawlhole in a ?party wall in the south of
The wall in section 27 had an overhang, at the Shrine 8.VIII, and a blocked doorway in the western
juncture of two types of mud brick and mortar (Fig. wall of House 2. Both of these features were blocked
14.9:0). This feature may represent a change of phase with bricks which differed from those of the original
in the wall construction which was subsequently wall. There is no overlying stratigraphy to deter-
continuously rendered, as one unit, employing the mine whether the blocking may have been conducted
extant top of the earlier wall as the basal face of the to provide structural support for later walls built on
upper wall inside the building below an overhang. the same alignment, or may have occurred during
Some buildings excavated by Mellaart had up to the use of the building.
three overhangs incorporated into a single wall,
which he suggested may be a relic from earlier tradi- Use of buildings
tions of construction of wooden houses (Mellaart 1964, The microstratigraphy and micromorphology of
85-6). floors and occupation deposits, and information on
use of space within buildings is discussed in Chap-
Modifications of buildings ter 15.
A second mud-brick wall was occasionally built in The total depth of floors and occupation se-
front of the plastered surface of an earlier wall. Ex- quences in the buildings studied is generally only 2-
amples of this occur both in the sections in Area E, 10 em, with the exception of the sequence in VIII.25
and in areas which have been planned by surface which, despite erosion, is still 37 em thick. Floors
clearance operations. The new interior wall was then and occupation sequences only represent 2-55 per
itself replastered on many occasions. In level VIB cent of the height of the preserved walls, which range
Shrine 61 section 2, a new wall was built in front of in height from 0.42-2.71 m (Fig. 14.23). The tallest
a previously burnt wall in the western half of the extant walls excavated by Mellaart stood to a height
building. In other instances the plaster on the origi- of 2.7 min Shrine VIA.lO (Mellaart 1963, 70), which
nal wall was very thick and irregular, and may have was increased to 3.3 m when he removed these floors
no longer have been suitable for plastering or deco- to find those of VIB.10 60 em below (Mellaart 1964,
ration, and was therefore blocked in behind the new 50).
wall which could bereplastered smoothly, as in build-
ings VIB.31 and east of VIB.29, section 7. Replacement of buildings
Such practices obviously extend the use-life of Four major types of infilled deposits have been iden-
buildings, and argue for considerable longevity in tified in the sections:
the history of individual buildings. Mellaart esti-
mated that some of the buildings at <;atalhoyi.ik were 1. Burnt structural debris
used for over a century, such as Shrine VII-VIB.31, Mellaart states that 'whereas the crowded settlements
on the basis of the numbers of layers of plaster, the of levels VII and VIB were not - or only partly -
general deterioration of the walls, and the presence destroyed by fire, all subsequent building-levels

294
1960s' Surface: the Stratigraphy of ~atalhoyi.ik

perished in conflagrations' E
(Mellaart 1966, 172).
The only extant ex-
ample of burnt infill in
1008 1008
buildings excavated by
Mellaart occurs in the
northern antechamber of
Shrine VIB.31. Only a
small disturbed section
survives through the
northern antechamber of
Shrine VIB.31, the room fill
includes collapsed baked 1007 1007
mud bricks in charred de-
bris. This burning may
have been localized be-
cause Mellaart states that
the main room was not
burnt.
A new series of five
burnt adjacent buildings in 1006 1006
a linear east-west align-
ment along the southern
Section 18
edge of Area E has been north facing
exposed by erosion. This
Figure 14.22. Shrine 27, levels VIII-VII. Section 18.

Figure 14.23a. Bar chart


3
illustrating depth of
floors(white) and height 2.5
rfJ
of walls (black) in metres Q)
.....
...... 2
in buildings which are Q)

sealed by later levels. For s


correlation of these
.s...... 1.5
.;::::
arbitrary building co
"Q)
1
numbers with section .;::::
0.5
numbers and 1960s'
levels and buildings 0
please see concordance ...-< ...-< ...-< ...-< N ('<") -.:!< Lf) \0
"'r:--: 00 0\
r:--: r:--:
table below. C'i c<"i
"' r:--: r:--: r:--: r:--: r:--:
building numbers for each section
r:--:

40
35
30
Figure 14.23b. Bar chart illustrat- -;::: 25
ing depth of floors as a percentage ~ 20
.....
of the height of walls in buildings ~ 15
which are sealed by later levels. For 10
correlation of these arbitrary 5
building numbers with section 0
numbers and 1960s' levels and
buildings please see concordance
table below. building numbers for each section

295
Wendy Matthews & Shahina Farid

burning occurs in buildings to the east and south of Tringham (1990) has suggested that burnt build-
VII.32 in sections 7 (Fig. 14.6) and 9 (Fig. 15.8), and ings in southeastern Europe may have been ritually
south of Shrines VII.27 and VII.10 (Figs. 14.9, 14.17 burnt at the end of their use. The localized, perhaps
section 12, and 14.22 section 18). Less than 30-50 em controlled, extent of the burning discussed here, may
of burnt infill survive in the latter westernmost four support the hypothesis that some buildings at
of these buildings. This fill includes very dark red- <;atalhoyiik were ritually burnt at the end of their
dish brown rubified deposits, with occasional or- use. This hypothesis may be further supported by
ange baked bricks visible in section. The platform the fact that the infill of the burnt building east of
and multiple layers of plaster in section 9 were VII.32 includes unburnt mud bricks which are iden-
rubified a homogeneous dark reddish brown. In thin tical in size, colour and texture to those used in the
section the vegetal stabilizers in the platform and construction of the adjacent new building in level
plasters were preserved charred in situ. VIB, and does not include the old debris from the
Mellaart records that the conflagration at the adjacent walls of the building east of VII.31 which
end of VIA was so intense the heat must have stood higher, and could have easily been
penetrated to the depth of about 3 feet or more pushed into the cavity in this burnt or burning build-
below the floor level of the buildings, carbonizing ing. On the contrary, the lens of charred debris which
bodies and burial gifts alike and preventing all overlies the wall stub of the building east of VII.31
further bacterial decay. To this we owe the conser- suggests this building was levelled before the build-
vation of numerous perishable materials: flesh and ing east of VII.32 was burnt. The burnt building was
tar on human bones, desiccated and carbonized not infilled as part of a levelling operation onto which
brains inside human skulls, traces of the wrap-
later buildings were constructed. The area was
pings - skin, leather, fur and textile - in which
the dead were shrouded, and the accompanying subsequently used for deposition of refuse. It is not
perishable burial gifts, wooden bowls, cups, plates, clear from the stratigraphic evidence in the section
boxes, staffs, as well as their contents, food etc .. whether the southern part of this building was only
(Mellaart 1964, 85-6) partially infilled, and later covered by refuse, or
whether it was cut by a large refuse pit.
The linear arrangement of the buildings might sug- Mellaart suggested that ritual practices may
gest a conflagration of this area, but, given the lim- have taken place with regard to the cessation of the
ited exposure of the buildings to the south of Shrines use of buildings. With regard to AIII.8:
27 and 10, the extent of the burning is not known. after its destruction by fire the entire building was
The localized but intensive burning within the build- remodelled with panelled walls, thickly coated with
ing east of VII.32, however, remains enigmatic. white plaster and then filled in with a clean fill of
The infill of the burnt building east of VII.32 decayed mud brick containing not a single object.
slopes steeply down from the north. This fill in- After the fire it had evidently not been used and
cludes baked mud bricks and mud brick fragments, the care taken over its remodelling previous to its
filling in suggests again that we are not dealing
orange and dark pink in colour, which constitute 20-
here with a private house, but a shrine. No burials
80 per cent of the deposits in layers within the fill, large were found below its floor. It is now clear that
beams of juniper wood more than 20 em in diameter, exactly the same procedure was followed after de-
collapsed wall plaster burnt pale blue, unburnt wall struction (by fire) of ... AIII.l, EIV.1, EVI.8 and
plaster, and unburnt mud bricks, which constitute EVI.7 ... In the case of the remaining shrines
0-50 per cent of the layers of infill. These components (EVI.l0,14,31, and AVI.l) the rooms were already
tend to occur in layers 10-50 em thick which are partly filled with masses of burnt brick of walls
interbedded throughout the fill. The burning in this fallen and fused within the fire which made clear-
exposed section of the building was confined to the ance with stone tools impracticable. Under those
circumstances more debris from other buildings
southern half of the building where it has baked the
was added and no attempt was made to clear and
plaster on the walls and rubified the outer face of the plaster them, so they were found intact with their
mud bricks behind. The northern wall was not burnt. contents. The others had been more or less cleaned
This localization is surprising, given the intensity of out. (Mellaart 1963, 47-8)
the burning which had baked more than half of the
bricks in the fill, and had burnt the wall plaster on
the western wall blue in colour. Similar areas of local- 2. Deposits with mud brick and plaster relief
ized burning within buildings have been observed fragments
during surface studies at the north of the mound. Only two buildings in the sections have sizeable

296
1960s' Surface: the Stratigraphy of <;atalh6yiik

fragments and quantities of mud brick and plaster for excavation currently exposed, includes deposits
relief fragments, namely the southern building in sec- rich in plant remains and bone.
tion 3, and the eastern building in level III, section 1. The infill of buildings at <;atalh6yiik was a com-
During excavation of some buildings, such as plex practice, and in the future will be studied in
E.VIB.31, Mellaart observed that: conjunction with analysis of the use of the buildings,
The large head, arms and feet are broken and as the extant height of walls, and duration of phases,
the building was not destroyed by fire it would for example.
seem that this destruction was a deliberate act, a
sort of desecration before it was abandoned and Construction of replacement buildings: levelling and
filled in. (Mellaart 1964, 47) foundation preparations
In other instances in buildings in levels VIB and VII Buildings were often constructed on top of the sur-
which had not been destroyed by fire but which had viving walls of earlier buildings, which on average
been abandoned after very long occupation artefacts, stood to an extant height of 1.25-2.15 m. Level V
beams and animal heads had often been removed, appears to have been generally less well represented
and were pulled down and filled in (Mellaart 1964, in the areas excavated by Mellaart. In section 2, it is
39 & 50) clear that the walls probably from level V had been
razed to a height of only 30-40 em, and were cut into
3. Homogeneous, well sorted mineral sediments by later buildings.
and flecks of white ?plaster The building east of VIB.31 was built on top of
Many of the buildings in the sections have a homo- cut down walls from an earlier building in level VII.
geneous well-sorted infill of mineral sediments with A thin burnt lens from the destruction of the large
flecks of white ?plaster of currently unknown origin, burnt building east of VII.32 overlies these cut down
which was almost certainly deposited within the walls, suggesting that the burnt destruction of the
buildings as a single event. This infill does not in- building may have taken place after the ground in
clude obvious pushed in walls or collapsed mud the adjacent plot had been cleared. The unburnt bricks
bricks. In some cases this is because the walls were in the room fill of the burnt building east of VII.32
left standing to considerable heights, and rapidly have the same distinctive large size, pale brown col-
infilled when the building was no longer in use. In our and clayey sand texture as those used to con-
other instances, however, the whereabouts of mud struct the building on the cleared plot east of
brick from walls which have been cut down to a VIB.31.
height of approximately one metre, remains enig-
matic. They may form the bulk of the aggregates in Stratigraphic sequences in open areas
the between-wall packing. This needs to be re- The major types of deposit which occur in open
checked in the field. areas are discussed in more detail in Chapter 15. The
bulk of deposits in open areas comprises ashy layers
4. Deposits rich in plant remains and bone derived from hearth and oven rake-out and refuse
Refuse deposits in buildings occur either as i) homo- swept or thrown out of the houses and shrines. The
geneous layers up to 0.8 m thick which accumulated sequences of clean floors within buildings must be
over a short period of time, perhaps imported as studied in the light of the large quantities of refuse
levelling fill, or as ii) thin more heterogeneous layers dumped in the open areas.
which probably accumulated over a longer period of The extant sequence of ashy deposits in Area A
time originating from periodic cleaning within build- is more than 2.6 thick, and in the south of Area E is
ings, hearths and ovens, and from roof tops. The 2.0-2.3 m thick. Many of the individual depositional
build up of heterogeneous deposits in level XI within layers are less than 10 em thick. The density of min-
the walls of XII.25 appears to have occurred after eral sediment derived largely from decayed build-
some collapse of the wall plaster onto the floors. ing materials varies, and in some layers in section 2,
Mellaart had observed that many abandoned build- predominates. Deposits in open areas either abutted
ings were used for rubbish disposal and sanitation. or were cut into buildings, and occasionally were cut
Refuse was not only deposited within upstanding into pre-existing refuse as in the western half of
walls, but on occasions was deposited within large section 2, close to the base. There is some evidence to
pits which cut through existing walls, as in levels V- suggest such deposits excavated in antiquity may
III section 2 (Fig. 14.3). have been used to infill buildings as a single
The extant building infill in the 20 x 20 m area depositional event (p. 299). The angle of depositional

297
Wendy Matthews & Shahina Farid

slope in these open areas varies from horizontal to The complexity of the site structure and phas-
steeply sloping. The relationship between open ar- ing is evident in section 7, in particular.
eas and buildings is discussed below (p. 299). The Only one completely horizontal phase of level-
deliberate scoops filled with highly burnt ash in the ling was identifiable, from level VIII-VII, along the
north of section 2, level III, may be the product of contours of the mound. This levelling can only be
industrial activities, and perhaps relate to contem- traced across the space of two buildings, due to the
porary changes in pottery technology (Chapter 9 by more limited exposure at the base of the section.
Jonathan Last on pottery). This levelling does, however, suggest some form of
co-operation between the new builders. Indeed, the
Site structure overlying buildings all share party walls, namely
buildings VII.29 and 31, and the two buildings ex-
Terracing posed by erosion to the east. There is a change in the
During the course of his excavations Mellaart identi- plan of the antechamber in VII.31 which is extended
fied 12 building levels. He records that buildings in north to take up space originally used by VIII.29.
each of these levels were terraced down to both the There is some evidence to suggest that there
west and the south (Mellaart 1964, fig. 3; 1966, figs. were shared material sources and more or less con-
5-6). The top of level V in the east of Area E for temporary rebuilding events in levels VIII-VII in
example was 5 m higher than the buildings on the particular. The bricks of Shrines 27, 10, 8 and 1 are all
lower western part of the slope. dark grey in level VIII and change to pale slightly
We have detected at least one stratigraphic ex- orange brown in level VII.
ample of terracing as attested by a later building Although the final levelling of level VIII build-
cutting into an earlier building. The eastern wall of ings was uniform, there had previously been some
Shrine VII.lO was cut into the western wall of Shrine organic development within the structures of the
X/VIII.27, along the juncture of an east-west terrace separate buildings. The materials and floor heights
identified by Mellaart (1966, 180). Similar sequential of building IX and VIII. 29 differed from those in
changes in mud bricks and mortars have been traced both Shrine IX and VIII.31 and the unexcavated build-
across four buildings on an east-west alignment ing to the east. The significance of this change in
through the terrace detected by Mellaart. The abso- material resources may relate to changes in land use
lute heights of the changes observed in the surviving off-site.
field sections, correspond with the levels in metres A similar difference in the organic develop-
above the plain recorded by Mellaart (Table 14.2). ment of adjacent buildings is attested in level VII.
These levels suggest Shrines 27 and 25 on the upper The building east of VII.31 was apparently rebuilt
terrace are 0.74-2.45 m higher than Shrines 10,8 and and the floors raised 1.4 m, whilst the adjacent build-
1 on the lower terrace. ing east of VII.32 continued in use, unmodified. This
The extant height of floors in House 2 suggests modification is similar to the raising of floors and
there was also a series of north to south terraces as insertion of walls in level VIB.29 and 31. Although
illustrated by Mellaart. The building at the southern the insertion of walls into the building east of VIB.29
end of section 3 which has rooms with fallen bucrania and building VIB.31 is at the same absolute level and
and a bench, was apparently on a higher terrace would appear to be contemporaneous, the relation-
from a building further to the south. ship of these modifications to those in the building
Given the evidence for some organic develop- east of VII.31 which was completely rebuilt is not
ment of adjacent houses discussed below, there evident from the surviving section. If our correla-
would probably have been steps of a metre or more tions with Mellaart' s levels are correct, the division
between buildings at irregular intervals. suggested by the terms VII and VIB is misleading for
there is more continuity in this section between VII
Organic development and major phases of and VIB than there is between VIB and VIA, as Mellaart
abandonment/destruction, levelling and replacement himself stated (1964,42). There appears to be consid-
Variations in heights of floors and buildings appear erable organic development within individual build-
to be the effects in part of organic development of ings from level VII-VIB, with up to three phases of
individual building units, and in part dictated by the modification evident in some buildings.
structure of the underlying mound, and may have
been to provide light in narrow upper windows Relationships of buildings and unroofed spaces
(Mellaart 1962, 46). The area occupied by the major context types in

298
1960s' Surface: the Stratigraphy of ~atalhoyiik

vertical sections through the mound has been traced 160

and calculated using a computer graphics tablet by 140

T. Ritchey, for the main north-south and east-west 120


sections, sections 1-3 and 6-8, (Figs. 14.23-14.24), 100
which include a greater range of deposits than the 'il 80
wall elevations which comprise the bulk of the com- 60
posite sections in the 20 x 20 m area cleared for 40
excavation. Buildings occupy 149.34 m 2, 64.4 per cent 20
of the total231.98 m 2 area in these sections, unroofed
0
areas 66.8 m 2 or 28.8 per cent and pits, many of them Buildings Open areas Pits
Classical, 15.84 m 2 or 6.8 per cent. Within buildings,
walls constitute the largest area component preserved Figure 14.24a. Area in m2 of the three major context
in the sections at 100.8 m 2, or 67.5 per cent, room fills types in sections 1-3 and 6-8.
41.79 m 2, or 28 per cent, and floors only 5.51 m 2 or 3.7
percent. 70
Much of the western half of section 2, and con-
60
siderable areas of section 3 comprise layers of ash
50
rich in charred plant remains up to 3 em in diameter,
and bone. These refuse deposits variously cut or
~
" 40
30
abut buildings, indicating that they accumulated both
20
in large open pits, and in open areas or courtyards.
10
The sediment input in the open layers of ash varies
0
in colour and concentrations according to the pres-
Walls Room fill Floors
ence and nature of the surrounding mud-brick walls.
The layers of ash in section 2, therefore, include red- Figure 14.24b. Percentages by area in m2 of walls, room
dish brown sediments from the cut stubs of Shrine fill and floor sequences within buildings as represented
VIB-VIA.61, brown sediments from the face of the in sections 1-3 and 6-8.
level V wall, and orange sediments from the face of
the wall for building 111.2. ashy deposits which are characteristic of some sec-
There is some alternation between areas used tors of this area of the site in VIA, when there are less
for buildings and open areas. The areas of the west- tightly packed buildings and more unroofed areas
ern end of Shrine VIA.61 in Area A, section 2 (Fig. (Mellaart 1964, 39-40).
14.3) and the large burnt building east of VII.32 in
Area E, section 7, for example, were subsequently Associations of buildings
used as an open area. Some areas appear to have There were no double walls between buildings 29
been more consistently used for buildings as in the and 31 in level VII which were excavated and planned
case of Shrine 10, whilst other areas appear to have by Mellaart. These two buildings are unusual for
been more consistently used as open areas such as they share the same party walls (Mellaart 1967, fig.
the western end of section 2. The distribution of 10). The buildings to the east which have been exposed
open areas should be compared to the distribution by erosion also share party walls with buildings 29
of buildings which have been infilled with either and 31. The walls consist of single bricks which are
massive deposits or deposits which accumulated plastered on both sides, facing west into building 31
more slowly through time, in order to understand and east into the unexcavated building behind. The
the full range of discard and deposition within the presence of these party walls may attest the presence
settlement, in conjunction with renewed excavations. of a larger more multi-roomed structure(s) closer to
When buildings were constructed on top of ar- the centre of the mound, and a greater degree of site
eas which had previously been open areas rich in complexity. Because these buildings have not been
layers of ash and refuse, there is occasionally evi- excavated, it cannot be ascertained whether the oc-
dence for terracing and extra foundation bricks as in cupants or activities within these four buildings were
section 7 levels VIA and V. related in any way, as the sharing of party walls
In section 7, the building to the east of VIB.31, might suggest. Mellaart observed that:
now exposed by erosion, has been severely trun- On the upper terrace itself, the close association of
cated in antiquity and is cut by pits and overlain by rooms 29 and 31 apparent in levels VI and VII is

299
Wendy Matthews & Shahina Farid

still more pronounced in level VIII (fig. 7), where with areas and levels excavated in the 1960s, provid-
room VIII.29 seems to have served as a meeting- ing a vital link to the information from the extensive
room and is provided with a long bench set against excavations conducted in the 1960s. The information
the north wall and several ovens. (Mellaart 1966, 180) recorded has enabled study of spatial and temporal
The change in the boundaries between these build- variation in the nature of the architecture, occupa-
ings from level VIII to level VII, therefore, may ei- tion sequences and structure of the site. The
ther be indicative of a complete break with the past, stratigraphy is complex, and some issues raised can
or of some form of co-operation or association be- only be resolved by future excavation. The informa-
tween the occupants, but must be considered within tion recorded from these sections will be entered
the context of the more widespread construction ho- into the site data base in the near future, enabling
rizon in these levels from Shrines 27, 10,8 and 1. computer-aided analysis of specific research ques-
tions, in conjunction with new evidence from forth-
Comparisons of buildings and shrine/non shrine coming excavations.
The character of mud bricks and mortar does not
appear to vary according to elaboration of the inside References
of buildings. Some variations in the microstrati-
graphy have been observed to suggest differences Butzer, K.W., 1982. Archaeology as Human Ecology: Method
between domestic and cleaner or ?more ritual activi- and Theory for a Contextual Approach. Cambridge:
ties (Chapter 15). The features within buildings have Cambridge University Press.
been plotted (Figs. 1.3-1.7). The height of walls and Courty, M.A., P. Goldberg & R.I. Macphail, 1989. Soils and
the nature of the fill and replacement of buildings or Micromorphology in Archaeology. Cambridge: Cam-
open areas could be plotted against this scale, al- bridge University Press.
Esin, U., 1991. Salvage excavations at the pre-pottery site
though the information in the sections only refers to
of A~1kh Hoyiik in Central Anatolia. Anatolica 17,
a few of the buildings which have been excavated, 123-74.
and to many which have not yet been. In some in- Hodgson, J.M., 1976. Soil Survey Field Handbook. (Soil Sur-
stances one side of a building may be much better vey Technical Monograph 5.) Harpenden: Soil Survey.
preserved than another, such as in the burnt build- Limbrey, S., 1975. Soil Science and Archaeology. London:
ing in section 7 east of VII.32, or Shrine VIB.31. Academic Press.
Mellaart stated that although Shrine VIB.31 was one Mellaart, J., 1962. Excavations at <;:atal Hiiyiik, first pre-
of the best preserved standing to a height of 6 feet, liminary report, 1961. Anatolian Studies 12, 41-65.
its north and the greater part of its eastern wall had Mellaart, J., 1963. Excavations at <;:atal Hiiyiik, second
preliminary report, 1962. Anatolian Studies 13,43-103.
been destroyed right down to the lower red dado on
Mellaart, }., 1964. Excavations at <;:atal Hiiyiik, third pre-
the platform (Mellaart 1964, 45). The western and liminary report, 1963. Anatolian Studies 14,39-119.
southern walls ofVIB.31 had plastered reliefs of bulls Mellaart, J., 1966. Excavations at <;:a tal Hiiyiik, fourth pre-
heads and goddesses. The remaining segment of the liminary report, 1965. Anatolian Studies 16, 165-91.
east wall had a plastered relief of a figure with hair Mellaart, J., 1967. (;a tal Hiiyiik: a Neolithic Town in Anatolia.
flowing out behind it (Mellaart 1964, 45-7, fig. 7). London: Thames & Hudson.
Rosen, A.M., 1986. Cities of Clay: the Geoarchaeology of Tells.
Conclusions Chicago (IL): University Press of Chicago.
Schiffer, M.B., 1987. Formation Processes of the Archaeologi-
We have recorded and drawn a total length of 160m cal Record. Alburquerque (NM): University of New
Mexico Press.
of sections through 11.04 m of the mound from the Todd, I., 1976. (;atal Hiiyiik in Perspective. Menlo Park (CA):
extant surface of the 1960s' excavations and new Cummings.
surfaces which have been exposed by erosion in the Tringham, R., 1990. Households with faces, in Engendering
intervening three decades. We have been able to corre- Archaeology, eds. J. Gero & M. Conkey. Oxford:
late many of the observed features and buildings Blackwell, 93-131.

