Bulletin on Sumerian Agriculture - 6 (1992): Trees and Timber in Mesopotamia

BULLETIN ON SUMERIAN AGRICULTURE Volunle VI

Cambridge 1992

General Editors J.N. Postgate Faculty of Oriental Studies University of Cambridge Sidgwick Avenue Cambridge CB3 9DA U.K.

CONTENTS M.A. Powell Department of History Northern Illinois University De Kalb Illinois 601 25 U.S.A.

Preface

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

Timber and trees: ancient exploitation in the Middle East: evidence from plant remains G. Willcox Levant trees and tree products N. Liphxhitz

. . . . . . . . . . .

v

1-31

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33116

Ethnographic evidence for wood, boats, bitumen and reeds in Southern Iraq E. Ochsenschlager

. . . . . . . . . . . . . . . . 47-78

LC bois dans l'architecture: premier essd pour une estimation des besoins dans le bassin misopotarnien . . . 79-96 J.-Cl. Margueron Dcndrochronological wood from Anatolia and environs P.I. Kuniholrn

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

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99-122

'Timber production in Presargonic Lag& M.A. Powell

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

h e r k u n g zu den neusumerischen Texten iiber Schilfirohr W. Sallaberger

'Rohr' und dessen Verwendungsweisen anhand der neusumerischen Texte aus Urnma FI. Waetzoldt Kced in the Old Babylonian texts from Ur M, Van De Mieroop

*

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

155-161

ISSN 0267-0658 Orders may be placed though bookseuers or direct to the Sumerian Agriculture Group, Faculty of Oriental Studies, Sidgwick Avenue, Cambridge CB3 9DA, U.K. OSumerian Agriculture Group 1992. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means without the prior permission of the publishers. Printed in EngIand by Aris & Phillips Ltd., Warminster, Wiltshire

. . . . . . . . . . 163-170

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

171-176

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

177-192

Wood, reed and rushes: a note on Neo-Babylonian practical texts G. van Driel

Notes on the reeds of Mesopotamia F.N. Hepper

. . . . . . . . . . . . . 125-146 147-153

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

Trees and timber in the Assyrian texts J.N. Postgate

123-124

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

Wood in the Old Babylonian texts from Southern Babylonia M. Van De Mieroop

Lc bois & Mari J.-R. Kupper

97-98

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

193-194

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

195-196

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

197-199

A corrective note on Pistaciatrees and resin

F.N. Hepper Some wood identifications from Mesopotamian sites P.R.S. Moorey & J.N. Postgate

Illustrations of trees by F.N. ~ k ~ ~ e r Pinus halepensis @. iv) - Populus, Buxus, Platanus (p. vi) - Quercus spp. (p. 32) - Juniperus, Tamarix @. 154) - Abies, Cupressus, Cedrus @. 162) List of contributors and addresses

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

iii

200

PREFACE

'Jc

lulnl.

h of thc Bulletin on Stmerinn Agriculture contains papers from the meeting of thc

,4gr-:culltart Group held an the sympathetic environment of the Inter~~ationales W~xscnccllallslir:um of !he Univenity of Heidelberg in July 1989. Our thxlks go to Prof. Dr. i l ; ~ ~ r n r Wactzoldt t~r for organizing the meeting, and to the Deutsche Forschungsgemeinschaft and 11ri- Ministry of Ck~ltureV J I ' I ~ C generously ~ defrayed the costs of the meeting.

SIIII~ r; tan

Some of those at Heidelberg were not presenting formal papax, but taking part in the cIi\cussrori, nnd some of bhe papers given were not destined for inclusion in this volume. 1';trlicipmts in h e meeting not represented here in print are: Dr. A. Cavigneaux, Prof. Dr. W. t:rcy, Dr. J.-P. GrCgoire, Dr. B. HruSka, Prof. Dr. J. Renger, Prof. M. Stol, Dr. G. Szab6, and last but not least Prot', Dr. W. van Zeist. Prof. J.-CI. Margueron, though unfortunately prevented from

ntlcnding in person, was able to send his paper.

In July 1990 the Group met in Barcelona to discuss sheep and goats, at the invitation of Prof. G. del Olmo Lete of the Institute Oriental. The papers from that meeting will form the next volurne(s) of the Bulletin, now in preparation. The time and place of the subsequent meeting, which will probably deal with cattle, have yet to be decided. The Editors would like to take this opportunity to stress that they are always ready to consider for publication contributions on agriculture in Mesopotamia which do not derive directly from a meeting of the Group, especially those concerned with the theme of forthcoming volumes.

As in previous volumes, bold type is used for Sumerian, italics for Akkadian words. Bibliographical conventions generally follow the usages of the individual authors, but Harvard system references are used for books and articles, whereas traditional Assyriological abbreviations are preferred for text editions. These can be tracked down in one of the following works: R. Borger, Handbuch der Keilschrifliteratur, W. von Soden, Akkadisches Hanmuiirterbuch, the Chicago Assyrian Dictionary. For Ur 111 texts see the works cited by Waetzoldt on p. 146.

Once again, the format and typography of this volume have benefited from the support of the Literary and Linguistic Computing Centre in Cambridge: our thanks are due equally to Beatrix Bown, John Dawson and Rosemary Rodd for their help and readiness to solve our problems. We are also grateful to Kirstie Shield (Girton College) for her assistance with the type-setting and to Trinity College, Cambridge, for financial support towards this. Pinus halepensis Nicholas Postgate Marvin Powell

January 1992

TIMBER AND TREES: ANCIENT EXPLOITATION IN THE MIDDLE EAST: EVIDENCE FROM PLANT REMAINS G. Willcox (C.rV.R.S., E.RA. 17, JaEs, France)

Populus euphratica

Plataiius orientalis

Introduction The exploitation of trees and shrubs (ligneous species) has been fundamental to man's development and dates back into the Palaeolithic to the invection of fire. This natural resource, which makes up an important part of the vegetation cover in many regions of the world, has, and is being, over-exploited by man. The study of the exploitation of ligneous species through analysis of charcoal remains from archaeological sites has relevance beyond historical reconstruction, because it can throw light on the progressive destruction of forest cover and the resulting effects on the natural environment. The history of vegetation in the Middle East, however, would be very poorly understood were it not for the survival of undisturbed stratified pollen in lake beds. Pollen analysis carried out principally by van Zeist and his team at Groningen has provided evidence revealing the sequence of vegetational changes which have occurred since the last glaciation, when severe steppe conditions existed (van Zeist & Bottema 1982, 1977). Most of the vegetation cover that we see in the Middle East today is degraded, but we may find relict associations in inaccessible and isolated regions where exploitation would be impractical (Davis 1971, Zohary 1973, Guest 1966). They are what remain of the original climax vegetation. The examination of these relict associations has greatly enhanced our knowledge of past vegetation and of past potential timber resources. The cedars of Lebanon, now almost completely destroyed, are an example. Charcoal is omnipresent on archaeological sites and is readily recovered by flotation (Williams 1973, Pearsall 1989). This charcoal (almost pure carbon, which is inert and does not decompose) preserves perfectly its anatomical structure and so may be readily identified when examined under the microscope with the help of reference material and the standard manuals (see Greguss 1955 & 1959, Schweingruber 1978). It is not uncommon on archaeological sites to find charcoal belonging to species which have long since been wiped out. This is particularly true of areas where intense settlement has occurred continuously over a long time span. With four lines of direct evidence, pollen from lakes, relict vegetation assemblages, seeds and charcoal from archaeological sites, one can begin to reconstruct the ancient vegetation and thus the environment of ancient civilisations. Charcoal from archaeological sites is of particular interest to the archaeologist because it gives direct evidence of the use of tree and shrub species in the immediate area of the site, although there is also some evidence for long distance transportation. On the other hand, for the botanist who wishes to reconstruct the vegetation, there is always the risk that ancient man was selective in his exploitation, and this must be taken into account. In this paper I have used scientific names of plants to avoid confusion, but readers who wish to refer to common names may consult the list (see below) where English common names are given. A word of warning: names in English for timber or lumber are not always the same as for

Timber and Trees

the tree and often include several different species. A good example is deal, which wuld br any close-grained Gymnosperm. Ebony is a very dark wood that may come from several different species. Teak and mahogany may also refer to more than one timber species. Rosewoods are named after their fragrance, and are not members of the plant family Rosaceae.

Phytogeographical Regions These regions are made up of distinct associations of plants and have been described by Zohary (1973) and Guest (1966). The following is a summary with reference to some of the major tree species which could have beer. exploited in antiquity for timber. Mediterranean This region concerns the coastal k l t of Turkey and the Levant where a number of imponant timber trees occur, the most important being Cedru.~libani, Quercur s r ~ pFr~.xinus ~~ spp., Pinus spp., Platanus orientalis, Abies cilicica, Cupressus sempervirem. Irano-Turaniun This vast area stretching from Turkey to China (including nonhem Iraq) contains an extremely varied Rora and a large number of woody species. However, in general this region is characterised by low rainfall and extreme temperature variation, which docs not lead to the production of good timber in the modern sense of the term, though one could consider Juniperus spp. Quercus spp. Jugkrss regia, Ulmus spp., Pistaciaz atlantica, Pinus brutio, and Tuxus haccata, as some of the possible sources of timber. These species are f~imndin the Zagros. and in the Taurus together with some Mediterranean species. Two other areas within this phytogeographical region should be considered as potential timber producers because of their special climatic conditions. These are the Euxine region where Fagus orientalis, Abies cilicica, Picea orientalis, Coryllls spp., Castanea sativa and Carpinus orientalis are to be found, and the Caspian region, where such species as Farrotiu persicta, Mnrus nigra, Z e l b v a sp. occur. These areas have been exploited in more recent times, but at presmt there is no evidence of their being exploited in the second or third millennium B.C.

