DOMESTIC ANIMALS OF MESOPOTAMIA PART I1
BULLETIN ON SUMERIAN AGRICULTURE Volume VIII
Cambridge 1995
General editors CONTENTS
J.N. Postgate Faculty of Oriental Studies University of Cambridge Sidgwick Avenue Cambridge CB3 9DA U.K.
M.A. Powell Department of History Northern Illinois University De Kalb Illinois 60115 U.S.A.
iv List of contributors and addresses ........................................................................................... Preface
........................................................................................................................... V-vi
Ethnographic evidence for water buffalo and the disposal of animal bones in southern Iraq: Ethnoarchaeology at Al-Hiba ....................................................................... 1-9 Bonnie Gustav & Edward Ochsenschlager Plough and Power: the economic and social significance of cultivation with the ox-drawn ard in the Mediterranean ................................................................................ 11-22 Paul Halstead How can modem food technology help to identify dairy products mentioned in Sumerian texts? ... 23-3 1 Michael Teuber Late Uruk Period cattle and dairy products: evidence from proto-cuneiform sources .....................33-48 Robert K. Englund Sheep and goat terminology in Ur I11 sources from Drehem ....................................................49-70 Piotr Steinkeller Plow animal inspection records from Ur 111Girsu and Umma .............................................. 7 1- 171 Wolfgang Heimpel
ISSN 0267-0658 Orders may be placed through booksellers or direct to the Sumerian Agriculture Group, Faculty of Oriental Studies, Sidgwick Avenue, Cambridge CB3 9DA, U.K.
Old Babylonian cattle ..................................................................................................173-213 Marten St01 215-240 Cattle in the Neo-Babylonian period ............................................................................... G. van Driel
O~umerianAgriculture Group 1995. 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 England by Aris & Phillips Ltd., Warminster, Wiltshire
iii
Addresses of contributors Dr. G. van Driel
Dr. R.K. Englund
Department of Assyriology, University of Leiden, POB 9515, 2300 RA Leiden, The Netherlands Seminar fiir Vorderasiatische Altertumskunde, FreieUniversitiit Berlin, Bitterstr. 8-12, D 14195 Berlin, Germany
Bonnie Gustav
Department of Anthropology, Brooklyn College, Brooklyn, NY 11210, U.S.A.
Dr. P. Halstead
Department of Archaeology and Prehistory, University of Sheffield, Sheffield S10 2TN, U.K.
Prof. W. Heimpel
Department of the Near and Middle East, University of California, Berkeley, Ca. 94703, U.S.A.
PREFACE In our second volume on the domestic animals of Mesopotamia we resume the topic of sheep and goats, in the shape of the article of Piotr Steinkeller, and move into the topic of cattle discussed at the 1993meeting of the Sunerian Agriculture Group at Wassenaar, The Netherlands. This therefore gives us the opportunity to express our considerable gratitude for the hospitality of our Leiden hosts at that meeting which took place in the ideal surroundings of the NIAS Centre for Advanced Studies at Wassenaar: we are delighted to acknowledge the substantial contribution towards the costs of mounting the conference and towards the expenses of the participants received from the Royal Dutch Academy, from the CNWS at Leiden and from the Mesopotamian funds of the Department of Archaeology, Leiden University. We are most grateful to the NIAS and its staff for their kind and generous reception of the participants; and to our colleagues Marten St01 and Govert van Driel in particular, very many thanks from us all.
Prof. E.L. Ochsenschlager Department of Anthropology, Brooklyn College, Brooklyn, NY 11210, U.S.A. Prof. P. Steinkeller
Department of Near Eastern Languages and Civilizations, Harvard University, 6 Divinity Avenue, Cambridge, Ma. 02138, U.S.A.
Prof. M. St01
Heivlinder 27,2317 JS Leiden, The Netherlands
Prof. M. Teuber
Department of Food Science, ETEI Zentrum, CH-8092 Ziirich, Switzerland
While on the subject of Leiden, it is very satisfactory for us to be able to report that Miguel Civil's edition of the Fanner's Instructions or Sumerian Georgica, presented to our meeting there in 1987 (see BSA 5, p. iv) has now appeared, through the co-operation of the Oriental Institute at Barcelona: M. Civil, The Fanner's Instructions: A Sumerian agricultural manual (Aula Orientalis - Supplementa 5; 1994, Editorial AUSA Apartado de C o m s 101, 08280 Sabadell (Barcelona), Spain. ISBN 84-8881003-2). The volume includes the reconstructed text and translation, a general commentary, discussion of some specific points, and an edition of new texts from Chicago.
On a sadder note, we have to record the death of Evan Guest in 1992, at the age of 90. Anyone who has worked on the natural resources of Mesopotamia must be familiar with the superb volumes of the Flora of Iraq, a series sadly suspended by the political and financial disasters which have overtaken that country. The first, and enormously informative, introductory volume of the Flora was mainly from the pen of Evan Guest, and during his long association with the Royal Herbarium at Kew he was instrumental in pushing the series along, contributing notes on the ethnography of the plants. Evan Guest attended the first meeting of the Sumerian Agriculture Group in Cambridge in 1983, and those who were present were privileged to hear an absorbing account of his early years in the service of the Iraq government agricultural administration. An obituary notice and a sketch appeared in the Daily Telegraph for the 3rd April 1992.
This is almost certainly the penultimate volume of the Bulletin. No further meetings of the Group are planned, but a ninth volume of the Bulletin 'is expected to include some further papers on sheep and goat, cattle, and pigs. We also hope to prepare an index of all volumes of the Bulletin for inclusion in Vol. 9.
ETHNOGRAPHIC EVIDENCE FOR WATERBUFFALO AND THE DISPOSAL OF ANIMAL BONES IN SOUTHERN IRAQ: ETHNOARCHAEOLOGY AT AL-HIBA Edward Ochsenschlager and Bonnie Gustav
As in previous volumes, bold type is used for Sumerian, italics for Akkadian words. Neo-Babylonian logograms are left in capitals. Bibliographical conventions generally follow the usages of the individual authors, but Harvard system references are used for books and articles, whereas more traditional Assyriological abbreviations are often preferred for text editions. These can be tracked down in one of the following works: R. Borger, Handbuch der Keilschriftliteratur, W. von Soden, Akkadisches Handworterbuch, and the Chicago Assyrian Dictionary. For Ur 111 references see M. Sigrist & T. Gorni, The comprehensive catalogue of published Ur 111 tablets (CDL Press, Bethesda, Maryland).
This volume was prepared on Macintosh computers and uses the font "Sippar" developed by Dr. Wilfred van Soldt of Leiden University, which, with his characteristic generosity, he has permitted us to adopt, ~ h a n E are s due to Jonathan Blanchard Smith of Cockroft Adams Ltd., for help with processing the texts, and to Trinity College, Cambridge for continuing material support in the production process. Nicholas Postgate Marvin Po well
July 1995
(Brooklyn CoUege, City University of New York)
WATERBUFFALO The Waterbuffalo of Iraq are river type buffalo (Bubalus bubalis). Some think they are most closely related to the Waterbuffalo of India, but a white patch which is rather common on the foreheads indicates some influence of the Nile breed. This latter influence, of course, could be quite recent. It was estimated in 1970 by an official from the agricultural department that cattle outnumbered waterbuffalo in Iraq by about 6 to 1. A well-cared for animal can live for about 18 to 20 years, and several have lived to be 25 years old or older.
Use The Mi'dan near al-Hiba keep waterbuffalo primarily for milk, dung and hides. Of the three, dung is the most important. One of the primary tasks for little girls, at a very early age, is to follow the waterbuffalo, collect their dung, and bring it back home carried atop their heads in baskets water-proofed with bitumen. It is mixed with straw or crushed reed, patted into thin disks by the women of the family, and then allowed to dry in the sun. These disks are used for fuel, when one desires to maintain a fairly even temperature over a period of time, for cooking, baking food or pottery and providing heat in cold, rainy weather. During summer dung patty fires provide an acrid smoke which keeps mosquito and fierce biting flies at bay for both owners and their animals. At the height of the insect season when the Mi'dan build a fire the whole herd jostles for the best position on the downward side in order to push their heads and faces into the smoke. In spite of their rugged appearance, waterbuffalo are very sensitive to large numbers of insects and an unhappy buffalo soon loses condition. The dung is also used for repairing leaks in a reed structure and for both waterproofing and sealing storage containers in the courtyard. Applied fresh, it dries to a cement-like hardness in the sun. It is applied to the forehead for headaches and used as a healing agent for burns. For cuts and wounds it is used to stop the flow of blood as well as for healing. Buffalo milk is served at room temperature, as well as heated and sweetened with sugar, as a drink, and is used in making gravies. In the heating process the heavy cream rises to the top and is scraped off and used on bread. The cream which rises to the top of unheated whole milk is made into butter. Butter is sometimes made in a hollowed-out gourd which is shaken from Bull. Sumerian Agric. 8 (1995)
1
Ochsenschlager and Gustav
Waterbuffalo and bone disposal at Al-Hiba
side to side and sometimes in a sheepskin. The apertures of a tanned sheep skin are sewn shut except for the neck. The cream is put in this leather bag, air blown into it (as into a balloon) and the neck tied tightly closed. Butter is produced by rolling the skin from side to side, shaking it briskly, and in the final stages kneading it with both hands. The buffalo are milked by both men and women, into wooden pails with tapered, knobbed or pointed, bottoms or, into sabbat, baskets coated with bitumen on the interior. Most people milk them from their flanks, but one man was seen milking from behind the cow. When questioned, he claimed it was easier to milk this way if the cow was as quiet and reliable as the old one he was milking. Younger ones were too nervous and could cause serious injury to the milker if they idcked. Calves are restricted in the access to their mother's milk at an early age. They stay tethered in the courtyard while the cow forages for grasses during the day, and wear muzzles made out of cord when the adults return from the marshes in the evening until the cow's owner has first drawn part of her milk for his own use. Meat from the waterbuffalo can always be found for sale in the market towns. This usually comes from male animals sold by the Mi'dan for cash. It is very unusual for a Mi'dan to slaughter a waterbuffalo himself although they will occasionally do so if an animal is injured past saving or if the family is very hungry during a period of food shortage. Male buffalo, of course, are as productive as females when it comes to the most important product of these animals - manure. No waterbuffalo in the area is ever used as a draught or pack animal. Even in other parts of Iraq where cattle are often used for draft, buffalo are kept for milk. A good cow will produce between 18 and 19 pounds of milk while a good buffalo will give as much as 25 pounds.
Feeding Water Buffalo exist almost entirely on grasses, sedges and young reeds from within or on the edge of the marsh. For most of the year they forage for themselves during the day while their owners collect fodder to feed them when they return from the marshes. During the rainy season, however, the marshes are sometimes too deep for the buffalo to reach the grasses and their owners and the owners' families must collect enough to feed them during the day as well. If the water is very high, the forage must be carried to the banks or house for the buffalo to eat. If it is only moderately high, the buffalo may accompany a member of their owner's family into the marsh. The harvester will cut the young growth beneath the water's surface, and the buffalo will eat it as it floats to the surface. This is probably also the way the animal is trained to crop forage from beneath the water's surface, as grazing beneath the water is not natural to the waterbuffalo. If the water is altogether too deep or if there are too few people in the household to cut forage for them, an owner may feed their herd barley and chopped wheat or barley straw for a brief time. Waterbuffalo are more stubborn than cattle about the introduction of a new food should a dietary supplement be needed. They will resist trying something new for days, often until their condition has deteriorated. Cattle, on the other hand tend to experiment with the new food after a few hours of hunger.
Housing Some owners of waterbuffalo provide a roofed structure or sitra to protect their livestock against inclement weather (see BSA 6, 1992, pp. 59-60). Occasionally an owner will tether his
Ochsenschlager and Gustav
Waterbuffalo and bone disposal at Al-Hiba
buffalo to stakes at night in the lee of his dwelling. Most, however keep all their livestock in a courtyard fenced with reeds or mud. During the chilly weather of late winter and early spring, the closer the buffalo are kept to the family's dwelling the warmer the interior of the house will remain. If the animals are not tethered to stakes but allowed to roam free in the courtyard, ditches, two to three feet in width, or holes of the same width are dug every two feet alongside interior walls of mud or reed to protect them against the lumbering strength of the waterbuffalo. Such protection is particularly needed in the spring when the animals seek anything solid against which they can scratch off their itchy winter coats. Husbandry In the villages around al-Hiba the largest herds of waterbuffalo numbered near 20. The ordinary owner kept closer to eight to ten. In the morning the cows are milked, calves allowed to nurse and then the adult buffalo are released from their evening's confinement to make their way across the mound to the marshes. The cattle seem always to follow the same paths from the villages emerging eventually into long lines on three main paths which the cattle from the surrounding villages seem to share. Most of the time the cattle set off without urging when released from their compound. If the water is still fairly deep or if a cow has a calf at home, the cattle may require some urging. Village boys or girls will urge them along with a stick until the animals are caught up in the mass movement and move off on their own. The breeding of the waterbuffalo seems to be very haphazard. No control is apparently exercised by the owners. The animals can mate indiscriminately in the marshes with any of the males from the surrounding villages. In most cases, an owner would be hard put to tell you who was the father of a newborn calf. Holes are often drilled through the end of the horns and pieces of yarn or cloth tied through them. These are talisman to prevent injury and bring luck but also function in making each of his herd recognizable to their owner, even at some distance. Cows carry their young for eleven to twelve months. When they are about to give birth it is not unusual for the owners to bring them inside their houses for the event, and some calves left behind when their mothers go to the marsh pastures during the day seem to have free use of the house interior wandering in and out at will. A cow sells for about 75 dinars and a bull for 45 to 50 dinars. Sick animals are usually treated with a hot shish and wounds are dressed with a concoction of reed ashes and salt. Waterbuffalo appear stolid, peaceful and content and most of the time they are so. They are capable, however, of building up great resentment against an owner if mistreated or against another waterbuffalo of either sex whether in or out of their own herd. Once they attack each other, the battle is almost impossible to stop until one of the animals is dead or dying. Such a confrontation between two cows broke out on the mound at al-Hiba one day and the owners were joined by sixty of our workmen who did everything possible to distract the cows' attention and herd them away from each other. All was to no avail. Lumbering beasts turned into skilled fighters which charged and maneuvered with deadly speed and grace, unfazed by blows from sticks and clubs, and apparently totally incapable in their frenzy of seeing anyone but their enemy.
Ochsenschlager and Gustav
Waterbuffalo and bone disposal at Al-Hiba Ochsenschlager and Gustav
Waterbuffalo can acquire a taste for bitumen. This can be very hard on bitumen covered boats which the Mid'an depend on for transportation. (see BSA 6,1992, p. 52). Although the Mi'dan and waterbuffalo live together in a symbiotic relationship one should probably not think of the waterbuffalo in this area as fully domesticated. Man offers some protection and at times food in return for what the animals produce naturally. There has been no attempt to "improve" them by controlled breeding programs or to train them as draft animals.
ANIMAL BONES Refuse pits are associated with each family's housing complex. A single pit serves the entire family and is located outside the courtyard and within 10 to 25 meters of the same. They consist of rather shallow holes about one meter in depth and from two to three meters in diameter. Household refuse is disposed of in these pits. When they are nearly full, they are covered over and a new pit dug close-by. Among the items consigned to these pits are the leftover remains from meals which usually consist only of some of the bones, cartilage and sometimes pieces of skin. Butchering of animals for meat is always carried out near the currently used pit and the remains from this process are cast into them as well. One would imagine that these pits would become quite odorous, especially in the summer, but that is not the case. Dogs, who have attached themselves to the family, stake out their independent territories around this pit and pick it clean of anything remotely digestible. Domestic birds such as chickens or turkeys, and domestic cats, which die of natural causes are also assigned to these pits, but not the larger domestic animals such as dogs, sheep or cattle. The latter are dragged off as far as possible so that the odor of their decay will not affect the quality of life. A large carcass such as this immediately becomes the focus of all the dogs in the neighborhood. They descend on it, fight over it, and if the carcass is particularly large, such as a waterbuffalo, establish temporary territories as close to it as possible which they maintain until the animal is fully eaten. Weaker dogs take advantage of an opportunity to race towards the carcass, grab some edible morsel and retreat with it to their permanent homes, the environs of their family's garbage pit. When the dead animal is reduced to bones, dogs will continue to frequent the area selecting an appropriate bone to chew on and often carrying it off to their permanent territory. The same is sometimes true of the remains of the dog's regular quarry, the wowie (jackal) or fox, if it is successfully hunted by one or two dogs. If a whole pack is involved in the chase, however, the entire animal disappears in an eating frenzy and the ground where it is captured is spattered with blood, not always just that of the captured animal. As a result of all this, the refuse pit and its environs take on a most interesting pattern or configuration. Our study of two of these pits shows that by the time the pit is buried, it contains less than 10% of the surviving fragments of bones that were deposited there by humans. About 20% more of the surviving fragments can be found within a radius of 10 meters. The remaining 70% have either been entirely consumed or are scattered all over the village and its surroundings. Family dogs carry them off to shady places for leisurely gnawing, and dogs belonging elsewhere make successful forays and raids. Furthermore, there are bones represented, sometimes in the pit itself, which are the results of scavenging and hunting by the dogs and are not the product of human activity. It is clear to us that we must take into account the possibility of similar bone dispersal in the study of the refuse pits at al-Hiba.
Waterbuffalo and bone disposal at Al-Hiba
AFTERWORD - 20 YEARS LATER Returning to the site in 1990 we found many changes. Among these one of the most striking was the complete disappearance of the Mi'dan, who no longer lived in the area. In 1970 the Beni Hassan had looked down with contempt on the Mi'dan for keeping waterbuffalo. They considered the waterbuffalo as at least partially unclean, for many would not knowingly eat of its flesh, and as somewhat useless when compared with the productivity of their domesticated cattle. Imagine our surprise then to find that despite the absence of the Mi'dan in 1990, the number of waterbuffalo in the area had only decreased by about 20 to 25% (about the same amount of decrease as had occurred in the size of the marsh land in the area). Furthermore fhe waterbuffalo were now being kept by the Beni Hassan who had so thoroughly disparaged them in the past. The reason for their change in attitude was made clear by some rather embarrassed owners who remembered elaborating twenty years before on the disgusting nature of the waterbuffalo and the Mi'dan who kept them. Waterbuffalo exploit an ecological niche in the marshes in which no other domestic animal can function. Clearly the economic benefits derived with but moderate effort (the marshes rarely flood to the depth where families must forage for their waterbuffalo), is enough to justify a change in attitude. Waterbuffalo meat is eaten more frequently in the area than it used to be, but it is still against the law for butchers to sell it labeled as meat from domestic cattle. During our stay in 1990 two butchers in the area were closed down by the police for doing so, and one who was said to have done so knowingly was put in jail.
Ochsenschlager and Gustnv
m.
-- --
-
Waterbuffalo and bone disposal at Al-Hiba
PLATE I
Waterbuffalo and bone disposal at Al-Hiba
Ochsenschlager and Gustav
PLATE I1
.-.
1.Women heating waterbuffalo milk
1, Sitra, or c o d for waterbuffdo, built with the owner's home in the center of the back wall.
2. Building a new sitra with the help of neighbors.
2. Woman using a gourd butter churn
Ochsenschlager and Gustav
Waterbuffalo and bone disposal at Al-Hiba
Waterbuffalo and bone disposal at Al-Hiba
Ochsenschlager and Gustav
PLATE 111
PLATE IV
1. A sheepskin butter churn
1. Dried dung patties stacked to shed water during the rainy season
2. Using dung patties to heat a tabag which serves as a kind of frying pan 2. Making dung patties
PLOUGH AND POWER: THE ECONOMIC AND SOCIAL SIGNIFICANCE OF CULTIVATION WITH THE OX-DRAWN ARD IN THE MEDITERRANEAN Paul Halstead (University of Shefield)
A rich and diverse body of archaeological evidence indicates that an 'ard' or 'scratch plough' drawn by a pair of cattle was known, and in some cases used, in many parts of Europe from at least the 4th-3rd millennium BC (Sherratt 1981; Chapman 1982). In the Mediterranean region, 3rd millennium BC models of a ploughing scene from Vounous in Cyprus and of yoked cattle ~ the ~ ard and of from Nemea in Greece (Pullen 1992) provide early evidence for m w l e d of animal traction; apparent osteological evidence for improved survivorship of male cattle at 3rd millennium BC Pevkakia in Greece may indicate the raising of oxen (i-e. castrated male cattle) for traction (Halstead 1987a, 81 Table 2), while Linear B texts from the late 2nd millennium BC palaces of southern Greece unambiguously record the existence of pairs of working oxen (Ventris and Chadwick 1973,212; Palairna 1989,91). In addition to its economic significance, it has been suggested that the introduction of the ox-drawn ard played a major causal role in the development of stratified society in bronze age Europe. Two principal arguments have been advanced. Gilman (1981) has suggested that ploughing with oxen necessitates fixed fields, cleared of obstacles such as boulders and tree stumps, and thus represents a major capital intensification of arable farming. This in turn enforces more permanent settlement and so prevents farmers from, literally, escaping the clutches of emerging elites. Gilman's case rests partly on the assumption, now almost universally rejected (e.g. Rowley-Conwy 1981; Behre and Jacomet 1991; Jones 1992), that early neolithic farming in Europe was characterised by short-term, swidden cultivation of temporary clearances and consequently by impermanent settlement. Sherratt (1981) draws on Goody's survey of ethnographic literature (Goody 1976), which demonstrates a clear cross-cultural association between plough-based agriculture and stratified societies, on the one hand, and between hoe cultivation and more egalitarian societies, on the other hand. Goody draws attention to two possible causes which may underlie this association. First, compared to manual cultivation, the plough raises output and so makes it possible to support non-producers (e.g. a ruling elite plus retainers). Secondly, the plough involves more extensive land-use and so increases the likelihood that land will become a scarce resource, distributed unevenly between rich and poor. Both propositions are inherently plausible and, in each case, Goody cites sparse supporting evidence of diverse geographical origin. On the other hand, area yields (i.e. yields per unit area) of ard agriculture may be lower than those of manual cultivation (e-g. Gallant 1991, 51; Sallares 1991,477; Bogucki 1993,498), though low area yields are of course off-set by the capacity of oxen to cultivate a large area. Moreover, in Bull. Sumerian Agric. 8 (1995)
11
Halstead
Plough and power
assessing the productivity of ard agriculture in terms of yields per unit labour, the cost of maintaining work oxen must also be taken into account (Bayliss-Smith 1982,107-8). In an attempt to clarify the social significance of the ox-drawn ard in Mediterranean prehistory and early history, this paper reviews a variety of evidence for the costs and benefits of ard agriculture and manual cultivation. Because both costs and benefits are likely to differ greatly between contrasting environments, the evidence reviewed is mostly drawn from the circum-Mediterranean region.
The costs of maintaining work oxen Large work animals such as oxen have both capital and running costs. The capital costs of raising or buying young oxen obviously vary, depending on the number of female cattle kept for breeding and dairy purposes (e.g. Bogucki 1993). In 19th century Greece, farmers paid the equivalent of the purchase price of 1.2 ha of land or 40-45 sheep for a pair of oxen (Psikhogios 1987, 48), while some early 20th century farmers who bred their own oxen from the family cow(s) waited several years to raise a full ploughing team (Halstead field notes [Assiros, Central Macedonia]). For many farmers, therefore, neither buying nor raising oxen was easy. Large animals also need significant areas of pasture, the availability of which inevitably varies in accordance with differences in climate and soil, rural population density, land tenure and land use. The animals may need to be taken to pasture throughout the year (e.g. Aschenbrenner 1972, 52), though such day to day herding may be undertaken by children without other major work commitments (e.g. du Boulay 1974, 246-7). In addition, work animals must be fed supplementary, harvested fodder, especially during the working season when their energy expenditure is high and time available for grazing is reduced. One study in Greece (Halstead and Jones 1989) has documented some of the labour costs entailed in providing harvested fodder for work animals. If crops are grown specifically for fodder, additional human labour is required throughout the arable farming cycle from ploughing and sowing through harvest to threshing and winnowing (fodder crops are usually threshed and winnowed, because farmers feed the grain and straw to different animals in different proportions at different times). Elderly farmers in northern Greece report sowing between 0.5 and 2 ha of bitter vetch specifically for the plough oxen, with the upper limit on the area sown being set not by the nutritional demands of the oxen but by the availability of human labour for harvesting this low and scrambling crop (Halstead field notes [Assiros, Central Macedonia]). Much of the fodder for large work animals may be derived from the processing byproducts of crops grown for human consumption. Such by-products are often treated by farmers as involving no cost (cf. Langdon 1986, 158-9), but they greatly increase the burden on human labour at harvest and threshing time. Cereal crops are cut low to recover much of the straw, making reaping both slower and more back-breaking and greatly increasing the volume of crop to be transported from field to threshing floor. The crop to be threshed now consists largely of straw rather than just of grain and chaff and, although much of the labour of threshing is effected by the trampling of work animals' hooves, this remains a laborious task and has to be undertaken at the hottest time of day when the midday sun makes the crop brittle. Finally, because the light straw and chaff are needed as fodder, the threshed crop can only be winnowed in a light breeze, thus significantly slowing down the cleaning of the crop (Halstead and Jones 1989).
Halstead
Plough and power
In the Kingdom of Naples, capital and running costs together made a day's work by a pair of oxen as expensive as that of ten labourers in the 18th century and of six labourers in the late 19th century (Delille 1977,147). Even if young male cattle, pasture and child labour for herding are freely available, and even if additional grain crops are not sown, the use of food crop residues as fodder enforces a substantial increase in agricultural work during the harvest and threshing period of maximum labour stress. In making this substantial investment of labour, the farmer is effectively 'storing labour' (cf. Rowley-Conwy and Zvelebil 1989, 49) by buying 'energy slaves' (Bayliss-Smith 1982,52-53) capable of considerable amounts of work during the critical season of tillage and sowing in the autumn and winter. This point may be made clear by comparing the areas cultivable by hand and by ox-drawn ard.