300
Chapter 15

Multiple Surfaces: the Micromorphology

Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

Introduction currently lies north of the Mediterranean woodland


climax zone, in the Konya plain within Zohary' s zone
This report is an assessment and evaluation of varia- of Xero-Euxinian steppe forest (Querco-Artemisietea
tion in occupation sequences within different build- anatolica) and dates to c. 6200-5500 be. Tell Brak is
ings and areas of the settlement as observed in located in the Mesopotamian steppe in northeastern
microstratigraphic sequences recorded in the field, Syria, and periods investigated range from c. 3600-
and large resin impregnated thin sections of undis- 1800 BC. Tell Abu Salabikh lay on an ancient branch of
turbed sequences of floors and deposits, studied in the Euphrates in the Saharo-Arabian desert vegetation
the laboratory. We are very grateful to The Directo- zone in southern Iraq, the period studied dates to c.
rate-General of Monuments and Museums in Tur- 2500 BC. Saar is in the Sub-Sudanian vegetation zone
key for kindly permitting export of these samples on the island of Bahrain, and dates to 2000-1800 Be.
for scientific analysis. Three samples from the earlier Aceramic
The research aims and procedures are discussed Neolithic site of A~1kh Hoyiik, in the foothills to the
before we examine the micromorphological evidence east of the Konya plain are also being examined, two
for analysis of i) the materials and technology of from storage bins, and one from an open space rich
floors/surfaces, ii) the impact of activities on floors in plant remains. These samples were kindly col-
and surfaces, iii) the type, taphonomy and contex- lected by Fiisiin Ertug, with the generous permis-
tual relationships of artefacts, organic remains and sion of Professor Ufuk Esin.
sediments in occupation deposits, and iv) post- Analysis of the microstratigraphy of occupa-
depositional alterations. Contextual variation in tion sequences both in the field and in thin section is
microstratigraphic sequences within buildings and enabling us to identify variations in the formation of
open areas is then examined. depositional sequences and uses of space in differ-
This research is part of a three year project sup- ent sociocultural and environmental contexts. At all
ported by the Natural Environment Research Coun- four sites we have observed differences in the type,
cil at The McDonald Institute for Archaeological thickness and frequency of floors and occupation
Research, University of Cambridge. The aim of the deposits which vary according to context.
project is to develop the application of micromorpho- Ethnoarchaeologists and geoarchaeologists
logy to the study of occupation sequences and the working in the Middle East have recorded similar
use of space in early urban sites in the Near East. The contextual variations in the character of occupation
research is being conducted by Dr W. Matthews and sequences. Kramer in particular noted that:
Dr C.A.I. French, in collaboration with Professor M.K. The floor of each area within a house compound is
Jones in Cambridge, and Mr T. Lawrence and Dr peculiar to that kind of area and therefore diagnos-
D.F. Cutler at the Plant Anatomy Section, Jodrell tic of primary function ... it is likely that an exca-
Laboratory, Royal Botanic Gardens, Kew. The objec- vator could readily distinguish between roofed and
tive is to examine the nature and formation of occu- unroofed areas, [and] identify stables, storerooms,
pation sequences in different sociocultural and kitchens and living rooms ... by evaluating varia-
tions in floors. (Kramer 1979, 148-9)
environmental contexts. The four sites in this project
lie on a transect through major geobotanical zones in Butzer has examined urban deposits from a wide
the Near East (Zohary 1973, map 7). <:;atalhOyiik range of geographical contexts and observed that:

301
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

The rate and type of build up differ on living floors, order to aid interpretation of deposits during future
in streets and alleys, or in and around community excavations. The aim in selecting field sections and
structures, such as civic buildings, walls, terraces, samples for analysis was to study a wide but repre-
and drainage systems. (Butzer & Freeman, in Rosen sentative range of context and deposit types. The
1986, xiii)
samples discussed in this report are principally from
They argue that all component materials in dep- the field sections exposed in Mellaart's Area (Chap-
ositional sequences, including sediments are poten- ter 14), and two palaeoecological sections (Roberts
tially informative about cultural behaviour and Chapter 2). The locations of the microstratigraphic
settlement history. sections studied are illustrated in Figure 14.1. The con-
Anthropological studies of the use of space have texts include a range of open areas from levels XI-III,
illustrated that space as defined by architectural units and, within buildings: floors, platforms, fire-installa-
is endowed with meaning through practice, through tions and architectural materials from levels XII-V.
the activities carried out in that space, which are In the field the sections selected for microstrati-
both informed by, and therefore representative of, graphic analysis were cleaned with an artist's pallete
sociocultural behaviour and conceptual schemes knife in order to cut rather than smear deposits, and
(Bourdieu 1977; Moore 1986). Analysis of micro- to bring out the character of the boundaries between
stratigraphic sequences enables study of the intended depositional layers. The sequences were then photo-
and actual uses of space by examination of the types graphed and drawn at 1:5 (Photographs 15.1-15.2 &
of floors or surfaces, the impact of activities on those Figs. 15.1-15.8 & 15.11). Each identifiable deposi-
surfaces, the relationship between sediments, arte- tional unit in the field was described adapting stand-
facts and organic remains in deposits, and post- ardized procedures employed by soil scientists,
depositional alterations. The McKellar hypothesis archaeological scientists (Hodgson 1976; Limbrey
states that it is the smaller artefactual remains which 1975) and micromorphologists (Bullock et al. 1985;
are 'more likely to become primary refuse' even when Courty et al. 1989). These descriptions are recorded
periodically cleaned (Schiffer 1983, 679). Maintenance in the ~atalhoyiik archive.
and discard practices are likely to vary according to The entire sequence of floors and accumulated
context. Analysis of entire sequences enables us to deposits in each of the rooms studied was sampled
study indications of behaviour through time and to for thin-section analysis. Many of the extant sequences
identify any changes which need not be represented are less than 10 em thick, and could be sampled in
in the architecture. Consistent repetitions in the se- one block, 13.5 x 6.5 x 8 em. Where sequences were
quence of floors and occupation deposits may sug- thicker, overlapping samples were collected. 51 thin-
gest continuity and/ or cyclicity in human behaviour section samples were collected from archaeological
and sociocultural practices. Any changes in the contexts, 4 from palaeoenvironmental contexts, and
microstratigraphic sequences may be expected to rep- 17 from ethnoarchaeological contexts. The block sam-
resent changes in behaviour or sociocultural or envi- ples were cut out from the section face using a Swiss
ronmental context. Army knife (Photograph 15.2). They were then
Previous applications of micromorphology to wrapped tightly in laboratory tissue and tape, and
studies of occupation sequences and use of space marked with their top, orientation in the field and
within archaeological sites include analysis of de- sample number. The blocks were air dried in the
posits within Palaeolithic and Neolithic caves field and packed for export in well padded wooden
(Goldberg 1987; Wattez et al. 1990; Courty et al. 1991), boxes. The thin sections were prepared in the
and a range settlements in the Near East (Matthews Geoarchaeology Laboratory in the University of Cam-
& Postgate 1994; Weiss et al. 1993) and elsewhere in bridge following established procedures (Murphy
the world (Courty et al. 1989; Davidson et al. 1992; 1986). The blocks were gently oven dried at 20-40° C
Courty et al. 1994; Macphail1994). The application of for at least one week. They were impregnated with
micromorphology to analysis of plant remains in an unsaturated crystic polyester resin under vacuum
particular is being developed (Wattez & Courty 1987; for 24 hours and were then allowed to harden over a
Goldberg et al. 1994), and ethnoarchaeological re- period of six weeks to prevent distortion during cur-
search undertaken (Goldberg & Whitbread 1993). ing. The thin sections are 13.5 x 6.5 em in size and
25-30 Jlm thick. They were cut, ground and polished
Procedures in method and interpretation using a Brot micrometric advance grinder polisher by
The objective in method design is to provide links Julie Boast (Geoarchaeology Laboratory, Department
between field and laboratory characterizations in of Archaeology, University of Cambridge).

302
Multiple Surfaces: the Micromorphology

Spot and block samples were collected for a sediments. These examinations in the field will be
range of complementary analyses including Scan- verified by analysis of large resin impregnated block
ning Electron Microscopy coupled with an Energy samples in thin section and samples under SEM with
Dispersive X-Ray Analyzer, chemical, physical and EDXRA.
mineralogical analyses, and phytoliths. Blocks of mud The deposits in thin sections are described us-
brick adjacent to thin-section samples have been sub- ing internationally standardized terminology and
ject to optically stimulated luminescence dating (see procedures which are based on morphological at-
Chapter 16). Samples for pollen analysis were exam- tributes, rather than genetic classifications, and can
ined by W. Eastwood, University of Wales (see be- therefore be applied to a wide range of natural and
low). Samples for a range of interdisciplinary analyses anthropogenic deposits (Bullock et al. 1985; Courty
were collected at 50 em intervals across floors dur- et al. 1989). The significance of each of these attributes
ing the first season of renewed excavations in 1995 to archaeological questions and investigations is dis-
(Matthews n.d.). Integrated organic and inorganic cussed in Matthews (1996). With the help of T.
analyses will be conducted in the near future by Dr Ritchey, we have designed one of the first relational
R.P. Evershed at the University of Bristol and Dr data bases for thin-section description, using
D.A. Jenkins, University of Wales. Microsoft Access. These descriptions are recorded in
All micromorphological scales of analytical fo- the interdisciplinary networked data base on the
cus are inter-related. The thin sections are first ~atalhoyiik workstation and will be accompanied
analyzed macroscopically in the hand and correlated by images of the main deposit-types. This data base
with the microstratigraphic photographs and 1:5 sec- is still being added to at the time of writing this
tion drawings and descriptions. They are then exam- report, due to the complexity of the ~atalhoyiik
ined at low magnifications of 5.8x-70x using a Wild microstratigraphic sequences where there may be as
M420 large field transmitted light macroscope with many as 75 layers of plaster and occupation deposits
Apozoom, before analysis at higher magnifications in each thin section or 160 layers of wall plaster, and
of 40x-600x using a Nikon Optiphot optical polariz- the fact that the NERC project is still in progress. The
ing microscope. The deposits are examined in plane- aim of this report is to illustrate the range of varia-
polarized, cross-polarized and reflected light at both tion in deposits and microstratigraphic sequences at
low and high magnifications. The natural autofluor- ~atalhoyiik. The detailed statistics comparing these
escence of organic and inorganic components and deposit types will be published in a later report. The
fine fabrics is being studied in reflected fluorescent application of image analysis and computerized cal-
blue light at magnifications of 40x-600x, excitation culation of the size, shape and abundance of the
wavelength 480 nm, with an excitation filter -450- heterogeneous components in anthropogenic depos-
490, dichroic mirror -510 and barrier filter -520. The its is currently problematic, but a range of applica-
autofluorescence of components in blue light varies tions are being developed by Professor D.A.
considerably from low excitation colours of dark Davidson, University of Stirling, and Dr T. Acott,
brown to high excitation bright green and yellow. University of Greenwich.
Although some excitation colours are characteristic There are four principal elements or stages in
of specific compounds, and higher excitation col- thin-section analysis of occupation sequences and
ours are associated with fresher organic remains or evidence for uses of space. These entail analysis of:
higher phosphorus content (Courty et al. 1989, 48- 1. the materials and technology of floors or surfaces;
50), specific correlations with material substances 2. the impact of activities on floors or surfaces;
and diagenesis and decay have not been fully explored. 3. the nature and taphonomy of component arte-
In future excavation seasons, techniques for facts, organic remains and sediments in occupa-
analyzing the micromorphology of deposits as they tion deposits, and their contextual relationships;
are being excavated in the field will be explored and 4. post-depositional alterations.
developed. This will include microscopic analysis of Interpretation of deposits and microstrati-
deposits on latex peels, in small blocks of micro- graphic sequences in the field and in thin section is
stratigraphic sequences, and spot samples in a mount- based on:
ing medium on glass slides. This will enable instant 1. analysis of the type, morphology and frequency
assessment of the type, size and abundance of of components and deposits in each sequence;
depositional components, and in the case of peels 2. contextual and comparative analysis of each se-
and blocks, some information on the contextual rela- quence and its location within different buildings
tionships between artefacts, organic remains and and areas of the settlement;

303
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

3. comparison with other archaeological informa- by a relatively homogeneously pugged fabric which
tion from the same and comparative contexts has an embedded (porphyric) related distribution
both within the settlement and in the surround- and a microstructure with 2-10 per cent pseudo-
ing environment by other analyses of artefacts morphic voids from plant remains which have since
and organic and inorganic remains and sediments; decayed. Siliceous plant remains adhering to the
4. comparison with thin section samples from other walls of these voids, and charred remains in burnt
archaeological sites and ethnoarchaeological and floors, suggest that these plant remains principally
experimental studies of settlements, soils and comprise Gramineae stem, leaf sheath and leaf frag-
plants, for example; ments. These are likely to have been deliberately
5. dialectic working hypotheses based on contex- added as stabilizers to the plaster, in order to in-
tual and comparative analysis of the above sets crease workability, tensile strength and cohesion,
of information and theories relating to activities thereby reducing cracking during drying and use
and depositional agencies and processes in set- (Norton 1986,32). Some floor plasters include a range
tlements and environments from other studies of anthropogenic debris such as heterogeneous burnt
in archaeology, ethnoarchaeology, soil science, aggregates, charred plant remains and fragments of
geology, botany and zoology, for example. bone.
One of the most distinctive materials used to
Results plaster the floors is a white calcareous silty clay (Pho-
tographs 15.10-15.12). In both the field and thin sec-
Materials and technology of floors and surfaces tion these plasters range in colour from slightly
A range of materials were used in the manufacture greenish white (2.5Y 8/2) to creamy white (10YR8/
of deliberately prepared 'mud' plaster floors at 2). In thin section, these sediments closely resemble
<::atalhoyiik. Identification of the sources of some of modern samples of soft lime deposits which com-
these materials is pending further collaborative prise up to 95 per cent pure carbonates of calcium
palaeoecological research. Modern soil maps are and magnesium, and are unconsolidated Neogene-
available for the area (Driessen & de Meester 1969; Quaternary carbonate deposits (Photograph 15.3;
de Meester 1971), but palaeoecological research has Driessen & de Meester 1969, 42-6; de Meester 1971,
established that the distribution of sediments and 57). These modern soft lime deposits today occur
soils in the region differed during the early Holocene within 5 km of the site and have been classified as
(Roberts Chapter 2). The mineralogy is related to the the Kasmham soil series to the west, and the Urii.nlii
geology in and around the great Konya Basin and series to the southeast. The modern deposits range
the <::ar~amba <::ay river catchment. The basin floor in colour from white (5Y8/1) to white and creamy
was covered with shallow calcareous Pleistocene lake white (10YR8/2). They have a clay loam particle size
deposits and beach ridges. It is surrounded by Up- and are very sticky and plastic. These deposits are
per Cretaceous limestone to the south in the Taurus locally known as ak toprak, and are used by modern
mountains, Palaeozoic limestone and schists in the villagers in the region for plastering the floors and
Anatolides mountains to the north and west, and walls, by mixing with water (Mellaart 1962, 48;
outcrops of Upper Miocene-Holocene volcanic ma- Nurcan Yalman pers. comm.). It has already been
terial in localiszd uplands. suggested that these deposits are likely to have been
In thin section it is evident that a range of de- used during the Neolithic for plasters (Mellaart 1962,
posits including alluvial and lake-derived deposits 48). The plaster samples analyzed in thin section
were used in the manufacture of plaster floors, rang- include sparse shell fragments and, pending investi-
ing in particle size from a silty clay to a medium- gations by the conservators and SEM with EDXRA,
coarse sandy silt loam (Photographs 15.8-15.11). do not appear to have been prepared by firing lime
Many silt loam sediments which are greyish brown to temperatures of 850-900 C (Moorey 1994), in con-
in colour include 10-20 per cent finely fragmented trast to samples analyzed from a number of Neolithic
plant remains which are dark pigmented and often sites in the Near East, including one sample from
are less than 15-50 f.!m in size. Floor plasters manu- <::atalhoyiik (Kingery et al. 1988). There have been
factured from these deposits are dark grey (10YR 4/ some difficulties in distinguishing between organi-
1) in the field, and, at less than 3-10 mm thick, could cally bonded lime plaster and burnt lime plasters
be mistakenly identified as thin layers of occupation particularly in weathered samples (Affonso 1996).
debris macroscopically in the field. Similar white calcareous silty clay sediments
In thin section floor plasters are characterized were used to plaster the walls at <::atalhoyiik. One

304
1014 E w 1014

~
s......
-u·
ro-
(f)
~
....
PI'
(")
'' : 93.10
. J l'1l
rn
(J.)
0
g.
Vl l'1l
-dl:-------=
~
1013.50 1013.50 ;:;·
....
Scale at 50cm intervals 0
3
0
....
'"0
:::>'
0
Figure 15.1. Microstratigraphy of two fire-installations in section 2, level V, south of V.61 (Fig. 14.3). 0
OQ
'<
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

fragment of collapsed wall plaster from Area E has to underlying deposits which are not affected by the
been analyzed in thin section (Photograph 15.4). This impact of activities. The reactive zone refers to un-
fragment is only 3 em thick, but includes 160 layers derlying deposits which are slightly altered by the
of plaster. In thin section it is evident that the se- impact of activities, exhibiting cracks or compaction,
quence comprises 80 couplets of plaster consisting for example. The active zone refers to the interface
of a preparation layer of very pale brown calcareous between surfaces and activities, where both floor
silty clay and a finishing coat of finer white calcare- and occupation deposits may be reworked by tram-
ous silty clay. The primary preparatory layer [1] is pling for example.
4-6 mm thick, all subsequent preparatory layers are Charred and siliceous remains of mats were
only 24-720 f..lm thick. The finishing coats [2] are 12- uncovered on floors during excavations at <;atal-
552 f..lm thick. The preparatory layers were stabilized hoyiik in the 1960s (Mellaart 1967, pl. 120). Impres-
with the addition of 2 per cent stabilizers, and had sions probably from mats or floor coverings have
slightly irregularly textured surfaces on to which the been identified on the surfaces of some plaster floors
finer finishing coats of white calcareous plaster were in thin section at <;atalhoyiik. Microlayers of silty
keyed in. The surfaces of the finishing coats by con- dust immediately above some plaster floors may
trast were very smooth and compacted, indicating have percolated through floor coverings and have been
that they may have been deliberately smoothed as subsequently compacted. Trampled surfaces with
Mellaart suggested, perhaps using white or green sub-horizontal cracks in the reactive zone and dis-
polishers which were found extensively at the site lodged aggregates of plaster in the active zone occur
(Mellaart 1962, 48). The materials and technology in sequences of domestic deposits in sections 3 and 4
employed throughout this entire sequence of wall (Photographs 15.8-15.9 [1]). Pale yellow salts had also
plasters are remarkably consistent. impregnated the surface of the plaster floor in section 3
Five blocks of mud brick and mortar have been (Photograph 15.8 [1]). Further ethnoarchaeological
sampled for analysis in thin section, from levels XII, and experimental research and sampling is required
VII, VI, and V. The source materials in mud-brick in examination of the impact of activities on surfaces.
samples from levels XII, VII and VI comprise silt
loam to sandy silt loam sediments with finely frag- Occupation deposits
mented dark pigmented and charred plant remains In thin section we are studying the type, taphonomy
and some fragments of bone and burnt aggregates and contextual relationships of artefacts, organic re-
(e.g. Photograph 15.5 [1]). The source materials in mains and sediments in each depositional unit. The
level V comprise alluvial and alluvial fan deposits buildings at <;atalhoyiik were kept remarkably clean.
with medium to coarse sandy clay loam to fine- Many of the layers of occupation deposits on floors
medium sandy silt loam particle sizes. There tends and surfaces are less than 1 mm thick, ranging from
to be fewer voids from plant remains in mud bricks some floors on which there are not even any micro-
than floor plasters, at <2-5 per cent. Mortars gener- scopic residues, to 240-800 f..lm and, in exceptional
ally include higher concentrations of heterogeneous cases, 2.9 em thick. Levelling debris in Shrine VIII.25
anthropogenic debris than mud bricks, including was up to 4.5 em thick between plaster floors. Open
charred plant remains, burnt bone, and occasional areas used as middens by contrast have several me-
flakes of obsidian (Photograph 15.5 [2]). Similar mor- tres of accumulated refuse (see Chapter 14).
tars rich in anthropogenic debris were used four thou-
sand years later at the site of Abu Salabikh in southern Plant remains
Iraq, c. 2500 Be (Matthews & Postgate 1994, fig. 15.7). In thin section it is evident that plant remains consti-
The plaster linings of fire-installations in sec- tute the largest class of residue from human activi-
tion 2 (Fig. 15.1) are generally made from well-sorted ties, both within buildings and in open areas at
fine and medium sandy clay loam alluvial deposits <;atalhoyiik, comprising more than 40 per cent of a
(Photograph 15.6). range of deposits. In the current three-year NERC
project we have been very fortunate to collaborate
Impact of activities on floors and surfaces with Mr T. Lawrence and Dr D.F. Cutler at the Plant
Few of the floor surfaces sampled in thin section Anatomy section of the Jodrell Laboratory, Royal
have boundaries as smooth as those of the wall plas- Botanic Gardens, Kew, who have collaborated in
ters. Different zones of impact of activities on floors other archaeological projects (Gale & Cutler forth-
and surfaces have been identified in micromorpho- coming). Plant remains in thin sections from all four
logical studies (Ge et al. 1993). The passive zone refers sites in the NERC project have been identified by com-

306
Multiple Surfaces: the Micromorphology

parison with the extensive reference collection of A wide range of plant types have been identi-
plant anatomical microscope slides and library with fied and characterized in the micromorphological
a wide range of plant anatomy atlases at Kew. thin sections. Identifications in the future will be
The variety of plant materials and anatomical aided by analyses of charred wood and plant re-
elements preserved in their depositional context in mains retrieved by full-scale excavation and wet-
the thin-section samples from <::atalhoyiik, is ena- sieving and flotation, as these are three-dimensional
bling study of a range of human-plant relationships specimens which can be fractured and observed with
at a critical stage in the development of agriculture surface illumination or sectioned with a microtome.
and complex settlements. A range of root, ?tuber, These analyses will provide larger samples and sta-
stern, leaf, leaf sheath, husk, and seed/fruit anatomi- tistics to complement analysis of rnicrostratigraphic
cal elements have been identified. These are vari- associations. Further modern reference collections
ously preserved as desiccated uncharred remains, need to be built up of woods and other plant materi-
siliceous remains (Photograph 15.13), charred re- als in the region.
mains (Photograph 15.7), calcareous ashes, and Interpretation of these rich plant remains re-
pseudornorphic voids in sediments surrounding quires ongoing ethnobotanical research, like that con-
plants which have since decayed (Photographs 15.10- ducted by Fiisiin Ertug, Washington University, at
15.11). The charred plants range in size from several the acerarnic Neolithic site of A:;;tkh Hoyiik and sur-
microns to more than 2.6 ern. Identification, how- rounding modern villages. A wide range of wild
ever, is not necessarily dependent on size, but on plant leaves, sterns, seeds, tubers, etc. are still used
whether diagnostic characteristics are present and in by modern villagers (Ertug pers. cornrn.). The chal-
the correct orientation for secure identification. Plant lenge is establishing how these might survive and be
anatomy identifications are usually based on obser- identified in the archaeological record (Hall 1987;
vations of a given plant organ viewed in two or Evershed et al. 1992).
three opposing planes. For example in leaves of The charred plant remains identified at C::atal-
grasses surface views of the epidermis, combined hoyiik include a wide range of anatomical elements
with transverse sections, are usually required to at- including: dicotyledonous and coniferous woods;
tempt an identification. In the case of woods of di- monocotyledon root, stern, leaf sheath and leaf; and
cotyledons and conifers transverse, tangential seeds and fruits which are listed below:
longitudinal and radial longitudinal sections need to
Wood
be examined in order to see all of the potential diag-
Celtis/Ulmus (hackberry I elm)
nostic characters (Cutler 1978). The orientation of Chenopodiaceae- similar to Sa/sola, Sueda etc.
plant remains in micromorphological thin sections is Pistacia/Ulmus/Celtis (pistachio/ elm/hackberry)
by contrast dependent on the way in which the plant Quercus (deciduous white oak type)
Salix/Pop/us (willow/poplar)
remains were deposited and have settled due to grav-
ity. Although only one view is available for each Unidentified coniferous wood
fragment of plant, there are often several fragments Unidentified dicotyledonous woods with characters as follows
of the same type of plant within any depositional (apparently all different species):
unit. If these fragments are in several different 1. bands of early wood vessels containing tyloses
orientations, they can provide a composite view of a 2. moderately large vessels
3. rays rather tall 4-5 cells wide
particular plant. Woody material is usually more 4. rays mostly 6 cells wide and relatively short
easily identified than herbaceous material because 5. semi ring porous, small- to medium-sized vessels and
there is a limited number of woody plants present in medium width rays ?Prunus
the region, which are fairly well represented in the 6. very thin walled cells
7. very small vessels
Kew reference collection. The anatomical reference 8. very small vesels in radial chains, similar to Reaumuria
material of the large range of native herbaceous plants (Tamaricaceae)
is not as comprehensive. Woods can, however, still 9. very rounded medium-sized vessels
present difficulties because botanically closely re- 10. wide rays ?Balanites/Tamarix
11. diffuse porous woody herbacious perennial
lated woods can be anatomically very similar and
very difficult to distinguish from one another, such Stem, leaf sheath and leaf
as Salix from Populus (Salicaceae) and Ulmus from Gramineae (grasses)
including 'reed': probable Phragmites
Celtis (Ulmaceae). Even Pistacia can be very difficult also Arundo donax
to separate from the unrelated Ulmus/Celtis in ob-
lique sections.