Saltaro-Sindion This forms a belt of sub-tropical desert steppe vegetation m i n g from west Africa, Egypt, Ardbia (including southern Iraq) to northwest India. This region is chamcicrised by high temperatures and extreme aridity. Evaporation greatly exceeds precipitation, which rnakes it a harsh environment for plant growth. The trees mentioned below can form open forest savanna relying on ground water and deep rooting systems. The northern Ximit of these trees would appear to be defined by winter temperatures and most of the endemic trees in this region are frost-sensitive. Common trees in this region are Acacia spp. Phoenix dactylifiera, Zizjphus spina-christi, Prosopis cineraria, Salvadora persica and Tanarix spp. These trees may reach a considerable height under favourable conditions, that is to say 5-10 meters. This region includes the southern part of Iraq, but little of the original vegetation remains.

Sudano-Deccaniun This region forms a belt of tropical vegcrstion running parallel but to the south of the Saharo-Sindian region; unlike the former it has a high diversity of species. The trees which most concern us here are the following. Dalbergin spp.. Bosweifia spp. Ficus spp., Avicennia marina, Acacia spp., Maerua sp., Moringa sp.

r~ortl~crn elevated areas. At high altitudes where temperatures are relatively lower and mr~il;ill I ~ ~ y l ispecics ~ r , such as Betula verrucosa and Pinus brutia are to be found (Guest 1966, 85). whlle 111 111clowlands one finds Phoenix dactylifera, Zizyphus spina-christi and Acacia iraqensis. The . I ~ I I N I I~mits ~ ~ of many species define the present-day natural distributions. However, past climatic c orrirlrons, even if only marginally different, would give rise to a different zonation and different ;~lrirudclimits.

IIIC

~(~c~r.c\t evolution during the holocene in the Near and Middle East Pollen evidence from 1.1he scdi~nentsindicates that during the latter part of the Pleistocene climatic conditions in the Mldtllc East were such that tree species survived only in rtfuge areas and much of the territory w .I, dry sleppe, probably rather like that which occurs on the h~gll! !iilln plateau at present. In gcrlcral, lakes at the beginning of the Holocene in the Middle East appeir to have k e n more cxtcr~siveand to 'have subsequently shrunk or dried up. Thus while the broad vegetational history II:I\ emerged, the climatic reasons for these changes in vegetation are not fully understood. Major refuge areas wherc deciduous forest species could survive the adverse climatic conditions during the late Pleistocene were probably restricted to the coastal areas of the Black Sca, the Mediterranean and the Caspian (van Zeist & Bottema 1982). In addition it is possible 11l;lt thcre were minor refuge areas where particular micro-dimates favoured tree growth. There is wrnc evidence that the vegetation zones migrated southwards but this has still to be confirmed. Some ten to fourteen thousand years ago a climatic change occured which favoured the expansion of tree species in Turkey. Iraq, Syria and Iran (van Zeist & Bottema 1982). Evidence lrom carbonised plant remains from early Neolithic archaeological sites suggest that t ~ species e were more widespread in Syria (see below). Similar changes occured in Africa and north America (Lkrine 1989). In the Middle East this amelioration apparently coincided with the transition from hunlcr gathering to farming communities. The expansion of tree species was gradual and reached a maximum some 6000-4000 B.P. It was not simultaneous in all areas because of two factors, the d~stanceof sites from refuge areas and the differing rates of co!onisation. Thus the expansion of forest species was later at Lake Van than at Lake Ghab (van Zeist & Bottema 1982). From our point of riew we should note that the birth of the city states in Mesopotamia coincides with the maximum forest cover in the region and timber resources would have been more available then than at any time since. After this period of maximum expansion of the forest species there is some evidence of decLinc, hut this is variable from area Lo area and it is not clear whether this was due to climatic factors or to deforestation by man, through over-exploitation. When one examines the vegetation of the Middle East today, there is little doubt that much is highly degraded. But it is not easy to establish when this deterioration began. Indeed it is probable that the degradation of the vegetation occured at different times in different places depending on population pressure. Evidence from charcoal at A ~ v a nin eastern Anatolia and Bosra in southern Syria would seem to indicate that deforestation was a late phenomenon.

Some man-made factors affecting forest deterioration

Forest zonation in Western and Central Asia Within eat.h phytogcognplural i~gi(.r)w r may observe different pliurt formatiorls resulting from differen[ c ~ iif~t1u.11 * : ! i d i , ~ l i t f i i i i ~ ~ri >i ,i i ~ d ixing , B

the mcxt important. 11.1 Iraq for caample, there is a markc,!

'3 7

:

? :

:. hc

7q,?~~-t

ftnC ++:d,;~!,, ar;J :b5

The role of fire m i l e wme forests depend upon fire for germination and healthy gmwth is ft;r ~ x ~ i m in p [ Austr,dia) ~ then. i l no cvidirncr lor i b ~ cin ih; Zvliddle East. On the contrary, one i t as h i n r datriwi.nt$ As man': population inazai.rd so must l l risk ~ of forest flrc. it:! o,ljv

Timber and Ttew

Timber and Trees

In many areas today this risk has been reduced by drastic thinning of the forests so that the self perpetuating forest fire could not develop. One can imagine that in the past the frequency of forest fires would have been greater than at present.

The role of grazers and browsers Man's herds, particularly sheep and goats, prevent forest regeneration because they browse on the young shoots of newly germinated seedlings. While this was also the case for wild ungulates, the fact that man became sedentary put increased pressure on particular areas, lessening the survival of seedlings, Soil erosion Removal of the ground cover by fire or grazing increases soil erosion. Added to this, the felling of trees would also leave the soil more susceptible to erosion.

Types of forest exploitation Dry (dead) wood gathered as fuel Freshly cut wood from a live tree or shrub is difficult to bum because it contains high levels of moisture. The gathering of dead wood is prefened if no

stocks have been made to allow wood to season. This method requires a large catchment area and is therefore best suited to hunterlgatheree and nomadic peoples. It has the great advantage of not destroying the vegetation. Timber for any purpose is best cut during the dormant period before the sap has risen (dry season or winter depending on the region). Rough timber for construction is not necessarily seasoned but when cut green it will undergo considerable shrinkage. Seasoning is essential for cabinet making. Pollizrding This is the cutting of higher branches which are then allowed to regenerate for a period of several years before cutting again This method is not often seen in the Middle East; it is however common in Europe and nonhwest India where Prosopis cineraria and Acacia spp. trees are cut in this way, the leaves being used as fodder while branches are kept for fire wood. Salk spp. may also be cut in this way to obtain young shoots for basket-making. Coppicing Trees are cut at ground level and the stocks are encouraged to mruub usually a feu trees are left as standards to encourage the upward growth of the young saplings. This method is less suitable in semi-arid environments and is only effective for fiawood production from Corylus, Carpinus and Betulus. Uprooting More common than one might expect, especially where tools are elementary and trees are scarce. Uprooting is also carried out when there is clearing for agricultural land. Felling The felling of large trees for timber could be accomplished using polished stone axes and does not require sophisticated tools. The division of large trunks was probably carried out using a splitting technique with wedges. Seasoning of timber With the development of settled village life it is probable that the systematic felling of timber and its subsequent storage and seasoning was adopted. Wood is usable as fuel after one year of drying; when green it smokes excessively and bums poorly. Where high temperatures were required this period may have been extended.

Specialised use of timber Architecture The principal timbers in buildings made from mudbrick are the roof beams which support the flat moves (Aurenche 1981). The span is limited by the dimensions of the timber available and by its strength. For example poplar and oak beams cannot satisfactorily span much more than 5 meters without some kind of other support either from a truss or from a vertical pillar. Populus beams are the most common in areas where this tree is available. They frequently occur preserved in the carbonised state after a conflagration, where burning timbers are covered by earth from the roof structure. In other cases Quercus. Pistacia, and Celtis, have also been identified in this context (see Willcox 1990a, 1991akb; van Zeist 1984). In southern Arabia it is not uncommon to find date palm trunks used as roof beams. At Pella in Jordan during the Byzantine period date was used in roof construction, (Willcox 1983, unpublished report). Furniture Fine-grained woods which are resistant to insect attack are the most suitable and include a large number of species. Boat building Modem boat builders are extremely selective about their timber and rely mainly on imported tropical woods. Coniferous woods for example are not considered good enough. However this may not have been the case in the past. In considering wood from the Gulf, Theophrastes in Hist. IV, VII 7-8 mentions boat-building wood from this area, interpreted by Bretzel (1903), as being wood from the mangrove tree which grew on the coast of Bahrain. This would seem be to be an unlikely timber tree. Could it be possible that he was refering to Acacia?