The benefits of oxen: tillage by hand and by ox-drawn ard To ask how much land can be ploughed by a pair of oxen in a day is rather like asking 'How long is a piece of string?'. The answer depends, inter alia, on the size and condition of the oxen, the type of plough and plough share, the nature of the soil and terrain, the weather before and during ploughing, when and how well the field was last ploughed, the spacing between furrows, and the size, distance apart and distance from the farmer's base of individual fields. Nonetheless, many cultures measure land in terms of a rule-of-thumb figure for the area which can be ploughed in one day (e.g. Evans 1960, 40-41). The classical Roman unit of land measurement, the jugerum (cf. jugum = 'yoke') represents 0.25 ha; in 19th century Greece, the stremma ('turning' and hence area which oxen can plow in a day) varied locally between 0.09 and 0.16 ha (Psikhogios 1987, 24-25); and in various parts of the East Mediterranean, the Turkish doniim (also 'turning') covered a similar range. Recent estimates of the area which can be ploughed in a day by a pair of oxen range from 0.1 to 0.4 ha in 20th century Palestine and Jordan (Russell 1988, 123 Table 27; Palmer and Russell 1993). Estimates from 20th century Greece range from 0.1 to 0.3 ha, though the higher figures are sometimes linked to the use of two or three pairs of oxen in rotation with one plough (Aschenbrenner 1972, 57; du Boulay 1974, 242; Halstead and Jones 1989; Halstead field notes [Assiros, Central Macedonia]). The norm for ploughing in Mediterranean conditions under traditional technology, therefore, may be regarded as roughly 0.1-0.3 hatday. By contrast, estimates of the area which can be dug by hand with a variety of tools in different parts of the world range between 0.02 and 0.05 hdday (Russell 1988, 114 Table 19). In 19th century Greece an axinari, the area cultivable by hand in a day, was reckoned at ca. 0.05 ha (Psikhogios 1987, 34), while one estimate of the area cultivable manually by early 20th century market ,gardeners in northern Greece suggests a range of 0.02-0.05 hdday (Halstead field notes [Assiros, Central Macedonia]). In other words, tillage with an ox-drawn ard is something like 2-15 times faster than manual cultivation. In effect, the greater speed of the oxdrawn ard allows farmers to break one of the principal bottlenecks in the agricultural year - the autumn and winter period of tillage. The increased scale of cultivation thus made possible has radical implications for agricultural methods at subsequent stages in the crop growing cycle (Halstead 1987b). The implications of tillage by ox-drawn ard for subsequent stages of crop husbandry First, a major increase in the area sown requires more rapid methods of sowing and, as a consequence, ard-based agriculture tends to be associated with broadcast sowing (Sigaut 1975,
Halstead
Plough and power
220-21), while small-scale manual cultivation tends to involve more precise but more labourintensive methods such as dibbling or planting. The disadvantage of broadcast sowing is that it tends to use up twice as much seed corn as the more intensive methods (Halstead 1990). Secondly, small-scale manual cultivation is often associated with relatively intensive husbandry such as weeding by hand, whereas weeds tend to be controlled under large-scale ard-agriculture by regular, ploughed fallowing. Thirdly, small-scale cultivation tends to be characterised by annual cropping of land, often involving cereal-pulse rotation, while large-scale ard-agriculture tends to specialise in the less labour-intensive cereal crops. Thus two broad strategies may be distinguished (Halstead 1987b; 1992a): (1) Manual cultivation, with intensive husbandry methods, is characterised by relatively high area yields, but is small-scale and has high labour inputs, and so produces relatively little surplus over and above what is needed to feed the human workforce involved. (2) Ard cultivation, with extensive methods, is characterised by low area yields, but is large-scale and has low human labour inputs, and so has the potential to produce a large surplus. The greater capacity for surplus of ard agriculture is facilitated by the fact that a significant part of the fodder for the critical work animals is made up by the processing byproducts of human food crops.
The harvesting problem While more extensive methods of crop husbandry can be adopted to compensate for the larger scale of ard agriculture, and while work animals can contribute to the transporting and threshing of the crop after harvest, potential improvements in the speed of harvesting were limited before the relatively recent introduction of true scythes and then of mechanical reaping machines. A series of estimates by Greek farmers for reaping with sickles suggests normal cereal harvesting speeds of UD to 0.1 hdman day. Slower rates are reported for pulses and also for elderly workers, while some of the faster estimates exclude the labour of a second person (often the wife of the reaper) engaged in binding the crop into sheaves (Halstead and Jones 1989,47; Halstead field notes [Assiros, Central Macedonia]; Wagstaff and Augustson 1982, 131). Figures from a variety of recent European and Near Eastern contexts (Russell 1988, 116-17, Tables 20-21) suggest harvesting and binding speeds for cereals cut with a small- or mediumsized iron sickle of ca. 0.05-0.1 hdman day. This compares with rather higher speeds for ploughing with the ox-drawn ard of 0.1-0.3 hdday (above). Russell has also assembled figures for possible reaping speeds with more primitive technology. Reaping experiments with flint, obsidian, copper and bronze sickles suggest speeds of ca. 0.02-0.05 hatman day (Russell 1988, 116 Table 20), accentuating the contrast between ploughing and reaping rates. In sum, an ox-drawn ard can plough more land in a day than one man can reap, but some of the faster tillage rates may involve the labour of two persons (one to plough and one to sow) and of course farming families commonly mobilise their entire work force at harvest time. Where the farmer has a large family including several adult offspring, therefore, the area which can be ploughed and sown per day with the aid of an ox-drawn ard and the area which can be reaped per day by the family may be in balance. Where the farmer commands a small family labour force, however, the implicatibn is that the services of additional reapers must be secured or that the harvest season must be substantially longer than the ploughing season (perhaps involving underuse of the oxen).
Halstead
Plough and power
Up to a point, the length of the harvest season can be extended by growing a range of crops with different maturation times, by staggering the sowing of each crop type and by sowing crops in several fields differing in altitude, aspect and soil type (e.g. Forbes 1982). Once the crop approaches ripeness, however, it must be harvested with some urgency, before the grain is shed or robbed by birds or rodents (Halstead and Jones 1989). In Greece, before mechanisation, the barley and wheat harvest might typically be spread over up to one month, and this period might be extended to up to two months if the crops to be harvested included the early-ripening pulses and the late-ripening oat, rye and millet. The pulses are usually sown on a relatively small scale on large farms, however, and oat and rye are not yet attested as crops in prehistoric Greece and millet only rarely so (Halstead 1992a, 108 Table 1; 1994,204-5 Table 7.1). By contrast, the winter ploughing season in Greece might typically extend over one to three or even four months between October and February, depending on how early the autumn rains began in a given year and how much time was lost because the ground was too wet, heavy or frozen for cultivation (Aschenbrenner 1972,50-51; du Boulay 1974,242; Kostis 1993,41; Halstead and Jones 1989; Halstead field notes [Assiros, Central Macedonia]). Consideration of the relative lengths of the harvesting and ploughing seasons, therefore, again accentuates the contradiction between the ploughing potential of a pair of oxen and the harvesting constraints imposed by a family labour force. More meaningful than a comparison between daily ploughing and reaping rates is a comparison of the area which can be ploughed and harvested per year, though estimates of these figures are of course greatly complicated by variation in family size and structure and in the length of the ploughing and harvesting seasons. Nonetheless, some regularities again emerge.
The scale of tillage and harvesting possible on a family farm For 20th century northern Greece, recent estimates by several elderly farmers suggest that the maximum area cultivable per family per year by hand was ca 3 ha and the maximum with oxen ca 10 ha, of which about half would be in winter cereals and half fallow or in summer crops (Halstead field notes [Assiros, Central Macedonia]). Figures from Byzantine times are broadly comparable: average allocations per ox pair of 8 ha and 15 ha (including fallow or summer crops) are recorded (Harvey 1989,50-51,140), while farmers without oxen were allocated one quarter the area of land assigned to those with a pair of oxen (Harvey 1989,52). An average allocation of 10 ha per ploughman is also recorded for a 17th century estate in Macedonia (Moskof 1979,60-62). For southern Italy, two parliamentary enquiries into late 19th century and early 20th century agriculture offer further comparative data (Delille 1977, 123). In the rich plains of Campania, where a pasture-maize-wheat-oat rotation was practised, a farmer with oxen might take on 10 ha (presumably 5 ha in winter cereals), while a tenant farmer without oxen might take on a maximum of 2-3 ha (Delille 1977, 128). In the less fertile Appenine hills and in large parts of Apulia, under more extensive conditions of husbandry, the areas which might be farmed are ca 10 ha with oxen and up to 4-5 ha with manual cultivation (Delille 1977, 129). In the area of Crotone, a family without oxen could not rent more than 4 ha (Delille 1977, 136 n 89). Finally, in the Auvergne region of France, where oxen were used for ploughing until ca. 1950, farmers were able to cultivate up to 5-6 ha of winter cereals with one pair of oxen (Halstead field notes wireflews village]).
,
-
Halstead
Plough and power
Again comparison with the area which can be harvested by a family labour force is illuminating. In theory, at a rate of 0.1 hdday, a recent Greek reaper armed with a sickle could cut up to 3 ha of wheat and barley during a harvest season of one month; a family of two active parents and two grown sons could cut up to 12 ha. As noted above, however, the maximum figure of 0.1 hdday was probably only attainable by the most fit, industrious and skilled reapers, and perhaps ignores the labour involved in binding the crop. Farmers suggest that 1-2 ha of winter cereals per reaper is a more realistic figure for a harvest season (Halstead field notes [Assiros, Central Macedonia]). Moreover, the family will have to divide its labour force between harvesting, transport of the crop to the threshing floor, and normal maintenance activity with livestock, while time may also be lost if the crop is wet by summer rain. In practice, estimates by recent Greek farmers suggest that the maximum area of winter cereals which can be harvested by a reasonably active family is about 4 ha, perhaps 6 ha for a large family with adult offspring (Halstead field notes [Assiros, Central Macedonia] ; also du Boulay 1974, 242). The maximum harvesting capacity of a family labour force, therefore, at a favourable point in the domestic cycle, barely matches the normal cultivation potential of a pair of oxen. Further light may be shed on these figures by attempting an even more problematic estimate of the minimum areas in cultivation necessary to support a family and to justify the keeping of a pair of oxen.
The minimum size of a viable family farm Delille (1977, 132) estimates that, in southern Italy, 4-5 ha in a two-year cereal-fallow rotation would suffice to support a family of 4-5 persons (including some allowance for payment of taxes and other expenses, but not for the consequences of a bad harvest): the Mediterranean smallholder working the land by hand can only cultivate enough for bare survival (Delille 1977, 135). On the other hand, it may be inappropriate (see above) to extrapolate the extensive husbandry methods and low area yields of large scale cereal agriculture to a smallholding cultivated by hand. With intensive hoeing, weeding and perhaps watering and manuring, and with cereal-pulse rotation, a small holder might well produce average yields approaching 1 tonlha. In a prehistoric context, therefore, a farmer unencumbered by the need to pay taxes in money, might well be able to feed a family of 4-5 persons (needing, say, 1.5 tons of graidyear) from as little as 2-3 ha of intensively cultivated land. Hopkins (cited in White 1970, 336) similarly calculated the minimum area to support a Roman farming family at 2 ha without oxen. A family subsistence requirement of 2-3 ha of winter cereals is reported from 19th century and 20th century Greece (Psikhogios 1987, 39 n 13; Halstead field notes [Assiros, Central Macedonia]). In other words, the area needed to support the family of a 'subsistence farmer' (ca 2-3 ha) is comparable with the areas which a family labour force can cultivate by hand (ca 2-4 ha) and harvest (up to 4 ha). Conversely, a farmer making full use of the cultivation potential of a pair of oxen (5+ ha of winter cereals) is likely to be supporting a very large family or producing a substantial surplus. Finally, is it possible to estimate the area in cultivation necessary to justify the keeping of oxen, given the substantial if highly variable costs involved? Recent farmers in a part of northern Greece traditionally dominated by cereals, and whose own experience dates to the period in the earlier 20th century when large estates were dissolved and the land redistributed to large numbers of smallholders, suggested that ownership of a pair of oxen was only economic on holdings exceeding 3-4 ha (Halstead fieldnotes); below this level, farmers would variously
Halstead
Plough and power
share or rent oxen, use cows or equids or cultivate by hand (see below). The minimum holdings of the classical Athenian hoplite class, the zeugitai (i.e. ox-farmers), have similarly been estimated at ca 4-5 ha (Isager and Skydsgaard 1992, 79; also Foxhall 1992; Jarneson 1992; Osborne 1992, 24), while Hopkins (cited in White 1970, 336) has suggested a minimum requirement of 5 ha for a Roman farmer with oxen. Delille (1977, 135) emphasises the inability of the smallholder, in the Mediterranean as a whole, to raise large livestock. As a result, ownership of oxen has often been highly restricted. In early 19th century Greece, only one third of farmers possessed oxen and in 1860 there was an average of only one ox per farmer; many farmers cultivated by hand (Psikhogios 1987,47). Conversely, much of the cereal land in 19th century Greece was in the hands of large, oxpowered estates (Vergopoulos 1975) or 'chifliks' - from the Turkish ~ i for t 'pair' [of oxen] (cf. Asdrakhas 1978,41-43). A similar contrast between numerous small-holdings, too small to justify the keeping of oxen, and a few large estates with oxen is apparent for Byzantine Greece (Harvey 1989,63-64,71). Likewise, while many 19th century Italian farmers cultivated small holdings by hand, in Calabria in 1811 70% of the 13,750 oxen were in herds of 100-1200 head (Delille 1977, 136), that is attached to very substantial agricultural estates.
The economic implications of keeping plough oxen The implication of these various calculations is that, as regards the growing of grain crops, the 'traditional' Mediterranean farmer has been trapped between the minimum scale needed to justify the keeping of plough oxen and the maximum scale of harvest manageable with the labour force of a normal family. In other words, for all but the largest families at the most favourable point in the domestic cycle, farmers were faced with the choice of growing grain crops by hand on a scale sufficient for subsistence, or of investing in a pair of oxen, hiring outside labour at harvest time and producing a substantial surplus of grain for the market. In practice, of course, a number of alternative strategies are possible between these two extremes. First, the preceding discussion has avoided defining a 'typical' farming family. A nuclear family household may only temporarily muster the harvesting labour needed to match the ploughing potential of a good pair of oxen (while offspring are grown up but still attached to the parental household), but more extended 'stem' or 'compound' families, consisting of two or three generations under the same roof, were not uncommon in 19th-20th century Greece (Psikhogios 1987, 97-1 12). Such households can muster a relatively large and continuous labour force, but may also include enough consumers to neutralise the surplus producing potential of a pair of oxen. Secondly, in recent decades, Greek farmers with too little land to maintain oxen have usually ploughed with a pair of cows, a cow and a donkey, or a mule rather than resorting to manual cultivation. In this case, the capital and running costs of oxen are reduced by yoking a dual purpose animal, also used for milking or transport. Elderly farmers are in almost unanimous agreement, however, that the use of non-specialist plough animals greatly limits both the quantity and quality of ploughing, while also jeopardising the productivity of the cow in milk and calves, and making the donkey or mule unavailable for other work. Some smallholders have instead shared or hired oxen, but both solutions again limit the possible scale of cultivation and also mean that the farmer has only partial control over when he ploughs and sows and so incurs greater risk of crop failure or low yields (Halstead field notes [Assiros, Central Macedonia; Naxos]). As a result, grain production in much of the Mediterranean in
Halstead
Plough and power
early modern times has tended to be undertaken either by smallholders producing, at best, enough for subsistence or by large estates producing on a large scale, and with minimal input of human labour, for market (Delille 1977; Psikhoyios 1987; Vergopoulos 1975).
The social implications of keeping plough oxen The preceding discussion has documented a close relationship between the use of plough oxen and large-scale cereal agriculture with extensive methods of husbandry. It is now possible to reevaluate, in a Mediterranean context, Goody's suggestions as to the causal relationship between the plough and social stratification. As Goody noted, the ox-drawn plough permits large-scale surplus production and so enables the support of non-producers, such as an elite and its retainers; extensive plough agriculture is also 'wasteful7of land (compared to intensive manual cultivation) and so promotes land shortage, which in turn makes possible uneven access to land. In addition, at least under the scheduling constraints of Mediterranean climate, ox-based agriculture geared to surplus production of grain requires the existence of an external force of landless or otherwise dependent labourers at harvest time. The use by plough owners of fallow as a substitute for labour-intensive, manual control of weeds also presupposes preferential access to scarce land, while the costs of raising and maintaining oxen, in most parts of the Mediterranean, are prohibitively high for most farmers engaged in 'subsistence' cultivation. Social stratification thus emerges,pace Goody, as a precondition as well as consequence of oxbased agriculture in the Mediterranean (also Gudeman 1977). Finally, while both the anthropological and archaeological literature tends to be concerned with the relationship between the plough (or ard) and social complexity (Goody 1976; Sherratt 1981), the foregoing discussion suggests that the critical variable is ownership of specialised plough oxen and not mere use of the ard (also Braudel1975,77; Spurr 1986,171). Conclusion In the context of European or Near Eastern prehistory, therefore, it would not be appropriate to assign great social significance to archaeological evidence for the existence of the ard or yoked pairs of cattle or even for the practice of ploughing (see Rowley-Conwy 1987). On the other hand, references in the Linear B records of the late bronze age palaces of southern Greece to pairs of working oxen are more significant. First, working oxen were concentrated near major centres and sub-centres, as were recorded palace land holdings and cereal harvests, suggesting a connection between oxen and palatial agriculture. Similarly, some palatial land was allocated in units of ca 5 ha, that is of a size suitable for cultivation with a pair of oxen. Secondly, palatial records of grain harvest are restricted to one type of cereal, suggesting extensive agriculture, in sharp contrast with the archaeobotanical record, which suggests a balanced mixture of various cereals and pulses more characteristic of small-scale, intensive husbandry. Textual and bio-archaeological evidence in combination, therefore, suggest that, while much of the population practised intensive horticulture, the palaces (like the land-owning elites of the early modern Mediterranean) were in large measure supported by the surplus grain produced on large, extensively farmed estates, cultivated by plough oxen (Halstead 1992a; 1992b). While prehistorians tend to emphasise the role of economic intensification in supporting early states, it is tempting to speculate, on theoretical grounds, that extensive ard-agriculture played a major role in financing the surplus necessary to support many Near Eastern
Plough and power
Halstead
civilisations. In 3rd millennium BC Mesopotamia, the contrast between texts overwhelmingly concerned with one cereal and a diverse archaeobotanical record (Halstead 1990) again hints at specialised administrative involvement in extensive agriculture. In Sumerian literature, the association of the plough with gods and kings likewise suggests an elite context for extensive agriculture (Vanstiphout 1984,242-3; Hruska 1992,73), while the disputation poem between the hoe and the plough concludes that, for Sumerian society as a whole, the hoe was more important (Vanstiphout 1984,247).
Acknowledgements Attendance at the Wassenaar meeting of the Sumerian Agriculture Group was assisted by a grant from Sheffield University Foreign Travel Fund. The British Academy funded fieldwork in Greek Macedonia in 1993, during which interviews were conducted with elderly farmers who had worked on Turkish chiflik estates in the first two decades of this century. I am grateful to Yannis Hamilakis, Nicholas Postgate and Tom Williamson for useful bibliographic suggestions, and to Carol Palmer for drawing my attention to the meaning of the Turkish doniim.
I
BIBLIOGRAPHY Aschenbrenner, S. 1972 "A contemporary community", in W.A. McDonald and G.R. Rapp (eds.) The Minnesota Messenia Expedition (Mmneapolis: University of Minnesota Press), 47-63. Asdrakhas, S. 1978 Mikhanismoi tis Agrotikis Oikonomias stin Tourkokratia (Athens: Themelio). Bayliss-Smith, T. 1982 The Ecology of Agricultural Systems (Cambridge University Press). Behre, K.-E. and Jacomet, S. 1991 "The ecological interpretation of archaeobotanical data", in W. van Zeist, K. Wasylikowa and K.-E. Behre (eds.) Progress in Old World Palaeoethnobotany (Rotterdam: Balkema), 81-108. Bogucki, P. 1993 "Animal traction and household economies in neolithic Europe", Antiquity 67,492-503. du Boulay, J. 1974 Portrait of a Greek Mountain Village (Oxford University Press). Braudel, F. 1975 The Mediterranean and the Mediterranean World in the Age of Philip 11, volume 1 (London: Fontana). Chapman, J. 1982 "The secondary products revolution and the limitations of the Neolithic", Institute of Archaeology Bulletin 19, 107-22. Delille, G. 1977 Agricoltura e Demografm nel Regno di Napoli nei Secoli 18 e 19 (Naples: Guida Editori).
-
Halstead
Plough and power
Evans, G.E. 1960 The Horse in the Furrow (London: Faber and Faber). Forbes, H. 1982 Strategies and Soils: Technology, Production and Environment in the Peninsula of Methana, Greece (Ph.D. dissertation, University of Pennsylvania).
Halstead
Plough and power
Jones, G. 1992 "Weed phytosociology and crop husbandry: identifying a contrast between ancient and modern practice", Review of Palaeobotany and Palynology 73: 133-43. Kostis, K. 1993 Aforiu, Akriviu kai Pina (Athens: Ekdosis Alexandrias).
Foxhall, L. 1992 "The control of the Attic landscape", in B. Wells (ed.) Agriculture in Ancient Greece (Stockholm: Swedish Eastitute at Athens), 155-59.
Langdon, J. 1986 Horses, Oxen and Technological Innovation: the Use of Draught Animals in English Farming from 1066 to 1500 (Cambridge University Press).
Gallant, T.W 1991 Risk and Survival in Ancient Greece (Cambridge: Polity Press).
Moskof, K. 1979 Istoria tou Kinimatos tis Ergatikis Taxis (Thessaloniki).
Gilman, A. 1981 "The development of social stratification in bronze age Europe", Current Anthropology 22, 1-8.
Osborne, R. 1992 "'Is it a farm?' The definition of agricultural sites and settlements in ancient Greece", in B. Wells (ed.) Agriculture in Ancient Greece (Stockholm: Swedish Institute at Athens), 21-25.
Goody, J. 1976 Production and Reproduction (Cambridge University Press). Gudeman, S. 1977 "Morgan in Africa", Reviews in Anthropology, 575-80. Halstead, P. 1987a "Man and other animals in later Greek prehistory", Annual of British School at Athens 82, 71-83. 1987b "Traditional and ancient rural economy in Mediterranean Europe", Journal of Hellenic Studies 107, 77-87. 1990
"Quantifying Sumerian agriculture - some seeds of doubt and hope", Bulletin on Sumerian Agriculture 5, 187-95.
199%
"Agriculture in the bronze age Aegean", in B. Wells (ed.) Agriculture in Ancient Greece (Stockholm: Swedish Institute at Athens), 105-16.
1992b "The Mycenaean palatial economy: making the most of the gaps in the evidence", Proceedings of Cambridge Philological Society 38,5746. 1994
"The North-South divide: regional paths to complexity in prehistoric Greece", in C. Mathers and S. Stoddart (eds.) Development and Decline in the Mediterranean Bronze Age (Sheffield: J.R. Collis), 195-219.
Halstead, P. and Jones, G. 1989 "Agrarian ecology in the Greek islands", Journal of Hellenic Studies 109,41-55. Harvey, A. 1989 Economic expansion in the Byzantine empire, 900-1200 (Cambridge University Press). HruIka, B. 1992 "Hacke oder Pflug? Zur Bodenvorbereitung in den sumerischen Literaturtexten", in J. Prosecky (ed.) Ex Pede Pontis (Prague: Oriental Institute, Czechoslovak Academy of Sciences), 70-79. Isager, S. and Skydsgaard, J.E. 1992 Ancient Greek Agriculture (London: Routledge). Jameson, M. 1992 "Agricultural labor in ancient Greece", in B. Wells (ed.) Agriculture in Ancient Greece (Stockholm: Swedish Institute at Athens), 135-46.
Palaima T.G. 1989 "Perspectives on the Pylos oxen tablets", Studia Mycenaea 1988 (Skopje), 85-124. Palmer, C. and Russell, K. 1993 "Traditional ards of Jordan", Annual of the Department of Antiquities of Jordan 37, 37-53. Psikhogios, D.K. 1987 Proikes, Foroi, Stafida kai Psomi: Oikonomia kai Oikoyenia stin Agrotiki Ellada tou 19 Aiona (Athens: Ethniko Kentro Koinonikon Erevnon). Pullen. D.J. 1992 "Ox and Plough in the Early Bronze Age Aegean", American Journal of Archaeology 96,4554. Rowley-Conwy, P. 1981 "Slash and burn in the temperate European Neolithic", in R.J. Mercer (ed.) Fanning Practice in British Prehistory (Edinburgh University Press), 85-96. 1987
"The interpretation of ard marks", Antiquity 61, 263-6.
Rowley-Conwy, P. and Zvelebil, M. 1989 "'Saving it for latet", in P. Halstead and J. 0' Shea (eds.) Bad Year Economics (Cambridge University Press), 40-56. Russell, K.W. 1988 After Eden: the Behavioral Ecology of Early Food Production in the Near East and North Africa (BAR Int. Series 391; Oxford: British Archaeological Reports). Sallares, R. 1991 The Ecology of the Ancient Greek World (London: Duckworth). Sherratt, A.G. 1981 "Plough and pastoralism", in I. Hodder, G. Isaac and N. Hammond (eds.) Pattern of the Past: Studies in Honour of David Clarke (Cambridge University Press), 261-305. Sigaut, F. 1975 LJAgricultureet le Feu (Cahiers des ~ t u d e Rurales s 1; Paris). Spurr, M.S. 1986 "Agriculture and the Georgics", Greece and Rome 33, 164-87.
Halstead
Plough and power
Vanstiphout, H. 1984 "On the Sumerian disputation between the hoe and the plough, Aula Orientalis 2, 239-51. Ventris, M. and Chadwick, 3. reek;) ~ u(Cambridge ~ University Press). 1973 Documents in M ~ c ~ M Vergopoulos, K. 1975 To Agrotiko Zitima stin Ellada (Athens: Exandas). Wagstaff, M. and Augustson, S. 1982 "Traditional land use", in C. Renfrew and M. Wagstaff (eds.) An Island Polity (Cambridge University Press), 106-33.
HOW CAN MODERN FOOD TECHNOLOGY HELP TO IDENTIFY DAIRY PRODUCTS MENTIONED IN ANCIENT SUMERIAN TEXTS ? Michael Teuber (Laboratory of Food Microbiology, Institute of Food Science, Swks Federal Institute of Technology)
White, K.D. 1970 Roman Farming (London: Thames and Hudson).
In principle two routes are available to us: 1. We can analyse and study products made by nomads in that area and compare the properties of these products with the ones mentioned in the ancient texts (Anderegg, 1894; Mair-Waldburg, 1974). 2. We can try to prepare such products in the laboratory to find out whether the necessary conditions coincide with conditions presumably prevailing in Mesopotamia 3000 to 5000 years ago. The reason why we can be quite confident about this approach is the fact that the biochemical composition of milk (of different species) has not significantly changed since that time. In addition, the scientific progress during the last 100 years in dairy science has yielded so much experience and knowledge that we are able to predict (even retrospectively) what happens to milk and its components when specific actions are taken, e.g. spontaneous fermentation with lactic acid bacteria, heating or boiling, churning, draining of whey (milk serum), ripening or drying of curd (Teuber, 1993; Teuber et al. 1987). First I shall give a brief explanation of the properties of milk. The composition depends on the animal species and is given in Table 1. The main chemical fractions determining the biochemical and nutritional properties are accordingly:
1 Water (about 85 to 90%)
2 Protein (3 to 3.5%) The protein fraction is composed of 2 completely different kinds of molecules: 2.1 The caseins They are organized in micelles together with minerals (mainly calcium phosphate and potassium), and citric acid. The casein micelles (50 to 300 pm in size, about loi4 per ml) are in colloidal solution and stable at boiling temperatures. Caseins, however, are precipitated (coagulated and denaturated)
Bull. Sumerian Agric. 8 (1995)
Identification of dairy products in Sumerian texts
Identification of dairy products in Sumerian texts
Teuber
under acidic conditions (lactic acid fermentation or addition of acetic acid) or by the action of specific enzymes like those present in calf rennet or lamb stomachs.
Table 1 Composition of milk from different animal species (average values, undefined lactation phases)
2.2 The whey proteins They are globular proteins which are soluble under acidic conditions but denaturated (precipitated) at the surface of the casein micelles by boiling. For that reason, the coagulum obtained from boiled milk by lactic fermentation is much finer than the one obtained from non heated raw or butter milk. In addition, denatured whey proteins act as a kind of molecular glue at the surface of casein micelles to yield a very dense, almost glass like structure of the curd after drying. This is important for the structure and stability of dried curd (cheese). Curd can be separated from whey (milk serum) by draining (e.g. through cloth, sieves or riddles) to yield fresh cheese.
animal species
dry matter
%
%
beef goat sheep buffalo zebu
12,7 13.2 19,3 17,2 13,5 17,3 29,9 15,O 11,2 11,7 18,8 233 17.6 31,O
yak reindeer camel horse donkey
3 Fat Milk fat is organized in small fat droplets surrounded by a fat globule membrane composed of lecithin and proteins which render the fat globules water "soluble". When the fat globule membrane is destroyed by mechanical stress (e.g. by churning) the fat globules melt together into large lumps of butter which still contain about 15% of water. Boiling of butter separates the remaining water from the fat, yielding butter oil (which is called ghee in India). The boiled water-free butter oil is very stable and can be kept for many months, and even years.