307
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

Cyperaceae - triquetrous stem We have compiled a basic series of morpho-


Unidentified monocotyledonous stems and leaves with charac- logical types of articulated siliceous plant remains.
ters as follows (apparently all represent different species): To date 16 morphological types, some with sub-types,
have been drawn up. Siliceous remains of cereal-like
1. large fibrous bundle sheaths, probably not a grass
2. large parenchymetous bundle sheath cells
epidermises have been identified. As yet, it is still
3. one midrib bundle, two lateral air spaces and square margin not possible incontrovertibly to distinguish morpho-
with enlarged cells at corners logically between wild and cultivated grasses. We
4. longitudinal ridges formed of thick walled cells have been able to identify siliceous and ?non-sili-
5. stem -lacking any distinguishing features
ceous fragments of the reed Phragmites and also
Root Arundo donax which are abundant in many deposits.
monocotyledon root The term 'reed' is used in this report to refer to
coarse Gramineae stems and leaf sheaths. Conical
Seeds and fruits
cereal grains, including barley
silica bodies possibly from Cyperaceae are also
unidentified fruits and seeds present.
Some of the siliceous plant remains have melted
A number of specimens of uncharred wood and plant in temperatures greater than 600-800° C, and have
remains have been identified. These include: formed vesicular lengths and globules of silica, which
Quercus (oak) are <0.5 mm in size.
dicotyledonous wood with rays mostly 4 cells wide The calcareous ashes at c;:atalhoyiik are often
Celtis pericarp lozenge-shaped, c. 30 J!m in size. Micromorphologists
seeds type 4.1 and 4.2 are developing identification of calcareous ashes us-
tuber containing possible starch
ing SEM with EDXRA on samples from experimen-
tal combustion, and compiling reference collections
In addition a range of well-preserved siliceous and (Wattez & Courty 1987). The lozenge-shaped ashes
non-siliceous plant remains are observable in thin at c;:atalhoyiik may resemble deciduous wood ash
section, often articulated. These include leaf and stem crystals (Wattez & Courty 1987, figs. 1-2; Courty et
epidermises with stomata, papillae, prickle hairs, long al. 1989, fig. 7.1a).
cells and short cells; parenchyma, vascular bundles, Nine samples of a variety of deposits within
bulliform cells; culm nodes, husks and awns (Photo- the settlement at c;:atalhoyiik were examined for pol-
graph 15.13). It has proven difficult to identify these len content by Warren Eastwood (University of
siliceous plant remains despite their good preserva- Wales, Aberystwyth). A range of Monocotyledon,
tion and articulation. Many of these are from mono- Dicotyledon and conifer pollen grains were identified
cotyledons, especially grasses which may exhibit (Table 15.1), but pollen concentrations are extremely
considerable variation in the anatomy between dif- low. Pollen grains are also sparse in thin-section
ferent parts of the same plant, but similarities in the samples. The significance of these grains is not
anatomy between different species. Many sheets of easily assessed for the following reasons. The pollen
siliceous plant remains in thin section lie in longitu- grains may have been (a) originally present in the
dinal, transverse or oblique sections, due to the ef- alluvial or lake deposits from which bricks or plas-
fects of gravity. These orientations usually possess ters were made, or (b) incorporated into the brick or
fewer diagnostic characters than surface views of plaster at the time of manufacture, or (c) incorpo-
epidermises, which occur less frequently. The study rated during the formation of occupation deposits,
of siliceous plant remains in thin section needs to be or (d) introduced later by soil fauna. Pine pollen is
complemented with analysis of articulated ancient widely dispersed, so its presence cannot be regarded
specimens extracted both from small bulk samples as an indication of the local presence of pine trees.
and microstratigraphic blocks which can be handled The other pollen taxa are also well dispersed but
and oriented during analysis. During the course of may be better indicators of local sources, including
the c;:atalhoyiik project, identification of all of these derivation through plant materials brought onto the
sample types will be aided by the compilation of a site (e.g. matting and roofing). The pollen assem-
reference collection of Gramineae and analysis of the blage may also have been affected by differential
range of agents and mechanisms by which they are preservation, as suggested by the presence of
likely to have been introduced to the site, including Compositae, which are generally decay-resistant. In
dung. A reference collection of phytoliths from plants short, these pollen data allow no firm conclusions
in the Levant is being established by Rosen (1992). regarding the environment or activities at the time
of occupation.
308
Multiple Surfaces: the Micromorphology

Table 15.1. Pollen analyses conducted by W. Eastwood at the Palynological Research Centre, Institute of Earth Studies, University of Wales,
Aberystwyth.
c;:atalhiiyiik pollen appraisal report

Archaeological Section Deposit type Palynological Volume Microscope slide pollen appraisal
sample no. no. ref. no. processed
29 Occupation C::HEARC.93-I 1 cm3 I Quercus undiff. (T); I Chenopodiaceae undiff. (H); I Compositae,
deposit Liguliforae (H); Compositae, Tubuliflorae (H); 19 Gramineae (H); I unknown;
4 indeterminate.
30 3 Occupation C::HEARC.93-2 0.4cm3 I Compositae, Liguliflorae (H); 5 Gramineae (H); 3 indeterminate.
deposit
31 3 Open area <;:HEARC. 93-3 lcm3 Pinus fragment (T); 2 Chenopodiaceae undiff. (H); I Compositae,
deposit Tubuliflorae (H); 2 Gramineae (H); 48 Sparganium type cf. (A).
32 Mud brick C::HEARC.93-4 lcm 3 2 Pinus undiff. (T); I Chenopodiaceae undiff. (H); I Compositae,
Tubuliflorae (H); 2 Gramineae (H); I indeterminate.
33 4 Occupation C::HEARC. 93-5 0.15cm3 1 indeterminate.
deposit
34 4 Occupation C::HEARC.93-6 1 cm 3 31/2 Pinus undiff (T); 3 Compositae, Liguliflorae (H); 5 Compositae,
deposit Tubuliflorae (H); II Gramineae (H); 5 Sparganium type cf. (A); 3 indeterminate;
1 unknown.
35 2 Open area C::HEARC.93-7 1 cm 3 No pollen.
deposit
36 2 Open area C::HEARC. 93-8 1 cm3 I Pinus undiff (T); I Chenopodiacea undiff. (H); 2 Gramineae (H).
deposit
37 4 Floor C::HEARC.93-9 0.8cm' I Pinus undiff (T); I Chenopodiaceae undiff (H); I Compositae,
Liguliflorae (H); 4 Gramineae (H)

Note: Indeterminate includes grains which are crumpled, broken or obscured.


T: Tree
H: Herb
A: Aquatic

nivorous animals such as pigs for example has dense


Dung and coprolites amorphous organic matter and may contain frag-
A range of dung pellets, digested plant remains, and ments of phytoliths and bone. The nature of dung
coprolites have been identified at <;atalhoyiik in a from specific species varies according to age, diet
variety of contexts including rooms, open areas and and season, and with careful study and reference
fire-installations. Modern reference material of burnt collections may provide information on animal be-
and unburnt dung from a range of animals and sta- haviour, relationships with humans, and ecology
ble deposits has been studied in thin section and (Anderson 1995; Brochier 1992; Stewart & Stewart
compared to ancient dung from different regions of 1970) .
the world, including the Konya plain as part of the At least three principal types of dung/ coprolite
current NERC project in collaboration with S.M. have been identified at <;atalhoyiik. Types 1 and 2
Anderson (Anderson 1995), and from <;ayonii Tepesi are characteristic of herbivore dung, type 3 resem-
in Turkey (Factor, J.L., 1991. B.A. Thesis, Harvard bles omnivore coprolites. Type 1 has more readily
University, cited in Brochier 1993, 47), in Iraq recognizable plant remains (Photographs 15.18-
(Matthews & Postgate 1994), Europe (Courty et al. 15.19), than those in type 2. Extant pellet fragments
1991; Davidson et al. 1992; Canti in press) and India of type 1 are up to 13 mm in length. Inclusions are
(Courty pers. comm.). Dung has also been identified unoriented and randomly distributed, and princi-
by comparison of components in bulk samples of pally comprise abundant digested siliceous plant re-
modern and ancient dung (Brochier 1985; 1992; 1993). mains, often of Gramineae, occasionally of tubers,
Dung from modern species in thin section can and abundant spherulites discussed below (pp. 322-3).
be distinguished on the basis of differences in the Type 2 has a fine fibrous structure, and includes
type, nature and size of components and their struc- finely fragmented and digested plant remains (Pho-
tural arrangement, and the external morphology of tograph 15.24). This dung type does not contain
dung pellets (Courty et al. 1991, table 1). Dung from spherulites. Extant pellet fragments are up to 7.7
cattle may be strongly layered and contain coarser mm long and 1.05-4.2 mm in height, and are ovoid
fragments than dung from sheep and goats which in shape, and often compacted. These dung frag-
tends to be more finely fragmented. Dung from om- ments resemble sheep/goat droppings in stabling

309
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

layers in the Early Middle Neolithic cave site of Arene charred wood Quercus (oak)
charred wood Celtis/Ulmus (hackberry I elm)
Candide, Finale Liguria, Italy (Courty et al. 1991, siliceous stem, leaf Gramineae (grasses, including
figs. 4-5). 'reeds', i.e. Phragmites, or
Type 3 comprises dense pale yellow amorphous Arundo donax)
organic matter and includes pseuodomorphic voids of siliceous dung pellet ?sheep I goat
calcareous ashes, lozenge ?deciduous wood
plant remains, fragments of phytoliths
<150 J..Lm, and bone <5 mm (Photograph 15.25) and The variety of types and range in calorific values of
is occasionally associated with hackberry pericarps <2.8 plant remains at <::atalhoyiik is much greater than at
mm. Extant fragments are up to 21 mm in size. The Abu Salabikh. The larger differences in calorific value
yellowish phosphatic organic matter is similar to that and nature and rates of combustion of different plants
in a human coprolite from Middle Saxon levels at at C::atalhoyiik are likely to produce a much greater
Maiden Lane, London, eighth century AD (Courty et al. range of juxtaposed burnt plant remains than at Abu
1989, pl. IVa). The organic aggregates from C::atalhOyiik Salabikh. The calorific values of oak, hackberry and
have less dark brown amorphous organic staining, and elm woods for example are much higher than those of
may possibly be derived alternatively from pig or dog Gramineae and dung. Further analyis of ethnoarch-
coprolites, or ?food residues. aeological and experimental burning conditions and
Positive identification of any of these dung re- remains is required.
mains must await compilation of a wider reference Within buildings and open areas at <::atalhoyiik,
collection, and correlation of the physical character- a number of contextual variations in plant remains
istics in thin section and under SEM with chemical and dung are emerging. These include variations
signatures, such as traces of bile acids, sterols and with regard to plant materials, anatomy, size and
coprostanols, which, if present, can be detected by abundance. The following table is a selective
gas chromatography /mass spectrometry, and in list of key plant remains which co-occur in at least
many instances are specific to, and therefore diag- two different examples of the same context type, and
nostic of, different animal species (Bethel et al. 1994; appear therefore to be characteristic of that type of
Evershed & Bethel in press). Dung which has been context. Other plant remains may be present in these
disaggregated or transformed by trampling or contexts, but are generally less abundant, and some-
burning for example may be more difficult to iden- times more ubiquitous. The preliminary list in Table
tify. 15.2 is presented here in order to illustrate the nature
of the results which are emerging. The contextual
Contextual variation in plant remains and dung correlations are discussed with reference to other
Contextual variation in plant remains was identified variations in microstratigraphic sequences in the fi-
in thin-section samples from the early urban settle- nal section of this chapter, and will be further tested
ment of Abu Salabikh in Iraq (Matthews & Postgate once the data base is complete.
1994). In particular, a strong correlation was observed
between types of plant remains and different con- Other components in occupation deposits
texts and regimes of combustion. The highest con- Other components in occupation sequences in thin
centrations of charred plant remains occurred in section include fragments of obsidian (Photograph
enclosed cylindrical walled clay ovens where condi- 15.26), grindstones, red ochre (Photograph 15.12),
tions are likely to have been reducing. Layers of and unburnt and burnt bone (Photographs 15.9 &
siliceous plant remains and ashes were abundant in 15.13), shell, diatoms, and a wide range of aggre-
open semi-circular hearths, where most of the car- gates from different building materials and natural
bon had been oxidized. Calcareous ashes were par- sources (Photographs 15.8, 15.12 & 15.14).
ticularly abundant in large ventilated chambered Obsidian fragments are surprisingly sparse in
ovens. Four fire-installations have been sampled at the samples analyzed. One fragment of obsidian, 2.45
<::atalhoyiik. The two installations in section 2, had mm in size, perhaps significantly occurs in one of the
been thoroughly swept out before replastering (Pho- refuse deposits with ?wood chippings and 'reed' re-
tographs 15.2 & 15.6), leaving only isolated aggre- mains (Photograph 15.26). Obsidian piece esquillees
gates of in situ fuel, less than 2 mm thick. A much are one of the most common tool types at C::atalhoyiik
thicker layer of fuel and burnt remains, 3.5 em thick, and are likely to have been one of the principal tools
survives in situ on the base of the fire-installation in used for wood working (Chapter 10). A wide range
Shrine VIII.25, phase 4. A surprisingly wide range of of sophisticated wooden vessels were recovered from
plant remains are present, these include: the site during the 1960s' excavation, preserved by

310
Multiple Surfaces: the Micromorphology

gates and 20-30 per cent


Table 15.2. Contextual variation in types of plant remains and dung.
charred reed and wood <1
Context type Deposit type Plant remains type Size range mm mm, in a discrete layer 0.2-
Room - few white floor plasters occupation deposits charred wood <2.1 4.9 mm thick, between layers
(Photographs 15.7-15.8! siliceous Gramineae including: <7.0
siliceous cereal-like fragments <1.0 with >40 per cent charred
'reed' and wood. Red ochre
Room- white floor plasters occupation deposits charred 'reed'-like fragments <0.2
(Photograph 15.15! also occurs as isolated single
?Stable accumulated deposits digested siliceous Grarnineae <1.5 grains in floor make-up and
(Photographs 15.17-15.19) dung types 1 and 2 <13.0 occupation deposits in a
spherulites 0.005-0.020
range of contexts.
Open areas dumped single unit
(Photograph 15.25! (?burnt wood chips charred oak <10.0
+ ?basketry I matting
manufacture residues
charred 'reed'
obsidian
Deposition of components in
<0.5

?used as fuel) occupation deposits


dumped single unit
Differences between depos-
(?food cooking) cereal grains, incl. barley <4.2 its which accumulated in situ
burnt aggregates of
fire-installation plaster <12.0 and those which had been
dumped single unit hackberry stones <2.8
dumped were observed at
dung type 3, omnivorous <21.0 Abu Salabikh (Matthews &
Postgate 1994, 190). Attributes
characteristic of trampling,
charring during extensive burning in level VIA in sweeping, and successive use and shaking of mats
particular (Mellaart 1964, 85-92). At least ten obsid- or rugs were also identified. A range of similar dif-
ian fragments, 0.25-5.5 mm in size, occur in a thin ferences is emerging at (atalhoyiik. Many of the
lens, 0.3-2.1 mm thick, in refuse deposits in an open thin lenses of occupation deposits at (atalhoyiik,
area in level XI Court 25. This lens clearly represents <0.2-1.0 mm, are compacted with a massive micro-
a discrete depositional episode, between layers of structure and an embedded (porphyric) related dis-
charred 'reed' and wood, and includes fragments of tribution (Photograph 15.16). By contrast, thicker
bone <2 mm and 5-10 per cent heterogeneous aggre- layers of accumulated deposits are less compacted
gates. The charred wood includes fragments of and have a complex packing void microstructure,
Quercus, Salicaceae and Pistacia/Celtis/Ulmus. and inter-grain aggregate (enaulic) or bridged
Red ochre has been identified in significant (gefuric) related distribution between coarse and fine
quantities in three contexts. The first context is as components (Photographs 15.8-15.9). Microstrati-
two lenses of red ochre in the top of a small hole dug graphic lenses occur within some layers of occupa-
into the top of large grave in section 3 (Photograph tion deposits. Steeply sloping deposits in open areas
15.12). These lenses of red ochre are 20-120 J..Lm and have been sorted by gravity.
24-360 J..Lm thick, and occur between two thin layers Spatial and cyclical variation in deposition in
of white plaster, which were subsequently capped ritual contexts has been observed and studied in the
by a thicker layer of white plaster. The ochre is bright temple at Saar, Bahrain (Matthews et al. in press).
salmon pink red in reflected light, very similar to The plastered surfaces on top of the altar were regu-
some of the red paints used in the wall paintings. It larly cleaned after periodic burning at Saar. Layers
is very dark reddish brown in colour in plane- of deposits on top of the altar are very thin, <200 J..Lm,
polarized light with high relief, and isotropic in cross- and principally comprise fragments of burnt date
polarized light. palm. Thicker layers of burnt date palm accumu-
The second context in which red ochre occurs is lated around the bases of the altars. Clean plaster
in the upper 5 mm of 3.7 em of in situ burnt deposits floors and layers of sandy levelling deposits built up
in the fire-installation of Phase 4 in Shrine VIII.25. away from the altars. The back rooms of the temple
The red ochre occurs as a cluster 1.44 mm in diameter were almost certainly used for storage, and had thick
and as aggregates <250 J..Lm in size comprising 2 per layers of packing and plastered scoops or bases for
cent of the deposit. The third occurrence is as discon- stored items.
tinuous micro lenses in a thin layer of refuse deposits
in an open area in level XI Court 25. The lenses of red Post-depositional alterations
ochre are less than 0.48 mm thick and 7 mm in length, There is evidence to suggest floors and occupation
and are deposited with heterogeneous mineral aggre- deposits may in some instances have been removed

311
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

from within buildings, resulting in a truncated hori- ganic diagensis and decay. These remains require
zon. The juncture between wall plasters and floor further specialist analysis both in thin sections and
plasters has been examined both in the field and in spot and bulk samples. Amorphous organic staining
thin section, in order to identify any differences in and organic aggregates are preserved in a range of
the frequency of replastering of floors and walls, contexts (e.g. Photograph 15.14), and exhibit bright
and to investigate whether there is any evidence for yellowish autofluorescent excitation colours in blue
truncation and removal of floors. There is no conclu- light, suggesting that organic residues and phospho-
sive evidence within the current sample of five rus may be present.
buildings with which to establish whether any
unconformities between the wall plasters and floor Contextual variation in microstratigraphic
plasters are the result of the irregularities in the wall sequences
face and the generally less frequent replastering of
floors, or the actual removal of floors. This problem There are marked differences between microstrati-
in stratigraphic interpretation is compounded by the graphic sequences within buildings and in open ar-
practice of raising some floor levels by up to 60 em eas (Figs. 15.9-15.10). No deliberately prepared floors
as in level VIB to VIA in Shrine 10 (Mellaart 1964, or surfaces have yet been identified in any open
47-50), or reusing the tops of earlier walls as the base areas. This is in contrast to within buildings where
of later walls (Chapter 14). In both of these instances up to 99 per cent of sequences may comprise deliber-
the stratigraphic sequence at the juncture with the ately laid floors, and only 1 per cent may comprise
later floors would show multiple layers of wall plas- accumulated deposits. Accumulated deposits within
ter for which there are no corresponding floors, until open areas may be several metres thick (Chapter 14).
underlying levels and sequences have been exposed. By contrast, the floors within buildings were kept
In the large area of deep refuse deposits in field extremely clean. In some instances there are not even
section 3, one depositional layer includes large re- any microscopic remains between successive layers
deposited fragments of floor plaster comprising of plaster, as in the case of many of the floors in
multiple layers of orange and white plasters. In one samples from sections 7 and 9 for example. Where
thin section of floors from the unexcavated building there are accumulations of occupation deposits be-
east of building 32, level VII, there is an isolated tween plaster floors they range in thickness from 0.1
island of floors 4.2-5 mm in height, where at least mm to 2-3 em. Many deposits are less than 0.5 mm
two layers of slightly sandy silt loam plaster, each thick. One exception to this is an accumulation of
with at least two multiple layers of white plaster heterogeneous aggregates and organic-rich deposits
finishing coat, have been removed from the rest of in a relatively inaccessible area to the east of fire-
this sequence. One informant during ethnoarch- installations in Shrine VIII.25, which has an extant
aeological research suggested that if the sequence of depth of 37 em.
floors within a building becomes too high, the floors
would be dug out and new ones laid (Nurcan Yalman Variation in microstratigraphic sequences within
pers. comm.). buildings
Many of the deposits at <::atalhoyiik are well The total depth of each of the microstratigraphic
preserved with little post-depositional disturbance. sequences within the buildings studied ranges from
There has been some bioturbation from root and 5.5-37 em (Fig. 15.2). With the exception of Shrine
insect activity, represented by channels and cham- VIII.25, each entire sequence of plasters and occupa-
bers, which in many instances have been infilled by tion deposits on floors and platforms is less than 10
disturbed sediments or soil faunal excrement. Sparse em deep, and can be studied in one large thin sec-
modern root fragments have been observed in chan- tion, 13.5 x 6.5 em.
nels in thin section, particularly in field section 6, The horizontal and sequential application of
where there are modern bushes on top of the section. plaster on platforms, floors and low dividing ridges
Reprecipitation of salts has resulted in local distur- is complex. The layers of plaster on the platform in
bance of the fabric of some deposits, including the field section 9 are thinner and were more frequently
plaster floor surface in field section 3 (Photograph applied than the thicker layers of plaster on the adja-
15.8). Relics of ancient microfauna! eggs, ?parasites, cent floor (Fig. 15.3). Eleven coats of plaster each
spores, microbial filaments and bacteria are present, with finishing coats of plaster were applied on the
giving some indication of pre- and post-burial con- platform, but only three layers of plaster were laid
ditions in the built environment, and agents of or- on the floor. The low ridges in field sections 3 and 6

312
Multiple Surfaces: the Micromorphology

40cm 40cm

30 30

20 20

10 10

0 0

10 10

20 20

E. of VII.32 S. of VII.32 S. of VII.32 VIII.25 Section 3 Section 3 Section 4 VIII.25 X/Vlll.25/15


(Platform) (centre) Platforms VIA/V VIA/V E. of VIB-A.47 (East of FI) ?Stable
+centre centreS. room Room with
bench

Plaster floors
Platform
Multiple layers of white plaster
Levelled debris
Domestic occupation deposits
Dung-rich deposits

Figure 15.2. Microstratigraphic columns illustrating the type, thickness and frequency of floors and deposits in
different contexts.

were plastered more frequently than the floors, per- aggregates and organic-rich deposits accumulated
haps because they were more exposed to abrasion in a relatively inaccessible area between the eastern
and wear (Figs. 15.5 & 15.7). The floors in field sec- wall and the fire-installations. In thin section (Photo-
tion 7 had irregular depressions which were patched graph 15.14), these deposits include large heteroge-
and plastered over and in some instances may relate neous plaster aggregates <5 em, burnt aggregates,
to truncated horizons (Fig. 15.4). charred plant remains <1 em, pale yellow salts, burnt
There is considerable spatial and contextual bone <5 mm, organic aggregates <1 em in lenses
variation in the type, thickness and frequency of more than 6 em in length, and dung (type 2: fibrous,
plaster floors and occupation deposits within some no spherulites). The plant remains include: charred
buildings. oak and ?pistachio wood, charred cereal grain frag-
ment, Gramineae epidermises, and hackberry peri-
Shrine VIII.25 carp. Fire-cracked stones occurred in the lowermost
Spatial variation is particularly evident in a section of these deposits.
through the southern half of Shrine VIII.25 (Figs. At least nine phases of periodic rebuild of
15.5-15.6). Thick irregular layers of heterogeneous the fire-installations have been identified from

313
~
('[)
::l
0..
"<
~
~
.....
.....
::r
('[)

<
,'JJ
1007 1007
n
::r
~
....
ro-
[fl

>Tj
....
('[)
::l
n
(;.)
.......
7....,
>!'>-
1006.50 1006.50 s·
0
.....
::r
"<
r'
Scale at 50cm intervals ~

....('[)<
::l
n
Figure 15.3. Microstratigraphy of platform and floors in the building south of VII.32, section 9. ('[)

~
0
~
<
~
n
~
ro-
....
N s

______ D G ~
s,...,.
>fj'
;:o
(fJ
r----- c
....
1009 L-___ 1009 Pr'
n
ro
Vl

w
I ,...,.
....... :::r
(.}1 ro
~
;:;·
....
0

0
8
....
'"d
:::r
0
0
OQ
'<:

I
_ _ j _ __ - ·- -

Scale at lOcm intervals


-----
Figure 15.4. Microstratigraphy of floors in the building east of VII.32,section 7 (Fig. 14 .6).
w

1007 1007
~
ro
::l
0..
'<
~
PJ
...,.
1006.50 1006.50
&
ro
~
Y'
Scale at SOcm intervals n
::l"'
...,~
;:o
Figure 15.5. Microstratigraphy of floors in Shrine 25, level VIII, section 6 (Fig. 14.20). Vl
'Tj
...,
ro
::l
n
VJ .?"
......
0\ ......J
§.
0
&
'<
1007.50 1007.50 r<
w ~
~
...,
ro
::l
n
ro
~
1007 1007 0~
<
0.:
n
c...,.
1006.50 1006.50
;:o
...,
Scale at 50 em intervals