Trees and shrubs as sources of fuel Wood was undoubtedly the major fuel in the Middle East until the twentieth century, and the vast majority of charcoal recovered from archaeological sites represents wood used as fuel. Most woody species can be used for fuel. Some are preferable because they burn slowly giving out a constant heat. These are the dense woods such as Quercus spp. Others bum quickly giving out an intense heat over a short period of time such as Populus spp., Fraxinus spp. and Acer spp. For domestic use such as cooking and heating, the most readily available wood is used and this probably means the nearest to the habitation, regardless of how it burns or its thorniness. While timber for construction may have been the object of trade over relatively long distances, this was probably not the case for fuel. Small branches are frequently used for fuel whereas large sections are used for timber. This difference is often visible among the charcoal remains. The presence of small twigs would appear to indicate that a species is local and is proposed as a criterion for evidence that the t m g a w in the neighbourhood of the site since it would be impratical to transport anythmg except large-section branches or trunks. Charcoal manufacture has not to my knowledge been identified from the Near East. Large quantities of wood are necessary, and the finished product is easily transported over long distances, being extremely light. Today one sees it used in the urban situation or for smelting where high temperatures are required. Possible alternative combustible materials include the following-:

Timber and Trcca

Timber and Trees

Dungcake fuel The use of dungcake fuel has been described by Miller (1984). In general the use of this material can be seen today in traditional societies from Turkey to Mia. Its utilisation may occur in areas where there is abundant timber. To make dungcake fuel cow dung is mixed with chaff or chopped straw and left to dry, often by sticking the cakes onto a vertical south-facing wall. The use of dung cake has been tentatively identified at Bosra (Willcox in prep) and at Malyan (Miller 1985). Straw Straw is sometimes seen to be used for ephemeral fires. It can be used for starting a fire or even for cooking flat Bedouin bread. Its continual use should theoretically leave a residue of silica phytoliths. Chaff This consists of the glumes, awns, rachis fragments, and chopped straw, left after threshing and winnowing ( H i b a n 1984). As mentioned above it is frequently mixed with mw dung to make dung-cake fuel. As a combustible by itself it tends to smoke excessively. Annuals When nothing else is available the dry stems of wild annuals are sometimes collected in the desert regions of the Middle East and used in domestic heanhs when the fire is required for a short period such as for heating water or mi&. Stems of cultivars such as cotton or sesame may also be used. Reeds Phragmites australis and Arundo dona both rorm stands of dense reeds in moist habitats. They have a multitude of uses, fuel being only one; I have observed the latter collected for firing brick kilns in southern Afghanistan where timber is extremely scanx.

(* possible or potential timber trees)

Long distance transportation of timber High quality timber could have been imported over large distances and there is indeed both archaeobotanical and epigraphid evidence for the importation of timber by the historical period and possible evidence for the prehistoric period (see below). Below is a list of potential timber trees from western Asia, including some species fmm the hdus and northern Oman, but not including shrubs which would have been used for fuel and would have been gathered on a more local scale (see above).

Taxacaceae Taxus baccata

List of trees and shrubs native to Iraq which could have been used for fuel

Mountain forest (Zagros foothills) Quercus iMectoria* Q. libani* Q. aegilops* Acer cinerascens Crataegus azarolus* C . monogyna Pyrus syriaca* Prunus mahaleb P. amygdalus* Cotoneaster numrnularia Amelanchier integrifolia Cornus australis* Rhamnus spp. Paliurus spina-christi Ulmus sp.* Celtis tournefortii Ceris siliquastrurn* Betula verrucosa Pinus halepensis* Juniperus oxycedrus J. polycarpos*

GaUery forest (TigrisEuphrates) Salix spp.* Populus euphratica* Fraxinus syriaca* Platanus orientalis* Juglans regia* Vitex agnus-castus Tamarix aphylla* T. pentandra

Desert steppe Acacia gerrardii* Hakoxylon spp. Zizyphus nurnmularia 2.spina-christi* Prosopis farcta Nitraria retusa Calligonum comosum

Steppe Pistacia utlanrica* P. khinjuk Prunus micmiarpa Y.arabica Ephedra spp.

Cypressaceae Cupressus sempervirens Juniperus oxycedrus J . excelsa

J . druuacea

J . polycarpos

Pinaceae Cedrus libani

Picea orientalis Abies cilicica

Yew. Asia Minor. This is a compact, durable wood, sought after for bows and for cabinet-making. Present-day distribution may be much reduced. Cypress. Eastern Mediterranean. Can grow very tall, specific timber uses not known. Juniper. Western Asia. Possible use as timber. Juniper. Asia Minor and the Levant. This tree can grow to a considerable size, and some authorities suggest it could have been the biblical cedar. Juniper. Asia Minor and north Syria. Has edible fruits. Juniper. Oman, Iraq (rare), southern Afghanistan, Pakistan. Similar to J. excelsa but with a more easterly and southerly distribution.

Cedar of Lebanon. Southern Asia Minor and the Levant. Normally thought of as the main timber tree of Western Asia, though rarely confirmed by identification. Now drastically reduced in terms of its former distribution in Lebanon. Spruce. N.E. Asia Minor. A conifer more usually associated with cooler climates. Produces high-quality timber. Fir. Taurus. Identified from Byblos and Bosra. Suggests that timber was being exploited in Asia Minor at an early date.

Timber and Trees

Azadirachta indiea

Neem. hnjab/Indus. Large tree, used for timber.

Tamaricaceae Tamarix aphylln, T. articulata Tamarisk. Bolh these trees grow in the more southerly part of Western Asia and grow to considerable height. They are frequently cut for timber, though the quality is poor. A vicenniuceae

Avicennia marina

Salvadoraceae Salvudora persica

Palmaceae Phoenix dactylifera

Mangrove. Gulf, Bahrein and Oman. A tree occuring in muddy salt marshes and tidal creeks. Bark used for tanning. Other uses not known. Tooth brush tree. Oman, Baluchistan and southern Iran. Used for the fabrication of tooth brushes. Identified from Mundigak outside its present-day distribution (Casal 1961). Date. Southern Iraq (see Guest 1966). Trunks of date palm are frequently used for roofing beams. Identified from Pella in the Jordan valley.

It is probable that southern Iraq lacked quality timber during the Sumerian period and for this reason timber had to be either cultivated or imported from elsewhere. Timber from the Taurus and Lebanon is the most obvious source but other areas should be considered. For example the Zagros mountains or the Caspian area could have supplied large quantities of high-quality timber. Timber trees available in the Zagros mountain zone have already been listed and there is little doubt that this area was the nearest region possessing timber of reasonable quality Timber from the south, that is to say, southern Arabia, is a possibility. For example the Omani mountains may have been a source of timber. Here Acacia spp., Maerua crassijliu, Prosopis cineraria, Zizyphus spinashristi, Tamarix aphylla, Juniperus polycarpos, Olea sp., and Teucella sp. would be candidates. In southern Oman and the Yemen we move into another vegetational zone with a great variety of small trees and shrubs, the majority of which do not produce good timber; the only species worthy of mention is Euclea schimperi, which has dad; hard red wood much prized for making small objects (Miller & Moms 1988). In Sind, that is the Indus region in the drier desert areas, the same group of trees are available which we find in northern Oman. In the areas further north we meet yet another zone of vegetation leading up into the Himalayan foothills which is beyond the scope of this survey. However, in the northern part of Sind there are several timber species mentioned in the list above. Methods of charcoal analysis Presence of charcoal. Charcoal occurs in archaeological sediments either associated with hearths, conflagrations or in rubbish pits. It results from incomplete combustion where the supply of oxygen has been cut off during burning. The most

common reasons for this are: when a fire is stifled by its own ash, when a fire is intentionally extinguished, or in the case of a burnt building, when burning beams are covered by earth falling from the roof which suffocates the fire. Hearths are by far the most common source of charcoal and are readily sampled. It is here that we find concentrations of charcoal. Frequently charcoal also occurs as dispelsed fragments in floor deposits, fill, destruction and even in mudbrick deposits. When charcoal is found thinly diffused throughout a sediment such as mudbrick there is a possibility of it being residual and it may be of earlier date than the layer in which it occurs. It is therefore of less value. Methods of recovery Flotation is the most suitable method for the recovery of all carbonised remains. Where there is evidence of in situ burning the archaeological unit should be subjected to flotation. It should perhaps be pointed out that sediments in the Middle East react well to simple flotation techniques because of the lack of humic and clay colloids, which in Europe tend to cause the sediment to cling to the carbon thus preventing it from floating. This is not the case in semi-arid areas. Flotation is preferable to hand sampling for the following reasons: 1) seed remains invisible to the naked eye often occur in the same deposits and so should be sampled at the same time. 2) some species of wood tend to break up more easily and with hand sampling excavators tend to introduce a bias by only collecting the larger fragments. 3) flotation separates the sediment from the charcoal with a minimum of mechanical s a s s . The most suitable technique for flotation and one which I have used on many sites in the Middle East is decribed by Williams (1973). Charcoal which does not float is caught in the wet sieve (3mm diameter). After flotation, the charcoal and other carbonised material such as grain should be slowly dried. Rapid drying causes the charcoal to fracture. Identification Identification is carried out using a reflected light microscope with dark field. The charcoal is fractured in the following planes, cross section, tangential section and longitudinal section, in order to expose the diagnostic features. In the Middle East identification is normally made to genus level only; however in Europe specific identifications are possible because of the reduced number of species (Vernet & ThiCbault 1987). While several publications may aid identification there is no substitute for modem reference material. This is collected during field excunions and is then prepared by heating to three to four hundred degrees centigrade and cutting off the oxygen supply by wrapping in aluminium foil leaving a small hole for the volatile gases to escape. After three or four hours one is left with nearly pure charcoal. Quantitative analysis On Near Eastern sites recovery is often only partial and the quantity of charcoal which survives is the result of the fortuitous circumstances of the carbonisation process. Neither the volume, weight, or number of fragments necessarily relate to the quantities burnt in the original fire, or the available timber. Thus to avoid bias, presencelabsence analyses are well adapted to the data when the percentage of archaeological units as a total of all units sampled with a given species is used as a measure of abundance. With this method it is necessary to have a minimum number of samples from each period or site before a quantitative estimate can be made on the availability of timber by comparing one site (or phase) with another. When the minimum number of samples is not available, phases or periods must be grouped together as in the case of Can Hasan (Willcox 1991b) where several phases were treated as a single period. A