4 Milk sugar (lactose, about 5%) Milk sugar is the component which is transformed into lactic acid by lactic acid bacteria. The concentration of lactic acid obtained in sour milk, sour cream or yoghurt (between 1 and 2%) is high enough to prevent the growth of pathogenic bacteria and of spoilage microbes. However, sour milk is still an excellent substrate for yeasts (Kefir!) and moulds (Roquefort and Camembert cheese). Yeasts and moulds can only be kept out of dairy products if the water content is reduced as in butter oil or dried cheese. Dried fresh cheese is stable for decades! The stability of cheese can also be increased by addition of high amounts of salt (sodium chloride). All classical and ancient techniques for the manufacture of dairy products use these few basic manipulations, however, in different combinations (see also Ryder 1993). The most important factor for the ancient dairy farmer was the desire and necessity to obtain a product which was stable enough to reach the market (consumer) and to have a preserve for the season of the year with low milk production. We can devise a technologydependent ladder of increasing stability against spoilage (at room temperature) as shown in Table 2. The only products which keep for much longer than about 2 weeks are hard cheeses, dried curd cheeses and butter oil. Whenever the ancient texts from Mesopotamia mention dairy products with presumable keeping times of more than two weeks we have to look for them and their identity within the short list given in that table. A strongly salted curd (at least 15% NaCl) could keep also for a few weeks. It seem questionable, however, whether nomads would be able to carry the necessary large amount of dry salt with them.
Teuber
pig dog cat dolphin
water
fat
casein
whey protein
lactose
%
%
%
%
%
87,3 86,8 80,7 82,3 86,5 82,7 70,l 85,O 88,8 88,3 81,2 76,5 82,4 69,O
3,7 4s 7,4 7,4 4,7 6,5 14,5 5,4
23 2,5 4,6 32 2,6
0,6 0,4 0,9 0,6 0,6 5,8* 1,7 1 ,o
43 4,l 4,8 4,8 4,9 4,6 32 5,1 62 7,4 5,5 3-1 4.8 0,6
19 1,4 6,8 12,9 43 18,O
8,l 2.9 1,3 1 ,o 2,8 53 3,7
-
12 1 ,o 2,o 2,l 3,3 9,4*
ashes
* : casein + whey protein
Table 2 Keeping times of dairy products
I
I
keeping time
1 milk and its protein fraction I
112 day
raw milk
1 day
boiled milk
1 week
!
sour milk
2 weeks
sour curdJfresh cheese
2-3weeks
soft cheese
1-2years
hard cheese
several years
dried cheese
fat fraction sweet cream
!
sour cream butter
butter oil
I
Identification of dairy products in Sumerian texts
Teuber
General procedures for the preparations of nomadic cheeses Very useful details on indigenous cheeses of the Middle East, their properties and production technologies can be be found in three recent publications by P.F. Fox (1993), Hamid A. Dirar (1993), and Abou-Donia (1978). Recent accounts of nomads still living in the Middle East are given by Baum (1989) and Janzen (1980). For the preparation of ripened soft and especially hard and semi-hard cheeses, we need precise timeltemperature regimes and management which probably were not available in ancient Mesopotamia due to climatic conditions. I believe that we can therefore strike such products with a complex production technology from our main considerations. However, fresh cheese curds stored in skin bags from goats and sheep may have ripened to Roquefort-like soft cheeses, e.g. in Anatolia, or in the Zagros mountains (i.e. under temperate climatic conditions with lower temperatures and air humidity than in the lower Euphrates and Tigris valleys). The substrates (raw materials) for nomadic cheeses can be either whole milk, skimmed or partially skimmed milk, whey or buttermilk. The milk can be mixed milk if different species of animals compose the flocks and herds.The following procedures are documented (see Fox, 1993, and Dirar, 1993): A.
1st step: 2nd step: 3rd step:
B.
1st step: 2nd step:
3rd step: 4th step: 5th step: C.
1st step: 2nd step: 3rd step:
Coagulation of milk with the aid of rennet (from young animal stomachs) Draining of whey through cloth with concomitant spontaneous lactic fermentation during coagulation and drainage. Resulting fresh cheese consumed directly or dried (with or without addition of salt and spices) Condensation of whole milk or butter milk by boiling. Coagulation of boiled, condensed whole milk or butter milk by thermophilic lactic fermentation (e.g. by using yoghurt culture), yielding sour rnilk or yoghurt. Removal of butter by churning if whole milk was used as raw material Draining of whey e.g. with cloth or through leaf mats Drying of formed cheeses Coagulation of whole boiled andlor condensed milk by fermentation as above under B Removal of whey by decantation or evaporation Addition of new milk until vessel is filled completely.
Hypothesis that the sign GA'AR in the ancient texts from Uruk I Warka stands for dried cheese This characteristic sign has a relation of the long side to the broad side of 2:l. The dot in the middle has been proposed by Nissen (as cited in an elaborate treatise by M. Stol, 1993) to
Identification of dairy products in Sumerian texts
'
Teuber
symbolize a string which has been used to hold small pieces of cheese. For reasons of mechanical stability these have to be dried cheeses. In the collection of our Institute made by the late Prof. Marc Bachmann we have a chain of small dried cheeses assembled on a string of 1 m length. The 18 rectangular small cheeses have a mean size of 15 x 18 x 35 rnm and a weight of about 5 g each. Under good manufacturing conditions, this amount of cheese represents the protein content of between 1 and 1.5 liters of cow's milk. This cheese-chain was given to Mr. Bachmann by nomads in Afghanistan about 25 years ago. The exact mode of manufacture is not known, in particular we do not know how and when during the production process the string was put through the cheeses. However, the cheese has most probably been produced from mixed milk (sheep, goats, cows and maybe even camels). The general appearance of the cheeses is close to Krut (or Qurilt) which is made by boiling buttermilk, draining of whey through cloth, kneading of the curd, drying in tablets, and cutting into the final size with a sharp knife giving sharp edges. I tried to reproduce such cheese in the laboratory. We are able to produce concave cheese surfaces, resembling the sign GA'AR as outlined in Fig 1. The 18,120 units of GA'AR mentioned by Nissen et al. (1990) when similar in size to the ones in the Afghanistan cheese chain would represent the yield from about 1,120 liters of rnilk coming from about 3 cows. This number could be supported by the amount of butter fat mentioned on the same tablet. It seems to be no problem to record the whole yield of a cheese season together in this form. In Fig. 2, I present photographs of the mentioned cheese chain from Afghanistan (2C), dried cheese curd from Afghanistan (2A) which probably was made as described by AbouDonia for Ekt cheese, and a cheese chain made in our laboratory from boiled buttermilk (2B) as outlined in Fig. 1. That the measures of GA'AR changed from number to volumes in later times (see Stol, 1993) could be an indication that production of cheese by sedentary farmers changed from production mode A or B (as practised by nomads) to production mode C. The mode C was still in use in Bavaria at the turn of this century to secure protein preserves for the wintertime ("Herbstmilch"). Production mode C was possibly the forerunner of the traditional ripened soft a d hard cheeses now so common in Europe and elsewhere.
Note added in proof In spring 1995, a PhD student of our University, a native from Bhutan, showed me a video he produced in the summer of 1994 in Bhutan when he observed a tribe of yak breeders and nomads in the Himalaya valleys at about 4000 m altitude. Dried cheese as shown in Fig. 2 C is a main product and source of income. About 1 to 2 litres of milk are taken from cows while their calves are kept in front of them (as in the famous El-Obed frieze). The fresh milk from several cows is put into a wooden barrel and beaten with a properly equipped stick from above. The butter is removed and placed into leaves until locally consumed. The skimmed milk is placed into a metal vat and slowly heated on an open fire to about 80" C. The milk is slowly stirred and several cups of whey from the previous day are added to aid acidification by lactic
Identification of dairy products in Sumerian texts
Teuber
fermentation. When the curd formation is completed, the coarsely coagulated milk protein is collected in a cheese cloth. Whey is removed by kneading. Surplus whey is given to the "yak dogs". The lump of curd obtained is formed into a square tablet and then pressed in the cloth over night between two milling stones. From the roughly square and flat piece of still moist cheese two kinds of products are made:
How to produce concave dried cheese pieces
1. By cutting the tablet into 4 square pieces, larger cheese loafs are obtained which are dried 4 on a rope by putting the rope through the middle of the loaf. A small wooden stick knotted into the string below each loaf keeps them in place and apart from each other. 2. By splicing a curd tablet along the flat side and cutting the obtained thinner tablets into smaller pieces like the ones seen in the cheese chain from Afghanistan (see Fig. 2 C), a similar product is obtained by putting 20 small cheese pieces with a needle onto one rope. Both cheese types are dried by these nomads in the air for several weeks by hanging them between the poles of their tents. Before sale, cheese chains are briefly dipped into boiling milk to provide for a white surface. Bhutanese nomads use the small cheese pieces mainly as a snack when climbing in the mountains. A dried piece of cheese is kept in the mouth like a candy and slowly dissolved by sucking. Cheese chains are also commonly used as a gift for visitors. Export of dried cheeses to India is of prime economic importance. During his summer term in Bhutan, our student will investigate and record the exact details of cheese making so that we will be able to calculate the exact yields of butter, cheese and whey in order to compare it with the numbers known from the ancient texts from Uruk.
Teuber
Identification of dairy products in Surnerian texts
1. cheese curd tablet 2.drying of the tablet for a limited time
-
to harden surface
3. cutting into small pieces
/p
4. pulling a thread through the middle along the
large axis 5. further drying leads to concave edges of the still soft curd
8
Fig. 1. Outline for the production of dried cheese curd pieces with concave edges
Identification of dairy products in Sumerian texts
Identification of dairy products in Sumerian texts
Teuber
BIBLIOGRAPHY Abou-Donia, S.A. 1978 'Microbial flora and chemical composition of Ekt cheese", Egypt. J. Dairy Sci. 6,49-52 Anderegg, F. 1894 Allgemeine Geschichte der Milchwirtscha# (Orell Fiissli, Ziirich) Baum, E. 1989 Nomaden und ihre Umwelt im Wandel (J. of Agriculture in the Tropics and Subtropics. Beiheft 38. Witzenhausen). Dirar, H.A. 1993 The indigenous fermented foods of the Sudan - A study in African food and nutrition (CAB International, WallingfordIOxon) (The chapter on dairy products contains descriptions of the composition and production technology of cheeses similar to nomadic cheeses of the middle east. The product kush-kush is obtained from fermenting sour milk by skimming the curd coming to the surface by strong gas formation. The curd is sun dried on palm leaf mats either as powder or after moulding into balls. Such balls are known from Akkadian seals as recently described by P. Steinkeller, "Early Semitic literature and Third Millenium Seals with mythological motifs", in Literature and Literary Language of Ebla (Quademi di Semistica 18; Florence 1992), p. 249). Fox, P.F. 1993 Cheese: chemistry, physics and microbiology. Vol. 2 Major cheese groups (2nd ed.; Elsevier, London). (with chapters and remarks on non-European varieties including examples from India, Afghanistan, Egypt, Syria, Saudi Arabia, North Yemen, Iraq, Israel, Turkey and Cyprus). Janzen, J. 1980 Die Nomaden DhofarsBultanat Oman - Traditionelle Lebensfonnen im Wandel (Bamberger Geographische Schriften Heft 3; Bamberg). Mair-Waldburg, H. 1974 Handbuch der Kase. Kase der Welt von A-2, eine Enzyklopedie (VollcswirtschaftlicherVerlag, Kempten). Nissen, H.J., Damerow, P., Englund, R.K. 1990 Friihe Schrift und Techniken der Wirtschaftsvenvaltung im alten Vorderen Orient - Informationsspeicherung und -verarbeitung vor 5000 Jahren (Verlag Franzbecker, Berlin). Ryder. M.L. 1993 "Sheep and goat husbandry with particular reference to textile fibre and milk production", BSA7,9-32. Stol, M. 1993 'Milk, butter, and cheese", BSA7, 99-1 13. Teuber, M., Geis, A., Krusch, U., Lembke, J., Moebus, 0. 1987 '%iotechnological Processes for the Manufacture of Foodstuffs and Fodders", in P.Prave et al. eds., Fundamentals of Biotechnology (Verlag Chemie, Weinheim) pp. 325-379. Teuber, M. 1993 "Lactic Acid Bacteria", in H. Sahm, ed., Biotechnology, Vol. I , Biological Fundamentals (2nd ed.; Verlag Chemie, Weinheim) pp. 325-366.
Fig. 2. Photographs of dried curd cheeses from Afghanistan (2A), cheese chain made in our laboratory from boiled condensed buttermilk as outlined in Fig. 1 (2B), and cheese chain from Afghanistan (2C) collected by the late Professor Marc Bachmann of the Institute of Food Science of the ETH Ziirich.
LATE URUK PERIOD CATTLE AND DAIRY PRODUCTS: EVIDENCE FROM PROTO-CUNEIFORM SOURCES Robert K. Englund (Free University of Berlin)
Introduction Of the two major uses of large cattle we may expect to find documented in proto-cuneiform sources, namely their exploitation as draft animals and as producers of meat and dairy fats, the former, in particular plowing, is not recognizable in its clear form known from administrative texts of the later third millennium. No texts known to me refer to numbers of oxen specifically assigned to plowing units, nor do the sources attest to the substantial costs-feed grainknown to arise with the use of oxen in plowing and seeding fields. Only several uncertain accounts register together the existence of both the plow represented by the sign APIN and oxen represented by the sign Gu4.1 Whether oxen played a large role in field work in the Late Uruk period is thus a matter of conjecture. It seems unlikely that large cattle were widely exploited as producers of meat any more in the Late Uruk period than in later times.2 As sources of dairy fats and cheese, however, cows were clearly prized and closely controlled. A relative abundance of documentation registers quantities of these products, from which herd sizes can be extrapolated that are comparable with those of the pre-Sargonic Lagash period, but that fall well short of the numbers of animals herded during the IIIrd Dynasty of Ur.3 Proto-cuneiform documentation of early cattle herding derives with few exceptions entirely from the large numbers of tablets and fragments unearthed during the German excavations of the southern settlement of Uruk.4 This Uruk documentation exhibits a clear dichotomy between texts dating to the Uruk IV (ca. 3200-3100 B.C.) and those dating to the Uruk IIIIJemdet Nasr (ca. 3100-3000 B.C.) periods. The earliest texts record numbers of cattle apparently assigned named officials or institutions, to the near exclusion of records of dairy produce, whereas among the texts dating to the Uruk I11 period, exceedingly few accounts of groups of cattle are found, but large numbers of records of dairy fats and cheeses, complemented with the existence of an involved metrological system seemingly developed to afford greater control of these products.
Identification of proto-cuneiform signs representing cattle Following the publication and analysis of a small number of Late Uruk documents between 1917 and 1927,s S. Langdon's work on the Uruk I11 period proto-cuneiform tablets from Jemdet NasI.6 resulted in 1928 in the first identification of large numbers of archaic signs with apparent later forms. It was readily obvious that in the case of large cattle pictograms representing the animals' heads, and, pars pro toto, the animals themselves, were the precursors Bull. Sumerian Agric. 8 (1995) 35-50
33
Englund
Late Uruk Cattle
of signs which in later periods represented animals differentiated according to sex and age using ideographic qualifications. Based on the primarily Uruk IV period texts excavated by German teams in Uruk in the years 1928-1931,7 A. Falkentstein noted in 1936 the paleographic development, beginning in the Uruk IV period, of the signs AB2, GU4 and A M A R . ~ Recent considerations about the possible semantic significance of a small number of so-called tokens found both alone and in association with sealed clay bullae, dating to the Late Uruk period prior to the development of writing,g may be considered highly speculative; one can only presume that there is no reason why such semiotic devices should not have been used together with tokens of obvious numerical significance, however the forms which were unearthed next to-none within-bullae in Uruk bear no more than a passing resemblance to the later signs GU4 or possibly AMAR. The Late Uruk sign AB2 depicted, seen from the front, the head of the domesticated female Bos with down-turned horns, the sign GU4 the head of the bull or ox with horns upturned,lO and the sign AMAR the hornless calf with ears held upright. This latter sign, which otherwise did not indicate the sex of the animal, could be qualified in the second writing stage Uruk I11 with the signs SAL and KUR, representing females and males, respectively.1 1 In the course of Uruk excavations following the publication of ATU 1 in 1936, a total of 24 Uruk I11 period fragments representing witnesses of an archaic lexical list dealing with large cattle were discovered and have been published in a volume containing all known archaic lexical material from Uruk and other sites.12 Although in a very poor state of preservation, these witnesses did allow a sufficient reconstruction of the text to confirm early suspicions that it represented the original version of a list which in a possibly redacted form was part of the scholastic curriculum in Early Dynastic Fara, Abu Salabikh and Ebla.13 The compendium contained in its archaic form three clear sections dealing in turn with AB2, GU4 and AMAR; each section consisted of a canonized sequence of sign combinations beginning with entries recording 'stall', hide colors ('reddish', 'white', 'black') and other both standardized and specific characteristics of the respective animals concerning above all their hides. A fourth section partially preserved in only one text consists of animals represented by the sign AM, (AMARgunfi), possibly corresponding to later am (GU4+KUR),"wild bull".14 A section of the well preserved archaic 'tribute' list-in all likelihood an early form of folkloristic literature-contains among standardized sequences of animals and agricultural products the notations INl4 AB2 I IN1 GU4, "10 cows, 1 bull". The same relation of 10:l is found in two following couplets of the text registering female and male sheep and goats, thus listing together the three major domestic animals in the early Babylonian economy.15
Records of heads of cattle In general, accounts of numbers of large cattle are well represented in texts dating to the period Uruk IV,l6 but are essentially unknown in texts from the following Uruk I11 period. The dairy products milk fat and cheese, on the other hand, are well attested in accounts from the latter period, whereas such accounts are seldom found in Uruk IV texts. Cattle as discrete objects were as a rule registered in proto-cuneiform texts in the sexagesimal system; the exceptional use of the sign N8 in the Uruk IV period to designate immature animals is discussed below.
Ikglund
Late Uruk Cattle
Small, characteristically pillow-shaped Uruk IV period tablets17 record the receipt by a named individual of one or as many as several head of cattle (see figure 118). Inscriptions in these accounts consist of numerical notations employing the sexagesimal system, one or more signs representing heads of cattle and one or more signs which seem to represent receiving individuals or officials and possibly, for example, in the case of the signs ERIM, and A L , ' ~the function of the animals.20 Numerical notations representing numbers of both male and female animals were explicitly qualified by the sign combination GU4 AB2 as in later periods both for large cattle and, using a related sign combination, for sheep and goats.21Reverse faces of the 'receipts' remained uninscribed. Simple records of single transactions were booked into larger accounts in a format represented by the two tablets W 9656,ev and W 7227,a in figure 2. Up to five columns on the obverse face of these texts contained from four to seven individual entries, each of which corresponded to one of the simple records discussed above. The numerical total of the cattle recorded in these entries was entered on the reverse face of the account (rotating the tablet around its 'horizontal' axis).22 This total was qualified with the ideograms GU4 or GU4 AB2, respectively, representing cattle and, in the case of the former text depicted, with a further sign combination possibly recording the distributing organization (? NUN,) and the purpose of the distribution (? 'GI' [ I). A third account in figure 2, W 9656,ex, demonstrates the use in the Uruk IV period of a bookkeeping mechanism which recorded an apparently complete herd of adult and young cattle probably separated according to the function of the individual animals. Employing the principle of listing most to least valuable, the text registers numbers of oxen, followed by counted cows and calves (AMAR), in some cases assigned named individuals. In accounts from the Uruk IV period, the calves could, just as is true of lambs and the children of dependent laborers who were probably too young to be put to work, be qualified using the sign N8 (o,that is, the basic sign N1 turned 90" clockwise) which in sexagesimal notations generally designated "112" of a discrete unit.23 Thus the fourth case of the text's second obverse column contains a notation 'Nll N8 representing one cow and one calf; the latter animal was included on the text reverse, column 1, case 3, among a group of four animals qualified as AMAR. Uruk I11 period accounts of herds of large cattle are very rare and register only modest numbers of animals. The preserved sections of text W 14275 in figure 2 contain notations representing just 8 head; it is just the same an interesting example of the correspondence in the archaic period between designations attested in the lexical cattle list and those attested in administrative documents. The sign combinations AB2 NE, (obv. i I), AB2 U4 (i 2), AB2 SU (ii I), AB2 GI6 (ii 2) and GU4 'NE,' (ii 4) in this account, all known from the canonized lexical text ATU 3, 89-93,24are found neither in accounts from the preceding Uruk IV period, nor in the administrative corpus of the following Early Dynastic period, including the age represented in the SIS 4-8 texts from Ur,25 during which the lexical list in its archaic form was part of scribal curriculum. Recorded categorization of animals according to age seems to have been more developed in the Uruk I11 period as well; the text W 14361 (figure 2, bottom right) registers in three cases of its second column notations representing oxen in their fifth, fourth and second years, respectively (sign combinations U4+5NS7,U4+4N57and U4+2N57G U ~ ) . ~ ~ These oxen were apparently in the charge of an official called GAL, UTUL, (obv. iii 2), "head cowherder".
Late Uruk Cattle
Englund
Examples of receipts (?) for one cow (sign
Englund
Late Uruk Cattle
@)
Examples of receipts (?) for one or at most several bulls/oxen (sign& fl
Examples of receipts (?) for one calf ( s i g n p ) and for mixed cattle (?; sign combination
W 9579,da
W %56,in
Figure I: Examples of receipts (?)for one or at most several cattle (all copies from R. Englund, ATU 5).
@)
W 9335,~
Figure 2: Examples of complex accounts of cattle. The upper three texts (copiesfrom R. Englund, ATU 5 ) date to the Uruk IV period and register numbers of animals totaled on the texts' reversefaces; W 7227,a books a total of 54 cows and bulls. The lower two Uruk 111period texts contain accounts of small numbers of cattle qualified with signs known from the lexical list ATU 3, pp. 89-93 (W 14275), and with sign combinations known to represent the animals' ages (W 14361; Jirst published in ATU 2, pl. 60).
Englund
Late Uruk Cattle summatfion
(sign DUGb) per two (W 20274,12) or four (W 20274,63) milk cows. We may compare these relationships with the amount of the dairy fat required as a yearly delivery of herders from the pre-Sargonic Lagash (ca. 2400-2300 B.C.) and the Ur I11 (ca. 2100-2000 B.C.) periods. Numerous accounts from the earlier period set this quota at 10 Old Sumerian sila or ca. 15 liters, accounts from the later period on the other hand assume a yearly delivery of just 5 neoSumerian sila or ca. 5 liters. If we assume a yearly milk production of reasonably well fed cows kept in semi-arid climates of between 700 and 800 liters, of which about half will have been given to their calves, the remaining milk with a fat content of ca. 4% would result in ca.
2 cows (reconstructed)
I heifer calf
b~rthof l calffw cows)
D~D
BQ ry S
designation of calf as "yearling"
Late Uruk Cattle
1:nglund
responsible herdsman
summation
1 heifer calf
4 cows
1 bull calf
"dairy fat"
responsible herdsman W 20274,63
Figure 3: Two Uruk III accounts of small herds of milk cows. The sign combinations in the lower left case of each account represent the herders responsible for the cows.
Records of archaic dairy products Related herding accounts from the Uruk I11 period, of which only two are preserved well enough to permit a reconstruction of their c0ntents,~7record small numbers of cows together with their offspring, qualified SAL+AMAR and KURa+AMAR ('heifer calf' and 'bull calf'), from the accounting year of the text. Both texts record a ratio of two adult cows per recorded calf.28 The two accounts depicted in figure 3 further serve as examples of the greater accounting effort invested in the dairy production of cows in the Uruk I11 period. In a fashion parallel to that employed in recording the amount of dairy fat to be delivered by herders of sheep or g0ats,~9the accounts book in the totals on their reverse faces one jar of dairy fat AB, 'cow'
6@@
GU, 'hull', 'ox'
AMAR 'calf
SALAMAR 'heifer calf
KURAMAR 'bull calf 2~7,200+3~1,200+1~120 = 18,120 units of
D
Figure 4: Proto-cuneiform signs representing cattle.
Figure 5: Simple accounts of dairy productsfrom the Uruk Nperiod (above; copiesfrom R. Englund, ATU 5 ) and a large account from the Uruk Illperiod (below; see ATU 2, pl. 55, and Archaic Bookkeeping, p. 94) of products from animal husbandry, including the signs for dairy fat (DUG,,) and cheese (GA'AR).
Englund
Late Uruk Cattle
Englund
15 liters of dairy fat ('clarified butter', 'ghee'). This suggests that the pre-Sargonic Lagash herders delivered all the produce of their cows, the Ur I11 herders in contrast perhaps a third. Since, moreover, the archaic sign DUGb seems to have represented a ceramic jar no larger than ca. 10 liters (see below), it may be proposed that the archaic herders delivered dairy produce at a rate comparable to that of the Ur I11 period and were thus akin to contract herders who kept a percentage of the produce of their herds in lieu of other compensation for their labor. The first eight lines of the archaic lexical list 'Vessels,' one of several compendia for which precursors from the Uruk IV period are known, in fact consist of entries with the signs DUGb, KISIM* and other signs which represent containers of fats used in the administration of ar, ~ ~often found inchaic dairies.30 These signs, including NI,, DUG, and U K K I N ~ + N I ,are scribed together in administrative documents beginning in the Uruk IV period (see figures 58). In particular, the association of the sign NI, with DUGb in such documents as W 9206,c, W 9579,ad and W 9579,ah, and of NI, in the same case with AB2 and with DUG, in the account W 9656,eq (all figure 5) demonstrate that this sign should represent a container of dairy fat from its first use in the Uruk IV period. The sign, the real referent of which is unknown, is in later cuneiform documents the general designation of fats of all kinds. Only indirectly associated with the sign representing dairy fat, DUGb, is on the other hand the sign GA'AR in such texts as W 20274,97 (figure 5). This sign, found as a general object designation in a section of the archaic vessels list following a long section on containers of fats and other products,32 is, as a clear precursor of the Fara and pre-Sargonic Lagash sign LAK 490-itself replaced in Ur I11 documents by the sign combination ga.kJARIUDgunli-, posited to represent a unit of cheese (Sumerian reading possibly lga'aral or Igamurl). Whereas oil vessels were counted with the sexagesimal system, GA'AR was reckoned in discrete units using the so-called bisexagesimal (120-base) system and so may be associated with the products GAR. (dry grain products) and KU6 (fresh? fish). We consider such objects to have formed a part of the archaic rationing system.33 Unfortunately, no archaic texts known to me document the arithmetical relationship between dairy fat and cheese well attested in preSargonic Lagash and Ur I11 accounts.34 Containers of dairy fat and other liquids were not only as discrete objects counted using the sexagesimal system, but were also as members of a liquid capacity metrological system divided into smaller units using one of three numerical conventions. In the first place, the sign N8 (T), discussed above as a designation of immature cattle in the sexagesimal system, as a rule qualified 'I2 of some discrete unit, above all the contents of vessels and baskets, including those represented by the signs DUG, and KAS, (beer), KISIM*, DUGb, DUG, and DUG+AS, (fats) and GA2,+kJI (a fish product).35 Notations in a number of Uruk IV period texts suggest that the sign N8 in the sexagesimal system could also represent a smaller fraction than 'I2 of an object, possibly lIlo; the objects so qualified in these notations are, unfortunately, not always clear, although DUG, seems attested in at least two of the accounts.36 A second, Uruk I11 period convention used explicitly and alone for dairy fats seems on its surface substantially more complex, yet shares the basic structure of 'I2 and 'Ilo of the unit 'container.' A large number of accounts, of which nearly all derive from the same locus W 20274 as the herding texts described in figure 3 above, contain notations in this metroloFigure 6: Accounts concerning dairy fat stored in the jar DUG,.