Figure 15.6. Phasing of floors in Shrine 25, level VIII, section 6 (Fig . 14.20) .
Multiple Surfaces: the Micromorphology

redistributed debris in Shrine VIII.25 (Fig. 15.6). Six to other types of buildings at <::atalhoyiik, which
of the bases are extant, and were re-plastered more have two or more rooms, one room of which may
than 2-6 times and reddened by burning during use. have fire-installations separate from a larger room.
A continuous lens of diatoms 150 !.liD thick may have When excavated, Shrine VIII.25 was 5.0 x 3.3 m,
been applied as a thin layer of white silty plaster on 16m2 (internal room dimensions), and had 2 plat-
top of 240 !.liD of burnt remains at the base of the fire forms, 3 fire-installations, wall paintings of nets
installation in Phase 4. The diatoms have been iden- which Mellaart suggests may have been hunting nets
tified as Epithemia ?adnata and E. sorex and Rhopalodia (Mellaart 1966, 180 & pl. XLIV), and two red ochre
gibba, which are epiphytic, occurring in fresh to brack- burials. Mellaart classified this building as a 'shrine'
ish water (Roberts pers. comm.). Layers of burnt fuel probably on the basis of the wall paintings and the
and remains still occur in situ on the last base of red-ochre burials.
some of the fire installations. All of the bases were In this limited sample of five buildings, Shrine
infilled with aggregates from the destroyed super- VIII.25 stands out firstly as the building with the
structure. In thin section the burnt fuel and remains greatest depth of occupation deposits, at more than
include: abundant charred oak and hackberry I elm 37 em. The surviving sequence is more than three
wood, siliceous Gramineae, calcareous ashes, dung times deeper than other buildings in the sample
pellets type 2, burnt bone, and heterogeneous aggre- group. Secondly, it has the highest number of multi-
gates. ple layers of white plaster laid without an interven-
The fire-installations were separated from the ing mud plaster, at 6 layers. Thirdly, it has the only
western half of Shrine VIII.25 by a low plastered periodic accumulation of structural levelling debris,
ridge on the floor. The floors in the western half coinciding with cyclical internal modifications.
were kept much cleaner and were frequently plas- Examination of the microstratigraphic evidence
tered with multiple layers of thin white plaster, <0.12- suggests that this building was used for multiple
2.8 mm thick. Occupation deposits between floors activities relating probably to both residence and
are very thin, often 0-250 l..lffi, occasionally up to 1 mm. burial of the dead. Apart from periodic modifica-
In thin section some plasters bear mat impressions tions, the fire-installations appear to have been in
on the surface. The thin layers of occupation depos- constant use, and there are continuous accumula-
its include c. 10-20 per cent charred 'reed'-like frag- tions of thin well-maintained plaster floors with thin
ments <10-300 !.liD (Photograph 15.16), amorphous layers of occupation deposits, which include some
organic staining <300 !.liD, and very sparse (<2 per domestic debris including Gramineae and, next to
cent) shell, red ochre grains and siliceous Gramineae. the fire-installations, bone, organic aggregates and
The periodic modifications of fire-installations charred wood. Humans were buried at several stages
in the eastern half of the room coincide with modifi- during the lifetime of the building. The application
cations in the western half of the room. These modi- of red ochre/paint to the platforms above human
fications are represented by periodic layers of levelled bones in the southwestern corner, and the presence
debris with aggregates of fire-installation superstruc- of red ochre in the last 5 mm of deposits in the fire-
ture and white plaster, with sparse charred 'reed' installation of phase 4, suggest some ritual activities
and wood fragments including probable hackberry associated in the first instance, at least, with burial.
wood, <6 mm, which abut platforms in the south-
western corner of the room. The platforms were Southern building section 3, level VIA/V
raised, rebuilt and modified each time. Human bones The only other complete section currently exposed
were found in the platform for phases 6-7. The lay- through a building in Area E occurs in the southern
ers of plaster immediately above these bones were half of section 3 (Fig. 15.7). This building has two
successively covered in red ochre/paint, close to the rooms, and is not on Mellaart' s plans, suggesting it
western wall face. A slightly bell-shaped grave had has been exposed by erosion since the 1960s. It has
been cut through all of the surviving layers of plas- been attributed to level VIA/V (Chapter 14). In the
ter and platforms. This grave contained a crushed field section, the smaller northern room is 1.64 m
human cranium. There are some traces of a fire- wide, and the larger southern room is 2.92 m wide
installation in this western corner in phases 6-7. (internal measurements). There is less evidence for
The range of contextual variations in micro- horizontal spatial variation in the microstratigraphy
stratigraphic sequences across any room will depend in this particular section through the building, than
partly on the number of rooms in each building. in the section through the southern part of Shrine
Shrine VIII.25 is a single roomed building, in contrast VIII.25.

317
~
ro
::s
0..
'<:
~
Pol
N .....
1013.50
.....
1013.50 ::r"
ro
~
!:!'
n::r"
Pol
>-t
1013 1013
ro-
Vl
>Tj
>-t
ro
::sn
UJ 7
...... 1012.50 ,...,
00 1012.50
§"
0
.....
::r"
'<:
r'
Scale at 50cm intervals Pol
~
>-t
ro
::s
n
Figure 15.7. Microstratigraphy of floors in a building in section 3, level VIA /V, east of VIA.42 court or V.9 (Fig . 14.4). ro
~
Cl
Pol
<
5:
n
~
ro-
'"'
Multiple Surfaces: the Micromorphology

The most striking aspect of this building in sec- wide and 10 em deep in section, and was filled with
tion 3 is a vertical change in the microstratigraphic dark ashy material. A series of deposits possibly
sequence which suggests there was a change in the related to ritual activities have been identified in a
use of space from domestic to cleaner /?ritual activi- sample from the top of this hole in thin section (Pho-
ties (Photograph 15.1). This change in the microstrati- tograph 15.12). This sequence of deposits comprises
graphic sequence was observed and recorded in 1993. a thin layer of white plaster, 84-200 J..Lm thick, which
On re-examination of the section in 1995, it was evi- was covered by a thin layer of silt loam ?plaster, 84-
dent that erosion had removed a further 2-4 em 250 J..Lm thick, with an undulating surface. This sur-
from the section face to expose more of an irregular face was covered directly by a layer of red ochre,
cut which was visible in 1993. It is now clear that this 20-120 J..Lm thick. Aggregates of white plaster, some
cut is the edge of a large grave with approximately with multiple layers, and flecks of charred plant
vertical sides which contains several human bones remains accumulated on this ochre, perhaps relating
which have been identified as toe and long bones by to modifications of features within the room. This
Dr L. Martin (Fig. 15.7, northern side of large room). layer of aggregates, 24-360 J..Lm thick, was subse-
The particular significance of this grave to our quently covered with a second lens of red ochre, 24-
dicussion here, is that the grave was cut from pre- 360 J..Lm thick, with deposits with 10 per cent charred
cisely the same level as the observed change in the 'reed' -like plant remains. The slight depression at
microstratigraphic sequence. the top of this hole was then covered by a thick layer
The earliest use of the building is represented of white plaster, 2.8 mm thick. No other layers of red
by a series of thick plaster floors and layers of occu- ochre were observed in any of the subsequent se-
pation deposits, associated with a plastered emplace- quence or samples in the room.
ment which was probably for a pot set into the floor The later use of the building is represented by a
(Photograph 15.1:1-2). This emplacement is 27 em in sequence of finely plastered floors which covered
diameter. Pots set into floors were found by Mellaart, the grave and were kept much cleaner. A collapsed
and contained small stones which probably served Bos jaw, three horn cores and a range of moulded
as pot warmers (Mellaart pers. comm.). These pot plaster features lay on the last floor. The 'mud' plas-
settings tended to occur in the central depressions in ter floors are thinner than those in the first phase,
rooms, and have been illustrated in association with only 1.0-6.0 mm thick, and usually have a finishing
straw bedding in what may have served as a tempo- coat of white plaster, 60-480 J..Lm thick. Occupation
rary refuse area (Time Life Books 1991,124-5). The deposits between floors are very thin and intermit-
plaster floors in this phase of the building are thick, tent, at less than 120-720 J..Lm. In thin section (Photo-
1.0-2.5 em, and do not have a finishing coat of white graph 15.15) these deposits do not include any burnt
plaster. Overlying layers of occupation deposits are aggregates, bone, or cereal-like remains, which are
also thick, at 2.0-3.0 em. In thin section (Photographs characteristic of the first phase, and of domestic ac-
15.8 & 15.13) these layers of occupation debris are tivities. The deposits in this second phase include:
not very compacted and include charred wood, sili- thin discontinuous lenses of 'fibrous' remains and
ceous Gramineae including fragments of cereal-like small fragments of possible compressed dung pellet
?husk, calcareous ashes, burnt and unburnt bone, similar to dung type 2, which as yet are unidentified
and heterogeneous aggregates. These thick floors and (see pp. 309-10). Dung pellets occur in deposits on
deposits rich in domestic debris associated with a top of the last floor in the building, where they are
?pot emplacement suggest the building was initially intact and variously compressed (Photograph 15.24).
used as a domestic residence. Similar floors and oc- The ?dung pellet fragments between plaster floors are
cupation deposits were identified in a small room in up to 7.3 mm in length, and 0.6 mm thick. The pellets in
Area A-E hence VIA-B (Fig. 15.8 & Photograph 15.9). the collapsed debris are up to 7.7 mm in length and
The next sequence of deposits in the building in 1.05-4.2 mm thick.
section 3 relates to the digging of the large grave, In the adjacent small room to the north, there
with some evidence for associated ritual activities. are few deposits relating to the first phase of activi-
The thick plaster floors and occupation deposits from ties in the building. The late phase is represented by
the first phase in the north of this room were cut by the construction of a bench, which was coated with a
the large grave which, unusually, was not placed thick layer of white plaster. There are few remains or
below a platform. A small hole had been cut into the indications of activities.
top of the grave close to the wall, shortly after the The coincidence of this change in microstrati-
grave had been dug and infilled. The hole is 12 em graphy with the act of burial and use of red ochre

319
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

N s 1013
1013

1012.50 1012.50

1012 1012

Scale at 50cm intervals

Figure 15.8. Microstratigraphy of floors in a building in section 4, level VIB/A, north of VIA.46 (Fig. 14.5).

suggests the use of these areas, and perhaps the rest In the current set of samples one of the most
of the building, changed from domestic to cleaner I distinguishing characteristics of microstratigraphic
?ritual activities, perhaps from a residence to an an- sequences within buildings is the presence or ab-
cestral shrine, after the death of human individuals, sence of finishing coats of white plaster, which
possibly the inhabitants. although simplistic, appears to be significant (Fig.
15.9).
Use of space attributes Where there are successive layers of plaster with
Although the current sample group is small, it is one or more layers of white plaster finishing coats,
possible to identify two general sets of micro- overlying occupation deposits are either very thin or
stratigraphic sequences, the first of which can be absent, <1 mm thick, as in sections 7 and 9 and the
associated with clean /?ritual activities, the second upper phase of section 3 (Photographs 15.10-15.11 &
with more domestic activities. The characteristics of 15.15-15.16). The principal inclusions in the thin lay-
these two groups, and current subdivisions are illus- ers of occupation deposits comprise charred 'reed'-
trated in a cladistic diagram with a preliminary list like remains, <10-200 )lm. These sequences tend to
of key attributes in Figure 15.9. The significant at- occur in large or elaborate buildings. The floors in
tributes and interpretations proposed here will be section 7 occur in a large building 9.44 min length,
reassessed once the data base is complete and the which has thick accumulations of white plaster on
statistical significance of each attribute can be fur- the walls. This building was heavily burnt and has
ther evaluated. As excavations proceed the sample large charred juniper timbers (Chapter 17). It cur-
group will be considerably expanded and compared rently stands to a height of 2.73 m . The floors in
to results from other interdisciplinary analyses. More section 9 are covered in extensive layers of collapsed
variation in microstratigraphic sequences and other wall plaster with multiple layers of red paintings
significant attributes are likely to emerge. visible in section. The floors and the wall plaster in

320
Multiple Surfaces: the Micromorphology

this building were burnt. The sequence in the upper Variation in microstratigraphic sequences in a
phase of section 3, as we have already mentioned is ?stable and open areas
associated with a collapsed Bos jaw, at least 3 horn A significant proportion of the site at <;atalhoyiik
cores, and plastered moulded features. comprises open areas rich in discarded refuse (Chap-
Where there are few or no finishing coats of ters 7 & 14). A range of deposit types in open areas
white plaster as in the early phase of section 3 or in has been sampled for micromorphological analysis.
the small room in section 4 (Fig. 15.8), the overlying The cladistic diagram in Figure 15.10 illustrates the
occupation deposits are generally thicker, at 0.5-2.0 principal microstratigraphic characteristics of the
mm. In thin sections (Photographs 15.8-15.9 & 15.13) major deposit types in the field. Detailed micro-
from both of these sequences accumulated deposits morphological characteristics await completion of
consistently include burnt and unburnt heterogene- the data base. In thin section it is possible to identify
ous organic and inorganic aggregates, small frag- individual depositional episodes and thereby pro-
ments of bone <1.5 mm, charred cereal grain fragments vide contextual links and associations between the
<5 mm, siliceous Gramineae up to 7 mm in length rich artefactual, organic and sedimentary components
including cereal-like siliceous plant remains <1 mm, in single depositional units in these open areas.
melted silica <360 Jlm, charred wood <2.1 mm and Plant remains constitute the largest class of resi-
calcareous ashes <30 Jlm. There are few floors to the due in open areas. Clear differences in the types of
east of the fire-installations in Shrine VIII.25. Here plant remains in depositional layers, corresponding
occupation deposits also include organic aggregates to different activities are emerging (Table 15.2, pp.
and charred plant remains and appear to be domes- 310-11).
tic in character (Photograph 15.14). Organic aggregates, ?omnivorous coprolites dung
No finishing coats of white plaster have yet type 3, are more common in refuse deposits in open
been encountered during new excavation of floors in areas than in occupation deposits within buildings.
a small room in a large building in a separate com- These aggregates include fragments of ?digested bone,
plex at the northern end of the mound, room 70 in hackberry pericarp, and phytoliths (Photograph 15.25).
building 1. This room has a well constructed fire- It is not entirely clear whether most of the organic
installation, and distributions of lentils, peas, and refuse was burnt within buildings in fire-installa-
acorns on the floor. The larger room, room 71 by tions before being dumped in open areas, or whether
contrast, has thick layers of white plaster on a large it was burnt in the open areas after dumping. There
platform, and ritual fixtures. are clear instances of both, and both have implications

Microstratigraphic sequence Observations Use of space interpretations Associated features Sample locations

few or no accumulated deposits cleanl?ritual - - - - - - - - - - - - s e c t i o n ?


building east of VII.32

clean - - - - - - platform - - - - section 9


building south of VII.32
Plaster floors with
white plaster finishing coats charred reed-like plant remains - - - cleanl?ritual - - - - wall paintings - - - section6
red ochre in fire-installation and ned ochre burials Shrine VIII.25: south-central
sparse fragments in occupation deposits platforms and western areas
red ochre/paint on platform

few accumulated deposits cleanl?ritual - - - - - Bas jaw and hom cores - section 3
red ochre lenses moulded sculptures 5 building, upper sequence
occasional ?dung pellets

Plaster floors with few or no - charred wood domestic - - , - - - - pot emplacement - - section 3
white plaster finishing coats cereal-like plant remains 5 building, lower sequence
fragments of bone
heterogeneous burnt and unburnt aggregates ' - - - - - - - - - - - section4
building N of VIA.46y

Few laid plaster floors - - - charred wood burning and ?cooking - - fire-installation - - - section 6
charred cereal grain Shrine Vlll.25, southeastern area
organic aggregates

Key
Plain text microstratigraphic field observations
Italic text micromorphological observations
Bold use of space interpretations

Figure 15.9. Preliminary guide to microstratigraphic attributes diagnostic of uses of space within buildings.

321
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

for analysis of the contextual associations of activity The spherulites in these deposits are sub-
residues. This will be examined by analysis of burn- rounded and small, 5-20 J..Lm in size. They have a
ing on all other artefactual, and organic and inor- moderately high birefringence and an extinction cross
ganic remains in open areas, retrieved during in cross-polarized light. These spherulites have been
excavation and controlled wet-sieving and flotation. examined at Kew, and by Dr M. Canti, Ancient Monu-
On the basis of attributes studied in the field, ments Laboratory, English Heritage and Dr D.
deposits in open areas can be divided into two major Samuel, McDonald Institute, University of Cam-
groups, those with depositional layers which are bridge. Identification is problematic because a range
more or less horizontal, and those which are steeply of organic and inorganic compounds crystallize in
sloping (Fig. 15.10). radial or concentric structures and exhibit an extinc-
Four distinctive types of horizontally depos- tion cross (Canti in press). A number of common
ited sequences in particular were observed, three of organic and inorganic compounds can be ruled out.
which have been studied in thin section. The spherulites in Court X/VIII.25/15 are definitely
One very distinctive sequence comprises dung not characteristic of any plant materials. They are
rich deposits in a roughly walled space in level X/ not calcium oxalate druses which occur in many
VIII Court 25/15, which current evidence suggests, dicotyledons and have a much spikier form. Nor are
may have been a stable for herbivores (Fig. 15.11 & they starch grains which have a lower birefringence,
Photograph 15.17). As yet, it is not possible to deter- and stain with iodine potassium iodide. The spher-
mine whether this area was roofed or unroofed, be- ulites did not stain on contact with iodine potassium
cause the deposits have been very heavily trampled, iodide. The spherulites do not resemble coccolith
any evidence of wind or water-laid layers, for exam- microfossils, which also exhibit an extinction cross,
ple, would have been largely reworked. The sequence but tend to have a plate-like structure, and are discs
comprises thin interbedded lenses of orange-brown, not spheres and may appear eliptical in thin section.
grey and white silt loam to sandy silt loam deposits, The size, shape and optical properites, of the
10YR 6/4light yellowish brown and 10YR 6/2light spherulites in the Court X/VIII.25/15 resemble
brownish grey (Photograph 15.17). In thin section spherulites which have been identified in the dung
(Photographs 15.18--15.20 & 15.22) these deposits princi- and gut of modern animals (Brochier 1992; Canti in
pally comprise interbedded layers of orange organic- press), and in dung pellets and dung related depos-
stained deposits rich in dung pellet fragments and its on a wide range of archaeological sites, including
dung-derived components which include partially the PPNB sites of Cafer Hoyiik and <;ayonii Tepesi
digested plant remains, occasional diatoms and abun- in Turkey (Brochier 1985; 1993), Abu Salabikh in
dant spherulites. The plant remains include digested Iraq and Tell Brak in Syria in the current NERC
Gramineae stems and leaves, with sparse ?tubers. micromorphology project, and regions elsewhere in

Microstratigraphic sequence Observations Use of space Sample locations


interpretations

interbedded layers of dung pellets, digested plant remains, - - ?stable section6


orange-brown and spherulites, organic staining, and salts Court X/Vlll.25/15
white deposits with strong parallel orientation and
linear parallel distribution

scooped hollows with pale ash and highly fired aggregates - high temperature - - section 3
rubified substrate burning, ?industrial northern area

1
horizontal deposits
heterogeneous layers - - - rich in plant remains with accumulating refuse section 3
strong parallel orientation Court42
linear parallel distribution
complex packing voids section6
CourtXI.25

thick layers with - - - - rich in plant remains - - - - - - re-deposited refuse - section 6


massive bedding unoriented and randomly distributed building X/Vlll.27 infill

steeply sloping deposits heterogeneous layers rich in plant remains with - - - - accumulating refuse - section 2
strong parallel orientation western area
linear parallel distribution
complex packing voids

Key
Plain text microstratigraphic field observations
Italic text micromorphological observations
Bold use of space interpretations

Figure 15.10. Preliminary guide to microstratigraphic attributes diagnostic of uses of space in open areas.
322
1006 1006

~
E.
~

.u·;:c
(J)
.::...,
Pr'
n
ro
rn
w ::;.
N
w ro
~
;:;·
...,
0
s
0...,
'"0
;:r
0
0
OQ
Scale at SOcm intervals '<:

Figure 15.11. Microstratigraphy of ?stable deposits in Court X/VIII.25/15, section 6 (Fig. 14.20).
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

the world (Courty et al. 1989; Canti pers. comm.). posits will be sampled in 1996.
Spherulites from Court X/VIII.25/15 were kindly The third and most common type of deposit
examined under SEM with EDXRA by M. Canti, in comprises heterogeneous layers of refuse rich in plant
order to examine their three-dimensional structure remains, and constitutes the bulk of deposits in open
and elemental composition. These spherulites are areas with horizontal layering. These deposits have
slightly dumbell shaped in one axis (Photograph strong parallel referred and related orientations and
15.21), and principly comprise calcium and silica. linear parallel referred and related distribution pat-
Further research is required in order to establish the terns, from periodic dumping and settling of depos-
relationship of spherulites to digestive processes, and its, visible both in the field and in thin section. The
variation according to animal species, age, diet, pas- deposits sampled in section 3 are dark reddish brown
ture substrate, and depositional and post-deposi- from in situ combustion of smouldering refuse rich
tional processes and environments. in plant remains and ?omnivorous coprolite frag-
This sequence of deposits in Court X/VIII.25/ ments. Some deposits have a much higher sediment
15 has been compared to thin-section samples of content, but these are less frequent, and include
modern dung and stabling in the region of Pmarba~1 weathered building materials.
in the Konya plain and A~1kh Hoyiik on the edge of The fourth type of deposit sampled is also rich in
Cappadocia in Turkey, which were collected by Seona plant remains and refuse, but has massive bedding
Anderson, for an M.Sc. dissertation at the University and unoriented randomly distributed components, vis-
of Sheffield (Anderson 1995). The sequence in Court ible both in the field and in thin section. The massive
X/VIII.25/15 most closely resembles two modern deposit in level IX infills a severely truncated build-
samples from winter stables for sheep and goats ing in level X/VIII.27, and represents redeposited
which were roofed (Photographs 15.22-15.23). The refuse, dug out from another open area elsewhere.
modern samples similarly contain digested plant re- The microstructure has complex packing voids which
mains, organic staining and spherulites, and exhibit are characteristic of many dumped deposits.
pronounced horizontal layering and cracking due to Many of the steeply sloping deposits in open
the compaction and trampling of moderately dry areas such as those sampled in section 2 have hetero-
droppings and decay of some organic remains. De- geneous layers from periodic dumping, and parallel
posits from Court X/VIII.25/15 were also sampled orientation and distribution from sorting and set-
for a range of comparative chemical, physical and tling by gravity. Some of the largest fragments of
mineralogical analyses, which should aid assessment wood and richest samples of plant remains come
of whether these deposits are derived from in situ from the post-level VI dump in section 2, including
stabling within the settlement. In particular, gas chro- those with ?wood chippings and basketry /matting
matography /mass spectrometry analyses will be con- remains. There are also periodic layers rich in weath-
ducted by Dr R.P. Evershed and colleagues at the ered sediment and eroded mud brick.
University of Bristol in order to detect any surviving
traces of bile acids, sterols and coprostanols, which Conclusions
are specific to, and therefore diagnostic of, different
animal species. Distributions of phosphorus and salts Micromorphology is yielding sensitive information
in these deposits will be analyzed by Dr D.A. Jenkins on spatial and temporal variation in multiple sur-
and colleagues at the University of Wales. faces and deposits at C::atalhoyiik. It is enabling mi-
If comparative and interdisciplinary analyses croscopic analysis of the depositional and contextual
confirm that these deposits are the remains of in situ relationships of components within the sequences of
stabling at C::atalhoyiik, the results will have important thin plaster floors and occupation deposits within
implications for studies of early agricultural settle- buildings, which are often less than 1 mm thick, and
ments and ecology, already suggested by research at from single depositional episodes within the multi-
the earlier PPNB site of C::ayonii Tepesi (Brochier 1993). ple layers of refuse in open areas.
The second distinctive deposit type in open ar- In particular micromorphology is enabling iden-
eas at C::atalhoyiik occurs in a series of shallow scoops tification of the source materials and technology
in the northern area of section 3 which are c. 65-80 employed in the manufacture of floors and architec-
em wide and 15-20 em deep. Deposits within these tural materials, and should be further integrated with
scoops comprise pale ash with highly fired aggregates, palaeoecological sampling. The micromorphological
and may derive from industrial activity. Underlying evidence strongly suggests, pending analysis by SEM
deposits have been rubified by burning. These de- with EDXRA, that the white floor and wall plasters

324
Multiple Surfaces: the Micromorphology

were prepared from Neogene-Quaternary soft lime organic aggregates and coprolites and amorphous
deposits which comprise up to 95 per cent pure car- staining in thin section, suggest organic residues and
bonates of calcium and magnesium and outcrop to- phosphorus may be present.
day within 5 km of the site, and were mixed with Although the current group of samples is small,
water and occasionally vegetal stabilizers, as sug- recurrent sets of microstratigraphic and micromorph-
gested by Mellaart (1962, 48). The white plasters in ological attributes are emerging which can be associ-
the samples analyzed do not appear to have been ated with different classes of activity (Figs. 15.9-
prepared by firing lime to temperatures of 850-900° 15.10), in particular domestic-related deposits on the
C in contrast to samples analyzed from a number of one hand and cleaner/?ritual deposits on the other.
Neolithic sites in the Near East, including one sam- Horizontal and vertical changes in microstratigraphic
ple from <;atalhoyiik (Kingery et al. 1988). sequences have been identified within several build-
The floors and surfaces exhibit a range of at- ings at <;atalhoyiik, providing sensitive indications
tributes characteristic of different impacts from hu- of variation in spatial and temporal uses of space
man activities, including sub-horizontal cracks and and activities, independent of architectural elements.
dislodged aggregates suggestive of trampling, peaked There is considerable evidence to suggest the use of
boundaries and fine lenses characteristic of probable at least one area of a building from level VIA/V in
mat/rug impressions, and impregnation with salts. section 3 changed from domestic to ritual activities,
A wide variety of plant types, anatomical ele- after the digging of a large human grave.
ments and materials have been identified, preserved The micromorphological information will be
within their depositional context in thin section, ena- more potent when it is drawn from a larger sample
bling study of taphonomy and a range of human- group and when it is combined with further inte-
plant relationships at a critical stage in the development grated multidisciplinary analysis both of components
of agriculture and complex settlements. A variety of in thin section and from other samples and remains.
herbivore and ?omnivore coprolite fragments occur In particular organic and inorganic analyses of de-
in thin sections of deposits in fire-installations, build- posits will enable integration of chemical and physi-
ings and open areas. A probable stable area has been cal characterizations and identification of traces of
identified in Area E. It is represented by a thick activities, and will aid identification of potentially
sequence of interbedded layers of dung, organic stain- 'hidden' components. Controlled volume wet-sieving
ing and salts, which resembles deposits in thin sec- and flotation will provide larger sets of information
tions of ethnoarchaeological samples from modern on three-dimensional contextual variation through
dung and stable deposits from the Konya region, in time. The new relational data base for thin-section
particular two samples from winter stables for sheep description and interpretation is greatly aiding data
and goats. Identification of the dung in ancient sam- sorting and statistical corroboration of the signifi-
ples requires integration of physical and chemical cance of observed attributes and interpretations in
analyses, in particular gas chromatography /mass different contexts. Further ethnoarchaeological and
spectrometry analysis for traces of bile acids and experimental sampling of occupation deposits is re-
coprostanols which, if present, are specific to, and quired for study of the dynamic effects of human
therefore diagnostic of, different species. A range of activities and natural agents on depositional se-
discrete single depositional episodes have been iden- quences in the built environment. It is hoped that we
tified in open areas, including discard of small wood have been able to illustrate some of the questions
fragments and obsidian, perhaps from wood-work- and issues which microstratigraphic and micro-
ing, and charred cereal grain and burnt fire-installa- morphological analyses can address at <;atalhoyiik.
tion plaster, probably from cooking.
Many of the deposits at <;atalhoyiik are well Acknowledgements
preserved with little post-depositional disturbance.
There has been some bioturbation from root and We wish to thank the Natural Environment Research
insect activity. The occasionally extensive charring to Council for funding this research and S.M. Anderson,
depths of up to three feet at <;atalhOyiik has resulted M. Bower, Dr J. Sunbury, Dr M. Canti, Dr M.A.
in remarkable preservation of organic materials, in- Courty, W. Eastwood, F. Ertug, Professor U. Essin,
cluding charred textiles, fur, human tissue and Dr. R.P. Evershed, Dr D.A. Jenkins, Professor M.K.
wooden vessels and their contents (Mellaart 1964, Jones, M. Nesbitt and Dr D. Samuel for kindly offer-
85-6). The bright autofluorescent excitation colours ing helpful advice and assistance on individual as-
of organic materials within charred plant cells, in pects of the research.