Willwx

Timber and Trees

Timber ad lLrr

Willwx

sample is equal to the sum total of charcoal from any one archaeological unit. This may vary from a few small fragments to several litres. To compare different periods or sites there must be a minimum of ten, preferably more, units sampled from each period. In European cave sites where preservation, recovery techniques and excavation offer more complete data, researchers have used volume, number of fragments, and weight in order to make a quantitative assessment (Chabal 1990, Vernet 1990, Heinz 1990). Their results show that these methods are valid for these particular sites and recovery methods. Similar methods were also used at Malyan in Iran with convincing results (Miller 1985).

Charcoal data from selected sites in Western and Central Asia In this section I have attempted to amass as much information as possible on charcoal identification from the Near and Middle East including much of my own unpublished material. In some cases a list of identifications has been given, in other cases the bibliographical reference. Each site is listed by country with a note on location and archaeological context. Presence analysis was used as a measure of frequency from each phase or site. Thus presence within a sample (= archaeological unit), regardless of quantity, is the criterion. Where sample numbers are less than 5, percentages have not been calculated.

+ ++

+++ ++++ *

= = = = =

present in 1-25% of samples present ** in 25-50% of samples " 50-75%" " '* w 75-100 w v presence where the sample size is ); also made of asal and ildag

kul gisalx, "handleH(?) of steering oar: no. 144, also made of asal and gulbu

bu-ba, reading and meaning uncertain: no. 6

"ma dim", probably part of a boat: no. 92 (transliteration only; cf. Salonen 1939191f.)

bar be-da, probably some type of pole: no. 22; also made of haShur and usuh

ma gii, part of a boat: no 92. (2 times; cf. Salonen 1939:94))

gi3 gid-da, probably spear shafts: nos. 6, 103; also made of gigid and usuh

MUSXMUS,reading and meaning uncertain: no. 92 (2 times)

giS na, reading and meaning uncertain: no. 6; cf. perhaps giS gi na made of usuh)

NI-du,, reading and meaning uncertain, in other contexts perhaps a house-like structure: no. 48

g s zi nh, "wall" of bed: no. 17; also made of gigid and ildag

umbin ellag, perhaps "beams" for "feet" or for "wheels": no. 103; ellag is also made from Sinig

ha-zi, axemandle): no. 44, also made of asal

zag 18 f

kul gisalx, "handle"(?) of steering oar: no. 6, 28; also made of asal and gigid

all four walls of a gi4 (ma!# in the preliminary version (HSS 3 42), but in the h a l version (DP 613) the shorter

sides are called by the mensurational term sag.

~ reading , and meaning uncertain: no.

92 (2 times)

gul-bu, reading and meaning uncertain. The proposal of A. Salonen (Agricultura Mesopotarnica 1968:502) to see in this tree the Sumerian version of dulbu has recently been affirmed by G. Selz (1989 p. 518 with lit.). However, as P. Steinkeller has pointed out to me, use of BU as a phonetic sign to write /bu/ is hard to document in third millennium texts, and, when one examines the

ma gid, pole of cart: no. 103; also made of Sedu ma gid har-hada-ka, pole of a harhadak-vehicle: nos. 17, 18, 22; also made of Sedu ma gid mar, pole of cart: nos. 17, 18, 22, 48; also made of asal ma gid nig Su-ka, pole of a ningSuk: nos. 17, 18, 22, 44, 45; also made of Sedu mar, cart or waggon erin, (or giS erin,) mar, yoke of cart: nos. 17, 18; also made of asal

Powell

Timber production

Powell

ki sag mar, "forehead"(?) of cart: no. 48; also made of asai

assumed to be a variety of poplar, though we have no means of substantiating this at present, and we have already noted above (under asal) the opinion of Flora of Iraq that only the Euphrates poplar is native to Iraq. The ildag tree stands at the beginning ~f the second "labor" imposed upon Enmerkar by the Lord of Aratta (Enmerkar 398-411 = S. Cohen [Diss. Pensylvania, 19731 p. 81 131f. + commentary p. 268-270, with blhliography of proposed identifications), namely, to produce a scepter which is not of any kind of wood c.r tree: it should not be of "ildagx (wr. GI~.A.AM), nor of d~im.gig (kanaktu), nor of @eren, nor of %u-fir-me, nor of "ha-Su-hr, nor of %iSkarin, nor of "esi, nor of gJza-ba-lum,nor of dasal lam the wood of chariots (gigir-ra), and not of flkid-da the wood of whips (usan,-na). Perhaps it is significant that, of the eight trees recorded as being grown or harvested in the Presargonic Girsu documents, only ildag and asal occur in this list. In the Presargonic records ildag is attested as producing firewood, u gibil (no. 35) and bundles of branches (nos. 4, 18, 35, 44, 92), as well as being used for the following specific purposes.

ri-gi4-bi-18, a cabinet-like structure(?): nos. 22, 28; also made of asal Sag, si ri-gi4-bi-16, crosspieces for a rigibilu: nos. 18, 22, 28; also made of asat "[ ...I

tu", reading and meaning uncertain: no. 92 (transliteration only)

haShur Identification of this tree with the apple seems fairly certain (Powell 1987b). It produces fruit (Deimel 1925, nos. B 1-6, 8-12) and the kind of wood one expects from the apple. "Trunks" are occasionally mentioned (no. 123), as are "trimmed branches" (nos. 48, 109, 123) and it produces large numbers of "bundles of branches" (nos. 4, 18, 23, 26, 28, 48, 107, 109, 118). suggesting that a lot of apple wood derived from winter pruning. Likewise characteristic is the frequent use of apple wood for various kinds of "pegs", some of which probably had only a vague resemblance to our idea of a peg. Specific uses are as follows. bar be-da, probably some type of rod (reading not entirely certain): no. 22; also made of gu~buand usuh

gag, "pegs" of various kinds eme gag zag, meaning uncertain (see gag KA lh): no. 6 gag ha-Bar-ra-na, meaning and reading uncertain: nos. 4, 6, 22, 26, 28, 111; also made of Sedu gag ha-har-ra-na gal-galltur-tur, "large/small gag haharana): no. 118 giS.ur/pa kud gag ha-bar-ra-na, "trunksltrimmed branches for gag haharana": no. 123 gag KA ki, reading and meaning uncertain: no. 4 (transliteration only; read zu or eme! for KA?) gag mas, meaning uncertain: nos. 109, 111 Cf. gag zag-ga gag m d , without specification of type of wood in nos. 22 and 28

giS.ur/pa kud gag maS: no. 123 gag ti mar, "pegs" for Cart ribs: nos. 4, 6, 26, 28, 118

Timber production

I 1

!

apin, plough a apin, ''arm" of plough: nos. 18, 35, 92 (2 times); also made of asal and gulbu dam apin, "wife" of plough: no. 123; also made of asal and gulbu bu-bu-ra, reading and meaning uncertain: nos. 6, 92; also made of gigid du, nig Su-ka, "dun for a ningbk-vehicle", perhaps part of the pole assembly: 110s. 44, 45; cf. Ser7 du,-na under Sedu

giS RU, reading and meaning uncertain: no. 75; also made of haShur giS ru.ur-ka, reading, order of signs and meaning uncertain: no. 35 g% zi nh, "wall" of a bed: no. 92; also made of gigid and gulbu nag-ku, reading and meaning uncertain, perhaps a container: no. 16; also made of asal and haShur

Cf. also, without specification of type of wood: pa kud giS ti mar, "trimmed branches as wood for waggon ribs"

sag, probably "head" of plough: no. 92 (2 times); cf. sag apin made of asal

(no. 109), and giS ti mar, "wood for waggon ribs" (nos. 22, 111)

ux- (or i-) Sub an-na har-ha-da, "upper uSub of a harhadak-vehicle": nos. 4, 18, 92 (2 times), 103

g%.urlpa kud gag ti mar, "trunks and trimmed branches for waggon ribs": no. 123 gag til, reading and meaning uncertain: no. 6 (cf. ti1 under a d ) gag zag-ga, meaning ~IxxXtain:nos. 4, 6, 26, 111 (of a waggon), Cf. gag zag-ga gag maS, without specilkation of type of wood (nos. 22. 28)

giS.ur gag zag-ga, "trunks" to make gag zag-ga: no. 123 pa kud gag, trimmed branches to make gag: no. 48 Sen gag, meaning uncertain: no. 6 giS RU, reading and meaning uncertain: no. 123; also made of ~ d a g B. Eichler, JAOS 103 (1983) 101f., offers the meaning "javelin" for giS.RU, but on the whole this does not sound like the kind of thing to be made from apple wood, and gi.4 RU may be an abbreviation for something else. See, in particular, under ildag.