Late Uruk Cattle
Late Uruk Cattle
Englund
gical system which exhibits the structure lx N1 vessel = 2x Nl+KU3a= 1 0 N2 ~ (corresponding to the basic unit N1 crossed by a horizontal stroke: B; see figure 9). W 20274,6 in figure 6 offers a simple summation of three entries with numbers of a container of fats represented by the sign DUG:^: 7N1 DUG, IN1 KU3, + 4N1 (DUG,) 1 KU3, + 5N1 (DUG,) = INl4 7N1 DUG,, that is, 7 112 + 4 112 + 5 = 17. The only known duplicate administrative texts from the archaic text corpus, W 20274,33 and W 20274,89 (see figure 6 and the transliterations below), contain a somewhat more involved account, yet the reckoning steps exhibited by both are easily recognizable as simple additions of whole numbers and fractions from the same metrological system.38
Obv. i l a lblal lbla2 lblb lb2 2 3 4
5 ii 1
2 3 4 5 6 Rev. i 1 2 3 4
5
6 ii 1
'5N11 ; 'DUG,' DUB, lN1 ;BA AN ZABALAM, lN1 ; SAL BA PINGbl ZATU75 1, DUR2 3N5, 'BU,+DUQ 3N1 ;BA 'KI; IN1 I lN1 KU, ;ZATU649 lN1 'KU3: ;'ZABALAM; AB2 lN1 ; SI,, NE, SE, GI GA, MUD NUN, 2N1 ;ZATU648 ABb P1NGbl+3N5, SU, 3N1 ; SI U4 'SIC? 2N1 ;TUR,, A GA, BU, 'Ijr BA GI+GI ZATU648
5N1 ;DUG, DUB, lN1 ; BA AN ZABALAM, lN1 ; SAL BA PINGbl ZATU75 1, DUR2 3N5, BUa+DU6 3N1 ; BA KI, lN1 I lN1 KU, ; ZATU649 lN1 KU,, ; ZABALAM, AB2 lN1 ; SI,, NE, SE, 'Gr 'GA; [MUD NUN,] 2N1 ;ZATU648 PIRIGb1+3NS7ABb SU, 3N1 ; SI U, SIG 2N1 ;TUR3, A GA, BU, HI [BA GI+GI ZATU6481
2N1 ; ZATU648 '3N11; 'SI U4 SIC? '2N11;'TUR,, A' lN1 I lN, KU, ;ZATU649 IN1 KU3, ; ZABALAM, AB2 lN1 ; SI,, NE, 'Gr SE, 'INl4 3N11;DUG, GI+GI BA
[2N1 ;ZATU6481 '3N11; SI 'U4 SIG' 2N1 ; TUR,, A lN1 I lN1 KU3, ; ZATU649 lN1 KU,, ; ZABALAM, 'AB,' lN1 ;NE, SI,, GI SE, INl4 3Nl ;'ZATU648' DUG, GI+GI BA
An example of an account containing the full structure of the second conventional system used to record fractions of containers of fat is found in figure 7. This and a small number of other accounts clearly demonstrate the relationships N1 KU, = If2, and N2 ( B ) = lIl0 N1 DUG, in this system39 and thus make plausible the assumption that it may represent a development from the Uruk IV system with, dependent on context, N8 equal both to N1 KU3, and to N2. The meaning of KU3, in this connection is, aside from the fact that it indicated a half measure, not obvious.40 The third means of designating fractions of oil jars is found in just one text, presented in figure 8, but here fully documented. W 21682 contains on its obverse face two columns with 5 entries, each of which consists of the numerical sign N1 together with the sign combinations S1LA3,+GARA2, or SILA3$GAa-the former explicitly written in the first four cases of the first column, the latter probably only in the lost first case of the second column-representing
Late Uruk Cattle
linglund reconstructionof the summation:
obverse
reverse I
memlogical relations:
ID$,
=
.D(
Figure 7: Reconstructed account of dairy far stored in jars demonstrating the metrological relations in the system DUG,.
units of a dairy product, the sign SI (meaning unknown) and further ideograms probably representing receiving individuals. The meaning of both signs GARA2, and GA, is not certain. Clearly, they represent vessels and are both invariably in context with dairy products. The former sign, a gunified variant of the sign DUG,, is attested in the archaic Ur (ED 1-11) version of the lexical list L6 A, 1. 20, as a
,,, >s
standardized vessel SILA, (pro@ly a Blumentopf wlth ca. 0.8 liter capacity)
Bd> E
relations between the standardized vessels:
,>
product
5 units
>s
+,
=I
grand total of the dairy products:
1 unit
S;?dgGy
Figure 8: Metrological relationship between SILA, and DUG,.
disbursement
>P
Englund
Late Uruk Cattle
variant of GA, in the combination GALa GARA2,, "head of GARA2,", and representing a product among notations for domestic animals and other agricultural products in the list ' T r i b ~ t e ' The . ~ ~ Uruk IV period form of the latter sign (see figure 10 below) is apparently the representation of a flat basket, the inner surface of which was probably coated with bitumen so that the basket could be used in the milking of dairy animals. The reverse face of the tablet contains in the first column (to the right) subtotals of each of the obverse columns, numerical notations representing five units qualified by the sign combinations SILA3,+GARAZaand SILAJa+GAa,in the second column the final total N1 DUGb qualified with SI and the sign GU7 (SAG+GAR = 'human' + 'rationing bowl'; the specific administrative meanings in the use in archaic texts of the two signs BA and GU7, both apparently qualifying distributions, are poorly ~ n d e r s t o o d ~representing ~) "disbursement". This equation proves that the sign SILA3,, to be identified as a pictographic representation of the mass-produced Blumentopf which followed and for some time in Late Uruk levels coexisted with use of the beveled-rim bowl GAR, assumed the same metrological function in the three systems discussed as Uruk IV period N8 and Uruk I11 N2, all representing a measure equal to 'Ilo of the amount of liquids or semi-liquids contained in some larger vessel.43 According to data derived from excavations, above all measurements conducted of the masses of beveledrim bowls found in Late Uruk settlements, and in accordance with textual analysis, the most plausible current working hypothesis of the absolute capacities of these various units is the following:44 GAR = IN8 (?; Uruk IV) = SILA,, = 1N2 1N8 = IN, KU,, IN, DUGb,, etc.
= ca. 4/5 liter
The sign DUGb, representing a ceramic jar without a spout, was consistently distinguished from the sign DUG, L, including the representation of a spout. N,, N , DUG, N, SILA,~ This fact and the contextual usage of -lo_ L both signs suggest that the former jar will most likely have contained semiliquids, the latter liquids, above all beers. A large number of signs were Figure 9: Metrological systems employed in dairy notations (cp. impressed in DUGh in archaic lexical Englund, Iraq 53 [I9911 101-4). The application o f the .. " upper system with dairy products is not proven, the lowtexts, to a lesser extent attested in ader two systems are only known from the Uruk Illperiod. -
N , DUG~NKKIN,+NI~
N, K U , ~
lhuk IV
sign name
~0 0
meantng
DUG*
beer
K A ~
beer
Uruk l V
Uruk Ill
srgn nome
00 00
meaning
DUGb
dairy fat
DUG.
dairy fat
5 ~ ~ 1 %dai, fat
p G
KISIMd
butter fat from sheep's milk
P E P
KISIM,
butter fat from goat's milk
0,&p
Uruk I V
Umk I l l
mme
0 0
dairy fat ?
b 8 $3
meanlng
milk ?
@@
dairy fat mixed with crushed barley?
BB>
GAR*,
irem ?
Figure 10: Probable archaic designations of liquid /semiAiquid products.
lninistrative texts, to specify the product contained in the jar represented by the sign, including among others SE ('barley'), NAGA ('an alkaline plant' ?), TI (?), MAS (male goat), KUR (a plant related to the grapevine ?), GIS ('wood'), KU6 ('fish') and S A ('pig').45 ~ ~ The sign K A S is ~ also one of these composita, consisting of dots impressed within the sign DUGb in its Uruk IV form, abstracted to vertical lines in the Uruk I11 writing phase. These impressed dots are in my opinion the same as those added to numerical signs from the grain capacity system (forming the system S*)46 to indicate amounts of barley groats used in the production of dry and liquid grain products.47 This K A S ~is also consistently distinguished from the related sign KAS,; whereas the latter sign is generally attested together with notations representing both rough-ground barley and malt, thus representing beer, the former is invariably found in a dairy context.48 The same applies for the sign KAS, (see figure lo).@
NOTES TO TEXT 1
2
N,
D>
Uruk I l l
= ca. 4 liters = ca. 8 liters
These measures would, moreover, suggest a yearly delivery in the Uruk I11 period of from two to four liters of dairy fat per milk cow (see figure 3 above) and be in general accordance with figures known from the Ur I11 period. With an expected yearly delivery of between 'I4 and 112 DUGb per cow-year, the notation of '26 DUGb' in W 20274,97 obv. i 1 (figure N8 5 ) , understood as part of an annual acNI4 N, 2 count, would correspond to ca. 50- 100 a - lo D -4,. N~ milk cows, or total herds of perhaps U 100-200 head. N,,
Late Uruk Cattle
I:nglund
3
See, for example, the two texts R. Englund, Archaic Administrative Texts from Uruk: The Early Campaigns (= ATU 5; Berlin 1994) pl. 86, W 9656,f, and pl. 100, W 9656,dr, with counted APIN GU, apparently assigned to temple households. The inscription of the latter text is duplicated in the second column of the obverse of the former. The very meager bone remains from Uruk of Bos taurus identified by J . Boessneck, A. von den Driesch and U. Steger, BagM 15 (1984) 170-172, were almost entirely of adult animals. Although the authors believed the crushed remains indicated the exploitation of cattle for meat, the numbers of bones-only 30 of the 73 specimens were from Late Uruk levels-permit no more than speculation as to whether the animals were selected for meat or were simply butchered in old age or after having died from some other cause. A general idea of the sizes of state-controlled herds in the pre-Sargonic Lagash period is offered by the text DP 93, which registers a total of 119 adult and juvenile animals in the care of dairy herders (cf. K. Maekawa, Zinbun 15 [I9791 122-123) in the year Enentarzi 5. The numbers in DP 93 are confirmed by further documents recording the delivery quotas imposed on the herders of the dairy fat i fib ssb.gald6.a and the cheese LAK 490 (DP 273-276; Nik. 1,257,259; RTC 64; TSA 37; VS 14,89) on the one hand, the amount of feed distributed to the herders (Nik. 1, 68, HSS 3, 37, etc.) on the other. This suggests total holdings, including draft oxen, of ca. 250 head in the economic unit represented by the archive from Girsu. State-owned herds in the Ur I11 period were substantially larger; the account MVN 15, 108, books for a single dairy herd supervisor from Umma over 300 milk cows, from which may be extrapolated a total of close to 1000 cattle. The supervisor Atu was, moreover, one of possibly several such managers active in Umma (see also the numbers in the text SET 130). The archaic texts from Uruk discussed in the
Englund
4
5
6
7 8
9
10
11 12
13 14 15
16 17 18 19
20
Late Uruk Cattle
present study document cattle herd sizes of between 50 in the Uruk IV period and possibly 100-200 in the Uruk I11 period. While it is true that with a current count of 5000 texts and fragments the corpus from Uruk can be expected to contain the majority of all early accounting genres, the ca. 450 archaic texts now registered from excavations of other sites, above all Jemdet Nasr (together with some few texts from Kish numbering 244 tablets and fragments), and from the antiquities market, do exhibit a broad spectrum of features known from archaic bookkeeping, in some cases containing the best documentation of particular economic activities. Still, the certainly uneven coverage of household activities in the proto-cuneiform documents, resulting from the vagaries of excavations, is the more to be underscored given the relatively modest number of texts and fragments from the Late Uruk period. V. Scheil, "Quelques signes originaux de 1'Ccriture cunCiforme," RA 14 (1917) 93-94; A. Deimel, LAK (1922) p. 73, no. 2; F. Thureau-Dangin, "Tablettes ? signes i picturaux," RA 24 (1927) 26-29. The inscriptions on the archaic stone tablets republished by I. J. Gelb, P. Steinkeller and R. Whiting, OIP 104 (1989) as nos. 1-11 and 13, of which the Blau pieces (10-11) were first published in 1885, and that on the tablet published by S. Langdon in Excavations at Kish I (Paris 1924) 99-101 and pl. XXXI, are in my opinion to be dated to Early Dynastic 1-11. The Herbert Weld Collection in the Ashmulean Museum: Pictographic Inscriptions from Jemdet Nasr Excavated by the Oxford and Field Museum Expedition (= Oxford Editions of Cuneiform Texts 7; London etc. 1928). See now the edition by R. Englund and J.-P. GrCgoire, The Proto-Cuneiform Texts from Jemdet Nasr (= Materialien zu den friihen Schrifzeugnissen des Vorderen Orients [MSVO] 1; Berlin 1991). See the introduction to ATU 5, pp. 11-23. ATU 1, p. 52+5-53+1.Falkenstein drew particular attention to the different graphic forms of the two signs AB2 and SIG (first noted by S. Langdon in OECT 7 s.v.),the latter a tend-form of the sign U4 which seems to designate "evening." R. Englund, review of D. Schmandt-Besserat,Before Writing, in: Science, from 11 June 1993, p. 1671; P. Damerow and H.-P. Meinzer, "Computertomografische Untersuchung ungeijffneter archaischer Tonkugeln aus Uruk: W 20987,9, W 20987,ll und W 20987,12," BagM 26 (1995) 9. I am not aware of a graphic differentiation in archaic documents between breeding bulls and castrated oxen, both apparently = GU4 (the few bulls kept for breeding in pre-Sargonic Girsu were called simply gu4 hb, "bull of the cow"). See figure 4 for a table of the graphic forms which represented large cattle in the Late Uruk texts. R. Englund and H. Nissen, Die lexikalischen Listen der archaischen Texte aus Uruk (= ATU 3; Berlin 1993), esp. pp. 22 and 89-93. SF 81; OIP 99, nos. 25-27; MEE 3, nos. 12-17, pp. 47-50 ("Lista di animali A"); s. Th. Krispijn, JEOL 27 (1981-82) 47-53, and J. Krecher, OrAnt. 22 (1983) 179-189, to the syllabic version MEE 3, no. 62. See ATU 3,2241. See ATU 3, 25-29. The numerical relationship probably mirrors the number of males kept in herds for purposes of breeding. Incidentally, these accounts were without exception uncovered during the first Uruk campaigns, published as photos in A. Falkenstein's ATU 1 and subsequently as copies and indexed in my ATU 5. For an overview of the text formats employed in such accounts see M. Green, "The Construction and Implementation of the Cuneiform Writing System," Visible Language 15 (1981) 345-372, esp. 349-356. All tablet copies found in this and the following figures are represented at 75% of original size. See the texts included in figure 1, W 9579,cz (upper row) and W 9579,dd (second row). ERIM, may mean in this context "for the yoke," and since both W 9579,dd and W 9656,ak depicted in the second row of this figure contain the sign combination NESAG, GAR3, it may be reasonable to assume that the sign AL in the former text represented a function of the ox (GU,) recorded there (see also ATU 3, 91, to line 5 1: U4 AL X GU,, with no apparent correspondence in the ED texts, assuming MEE 3, p. 53,ll. 51-52 = ATU 3 , 91,ll. 50 and 52). While this is not the place to discuss the methods available for the isolation and identification of object designations, in the cases involved here of cattle, as against the other ideographic representations found in
1;nglund
Late Uruk Cattle
syntax-free archaic accounts, it may be stated that as a general rule signs representing counted objects are situated closest to the numerical notation, inscribed, insofar as this is discernible due to the existence of sign distortions caused by subsequent inscription, immediately after the numerical notation and before the impression of the accompanying ideograms. This is true both for receipts, for example, W 9335,c in figure 1, lower right, and in such large accounts as W 7227,a in figure 2. The largest attested total of adult animals is '54' contained on the reverse of W 7227,a in figure 2. Although the bundling step '60' was thus not crossed, there can be little doubt that this notation is from the sexagesimal system, since the only other candidate known from the early Near East used to count animals, the decimal system from proto-Elamite Susiana, was not used in Babylonia (see P. Damerow and R. Englund, The Proto-Elamite Texts from Tepe Yahya [= American School of Prehistoric Research Bulletin 39; Cambridge, MA, 19891 24 and 53-55 with footnotes) and since other animals and humans were demonstrably counted sexagesimally in the archaic period. See A, Vaiman, "Die Bezeichnungen von Sklaven und Sklavinnen in der protosumerischen Schrift," BagM 20 (1989) 121-133, and the comments of P. Damerow and R. Englund, BagM 20, 137-138. See above, notes 12-14. The sign combination AM, KUR, of obv. ii 3 may bear some relationship to the presumed AMb section of the archaic cattle list cited above. E. Burrows, Archaic Texts (= UET 2; London 1935). For a description of archaic designations of years and animal ages see ATU 2, 146, and R. Englund, "Administrative Timekeeping in Ancient Mesopotamia," JESHO 31 (1988) 140-148 and 156-162. The standard age sequence for Ur I11 bullsloxen attested, for example, in the theoretical account TCL 2, 5499 , --), mu.2, mu.3, gu4 gal, "milk (I. J. Gelb, JCS 21 [I9671 64-69), was gu4 amar ga, gu4 m u . l ( ~ S sign bull-calf, one-year bull, two-year bull, three-year bull, large (full-grown) bull." First published by M. Green, "Animal Husbandry at Uruk in the Archaic Period," JNES 39 (1980) 32, nos. 35-36; see now H. Nissen, P. Damerow and R. Englund, Archaic Bookkeeping (ChicagoLondon 1993) 89-93 with figure 7 1. The same ratio is attested in the Ur I11 account cited in note 26 above; this reference should be understood as hard evidence for an administrative rule, since as a theoretical exercise (see Archaic Bookkeeping, 97102) the text mirrored traditional quotas and since those quotas and value equivalencies-yearly deliveries per milk cow of dairy fat and cheese, and silver values of these products, respectively-also underlying numerical entries in the text are found to be exactly the same in contemporaneous administrative documentation. Insofar as an adult cow should bear one calf per year, the ratio of 1:2 may result from a set number of calves either culled from the herd for meat, sacrifice or some other purpose, or a number accorded hired herders as their income. However, it is impossible to derive a similar rule for the archaic period based on just two small accounts, in particular since the production of dairy fat per cow registered in the two texts seems to be inconsistent. M. Green, op.cit., pp. 12-13 (the last column of this table records the fraction of the oil contained in the jars KISIM& delivered per adult female animals; for example, the first text W 20274,55 books obv. i 1 70 ewes and rev. i 2 a fat delivery of 3Nl IN8 = 3 11, KISIM, for a quota of 'I2,, or, with the author, 0.050 KISIM, fat per ewe). See ATU 3, p. 29, figure 12, and pp. 30-32. Of the proto-cuneiform signs representing ceramic vessels, only NI, may have been a two-dimensional depiction of clay objects found in the pre-literate accounting devices known as clay bullae; see now the summary of D. Schmandt-Besserat's important work on these symbols i i her Before Writing, vol. 1 (Austin 1992) 108-128, and my review of her treatise in Science, 11 June 1993, p. 1671. The 'oil tokens,' believed themselves to have represented concrete containers, have been found in bullae from Uruk and from Habuba Kabira in Syria. ATU 3, pp. 29-32. See P. Darnerow and R. Englund, ATU 2, 132-134. Alone the use of different numerical systems with the two products in archaic texts would obscure a possible connection between DUG, and GA>AR.Cheese was in later periods quantified not as discrete units but in capacity measures. The specific contents of dairy milk (87% water, and 13% solids, of which ca.
Englund
35 36
37 38 39 40
41
42
43
44 45
46 47
48
49
Late Uruk Cattle
3 3 % was fat, 33% casein, and 5% lactose) may have played a role in determining that herders from the pre-Sargonic Lagash period were required to deliver 18 sila, from the neo-Sumerian period 7 'I2 sila of cheese per year, i.e., ca. 1.5 x the norm set for dairy fat. This roughly corresponds to the volume of kaSk cheese which may be derived from a measure of milk relative to its fat content. I assume that the norm of 7 '1, sila derived from a nice relationship to the norm of 5 sila of i nun fat from Ur I11 times, and that 18 was also a "nice number" in pre-Sargonic Girsu bookkeeping: 3 bhn, or 11, of the barig consisting of 36 sila. ATU 2, 128 c. Cf. the texts W 19466,a (unpubl.) obv. i 1 ('5N1 5 N i ; DUG,) and W 20652 (unpubl.) obv. i 1 (4N1 [ ] '6N; ; DUG, [ I). The notation 3N1 9N8 in ATU 5, pl. 111, W 9656,gl (cited ATU 2, 129 d, as ATU 1, no. 490) refers to an object not preserved in the second case of the tablet, and this and the preceding two notations could in principle derive from a number of other numerical systems. Clearly sexagesimal, however, is the notation r3N342N11[ 1 '4N; in ATU 5, pl. 64, W 9 5 7 9 , ~rev: 1 (cited ATU 2, 129 d, as ATU 1, no. 352); the apparent object represented by the sign combination SUHUR KAS,, literally 'jar of dried fish flour oil,' must at least be admitted as a weak reference for the use of N8 < 11, in a sexagesimal notation of oil jars. Note further the possible value N8 = 1Il0 'iku' in the area system attested in the two Jemdet Nasr texts R. Englund and J.-P. Grkgoire, MSVO 1, nos. 2 obv. i 2d, and 4 obv. i 5d. See R. Englund, "Archaic Dairy Metrology," Iraq 53 (1991) 101-104. Including only the 3 units qualified as BA KI, in the second subcase of the first case of each text's obverse face, the addition is: 3 + 1 '1, + 112+ 1 + 2 + 3 + 2 = 13 DUG,. See ATU 2, 131 c. No dairy accounts known to me contain a notation 5-9N2, in compliance with the expected replacement of 5N2 with IN1 KU,,. I might draw attention to the fact that tokens often related to this sign have been found in clear association with sealed bullae in Uruk (see D. Schmandt-Besserat, Before Writing, vol. 1, p. 119, to W 20987,27; unmarked crescent) and possibly within still complete bullae &om Susa (identified tomographically; this information was given me by J. Friberg). See ATU 3, pp. 73 and 114-116, respectively; in 'Tribute' followed by a notation of '10 cows'. See preliminarily my "Grain Accounting Practices in Archaic Mesopotamia," in J. HOyrup, ed., Changing Views on Ancient Near Eastern Mathematics (Berlin forthcoming). The text W 20274,72 (unpublished) seems to contain an addition '2N11[ ] + '2N1 SILA,,+GARA,,' + IN8 = IN, DUG,, implying that, as might be expected, N8 also served in this system to represent 'I2 of a basic unit and 5x N, SILA,,. The use of N2 and N, SILA,, together in W 20274,92 (unpubl.) with an apparent summation on the reverse is not otherwise attested among dairy accounts and may so represent a scribal error. See ATU 2, 15360. See 11. 21-61 of the archaic lexical list 'Vessels', ATU 3, 125-129. Far fewer examples are known of signs impressed in the beer sign DUG,, including KASKAL, LAM,, NAGA, and U2, (types of herbs ?), all of which are attested only in administrative texts. ATU 2, 140-141. This interpretation should be noted to the suggestion of R. Michel, P. McGovern and V. Badler, "Chemical Evidence for Ancient Beer," Nature 360 (5 November 1992) 24, that the grooves found on the inside surface of vessel shards excavated in Late Uruk levels of Godin Tepe were represented by the internal markings in the DUG signs. Cf., for example, the unpublished texts W 15774,b obv. i 1 (KAS, together with DUG, representing a container of dairy fat) and W 2051 1,2 obv. v 3a, vi lb3, 2b3 (registering counted KAS, SILA% BA in a large account of dairy fats). Cf., for example, the texts'W 19408,5 (KAS, together with DUG,), 12, 17, W 20044,40 (all unpublished), and M. Green, Visible Language 15, 355, fig. 4a, W 21049 (together with DUG,), MSVO 1, 179 obv. i 3 (KAS, following DUG,) (in no instance together with DUG,; to be noted to ATU 2, 131).
SHEEP AND GOAT TERMINOLOGY IN UR I11 SOURCES FROM DREHEM Piotr Steinkeller (Harvard University)
1 Introductory remarks First, for the benefit of readers not versed in Assyriology, a few words must be said about Drehem and its written sources. Drehem, ancient Puzrish-Dagan, was founded by king Shulgi in the middle of the 21st century BC as a center for the collection and distribution of domestic and wild animals that were delivered to the state as tax revenues or were obtained as booty during the course of military campaigns.l In fact, Drehem was not the only such center in Babylonia; in several other cities (Ur, Uruk, and possibly Eshnuna) there existed similar but smaller animal centers, which functioned as local branches of Drehem. This enterprise operated for some thirty years. Among the animals that passed through Drehem and its subsidiaries one finds primarily cattle, sheep, and goats, both domestic and wild, but also gazelles, deer, bears, pigs, marsh boars, and various types of birds. As far as it can be determined, the Drehem enterprise was a completely unique occurrence in Mesopotamian history. Unlike other livestock operations known from Mesopotamia, Drehem was not involved in animal production, its sole purpose having been the collection and distribution or allocation of animals. Furthermore, the overwhelming majority of animals handled by Drehem and its branches originated not in Babylonia itself but on its northern and eastern peripheries: the Assyrian piedmont, the Zagros mountains, and the western portion of the Iranian plateau. The special character of the Drehem enterprise gave rise to the development of an exceptionally rich animal terminology. The importance of this terminology to the modern scholar lies in the fact that it was used by the same group of scribes during a comparatively brief period of time. This assures that a single classificatory system is involved, which in turn allows highly reliable identifications of species and breeds. The present study deals specifically with Drehem terminology pertaining to domestic and wild caprinae. The types of terms considered are only those which, however broadly, relate to biological characteristics of animals. Not treated here are the numerous terms specifying the intended use of animals, such as kin-@&a,which identifies the lamb suitable for extispicy? or the various descriptions of meat preparation and processing, such as mu-du-lum, "sated",3 uzuhhd, "dried": urn-a-bala, "meat broth"? ka-izi, " r ~ a s t e d "and ~ ne-mur-ta ba-leg6, "roasted in hot embers."7 Because of limitations of space, I will not discuss the historical evolution of the Drehem terminology and the distribution of terms depending on text types. Nor shall I offer a Bull. Sumerian Agric. 8 (1995)
49
Steinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
systematic comparison of the Drehem usage with that documented in other Ur I11 centers; comparisons will be made, but only selectively.