325
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

Photograph 15.1. The microstratigraphic sequence in this building suggests there was a change in the use of space
from: i) domestic activities attested by relatively thick plaster floors and burnt occupation deposits [1], associated with
a plastered emplacement perhaps for a pot [2] , to ii) clean/? ritual activities attested by the sequence of fine white and
orange plaster floors [3] which are associated with a Bos jaw, horn cores and elaborate plastered features [4]. Building
in the south of section 3, level VIA /V. Scale = 20 em.

Photograph 15.2. Block sample for micromorphological analysis in the process of being cut out from a sequence of
multiple layers of burnt plaster in a fire-installation. Building from level V in section 2, Area A. Sample (:h93 .08,
H:13.5 em, W:6.5 em, depth 8-10 em.
326
Multiple Surfaces: the Micromorphology

Photograph 15.3. Modern sample of soft lime deposits which occur within 5 km of c;atalhoyiik and closely resemble
materials used in Neolithic white plasters . These deposits comprise up to 95 per cent pure carbonates of calcium and
magnesium, and are used today for plastering floors and walls in surrounding settlements. Sample (h94.50. Cross-
polarized light (XPL). Frame width= 7.2 mm.

Photograph 15.4. The first 41 layers of a sample of wall plaster are visible here in sequence from left to right. These
plasters were laid as couplets comprising: [1] a pale brown silty clay preparatory layer, 24-720 Jlm thick, with vegetal
stabilizers which have since decayed leaving voids in the sedimentary microstructure, and [2] a finer denser finishing
coat, 12-552 Jlm thick, which was keyed into the irregular surface of the preliminary coat. The outer surfaces of these
finishing coat are very regular and sharp, and were probably smoothed or burnished. Sample (h94.24. Collapsed
fragment in Area E. XPL. Frame width = 7.2 mm.
~ -
,, .

::2
(!)
:::::
0..
'<
~
Q.l
.....
.....
::T
(!)

~
!:!'
n::T
Q.l
'"t
ro-
CJJ
'Tj
'"t
(!)
:::::
(')

w 7
N
CXJ >-l
§'
0
.....
::T
'<
r'
Q.l
~
'"t
(!)
:::::
(')
(!)

~
tj
Q.l
<:
0.:
n~
.....
ro-
'"t
Photograph 15.5. Mud brick [1] and mortar [2] from the Photograph 15.6. Many fire-installation plasters have a
earliest building uncovered by Mellaart, Xll.25. The mortar coarser particle size than floor plasters, and closely resemble
comprises heterogeneous burnt and unburnt anthropogenic alluvial deposits sampled in a palaeochannel adjacent to the
debris including fragments of charred plant remains, bone, site. Each plaster has been baked up to 1-2 em from the
and volcanic glass. Sample (:h94.13. Area E, section 6. Plane surface [1]. Only discontinuous lenses of burnt remains
polarized light (PPL). Frame height= 7.2 mm. <2 mm thick [2] survive between each layer of plaster [3].
Sample (:h93 .08. Area A, Section 2, level V. PPL. Frame
height = 7.2 mm.
·Multiple Surfaces: the Micromorphology

a b

Photograph 15.7. Identified plants in thin section include:


a) transverse section , charred wood, Quercus, deciduous white oak type. Sample (:h93.12. Open area, Area A, section
2. PPL. Frame height= 7.2 mm;
b) transverse section, charred wood, Salix/Poplus, willow/poplar. Sample (:h93.12. Open area, Area A, section 2.
PPL. Frame width = 7.2 mm;
c) transverse section, charred wood, Chenopodiaceae similar to Salsola and Sueda for example. Sample (:h94.10. Open
area, Area E, section 6, level IX. PPL. Frame width= 4.4 mm;
d) charred triquetrous stem of Cyperaceae, Cyperus/Carex/Scirpus. Sample (:h93.07. Open area, Area A-E, section
3. PPL. Frame width= 1.76 mm.

329
~
(!)
::::
0.
'-<
~
PJ
,..,.
,..,.
::r"
(!)

~
!!'
n
::r"
PJ
.....
ro
(fJ

>Tj
...,
(!)
::::
n
w 7
w >-1
0
§'
0
:;
'-<
rPJ
...,~
(!)
::::
n
(!)

~
a
PJ
<
5:
nc
.....
ro
...,
Photograph 15.8. Initial sequence of thick plaster floors [1 & Photograph 15.9. A similar sequence of thick plaster floors
31 and occupation deposits [2] in section 3, probably associ- [1 & 3] and occupation deposits [2] probably related to
ated with domestic activities. These deposits include charred domestic activities, includes bone and charred and siliceous
wood, siliceous plant remains, calcareous ashes, bone, and plant remains, including a cereal-like remains. Sample
burnt aggregates. The surface of the plaster floor [1] has sub- (h93.01. Area A, section 4, building north ofVI.46. PPL.
horizontal cracks and is impregnated with pale yellow salts. Frame height = 7.2 mm.
Sample (h93.03. Area A-E, section 3, southern building,
level VIA-V. PPL. Frame height= 7.2 mm.
~
~
.u·
ro
[Jl
~
.....
.....,
ll>
()
ro
'!!
(.;J
(.;J &
>--' ro
~
r:;·
.....
0
s0
.....
'1j
::>
0
0
OQ
'<

Photograph 15.10. Later sequence in section 3 of thin plaster Photograph 15.11. Comparative sequence of thin plaster
floors [1] with finishing coats of white plaster [2] and occupa- floors with multiple couplets of white plaster, and sparse
tion deposits (<0.2 mm thick), associated with clean/?ritual occupation deposits. Linear and curvilinear pseudomorphic
activities. Sample (:h93.03. Area A-E, section 3, southern voids from decayed vegetal stabilizers are clearly visible in the
building, level VIA-V. PPL. Frame height= 7.2 mm. plasters [1]. Sample (:h94.21. Area E, section 7, building east
of VII.32 . XPL. Frame height= 7.2 mm.
(1)
<
::s
0..
'<:
~
Pl
.....
.....
7 ::s"'
(1)

~
!!'
n::s"'
Pl
'"'!
ro-
Vl

~
(1)
::s
(')
UJ 7
UJ ....,
N 6 §'
0
5 .....
::s"'
'<:
r'
Pl
4 ~
'"'!
(1)

3 ::s
(')
(1)

2 ~
Cl
Pl
1 <
0.:
n~
.....
ro-
>-t

Photograph 15.12. Microstratigraphic sequence of ?ritual deposits in the small hole in the top of the grave in section 3, comprising lenses
of: [1] thin white plaster, [2] silt loam ?plaster, [3] red ochre, [4] aggregates of white plaster, [5] red ochre, [6] deposits with finely fragmented
charred plant remains, and finally [7] a thick capping of white plaster. Sample (:h93.06. Area A-E, section 3. Reflected light (RL). Frame
width = 7.2 mm.
Multiple Surfaces: the Micromorphology

. . 1.
. . . •, J.l:'
.. ...

~ . .
~

Photograph 15.13. Detail of probable domestic occupation deposits in section 3, including fragments of bone [1],
cereal-like husk [2], and siliceous grass epidermis [3]. Sample (h93.03. Area A-E, section 3, southern building, level
VIA-V. PPL. Frame width= 1.3 mm.

Photograph 15.14. Deposits rich in organic staining [1], salts [2], aggregates of burnt fire-installation plaster [3],
and charred plant remains, east of the fire-installations in Shrine VIIJ.25. Sample (h94 .02 . Area E, section 6. PPL.
Frame width = 7.2 mm.
333
~
ro
::1
0..
'<
3::
~
......
::::1"'
ro
~
!:!'
n
::::1"'
lll
>-t
ro-
w
'Tj
>-t
ro
::1
(")

~ .
··..· ' • .·. · . ·'. •.('V~ ·-;·
,.···~ -~ - •.
' ,... ~ · . '!' ... . · • . ' ::::1"'
~ ·. .0>. •
.... »!L"'~'
~.,;. W·@ · . .•, ..;.· ;.- • •• ., .~~ 1; '• · .. ·. _ . •. ··. ..li!i:.i
·. .,;~. ··.··;· ; ·.·.
~ ~ <.ol: •:· 1 ' · · \:•·· - . _-Iii:- --! i t _ - - ' - -" 4 \iJ
''*' 'L:.,.'f::;
,.....~, Jt":)'_~N; ·- ·-- >•,_\ ~- ~ - ·- . . ·:'+. ._.L_..,.;,_•' ~
.~ . ' §'
0......
::::1"'
'<
r'
lll
~
>-t
ro
::1
(")
ro
~
tJ
lll
<
5:
n
.:::
......
ro-
.....
Photograph 15.15. Thin layer of compacted fibrous deposits Photograph 15.16. Thin lenses of sandy silt loam [2] and occupation deposits rich in
[1] below a thin layer of white plaster [2] in the later sequence charred 'reed' -like plant remains [3 1, between white plaster floors in Shrine VIII.25 [1
of deposits in section 3, southern building. Sample (h93.03. & 4]. Sample (h94.07. Area E, section 6. PPL. Frame width= 1.3 mm.
Area A-E, level VIA-V. PPL. Frame height= 1.3 mm .
Multiple Surfaces: the Micromorphology

Photograph 15.17.lnterbedded layers of dung-rich deposits and salts in Area E, Court 25/15 level X/VIII. Sample
(:1194.15. Section 6, west. Scale= 50 em.

Photograph 15.18. Lenses of salt [1], dung pellet type 1 [2], and organic stain ing [3] in Court 25/15 level X/VIII.
Sample (:1194.15. Area E, section 6, west. PPL. Frame width= 7.2 mm.

335
~
ro
::s
0..
'<
~
C)
......
......
::s-
ro
~
!!'
n::s-
Cl
.....
(i)
rn
"11
.....
ro
::s
n
VJ ?"
VJ
0\ >-l
§"
s.::s-
'<
r'
C)

~
.....
ro
::s
n
ro
~
CJ
C)
<:
0.:
nc
......
(i)
Photograph 15.19. Detail of dung pellet in Photograph Photograph 15.20. Detail of dung pellet, same view as Photograph .....
15.18, illustrating partially digested plant remains, including 15.19 in cross-polarized light illustrating abundant spherulites. The
fragments of Gramineae stem and leaf, and ?tubers. Sample spherulites are 5-15J.l1n in diameter, and have a moderate
(h94.15. Area E, section 6, west, Court 25/15 level X/VIII. birefringence with a cross of extinction, in contrast to the isotropic
PPL. Frame height= 1.3 mm. plant remains. Spherulites have been found in the gut and dung of
animals such as sheep, goats, cattle, gazelle and pig, but further
research is required, before any associations can be confirmed.
Sample (h94.15. Area E, section 6, west, Court 25/15 level X/VIII.
XPL. Frame height= 1.3 mm.
Multiple Surfaces: the Micromorphology

Photograph 15.21. Scanning electron microscope image of a spherulite from a bulk sample of deposits from Court 25/
15 level X/VIII in collaboration with M. Canti, English Heritage . Sample c;h95.85. Area E, section 6, west.

337
~
ro
:;
0.
'<
~
Q)
......
......
::T
ro
~
!:!'
n::T
Q)
>-t
ro-
[Jl

"ri
>-t
ro
:;
()
w 7
w
(X;
....,
§.
0
~
'<
r<
Q)

~
>-t
ro
:;
()
ro
~
t:l
Q)
<
0.:
n
>=
......
ro-
Photograph 15.22. Sequence of dung-rich deposits in Court Photograph 15.23. Comparative sample of modern sheep/ "'
25/15 level X/VIII. PPL. Frame height = 7.2 mm. goat stable deposits from the Konya region, collected by Seona
Anderson, University of Sheffield, illustrating similar but less
compacted horizontal layering and dung pellet fragments.
PPL. Frame height = 7.2 mm.
~
s......
-c;·
ro
(fl
.::>-;
.......
PJ
n
(1)

(.).)
"'......
(.).) .-.~~i~~~~"~~-~1'·~~·& ::r
(1)
~-·,.\ .•
\0 '
,.~~L . ""....::..~"'.<
.. ·.·· l' . . . . ....... . ,... ....1~ : .. _:t '~
: ~~..
~ · . ··~
:t.:.~-.\....,. - ·" ~

>-;
0
9
0
>-;
'ij
::r
0
0
(1Q
'<

Photograph 15.24. Dung pellets type 2 [1] in debris lying Photograph 15.25. ?Omnivore coprolite type 3, with
on the latest floor in the southern building in section 3. These yellowish organic fine fabric , and fragments of bone, and
dung pellets have a fibrous structure with few recognizable charred and siliceous plant remains and pseudomorphic
plant remains and do not contain any spherulites. Some voids. Sample (h93.16. Area A-E, open area in south of
pellets are more compressed than others. Sample (h93.04 . section 3, level VIA-V. PPL. Frame height= 1.4 em.
Area E. PPL. Frame height= 7.2 mm.
~
ro
::l
0..
'<
~
PJ
.....
g.
ro
~
,"'
n
::;-
PJ
...,
;:o
"'...,'Tj
ro
::l
n
~ ?'"
..,. >-l
0
§'
0
g.
'<
r
tu
~
...,
ro
::l
n
ro
~
CJ
PJ
<
5:
n
~
;:o
...,

Photograph 15.26. Obsidian flake [1] and small tapered fragments of charred oak wood [2] in a layer in an open area with charred 'reed' and willow/poplar
fragments, which could perhaps represent remains from wood working and ?basketry/matting. Sample (:h93.12 . Area A, section 2. PPL. Frame width= 7.2 mm.
Multiple Surfaces: the Micromorphology

References Evershed, R.P., C. Heron, S. Charters & L.J. Goad, 1992.


The survival of food residues: new methods of analy-
Affonso, T., 1996. Identification of lime plasters. The Old sis, interpretation and application, in New Develop-
Potter's Almanack 4(1), 1-6. ments in Archaeological Science. Joint Symposium of the
Anderson, S.M., 1995. Fuel, Fodder and Faeces: an Ethno- Royal Society and the British Academy, February 1991,
graphic and Botanical Study of Dung Fuel Use in ed. A.M. Pollard. Oxford: Oxford University Press,
Central Anatolia. Unpublished M.Sc. Dissertation, 187-208.
University of Sheffield. Fedoroff, N., L.M. Bresson & M.A. Courty (eds.), 1987.
Bethel, P.H., J. Ottaway, G. Campbell, L.J. Goad & R.P. Micromorphologie des sols - Soil Micromorphology.
Evershed, 1994. Biomarkers of human activity: the Plaisir: Association Fran<;aise pour !'Etude du Sol.
use of coprostanol in the soil as an indicator of Gale, R.M.O. & D.F. Cutler, forthcoming. Plant Remains in
human faecal material. Journal of Archaeological Sci- Archaeology. London: Archetype Books & H.M.S.O ..
ence 5, 577-84. Ge, T., M.A. Courty, W. Matthews & J. Wattez, 1993. Sedi-
Bourdieu, P., 1977. Outline of a Theory of Practice. Cam- mentary formation processes of occupation surfaces,
bridge: Cambridge University Press. in Goldberg et al. (eds.), 149-64.
Brochier, J.E., 1985. Recherches de fumiers d'ovi-caprides Goldberg, P., 1987. Soils, sediments and Acheulian arti-
dans le neolithique de Cafer Hoyiik (Turquie). facts at Berekhat Ram, Golan Heights, in Fedoroff et
Cahiers de l'Euphrate 4, 121-2. al. (eds.), 583-90.
Brochier, J.E., 1992. Shepherds and sediments: geo- Goldberg, P. & I. Whitbread, 1993. Micromorphological
ethnoarchaeology of pastoral sites. Journal of An- studies of a bedouin tent floor, in Goldberg et al.
thropological Archaeology 11, 47-102. (eds.), 165-88.
Brochier, J.E., 1993. C::ayonii Tepesi. Domestication, rythmes Goldberg, P., D.T. Nash & M.D. Petraglia (eds.), 1993.
et environment au PPNB. Paleorient 19(2), 39-49. Formation Processes in Archaeological Context. (Mono-
Bullock, P., N. Fedoroff, A. Jongerius, G. Stoops & T. graphs in World Archaeology 17.) Madison (WI):
Tursina, 1985. Handbook for Soil Thin Section Descrip- Prehistory Press.
tion. Wolverhampton: Waine Research. Goldberg, P., S. Lev-Yadun & 0. Bar-Yosef, 1994. Petro-
Canti, M., in press. An investigation of microscopic cal- graphic thin sections of archaeological sediments: a
careous spherulites from herbivore dungs. Journal of new method for palaeobotanical studies. Geo-
Archaeological Science. archaeology: an International Journal9(3), 243-57.
Courty, M.A., P. Goldberg & R.I. Macphail, 1989. Soils and Hall, A., 1987. In pursuit of the uneatable (some thoughts
Micromorphology in Archaeology. Cambridge: Cam- on the search for archaeological evidence of vegeta-
bridge University Press. bles). Archaeology in York: interim 12, 1-7.
Courty, M.A., R.I. Macphail & J. Wattez, 1991. Soil micro- Hodgson, J.M., 1976. Soil Survey Field Handbook. (Soil Sur-
morphological indicators of pastoralism: with spe- vey Technical Monograph 5.) Harpenden: Soil Survey.
cial reference to Arene Candide, Finale Ligure, Italy. Kingery, D.W., P.B. Vandiver & M. Prickett, 1988. The
Rivista di Studi Liguri A.L VII, 1-4, 127-50. beginnings of pyrotechnology, part II: production
Courty, M.A., P. Goldberg & R.I. Macphail, 1994. Ancient and use of lime and gypsum plaster in the pre-
people: lifestyles and cultural patterns, in Transac- pottery Neolithic Near East. Journal of Field Archaeol-
tions of the 15th World Congress of Soil Science, eds. L. ogy 15, 219-44.
Wilding & K. Oleshko. Acapulco: International So- Kramer, C., 1979. An archaeological view of a contempo-
ciety of Soil Science, 250-69. rary Kurdish village: domestic architecture, house-
Cutler, D.F., 1978. Applied Plant Anatomy. London: hold size, and wealth, in Ethnoarchaeology: Implications
Longman. of Ethnography for Archaeology, ed. C. Kramer. New
Davidson, D.A., S. Carter & T.A. Quine, 1992. An evalua- York (NY): Columbia University Press, 139-63.
tion of micromorphology as an aid to archaeological Limbrey, S., 1975. Soil Science and Archaeology. London:
interpretation. Geoarchaeology: an International Jour- Academic Press.
nal 7(1), 55-65. Macphail, R.I., 1994. The reworking of urban stratigraphy
de Meester, T.D., 1971. Highly Calcareous Lacustrine Soils in by human and natural processes, in Urban-rural
the Great Konya Basin, Turkey. (Agricultural Research Connexions: Perspectives from Environmental Archaeol-
Reports 752.) Wageningen: pudoc Centre for Agri- ogy, eds. A.R. Hall & H.K. Kenward. (Symposia of
cultural Publishing and Documentation. the Association for Environmental Archaeology 12.)
Driessen, P.M. & T.D. de Meester, 1969. Soils of the Cumra Oxford: Oxbow Books, 13-43.
area, Turkey. (Agricultural Research Reports 720.) Wa- Matthews, W., 1996. Micromorphological characterisation
geningen: pudoc Centre for Agricultural Publishing. and interpretation of occupation deposits and micro-
Evershed, R.P. & P.H. Bethell, in press. Application of stratigraphic sequences at Abu Salabikh, southern
multi-molecular biomarker techniques to the identi- Iraq, in Archaeological Sediments and Soils: Analysis,
fication of faecal material in archaeological soils and Interpretation and Management, eds. T. Barham, M.
sediments, in Archaeological Chemistry Symposium. Bates & R.I. Macphail. London: Archetype Books,
(ACS Symposium Series.) 41-74.

341
Wendy Matthews, Charles French, Timothy Lawrence & David Cutler

Matthews, W., n.d .. <::atalhi:iyiik 1995 archive report: sam- ton Spa: Salvo Print.
pling strategy. Rosen, A.M., 1986. Cities of Clay: the Geoarchaeology of Tells.
Matthews, W., C.A.I. French, M.K. Jones, T. Lawrence, & Chicago (IL): University Press of Chicago.
D.F. Cutler, in press. Activities inside the temple: Rosen, A.M., 1992. Preliminary identification of silica skel-
the microstratigraphic evidence, in The Dilmun Tem- etons from Near Eastern archaeological sites: an ana-
ple at Saar, eds. H.E.W. Crawford, R. Killick & J. tomical approach, in Phytolith Systematics, eds. G.
Moon. London: Kegan Paul International. Rapp & S.C. Mulholland. New York (NY): Plenum
Matthews, W. & J.N. Postgate, with S. Payne, M.P. Charles Press, 129--48.
& K. Dobney, 1994. The imprint of living in a Meso- Schiffer, M.B., 1983. Toward the identification of forma-
potamian city: questions and answers, in Whither tion processes. American Antiquity 48, 675-706.
Environmental Archaeology?, eds. R. Luff & P. Rowley Stewart, D.R.M. & J. Stewart, 1970. Food preference data
Conwy. Oxford: Oxbow Books, 171-212. by faecal analysis for African plains ungulates.
Mellaart, J., 1962. Excavations at <::atal Hiiyiik, first pre- Zoologica Africana 5, 115-29.
liminary report, 1961. Anatolian Studies 12,41-65. Time-Life Books, 1991. The Human Dawn. (Time Life His-
Mellaart, J., 1964. Excavations at <::atal Hiiyiik, third pre- tory of the World.) Amsterdam: Time Life Books.
liminary report, 1963. Anatolian Studies 14, 39-119. Wattez, J. & M.A. Courty, 1987. Micromorphology of ash
Mellaart, J., 1966. Excavations at <::atal Hiiyiik, fourth pre- of some plant remains, in Fedoroff et al. (eds.), 677-
liminary report, 1965. Anatolian Studies 16, 165-91. 83.
Mellaart, J., 1967. (:atal Hilyilk: a Neolithic Town in Anatolia. Wattez, J., M.A. Courty & R.I. Macphail, 1990. Burnt
London: Thames & Hudson. organo-mineral deposits related to animal and hu-
Moore, H.L., 1986. Space, Text and Gender: an Anthropologi- man activities in prehistoric caves, in Soil Micro-
cal Study of the Marakwet of Kenya. Cambridge: Cam- morphology: a Basic and Applied Science, ed. L.A.
bridge University Press. Douglas. Amsterdam: Elsevier, 431--40.
Moorey, P.R.S., 1994. Ancient Mesopotamian Materials and Weiss, H., M.A. Courty, W. Wetterstrom, F. Guichard, L.
Industries: the Archaeological Evidence. Oxford: Senior, R. Meadow & A. Curnow, 1993. The genesis
Clarendon Press. and collapse of third millennium north Mesopota-
Murphy, C.P., 1986. Thin Section Preparation of Soils and mian civilization. Science 261, 995-1004.
Sediments. Berkhampsted: A.B. Academic Publishers. Zohary, M., 1973. Geobotanical Foundations of the Middle
Norton, J., 1986. Building with Earth: a Handbook. Learning- East. Stuttgart: Gustav Fischer Verlag.