nag-ku, reading and meaning uncertain (perhaps a container): nos. 16, 22, 28, 48, 109; also made of asal and ildag

ru-gh-ba-ka or perhaps giS ru-gu-ba-ka, meaning uncertain: nos. 22, 28; cf. gi.4 RU ti mar, "ribs" for a cart or waggon: nos. 22, 111 ildagx This tree, borrowed into Akkadian as ildakku and also equated with adiiru, is usually

(-Sub)

Se-du,, The reading is purely conventional, since we can neither identify the tree with certainty nor is an Akkadian equivalent known. The only thing that can be said in favor of the "literal" interpretation of the signs as "sweet grain (tree)" currently popular among Sumerologists is that Se is not normally used as a phonetic sign in these texts. The Presargonic evidence suggests a small to medium-sized tree, but finding a suitable candidate is far from easy. "Trunks" (no. 109) and bundles of branches (no. 4) are rarely mentioned. It is used for cart/waggon poles and other parts, and it is the only tree used to make bows. It is barely possible that there is some connection between Se-du,, and Se-du-a, equated with supiilu, which is usu4dy thought to be a type of conifer. One naturally thinks of acacia (Acacia nilotica or arabica), which was used in ancient Egypt to make bows (Meiggs 1982:404, citing literature and two examples in the Ashmolean) and which is known as the babul tree in India (Gamble 1922:292-294; TRADA 1979:265f.), where it sometimes reaches a height of 15 to 18 m. and a diameter of 0.75 m. with a clear bole of ca. 7 m. and is used for a wide range of objects (agricultural tools, handles, oars, pounders, anvil blocks, boat parts, railway sleepers, etc.) requiring a durable strong wood. However, W. van Zeist expressed skepticism in Heidelberg about the possibility of acacia growing in Iraq in ancient times, and the Flora of Iraq (l'ownsend/Guest 1974:44-53) regards only one species as being native

Powell

Timber production

(Acacia negevensis as being definitely native); moreover, even though Flora seems to entertain the idea that A. nilotica could also be native to Iraq, both of these species are said to be extremely rare. This leaves us in a bit of a quandry, and I fear we shall simply have to await botanical evidence from methods that give a more representative sample than we have at present. The specific uses attested for Se-du,, ape the following. ban, bows: nos. 22, 27, 28, 132 gag, pegs gag ha-har-ra-na, rI-Ieaning ~ n ~ e r t a i nos. n : 4, 22, 27, 28; also made of haShur 18 zag-ga, meaning uncertain (cf. perhaps gig la = namzaqu, "key"): no. 103 ma gid, pole of cart: nos. 16, 27, 40, 103; also made of asal and gulbu ma gid ha-harda-ka, pole for a harhadak-vehicle: no. 132 (cf. no. 103); also made of gulbu ma gid nig Su-ka, pole for a ningsuk-vehicle: nos. 18 (), 132; also made of gulbu mud, probably part of the pole assembly of a cartlwaggon: no 40; also made of a s d sahar-gi,, probably "fender" or the like: nos. 22, 28, 132 (copy: gi)

Powell

Timber production

Slightly disturbing to me at the Heidelberg Meeting was the observation by N. Livshits that the spread of Pinus halepensis in the Levant is a phenomenon of fairly recent times, which makes one wonder whether the natural stands of Pinus brutia in northern Iraq are also of recent date or at least more recent than the third millennium. After all, adequate descriptions of the distribution of flora in Iraq do not go back past the present century, and significant changes in distribution patterns are possible over this long period of time. Unfortunately, the botanical evidence discussed at the Heidelberg meeting was completely ambiguous on this issue. It is curious - and perhaps significant - that in poetic passages enumerating trees, the usuh rarely occurs. However, one passage in particular deserves attention. In "Home of the Fish" the usuh is used as a metaphor for the SE+SU~UR.SUMfish: giri, pad-ra-bi giS u-suh, an-na "his skeleton is a tall usuh tree" (line 82, M. Civil, Iraq 23 [I9611 160). Unfortunately, the identification of the fish is even more obscure than the tree, but what can be meant if not the spine with ribs branching out at regular intervals all along its length? This does not sound like Pinus brutia. Another curious feature of the usuh tree that perhaps deserves notice is the rarity with which bundles of branches are mentioned in the Girsu timber texts (only no. 48). The specific uses attested for usuh are insufficient to establish the identity beyond doubt.

"Fender" is based on assumed identity with saFprgli (CAD: "dust guard"), which occurs among parts of the chariot in Angim 51-62 (J.S. Cooper, Return of Ninurta [I9781 pp. 6245, 110).

Sag, si, probably some type of crosspiece; cf. hgsi for a rigihilu under asal and gulbu Sag, si mar, crosspiece or the like for a cartlwaggon: no. 4 B

Ser7 du,-na, probably part of the pole assembly of a cartlwaggon:

no. 40; cf. ako "kr, under asal and

du,-n under ildag

u-suh, Identification of this tree remains problematic. Until recently it has been supposed to be the fir, primarily on the basis the identification of it with Ahies cilicica by I. Low (Flora der Juden V O ~ . 3 [i924] 349f.). However, the evidence presented by Liiw for this identity is of the most tenuous nature, and the fir does not seem to grow in lowland Iraq even under cultivation. Thus, M. St01 (Trees, Mounfains CUI(I Millstones [I9791 68) has made the reasonable proposal to identify it with a tree that has been cultivated there in modern times and which seems to be native at least to the northern region of Iraq, the Calabrian pine, Pinus brutia, a close relative of Pinus halepensis, the Aleppo pine.

B. Landsberger had apparently come to a similar conclusion, for in his work The Date Palm (BAm 17 [I9671 p. 9 nos. 8 and 19), published shortly before his death, he translated u-suh, as "pine" without further comment. In favor of identification with pine is also the fact that one of the characteristic uses of u-suh, is for ship timbers (Salonen 1939:138-142), and in the ancient Mediterranean the trees most often used for ship timbers were fir, pine, cypress, cedar, and larch (Meiggs 198256f.. 86, 118-120). The last is a European tree that becomes important for ship building only in the Roman period. For cypress and cedar we have other Sumerian and Akkadian names that seem more likely than usuh. Moreover, as Meiggs has noted (1982:57), Theophrastus (Hist. plant. 5.7.1) says warships were made of fir because it was light, whereas merchantmen were made of pine because it did not decay, and this observation has been confirmed archaeologically in a particular instance by the recovery of a merchant ship which sank off the coast of Cyprus in the fourth century BC and which was made almost entirely of pine (Meiggs 1982:14). Thus, Pinus brutia seems a viable candidate, but for the benefit of those more distant from the problem, it is perhaps well to underscore the ambivalent character of the botanical evidence.

am-ra, ship timbers: nos. 115, 117, 119 (dagal "wide"), 135 (sig, "narrow") am-ra ma da, timbers for the side(?) of a ship: 50a am-ra ma gur,-ra, timbers for a magur ship: nos. 48* (usuh inferred from context), 50a (transliteration: "-Su, i.e.,

Sh, for ra)

am-ra ma lu-ub-ka, timbers for a malumbak boat, reading not entirely certain: no. 5Oa am-ra ma 50 sila, reading and meaning uncertain: no. 50a The text records 30 of these timbers out of a total of 120 "large and small" (gig h-suh, tur-mah-ba) cut from the garden of Eku. This text demands careful collation, because, if Deimel's reading is correct, it would be not only the oldest attestation of timbers measured in sila but also would imply remarkably large trees (5 cubits in girth or about 80 cm in diameter) being grown in significant numbers. For this method of stating girth of logs. see Powell, " M A und Gewichte" 5

mB,RLA 7 (1990) p. 491, and note that the "formula" for transforming

sila into cubits and vice-versa is: number of cubits-of-girth-squared times 2 = sila, i.e., x sila = 2(y2 cubits). In the present case, 50 sila = 2y2, thus the girth y would be 5 cubits. The text seems to list the timbers from largest to smallest, of which these are the first, which would favor Deimel's reading. If the reading and my tentative

interpretation are correct, the implications of this text are far-reaching indeed, because trees of this size can only have been grown over a period of several generations, which would imply long-term cultivation and administrative practices and would profoundly affect our conceptions of what a kiri actually looked like, and much else besides. The text is dated to the 4th lugal year of Uruinirngina.