2 Species Beginning with the question of species, four species of caprinae can be confidently identified in Drehem sources: the domestic sheep (generic udu), the domestic goat (generic maS), the wild sheep or the Asiatic mouflon (udu-hur-sag),8and the wild goat or the bezoar (dard~ldara~).~ The identifications of udu-hur-sag and dar* as the mouflon (Ovis orientalis) and the bezoar (Capra hircus aegagrus), respectively, are assured by the existence of hybrids of these animals with their domestic counterparts.10 At the same time, it is striking that the Drehem terminology and, for that matter, Mesopotamian animal taxonomies in general, lack specific terms for wild caprinae such as the urial (Ovis vignei), the ibex (Capra ibex), and the markhor (Capra falconeri), which, as we can judge from surviving pictorial representations,ll were well known to ancient Mesopotamians. It is possible, therefore, that, in its broad sense, the term udu-hur-sag includes also the urial, whereas the term dar* designates also the ibex and the markhor. Other species that need to be discussed in this connection are mBS ~ a - ~ a n ("Makkan ~~), goat", and maSda (A&. ~abitu),"gazelle." Since the "Makkan goat" occurs infrequently in Drehem sources,l2 and since this animal came from (or was at least associated with) the land of Makkan, which can positively be identified as the Oman peninsula,l3 it almost certainly was not simply a breed of domestic goat, but rather a distinctive exotic species, whose physical appearance was somewhat goat-like. In my opinion, the best candidate for that species is the Arabian white oryx (Oryx leucoryx), a type of antelope, which is by far the most characteristic animal native to Oman.14 That the oryx was known in Mesopotamia, is proved by ancient iconography.15 An alternative, but, in my view, less likely identification,l6 would be the Arabian tahr (Hemitragusjayakari Thomas 1894), a type of small mountain goat.17 Turning to the mdda, there is no doubt that we find here the "gazelle." In all probability, maSda does not designate any specific type of gazelle, but rather is an inclusive label for all gazelles native to the region, e.g., the goitered gazelle, the mountain gazelle, and the Dorkas gazelle. Finally, we should consider the question of Segg-bar,18Akk. sapparu. The only reason why it is at all necessary to address this issue here is the fact that this animal is thought to be either a wild goat or a wild sheep.19 However, it is quite certain that both these identifications are wrong. That the legg-bar was neither a sheep nor a goat is demonstrated quite conclusively by the fact that, in Drehem summary accounts, the legg-bar are grouped together with deer (lulim), equids, and bovines,20 and not with sheep and goats, domestic or wild. Equally telling is the fact that the age of Segg-bar is occasionally identified,al and that its young are described as amarJ2 both features never being attested in connection with sheep and goats. Since Segg-bar are commonly associated with deer and gazelles,23 and since this species very likely had antlers,24 it appears highly probable that the Segg-bar was a type of deer. Assuming that lulim is the red deer (Cewus elaphus), the Segg-bar should in all probability be identified as the Mesoptamian fallow deer (Duma mesopotamica),25 which has characteristic palmated antlers,26 and whose coat is spotted with white.27 In this connection, it may be significant that, according to the Kesh Temple Hymn, lines 48-49,28 the Segg-bar was "spotted" or "dappled" (an-a).
Steinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
3 Breeds 3.1 Sheep 3.1.1 Native breed of sheep It is safe to assume that the sheep identified simply as udu, without additional breed designations, represents the standard native breed. A specific name of this breed is udu Ki-engir,(GI), the "native" or "Sumerian" sheep, which appears only in three Drehem documents.29 It is noteworthy that in two of those sources (AUCT 1 405; BIN 3 350), both concerning the Suaid of the city URU.SAG.RIG7, udu Ki-en-gir, are listed in apposition to udu Lu-lu-bu, the "Lullubean" sheep. In the Ur 111 sources from Lagash, Urnrna, and Ur, the native breed of sheep is described by the appellative udu-gir, or udu-em-gir,, "native sheep."30 3.1.2 Fat-tailed sheep Ilrehem sources mention at least two types of fat-tailed sheep: gukkal and udu kun-gid. The gukkal, which is first documented in Babylonia in the Uruk I11 period,31 is very common in Ilrehem documentation. There, the term g ~ k k a letymologically ,~~ "big tail" (*kun-gal),appears to describe specifically a sheep with a short, broad fat-tail.33 The udu kun-gid, a "long-tail", which is considerably rarer than the gukkal at D ~ - e h e mis, ~in~all likelihood the variety with a long fat-tail.35 This type of sheep, otherwise unknown in Mesopotamian documentation, is probably identical with the udu-gukkal-ib-18= zib-ba-nu of tlh. XI11 26,36 which, according to the Akkadian translation, was a gukkal with an extra long tail. An alternative, but less likely, identification might be udu-gukkal-TU = gu-uk-kal-la-nu, listed in Hh. XI11 27,37 which probably denotes the gukkal with an extra big or fat tail. It appears that yet another term for the fat-tailed sheep is udu gal-tab-bu-um. In contrast to the gukkal and udu kun-gid, this type of sheep is extremely rare, as it is found in only five published documents.38 The possibility that udu gal-tab-bu-urn was a type of fat-tailed sheep is suggested by the text FAOS 16 1052, where, in three separate sections, udu gal-tab-bu-um are grouped together with the udu kun-gid and the gukka1.39 It is also significant that, in two other documents, udu gal-tab-bu-urn are associated with the udu kun-gid and gukkal, respectively.40 Given the great rarity of this type of sheep, it is possibly the "fat-tail" with an extra big tail. For this possibility, note further that Hh. XI11 names two specific varieties of the gukkal, called zibbanu and gukkallanu (see above). If, as I suggested earlier, udu kun-gid is the zibbanu, then udu gal-tab-bu-um is perhaps to be connected with gukkallanu = udu-gukkalT U . ~ ~ The term gal-tab-bu-urn, which appears to be a Sumerian loanword, may conceivably be identical with kaltappu, a variant form of kilzappu, "footstool", also "threshing board."42 The problem with this solution, however, is that the form kaltappu is not documented before NeoAssyrian times.43 No less important, it is difficult to think of a reason as to why a fat-tailed sheep should be associated with a footstool or a threshing board.44
Steinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
3.1.3 udu aslum,(A.LUM),"long-fleeced sheep" The breed called udu A.LUM is very common in Drehem sources. Outside Drehem, this breed is extremely rare, as it is attested in only two other Ur 111texts: one from Umma (TLB 3 44: 1,3) and one from Lagash (Barton, H a v e ~ o r d2, pl. 75 HLC 57 ii 5 , l l , iii 3, 12, iv 1, 4).45 Although udu A.LUM does not appear in economic tablets before Ur III times, the term is attested already in an Early Dynastic list of animals, preserved in manuscripts from Abu Salabikh and Ebla.46 Following the Ur I11 period, udu A.LUM is documented in the early OB texts from Isin$7 at Old Babylonian Mari, and in the inventories from Qatna, as well as in Sumerian literary texts.48 There are no clear attestations of this term following OBMB times.49 The term udu A L U M is listed in the forerunners to Hh. XIII, where it appears with the qualifications niga, niga sigg, and gukkal (MSL 811, p. 83 lines 8-11). In the standard recension of Hh. XIII, udu A.LUM is replaced by udu asq(LAGABxA)-lum,with a translation pasillu: udu L A G A B X A ~ - ~ ~ L=pa-sil-lum UM (Hh. XIII 12; see also lines 13-15).50 In the Hg. commentary to Hh. XIII, pasillu is explained as immeru, "sheep", and udu L A G ~ X A . L U M is further glossed sulul~u/ sulum~ii,"long-fleeced sheep": udu aSLAGABxA.[LUM] = pa-si-i[l-lum] = [im-me-ru], udu S ~ - ~ ~ L A G A B X A ~ = ~ .su-lum-ku-ri ~UM] = MIN (MSL 811, p. 54 A II 235-36). Based on the writing asq-lum of Hh. XIII, and the fact that this writing replaces the earlier A.LUM, one must conclude that ALUM is to be read aslum,. aslum,, in turn, must be identical with Akkadian a ~ l ua,well-known ~ ~ term for "sheep."52 Since ALUM is first attested in an ED lexical text of undoubtedly northern Babylonian ("Kishite") background (see above), this term is very likely an abbreviated / defective Akkadogram, similar examples of which are welldocumented in the ED script of northern Babylonia.53 As can be concluded from the fact that it was a rare word already during OB times, aslu had ceased to be a viable breed-designation sometime in the Isin-Larsa period, that function having been taken over by pasillu (see Hh. XIII 12-15).54 Turning now to the question of the identity of this breed, it appears virtually certain that aslum, (and likewise the Akkadian aslu) is a word for the "long-fleeced" or "hairy" sheep, usually identified by the term SU~@U(S~G.G~D), Akk.sulumhii.55 Among the evidence pointing to such a conclusion, we must mention, first, the aforementioned Hg. commentary to Hh. XIII, which provides AS4.LUM with the gloss lsulu~ul. Second, in the leather texts from Isin skins of udu ALUM are consistently distinguished from those of ordinary sheep.56This shows that there was something distinctive about the skin of the aslum,, and, since that characteristic was not its color,57 it must have been long hair that set it apart from ordinary sheepskin.58 And, in fact, skins of aslum, sheep are often described as "hairy" (siki-mu) in the very same sources.59 Third, in the composition Curse of Akkade, line 23, the aslum, sheep is described as "longfleeced": ur-girls ur-nim darab kur-ra udu A.LUM suluhu(~fG.Gf~) si, "dogs, lions, mountain bezoars, the a l u m sheep richly provided (si = malii) with long fleece."60 And last but not least, the identification of aslum, as the "long-fleeced" sheep is strongly indicated by the very absence of that breed in Drehem sources. Considering that these employ the most extensive taxonomy of sheep breeds known from ancient Mesopotamia, it is precisely here that some mention of this breed is expected to be found.
Steinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
3.1.4 Shimashkian sheep The sheep known as udu S i m a ~ k i ( L u . ~ ~ ( . A which ~ ) ) , is ~ ~exceedingly common at Drehem, represents a domestic breed that was indigenous to the lands of Shimashki. As we now know,62 those lands comprised the whole western portion of the Persian plateau, extending from the Caspian Sea in the north to the borders of Fars in the south. The term udu Simagki is found exclusively in the Ur 111documentation from Drehem. The only exception is the occurrence of this term in the Lagash text Barton, Haver$ord 2 pl. 75 HLC 57.iii.17, iv. 1, 8. As noted ear lie^-,63 the use of Drehem terminology in that document is due to its recording deliveries of livestock from Drehem and Ur. 3.1.5 Lullubean sheep Another certain example of a foreign breed is the "Lullubean" sheep. This breed is frequent at Drehem, although not as common as the Shimashkian sheep. It is usually identified by the term udu L ~ I I ~ ~ , ( L u . u L uwhich, - ~ ~ ) ,with one exception (Barton, Haverford 2 pl. 75 HLC 57.iii.14, iv. l , 5 - L a g a ~ h )is, ~attested ~ exclusively at Ur 111Drehem. In two Drehem sources, the same breed is identified as udu Lu-lu-buythe "Lullubean" sheep (AUCT 1 405; BIN 3 350).65 As indicated by its name, this breed was native to the country of Lullubum, situated in the Zagros ranges, probably in the area of present-day Kurdistan.66 I assume that L U . U L U - U-~to be read L U I I U ~ , ( L U . U L Uor - ULu-ulu-um ~) - is a fancy writing of Lullum,67 a byform of Lullubum, which is documented in the Old Babylonian texts from Shemshara, Mari, and Tall Rimah, as well as in later sources.68 There is independent evidence that this variant was known and used already in Ur 111times.@ 3.1.6 Sharurnijum sheep This breed, which is identified by the term udu Sa-ru-rni-~m,~~ is documented exclusively at Drehem. With the exception of the udu gal-tab-bu-um, it is the rarest of the breeds documented there.71 Since Sharurnijum is a gentilic formation, we apparently find here yet another foreign breed. It is troubling, however, that neither Ur 111 nor any other third millennium sources record any toponym from which the form Sharumijum could plausibly be derived. To my knowledge, the only toponym that would permit such a derivation is Sa-ru(?)-maki, which is uniquely attested in an Old Babylonian geographical list from Tell Harmal (MSL 11, p. 58 iv 122). 3.1.7 To conclude this part of my discussion, the following summary may be offered. If the suggested identifications are correct, eight specific breeds of sheep can be identified in Drehem sources: a) the native breed (generic udu / udu Ki-en-gir,); b) three breeds of fat-tailed sheep: the short broad tail - gukkal, the long tail - kun-gid, and the extra big tail - gal-tab-bu-um; c) the long-fleeced sheep (aslum,); d) three foreign breeds: the Shimashkian or Persian sheep, the Lullubean or Kurdistani sheep, and the Sharurnijum sheep.
Steinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
The breed terminology used outside Drehem is much simpler. Thus, at Lagash, sources differentiate only between "native" (udu or udu-gir, or udu-eme-gir,) and "fat-tailed" (udugukkal) ~ h e e ~At . 7Ur, ~ the distinction is between "native" sheep (udu-eme-gir,), "fat-tails" (udugukkal), and "fat-tails of the Upper Country" (udu-gukkal-igi-nim-ma).73 And at Umrna, sheep the latter are classified as either "native" (udu-eme-gir,) or "foreign/mountain" (udu-k~r-ra)?~ i ~ indicated ) . ~ ~ by further subclassified as either "fat-tailed7'(udu-gukkal) or "black" ( ~ d u - ~ As these data, outside Drehem one distinguished basically only between native and fat-tailed sheep, with "fat-tail" being a virtual synonym for "foreign" breeds. This contrasts sharply with Drehem practice, where fat-tails (gukkd, kun-gid, and gal-tab-bu-um) are always distinguished from foreign sheep (Shimashkian, Lullubean, and Sharumijum), with no instances of the former included or subsumed among the latter.
3.2 Goats In regard to breeds of domestic goat, the picture is considerably simpler. Only two such breeds are documented at Drehem: the "native" (generic mag) and the "Shimashkian" (mag Sima~ki),76 the latter being very common. Outside Drehem, only the native breed is attested.77
3.3 Hybrids As noted above (see under $2), Drehem sources mention two types of caprine hybrids: a hybrid and a hybrid of the domestic goat of the domestic sheep with the mouflon (udu-a-udu-hur-~a~),~~ with the bezoar ( m i l - a - d a r ~ d a r a ~Both ) . ~ ~hybrids are very common at Drehem. The terms audu-hur-sag, "mouflon seed", and a-dara?l/dara4,"bezoar seed", are limited to Drehem sources. We otherwise find no references to either hybrid in cuneiform documentation. Yet another example of a caprine hybrid, uniquely attested in CMET 7 72:7, is possibly udu-a-gukkal, "sheep (with) the seed of the fat-tailed sheep."
4 Sex and Age 4.1 The basic terminological distinction is between adult and young animals. The former are identified as udu (nita), "ram", ug, "ewe", ma6 (nita) or mag-gal (nita), "buck, and UZ,~O "nanny goat." Unless qualified by the designation gig-dii, "breeder" (see under $4.2), udu (nita) and mi6 (nita) 1mag-gal (nita) invariably describe castrated a n i m a l ~ .The ~ l terms for the young are sua4, "male lamb", kirll, "female lamb", mi& "male kid", and "iri~-~hr, "female kid." The young are further subclassified as ga, "suckling",82 gaba, and gub, the latter two describing weaned or semi-weaned animals. The only terms warranting discussion here are gaba and gub. In his contribution to this volume, Heimpel considers gaba and gub to be two separate terms. As understood by him, gaba, "Brust", describes young animals that were small enough to be clutched against one's breast,83 whereas gub, "Bestand", signifies animals that were brought for their first shearing and "werden gleichzeitig zum ersten Mal im Bestand gezZhlt."84 However, since gaba never occurs together with gub,85and since both terms describe animals older than ga, "suckling", but younger than fully mature, and thus refer to the same agecategory, I am inclined to assume, following Landsberger's original suggestion,86 that these are variant spellings of the same word, with gub being a syllabic realization of gab(a).
S Ieinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
However, regardless of whether gub is the same as gaba, the interpretation of gaba presents certain difficulties. This term, lit. "breast", is almost certainly an abbreviation of the descriptions dla4 gaba ug and m h gaba iiz, which are recorded in an ED lexical source.87 Accordingly, gaba means "(animal at the) breast (of its mother)", a further corroboration of which is provided by Hh. XIII 84, where udu-gaba is translated im-mer ir-ti, "sheep of the breast." However, such a translation obviously clashes with the commonly-held view that gaba describes fully weaned a n i m a l ~ . ~On 8 the other hand, the possibility that gaba could be an alternative term for "suckling" (as in well-known (dumu-)gaba"suckling infant") is precluded by examples in which this term appears together with As a possible solution to this dilemma, I propose that gaba (and similarly gub) serves to describe semi-weaned animals. As is practiced by some pastoralists, the young over onemonth-old are separated from the ewes and grazed in a separate flock, though they are still allowed to suck the ewes once or twice a day, till the weaning is completed.90 In support of this interpretation, note that the animals classified as gaba or gub are never qualified as niga, "barleyfed", or u, "grass-fed", which suggests that milk was still the primary source of their nutrition.
4.2 In addition, Drehem sources use the following ageisex related terms: a) h-tu-da, "newborn"; b) gig-dii, "breeder";91 c) en-zi, "lead animal";92 d) us sila4 nir/nu-a1hz miS nWnu-a, "pregnant ewe / nanny goat";93 e) ug silaq dii-a I iiz ma6 dii-a, "pregnant ewe / nanny Rejecting the long-accepted interpretation of nsnu-a sila4/m6S as "pregnant (with) a lamb/kidm,95Heimpel now proposes the translation "infertile", taking nulnu-a to mean "without."96 In my view, this interpretation is extremely unlikely, primarily because this description is very common at Drehem. By contrast, ug silaq dii-a / iiz m6S dii-a, the other known term for "pregnant", is attested in only two Drehem texts (see above). If the former expression were to mean "infertile",97 we would be forced to conclude that, at Drehem, infertile animals were regularly identified (by what means and for what purpose?), whereas pregnant ones were not. This I find impossible to accept. As for Heimpel's objection that it is difficult to think of any rationale for "slaughtering" pregnant animals, it should be pointed out that, at Drehem, the term ba-liS means not only "slaughtered", but also "dead (of n.atural causes)." As for silaq/maiS dh-a, this term is probably to be analyzed as "planted (dii-a = zaqpu) (with) a lambkid." Accordingly, its specific meaning seems to be "impregnated."
5 Various physical characteristics 5.1 Color Five color designations are attested: babbar?8 "whiteV,99816, "black",100 si4, "brown, red",lol sig17(GI), "yellow, tanm,102and giin-a, "spotted, mottled."l03 With the exception of "brown, red", which, as far as I know, is attested only for goats,lo4 these descriptions are documented for sheep and goats alike.
Steinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
Among these terms, only sigl7(GI),"yellow, tan", requires further comment. This term has uniformly been read as gi by scholars, with the translation "(fed with) reeds."l05 However, that interpretation is obviously erroneous, as shown by the examples where GI is combined with such qualifications as niga, "barley-fed",l06 and gay "suckling",lO7 which make it absolutely clear that GI cannot describe a type of feed. On the other hand, yellow or tan color is given as a description of both sheep and goats in Hh. XI11 (see above), and, since sig17 is otherwise attested as a graph of /sig/ "yellow, greenV,l08 there can be no doubt that this is how GI should be read in this context. 5.2 Condition of fleece I goat-hair Drehem texts use four terms to describe the condition of the sheep's fleece: bar-gal, "with fleece",l09 bar-su-ga, "without fleeceWy1 lo sila4/kirll AS(=dili?)-ur4,"lamb plucked for the first time9',lll and si-2-16, "plucked twice."l12 The first three are frequent in contemporaneous sources from other places. The fourth term, si-2-laYwhich is also attested for goats, occurs only at Drehem.
5.3 Physical deformities Finally, we find a number of terms describing various physical deformities: a) ba-za, "cripp1ed;l l3 b) gu-gu-turn;l14 c ) umbin TAB .TAB ;I d) G U R ~G. U R ~ . Of these, only the meaning of ba-za, "crippled", is beyond question. Assuming, as suggested by CAD G, p. 123a, that gu-gu-turn, which is attested exclusively of lambs, is a loanword from Sumerian kud-kud-du, Akk. /jurnrnuru,l l7 its meaning would be "crippled" or "deformed." More likely, however, gu-gu-turn is a type of meal or a ritual designation. This is suggested by the text MVN 2 1595, which records 1 udu niga mu udu mu-du-lum uzu-had ii silaq gu-gu-turn-ma&, "1 barley-fed sheep (earmarked) for salted (and) dried sheep and the gugu-turn lambs." The meaning of the term umbin TAB.TAB, which is found only in connection with goats, is , ~ ~ of ~ somewhat obscure. If translated literally, umbin TAB.TAB would mean " f ~ u r - h o o f e d " but, course, such a description would be meaningless in reference to quadrupeds. One possibility is that umbin is to be understood here not as "hoof' but as "digit" or "toe." In that case the term could describe a goat with the overdeveloped remnants of its back toes. A more plausible solution, however, is that TAB.TAB is to be analyzed as an early graph of adq, i.e., ZA-tenli or EREN-tend,Akk. kubbulu, 'lame."l l9 Accordingly, umbin TAB.TAB would mean "(goat) with a lame hoof." The term GUR8.GUR8, which is documented also of cattle,120 is even more difficult. Possibly, it is to be connected with [udu-gur4-gur4]= kur-kur-ra-nu (Hh. XIII 48); udu-gur4gurq and [uldu-gur-gur in Forerunners (MSL 811, p. 83 line 33, p. 91 line 3'). Assuming that kurkurrgnu derives from kurkurru, a type of bulky jarY121this term could describe an animal with a swollen or enlarged belly, and should be translated "pot-bellied" or the like.
Stcinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
Grades of quality 'I'he following terms grade sheep and goats according to type of feed and feeding rates: a) niga sigg, "barley-fed, top quality (first grade);122 b) niga sigg-us, "barley-fed, top quality, next grade"; c) niga 3-kam-us, "barley-fed, third grade"; d) niga 4-kam-us, "barley-fed, fourth grade"; e) niga, "barle~-fed";l~~ f) niga gud-e-us-sa,"barley-fed, 'following the oxen"'; g) u, "grass-fed";124
6
The basic distinction is between barley-fed and grass-fed animals. The former are further subdivided into six categories, which can be grouped as follows: four superior categories (a, b, c, and d), the standard category (e), and the inferior category (f). The sequence of these categories is assured by texts such as BIN 3 239 and 240, where all seven categories appear, always in the same order. Except for gud-e-us-saythe meaning of these terms is clear. Contrary to the common belief that gud-e-us-sa designates sheep and goats of the highest quality, whose value approximated or was closest to that of cattle,l25 a systematic examination of this problem shows that, in fact, it was the least valued category of barley-fed animal~.l2~ If, as seems likely, gud-e-us-sa is to be taken literally as "(an animal) that follows oxen", the term probably describes sheep and goats that were fed with barley left over from cattle feeding. It is certain that this six-tier classification of barley-fed animals is based primarily (if not exclusively) on the daily rates of feeding. Such feeding rates are, of course, well-documented in sources from Umma, Lagash, and other Ur I11 centers. For this conclusion, see, especially, MVN 15 64 i 1-3 (Drehedummal), which records 111udu niga sigg at the rate of 1112 sila of barley per sheep, followed by 196 udu niga at 1 sila, and 90 udu at 112 sila. Cf. also UET 3 1197 (Ur), recording udu niga sig5 and udu niga <sig5->usat 1112 sila of barley, and udu niga at 1 sila.
7 Wool and goat-hair The following are designations of wool and goat-hair found in Drehem sources: a) siki gir, (= siki udu-eme-gir,), "wool of the native sheepW;127 b) siki aslum,(A.LUM) (= siki udu aslum,), "wool of the long-fleeced sheepn;128 c) siki udu S i m a g k i ( ~ O . ~"wool ~ ) ~ ~of, the Shimashkian sheep"; 129 d) siki kur(-ra) (= siki udu-kur-ra), "wool of the foreign/mountain sheepm;130 e) siki gi6 (= siki U~U-gi6), "wool of black sheepW;l31 f) siki gir-gul, "wool (plucked with) the aritu-knife" or, less likely, "wool (of the sheep slaughtered with) the arttu-knife7';132 g) siki mug, "wool of poorlinferior quality7';133 h) siki i i ~"goat-hair."l34 ,
Steinkeller
Sheep and goat terminology in UI I11 sources from Drehem
The Drehem terminology for wool and goat-hair comes from the texts that originated in a separate office of the Drehem establishment, whose operations appear to have been largely independent of the animal office. It is characteristic that this terminology is essentially the same as the one used at Umma, although it is also influenced by the breed designations employed at the Drehem animal office. The dependence of Drehem wool terminology on Umma is reflected in its adoption of the terms gir, (for udu-erne-girls), kur(-ra) (for udu-kur-ra), and gi6 (for udugi6), which are basic classificatory terms for sheep in Umma sources (see above under §3.1.7).135 On the other hand, influence of the Drehem animal terminology is evidenced in the occasional use of the terms siki aslum, and siM udu Gima~ki,two breeds that are documented only in Drehem sources.
Slcinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
TABLES OF TERMINOLOGY
A. SPECIES a. udu - domestic sheep (Ovis aries). b. udu-hur-sag - (Western) Asiatic.mouflon (Ovis orientalis); secondarily may also describe the urial (Ovis vignei). c. mi6 - domestic goat (Capra hircus). d. darawdara4 - bezoar (Capra aegagrus); secondarily may also describe the ibex (Capra ibex) and the markhor (Capra falconeri). ( ~ ~ ) goat; probably to be identified as the Arabian white e. m G ~ i - ~ a n- "Makkan" oryx (Oryx leucoryx).
13. BREEDS
a. Breeds of domestic sheep 1. udu I udu-Ki-en-girx(GI) - native breed. 2. gukkal - fat-tailed sheep (short broad tail). 3. udu kun-gid - fat-tailed sheep (long tail). 4. udu gal-tab-bu-um - fat-tailed sheep (extra big tail). 5. udu aslumx(A.LUM) - long-fleeced sheep. 6. udu S i m a ~ k i ( ~ u . ~ -~Shimashkian ( . ~ ~ ~ ) ) sheep. 7. udu ~ u l l u m ~ ( ~ ~ .I ~Lu-lu-bu ~ - u m - Lullubean ) sheep. 8. udu Sa-ru-mi-um - Sharumijum sheep. b. Breeds of domestic goat 1. m6S - native breed. 2. mi5 S i m a ~ k i ( I , o . ~ ~ (- .Shimashkian ~~~)) goat.
C. HYBRIDS a. udu-a-udu-bur-sag - hybrid of the domestic sheep with the Asiatic mouflon. b. m8-a-daraWdara4 - hybrid of the domestic goat with the bezoar. c. udu-a-gukkal - hybrid of the domestic sheep with the fat-tailed sheep.
Steinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
D. SEX AND AGE Adult sheep goat
Male udu(-nita)
Female u8
sila4
miis(-nita) m6H-gal(-nita)
uz
m6H
Young Male Female kin1 ga gabalgub mi66-ghr ga gabalgub
ga gabdgub ga gabdgub
a. a-tu-da - newborn. b. gig-dii - breeder. c. en-zi - lead animal. d. ug sila4 nunu-a 1 hz m6g nunu-a - pregnant ewe 1 nanny goat. e. ug sila4 du-a 1 iiz m6E du-a - pregnantlimpregnated ewe 1nanny goat.
E. VARIOUS PHYSICAL CHARACTERISTICS a. Color 1. babbar 1 barg-bar6 - white. 2. gi6 - black. 3. siq - brown, red. 4. sigl7(GI) - yellow, tan. 5. gun-a - spotted, mottled. b. Condition of fleece / goat-hair 1. bar-gail - with fleece. 2. bar-su-ga - without fleece. 3. sila4 / kirll ~ S - u- rlamb ~ plucked for the first time. 4. si-2-16 - plucked twice. c. Physical deformities 1. ba-za - crippled. 2. gu-gu-turn - ... 3. umbin TABTAB - lame(?). 4. GUR8.GUR8 - potbellied(?).
suckling (semi-)weaned suckling (semi-)weaned
1
Sheep and goat terminology in Ur I11 sources from Drehem
St cinkeller
I:. GRADES OF QUALITY a. niga sig5 - barley-fed, top quality (first grade). b. niga sigg-us - barley-fed, top quality, next grade. c. niga 3-kam-us - barley-fed, third grade. d. niga 4-kam-us - barley-fed, fourth grade. e. niga - barley-fed. f. niga gud-e-us-sa - barley-fed, "following the oxen." g. - grass-fed.