342
Chapter 16

Luminescence Dating of Mud Brick from ~atalhoyiik

Romola Parish

The technique of luminescence has been widely used the sieve. The sample was then washed several times
for dating a range of fired materials such as ceramics with distilled water, then acetone in order to remove
and hearthstones, and for sedimentary deposits all traces of chemicals from the previous treatments.
which have been bleached by exposure to sunlight. The sample was then dried. The 2-11 f..LID fraction
The fact that the fine silt component of mud brick was isolated by settling in 6 crn3 acetone; first for 2
contains quartz and feldspar which are suitable for rnins. after which the supernatant was kept, then for
luminescence dating, and that the process of manu- 20 rnins. when the supernatant was discarded.
facture involves exposure of completed bricks to in- Twenty-four aluminium discs (1 ern diameter) were
tense bleaching, indicates that such materials would cleaned and each one was placed the base of a flat
be directly datable by luminescence. This applica- bottomed tube. The sample was suspended in ac-
tion represents an important development in Mid- etone and 1 rnl added to each tube. These were left
dle and Near Eastern archaeology where most of the undisturbed to evaporate off leaving the grains ad-
buildings can only be dated indirectly by artefacts hered to the discs.
and documentary materials, or by C-14 on timbers
which are often rare. Measurement
A pilot study was carried out dating two mud-
brick samples taken from <::atalhOyiik in 1994. One The ED was determined by the additive dose tech-
sample was from level XII, the lowest levels reached nique. Four discs of the natural signal were meas-
by Mellaart's excavations in the 1960s, and the sec- ured, and four discs for each of 5 additive dose points.
ond from level III/IV, which are younger. Uncalibrated Irradiation was by a manually-operated Sr90 source,
radiocarbon dates exist for comparison for level III/ delivering 14.3 Gy /min. Samples were measured in
IV but not for level XII (Mellaart 1967). a Rise automated reader, using a preheat of 220°C
for 5 rnins. for each disc and normalization by a 0.5
Preparation sec. exposure to the green light source and an expo-
sure of 30 sees. to stimulate the OSL signal. Detec-
Initially a sample was taken from the interior of the tion was by a filtered photomultiplier tube with a
brick, however the apparent age was much older combination of GG420 blue/ green filters for excita-
than expected if the interior had not been adequately tion and U340 for detection.
exposed during manufacture (see results below). A
second sample was taken by carefully separating the Dosimetry
mortar from the brick surface and scraping off the
outer clay surface, which would certainly have been The dose rate was determined by alpha counting
bleached during drying of the brick. and thick source beta counting to determine the U
Fine-grained samples were prepared from each and Th contents and the total beta activity. The K%
brick using standard techniques. These involved first content was derived from the U and Th ppm (Narnbi
deflocculating the sample with Calgon, removing & Aitken 1986). Water content was measured as a
organic material with H 20 2 and carbonates with percentage of dry weight. The cosmic component
H 2S04 • The sample was washed through a 64 f..LID was calculated to be 0.2 f..LGy (Prescott & Hutton
sieve, retaining the material which passed through 1988).

343
Romola Parish

Results

The results are shown below of the samples from the These initial results are very encouraging, indicating
interior of the brick, the surface samples of both that the technique holds significant potential for es-
bricks, and the C-14 dates from Mellaart (1967) which tablishing a chronology for the site based on direct
have been calibrated for direct comparison. dating of the buildings.

Sample OSL age ED C-14 date BP


(surface) (interior) (Mellaart 1967)
References

level XII 9500±1500 14900±3360 not yet dated Mellaart, J., 1967. (:atalhoyiik: a Neolithic Town in Anatolia.
level III/IV 8000±1750 not dated 8695±202/8595±112 London: Thames & Hudson.
Nambi, K.S.V. & M.J. Aitken, 1986. Annual dose conver-
Sample U (ppm) Th (ppm) K% water%
sion factors for TL and ESR dating. Archaeometry 28,
level XII 2.07±0.32 7.85±1.03 2.16±0.28 1.30±0.5 202-5.
level III/IV 2.85±0.39 7.53±0.93 2.12±0.27 1.02±0.5 Prescott, J.R. & J.T. Hutton, 1988. Cosmic ray and gamma
ray dosimetry for TL and ESR dating. Nuclear Tracks
An estimated a-value of 0.05±0.01 was used for both samples. and Radiation Measurements 14(1 /2), 223-7.

344
Chapter 17

Interim Dendrochronological Progress Report 1995/6

Peter Kuniholm & Maryanne Newton

The work of the Aegean Dendrochronology Project the earlier level VII. Note, however, that with the
has been for 23 years the building of long tree-ring absence of bark the end dates for the different levels
chronologies for the eastern half of the Mediterra- are not in fall-back order as one might expect.
nean with the aim of helping to bring some kind of Nevertheless, several interesting conclusions
rational order to Aegean and Near Eastern chronol- may be drawn. Previous reports state that all wood
ogy from the Neolithic to the present (Kuniholm from contexts after level XI was worked (Mellaart
1993; Kuniholm in press; Kuniholm & Kromer in 1966; Todd 1976). Fortunately, this is apparently not
press; Kuniholm & Newton 1990; Kuniholm & Striker the case. Sample CTL-16 (a post from level VII) was
1983; 1987). The following note sums up the collected in 1994 from an exposed scarp contiguous
dendrochronological situation for <::atalhoyiik and to Mellaart' s level VII. This post was unworked, i.e.
other Neolithic sites as of February 1996. it was stripped of bark but not otherwise shaped.
This provides a terminal year designation, impor-
KEY SITES: <::atalhOyiik (570 and 60 year chronolo- tant for dendrochronology since only with terminal
gies); Ko~k Hoyiik (224 year chronology and grow- years can the felling dates of timbers from different
ing); A~1kh Hoyiik; Hallan <::emi; <::ayonii; Hacilar; contexts be related to each other with a precision of
Kuru<;ay; Kastoria/Dispilio (300+ Late Neolithic sam- one year. The terminal year for CTL-16 at relative
ples expected in 1996); Yumuktepe/Mersin; year 1437 is 132 years earlier than the last preserved
Bademagac; A~agipmar; Ihpmar; Kumtepe; Amuq ring (relative year 1569) of CTL-6 from level VIB.
Survey. This tells us that the difference, in calendar years,
between the two timbers is at least 132 years, and
Central to any dating of the Neolithic in Anatolia, suggests that the life span of level VII and part of
either dendrochronological or radiocarbon, is VIB is likewise at least 132 years. The observation of
<::atalhOyiik. With completion of the analysis of the CTL-16 in the field furthermore suggests that care-
carbonized wood from the 1960s' excavations, we fully collected samples from the renewed excava-
have built a 570 year tree-ring chronology for the tions can be expected to preserve terminal growth
site, against which chronologies from the renewed rings (i.e. bark dates), thus enabling, through
excavations, as well as chronologies from roughly dendrochronology, annual resolution to the intra-
contemporary sites, can be compared. site chronology.
Some five hundred charcoal fragments left over The long tree-ring chronology from <::atalhoyiik
from the radiocarbon work at MASCA/University is equally helpful for refinements in external dating.
of Pennsylvania (Stuckenrath & Ralph 1965; Stucken- Many of Mellaart' s radiocarbon dates were on wood
rath & Lawn 1969; Mellaart 1967) were kindly fur- that we can now demonstrate to have been long-
nished to us by the excavator, James Mellaart. The lived. The dates obtained in the 1960s did not take
570 year tree-ring chronology, with a possible 60 into account the age of the tree, sometimes leading
year extension on the early end, is internally well (as Mellaart has pointed out) to erroneously early
replicated, comprising nine trees from proveniences dates. By knowing which rings are actually dated,
spanning Mellaart' s levels IV through VII. The graph we hope, by combining dendrochronology with ra-
in Figure 17.1 presents the individual series in diocarbon analysis, to pinpoint the absolute dating
stratigraphic order from the more recent level IV to of the site.

345
Peter Ian Kuniholm & Maryanne W. Newton

Catal Hoyi.ik Juniper Tree-Ring Sequences


Data Curves normalized by a twenty year running mean

CTL-4 (level VIB, House 28, Post)

CTL.Q (Level VI B, Shrine 1, Rootbeam)

CTL-12 (Provenience unknown)

CTL-17 (Level VII , Rootbeam)


Figure 17.1. 570 year
juniper ring-chronology
from several occupation
levels at Neolithic
1000 1100 1200 1300 1400 1500 1600 (atalhoyiik . (Material
Relative Years courtesy f. Mellaart.)

8400
-200 1 200
8300 DENDROCHRONOLOGICAL 8300
0_

co 8200
co ()~
'\"'
Cotol Hoyuk
8200
v Ring 1 = 7020 B.C.
())
8100 8100
CL
<(
w 8000 8000
>--
z 7900 7900
0
co 7800 7800
CL
<(
u 7700
- - - INTCAL93
7700
0
0 7600
<(
CL 7500
Col B.C.
-7200 -7000 -6800 -6600 -6400

346
Dendrochronological Progress Report 1995 I 6

Since our dendrochronology of the eastern Humanities, the National Science Foundation, the
Mediterranean is not complete back to the eighth Malcolm H. Wiener Foundation, the National Geo-
millennium, and we cannot yet convert our relative graphic Society, the Samuel H. Kress Foundation,
dating system to calendar years, we have submitted the Wenner-Gren Foundation for Anthropological
a series of decadal samples from the <;atalhoyiik Research, and individual Patrons of the Aegean
chronology to the AMS (Accelerator Mass Spectro- Dendrochronology Project. We thank the radiocarbon
metry) Laboratory at the University of Arizona for laboratories at Heidelberg (conventional) and Ari-
wiggle-matching against the radiocarbon curve. Lots zona (accelerator) for their help with wiggle-matching.
A and B from the extreme ends of the <;atalhoyiik
sequence (Fig. 17.2) fit roughly at 7020±50 years BC References
and 6500±100 years Be. We have selected eight more
decadal samples (Lots C through J) to see whether Kuniholm, P.I., 1993. A date-list for Bronze Age and Iron
the results obtained by the Arizona physicists will Age monuments based on combined dendro-
allow us to pin down the <;atalhoyiik chronology chronological and radiocarbon evidence, in Aspects
more precisely. When the results are available we ex- of Art and Iconography: Anatolia and its Neighbors:
pect to be able to report a wiggle-matched series that Studies in Honor of Nimet Ozga~, eds. M.J. Mellink, E.
is pinned down to ±40 years (±30 if we are fortunate). Porada & T. Ozgii~. Ankara: Tiirk Tarih Kurumu
Bastmevi, 371-3.
The dendrochronology of at least 11 other
Kuniholm, P.I., in press. Long tree-ring chronologies for
Aegean and Near Eastern Neolithic sites is being the eastern Mediterranean, to appear in the theme
actively investigated. Last summer we collected 222 volume of the 29th International Symposium on
samples from Late Neolithic Ko~k Hoyiik (Oztan & Archaeometry. Paper delivered in May, 1994.
Faydal in press), and the ring sequence thus far is Kuniholm, P.I. & B. Kromer, in press. A 285-year wiggle-
224 years and increasing as more samples are meas- matched chronology for the Early Bronze Age at
ured. It may eventually be long enough to overlap Kiten, Bulgaria, in Studia Praehistorica 15 in Memoriam
with sequences from Chalcolithic Can Hasan, or James Harvey Gau. Sofia: Academy of Sciences.
roughly contemporary sequences from Hacilar, Kuniholm, P.I. & M.W. Newton, 1990. A 677 year long
tree-ring chronology for the Middle Bronze Age, in
Kurw;ay, and Ihpmar. We expect some day to be
Studies in Honor of Tahsin Ozga~, eds. K. Emre, B.
able to link these sites and <;atalhoyiik together Hrouda, M.J. Mellink & N. Ozgii~. Ankara: Tiirk
dendrochronologically, forming a single regional Tarih Kurumu Bastmevi, 279-93.
chronology for the Anatolian Neolithic. Kuniholm, P.I. & C.L. Striker, 1983. Dendrochronological
Even at this preliminary stage, the tree-ring investigations in Greece: 1977-1982. Journal of Field
dates for <;atalhoyiik offer a remarkable improve- Archaeology 10(4), 411-20.
ment on what has been previously available for any Kuniholm, P.I. & C.L. Striker, 1987. Dendrochronological
stratified Neolithic site (cf. van Andel & Runnels investigations in the Aegean and neighboring re-
1995, where their best starting dates for the Thessalian gions, 1983-1986. Journal of Field Archaeology 14(4),
385-98.
Neolithic are EN 7000, MN 6000, LN 5500). We think
Mellaart, J., 1966. Excavations at <;:a tal Hiiyiik, fourth pre-
tree-ring dates can improve significantly our under- liminary report, 1965. Anatolian Studies 16, 165-91.
standing of the time involved in the transition from Mellaart, J., 1967. Catal Huyuk: a Neolithic Town in Anatolia.
hunting and gathering to settled communities, of the New York (NY): McGraw-Hill.
dispersal of such communities, and potentially of Oztan, A. & E. Faydal, in press, VI. Muzeleri Kurtarma
the rates of change within them. We thus look for- Kazlian Semineri. Ankara: Tiirk Tarih Kurumu.
ward to a level of precision in archaeological dating Stuckenrath, R., Jr & B. Lawn, 1969. University of Pennsyl-
in the eastern Mediterranean hitherto inconceivable. vania radiocarbon dates XI. Radiocarbon 11(1), 150-
62.
Stuckenrath, R., Jr & E.K. Ralph, 1965. University of Penn-
Acknowledgements
sylvania radiocarbon dates VIII. Radiocarbon 7,187-99.
Todd, LA., 1976. c;atal Hayak in Perspective. New York
The Malcolm and Carolyn Wiener Laboratory for (NY): Cummings.
Aegean and Near Eastern Dendrochronology is van Andel, T.H. & C.N. Runnels, 1995. The earliest farm-
supported by the National Endowment for the ers in Europe, Antiquity 69,481-500.

347
Chapter 18

<;atalhoyiik: the Anthropology of an Archaeological


Presence

David Shankland

Between 1992 and 1995, whilst Assistant Director, empiricists, oblivious that there is such a thing as
then Acting Director, of the British Institute of Ar- social theory at all. This sometimes gives rise to an
chaeology at Ankara, much of my time was taken up over-emphasis on the social creation of reality in the
with the re-opening of <;atalhoyiik, an event which one camp and on the minute and bare presentation
caused ripples throughout the archaeological com- of the archaeological record in the other. 3 It is axi-
munity in Turkey. However, as well as possessing omatic today that an anthropologist must examine
an administrative interest, these developments in- the power relations between researcher and re-
trigued me professionally. I am trained as a social searched, be aware that the creation of knowledge
anthropologist in Britain. In Britain, unlike America, may be gender-influenced, that the investigation may
departments of anthropology and archaeology are have a profound effect on the community with whom
often separate. Perhaps trammelled by this institu- they are living, and that their findings may be used
tional separation, it appeared that anthropologists for political purposes. However, some anthropolo-
and archaeologists from Britain who come to work gists seem to have been able to accept the need to
in Turkey often have very different ways of approach- face these problems without losing faith in the em-
ing the study of human societies. pirical basis of their results. This willingness to sit on
In spite of these differences, it seemed to me the fence: to accept both that the world is real and
that social anthropologists might have certain assets that knowledge is invariably context-bound/ I also
if pressed into an archaeological enquiry. They are thought might be useful, especially when applied to
good at social contexts: at appreciating the inter- an investigation of the way that archaeologists go
connectedness of the institutions of social life. They about obtaining their data.
are also adept at sensing the difference between what These impressions were gained through work-
people say goes on, what people say should go on, ing with archaeologists in an intense though un-
and what is actually found on the ground. This ex- systematic way. In order to put them to the test, I
perience could be potentially useful, for example, decided to search for a village in which to conduct
when attempting to use comparative DNA analysis fieldwork, but one close to a current excavation. Dr
of domestic bone remains to speculate how a kin- Hodder kindly gave his encouragement to the enter-
ship structure may have been: anthropologists have prise. Accordingly, I conducted fieldwork in 1995 at
often faced the problem that whilst kinship systems Kii<;iikkoy, the closest village to <;atalhOyiik. Whilst
are by definition based on biological relations be- I followed the normal practice of living with the com-
tween people, the fit between the biological and the munity (who generously provided me with a house),
social patterns is very difficult to clarify. 1 fieldwork was deliberately timed to coincide with the
More abstractly, I have a sense that anthropolo- third recent season of fieldwork at the site, which took
gists as a whole have reacted rather differently from place between 15th September and 30th October 1995.
archaeologists to the contemporary preoccupation The remarks below are based on that season's work,
with the conditions within which research is con- but they must inevitably be preliminary. They are
ducted.2 Archaeologists seem often split between offered tentatively, with the full knowledge that they
politically-correct post-modernist relativists, and will need revising and refining in the future.

349
David Shankland

Kiic;iikkoy is supplied by the household, and the household


live off the products of their labour, whether by
Kiic;iikkoy is a village of some one hundred house- direct consumption or by disposing of them for
holds set in the midst of the Konya plain, about cash.
twelve kilometres away from <;umra, the nearest In the centre of the village lies a mosque, whose
sub-province centre. The villagers' main economic dome and minaret can be seen from afar across the
activity is agriculture, there being little industry and fields. Near the mosque is also a primary school, and
as yet no significant tourism in the immediate area. opposite that, a cemetery. There are no large public
The winters are cool and wet, but the summers are buildings but for the mosque, though facing it are a
long and arid. Their fields have been systematically row of grocers, where dry goods such as soap, razor
irrigated since early this century, though for the last blades and cigarettes are sold, and a tea-house, which
three years the water supply has been disrupted, in opens in the afternoon and continues service until
part due to a major new irrigation system in the far into the night. Women avoid walking across this
process of construction (Konya Ovasz Sulama Projesi). open space when they can, keeping to the back al-
In compensation, individual wells increasingly draw leys between the houses. Men occupy it freely, the
ground water to spray directly on to crops. The wa- older, more pious men sit around the mosque, and
ter thus obtained is sufficient at present, but the the younger men the tea-house. Overall, when cir-
water table is falling rapidly. cumstances permit, men occupy the public spaces of
The land is flat and extremely fruitful, giving the village: they operate machinery, plough the fields
yields of wheat up to I in 40. Most households own and tend its crops, and women work in the gardens,
some land, a typical holding being about 30 doniim 5 look after the homes and children.
distributed unfenced around the village settlement. In short, Kiic;iikkoy is in many ways a typical
As well as wheat, other primary crops are sugar central Anatolian Turkish village, fortunate that it is
beet, melons and beans. Many households also main- the possessor of such very rich fields and adequate
tain flocks of sheep or herds of cattle, the main em- water supply, better integrated into the national
phasis being on sheep. The number of beasts owned economy than villages further to the east, but still
by a household varies greatly, but is often large: two less exposed to the outside than any villages in the
hundred is regarded as being a good-sized flock, but Aegean coast, pious but not aggressively so. Many
by no means unusual. The flocks are culled regu- of the villagers have migrated to Germany, Konya or
larly, and the beasts so selected sold in local markets Istanbul, but those who are left are actively pursuing
in <;umra or Konya. household-based agricultural production more lu-
Nearly every house is surrounded by a large crative but similar organizationally to the traditional
wall, within which are usually found a barn to keep subsistence economy practised so commonly by other
straw, a separate out-house in which to keep sheep Anatolian settlements until recently. 6
and a garden in which are grown tomatoes, cucum-
bers, apples, figs, grapes, large radishes, melons and Kiic;iikkoy and ~atalhoyiik
marrows. Some household complexes include also a
bryer to rear cattle, and a guest room set apart from The villagers have had a long association with
the main dwelling in which a man may entertain <;atalhoyiik, which lies within the village boundary,
visitors without disturbing his household. Whilst about a kilometre away from the main settlement.
we had insufficient time to conduct a detailed sur- When the site was first surveyed and dug in 1961-
vey, it is clear that the great majority of households 65, the excavation team lived in the village school
are patrilineal and patrilocal, and are usually com- and in the school-teachers' lodgings. Though James
posed of a married man, his wife, and their children, Mellaart brought his own skilled workmen to work
often with one or both of the man's parents. It is at the site, some people at the village were also em-
unusual to find people who are not related living in ployed there at that time. The villagers recall the
the same household, though the occasional shep- finds being shown in the schoolhouse as they were
herd, usually a migrant from a poorer area of the dug up, and some men remember as children going
country, may be offered accommodation in the to the site and watching as the excavation took place.
bryer during the winter months. Whilst house- During the years that the site has been untouched,
holds interact and cooperate with their relatives its watchman was drawn from the village, and, on
and neighbours, each acts as a largely independ- his retiring in 1992, he was replaced by two younger
ent agricultural unit: nearly all the labour required men, both also from Kiic;iikkoy.

350
Anthropology of an Archaeological Presence

Since the site has been re-opened in 1993, there A template


have been on average about fifteen workmen em-
ployed at the site during the season. The present When approaching such a small-scale community, it
dig-team have lived at C::umra, twelve kilometres is very frequent that an anthropologist examines one
away during the excavation, but relations are en- particular institution or aspect of social life and at-
tirely cordial between the village and the excavators. tempts to show in as much detail as they can the way
The first initiation into the life of a community can that it fits in with the activities of the community
be misleading, but at the end of the period of field- where they are researching. It would seem in theory
work in 1995, it seemed to me that in spite of the possible that a detailed picture of the way of life in
cordiality, and welcome, that the villagers have the village related to one important part of their
shown toward the excavation, the overall inclination agricultural activity might provide a template with
of those villagers who have not become part of the which information from the site can be compared
excavation team is one of polite but rather mild curi- when it is available. By 'template', I do not mean a
osity: that the site is there, and it is perhaps interest- direct comparison between activities, or between the
ing, but that, so far, it plays little significant part in material culture of the excavation and the village: I
the life of the village. mean rather an investigation into every aspect of
care and upkeep of their beasts from birth to death,
Lines of enquiry the division of labour (including gender division)
and ownership when nurturing young, the division
Whilst time was short, two broad lines of enquiry of labour when culling, pasturing and milking, and
were pursued in 1995. The first, a consideration of the role that the animals play in ritual activity, and
whether village life can help our understanding of finally the use to which their products are put after
life at C::atalhoyiik; the second, an examination of the death, showing in each case how they inter-relate
setting of the archaeological research at the site, what with the other activities in the village which are
might be called the anthropology of archaeology. By going on at that time. The end results would be a
'village life', I mean any aspect of the village which series of patterns of activity, showing where people
may be of use. This includes information that the at different times of the day, at different times of the
villagers may be able to provide on the whereabouts year conducted what activity, the focus being al-
of other sites in the area. Whilst prosaic, it is an area ways the relationship between animals and humans,
where anthropology may be of the greatest assist- but the ethnography to include also a wider indica-
ance. Anthropologists are trained to create a relation tion of the activities of the community at any point
of trust between a community and themselves, one in these cycles.
based on living in the community and on speaking There is little detailed, systematic information
their language. Archaeologists rarely have the op- published on the daily life of an Anatolian village,
portunity to become fluent in the language of the therefore such a study has the definite merit of being
people where they excavate, cultivating rather a serv- a potentially useful contribution to social anthropol-
iceable patois in order to communicate with work- ogy.7 Theoretically, as an investigation in social an-
men and officials. Nor do they have often the thropology and archaeology, it could be justified
inclination to billet in villages for long periods, con- like this: using the present to understand the distant
structing instead dig-houses in which they can exist past is today largely the province of ethno-
in isolation from the communities among whom they archaeology, having been driven from social anthro-
are excavating. This separation and loss of intimacy pology by the successful efforts of the early modern
is particularly significant in Anatolia, where the rich- practitioners: Malinowski, Radcliffe-Brown and
ness of the archaeological record is only matched by Evans-Pritchard in particular. 8 That battle may have
the oft-found desire of the villagers to hide what been necessary. Today, however, this need is over
they can from authority, either from caution or the and the study of the distant past can be re-incorpo-
aim to exploit the sites themselves. Indeed, whilst at rated into the discipline without loosing the theo-
Kii<;iikkoy, I was surprised at the enthusiasm and retical insights which have been gained.9 One (though
the regularity with which the villagers showed me certainly not the only) way that this re-incorporation
otherwise quite inconspicuous sites in their fields, could be done is to stress that the pattern-like nature
adding that they would not dream of being so con- of social events in rural communities is comparable
fiding with someone with whom they could not com- by epoch. Life at C::atalhOyiik was clearly connected
municate directly. with animals in many respects. In order for them to