a si-ge,, meaning uncertain: no. 12 (a si-ge, gal-gallus-bi/3-kam-ma us, "large / second / third sized asige"); cf. h-suh, gi gid 8-suh, a si, gal-gal (no. 29)

bar bC-da, some type of rod or pole: no. 29; also made of gulbu and h d h u r u-sub, gal-gal bar bbda ig-@, "large usuh for the post(?) of a doorlgate": no. 116 u-suh, bar be-da 4, "usuh for the barbeda of a house": no. 116 gar,, uncertain: no. 119 gi mug, steering oar, oar, or punting pole: nos. 77, 128 (2 times) gi muS seems to denote both "punting pole" and "steering oar", which is not wholly surprising, since both are used for steering boats. Illustrations of both have been collected by Salonen (1939, pl. XXXVI-XXXVII for the punting pole, pl. XLIII.1 for the steeing oar). The meaning punting pole seems to be implied in "Home of the

Powell

Powell

Timber production

Fish" (lines 77-79, M. Civil, Iraq 23 [I9611 161f., 170f.), where ''"gi-muS gi,, "black gimuS" is used metaphorically for the eel (@-bi). Punting pole must also be the meaning of parisu in the Old Babylonian version of Gilgamesh crossing the Waters of Death (see my comments in ZA 72 [I9821 93f. with literature); however, even though parisu is attested as an Akkadian equivalent of gi mug, the loan word gimu33u does not seem to have this meaning in Akkadian (against CAD girnd3u etc. and with AHw and CAD MI1 p. 68 under ma@u 8e). Moreover. "14 gimuS for a big boat (singular!)" @P 476 = no. 128 cited below) do not sound like punting poles. Also "23 gimuS for a (or is it more than one?) magur boat" (no. 50a below) seem odd if these are punting poles. However, the magur boat had a biconvex shape like the gibbous moon (CAD makurru mng. 3, p. 142) that gave its name to a geometrical figure (shown in photo: RA 54 133 lower left = no. Q). Presumably it was a sailing boat. What are they doing with all these gimuS?

gi mu3 ma gu-la, "gimuS of a large boat": no. 128 gi mug m6 gur,, "gimuS of a magur boat": no. 50a gi muS ma Su-ha, "gimuS of a fishing boat": no. 50a gi-na or giS gi-na, uncertain: no. 12 Cf. perhaps rnukiinu, cloth beam(?) of a loom, and giSgun.2, some kind of weapon.

1

giS gid-da, probably spear shafts: nos. 19, 116; also made of gigid and gulbu ma, boat, ship, see am-ra, gi mug, u ma dB, probably the mooring apparatus of a ship: no. 136 This seems to be part of the tarkullu apparatus (see the discussion of "kr, under asal above), interpreted by Salonen (1939:lll-113) as "mast", which seems doubtful. The Sulgi hymn R (STVC 60 obv. 35) mentions

m4-dh alongside the "upper targul" (""m4-dh-zu u targul sag an-na-ka), and there is a variant thr-dh to targul in the Curse of Agade, which is interpreted by J.S. Cooper (line 74. Curse of Agade [I9831 p. 55) as "mooring pole". It seems likely that the m4du (how ever it is to be read) and the targul formed two parts of Enkidu. and the Netherworld (line 295. A. Shaffer. Swnerian Sources f o r . . . Gilgamesh [Diss. Pennsylvania, 19631 p. 94. 119). 1u '"ma-dh ra 8-ni-in-Su-Su, "a md-du striking someone after he had let it fall" corresponds in the Akkadian version (Gilgamesh XI1 144) to 36 ina th-kuJ-Ju [malzsu(?)]. "someone [struck?] with a tarkullii. u ma,

probably plank of a ship or something similar (reading of u not entirely certain) u ma ma du, u for mooring apparatus(?): no. 119 u ma sig ba meaning uncertain: no. 119 u ma U1-la meaning uncertain: no. 119 u ma us meaning uncertain: no. 119 u ma SrZ.ir~,u for a goat-boat(?): no. 119

Rare trees and the botanical data We conclude our survey of Girsu arboriculture by returning to the problem of the rare trees. "White wood" (giS bar,-bar,) was obviously cultivated, because it is attested in a cutting from the garden of Urki used to make rigibilu (nos. 44, 45), and, in another context, but probably also deriving from some garden, it is used for zi-rl-lum (no. 107). Another rare tree also grown in gardens was the giparx, attested in a cutting from the garden of Eku (no. 92) in the form of "trees" (giS), one "trux-k" (giS.ur) and bundled branches (pa sa 18-a). Bundled branches of this tree are also recorded as being stored in a garden warehouse, ganun kiri, (no. 107), and it appears smeared with bitumen (@parxesirx) in a record of boat equipment (no. 135). To round out the picture, we should note that, in addition to the trees clearly attested in documents as being harvested, Sumerian literature seems to imply the cultivation a number of "exotic" trees. These passages raise the same heuristic problem already noted with respect to the

tamarisk and other trees, namely: does presence or absence from the "natural" flora attested in modern times mean that a particular tree was or was not grown in ancient times? Of course, factors like climate and soil conditions will have excluded certain trees, but, where sufficient desire and resources have been brought to bear on a problem, surprising solutions have often been found. Obviously we do not have the facts to solve this problem nor can we discuss it in detail here, but we can illustrate the dimensions of it. Ornamental trees must have been a feature of the Sumerian (and later Mesopotamian) landscape, because, otherwise, the recurrent theme of trees being destroyed would be inexplicable. This theme is very explicitly treated in first millennium sources, but it is also implicit in works like the Lamentation over the Destruction of Sumer and Ur which speaks of "Gutium" (a literary figure for enemies in the east) uprooting and cutting down trees, and the plaintive lament that "the holy dwelling of Nanna, a grove (which had a fragrance) like the perfume of an eren tree, its fragrance is destroyed (lines 87f., 425, P. Michalowski, Mesopotamian Civilizations 1, 1989). In the Old Babylonian period, Harnrnurabi is described in the Prologue to his Code as: "the one who clothed with green the gigund of Aja (in Sippar)". CAD (G p. 68-70 S.V. giguni) came to the conclusion that there was no evidence for planting trees on the giguru2 and thought that "green" must refer to hangings or bricks. However, aside from the fact that glazing of bricks has yet to be demonstrated for the Old Babylonian period, the giguna is frequently associated with tir, "woods", Sumerian literature evinces a close association between the giguna and trees. Thus, the story Naram-Sin's "crimes" against Enlil's temple in Nippur (J.S. Cooper. Curse of Aga& [I9831 p. 56f.) tells that Nararn-Sin ''eren sirSu-k-min%a-ba-lum B't~karin,giS gi-gun,-na-bee&GUM ba-an-sur-sur did GUM sur all around to its eren, Surmin, zabalum, and Ukarin trees, its giguna-trees.

I

the mooring apparatus, so inseparably linked that either could be used as pars pro toto. Thus, in Gilgamesh,

Timber production

I

The "trees" are usually identified as cedar, cypress, juniper, and boxwood, though none of these identifications is beyond debate, but here the question which concerns us is: were these trees planted in the temple precinct or was it wood out of which parts of the temple were made? Both interpretations are theoretically possible, because the precise meaning of the verb remains obscure. I am inclined to believe that it means "uprooted", and, if so, that would solve the question of whether there were trees growing on or at least around the giguna, but, unfortunately, I cannot substantiate this. There is also some rather equivocal evidence for importation and acclimatization of foreign trees. For example, in Nanna-Suen's Journey to Nippur (line 337, Ad. Ferrara, Studia Pohl SM 2, 1973, pp. 78, 105, 155), Nanna-Suen asks Enlil to give him the mag gurum(= GAM)tree (= Akkadian darmadu), which apparently grew in the foothills of the mountains (the alluvial hinterland seems less likely), so that he could take it back to Ur with him: an eden-na gig m d gurum sum-ma-da-ab urii& ga-de, Give me the m d gurum tree in the upper eden, so that I can take it to Ur.

This could, of course, be some literary topos, but I think it is more likely that it contains a double allusion to some specific political and arboricultural activity associated with one of the Ur 111 kings. Finally, we should mention the problems posed by the rather frequent tree metaphors in Sumerian hymns to gods and kings, such as this one placed in the placed in the mouth of ISme-Dagan by some educated courtier (W.Ph. Romer, Swnerische Konigshymnen der Isinzeit p. 51):

Powell

Timber production

'"me6 mah 6r gur-ra pa mu1 dagal-la me-en, Greatest of mes trees, thick of trunk and broad of crown, am I.

Of course, one could argue that these are literary figures of speech. But this merely postpones the problem of explaining how they got into the language in the first place. Would such "tropes" have ever attained any vogue at all if the mes had not been a fairly familiar sight? Here, however, the modern evidence is a bit disheartening, to say the least. As I have pointed out (Powell 1987a:149), the most likely candidate seems to be Celtis australis, but the Flora of Iraq (Townsend et al. 1980372f.) gives its distribution in Iraq as the mountain zone (550 m. and higher), and, while noting that the tree does grow up to 25 m. in Europe, it is said to be only 7-9 m. tall in Iraq. Hardly the sort of thing that would have occurred to a poet of the Isin period trying to impress the king! Thus, if the mes tree is indeed to be identified with Celtis australis, we find ourselves in the all too familiar dilemma: (1) it was being grown in southern Iraq in the OB period outside of its "natural" habitat; (2) it must have been cultivated under conditions sufficiently favorable to allow it to reach a size far beyond that of its modern counterparts.