G . WOOL AND GOAT-HAIR a. siki &,(GI) - wool of the native sheep. b. siki aslumx(A.LUM) - wool of the long-fleeced sheep. ~ ) ~of~ the Shimashkian sheep. c. s k i udu S i m a H k i ( ~ f i .-~wool d. siki kur(-ra) - wool of the foreignlmountain sheep. e. siki gi6 - wool of the black sheep. f. siki gir-gul - wool (plucked with) the aritu-knife. g. siki mug - wool of poor/inferior quality. h. siki iiz - goat-hair.
NOTES Abbreviations used are those of the Chicago Assyrian Distionary and the Philadelphia Sumerian Dictionary with the following additions: BAOM
Bulletin of the Ancient Orient Museum (Tokyo)
CMET
Catalogo del Museo Egizio di Torino (TurinJMilan)
Gomi-Sato, NSAT
T. Gorni and S. Sato, Selected Neo-Sumerian Administrative Texts from the British Museum (Tokyo, 1990)
Hirose Coll.
T. Gomi, Y. Hirose, and K. Hirose, Neo-Sumerian Administrative Texts of the Hirose Collection (Potomac, MD, 1990)
Lafont-Yildiz. TCT
B. Lafont and F. Yildiz, Tablettes cune'ifonnes de Tello au Muse'e d'lstanbul (Leiden, 1989)
Lau, OBTC
R. J. Lau, Old Babylonian Temple Records (New York, 1906) H. Limet, Textes sume'riens de la Ille Dynastie d'Ur (Brussels, 1976) I . Spar, Cuneiform Texts in the Metropolitan Museum of Art, vol. 1 (New York,
Limet, TSDU Metropolitan Museum
1988) SACT
Sumerian and Akkadian Cuneiform Texts in the Collection of the World Heritage Museum of the University of Illinois (Urbana)
S teinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
See, in general, T. B. Jones and J. W. Snyder, Sumerian Economic Texts from the Third Ur Dynasty (Minneapolis, 1961), pp. 212-38; G. Buccellati, The Amorites of the Ur 111 Period (Naples, 1966), pp. 274-302; Steinkeller in The Organization of Power: Aspects of Bureaucracy in the Ancient Near East, ed. by McG. Gibson and R. D. Biggs (Chicago, 1987), pp. 28-40; M. Sigrist, Drehem (Bethesda, 1992).
Slcinkeller 15
For excellent representations of the oryx, see Boehmer, Entwicklung, pl. 15 fig. 165; D. Collon, First Impressions: Cylinder Seals in the Ancient Near East (Chicago, 1987), p. 189 fig. 912.
16
As suggested by W. I-Ieimpel, 2.4 77 (1987) 64.
17
See P. N. Munton, ''The Ecology of the Arabian Tahr (Hemitragus jayakari Thomas 1894) and a Strategy for the Conservation of the Species", Journal of Oman Studies 8 (1985) 11-48; Uerpmann, Ungulate Mammals, pp. 111-12. In my opinion, this animal is too ordinary to have been imported to Babylonia. Uerpmam, op. cit., pp. 126-27 n. 9, cites a certain H Kauch, animal dealer at Dubai, who, though familiar with various types of wild caprinae and gazelles, "never received any thars, either dead or alive, nor has seen them in captivity. He thinks that the thar is considered as too ordinary and too similar to a domestic goat to receive much attention.''
18
For practical reasons, I retain the traditional reading Beg9-bar. However, it should be pointed out that the correct reading is Begbarx(S~~9)bar. I hope to discuss S E G ~ (= LAK-263) and related signs in a separate study. AHw., p. 1027a: 'Wildbock"; CAD S, p. 166: "wild sheep" in the occurreuces from Etana Epic and in bilingual examples, but, inexplicably, "a bovid", in the heading.
See Heimpel, BSA 7, 131-33. E.g., [x] 'udul niga mu-du-lum-Sb, "[XI grain-fed sheep for salting" (AUCT 2 338:4). See CAD Ml2, p. 162b. E.g., 1 silaq uzu-h6d-Sb, "1 lamb for dried meat" (AUCT 1 95:l). E.g., 1sila4 ga uzu-a-bala, "1 suckling lamb (for) meat broth" (AUCT 1 441:14). uzu-a-bala = ummar me^Jiri (CAD M12, p. 155). See [tu7]-a-uzu = rum-mar7me-e k-e-ri (MSL 11, p. 69 frg. c 1); tu7-uzubala, tu7-a-uzu (MSL 11 p. 152 lines 29-30); tu7-uzu-bala = SU-u [...I, tu7-im-me-uzu-bala = SU-u [...I (MSL 11, p. 89 Hg. B, VI 103-04). E.g., 9 sila4 ga ka-izi-R, "9 suckling lambs for a roast" (BIN 3 369: 1-2). See K A ~ ~ - ~ = - ~MIN ~ N(=E krrziifu) Zci NE (Antagal VIII 112 = MSL 17, p. 173); [uzul-ka-izi = Ji-ir Su-me-e, [uzu-kla-izi = pu-ut-tu-u' (Hh. XV 260-61 = MSL 9, pp. 13-14); uzu-ka-izi = :[urn-mu-u], uzu-ka-izi-tab = [MINI (Nabnitu XXIII 172, 175 = MSL 16, pp. 216-17).
I9 20
Thus we find the Beg9-bar listed: following bovids and deer and before equids (RA 63 [I9691 102 A 0 19548, 104 A 0 19550 = FAOS 16 1359); following bovids and before equids (FAOS 16 958); following bovids and deer, and before sheep (C'I' 32 pl. 14); followed by equids and sheep (CT 32 pls. 28-29 viii).
21
E.g., 1 silaq ga ne-mur-ta ba-Segg, "roasted in hot embers" (BIN 3 74:l-2). neldb-mur = tumru, p2mtu; Seg6 = ba3iilu. Cf. 45 sheep and goats gird-ta ba-Seg6, "roasted in an oven" (SACT 1 171:l-4). See H.-P. Uerpmann, The Ancient Distribution of Ungulate Mammals in the Middle East (Wiesbaden, 1987), pp. 126-27; J. Clutton-Brock, A Natural History of Domesticated Mammals (Austin, 1987), p. 53. For a good representation of the mufflon, see E. Porada in Investigating Artistic Environments in the Ancient Near Eust, ed. by A. C. Gunter (Washington, DC, 1990), p. 76 figs. 7 and 8.
Sheep and goat terminology in Ur I11 sources from Drehem
See Steinkeller, SEL 6 (1989) 3-4 and n. 5.
22
See, e.g., 1 amar Beg9-bar nita mu 1 (PDT 526 ii 4); 1 amar Begybar mi mu 1 (ibid iii 9); 2 Beg9bar nita mu 1 (BIN 3 234:4); 2 Segg-bar nita mu 2 , 3 Segg-bar nita mu 1 (ASJ 12 [I9901 41 no. lO:5-6). See, e.g., AUCT 2 294:2-3; PDT 526 passim.
See under 53.3.
23
With lulim: PDT 40; with lulim and maBda: BIN 3 234,260; with rnaSda: BIN 3 218; AUCT 2 294.
For the ibex, see Porada in Artistic Environments, p. 78 fig. 9; R. M. Boehmer, Die Entwicklung der Glyptik wahrend der Akkad-Zeit (Berlin, 1965), pl. 6 fig. 54; for the markhor, see Boehmer, op. cit., pl. 14 fig. 153. Due to its great similarity to the moufflon, the urial is more difficult to identify; a possible representation is found in Boehmer, op.cit., pl. 16 fig. 175.
24
This is suggested by the following literary passage: Begg-bar-resi-mui-bi nam-ta-an-i?,"the S. raiszs i sticks out its ... " (TCL 16 89:14), where si-mug, lit. "serpentine/coiled horns", is very likely a term for "antlers". For the sense "serpentinc, coiled" of mug, cf. (id-)a-muB, "serpentine water (course)", which describes small meandering canals in Ur 111 documentation from Urnma (H. Sauren, Topographie, pp. 44-45). In another source, the horns of the Segg-bar are described by the term a: 2 6 Begbarx(S~G9)(DP 490 v 2 - Pre-Sargonic). While si is a general word for animal horns, 8, or more commonly, a-darah (Akk. adrii), denotes the elaborate horns of wild goats and deer antlers. See ii-darahq6-ar-nu-umrfa(7)' a-[a-lim] (Proto-Izi I1 20; cited in CAD Q, p. 134b); 53 6-darah, followed by 96 si am and 90 si maBda (NATN 855:1-3 - Ur 111).
25
For this species, see Uerpmann, Ungulate Mammals, pp. 58-63.
26
See ibid., p. 58: "The Mesopotamian fallow deer shows a different form of the basal part of mature antlers, which only have a small brow tine immediately followed by a larger second tine, flattened at its base and sometimes palmated itself. It extends into a palmation along the whole beam, which gives the stags of this species a completely different appearance."
27
For ancient representations of the fallow deer, see Collon, First Impressions, p. 67 fig. 277, p. 189 fig. 915.
28
C an-Sk Beg9-bar ki-Bi?darah-mag, 6 an-Sb Seg9-bar-gim gun-a ki-Si?darah-mag-gim sig7-ga, "temple,
The attestations known to me are: msS M. (AUCT 1 786:3; PDT 130:1, CMET 7 361: 1; Hirose Coll. 28:1), m6L-gal M . (Hirose Coll. 75:2; Jones-Snyder 295). uz M. (SACT 1 94:2; FAOS 16 1347:4; Hirose 266:1), m6S ga M. (SACT 1 91:1), mi6S-g8r M. (AUCT 1 786:4), mi6S-ghr ga M. (Nesbit, Drehem IX:3). As demonstrated by examples in which the animals are qualified as ga, "suckling", this species could be raised in captivity. See, most recently, W. Heimpel, ZA 77 (1987) 30-31; D. T. Potts, The Arabian Gulf in Antiquity, vol. 1 (Oxford, 1990). pp. 133-50. A very important datum for this issue is provided by a recently-published Ur I11 tablet (Metropolitan Museum 17:84 - AS 4lvii/l0), which records the name of a ruler of Makkan: WA-du-um Ili-kin-gi4-a Na-du-be-li Cnsi ~ 6 - ~ aThe n ~fact. that Na-du-be-li I nadiib-'el-i 1 is an unmistakable Amorite name (see G. Buccellati, Amorites, p. 175), assures - in my opinion - that Makkan is to be sought on the southern coast of the Arabian Gulf. Here it may be highly significant that Mani'um, the name of the ruler of Makkan who was captured by Naram-Sin (Ma-ni-u[m] EN ~ c i - ~ a n ~ [ ~ ] - I. J . Gelb and B. Kienast, FAOS 7, p. 89 ii 4-6). is likewise documented as an Amorite name (Gelb, AS 21, p. 223, p. 621 no. 4525). See H. Jungius, "The Arabian Oryx: Its Distribution and Former Habitat in Oman and its Reintroduction", Journal of Oman Studies 8 (1985) 49-64; Uerpmann, Ungulate Mammals, pp. 80-83.
at (its) top a S., at (its) bottom a deer; temple, at (its) top spotted like a S., at (its) bottom yellow like a deer" (G. B. Gragg, TCS 3, p. 170).
Steinkeller 29
30
Sheep and goat terminology in Ur I11 sources from Drehem
Steinkeller
-
2586 udu Lu-lu-bu, 1051 udu Ki-en-girx (include 42 goats) (AUCT 1 405 S 46liv); 70 udu Lu-lubu, 902 udu Ki-en-gir, (include 234 goats) (BIN 3 350 - S 47lix) - both texts involve the Su-gid of URU.SAG.RIG7 and name the same officials; 15 sila4 ga Ki-en-girx, 1 6 kirll ga Ki-en-gir, (Buccellati, Arnorites no. 24:3'-4').
In the early Ur I11 sources from Lagash, the term is udu-girx; in the beginning of Amar-Sin's reign, it was replaced by udu-erne-girx. See the text Barton, Haverford 3 pl. 122 HLC 260, which tallies the native sheep that were available in the Lagash province from 3 32 to 3 47, or perhaps even later: 10,853 udu-girx @-a (i), 11,245 udu-gir, @-a (ii), and 22,653 udu-girx @-a (iv). A similar text, Reisner, Telloh 27, dated to AS 2, lists over 22,300 udu-erne-girx (@-a) (iv 10, v 5). See also sipad udu-girx (MVN 6 48: 1 [before Shulgi], 496:7 [before Shulgi]; ITT 4 7048:l [before Shulgi]; CT 1 6.iv.8 [S 431; Chiera, STA 17 i 8 [S 461; Lafont-Yildiz, TCT 741 edge line 7 [AS 271) and udu-ern[e-g]irx (LafontYildiz, TCT 620 viii 8 [SS 81); udu-erne-gir,-ra (I'IT 3 49152 [IS 21). The only exception to this rule is possibly found in Lambert, TEL 96:6 (transliteration only), which dates to IS 2, and appears to have sipad udu-gir,. At Umma and Ur, the term for "native breed of sheep" is consistently udu-erne-girx. See sipad uduerne-girx-ra-ke4-ne (Jones-Snyder 130:72; MVN 13 618 x 10; UET 3 1504 ii 26,1505 ii 39), udu-ernegir, (YOS 4 237:30,205), x GAN kud-r8 udu-erne-girx-ra (AnOr 1 303:4 - following x GAN kud-ra udu-gi6). For the meaning "native, domestic, indigenous", of girxlgirls (as in durnu-gir15/girx, "native sonldaughter", Ki-en-girx, "native land, erne-girls, "native tongue", and ur-girls, "domestic dog"), secondarily expressed also by erne-girls (as in sila erne-girls, "locallindigenous sila measure"), see my discussion in "Early Political Development in Mesopotamia and the Origins of the Sargonic Empire", in Akkad - The First World Empire: Structure, Ideology, and Traditions, ed. by M. Liverani (Padua, 1993), p. 110-11 n. 9.
In spite of the gloss EMEU-~~GI = [...I in Nabnitu XXIII 231 (MSL 16, p. 218), which led Th. Jacobsen, Studia Pedersen, p. 174 n. 10, to introduce a reading uli-gi-(r) for erne-girx, there is no doubt that the first sign is to be read erne. See sila me-girls in BIN 8 338: 1, which replaces sila erne-girls of other Sargonic sources (e.g., BIN 8 303:4, 306:3; 312:7, 320:1, 341:l). While the gloss u-li of Nabnitu XXIII 231 remains a mystery, it is quite certain that the lost Akkadian equivalent was Suweru or Sumeru. This follows from the fact that the respective section of Nabnitu lists such words as ier:erli, s'urrfi, ZZru, and iarierru. Cf. erne-girls = Su-we-rum in Sag iv 30 (MSL SS 1, p. 24). 31
See M. W. Green and H. J. Nissen, Zeichenliste der archaischen Texte aus Uruk (Berlin, 1987), p. 219 nos. 240-41. For other pre-Ur I11 attestations, see CAD G, p. 126a. Add the Sargonic examples CT 50 1251; OIP 14 143:4, 9, 10, rev. 7'; UET 2 Suppl. 10 i 4, iii 1. In lexical sources: gukkal (MEE 3, p. 67 nos. 18-25 line 52); u ~ u - ~ u - u ~ - ~ ~ ~ U =Dgu-uk-kal-lurn U.@L (Hh. XI11 23); the forerunners have UDU+@L (as in third mill. sources) (MSL 811, p. 8 n. to lines 23-25).
32
, rarely udu gukkal (as in TLB 3 The term for the male sheep is simply gukkal = U D U + ~ very 133:2). The ewe is usually identified as ug gukkal, sometimes as u ~ + (e.g., ~ LAnOr 7 17:9; AUCT 1 499:2, 3; MVN 13 60:6, 7), where U8 replaces the UDU of UDU+@L. The terms for male and female lamb are regularly sila4 gukkal and kirll gukkal, respectively.
33
For this breed, see M. L. Ryder, Sheep and Man (London: Duckworth, 1983), pp. 120, 148, 172, 246, 256, and passim.
34
Attested in, e.g., BIN 3 138:7, 209:1, 2; SACT 1 159:2; AUCT 1 530:4, 6; MVN 13 447:l; FAOS 16 1052 and passim in this source; BAOM 5 (1983) 31 no. 2:3. That udu kun-gid and gukkal are two distinct breeds is demonstrated by texts that list both types together: 1 sila4 gukkal, 1 ug kun-gid
Sheep and goat terminology in Ur 111 sources from Drehem
(BIN 3 138:6-7); 1 gukkal niga sig5, 1 gukkal, 1 kirll kun-gid (BAOM 5, p. 31 no. 2:l-3); and passim in FAOS 16 1052. 35 36
See Ryder, op. cit, pp. 120, 172, 263, and passim. udu-gukkal-ib-18-e in Nippur Forerunner (MSL 811, p. 83 line 15). udu-TU in Nippur Forerunner (MSL 811, p. 83 line 16). PDT 21:l (1 sila4 gal-tab-bu-urn gig-du), 644:9 (1 sila4 ga gal-tab-bu-urn); CMET 7 291:l (1 ug galtab-bu-urn niga), 305:6 (1 ug gal-tab-bu-um); FAOS 16 1052 iii 7 (36 sila4 gal-tab-bu-urn niga) and passim in this text, 1086:3 (1 udu gig-du gal-tab-bu-urn).
5 udu kun-gid gig-du, 1 udu gal-tab-bu-urn giS-du, 2 gukkal gig-dh (iii 19-21), 2 sila4 kun-gid, 6 sila4 gal-tab-bu-urn gun-a, 5 sila4 gukkal (iv 5-7), 3 kirll kun-gid, 3 kirll gal-tab-bu-urn gun-a, 5 kirll gukkal (iv 13-15). 1 US kun-gid niga, 1kirll gal-tab-bu-urn (FAOS 16 1067:2-3); 1 gukkal niga, 1u s gal-tab-bu-urn (CMET 7 3055-6). However, if ib-18 in udu-gukkal-ib-18 = zibbiinu is to be taken as "belt", and connected, following A. L. Oppenheim, JNES 4 (1945) 157, with the belts used to support the tails of fat-tailed sheep (see below n. 44), then a better match for udu gal-tab-bu-urn would probably be zibbiinu. See CAD K, pp. 361-63, under kilzappu (with variants kilizappu, kiltappu, kaltappu, galtappu, gis'tappu, kissappu, and gissappu). To be added is the form girtappu, appearing in ARM 19 104:1-2: 8 KUS UDU.GAL is' gir-tab-ba-tim,(DIN). For kilzappu as a threshing device, see most recently Steinkeller, Iraq 52 (1990) 20. kaltappu, a gold ornament of NB sources, and kaltappu, a writing for GI$.G~R.GUBin a NA royal inscription (CAD K, p. 362 c) and d)). Note, however, that AHw., p. 427a, considers the former item to be a separate lexeme - kaltappu, "ein Ggst. aus Gold." The only way such a connection could be made - and I offer this suggestion with due reservation - is to relate gal-tab-bu-urn to the "carts" that, according to the famous account of Herodotus, were used to support the tails of fat-tailed sheep in ancient Arabia: "One other thing is remarkable enough to deserve a mention - the sheep. There are two kinds, such as are found nowhere else: one kind has such longs tails not less than 4 112 feet - that if they were allowed to trail on the ground, they would develop sores from the constant friction; so to obviate this, the shepherds - who, fortunately, have sufficient knowledge of carpentry - make little carts and fix one of them under the tail of each sheep, to keep it clear of the ground. The other kind have flat tails, eighteen inches broad (Histories 3.115; translation after A. de Selincourt, Herodotus, The Histories [Harmondsworth: Penguin Books, 19541, p. 221). This account was repeated by Aristotle in Historia animaliurn 8.27. Although we lack evidence that would corroborate the existence of such "carts" in ancient times, the use of various devices, such as belts, nets, wooden rods, and even trolleys, to support the fat-tail is well-documented in the Middle East. For example, in his account of the Kazak fat-tails, C. D. Forde, Habitat, Economy and Society: A Geographical Introduction to Ethnology (London: Methuen, 1934), p. 340, writes: "The majority of the Kazak sheep belong to the fat-rumped breed, a large, rather goat-like variety derived from the wild Urial sheep of Asia ... [Its] most remarkable characteristic is the large double cushion of fat which covers and hangs from the buttocks, obscuring the short tail; in a grown sheep after good feeding this fat hangs down in two great ropes dragging on the ground, and may weigh over thirty pounds .... It is greatest in autumn when little wooden trolleys are often fitted to prevent the hanging cushions from dragging on the ground." Cf. also M. Hilzheimer, Antiquity 10 (1936) 202; A. L. Oppenheim, JNES 4 (1945) 157 n. 26. For representations, see Ryder, Sheep and Man, p. 230 fig. 5.15b, reproduced after An Illustrated History of the Sheep and Wool Industry, South African Wool Board (South Africa, 1970) [not accessible to me].
Stei~nkeller
45
Sheep and goat terminology in Ur 111 sources from Drehem
Stcinkeller
Sheep and goat terminology in Ur I11 sources from Drehem
Conceivably, then, gal-tab-bu-um, understood as either a "footstool" or, perhaps more likely, a "threshing board", might denote such a tail-supporting contraption.
Cf. J. S. Cooper's translation, The Curse of Agade (Baltimore, 1983), p. 51, "alu-sheep with long wool."
The same source also mentions Shimashkian (udu SimaSki) and Lullubean (udu L U . D L U - U ~sheep, ) which, like the u d u A.LUM, never appear in Lagash texts. The use of the distinctive Drehem terminology in this document is explained by its recording the deliveries of livestock from the distributive centers of Drehem (Nibruki-ta - ii 2, 12, iv 12, iv 14, vi 4) and Ur @rimki-ta - iii 5). Cf. Steinkeller, JAOS 108 (1988) 199 n. 15.
For the reading of LU.SU(.A) as SimaSki, see Steinkeller, JAOS 108 (1988) 197-202. The conclusive proof will be offered by M. Civil in a forthcoming note (to appear in NABU). See Steinkeller, NABU 1990113. See above n. 45. For this document, which records also Shimashkian sheep, see above n. 45.
46
A.LLJM, following gukkal and preceding udu-kur (MEE 3, p. 67 nos. 18-25 line 53).
47
kul udu A.LUM plus various designations of color and the ways of processing, passim in BIN 9 and 10. See also below nn. 56 and 59.
See H Klengel, RlA 713-4 (1988) 164-68.
48
CAD All, p. 374.
49
The alleged occurrences of *alu in Middle Assyrian sources - written e-lu and a-lu (ibid., p. 374b), almost certainly involve ajalu /jalu, "stag."
This writing is obviously based on the Akkadian lullfi, "man" (CAD L, p. 242a), itself a loanword from the Sumerian Iu-ulu (c*lu-lu), "men."
50
But note that the recension from Emar retains the writing udu A.LUM (Emar VU4, p. 107 lines 9'-12').
51
CAD A12, p. 336.
52
Once udu A.LUM is reclassified as aslu, then the lemma alu of CAD All, pp. 374-75, simply disappears, for its existence is not supported by any other evidence. Note that the examples of "udu-a-luhur-sag", cited ibid., p. 374b, represent actually udu-a-udu-hur-sag, for which see below under 93.3. Further, 1 GAL A.LU ZABAR of an OB Mari text, cited ibid., p. 374b, is a type of container, probably the same as alu D, ibid., p. 378b. For the alleged occurrences in MA sources, see above n. 49.
53
See, e.g., SU.ME for S U . ~ R . M Ecited , by W. G. Lambert, JCS 41 (1989) 27, and HA.ZI.URUDU (e.g., BIN 8 39 i 10) for @A.ZI.IN.URUDU.
54
Nora hene, in the modified spelling as4-lum, the rare value as4 = LAGABxA is very likely a (graphic) reminder of the archaic spelling with A. The use of asq-lum as a writing of aslu, a type of cubit measure (CAD Al2, pp. 336-37), is apparently a secondary development.
55
56
See Klengel, RlA 713-4, pp. 165-66. Add the following occurrences from Shemshara letters: LO Lu-ul-liirn (JCS 42 [I9901 138 SH. 913:10, 149 SII. 859+881:28) and LU LU-ul-li-imki (ibid.. p. 157 SH. 888:?7). See Lu-lu replacing Lu-lu-bi-im, discussed by Steinkeller, FAOS 17, p. 309 note to 1. 14. The alleged spelling Sar-ru-mi-um, occurring in Jones-Snyder 32:1, is almost certainly a misprint. The occurrences I can furnish are: 1 udu S. niga (PDT 1 596:3); 15 udu S. niga (BIN 3 539:ll); 1ug S. (FAOS 16 1168:5); 1 sila4 S. (Jones-Snyder 32:l); 1 sila4 S. (MVN 13 540:l); 1udu S., 1 udu S. gig-du, 4 US S. (AUCT 1 301:4-6 = StOr 9 no. 26); 2 udu 1 sila4 1ug S. (AUCT 2 366 i 7); 2 udu S. gig-du (FAOS 16 1052 iii 25), 5 sila4 S. (ibid. iv lo), 13 kirll S. (ibid. iv 18); 1 udu g. gig-du, 4 ug S. (Fish, Catalogue 295:7-8). See Reisner, Telloh 27; Lau, OBTC 161; ITT 4 7048:l-2 (a-Sag4 dabs-ba sipad udu-gir,, dabs-ba sipad udu-gukkal).
a-Sagq
See UET 3 1504 ii 23, 26, iii 7, 1505 ii 35, 39, iii 17.
Attested already in ED texts: AN-ur-Sb udu-sululp~,"A., (one of) the suluhu-sheep" (WVDOG 45 18 xvi); 1 udu-suluhu (TuM 5 88 i 1). This breed is not mentioned in Ur I11 texts, except for the Lagash PN Lugal-suluhu (H. Limet, Anthroponymie, p. 472; MVN 6, p. 356; 7, p. 180; Gomi-Sato, SNAT 154:3; etc.). However, it does occur in OB texts from Ur: UDU.NITA S ~ G . G (UET ~ 5 111:17, 817:7 courtesy M. Van de Mieroop). In lexical sources: u d u - s f ~ . ~ U(MSL D 811, p. 84 line 57 - Nippur S U D(Hh. XI11 16). Cf. Forerunner to Hh. XIII); U ~ U - S ~ G ~ U ' ~ U ' ~=USU-u = SU-u (= sikidu or the hke) (Hh. XI11 17), which indicates that yet another term for the "long-fleeced sheep" was udu-siki-du. However, this term is not otherwise documented in Mesopotamian sources.
See Jones-Snyder 130; MVM 13 618. The term udu-kur-ra appears already in an ED lexical text (MEE 3, p. 67 nos. 18-25 line 54 - written udu-kur). Cf. udu-kur-ra = im-me-ri 56-di-i in Hh. XI11 35. In economic sources, this term is documented from Sargonic times onwards (OIP 14 127:3, 8, 134:l [u8kur]; CT 50 62:l; HSS 10 171:2, 172 rev. 6'; RTC 244, 245, 246; etc.).
See BIN 9 281, 293, and 440, listing both kuS udu A.LUM babbar and kuS udu babbar, BIN 9 334 and 430, listing both kuS udu A.LUM gi6 and kul udu gi6, and BIN 10 128, listing both kuS udu A.LUM u-haib and kuS udu u-hab.
For the reading of L~.sU(.A)as SimaSki, see above n. 61. This breed is possibly identical with magkur, which appears in the Sargonic texts CT 50 125:3 and HSS 10 176:4.
57
See the preceding note.