351
David Shankland

maintain this aspect of their society, there must have altered, whether they wished for the excavation to
been certain regular, cyclic activities within their com- take place at all, whether political battles have been
munity which were concerned with animals. Some fought over the heritage being explored, how the
of these activities will be discernable in the archaeo- knowledge from the site has been used, who has
logical record. By presenting a model of how a benefitted from the excavation, all this is unrecorded
present-day society runs this aspect of their lives, and unexplained.
and by having to hand the infinitely suggestive re- These considerations are acute in the case of
sources of a living, local society to supply clarifica- C::atalhOyiik, where the finds at the site relate di-
tion of any particular point, the understanding of rectly to debates concerning nationalism, cultural
the way that this was achieved at C::atalhoyiik can be heritage and politics in Turkey as a whole. The mas-
forwarded. 10 Indeed, if it turns out that the people at sive (and in retrospect, still astonishing) changes
C::atalhoyiik lived in sedentary, repeated homogenous which Atatiirk implemented between 1923 and his
units, where the division of economic, kinship and death in 1937 transformed the Ottoman Empire into
ritual labour is largely dominated by gender and not a secular republic. In an attempt to provide theRe-
by intra-sex professional specialization or class, public's citizens, of very diverse background, with a
then, whilst the ideological differences between the new collective identity, Atatiirk stressed the Ana-
two communities may be great, the similarities in tolian, rather than the Islamic roots of this new na-
social and economic organization may be striking tionY The systematic excavation of the Hittite,
indeed. Neolithic and earlier periods combined with the es-
tablishment of Institutes of Pre-History in the Uni-
The re-opening of ~atalhoyiik versity of Istanbul and the opening of the Hittite
Museum in Ankara (later the Museum of Anatolian
The enquiry outlined above has to recommend it Civilizations) which the finds from C::atalhoyiik have
that it contains the possibility of its own failure: it helped to stock, all testify to this emphasis.
will have been unsuccessful in its primary aim, how- Today, whilst Republican Turkey has changed
ever disappointing that may be, if in fact it is found greatly since its inception, the protection and inves-
to be not at all relevant to the archaeological investi- tigation of its diverse cultural heritage is still taken
gation at the site. Investigating the setting in which very seriously. There is an extensive network of ar-
research takes place, our second line of enquiry, is chaeological museums throughout the country,
more difficult to justify in the sense that even after within which are recorded and displayed the finds
every aspect of the operation of the site has been of their respective provinces. The General-Directo-
examined from every point of view, the way that rate of Monuments and Museums, part of the Minis-
such knowledge would change or enhance interpre- try of Culture, coordinates these museums, and
tation of the results obtained there is not yet clear. devotes a considerable part of its resources to exca-
It may be sufficient encouragement, however, vations run by them. The General-Directorate also
to note that at present the circumstances in which considers all applications by foreign teams who wish
archaeological research is conducted are shrouded to work in Turkey, checks that they have sufficient
in obscurity. Excavations in Turkey started more resources to maintain their projects, encourages them
than a century ago and are continuing apace today. to publish their results and to consider the conserva-
The results of these labours are presented in site tion aspects of any intrusive work which is done.
reports, monographs and papers. The finds are placed C::atalhoyiik is very much a part of this overall,
in museums within Turkey. Beyond this we know supervised endeavour, and the site's continuing im-
almost nothing of the circumstances or consequences portance is indicated that in the Anatolian Civiliza-
of these digs. Anecdotal evidence tells of a complex tions Museum, the amount of space given to the site
juggling of academic renown and influence, finan- has been increased, and a female figurine from it
cial strictures, local political prestige, national eco- featured on the front of a recent catalogue (Fig. 18.1).
nomic and political considerations, strikes, good and An evaluation of Atatiirk' s reforms lies outside
bad cooks, and struggles with local labour forces. By the scope of this chapter. However, it is certain that
the time research has reached the official presenta- he struck an answering chord in many people, not
tion stage all is streamlined into concise explanation just those in authority. Even in the remotest villages,
accompanied by slides showing neat trenches and people recall learning about the site from their school
labelled artefacts. Whether the economic circum- text books or from television programmes about
stances of the local people have been radically Anatolia's past, and appreciate its importance. Finds

352
Anthropology of an Archaeological Presence

from the site are used by people of different persua-


sion all over Turkey, feminist, left-wing, intellectual
and Republican to demonstrate the truth of their
way of viewing the world. Media interest in its re-
opening has been intense. It is likely that interest
will grow even more as the excavation moves into
areas of the mound which have not yet been explored.
Independently of the media, bureaucracy and
large urban centres, <::umra, the local town, also
wishes to link its name with the site, and indeed, has
done so already. The town municipality have three
times held a '<::atalhoyuk Melon Festival': twice dur-
ing the late seventies, and for a final time in 1981.
The people in the town who organized this festival,
some of whom I was able to meet in 1995, explain
that they sat down, and tried to envisage the things
for which <::umra was most famous. They concluded
that these were <::atalhoyuk and melons. They re-
paired the back-road from Konya to <::umra, which
passes by the site. They invited dignitaries from An-
kara and Istanbul to participate in celebrations last-
ing two days. During the course of these days, they
held an opening ceremony at which melons were
handed out free. They also held sports races, a com-
petition for the best-tasting melons, and brought in
folklore groups and musicians from the surround-
ing area. For the invited dignitaries, they also ar-
ranged a visit to the site, which culminated with a
lecture given by the then head of the Director-Gen-
eral of Antiquities. Prizes for the races consisted of a Figure 18.1. Imagery from (atalhoyuk on the front
medal with a figure of a seated figurine from cover of a recent catalogue of the Anatolian Civilizations
<::atalhoyuk juxtaposed with a melon (see Fig. 18.2). Museum in Ankara.
Today, the same design is still used by a youth radio
station, <::umra radio (Fig. 18.3).
All this means that running the site will have to
be a tactful balancing between sometimes conflict-
ing interests. It also means that the physical conse-
quences of the excavation, for example where and
how the finds are housed, will be potentially a con-
troversial issue. The site itself will receive a growing
number of visitors, and permanent facilities will con-
tinue to be built there to meet the demand of re-
searchers and tourists. The local population also will
perforce be involved in this development, and it is to
them that we now turn.

Excavations and local villages

Whilst not all archaeological sites integrate with their


host communities in the same way, it is likely that
they often follow a pattern. If not known already,

Figure 18.2. Medal from the (umra melon festival.

353
David Shankland

development begins. Even so, the way small rural


communities react to social change is incredibly com-
plex, and it is important not to over-simplify the
way any intrusion leads to different forms of adjust-
ment and accommodation. 12
At <:;:atalhoyuk, whilst work is now being un-
dertaken to develop the site in various ways, includ-
ing the construction of permanent accommodation
and research facilities, the site was little permanently
changed by the presence of the archaeologists. Only
a small fraction of the site was dug. A fence was
built around the mound, and a small, though rather
intrusive, watchman's house constructed next to the
mound. Tourists did not come in large numbers, and
those who did were able to look at the remains of the
early trenches, and they had no opportunity to spend
money. Now that the site has been opened up again,
the effect on the local community is still economi-
cally rather small. The working season has been short,
only six weeks. The fifteen or so workers who are
drawn from the village represent almost the sole
Figure 18.3. Logo of the C::umra youth radio station .
extent of its surplus labour: the remainder of men of
suitable age being either busy in their fields or hav-
the site is identified through survey, or through a ing emigrated. Whilst the prospect of cash is always
local museum alerting the authorities to the pres- welcome, and there are always a few poor folk, the
ence of artefactual evidence which is worth preserv- village's comparative wealth means that the money
ing . Excavation begins. If the site is large, or coming in through workmen's wages is of little sig-
interesting enough, excavation carries on for a nificance in the village economy as a whole. This, of
number of years. A dig-house may be erected at that course, may change very quickly if land needs ex-
time, and a watchman is appointed. If the site ap- propriating, or if the villagers perceive a chance to
pears important, it begins to gain close attention invest in service industries around the site.
from the authorities and the press as a potentially
significant part of the nation's cultural heritage. Fa- The villagers' and the archaeologists' world-views
cilities and even a museum may be erected at the compared
site. If it is attractive and well-placed it becomes
visited by tourists and, as it takes its place in tour The ideological aspects of the villages' interaction
guide schemes, it may draw an ever increasing with the excavators and the research are difficult to
number of people. Finally, large hotels provide for a describe succinctly. However, it is interesting to con-
constant stream of visitors to the site, particularly if trast the two world-views. Archaeologists come from
attracted by the presence of incidental natural at- outside the area, excavate and evaluate every part of
tractions, such as the Christian Churches at Capa- the mound. An intensive survey around the site at-
doccia or the Roman remains at Pamukkale. tempts to do likewise with the other sites and mounds
Different people and organizations become in- in the area. Ultimately, and inevitably, the archae-
volved, therefore, at different stages in a site's devel- ologists will sift, ignore and overlook much of the
opment: the archaeologists from the outset, the press evidence of the past in the area which does not ad-
a little later, senior bureaucrats and politicians later dress their immediate enquiry, but those which they
still when any national significance is clear, and fi- do take note of will be used to support the archaeolo-
nally investors and businessmen. The local village gists' collective view of the world and its past epochs.
inhabitants are involved from the very outset of It is easy to overlook that the sites lie in the
the project. They will be acutely sensitive to any villagers' home environment, where they have lived
opportunities that such direct contact with the out- for generations, and that, just as do the archaeolo-
side world offers, though I suspect that the effect on gists, they possess a coherent sense of the past of the
their community is often rather small until touristic area, invest some of the local mounds with meaning

354
Anthropology of an Archaeological Presence

and ignore others and, those mounds that they do rabbit, but did not think much about it. When they
take notice of, interpret in such a way that fits in tried to dig, they found that their way was barred by
with their collective conception of the past. For ex- a huge stone, and when they tried to get past this,
ample, the question of a possible link between the they heard voices howling telling them to stop. They
present inhabitants of Kii<;iikkoy and the people who gave up. My informant explained that the white rab-
lived at <;atalhoyiik occasionally arises. Some men, bit was a talisman, placed there to protect the mound
are reluctant to feel that they, as Muslims, should be by those who used to live by it. If they had shot it
linked to the figurative art at <;atalhoyiik. Rather, before they had started, then there would have been
they say that Greeks (Rum) used to possess fields in no impediment to their digging up the mound.
the area, and that though these people departed for
Greece as part of the population exchanges of the Interaction between the villagers and the
twenties, they may be the true descendants of the excavators
people who used to live at the mound.
In contrast, the villagers positively affirm a con- We have noted that the villagers do not usually show
nection with remains which partly overlap with the an intense curiosity as to the results which will come
site at <;atalhoyiik, remains which we would regard from the site. In part, this reluctance appears to be
as being part of a Byzantine settlement. They refer to due to a strong anti-intellectual streak in the village,
the area where they are found as 'old village', Eskikoy, one which discourages systematic speculation about
and say that the people of Kii<;iikkoy are the de- the past. For example, a man told me that he went to
scendants of that community. The story as it is usu- the site before it was discovered by survey, and no-
ally told goes as follows: at the time when Anatolia ticed that there were old artefacts there. He thought
was occupied by small, nomadic Islamic states, there that they might come from the Stone Age (ta~ devri),
used to live there a leader (bey) and his people. He but no-one was particularly interested in this thought.
had two sons. The younger of those two sons fell out When the excavation began, he was shown to be
with his elder brother, and left the community. He correct in his speculation, and told his friends, 'Look,
moved to the site now known as Kii<;iikkoy, and set I was right!', whereupon they laughed at him, and
up home there. When some time had past, his father said, 'What a philosopher we have among us!'.
called him back, and told him that it was time to Any way of thinking must filter the reception
make peace with his elder brother. He did so. Ac- of new views otherwise it would be unable to retain
cordingly, his father said that there was now no cohesion at all. To this extent, it is not surprising that
impediment to his moving back permanently to his such checks on enquiry are part of the village life.
community. The younger refused, saying that he However, it is important to note just how very co-
had now established himself with house, horses and herent the overall collective memory of the villagers'
men in his new settlement. Kii<;iikkoy today takes its past remains, even in the face of the radically differ-
name after the younger brother who was its founder. 13 ent approach of the archaeologists. I think that this
The villagers often add that hindsight has shown stems in great part from the fact that they are al-
their forebear to be justified in this decision, because ready aware of the multiplicity of past civilizations
Kii<;iikkoy has grown and flourished, whereas in their land and have linked these already in a
Eskikoy has fallen into ruins. coherent way with their sense of being Islamic. An-
Other sites have a distinct place in a wider cos- swers are already in place which cope well with the
mology which includes the supernatural, and the challenge brought by the presence of archaeologists
spirits of those gone. Many of the larger mounds are in their midsts.
regarded as being protected by the previous inhabit- That the villagers are not deeply affected by the
ants of the area. Several men told me of occasions findings of the site may seem unimportant: in any
when they have tried to dig into them but were case the archaeological presence is intermittent, and
prevented by a feeling of terror coming across them, the villagers have to get on with their lives whether
or by a terrible rushing wind which nearly blew the academics are there or not. However, this issue
them off their feet. They link these frequently to a is more important than might first appear. The ar-
talisman. For example, one man explained that his chaeologists are few, but they are an active part of a
father decided to investigate a local mound, along- rational, enlightenment world-view which is very
side which was a ruined house. He took along with close to that originally conceived, and imposed on
him a friend to help him dig, and a mosque hoca to the population of Turkey by the Kemalist revolu-
lend spiritual support. On the way, they saw a white tion.14 Administrators, intellectuals, politicians, both

355
David Shankland

local and national are aware of this and are happy to which of a number of factions comes to control the
use the material from the site to endorse that world management of cultural heritage in the country. Fur-
view. The local politicians at <;umra, who publicized ther, as well as any ideological influence, excava-
the site as part of their melon festival were of course tions, particularly large ones, must have a very
doing so in part to increase the attention given to significant effect on the economy of village commu-
their town by Ankara, but they were of the Republi- nities as they develop, particularly on those less well-
can People's Party, and it is no coincidence that they off than at Ki.i<;i.ikkoy. A sustained social interest
should have chosen a local prehistorical site nearby into these side-effects of excavating, even if ostensi-
as a key part of their campaign. Nor is it a coinci- bly not part of the archaeological project, would at
dence that the only group apparently to use the sym- least be a way of rendering the context of archaeo-
bol today are a youth radio playing secular pop logical enquiry more open, and perhaps lead to a
music. The villagers' response is thus more than fuller discussion of its place in Turkey today.
simply the passive reaction of a small community to
an academic world which is foreign, it is indicative Notes
of the scale of differences between two world-views,
world-views which are being played out in Turkey 1. Cf. the debate sparked by Gellner (1987).
as a whole as this chapter is written, and it is not at 2. I should make explicitly clear that my comments
all clear which will dominate. <;atalhoyi.ik has al- in this article refer to social anthropology as it
ready played a role in this struggle, and, situated appears to be developing in Britain.
between the village and the outside world, will con- 3. Cf. Mitchell (1995, xiii-xiv) for a similar observa-
tinue to do so in the future. tion referring to the aridity of some classical ex-
cavation reports.
Conclusions 4. Gellner was a leading figure supporting this view
(e.g. Gellner 1992), but it also is held more widely.
To conclude: the initial aim of the research project See, for example, the recent volumes published
was to conduct conventional anthropological field- in celebration of the fourth decennial conference
work within a village community whilst bearing in of the Association of Social Anthropologists of the
mind any possible relevance to the archaeological Commonwealth: among them, Strathern (1995).
enquiry going alongside it. In the event, the pres- 5. Officially 940 m 2, traditionally, the amount one
ence of the excavation forced me to take into account man can conveniently plough in a day with one
the interaction between the archaeologists and the pair of oxen.
village community, leading in turn into an interest 6. Cf. Sirman (1988), who describes the successful
in the socially intrusive aspects of the archaeological adaption to the market economy of household-
enterprise. Whilst the results of the ethnographic based farming in a village in the west of Turkey.
investigation will not be clear for some time, and the 7. The standard work on Turkish villages in
anthropologist's greatest use may be as a cultural Anatolia remains Stirling's Turkish Village (1965).
dragoman liaising between the local people and the Since then studies have usually concentrated on
archaeologists, I am convinced that almost any rea- exploring one particular theoretical point rather
sonable attempt to bring the two disciplines together than attempting a general picture; for example,
must be good for both. Delaney's (1986) investigation into gender, au-
However aware archaeologists are of the social thority and religion in a village near Ankara (cf.
context of the research they are conducting, they Hann 1993).
have not integrated this as yet into the way that they 8. Evans-Pritchard (1981) vividly evokes this struggle.
conventionally present their excavation results in 9. This could be stated even more strongly: Evans-
Turkey. It is impossible to say as yet whether an Pritchard's teacher, Seligman, was prepared to
attempt to take these things into account at <;atal- write social anthropology in archaeological vol-
hoyi.ik will in fact significantly change the academic umes as a matter of course (e.g. Seligman 1932).
results from the site, but the very large gulf between The anti-historical interdiction which came into
the complexity of the research and the simplified force after that, making it difficult for social an-
way results are usually presented is disconcerting. thropologists to feed their rapidly developing
Much archaeology practised within Turkey has po- conceptions and investigations of society into
litical and national heritage interest, and the future either classical or prehistory, or, on the other
of all excavations in the country will depend on hand, to draw on their findings, has contributed

356
Anthropology of an Archaeological Presence

toward a quite needless na:ivety on both sides. ogy in Turkish Village Society. Berkley (CA): Univer-
10. Dr Louise Martin, of the Institute of Archaeol- sity of California Press.
ogy, London, herself the bone and animal spe- Evans-Pritchard, E., 1981. A History of Anthropological
cialist at <;atalhoyiik, has kindly agreed to help Thought. London: Faber.
Gellner, E., 1987. The Concept of Kinship. Oxford: Basil
co-ordinate this part of the research.
Blackwell.
11. Lewis (1961) is still the best introductory account Gellner, E., 1992. Post-modernism, Reason and Religion. Lon-
of these reforms. don: Routledge.
12. Cf. Stirling (1974; 1993). Hann, C., 1993. Culture and anti-culture: the spectre of
13. Kiir;iikkoy literally would normally mean 'small' Orientalism in new anthropological writing on Tur-
or 'little' village, but kiir;iik also may mean key. Journal of the Anthropological Society of Oxford
'younger' in the sense of 'younger brother', and 24(3), 223-43.
it is in this sense that the villagers interpret the Lewis, B., 1961. The Emergence of Modern Turkey. Oxford:
name when telling the story. Oxford University Press.
Makal, M., 1954. A Village in Anatolia. London: Vallentine
14. Cf. the famous work of Mahmut Makal, a
Mitchell & Co ..
Kemalist school teacher who went to work in an Mitchell, S., 1995. Cremna in Pisidia. London: Duckworth
Anatolian village, and described his difficulties &Co ..
he encountered there (Makal 1954). Also Seligman, C., 1932. Egyptian influence in negro Africa, in
Shankland (1994). Studies Presented to F. LL Griffith, ed. S. Glanville.
London: Egyptian Exploration Society, 457-64.
Acknowledgements Shankland, D., 1994. Social change and culture: responses
to modernisation in an Alevi village in Anatolia, in
This research was funded by the British Institute of When History Accelerates, ed. C. Hann. London: Ath-
lone.
Archaeology at Ankara and the Charlotte Bonham-
Sirman, N., 1988. Peasants and Family Farms: the Position
Carter Trust, whom I would like to thank for their of Households in Cotton Production in a Village of
support. I would like also to thank the General-Di- Western Turkey. Unpublished Ph.D. thesis, Univer-
rectorate of Museums and Antiquities for awarding sity of London.
permission to research, and my ministerial repre- Stirling, P., 1965. Turkish Village. London: Weidenfeld &
sentative, Mr Liitfi Onal, for his great and cordial Nicholson.
assistance in the field. Stirling, P., 1974. Cause, knowledge and change: Turkish
I am grateful to my colleagues at the Department village revisited, in Choice and Change: Essays in Hon-
of Archaeology, Lampeter for their very valuable com- our of Lucy Mair, ed. J. Davis. London: Athlone, 191-
229.
ments when this paper was presented as a seminar in Stirling, P., 1993. Introduction: growth and changes: speed,
December 1995, and to Dr Hodder for his constant scale, complexity, in Culture and Economy: Changes
inspiration during the preparation of this project. in Turkish Villages, ed. P. Stirling. Huntingdon:
Eothen, 1-16.
References Strathern, M. (ed.), 1995. Shifting Contexts: Transformations
of Anthropological Knowledge. (ASA Decennial Con-
Delaney, C., 1986. The Seed and the Soil: Gender and Cosmo[- ference Series.) London: Routledge.

357
Chapter 19

Conclusions

Ian Hodder

C::atalhoyiik is situated on the largest alluvial fan in allow the appearance of <::atalhoyiik to be placed
the Konya plain. Prior to twentieth-century inten- within a longer trajectory of settlement, economic
sive water management regimes, soils around the and cultural change.
site were waterlogged for much of the year and in
the Neolithic the two mounds (east and west) were Site formation
probably situated either side of the <::ar~amba river.
We are still a long way from understanding The lower flanks of the mounds have been buried by
even the most basic characteristics of the subsistence 2-3 m of post-occupation alluvium. There are about
economy at <::atalhoyiik and surface work cannot 4.5 m of cultural material beneath the base of the
contribute substantially to this issue. We know that deepest trench left by Mellaart (that is below level
there are domesticated cereals (Helbaek 1964), but XII on the eastern mound). The Neolithic eastern mound
the relative paucity of grinding artefacts and the (at present 13.5 ha) comprises three eminences - a
initial trace element and teeth microwear studies main southern eminence clearly separated from a
(Chapter 13) indicate a diet based on pulses and smaller northern eminence. The eastern eminence is
tubers. Uncertainty remains about the domestica- low and is part of a large shelf which may continue
tion of animals at the site (Ducos 1988; Grigson 1989). beneath alluvium round much of the mound. Cer-
The regional survey has identified a wide range tainly a flank of 'off-site' activity was discovered by
of sites of different dates on the <::ar~amba fan. In coring to the west of <::atalhoyiik East (Chapter 2).
collaboration with the geomorphological survey it is The Chalcolithic western mound (8.5 ha) has not
clear that the gradual spread of the fan can be been fenced in recent years and so has suffered much
mapped. In addition, an important find has been worse erosion and disturbance. The cross-sections of
that sites completely buried by alluvium can be iden- both mounds conform to the profiles predicted as
tified by studying dyke sections. Research into GIS the result of long-term uniform erosion (Chapter 5
analysis of satellite images is being continued in and Rosen 1986), perhaps with minor variation re-
order to evaluate the potential for identifying buried lated to prevailing wind directions (Chapter 9).
sites. The surface collections and scraping have al-
Excavations at Pmarba~1 have allowed parallels lowed an insight into the post-depositional proc-
and contrasts to be drawn with <::atalhoyiik, at least esses at work on the site. The intensity of surface
with regard to the sixth millennium BC. The different collection on the two mounds, despite the small per-
architecture in a rock-shelter context is perhaps un- centages of the total areas from which collections
remarkable, but the lack of pottery and ground stone were made (0.88 per cent of the eastern mound, and
indicates the potential for identifying differences be- 1.00 per cent of the western), allowed clear patterning
tween sites in the use of this landscape. Of great to be identified which could be related to general
interest in regional terms is the continuation of a sub-surface trends and site formation processes.
microlithic industry well into the Holocene in cen- High densities of wheelmade pottery in the south-
tral Anatolia. Future work at this site will allow a ern part of the eastern mound were shown by scrap-
developmental history to be built which appears ini- ing (see below) to relate to a concentration of Classical
tially to differ markedly from the Levan tine sequence. occupation in this area. On the western mound, later
Understanding the early Holocene sequence will occupation is concentrated in the eastern part of

359
Ian Hodder

the mound. Date of latest Neolithic activity in different parts


2.44 per cent of the eastern mound was scraped of the site
to reveal surface architecture and features over a
total of 0.33 ha. On the eastern mound Classical and One question that does seem to have found at least a
Byzantine buildings and burials are concentrated to- provisional answer is the relative date of the occupa-
wards the south. In addition, three 10 x 10 m areas tion on the top of the northern eminence. The ceram-
were scraped on the western mound although here ics, obsidian and figurine data concur in giving a
the technique was less successful due to later distur- position between Mellaart's levels V and VI. But this
bance. Overall, a clear pattern could be observed conclusion also raises new questions. In particular,
regarding fragmentation at different depths within was the northern eminence abandoned while occu-
the soil in the scraped areas. Both bone and ceramics pation continued on the higher and larger southern
have larger fragment size and thus better preserva- mound in levels IV and above? The answer to this
tion with greater depth, where they are protected question hinges on how much erosion has occurred
from weathering, trampling etc.. Similar factors may on the northern eminence. Were there levels above
have affected the figurines (Chapter 12). the area mapped in Figure 7.2 which have since been
Study of the movement of Classical and wheel- lost? Since this eminence was not so well protected
made pottery has been of special value in under- by Classical occupation, perhaps the observed build-
standing erosional processes. High correlations ing level was not the last on the northern eminence?
between wheelmade pottery and other artefact classes Such a possibility is supported by the fact that the
in the northern scraped area suggest that much of lithic and ceramic material found on the surface on
the contemporary detailed patterning is historic (post- the northern eminence includes material later than
Classical) in origin. In more general terms, Classical that found in rooms 70 and 71 during their initial
occupation has led to the masking of some distribu- excavation in 1995. Later levels may have occurred
tions (bone densities, lithic artefact densities) and to on this part of the site and been eroded off. How-
a diminution in size of bone fragments. Downslope ever, the amount of this later material is slight and
movement of sherds has been mainly to the south, the balance of evidence (Chapter 7) indicates the
perhaps because of prevailing winds (Chapter 9), or uppermost level planned on the northern eminence
because of the shape of the mound in this part of the was probably the last.
site (Fig. 5.7). Smaller sherds and pieces of obsidian The latest ceramic assemblage of the Neolithic
and fragments of bone tend to occur towards the on the east mound derives from the eastern plateau.
lower and flatter slopes of the mounds. This may be
partly the result of the greater distance travelled by The spatial and social organization of the site
smaller artefacts and partly the product of longer
periods of weathering and fragmentation. The magnetometer survey (Chapter 8) has produced
The obsidian data suggest a picture rather dif- a remarkable amount of detail over a large area (Fig.
ferent from the ceramics and the bone and this is to 8.3). The main problem in interpreting these plots is
be expected given the different size, fragmentation the difficulty of ascertaining the dates of the features
and weathering characteristics of this material. Ham- identified. At least some of the walls in Figure 8.5
ilton (Chapter 12) suggests that larger items such as are undoubtedly Classical in date, and the extensive
clay ball and figurine fragments may not have moved ashy tip shown in Figure 8.4 is probably late also (as
as far as the other, smaller materials. The different identified in the scraping). Many of the features in
fragmentation patterns for different types of figu- the southern part of Figure 8.3 are probably post-
rines suggest that post-depositional factors must Neolithic, while in other areas the degree of correla-
inter-link with varied and specific depositional prac- tion between known Neolithic buildings and features
tices. on the magnetometer plot allows some interpreta-
R. Matthews (Chapter 7) estimates that prob- tion of the plots as a whole.
ably no more than 0.5-1.0 m of archaeological de- On the northern eminence of the east mound, a
posits have been eroded from the surface of the total area of 1900 sq.m has now been scraped to
northern eminence of the eastern mound. Overall it reveal a continuous pattern of over 30 Neolithic build-
is clear that while detailed patterning of artefacts ings. The street identified here could not be followed
has been lost through displacement and weathering, to the summit of the northern eminence because of
general patterning can be discerned as long as site classical disturbance. The buildings in this area are
formation processes are taken into account. largely contemporary, with the possible exception of