Conclusion My review of the material leaves me rather pessimistic about our ability to identify ancient trees with modem species. Our only real hope for convincing identification of species is better methods of collecting botanical evidence which can give us a reliable picture of what kinds of trees and other plants were actually present in southern Iraq in the third millennium BC. The present evidence is not even minimally adequate for this task. The written evidence on the other hand suggests the existence of a systematic arboriculture of both fruit and timber trees in what must have been rather large gardens. These gardens included the date, fig, grape, and apple, the latter of which was extensively used for wood, probably both from pruning and from cutting out old or non-productive trees, as well as seven other trees regularly harvested for timber and a few others that are rarely mentioned. Tamarisks seem to have been systematically planted on the banks of irrigation ditches and on the dikes surrounding farms, probably to serve as windbreaks and to hold the banks of the canals or dikes, as well as for timber. Riverain woodlands were probably "wild" only in the sense that they were not planted, for those about which we have information were clearly the property of the state or temple, and we may assume that cutting timber was a prerogative that was just as jealously guarded in the Presargonic period as in Hammurabi's time, when unauthorized felling of trees could cost one's life. Like the tamarisk plantations, the riverain woodlands were administered as a part of the temple or state economy, and, although the documentation is insufficient to trace this in detail, it is clear that specific individuals - or, more likely, families who lived in the immediate proximity - were responsible for guarding, tending, harvesting, and supplying certain quotas of these resources. Thus, although there are many blind spots in the documentation, as a whole, the Girsu timber texts give us an illuminating glimpse into a system of arboriculture which involved rational exploitation of woodlands as well as a remarkably diversified "garden" culture.

Timber production

Powell

BIBLIOGRAPHY Bauer, J. Altsumerische Wirtschaftstexte aus Lagarch (Studia Pohl 9; Rome). 1972 Bor, N.L./Guest, E. Grarnineae. Flora of Iraq, vol. 9, edited by C.C. Townsend, E. Guest, & A. Al-Rawi 1968 (Baghdad, Ministry of Agriculture). Deimel, A. "flma-nu und ii", OrSP 5,42-44. 1922 "Die althmerische Baumwirtschaft", OrSP 16, 1-87. 1925a Desch, H.E. Timber. Its structure and properties. (Fifth edition; London). 1973 Gamble, J.S. A manual of Indian timbers. An account of the growth, distribution, and uses of the trees and 1922 shrubs of India and Ceylon with descriptions of their wood-structure (Reprint of second edition with some additions and corrections, London). Guest, E.R. & Al-Rawi, A. Flora of Iraq. Volume one: Introduction to the Flora. An account of the geology, soils, 1966 climate, and ecology of Iraq with gazetteer, glossary and bibliography (Ministry of Agriculture, Republic of Iraq). LaPlaca, PJ. and Powell, M.A. "The agricultural cycle and the calendar at Pre-Sargonic Girsu" Bulletin on Sumerian 1990 Agriculture 5, 75-104. Meiggs, R. Trees and timber in the ancient Mediterranean world (Oxford). 1982 Powell, M.A. '"The tree section of ur (= BAR)-ra = @bullu", Bulletin on Sumerian Agriculture 3, 145-151. 1987a "Classical sources and the problem of the apricot", Bulletin on Sumerian Agriculture 3, 1987b 153-156. Salonen, A. Wasserfahzeuge (Studia Orientalia Fennica 8, no. IV). 1939 "Bemerkungen zur sumerisch-akkadischen Brennholz-Terminologie", JEOL 18:331-338. 1964 Selz, G.J. Altsumerische Stuttgart). Venvaltungstexte aus LagaS, Teil I (Freiburger altorientalische Studien 1511; 1989 Steinkeller, P. "The Foresters of Umma" in M.A. Powell (ed.), Labor in the Ancient Near East (AOS 68). 1987 pp. 73-115. Thompson, R.C. A dictionary of Assyrian botany [= DAB] (London). 1949 Townsend et al. Flora of Iraq, Vol. 4, Part 1: Cornaceae to Rubiaceae, Vol. 4 , Part 2: Bignoniaceae to 1980 Resedaceae. Edited by C.C. Townsend & E. Guest, with S.A. Omar et al. (Baghdad: Ministry of Agriculture & Agrarian Reform). TRADA (Timber Research and Development Association) Timbers of the world. [Vol. 11 (LancasterINew York). 1979 Timbers of the world. Vol. 2 (LancasterINew York). 1980

Powell

T i r production

ANMERKUNG ZU DEN NEUSUMERISCHEN TEXTEN

Addendum (November 1991): While it seems clear that identification of the usuhJaJu@u tree with fir cannot be correct, identification with a specific species of pine is also problematic. The following passage from L.H. Bailey (The cultivated conifers in North America [New York, 19331 pp. 41f.), which is directed primarily at growers of conifers, will perhaps help the humanist understand why we have so much difficulty in identifying species: "However much the planter may desire an offhand way of recognizing the different species of pines, wild and planted, he will not find it. Long experience in propagating and planting enables a person to name the species with which he deals by acquaintanceship, but this will not aid him greatly in identifying other kinds. The distinguishing of the kinds in any important growing collection requires the habit of close observation and particularly the presence in the hand of foliage and cones, and then patience to compare with keys and descriptions. It is impossible to make a key without exceptions; and the use of a key presupposes a knowledge of the organs or parts with which it deals. The best keys are naturally the most difficult, for they require a study of covered rather than superficial characters".

UBER SCHILFROHR Walther Sallaberger (Miinchen)

Beim 2. Grazer Morgenladischen Symposium zum Thema "Der orientalische Mensch und seine Umwelt" (2.-5. 3. 1989) (die Vortrage wurden publiziert als Grazer Morgenliindische Studien [= GMS] Band 2, herausgegeben von B. Scholz) hatte ich ein Referat "Zum Schilfrohr als Rohstoff in Babylonien" gehalten und dabei versucht, den Weg des Schilfrohres vom Rohricht bis zum Handwerker v.a. anhand von Ur-111-Urkunden zu verfolgen. Auf die von H. Waetzoldt ausfiihrlich diskutierten Termini fiir Rohrarten und Rohrernte braucht hier nicht mehr eingegangen zu werden. An dieser Stelle kann ergwend auf zwei weitere Gesichtspunkte der Rohrgewinnung kurz hingewiesen werden (ausfiihrlicher mit Belegen in GMS 2 [1989], 311-330).'

1 SchilP6felder" Mal3angaben zu einzelnen Fluren der Provinz Umma bezeugen oft mehrere Kektar groBe Sumpfgebiete mit Rohrkstand (z.B. MVN 4 19, 2 1; TCL 5 5675.vi.3-12: insgesamt 29256 sar = ca. 105 ha). Eine besondere Rolle in der Versorgung der Provinz Umma und seiner Handwerksbetriebe mit Schilfrohr kommt der EngabaDU-Flur zu. Vereinzelt ist zwar Gerste oder Weizen vom EngabaDU bezeugt, meist wird es aber als Quelle von Rohr, seltener Binsen und Grasern genannt. Dal3 es sich somit um ein ausgedehntes Sumpfgebiet handelt, beweist auch seine Eage and drei Kanden oder eine Lieferung von Fischen vom EngabaDU (JCS 28, 215 Nr. 26). Genauere Auskunft iiber seine GriiBe gibt ein Text aus dem 4. Monat im 8. Jahre Arnar-Su'enas ( ~ N 465): S

a

10 (bur) 4 (bur) 2 (e3e3) ;iku, gig-@ izi ku-a, en-gaba-DU, 1 sar ba-zk, 8 sa i-ghl-Am, gu-kilib-bi 3-am, gi-bi 211550 sa, gu-kilib-bi 17630-Am 2

: iku

(= 0,95 km2)R(lhricht, vorn Feuer verzehrt, im EngabaDU-(Feld). Wird 1 sar abgemght, ergibt es 8 Biindel, also eines Ballens. Das Rohr davon (= der Flgche) sind 211550 Biindel, das sind 17630 Ballen.

264

3

Eine inhaltliche Bestatigung des Rohrichtbrandes bietet der Text Owen, JCS 24, 172 Nr. 93.9. aus demselben Jahr AS 8: guru8 udl-82 en-gaba-DU, izi-ku-a te-na 8 Arbeiter fiir einen Tag (im) EngabaDU-(Feld) den 'FeuerfraB7loschen.

$

Zudem gibt T ~ N S465 auch Einblick in die Verwaltung der Schilfrohrbestiinde, deren Ertrag vor der Ernte abgeschatzt wurde. Im vorliegenden Fall wurde aber das zu erwartende Rohr noch vor dem Schnitt durch einen Brand vernichtet.

Zur Rohrernte Fur die bei der Rohrernte verwendeten Werkzeuge gibt es nur wenige Hinweise. Bronzesicheln zu je ca. 170 g, um Futterrohr zu schneiden, nennt NATN 469: 10 d u KUV 31 ma-na-ta, gi-zi z6; Lieferungen von Holz f i r Sichelgriffe (zum Rohrschneiden?) an Agu p. Steinkeller, AOS 68, 106f. Doc. Nr. 31, 40), einen fiihrenden Beamten im rohrverarbeitenden Handwerk vori Umma, weisen in dieselbe Richtung. Literarisch (EmeS und Enten 207f.) und in lexikalischen Listen (PDS B, 123) ist dagegen ein bar-huda-Instrument (Civil, AOS 67, 45: "Machete") zum Schneiden (ebenso z k ! ) von Futterrohr bezeugt. 2

!