58
Here note that the fleece of the long-haired sheep was treated as a distinctive type of fleece, which, like the sheep itself, was called suluhu in Sumerian and sulumhii in Akkadian. See S ~ G . = G su-lum-ku-li, ~ it-qum, "fleece" (Proto-Din 414-15), zu-lum-@ TUG.S~G.SUD= SU-u, it-qu (Din V 131-32) (both cited "5, shekels of in CAD IIJ, p. 299b); 5 GfN ga-bi-im a-na sa-ra-up 5 zu-lu-bi a: T ~ G . N ~ G . B A R A G 'alum' [black-dying agent - same as im-KUG.GI] to dye 5 sulumhii-fleeces for the u~siigarments" (ARM 23 142:l-3).
59
See above n. 29.
E.g., BIN 9 233:10, 334:3, 8.
See YOS 4 237, where the total of 4378 sheep is broken into 1939 udu-gukkal (1. 199), 2150 udueme-gir, (1. 205), and 289 udu-gi6 (1. 21 1). For udu-gi6, see also AnOr 1 303:3-4, which lists the areas of (pasture?) land assigned to the udu-gi6 and udu-eme-gir,. For the conclusion that, at Urnma, udukur-ra are virtually identical with udu-gukkal, see also IIeimpel, BSA 7, 137-8.
In this regard note that mag ~ u l l u m ~ ( ~ U . ~ ~ solely ~ - u r attested n ) , in the Lagash tablet Barton, Haverford 2 pl. 75 HLC 57.iii.16, iv.1, 6, is assuredly an error for m i 5 SimaSki. Since the goats in question were part of the delivery from Drehem (Nibruki-[ta], iv 14), which also included Shimashkian and Lullubean sheep, the Lagash scribe, unfamiliar with the Drehem terminology, apparently confused the two geographical designations. For this document, see also above n. 45. See Steinkeller, SEL 6 (1989) 4-5. See ibid., pp. 4-5. More correctly, the word for "nanny goat" is uzudx@z) (see Steinkeller, MC 4, p. 42), but I use the traditional reading uz for convenience.
Sheep and goat terminology in Ur I11 sources from Drehem
Steinkeller
Cf. uzu-umbin-TAB.TAB = MIN (= e-pi-r[i-turn]) (Hh. XV 235 = MSL 9, p. 13), where the meaning of epiritu (if restored correctly) is unfortunately unknown. See udu-ad4 = kub-bu-lu (FIh. XI11 28); at-t?l ERIN-ten6 = 3 te-nu-u = k~i-bu-lu,ad-*da GRIN-tenii = 4 te-nu-u = KI.MIN (Ea I1 230-31 = MSL 14, p. 257). Cf. Iu-ad4 = hu-um-mu-ru-um (MSE 12, p. 201 line 10).
PLOW ANIMAL INSPECTION RECORDS FROM UR I11 GIRSU AND UMMA
E.g., 1 6b GUR8.GUR8 (MVN 2 168:3); 2 gud GUR8.GUR8 (MVN 13 816:33); 1 gud GUR8.GUR8 babbar niga, 2 gud GURg.GUR8 (Legrain, TRU 80:3, 8).
Wolfgang Heimpel
See CAD K, pp. 563-64; Steinkeller, MC 4, pp. 53-54. In favour of such a connection, note that both gur8-gurg and gur4-gur4 are attested as the spellings of kurkurru (see MC 4, pp. 53-54).
(University of California, Berkeley)
See udu niga sig5 = MIN MIN (= im-me-ru ma-ru-u) darn-qa (Hh. XI11 3). See udu
N - ~ ~ =S MIN E (= im-me-ru) ma-ru-u' (Hh. XI11 2).
Contents
See udu u = im-me-ri :am-mu (tb. XI11 10). So Jacobsen, Copenhagen, p. 8 n. 2; also this author, ASJ 3 (1981) 88. For this conclusion, see already A. Archi and F. Pomponio, CMET 7, p. 89. E.g., AUCT 3 191:1, 1921, 212:l. For girx(GI) as an abbreviation of udu-erne-girx, see H. Waetzoldt, Textilindustrie, p. 6. E.g., PDT 634:15; AUCT 1 492:3. PDT 627:2; Nies, UDT 44:2. P D r 429: 1. E.g., PDT 429:3, 634:7. E.g., AUCT 1 910:l; 2 111:1, 151:l; 3 361 : l . I assume that -gul is syllabic for -gu-lalgal. See udu gir-gu-la = MIN (= im-me-ru) ar-ri (Hh. XI11 4); gir-gal, gir-gu-la = a-ri-tu (CAD ,412, p. 270b). Cf. also Waetzoldt, Textilindustrie, pp. 53-55. E.g., PDT 634:9. For siki-mug, Akk. mukku, see H. Waetzoldt, Textilindustrie, pp. 56-57; CAD M12, pp. 187-88. Note SIG.MUG =: mu-gu in VE 973 (MEE 4, p. 307).
Introductory remarks I Introduction to the Girsu texts 1 Agricultural units 2 Households 3 The partition of households into two divisions and the top administrators of the households 4 Distribution of units among households and diachronic fluctuation of numbers of units 5 Cultivators, bull lieutenants, ten-bull-scribes, and plow bull scribes
11 Types of guq-apin-gub-ba texts 6 The three text types and texts of type B 7 Texts of type A 8 Texts of type C 9 Remarks to individual texts 10 Sample text 12
E.g., PDT 381:2, 383:4, 634:lO. Quite likely, the Umma classification is based not so much on types of breed but rather on types of wool.
111 Oxen and donkeys 11 Animal designations (A), sequence of listing (B), and quantities (C) 12 Excursus on Maekawa's identification of ade-kunga with the Persian Onager 13 Oxen and donkey groups 14 Classification of animals in totals, grand totals, and summaries 15 Working and not working animals 16 Norm of working animals in a group 17 The ideal composition of a group of working animals 18 Not working animals in plow groups 19 Barley of the calves
IV Loss and replacement, withdrawal and supplementation 20 Deaths 21 Death rates 22 Replacement, including a note on the nu gloss 23 Causes of replacements 24 Deficit replacements 25 Substitution Bull. Sumerian Agric. 8 (1995)
Heimpel
Plow animal inspection records 26 Withdrawals 27 Supplementation
V Other administrative procedures of plow animal management 28 Intramural redistribution 29 Redistribution and borrowing between households 30 Redistribution within the household of Ningirsu 3 1 Dissolution of groups and change of managers 32 Inspections 33 gu4-apin-dib-batexts 34 The term dib-ba in gu4-apin-gub-batexts
VI Umma 35 Umma texts 36 Inspections of administrative units and field inspections 37 The districts 38 The inspection sites 39 The bull of Gula 40 guq-Suhub 41 Month dates 42 Text formats 43 The persons of text 8 44 The norm of working animals in Umma 45 Supplements 46 Withdrawals and a deficit replacement 47 Deaths and death rate 48 Replacement 49 Registers of outstanding replacements
VII Conclusions
Plow animal inspection records
l lcilnpel
from Girsu have been noted sporadically but not sytematically treated before. They were called irlter their subscript guq-apin-gub-ba texts by I. Gelb.7 I have found 48 such texts, a number which may seem impressive; but considering that about 100 such texts must have been written I I I a single year, they are really a pitiful lot. Many are also badly preserved, and the chronological distribution of the well preserved texts is not very happy as there are no exemplars of one and the same administrative unit from consecutive years; one text records the animals of all administrative units of the province, but it is a fragment; there is no overlap t~etweentexts of different types; there are few links with the well-researched seed-and-fodder tcxts. The texts from Urnrna are few and much shorter. If there is one wish that emerged from trying to extract knowledge about the management of plow animals in the 21st century B.C. in [he lower alluvium of Tigris and Euphrates, it is that more guq-apin-gub-ba texts will be identified and published. Numbers written with the number stylus are followed by an asterisk * in transliterations. Abbreviations follow M. Sigrist and T. Gomi, The Comprehensive Catalogue of Published Ur I11 Tablets, pp. 81-95, right hand column, without year dates. Abbreviations not found there are: ATL
Agricultural Texts of Ur I11 Lagash published by Maekawa in a series of articles. Texts 1-10 = ASJ 3 (1981) 50-61; 11-23 = ASJ 4 (1982) 108-127; 24-36 = ASJ 8 (1986) 104-120; 37-45 = Zinbun 21 (1986) plates I-XI1 following p.157; 46-58 = ASJ 9 (1987) 113-129; 59-68 = ASJ 11 (1989) 129-141; 69-78 = ASJ 13 (1991) 222-235; 79-92 = ASJ 14 (1992) 228-243; 93-99 = ASJ 15 (1993) 122-129.
Buffalo Bulletin of the Buffalo Society of Natural Sciences 11-2 (1915). SLTNi S.N. Kramer, Sumerian Literary Texts from Nippur (1944). SNAT T. Gomi and S. Sato, Selected Neo-Sumerian Administrative Texts from the British Museum (1990). UTI
F. Ylldlz and T. Gomi, Die Umma Texte aus den archaologischen Museen zu Istanbul, Band (ID) (1993).
In transliterations the sign GANA2-tenfi is abbreviated as Gt. Where deemed helpful the origin and date of a text is indicated behind the quotation. G = Girsu, U = Umma, D = Drehem.
Introductory remarks
A survey of Ur 111 texts shows that oxen and donkeys were used for draft in deep- and seedplowing, harrowing, and the unidentified task designated as "wet place" (ki-durs),l and donkeys and mules for pulling the vehicles of gods, nobility, and royal messengers.2 There exist representations of oxen pulling a threshing sledge from around 3000 B.C.,3 and textual corroboration from the Old Babylonian Farmer's Almanach? but the information on threshing in Ur 111 texts is silent on the use of animal labor.5 References to oxcarts are rare;6 grain and reed were moved by porters and boat. Evidence for the use of any other animal draft, as for example in lifting water, is completely lacking. The aim of this contribution is not so much to amplify on these uses of animal draft but to present information on draft animals in grain cultivation that is contained in records of inspections of such animals by the provincial governments of Umma and Lagash. The texts
Allen Estes checked the references and made many useful suggestions to improve this study. All remaining errors are mine.
Heimpel
Plow animal inspection records
I Introduction to the Girsu texts 1
inspection of the personnel of the household. The personnel was divided into sections with the li)llowing subscripts.
Agricultural units
Ur I11 administration used a basic agricultural unit of approximately 130 hectar (20 bur) of which one half lay fallow during a given year and the other half was cultivated by one cultivator with his three oxdrivers and a group of animals which was mainly used to pull the plow.8 Accordingly, the unit was called "plow"9 or, more often, "bull" in the extended meaning of the word as "plow animal group". The agricultural land of the province of Lagash that fell within the framework of the provincial administration, which alone is the subject matter of Girsu texts, consisted of 600 units: according to ATL 53 (undated) 11,652 bur, or 755 k m 2 , land was designated as ghn-guq kurg-dabs-ba gir-sb-ga guq 600-kam, that is "bull land", or land that was worked with plow animal teams, including the allotments of the personnel of 600 "bulls". In the itemization of assets of land of the province in ATL 52 of IS 5, the 600 unit group was called the "old 600 bulls" and augmented with a few other tracts. A distinct subgroup was formed by 480 units. It was designated in ATL 53 as "(the land worked by) the personnel of the 480 bulls" (gir-s&-gagu4 480), and represented an area of 9424 bur. It appears also in TCT I 629, a balanced account of income and expense of barley of the province in AS 7. The fifth expense item of that text amounted to 8640 gur, or 25380 hl, of "seed, fodder, and wages for the 480 bulls" (Se-numun mur-guq a-hug-gh guq 480-kam).1° In the text fragment TUT 27 (date broken) the 480 unit group was connected with plow animals: 3679 guq-ade-@-aapin-bi 480-iun "3679 various oxen and donkeys representing 480 plows". The high numbers of oxen, sheep, and goats that are recorded in the text suggest that it contained a count of the total livestock under management of the province. If this is correct, the oxen and donkeys of the 480 plows represented the total of plow animals that fell within the frame of the Girsu archlve. Maekawa concluded that the 480 unit group was "managed by the main public institutions headed by the sanga- and the ~abra-officials".llThe units of these "public institutions" were itemized in TUT 5, a record of the expenses for cultivation for the province in S 47. They add up to 419.12 It is not known to which administrative unit or units the remaining 60 or so units belonged. 2
Plow animal inspection records
l lcirnpel
Households
The "main public institutions headed by the sanga- and the Sabra-officials" were administrative entities that were designated with the word 6 "house" and the name of a divinity, a city, or a person. They are usually called temple households,l3 but the term may not fit the household of Namhani, whose name was never written with the divine determinative. The household of Lugal-URUxGt was often called 6 URUxGt, or simply URUXG~M. The households cannot simply be identified with one temple or residence and its possessions, personnel, and land holdings; they were composite administrative entities. Some consisted of a main temple and a number of small temples, or at least architectonic units called 6 "house", that were dispersed throughout the territory of the province. An instructive example is the household of Nanshe as itemized in UDT 58 43 (cf. Fig. 1). The text is the record of an
Figure 1 kS-didli 6 dNanSe
C dNanSe 6-a-ga-su-lim-ka 6 dHendur-sag 6-Sh-ge-ph-da 6 dHendur-sag g ~ n a~! a - d a - s a k - l a ~ ~ 6 d[ ]-ab 6 dNin-gig-zi-da u 6 dAma-gulO-g6 6~ M ~ S . Z A . A N - ~ ~ C dNanSe Pas-enku E-ning-e-gar-ra 6 dNin-x[ E-A-dam-dun Tir-ba-bil-la E-d~an~e-SeS-e-gar-ra 6 Ki-sa-la
ks-didli Sh Ninaki
Sh Gir-suki
was, according to its position at the beginning of the The "house of Nanshe" (6 d~an?ie) list and the fact that the section relating to it is by far the longest, the Sirara, the main temple of Nanshe in Nina.14 Other Nanshe temples were located in other cities and settlements, despite the misleading subscript "various shrines in Nina" (G-didli Sa Ninaki): one in Pas-enku (location unknown), the Eshagepada in Lagash,ls and the Enanshesheshegaral6 and the "house Kisala" in Girsu. In addition to Nanshe temples the list included temples of divinities that were closely associated with Nanshe: according to the Nanshe Hymn Hendursag was her "herald" (nimgir) and the sag-kn-tar (-keeper) of her house17 and according to the Hendursag Hymn he brought butter and cheese for the wedding rite of Nanshe and Nindara.18 The household of Hendursag is attested by itself in IT'I' 211, 865 (cf. section 5 for this text). The Eninegara was the temple of Nanshe's "brother" Ningirsu in Nina.19 The guq-apin-gub-ba texts contain several references for similar household structures. Maekawa noted that one of the top administrators of the household of Ningirsu, a certain Utumu, was titled bishop of Bagara, Bagara being Ningirsu's temple in the city of Lagash.20 The subscript of text 1 listed this and other sub-households: guq-apin gub-ba standing plow bulls of the house of Ningirsu 6 d~in-gir-su-ka house of the commander 6 Sabra
Heimpel
Plow animal inspection records
Plow animal inspection records
1 leimpel
C Ba-g6ra
house Bagara Bbaibbar White House ii 6 d~a-bil-sajj and house of Pabilsag ugula ~ r - ~ ~sanga a - u supervisor bishop Urbau The "house of Ningirsu" was presumably the Eninnu in Girsu; house Bagara Ningirsu's temple in the city of Lagash; White House was also a temple of Ningirsu, its location is unknown;21 Pabilsag was an alter ego of Ningirsu according to the Old Babylonian hymn SLTNi 61 which consists of characterizations that refer equally to Ninurta, Pabilsag, and Ningirsu. An additional sub-household was the house of Ninhursag. It consisted of one unit.22 Maekawa noted also that one of the administrators of the household of Igalima, a certain Urlamma, was titled "bishop of the house of Ninshubur" (sanga ~ . ~ ~ i n - i u b uItr ) . demonstrates that this household, too, was composite. The other administrator, a certain Adda or Addamu, was bishop of Inana.23 The section of text 4 for which this Adda was supervisor included 1 or 2 units that was, or that were, subscribed as guq d~nana"bull(s) of Inana". It demonstrates that not all units under supervision of the bishop of Inana belonged to the subhousehold of Inana.24 The same may have been true for the bishop of Ninshubur. It is strange bat no administrator of the household of Igalima was a bishop of Igalima. I cannot place the following sub-households: "house of Ninsun" with 4 agricultural units according to text 37, "house of Gishbare" with two or more units according to text 34, and "house of Shulpae" also with two or more units according to the same text. In addition to the households listed in Table 4 a household of Shusin is mentioned a few times.25 It is not known whether one of the established households was renamed after Shusin as happened with the household of Namhani which became the household of Arnarsin in the years from AS 3 to 3s 1,26 or whether the household of Shusin was a new creation. An alltogether different type of household may have been that of the governor. It was not divided into two divisions (see next section); the only administrator so far known was a certain Baddari whose title is unknown.27
3
The partition of households into two divisions and the top administrators of the households
For the purpose of administering the agricultural units each household was divided into two parts;28 they will be called "divisions" in this study. Each division was led by a bishop (sanga) or a commander (Sabra). Each administered one half of the number of agricultural units and their plow animal groups (cf. Fig. 2).29 The texts contain several examples that demonstrate strict halving between the divisional heads not only of plow animal groups but also of individual plow animals that came from outside into a household: according to text 2 the household of Gatumdu held 2 donkeys which had been found the year before. One had been assigned to each division (cf. the description of
the case in section 23). Of a transfer of 8 donkeys from the household of Ningishzida to that of Lugal-URUxGt 4 each went to the two divisions (see section 29). Figure 2 text
divisional administrator Urshagamu Urbau Urlamma Adda Dugazida Urigalirn Urbau Lugaluruda Urnumushda Nammah
units
The documentation of the late years of Shulgi allows an almost complete roster of the divisional heads of households. It is assembled in Table 4 together with the uncomplete data from the the guq-apin-gub-ba texts of AS 2. The distribution of bishops and commanders deserves a study of its own. I will only mention the curious case of Utumu. The household of Ningirsu was led in 3 41 by two bishops. Utumu was called expressly bishop of Bagara in ATL 10 of 41 ;31 yet in text 5 of 3 46 a transfer of donkeys to the the division of Urbau of the household of Ningirsu came "from the plows of commander Utumu" (apin d~tu-gu10iabrata).32 The format of texts 6 and 9 indicates the possibility that the two divisional heads of a household were sometimes temporarily subjected to the supervision of a third administrator (cf. section 7).
Distribution of units among households and diachronic fluctuation of numbers of units
4
In terms of agricultural units, the households were of different size, ranging from a high of 98 units of the household of Ningirsu in AS 2 according to HSS 4, 4 to a low of 14 of that of Nindara in 3 47 according to TUT 5. The information is assembled in Table 3. The number of units of a household fluctuated over the years. The numbers for the household of Ningirsu seem to indicate a peak in AS 2 that flattened out one year later. HSS 4, 4 of AS 2 listed 98 cultivators, corresponding to that many agricultural units, under the supervision of a single commander. In light of the numbers of the other texts, it is impossible that he administered only one division of the household. It would mean that the household had a total of 194 units. I assume that the commander of the text served as supervisor of the two divisional heads of the
Heimpel
Plow animal inspection records
household and that the 98 units were the total of the household. The numbers for Nanshe show a similar increase, albeit less extreme, from 50 in 3 47 to 70 in AS 2. By AS 5 they had fallen to 60 (only one division of the household is documented in text 31) at which level they stayed until [SS 11.33 The household of Ninmar experienced a steep increase in units from S 47 to AS 2 if the number of HSS 4 , 4 is not misleading. But already one year later they were back down at 63, and the number 30 in text 32 of an unknown year of AS, or less likely SS, fits perfectly with the 60 units of 3 47 if it referred to one division. Toward the end of the reign of AS, in his 7th year, the number had fallen to 50. The development in the household of Shulgi was similar but the peak in AS 2 was flatter and so was the reduction by [SS 11. The numbers in the households of Dumuzi, NamhanilAmarsin, and Igalima did not fall off after AS 2, but kept increasing, modestly in the case of Dumuzi, substantially in that of Amarsin, and spectacularly in that of Igalima. Those of the households of Gatumdu and Ningishzida did not change. On the whole the numbers indicate an expansion of agricultural land that peaked in AS 2 and declined subsequently. The individual households were affected quite differently by the changes. 5
Cultivators, bull lieutenants, ten-bull-scribes, and plow bull scribes
The great majority of persons appearing in the plow animal inspection texts were cultivators (engar) and bull lieutenants (nu-banda-gu4). Gelb understood them as two "classes of temple personnel", noting that they formed a unit as expressed in the summarizing formula engar nubanda-gu4-me "they are cultivators and bull lieutenants" in TUT 111 and elsewhere, and that the level of their earnings was not dependent on the size of the household as it was for the higher ranks. He also noted the different ratios between their numbers: one bull lieutenant supervised 5 cultivators according to HSS 4, 4 and also 4 or 2 according to the 9 personnel lists that he tabulated in chart 11.34 The evidence concerning these ratios in the guq-apin-gub-ba texts is summarized in Table 2. The ratio of lieutenants to cultivators was specific to households: in those of Ningirsu, Nanshe, Shulgi, Namhani, Lugal-URUxGt, and Dumuzi it was 1 5 , in those of Igalima and Ninmar 1:4, and in those of Gatumdu, Nindara, and the governor 1:2. In the household of Ningishzida it was 1:2 in an unknown year of Shulgi according to text 10 , but 1:5 in AS 2 according to texts 45 and 46. There exist minor deviations from these ratios. For example: in text 5 the first bull lieutenant supervised 4 cultivators, the next 3 bull lieutenants supervised 5 cultivators each, and the last bull lieutenant 6 cultivators; among the 14 bull lieutenants supervising 28 cultivators in text 10 was 1 bull lieutenant who supervised 3 cultivators and another bull lieutenant who supervised only 1 cultivator. The preponderance of the 1:5 ratio agrees with the fact that the number of units in many households was a factor of 5. The evidence of the texts used by Gelb deviates in many points. HSS 4, 4 of AS 2 recorded a ratio of 1 5 , or nearly 1 5 , for all households. For that of Ninmar it was 19 bull lieutenants and 95 cultivators, which is closer to a ratio 1:5 than 1:4 while the ratio 1:4 is attested in no less than 6 plow animal inspection texts from the same year; and in the 4 late personnel lists (AS 8 and SS 1) on the left hand of chart I1 the ratio was uniformly 1:2. It compares with the evidence of the plow animal inspection texts as shown in Fig. 3:
Plow animal inspection records
l lcimpel
Figure 3 household Hendursag Ijagara URUxGt Igalirna
ratio in Gelb and here 1:2 and 1:5 1:2 and ? 1:2 and 1:5 1:2 and 1:4
remarks as part of household of Nanshe Bagara is included in text 1
For the time being it looks as if the ratios were changed to 1:2 after AS 2 and before AS 8.35
Cultivators usually managed one unit and its plow animal group, but occasionally also two. Luurasari and Lugalsipa managed one group in the division of Urnumushda in the household of d ~ u ~ a l -in~AS~ 2 ~andx took ~ tcharge of a second in the other division of the household at the same time according to text 12 (cf. Table 5). The second cultivator in text 5 and the first in text 29 had a double group (6-bi 2 - h ) , and so did the second of text 43 (Ur-la6-ga engar 2) and Lugalshala in text 33 V 1'-5' (gu4-apin-bi 2-am). Cultivators could become supervisors and function like bull lieutenants occasionally. The cases known to me are quoted in section 49. It is not known whether they could become bull lieutenants. "Plow bull scribes" (dub-sar-gu4-apin)and "lo-bull-scribes" (dub-sar-gu4-10)took an intermediary position between bull lieutenants and divisional heads in the hierarchy of the field administration. The former was the higher rank. According to BM 12232 he belonged to the rank of the "elders" (ab-ba-ab-ba)while the 10-bull-scribedid not.36 The latter is well attested in the so-called Round Tablets that refer to the household of Ninrna~-37and in Umma. The texts from Umrna allowed Maekawa to demonstrate that a 10-bull-scribe supervised two bull lieutenants which agrees with the title since the two bull lieutenants in turn supervised 10 cultivators who managed 10 plow animal groups that were called "bullsn.38 In the gyq-apin-gubba texts 10-bull-scribes and plow bull scribes were mentioned occasionally as persons who had to replace plow animals. They may have written the texts that are discussed here, the plow bull scribe the large texts of type A, the 10-bull-scribes the small texts of type B. Bull lieutenants wrote occasionally themselves, as is indicated by the seal legend of Lugalkuzu which identifies him as scribe (see section 49). The involvement of plow bull scribes in recording the receipt of skins of fallen animals is described in section 20. Gelb noted that the ranks of household personnel above the bull lieutenants and cultivators were called "elders". The elders that occur in the gu4-apin-gub-ba texts are the bishops, commanders, and an occasional plow bull scribe. They will be referred to as "adnlinistrators", while the 10-bull-scribes, bull lieutenants, and cultivators will be called "managers".
Heimpel
Plow animal inspection records
I1 Types of gu4-apin-gub-batexts 6
The three text types and texts of type B
The corpus (Table 1) consists of published texts that include records of plow animal groups in the form of lists of oxen and donkeys each of which was associated with the name of a cultivator. They are easily recognized by the round and rounded impressions of the number stylus. Most included the phrase "standing plow bulls (guq-apin gub-ba) of household x" in their subscript.39 They were classified under that phrase on the tags of tablet boxes that contained such texts. Cf. e.g. CT 3, 10 BM 14608 SS 4: pisan-dub-ba tablet box guq-apin gub-ba standing plow bull (records) 6 d~in-marki (of) house Ninmar i-gail are in it The context of the tablet box label inscription shows that the phrase guq-apin-gub-ba could refer to the tablets on which the "standing plow bulls" were registered. This extended meaning may also apply to the subscript of the texts. The scope and format of the texts exhibit variations that divide them into 3 different types. They will be described below separately. Short texts, classified here as type B, recorded groups of plow animals and their cultivators that were managed by a single bull lieutenant. Most of these texts were dated AS 2; several were undated, and one, text 8, was dated S 47. Originally, the texts from AS 2 would have been stored together in their tablet boxes, arranged according to households. The boxes disintegrated over time, but the tablets kept their physical proximity in the ground, loosing it step by step and more or less as they were transferred to the saddlebags of their illicit excavators, the new boxes of antiquities dealers, the shelves of museums, and the publications of Assyriologists. It so happens that most found their way into museums in England. The subscript identified the bull lieutenant of the cultivators that managed the plow animal groups that were registered in the text, the household in which he was serving, and occasionally also the divisional head of that household. It did so in various forms. They are tabulated in Fig. 4. Column 2 refers to the identification of the numbers of plow animal groups that were registered in the text. In text 8 it constituted the first element of the subscript; it was phrased 5 guq, and the numeral was written with the letter stylus in vertical notation. It does not occur in the texts dated AS 2. In texts 41 (not tabulated here because the rest of subscript is not preserved), 45, and 46 it was written on the left edge as last entry, again with the letter stylus, but in horizontal notation. Column 3 refers to the formula guq PN which represented the first statement of the subscript in a few texts. The persons were Ab-ba-gi-na (22), Ab-ba-Sag-ga (24), and ~ r - d u t u dumu ur-d~amma(28). The last person cannot be the bull lieutenant or the head of the division of that name as both were mentioned further on in the subscript. Abbagina and Abbashaga should be the bull lieutenants as these would otherwise not have been recorded.