360
Conclusions

spaces 19,42 and 44-50 (Fig. 7.3), which may belong 'houses' may have included buildings with a wide
to an earlier level. range of functions, the population, even at its maxi-
The radial pattern and the straight lines in the mum extent, may have been well below 5000.
plan (Fig. 7.3), which are shown to extend further in In considering the artefacts from the buildings
the magnetometer survey (Fig. 8.7a), suggest some excavated in the 1960s (Chapters 9 to 13), consider-
organization- but these characteristics of the build- able caution must be exercised. The records avail-
ing layout could have been the product of a need to able are lacking in detail; they are often incomplete
locate buildings on terraces encircling the summit, and inconsistent. But even if full records were avail-
and of the need to provide for water runoff (rain, able, it is not clear what the occurrence of artefacts in
snow meltwater, or even drainage) from the top of buildings represents. The study of floors in sections
the mound. Certainly the removal of refuse from has demonstrated that they are normally swept scru-
buildings must have required a considerable degree pulously clean. Burnt buildings have often been
of organization. And there were parts of the site (the cleared out prior to a controlled process of burning.
so-called courts) given over to refuse etc. for long Even in cases where accidental fires have trapped
periods (Chapters 14 & 15). artefacts on floors, it is unclear what the artefacts in
So far we have encountered no evidence for the room 'fill' represent. Artefacts in fill deposits may
existence of large-scale public buildings. The six par- derive from the roof of a building, but they may also
allel walls and the lack of wall plaster in scrape area derive from general surrounding deposits. The
980,1080 perhaps indicate something different and micromorphology and study of the sections (Chap-
distinctive, but this area on the north part of the ters 14 & 15) have demonstrated a variety of proc-
main mound needs further investigation. The site as esses leading to room infill.
a whole appears to consist of two main elements: It might thus be surprising if any correlations
houses, with varying degrees of elaboration of inte- between buildings and their contents could be
rior features, and open spaces, generally covered by gleaned from the museum material. Nevertheless,
spreads of rubbish. This simple settlement constitu- some initial indications of patterning can be observed
tion, and the evidence for crop processing, animal which will need to be explored during the excava-
tending (thick layers of dung within the settlement) tion phase.
and obsidian manufacture interspersed amongst The symbolic importance of the house in con-
houses, all support an interpretation of the site as an trast to areas outside the house is indicated by the
elaborate village rather than as an urban settlement prevalence of human or schematic figurines inside,
with differentiated functions. and animal or humanoid outside (Chapter 12).
Certainly there must have been specialized func- Initial indications are that the fine bifacially
tions amongst the population at the site. The obsid- flaked obsidian points are concentrated in the more
ian mirrors and dagger blades would have required elaborate 'shrine' buildings (Chapter 10). The fact
a high degree of skill for example (Chapter 12). The that cores too concentrate in the more elaborate build-
stone vessels and the beads obtained in exchange ings suggests some preferential access to or involve-
may also indicate some degree of specialization. Con- ment in obsidian. It is, of course, possible that this
centrations of similar figurines in certain buildings type of patterning is the result of biasses in the col-
perhaps indicate specialized activities although the lection of data in the 1960s - for example, perhaps
variation may relate to abandonment processes (see greater effort was put into the recovery and record-
below). ing of artefacts in buildings thought to be 'shrines'.
The size of the site must not be exaggerated. Only further excavation can resolve such problems.
We have seen that at any one time not all of the site But for the moment, the data do suggest some corre-
may have been occupied and certainly the focus of lations between elaborate buildings, cores and
later occupation seems to have shifted away from bifacially flaked obsidian artefacts. Given that many
the northern eminence to the south and east. But non-complex buildings also have these artefacts, we
even in levels V and VI we must presume that large cannot argue for total control of production by an
areas of the site were given over to open areas for elite. Rather, the picture is one of preferential access
refuse deposition etc .. In Chapter 7, R. Matthews and involvement.
estimates total population figures for the eastern The ceramic material does not indicate clear
mound ranging between 5000 and 10,000 persons. repeated differences between the architecturally more
Given that even within the built-up area houses may and less complex buildings. Although assemblage
have gone in and out of use, and given that the composition does vary with architectural complexity,

361
Ian Hodder

the nature of the difference is itself variable. How- include 70-72 with a plaster feature in the centre of
ever, there is some indication that the more elabo- the north wall and 86-88-89 with a large pair of
rate buildings are also the most 'progressive' in that cattle horns in space 89. This suggests that 'shrines'
they contain innovative or unusual ceramic forms are not confined to the area excavated by Mellaart.
(Chapter 9). The overall accretionary or organic process is clear
The largest numbers of figurines come from in a number of ways. No two buildings in the north-
very elaborate buildings, but on the other hand many ern scraped area in Figure 7.3 are identical. The over-
complex buildings have no figurines and many figu- all impression is of buildings packed against each
rines derive from non-complex buildings (Chapter other so that all available space was used in an accre-
12). tionary and opportunistic manner. Some of the
Wason (1994) has argued for a low-level of so- stratigraphical evidence too (Chapter 14) suggests
cial inequality at ~atalhoyiik based on the variation that buildings might be added next to earlier ones as
in buildings and burials. He suggests that the pres- part of the gradual growth of the mound. All this
ence of well-furnished child burials indicates some appears to suggest a decentralized pattern of growth
degree of hereditary status. However, he notes that and organization, within which differences can be
there is no simple relation between 'rich' burials and identified between more or less elaborate buildings,
elaborate buildings. As Mellaart (1967, 82) points and which is also structured by general requirements
out 'there is no relation between the size of a build- such as drainage and access.
ing or the abundance of its decoration and the pov- The microstratigraphic evidence (Chapter 15)
erty or richness of the burial gifts of the dead below has distinguished two types of occupation within
its platforms' (see also Hamilton, Chapter 12). buildings, although these types can both occur within
The faunal evidence surviving from the 1960s' one building and do not correlate in any simple way
excavations is not adequate to allow variation be- with 'shrine' versus 'non-shrine'. One of the most
tween building types to be studied. But a lack of distinguishing characteristics is the presence or ab-
clear distinction between 'shrine' and 'non-shrine' is sence of finishing coats of white plaster on the floors.
also suggested by the initial micromorphological Where such finishing coats occur, overlying occupa-
work, which indicates that in 'Shrine' VIII.25 there tion deposits are very thin or absent. Such floors
were multiple activities relating to probable resi- occur in rooms with elaborate sculpture or painting.
dence and burial of the dead (W. Matthews, Chapter Where there are no finishing coats of white plaster
15). The fire-installations in the building appear to on the floors, the overlying occupation deposits are
have been in constant use, and there is a continuous generally thicker and include a wider range of 'refuse'
accumulation of thin well-maintained plaster floors such as burnt and unburnt organic and inorganic
with thin layers of occupation deposits which in- aggregates, small fragments of bone, charred cereal
clude some Gramineae and, next to the fire-installa- grain fragments, silicified plant remains, melted silica,
tions, bone, organic aggregates and charred wood. charred wood, and calcitic ashes.
Burials occurred at several stages in the lifetime of It is too early to be able to classify such differ-
the building. ences as ritual and domestic. We need much more
Even on the cleaner (more formal or 'ritual') detailed contextual understanding of the use of space
floors in the section 3 building (W. Matthews, Chap- at ~atalhoyiik before variation in the amount of oc-
ter 15) dung pellet fragments of herbivores were cupation debris left on floors can be understood.
found. This evidence indicates that at least young But, for the moment, it is clear that although differ-
animals were brought into even more elaborate build- ences between buildings which are more or less com-
ings, and it serves further to blur the distinction plex architecturally can be identified, there is no
between 'shrines' and 'non-shrines'. Reconsideration overall and absolute separation of functions which
of the 1960s' data has indicated no clear separation might be termed ritual ('shrine') versus domestic
between 'shrines' and 'non-shrines'. The complexity ('non-shrine'). A general lack of clear rules or struc-
graphs in Figures 1.3 to 1.7 indicate a continuum of tures is suggested by Hamilton (Chapter 12) in rela-
variation between more and less complex buildings tion to both the spatial patterning of burials in houses
in terms of architecture and internal fittings. and the relationships between the sex and artefact
The scraping (Fig. 7.3) indicates an overall dif- associations of burials. The overall fluid structure of
ference between more and less elaborate buildings. the occupation is also seen in changes of use of space
More elaborate buildings on the northern eminence over time.

362
Conclusions

Change through time not seen by Mellaart because he did not systemati-
cally remove plaster from the walls. The presence of
The mound at <;atalhoyuk is approximately 23 m a 'door' between Houses 2 and 12 suggests that roof
high (including the unexcavated lowest levels and entry may not have been the only form of access, at
those below the level of the present plain). It was least in some houses. The blocking of entrances may
thought by Mellaart to be divided into 12 levels, have been a constructional feature, or it may have
with further unexcavated levels below. been associated with the change in use of buildings
In terms of overall change, the work on the from domestic to (ancestral) ritual.
museum collections has provided useful results. Clear The notion of cycles of house use is re-inforced
ceramic sequences have been identified for the first by an observation of Mellaart (1966, 182-3) concern-
time. In the lowest levels, thick, low-fired and usu- ing two burials in the 'Red Shrine' (VIII.31). 'What is
ally vegetable tempered pottery is found. From level clear is that these two burials took place before the
VIII this begins to be replaced by a thin, dark-faced erection of the shrine and the unusual features, the
pottery and increasing numbers of closed forms painted platform and the orange panel with libation
which reach a peak in level V. Above this, open hole, mark the position of the graves.' Mellaart (1966,
forms increase and there is a wider variety of types. 182-3) suggests that the building was constructed
The obsidian indicates a trend from a flake-based to over the burials, in memory of the interred.
a blade industry; there is also a potential decrease The placing of animal and other figurines in the
through time in the prevalence of complete bifacial walls of at least some buildings (Chapter 12) might
retouch. Another possible trend through time is that at least indicate a symbolic importance attached to
clay balls appear to be concentrated in the lower the construction of house walls. The overall impres-
levels (Chapter 12), whereas figurine representations sion is of an extremely complex and varied set of
of women perhaps increase through time. processes by which buildings were built, used and
All this might be seen as confirming the overall (as we shall see below) abandoned. The stratigraphy
division of the east mound sequence into a series of is built up as a product of these intentional and
distinct levels. There is some evidence (Chapter 14) varied strategies.
that adjacent buildings used similar bricks as they
moved from phase to phase in concert. However, as Stratigraphy as a social process
Matthews and Farid (Chapter 14) point out the
stratigraphic evidence is unclear except over small Last (Chapter 9) suggests an emulative process by
areas. It remains possible that the overall stratigraphy which certain more elaborate buildings take the lead
of the site is built up from small-scale, relatively in a ceramic 'style wars'. Given the apparently low
independent sequences of individual buildings. In degree of social complexity at the site (Chapter 12),
some cases there is clear evidence for organic (var- symbolic competition could have been an important
ied) and independent growth and transformation of process for creating social distinction. But was the
buildings. The rarity of party walls at the site is maintenance of a more elaborate building over time
perhaps explained by the independence of each build- also a matter of mobilizing resources? The lithics
ing - each architectural cell being repaired and re- suggest this may have been the case in that more
built separately. elaborate buildings demonstrate access to more com-
In at least one building identified in section 3 plex technologies (Chapter 10).
there is clear indication of a change of use (W. If the more elaborate buildings entail greater
Matthews, Chapter 15). On the earlier floors there is mobilization of labour, both in construction (archi-
a wide range of domestic activities. After cutting a tectural complexity) and use, to what extent was it
burial through these earlier floors, the building is possible to pass on privileged access to labour from
used for a narrower range of activities and there is generation to generation? Were more elaborate build-
evidence for 'ritual' features. The idea will be ex- ings in one phase replaced by more elaborate build-
plored in future seasons that at least some buildings ings in the next? How structured and institutionalized
go through a cycle of use from domestic to ancestral was power?
shrine. The differences beween 'shrine' and 'non- To begin to answer these questions it is possi-
shrine' on the site may be as much temporal as spatial. ble to return to the Mellaart data. There are many
In the Mellaart area, 'doors' and crawlholes had difficulties in evaluating the degree of continuity
been blocked (Chapter 14). This blocking was often between houses. Perhaps the most important is that

363
Ian Hodder

different areas of the site were excavated at different continuities in buildings and the numbers of conti-
levels, and the recording of each level is uneven. nuities in elaborate and non-elaborate buildings. It
Nevertheless, the degree of continuity between lev- is apparent from this table that, of all the buildings
els is remarkable. Walls are built on previous walls in level n that can be seen as continuing in plan from
wherever possible (Chapter 14). There is some sub- level n-1, the vast majority retain the same classifica-
division and redefinition of space but very little tion as either elaborate or non-elaborate (in the terms
change in house lay-out. The very high degree of just defined).
continuity might be the product of building prac- This stability in the use of space is also seen in
tices in that, without the use of stone foundations, Table 19.2. Here the percentages of elaborate build-
mud-brick walls were best built on previous mud- ings that continue as elaborate buildings from the
brick walls. previous phase (n-1), or that continue into the next
But walls were often not rebuilt completely. It phase (n+ 1) are shown. Given that, as already noted
has become clear in the Mellaart area that the re- different parts of the site were excavated in each
building of walls often occurred in a gradual way, level, the degree of continuity in the location of elabo-
with only the upper parts of the walls repaired (Chap- rate buildings is remarkable.
ter 14). Thus the floor of, say, a level VIII room might In small parts of the area excavated by Mellaart,
be several metres below its corresponding phase of buildings can be followed through multiple levels.
upper walling. A wall might thus have overhangs as Here again, the degree of stability in the continued
a result of inaccurate alignment on earlier walls, with reconstruction of elaborate buildings is striking. For
the join plastered over (Chapter 14). example, E8 continues in use as an elaborate build-
Nevertheless, there are changes in the use of ing through 5 levels (VIII to V where it is renamed
space through time, even if walls themselves are E6). E14 is another elaborate building which contin-
continually rebuilt. Thus, buildings become used as ues through these 5 levels. Several other elaborate
open areas for refuse and vice versa. Buildings be- buildings continue through at least 3 or 4 levels (E1,
come more or less complex. In order to evaluate the ElO, E12, E4).
degree of continuity between levels in the location of In overall terms there is a high degree of
the more elaborate ('shrine'-like) buildings, a more consistancy or 'routinization' that cuts across 'lev-
elaborate or complex building was defined as one in els'. There are also specific continuities through time
the upper third of the complexity variation in Figures in the depositional practices found in particular build-
1.3 to 1.7. The following table shows the numbers of ings and in the art. For example, from level XII to
VIII building 29 was the focus for the deposition of
Table 19.1. The degree of continuity of elaborate and non-elaborate ceramics within the area excavated by Mellaart (Last,
buildings between different levels. Chapter 9). As regards the art, Dorothy Cameron
Levels
(pers. comm.) has noted a number of specific conti-
VIII-VII VII-VIB VIB-VIA VIA-V V-IV nuities in the motifs used from level to level in one
Number of continuities building. Certainly the elaborate building E10 has
of building 9 27 33 8 3
very similar bull and 'goddess' sculpture in VIB and
Number of continuities of
elaborate buildings 2 7 15 3 0 VIA. There are also continuities from VII to V in E1
Number of continuities in and E8, especially when contrasted with very differ-
non-elaborate buildings 5 12 17 0 2 ent painting in a different area at FV.l.
Given all this evidence for the handing on of
Table 19.2. The percentage of elaborate buildings in level n that resources, we might expect a great emphasis on an-
continue as elaborate buildings (a) from level n-1, and (b) to level cestor cults- that is cults which legitimate an indi-
n+l. vidual's access to certain stable and institutionalized
(a) VIII VII VIB VIA v IV rights. Indeed, many of the activities and art within
Number of elaborate
buildings 13 19 15 6 4
the buildings can be interpreted in terms of links
Number of elaborate with past inhabitants of buildings. For example, the
buildings which continue 2 7 14 3 2
Per cent 15 37 93 50 50 location of burials below platforms within buildings
(b)
links present inhabitants with the ancestors. The over-
Number of elaborate all focus on death and death rituals in the art (e.g.
buildings 4 13 19 15 6
Number of elaborate the vulture paintings) perhaps points in a similar
buildings which continue 2 7 15 5 2
Per cent 50 53 79 33 33
direction. The evidence of figurines with removable
heads perhaps suggests some link with death too,

364
Conclusions

since headless figures are shown in the vulture paint- possible that through time, more and more effort
ings. The association between women and danger- was invested in identifying separate households and
ous wild animals has often been interpreted in terms their ancestral rights.
of some form of death and renewal symbolism. A A continuing process of competition is seen
concern with continuity is perhaps also seen in aban- through the sequence in ceramic innovation. But it is
donment rituals perhaps also seen in the symbolic elaboration of
buildings. Returning to the complexity graphs in
Abandonment rituals? Figures 1.3-1.7, it is evident that there is no overall
increase or decrease in the degree of difference be-
Matthews and Farid identify four ways in which tween more and less elaborate buildings. A highly
buildings were infilled. The first involves burning differentiated set of buildings would have a concave
and infill with burnt structural debris. They note graph; less marked differentiation would result in a
that this is often localized and structured, suggest- smoother, more even graph. Looking though the se-
ing possibly ritual 'cleansing' or 'closing-off'. R. ries of graphs we see, if anything, a continual move-
Matthews (Chapter 7) notes local areas of burning in ment back and forth between more and less
the northern scraped area, and suggests a strictly differentiated patterns. The lowest levels, especially
controlled use of fire within discrete buildings. The the first large sample in level VII, show smooth gra-
second type of infill involves deposits with mud dations between more and less elaborate buildings.
brick and plaster relief fragments. Mellaart (1964) But in VIB and VIA there are concave graphs and a
had already suggested a deliberate process of aban- distinct, small group of very elaborate buildings can
donment in such cases. The third and fourth types of be differentiated. In level V the graph smooths out
infill sometimes involve short deliberate acts of depo- again, to be replaced in level IV by differentiation.
sition of mineral sediments or refuse. Level III shows a more even distribution of 'elabo-
The recovery of groups of figurines on the floors rateness', which becomes more differentiated in level
in a number of more complex and burnt rooms may II. It may well be that different parts of the site will
be interpreted as part of an abandonment ritual produce different sequences and it is likely that dif-
(Hamilton, Chapter 12). The remarkably clean floors ferent measures of complexity will produce differ-
(as confirmed in the micromorphological analysis) ent results. But the overall impression is of a
also could be taken as indicating the deliberate na- continually competitive and unstable process.
ture of the abandonment process. If, within this continual process, there is any
increased concern with access to rights, we might
Conclusion expect to be able to see such changes within the
realms of production and the economy. The shift to
We might end on a note of conjecture- a luxury an obsidian blade technology in the upper levels
perhaps justified by our continued involvement in indicates social as much as technological changes.
an excavation project which over the long term might Contextual analysis will be needed in order to see
allow us to answer some of the questions that have whether this technological shift is related to increased
been raised. The conjecture concerns change through specialization of production. In Chapter 10 Conolly
time. Mellaart (1966, 168) argued that there were suggests that, in terms of the distributions of obsid-
party walls in levels XII and XI. These rarely occur ian cores and bifaces, the differences between more
higher in the sequence. While there are good struc- and less elaborate buildings become more marked in
tural reasons to support one house and roof against the later levels. Economic changes through the se-
another, the problem could have been solved, as it is quence are very poorly understood and, as a result,
today in the region, by simply building thicker walls. have not been discussed in this volume. One inter-
That separate thin double walls were retained im- pretation of the faunal evidence is that cattle were
plies some concern with identifying one's own domesticated in the upper levels (Perkins 1969, 178),
boundaries. This emphasis on separate buildings is although Ducos (1988) and Grigson (1989) have now
supported by all the evidence for variability in lay- questioned the domestic status of the cattle. The in-
out and contents of buildings. Mellaart (1966, 172) creasing investment in ceramics could be taken to
also suggests that later levels (above VIB) were de- suggest an increasing specialization of production.
stroyed in conflagrations. It has been suggested here Hamilton (Chapter 12) notes possible changes in gen-
that much of this burning was controlled and ritual- der roles or ideologies in the upper levels which
ized as part of abandonment procedures. It is again may be related to increasing concerns about

365
Ian Hodder

continuity, competition and ancestry in relation to activity, and yet a way of life with the simplest of
the house and domestic production and reproduction. structures. We struggle with the contradictions be-
Overall, the tendency of our work so far has tween stability and individual diversity, between
been to demystify. By this I mean that the evidence continuity and change, between egalitarian and
points to a relatively low level of socio-political com- ranked. Demystification in one direction has only
plexity. There is much variation from building to opened up challenges in others.
building, but there are so far no clear public spaces
or ritual centres or specialized elites. On the other References
hand, there is remarkable stability and continuity.
Long-term change occurs but within a continuity of Ducos, P., 1988. Archaeozoologie quantitative- les valeurs
structure which is remarkable. Institutionalized dif- numeriques a <;atal Huyak. (Les Cahiers du Quater-
ferentiation could be passed across the generations naire.) Bordeaux: CNRS.
with great efficiency. In understanding how this con- Grigson, C., 1989. Size and sex: evidence for the domesti-
cation of cattle in the Near East, in The Beginnings of
tinuity of power was exercised, maintained and per-
Agriculture, eds. A. Milles, D. Williams & N. Gardner.
haps increased through time we need to look into a Oxford: BAR, 77-109.
complex symbolic world in which the buildings were Helbaek, H., 1964. First impressions of the <:;atal Hiiyiik
the centre of rites of continuity and ancestors. We plant husbandry. Anatolian Studies 14, 121-3.
may have demystified the site such that it looks Mellaart, J., 1966. Excavations at <:;a tal Hoyiik, fourth pre-
more like a large village than a town or urban centre. liminary report, 1965. Anatolian Studies 16, 165-91.
But on the other hand, we are on the edge of a new Mellaart, J., 1967. c;atal Hayak: a Neolithic Town in Anatolia.
type of understanding of a mythical world deeply London: Thames & Hudson.
embedded in a complex social system which bal- Perkins, D., 1969. Fauna of <:;atal Hiiyiik: evidence for
early cattle domestication in Anatolia. Science 164,
ances somewhere between egalitarian and differen- 177-9.
tiated, and which also transforms itself through time. Rosen, A.M., 1986. Cities of Clay: the Geoarchaeology of Tells.
We struggle to reach understanding of something Chicago (IL): University of Chicago Press.
which seems so unfamiliar- a large and complex Wason, P.K., 1994. The Archaeology of Rank. Cambridge:
centre containing an enormous density of cultural Cambridge University Press.

366
On the surface: <;atalhoyiik 1993-95

After the excitement of its discovery and excavations in the early 1960s, the
world important site of <::atalhoyuk has remained dormant for 30 years. This
volume describes the first phase of renewed archaeological research at the site.
It reports on the work that has taken pla\e on the surfaces of the east and west
mounds and in the surrounding regions. It also discusses the material from
the 1960s' excavation in museums, which has been re-examined. The result is
that new perspectives can be offered on the internal organization and
symbolism of a site which is central to our understanding of the earliest
development of complex societies.

Editor:
Ian Hodder, Department of Archaeology, University of Cambridge,
Downing Street, Cambridge, CB2 3DZ.

Cover picture:
The number of wheel made sherds collected from each 2 x 2 m square on the east
mound, c;atalhoyiik.

Published jointly by the McDonald Institute for Archaeological Research, University of


Cambridge, Downing Street, Cambridge, CB2 3ER, and the British Institute of Archaeology
at Ankara, c/o British Academy, 20-21 Cornwall Terrace, London, NWl 4QP.

Printed by Short Run Press.

Distributed by Oxbow Books.

Cover design by Margaret Downmg.

You might also like