1

Sallaberger

Anrnerkung zu den neusumerischen Texten Uber SchiEohr

'ROHR' UND DESSEN VERWENDUNGSWEISEN ANHAND DER

W m n d bei der Rohrernte ein GroBteil der Arbeiter beim schneiden2 eingesetzt ist, sind fir Biindeln und Abtransport deutlich weniger Leute erforderlich; die tagliche Arbeitsleistung (Schnitt und Transport) betragt 2-3 Ballen (gu-kilib) zu je 12-15 Biindeln (sa) Rohr (gi), 1,5-3 Ballen zu je 5-6 sa ~utterrohr,~ 1,5-2 Ballen bzw. 13 sa gi-NE und 10 sa gi-$ID.

NEUSUMERISCHEN TEXTE AUS UMMA Hartmut Waetzoldt

Der Rohstoff Rohr wurde - oft nach Zwischenlagerung in Speichern - zum Transport auf Boote geladen (gi ma-a gh-ra) oder zu FWBen gebunden (ma-U-a k8-ri). Bei der (ifters genannten Gruppe gespeicherter Materialien von 12 bis 20 gigaim, "langen Rohxpfosten", 4-8 gii-dal, "Querholzern" (komen auch fehlen) und etwa 330-7400 (transportierten) Rohrbiindeln, handelt es sich wohl um groBe, am Zielort aufgeloste FNBe (vgl. besonders Gomi, Orient 16, 140.).

(Heidelberg)

1

ijber Rohr und damit zusammenh2ngende Fragen gibt es bisher keine umfangreichere Untersuchung. Jedoch bietet Walther Sallaberger (1989) "Zum Schilfrohr als Rohstoff in Babylonien" einen guten h r b l i c k . Auch J.W. Turner (1981) diskutiert in seiner bedauerlicherweise unpublizierten Dissertation (S. 199 - 227) die wlchtigsten Termini. Soweit wie moglich wird hier auf diese beiden Studien verwiesen und das dort gebotene Belegmaterial nicht wiederholt.' In Mesopotamien spielte das Rohr eine grol3e wirtschaftliche Rolle. Verwendet wurde es besonders als Viehfutter, zum Flechten von Korben und Matten, beim Haus- und Kanalbau und schliefilich als Brennmaterial. Zur Verwendung von Rohr und anderen Grasern in iigypten s. Keimer 1984, 72ff. Die zentrale Bedeutung dieses wichtigen Rohstoffes dokurnentieren nicht nur die Verwaltungstexte, sondern sie wird auch durch zahlreiche Funde bei Ausgrabungen bestatigt. So kamen z. B. Reste von Rohrmatten, -tiiren, -korben und -seilen ans ~ageslicht.~ Das Rohr genoB in Mesopotamien eine derart hohe Wertschatzung, daB es dafiir eine eigene Rohrgottheit gab (dingir-gi, M. Krebernik, ZA 75, 166). Verschiedene Henscher riihmen sich am "Rohr-Heiligtum des Gottes Ningirsu" gebaut zu haben (FAOS 6, 74:6). Diese besondere Wertschatzung des Rohrs spiegelt sich auch in der Literatur wieder, so preist 2.B. UrnanSe das Rohr mit folgenden Worten: "Heiliges Rohr, Rohr aus dem Rohricht des Engur, Rohr, deine Zweige sind (rot)braun. Deine Wurzel (hat) Enki (im) Erdloch(?) vorhanden sein lassen. Deine Zweige sind ... . Dein Bart (ist aus) Lapislazuli. Rohr, (dein) Antlitz reicht (bis) ins Fremdland, Rohr, die EM(-Gotter) (und) die Nunki(-Getter) miigen (dir alles) gut machen!" (FAOS 6, 137). Nach einer Hymne auf Ktinig Sulgi (F) gehorte es unter anderem zu den Aufgaben des KOnigs, dafiir Sorge zu tragen, daB reifes und junges Rohr (in geniigender Menge) gebiindelt wurde. Das Streitgespriich zwischen dem Baum und dem Rohr (Tree and Reed) ist nicht vollstiindig erhalten und blieb wohl deshalb' bisher unbearbeitet. Am Ende der Komposition von rund 250 Zeilen erklm Ktinig h l g i den Baum zum Sieger und nicht das so vielseitig verwendbare ~ o h r . 'In vielen anderen Texten findet man ebenfalls ErwWungen von Rohr oder ~ohrgegenstiinddn.~

NOTES 1

Fiir sein freundliches Angebot m6chte ich M. Powell sehr herzlich danken. [Editor's note: M.A. Powell learned of W. Sallaberger's work on reeds through C1. Wilcke in Miinchen and Herr Sallaberger kindly consented to provide us with a copy of his paper. We thank both of these colleagues most heartily and here publish this brief note by Herr Sallaberger to call our reader's attention to his longer work on the subject.]

2

Es handelt sich um die Termini SIC, (meist bei Fkichen; nur Binsen, Graer und Rohr: gi, gi-zi, gi-NE, Monate i, iv-viii, x-xii), kud (meist bei Personen; Dornstriiucher, Kameldorn und Rohr: gi; Monate i, iv-ix, xiii) und zd (bei Rohrarten gi, gi-NE, gi-$ID, gi-zi; Monate v-vi, viii-x; v.a. Lag&)

3

Zu den Belegen in GMS 2 Anm. 64 ist hinzuzufiigen: F. Pomponio, " A Neo-Sumerian Account about Reeds", Orientalia 58 (1989) 230-232 (3 Ballen zu je 5-6 sa gi-zi), PDT 1 368 (3 gu gi-zi). BIBLIOGRAPHICAL ABBREVIATIONS

AOS 67 F. Rochberg-Halton ( 4 . ) Languages, literature and history: philological and historical studies presented to Erica Reiner (New Haven, 1987). AOS 68 M. A. Powell (ed.) Labor in the ancient Near East (New Haven, 1987). GMS

Grazer Morgenlandische Studien

MVN

Materiali per il Vocabolario Neosumerico

Vorbemerkung

NATN D. I. Owen, Neo-Sumerian archival texts primarily from Nippur (Winona Lake 1982).

~ % N S M. Sigrist, Textes kconomiques nko-sm'riens de I' Universite' de Syracuse (F2u-i~1983).

I I

2

Das Wort gi

Bei den hier zu behandelnden Rohr-Texten ergaben sich relativ rasch zwei unterschiedliche Verwendungsbereiche. In Verwaltungstexten iiber Arbeiten auf Feldern, an Kanalufern und in Gwen bezeichnet gi das normale Schilfrohr (in der Regel woM Phragmites communis bzw. australis "Gemeines Schilfrohr"), in Abrechnungen von Kaufleuten und Listen mit

Waetzoldt

Rohr und dessen Venvendungsweisen

Nahrungsmitteln und Essemxn jedoch eine wohl eBbare Substanz ('Wun'-Roh~-1.Das R o t i wird in Traglasten, Biindeln und Gebinden gelieferl, wiihrend man day 'Wiire'-Rohr in Minen wiegt. Demnach wird gi iri den Umma-Texten als eine Art Oberbegriff herwendet.

3

Bei Rahr verwendete MaReinheiten

sa "Bund, Bindel" Die 1x5 weitem wichtigste Einheit ist sa "Biindel", die in &cn Prcivi~lzenwdxenti dsr Herrschaft der 3. Dynastie von Ur Verwendung fand. 3,1

3.2 gu-kilib(-ba j / gu-nigh(-na) "Gebinde" Den Ausdruck gu-kilib(-ha) benutzte man in ~3mma: wiihrend in den anderen Provinzen cher grr-nigh(-na) iiblich war.6 Je nachdem, s b es sich um frisches oder das vie1 leichtere trockemie Rohr ha~delte,konrlte 1 Gebinde aus nur 5 oder such aus 30 ~ u n c Rohr t ~ kstch~r~ MaBgebl~ch . fiir die GroBe des Gebindes war das Gewicht des Rohres, da 1 Person in der Lagc sein rnuBte, es ubcr eine gr6Bere Strecke zu transportieren. Das 0ffnen dieser Gebinde bezeichnete man ofknsichtlich mit dem Verb P "herausgehen lassen" (MYN 14, -1722 gdma-nugu-kilib-h +-a; rnit Rohr Sigrist, Toronto 390:lff.).

gu - rraglast" Dieses gC ist mit 'Traglast" und nicht rnit "Talent" (Gcwichtseinheit von ca. 30 kg) iiberselz~n.~ Es wird nicht wie das Talent in Mincn. sondern in sa "Burid" (JCS 10, 29. 6:lff.j unterteilt. Gelegentlich findet man auch 112, 113 ode! 5/6 gh.g Diese Makinheit ""Traglast" benutzte man bereits in der prasargonischen =it. In der Ur-III-Zit kommt sie z.B. auch hi Fischen, Zwiebeln und Holzern vor.lo 3.3

Wie sich anhand mehrerer Texte zeigen l a t , ist gu 'Traglast" mit gu-kilib "Gebinde" f's. 3.2) mehr oder weniger austauschbar, z. B. MCS 4, 10 BM 105429: 586 f i U