Plow animal inspection records
l lcimpel
Column 4 refers to the supervising bull lieutenant. He was identified as supervisor (ugula) and by name; his title was occasionally added (the cases are marked as "nbg"). Column 5 refers to the presence or absence of the formula guq-apin gub-ba. In text 23 p4-apin-dib-ba is found instead. The term is discussed below in section 33f. Columns 6-8 refer to the identification of the household in one of 3 forms: by recording in text 14; or by placing the word e the name of its divinity or person, for exainple d~in-marK or gu4 in front of it, for example 6 d~anEein text 8 and guq d~in-marW in text 13. Columns 9 and 10 are complementary, recording the name of the supervising divisional head with the formula ugula "supervisor" or i-dabs "took in charge". Figure 4 text
#
guq PN
ugula 1
gag
B
gu4-
N
i-dabs ugula 2
year
Heimpel
Plow animal inspection records
Plow animal inspection records
l leimpel
[guq-apin gulb-ba [(@ Nam-b]-ni 1-x (not -la, i.e. [A-kall-la,or -sum, i.e. wa-an]-sum) [ [~r-~~almma [enlsi / date
The copies of copyists who appreciated the physical aspects of tablets indicate many erasures of lines, and Reisner observed that one line of text 35 was scratched into the dry clay. It appears that the process of recording these texts was accompanied by much change of information. The last lines were occasionally written onto the left edge. The scribe, presumably the plow bull scribe (dub-sar-gu4-apin),of the household of Ninmar of AS 2 used the left edge for inscribing the name of the supervising bull lieutenant. It may have helped to identify a tablet after it was stacked among others in a tablet box. The sample of texts is too small to determine whether individual plow bull scribes can be recognized by a specific use of the various elements of the subscript.
7
Texts of type A
Texts of type A are long; they recorded the groups of an entire household or, less often, of one of its divisions. Text 12, which is perfectly preserved, has 340 lines in 10 columns; text 31, almost perfectly preserved, has about the same number of lines in 14 columns. Texts 5,9, and 10 are also long and of decent preservation. The rest are fragments large and small. They span the 13 years from 41 to AS 5 rather spottily. One of the few texts from the reign of Urnammu, a small fragment of a large tablet, labeled here text a, belongs to this type. Text formats are tabulated in Fig. 5. Texts registering the plow animal groups of two divisions have a threefold structure. First the groups were itemized according to the divisions (A1 and A2), then their numbers were totaled (B), and finally the text was characterized in a subscript (C). The itemization consisted of a definition of the composition of the plow animal groups in the division and the status of its animals (a sample is presented in section 10). The number of animals was then totaled and the number of groups, separated into oxen and donkey groups, formed by these totals was given. Finally the name of the divisional head was recorded, usually with the formula ugula PN, in texts 6 and 9 (column 4) with the formula guq PN "bulls of PN". Part B consisted of various totals. Texts 6 and 9 adhered to a rigorous separation of itemization and totaling and thus grouped the totals with the grand totals and summaries in part B, while the other texts kept the totals close to the itemization and thus left them in part 1. Totals, grand totals, and summaries were three categories of totaling. They are described in section 14. The subscript contained the key phrase of the texts "standing plow bulls of household X", the name and title of the ruling governor and the year date. In the texts of column 1 and 2 the names of the divisional heads were repeated; the scribe of text 12 (column 3) thought that was unnecessary. Texts 6 and 9 (column 4) pose a problem. Text 9 concluded section A1 with the entry guq Ma-an-sum in V 11, and section A2 with the entry guq A-kan-la] in rev. I11 2'. We expect the two names to be repeated and connected by u "and" in the subscript as in the texts of groups 1 and 2 of Fig. 5. Yet the traces are nor compatible with this expectation. The relevant section (rev. VI) reads as follows:
Figure 5
2 divisions A1
A2
B
c
contents itemization totals of animals # of plow animal groups ugula PN1 bulls of 1st division itemization totals of animals # of plow animal groups ugula PN2 bulls of 2nd division totals grand totals summaries total of groups total of 164 guq-apin gub-ba household PNl u PN2 PN governor Year
1
2
3
+ + + +
+ +
+ +
4
1 division 6 5
+
+
+ + + +
+ + + + - + + + + + + + + + -
-
-
+ - +
+
+ + +
+ + +
+ + +
+ + +
-
-
-
+ + +
+ + +
+ + +
+ +
+ +
-
-
-
+ +
+ +
+ + + + +
+ + + + + + +
+ +
+ + +
+ + +
-
-
-
-
-
-
-
-
-
-
-
-
-
column 1 = texts 2, 4, 11; 2 = 10; 3 = 12; 4 = 6, 9; 5 = 5; 6 = 31, 32
In text 6, the end of part A2 is preserved, and we find, as expected, one of the divisional heads of the household of Shulgi mentioned in the phrase guq ~ u - ~ i Iabra n a ~"bulls ~ of the commander Lunina". The supervisor of the first division should have been Urshugalama who was still divisional head in AS 2. In the subscript we expect again Urshugalama and Lunina,
Heimpel
Plow animal inspection records
but find Lugalsigsu instead. 3 47, the year of text 6, was the year in which Lugalsigsu is first attested as a top administrator in the household. In this text he appears to have been placed above the divisional heads. It must have been a temporary solution because by AS 2 he had succeeded Lunina as colleague of Urshugalama. The tripartite administrative top of the household would explain the make-up of text 9 where the person "( 1-x" was placed above Akala and Mansum, and of the evidence of HSS 4 , 4 where one administrator was in charge of all agricultural units of the household of Ningirsu (cf. section 4). Texts that included only the groups of one division differ from the rest by indicating 16-i amounts (cf. section 16 for their meaning). The texts of type A are not just long, they were also skillfully written and make an aesthetically pleasing impression which is enhanced not little by the use of the number stylus. Their importance was stressed by the choice of the unabbreviated version of the year dates that included the full titles of the reigning king.
8
Texts of typeC
Texts 30 and 37 differ in outward appearance as their scribes did not use the number stylus. In contents, they differ in not including itemizations of plow animal groups and mention of their managers and administrators. Recorded were the totals of animals of a household in sections of assets (gub-ba-h), withdrawals (zi-ga), deaths (ri-ri-ga), and supplements (libir-h), the count of oxen and donkey groups in a household, and the name of household. The designations of the totals correspond to those of the grand totals in texts of type A; yet they differ in including the supplements which were subsumed among the assets in texts of type A. It follows that a text such as 37 could not have been produced by mechanically copying the sections of grand totals of the tablets of the households. Text 37 included some unique features: it used the term i-bi-za "loss" (VII 29-30) and it mentioned several mules besides oxen and donkeys. The text was inscribed on a tablet of 16 columns that probably registered the plow animals of all households, or even the entire 480 unit group. Column I, XV, and XVI are lost, and of many columns only little remains. The date is broken, but there fortunately exists very good circumstantial evidence for it: column VIII 23 listed a dead female mule in the household of Nindara. It is very likely that it is the female mule of the household of Nindara whose skin should have been, but was not, received by the administration in SS 1 according to MVN 5, 180 (cf. section 20 II). Text 30 used the same terminology and a similar format. It recorded the plow animals of the households of Ningirsu and Ninmar in AS 3. In the first section assets and withdrawals were distinguished, in the second there is no distinction. Presumably the recorded animals were all assets. Texts of type C were probably used to establish the number of all plow animals in the province. That this was done is evidenced by TUT 27 (cf. section 1).
l lcirnpel
9
Plow animal inspection records
Remarks to individual texts
'the quality of publication of the texts is very uneven. Text 2 is almost completely preserved and thus a very important source, but it was published with an unreadable photo, no copy, and a problematic transliteration which includes impossible restorations. I have attempted to reconstruct the numbers by comparing itemizations, totals, grand totals and summaries with the following results: obv. I11 9 should be 1hb sag guq-gig-g& because 12 hb-ma + 8 8b-[mhh] of the sccond total = 20 6b-mhh of the grand total. Line 16 of the grand total must be 1 hb-[3]; line 18 6 guq-[I]; line 19 1hb-[I]. The milk calves pose unsolvable problems. 7 + 2 nu mar-ga su-su of the lirst and second total = 9 mar-ga [so-su] of line 20 of the grand total. But there is no equivalent for the 2 6b-amar-ga of the second total in the grand total. I can find only half of the adult donkeys of the grand total in the itemizations and totals. The subscript of text 2 in rev. V 7-10 should be read as follows: gu4-apin gub!-ba!d[gh-tiirn]-du10ugula Ur-Sag-ga[-gulo]ii! ~ r - ~ B a - [ u ] . 'The restoration of the name Gatumdu can be derived from the fact that ~ r - ~ ~dumu a - uUN-il and 1~r-gag-ga-gu10 were the administrators of the household of Gatumdu in S 43 according to ATL 27 (see Table 4). Text 6 was published by Th. Pinches in the Amherst Tablets and again by H. Limet in pigr rap hie I, as # 86. Limet quoted the Amherst text (p.78) but did not realize its identity with the text he was copying. Pinches' copy is better for details, Limet's for a view of the large fragment in its entirety. Text 32 is available only in Lafont's transliteration which contains some problems. oh]-i su-ga LLreplaced deficit" in I 21 and I1 4 cannot be correct. It makes no sense in the context and was never used in wapin-gub-ba texts. Read [duq instead. Column IX must contain the sukilib section, i.e. SU~LAGABXLAGAB (see section 14). To judge by the often hazy distinction between S U ~ L A G Aand B SUXLAGABXLAGAB in copies, the ligatures are not easily distinguished. Column X contained the end of the kilib-ba section; lines 10'-11' must be [guqapinl-bi 22 / [anSe-apinl-bi 8; line 12' should contain the total of deficits as in text 31 in the same position, i.e. @h-i]71 guq 06-i] 11ange. Line 13' is of course [guq-apinl-gub-ba.Line 14'-15' must contain the identification of the head of the division. Text 46: The first sign in line 21 was copied as two verticals written with the letter stylus. The format requires the name of the cultivator for the line. Read therefore A!. The cultivator, A-ab-ba, was dead, ba-TIL. This is the only such record in the corpus. Restore the lines on the left edge: [(guq-16.) d~in-gi]~-zi-da I [ugula U]r-6-ninnu. The numeral on the edge was copied as 2 by Radau. It should have been 5.
10
Sample text 12
This is an almost perfectly preserved text from AS 2 with a record of all plow animal groups of the household of Lugal-URUxGt. Its structure is very even. It is shown in Table 5. The household was divided into two equal divisions that were supervised by Urnumushda and Narnrnah. It will be seen from the material collected in Table 4 that Nammah was already divisional head in S 41. Urnumushda must have suceeded Niurum after S 47. Nammah or Urnumushda was bishop, or both were bishops, because text 12 registered animals that had
Heimpel
Plow animal inspection records
been replaced by "the bishopM.40If the unnamed bishop, or bishops, had been from a different household we would expect that to have been mentioned. The itemization within each division consisted of a list of decriptions of plow animal groups. Each was connected with its managing cultivator whose name and title was recorded. After a block of five groups the name and title of the supervising bull lieutenant was given. 10 cultivators and 2 bull lieutenants represented one division. In four cases, the formula PN engar, which identified the manager of a plow animal groups, was replaced by the formula g w a n i e PN1(-kam) PN2 i-dabs "(they are the) bulls/donkeys of PN1, PN2 took charge of them;" in one case (group 8 in the second division) it was replaced by the formula PN1 i-dabs. The formulas designate change of manager of a group. It is discussed in section 31. The itemization in text 12 follows the basic format that was used for all yearly livestock censuses: assets were recorded first, withdrawals, deaths, and supplements followed in that order. In Table 6 three samples of group descriptions are transliterated and translated. It will be seen that the numbers in the assets and supplements sections were written with the number stylus and those of the withdrawal and deaths sections with the letter stylus. The term "assets" is here used for gub-ba-Am, literally "it is standing", rather than the more common translations "on hand", "available", "present". None of these translations is completely satisfactory because the sections included also animals that were "yet to be replaced" as the "two year" cow in group 9. The designation implies that the animal was not "present", "on hand" etc. at the time of recording. Since the administration was not liable for its replacement the cow counted among its assets. The special status of absent assets was expressed in most texts of type A by placing the sign nu "not" close to the number sign, and in texts of type B by writing the number with the letter stylus. A full treatment of this category of animals is given in connection with the topic of replacement in section 22 below.
I11 Oxen,and donkeys 11
Animal designations (A), sequence of listing (B), and quantities (C)
A The gender and age designations of oxen are well known. I. Gelb demonstrated in his article "Growth of a Herd of Cattle"41 the basic fact that amar-a designated the full first year of life regardless of the time of weaning, mu-1 the second year, etc. Beginning with the fifth year of life the age of oxen (and donkeys) was not expressed anymore; females were Bb-ma and males, in the Drehem text which he discussed, guq-gal. In the Ab-gub-ba texts from Girsu, the males were guq-aib or gyq-giS.42 guq-aib designated the breed bull, the latter a male adult ox which was not used for breeding. In the plow animal inspection texts, guq-gig was the exclusive full designation of the adult male ox. It was often, but not consistently, abbreviated to gu4, preferably when other qualifications followed. Cf. e.g. the following sequences in text 37: hbm a L'cow"..db-mah-su-ga "replaced cow".. guq-gig "yoke bull" .. guq-gig su-ga "replaced yoke bull" .. gyq su-su "bull to be replaced" (III 12-16), or i b - m a .. Bb-ma su-ga .. I b - m a su-su .. g q -
l lcirnpel
Plow animal inspection records
fils .. guq su-ga .. guq su-su (IV 16-21). In text 4 supplemented bulls were called guq 6-tiir na-gabtum-ta "bulls of the pen house, from the NaGABtum" in the itemization while the very same
irnimals were designated gu4-gii na-gab-turn-ta "yoke bulls from the NaGABtum" in the total. While gu4-gii designated always individual animals, guq was also used for general and extended meanings. In the phrase Bb-guq-hi-a it designated any male ox, in guq-apin any plow animal including donkeys, in guq d~nanaplow animal groups. The element gig cannot have been an abbreviation of gig-du which qualified uncastrated male adult animals including oxen in the texts I'rom Drehem.43 It is contradicted by the opposition of guq-gii and guq-6b in the hb-gub-ba texts; we also must assume that the normal plow ox in Sumer was castrated. We might consider that fils in guq-gig meant "penis" and was abbreviated from some phrase which expressed castration. Yet, the term ade-gig contradicts this because ade-giS designated a collective of working plow donkeys including females. It is necessary to stress this point because B. Landsberger and others were, and are, convinced that castration must have been expressed in the terminology of domestic animals, spending much effort to find the expected designations and proposing to have found them. I do not know a single designation whose function was the identification of castrated animals as opposed to noncastrated ones. The distinction, in German, between "Stier" (uncastrated)and "Ochs" (castrated) seems to be restricted to a variety of Hochdeutsch which is somewhat removed from the reality of dialectal German as spoken by those who handle these animals. My attempts to understand the terminology of castration from farmers in the Schwarzwald (southwest Germany) always elicited head-shaking. They divided their oxen not into anatomical, but functional, categories, namely, among the males, draft, meat, and sires; and while these functional categories overlap the distinction of castrated and noncastrated they do so only more or less, but are not identical with them. I suspect that the situation is not much different in the case of "bull" (uncastrated) and "steer" (castrated) in Webster's English as opposed to the terminology of the countryside. In this study I use "bull" as translation of guq and "ox" as designation for a bos taurus. giS in guq-gig is the word for yoke. It is attested in lexical tradition in the equation gii = niru and in Ur 111terminology in the expression gig-i&"(fit) for the yoke". In the hb-gub-ba texts guq-gig can hardly have designated yoke bulls. It was presumably the male adult ox which was usually used as plow ox and kept, for reasons unknown to me, among the dairy cows. "3 year bulls" and "3 year stallions" do not occur in Girsu texts. As far as working is concerned they were functional adults so that the designation was really superfluous. "3 year" females are mentioned, yet their number creates a dip when plotted on a line between "2 year" and adult females, indicating that some of the "3 year" females were designated as adults. In the Umma texts "3 year" males and females occur routinely and in numbers that indicate consistent differentiation from adults. The term iu-giq poses a problem. There exists unambiguous lexical evidence equating guq-iu-giq with Sibu and Bb-iu-giq with Sibtu "old"$4 and this meaning fits the evidence very well. But text 10 VIII 31 recorded a "1 year" iu-gi4 stallion; a "2 year" iu-gi4 stallion is attested in text 1 I1 11-12, and a "2 year" Su-giq mare in the same text XI1 34.45 SU-giq must have designated a quality that applies typically, but not exclusively, to old age, such as "weak". iu-gi4 animals were always listed last, which indicates that they were the least valuable; they are typically found in withdrawal sections; and they did not work except for a single mare which
1 lcirnpel
Heimpel
Plow animal inspection records
was declared fit for the yoke (4 I1 8). The term Su-gi4 was not used in the Urnrna texts; instead, Su was placed in front of the animal designation.46 The term siKKurn stallion designated, in my opinion$7 a mule or donkey that was used to pull the coaches of royal messengers. Entries of siKKum stallions were, with one exception, marked with the "nu gloss" which indicates that they were not present before and at the time of recording (cf. section 22). In two Umma texts the designation za nu-Su "not branded" was added, once to a milk class male donkey (text 7 I1 2: 1dur-ga d nu-Su) and once to a replaced yoke bull (4 111 17'). These cases indicate that plow animals were usually branded. D. Foxvog, in a forthcoming study on branding in Sumerian terminology, quotes texts from Girsu (ITT V 6796; SNAT 25 1; 146) that mention the branding 'coppers' (urudu si-im-da) of individual households. Plow animals were occasionally transferred from one household to another (see section 29). SNAT 146 SS 8 seems to record a case where the question of branded animals arose in such a transfer. [n] anSe [nl donkey(s) guq-apin Ba-g8ra-ka ba-an-si were entered among the plow bulls of the Bagara ki LU-d~tu-ta from Luutu bull lieutenant Urbagara received them Ur-Ba-g8ra nu-banda-gu4 '1-dabs d u ~ ~Ba-ghra supervisor U., bishop of Bagara ugula ~ r - ~ g h - t i i m - sanga si-im-da u ~ u x ~ t - i-ak ba~~ the (branding) mark of48 URUxGt had been applied nu-bhnda-guq-keq igi-na i-lag the bull lieutenant ... Surprising is the absence of the term guq-hb-ur-ra which is well attested in texts recording the purchase of plow oxen (see section 27 below). It is known from a paragraph of the laws of Lipit-Ishtar where it was contrasted with guq-8b-sag-muru4.49For the guq-8b-ur-ra one had to pay 2400 liter barley rent for 2 years while 1800 liter were the rent for guq-ib-sagmuruq. CHs242f. treats the same topic. The first term was then written guq-8-ur-ra, and this is the form that entered lexical tradition where it was translated into the ungrammatical Akkadian form a-lap ar-ku (one expects ar-ku-ri) which served as the basis for the conventional English translation "rear ox".50 The passages quoted in section 27 confirm that the guq-hb-ur-ra was the most valuable plow animal and fetched the highest prices. Civil's interpretation of a plow ox, bull or cow (because of the element gu4-Ab),which was not found in the front (sag) or middle (muruq) position in a team of 3 pairs and thus in the rear as confirmed by the Akkadian translation and the corresponding synonym guq-eger-ra,is very difficult not to accept. But there are problems: the interpretation of the element guq-ab as bull or cow fits the legal context but not the references in section 27 which speak of 1 particular guq-8b-ur-ra;the forms guq-8-ur-ra and guq-ur-ra in post Ur 111tradition indicate that the form guq-8b-ur-rawas not anymore understood at the time and re-interpreted; the form a-lap ar-ku or ar-ki is, as noted, ungrammatical; the silent assumption that the pair closest to the pulled plow should require the best draft animal is not convincing; the OB tradition which seems to include the kind of information that would
Plow animal inspection records
help clarify the nature of the guq-8-ur-ra,as can be seen from the data assembled in M. Stol's contribution in this volume, has not been subjected to contextual interpretation. Note, for example, the letter CT 29,28f. 80186, a reference given by CAD S.V. arkci. The writer says that hc has 3 oxen but that there is no guq-ur-ra. He asks his father to look among his oxen, find a good guq-ur-ra, set its price, and send it. The implication seems to be that one good gu4-ur-ra will make it possible for the writer to use his three oxen, among which there is no gu4-ur-ra, effectively. The hypothesis of guq-iir-ra as one of a pair of rear oxen is not confirmed. #
11 The sequence of listing followed the principle of decreasing value and is mirrored in the
arrangement of Table 7: adult animals were listed before the young, and among the young the older animals came first. The Su-gi4 animals came in last place after the milk class. Within each age group, female preceded male as in the pasturing environment and not as in the fattening cnvironment.51 Female draft oxen may be expected to have ranked below males because of their inferior strength. Yet their potentially higher value is confirmed by SNAT 135 according to which a cow fetched the highest price (8 shekel silver) while 3 bulls fetched 7 each, and 2 more 6 each. According to SNAT 136 a cow and a bull fetched 7 shekel each. The term for oxen of various ages and gender in Girsu texts was 8b-guq-I$-a and thus conformed to the convention of listing the females first; in Umma it was guq-8b-@-aeven if there, too, females preceded males in itemizations.
C For the numbers in column 2 of Table 7 I have used all recoverable numbers from the Umma texts and the numbers from the assets and supplements sections of texts 3,4, 10-12,31, 32, and the assets sections of 5 and 9, from Girsu. Representative data for distribution of species in Girsu are discussed in section 13. In the Umma documentation there were 426 oxen and 157 donkeys, or 73.07% oxen. The ratio of bulls (guq-gig) to cows (bb-mih) in Girsu was 641 to 171, or 78.84% bulls; in Umma it was very similar: 169 to 47, or 78.24% bulls. The ratio of stallions to mares was 379 to 254, or 59.87% stallions; in Umma it was again close: 47 to 43, or 52.22%. 12
Excursus on Maekawa's identification of anSe-klinga with the Persian Onager
The translation "mule" for age-kunga refers to the offspring of a donkey and an onager, as was proposed by J.N. Postgate and is widely accepted now. Maekawa believes that it designated the Persian onager. He discussed his identification of the kunga with the subspecies equus hemionus onager, as opposed to the subspecies hemippus, the Syrian onager, repeatedly, last in ASJ 13 (1991) 206-210. His first argument refers to the anle-kunga that occur as plow animals in the texts Fortsch 195 and 66 of Lugalanda 4 and 5 (tabulated by Maekawa in Zinbun 15 (1978) 100f.) where "about half ... were either completely blind (igi-2-na-bi nu-dug) or wounded in one eye each (igi-I)". He observed: "Injury to the eyes of the AN€E.BARXAN equids should not be regarded as fortuitous, since almost all the male donkeys (AN€E.DUN.GI)and oxen (gu4) which were used as plow animals by the same household had healthy eyes in Lugalanda 4 and 5". And he concludes: "A possible interpretation is that (they) were wounded in their eyes when captured or during being trained as draught animals afterwards".
Heimpel
Plow animal inspection records
The status of the eyes of ade-kunga was also occasionally specified in Ur I11 records. BJRL 64,63 D S 44 IIm'n: 93 a g e kunga-nita 93 male mules 2 ange kunga-nita igi-1 (AS) 2 one-eyed male mules 124 a d e kunga munus 124 female mules 1 anSe kunga-munus igi 2 (DIS.DIS)g6-gar-ra 1 two-eyed female mule mu-DU lugal income of the king In the badly preserved Girsu text f i98 ~is found the entry 1a g e igi-2 (I1 4). While these references indicate that there existed a problem with the eyes of these equids, I cannot follow Maekawa's explanation. It seems to be derived from the underlying assumption that the adult "captured onagers", as he calls them, injured their eyes in the process of being captured or trained. Yet is there any knowledge about whether and how.onagers of any subspecies were caught and trained for the yoke? I find a statement in Brehm's Tierleben (where much information about practices now extinct survives) about the subspecies Kulan which may be of relevance: "Versuche, den Kulan zu ziihmen, sind in seinem Vaterland (Innerasien) selten und nie mit vollstandigem Erfolg angestellt worden. ...Sie Cjung eingefangene und mit Stuten als Stiefmiittern aufgezogene Tiere) ans Einspannen zu gewohnen war unmoglich".52 If it was hopeless to try to train onagers for draft that were caught as foals before weaning it must have been even more hopeless to try it after capturing adult animals as Maekawa assumes. Eye troubles may not have anything to do with capturing and training and may not have been limited to the ade-kunga. M. Greenberg referred me to a passage from the Mishna which implies that eye troubles with donkeys were anticipated: "He who hired a donkey and it went blind (hibriqa) or it was seized for state service. To him he (the owner) says: it is your problem".s3 Greenberg remarks: hibriqa is explained by "(dazzling) light" as though from the root brq "lightning". This has been traditionally taken - I think questionably - as euphemism for blind. I think that it may simply mean "dazzled". Maekawa's three remaining arguments are negative: (1) there exists no positive evidence that kunga were born by donkeys; (2) or that ange-eden-na was the father; (3) and references to kunga foals are almost nil. All three points are correct, but there is circumstantial evidence which supports the existence of mules whose parents were onager stallions and donkey mares. All three animals are attested as being found together in groups as has been mentioned by Postgate, J. Zarins, and Maekawa.54 I find most compelling the evidence of Buffalo 3 U [AS 81, a count of large cattle of the 6-Su-sum-ma55 of the governor of Umma. Among the animals is the following group of which Urshara, a well attested donkey herder (sipa-an~e),~6 was in charge @dab5) : 2 anSe-nita eden-na 2 male onagers 1 a g e kunga-munus mu-2 1 female 2 year mule 2 a g e koinga-nita mu-1 2 female 1 year mules 1 a g e kunga-nita-ga 1 male milk mule foal 3 em-ma 6 donkey mares
Plow animal inspection records
Heimpel
2 erne6 Su-giq 1diir Su-giq 1diir amar-ga
2 old donkey (mares) 1 old donkey stallion 1 male milk donkey foal
The text does not indicate that the mules were the offspring of the onager stallions and donkey mares. In fact, we can trace the male milk mule foal as it was mentioned also in UTI 2135 U AS 8:57 1 male mill< mule 1 a d e kunga-nita-ga ii-tu-da gig.apin 'plow birth' from cultivator Uana ki U2-6-na engar-ta uS-gu10i-dabs uS-.guloreceived it (now) in the E. Sa 6 Su-sum-ma A-a-kal-la 6nsi-ka of governor Aakala @rSa-ta-ku-zu intermediary Shatakuzu sukkal LkdNanna sukkal (sic) vizier Lunanna vizier and scribe Lunanna ii ~ u - ~ ~ a dub-sar nna The fact that the animal was received "from" the cultivator Uana according to UTI 2135 and was a charge of the donkey herder Urshara according to Buffalo 3 clearly indicates the direction of its movement from the cultivator to the donkey herder. Thus it was not born from parents in Urshara's herd. Unfortunately, the phrase 'plow birth' is not clear. Still, I regard the combination of animals in Buffalo 3 as good circumstantial evidence: male onagers, female donkeys, and mules were kept together. Why would they have not crossbred them in a time and in an environment where also donkeys and horses began to be crossbred? For the time being, there is little information on the availability of male onagers, and no information on the techniques of obtaining these animals. A glimpse into the world of wild animal trade, or tribute, is furnished by PDT 560 D &3 9 XI 12: 80 a d e eden-na nita 197 mag-dh nita Ab2-ru-um-ma-ba-an.The same person occurs repeatedly as originator of wild animals. His name was written in almost as many ways as there are references. 20 female wild goats (dara4)58and 20 shapparu animalss9 were received from "Ab2-ru-ma-KU-ba" in AS 5; a young of a gazelle from "Ab2-ru-um-ta-KU-ba"in AS 7;60and another already in 5 47 from king (lugal) '~a-bu-ru-um-rna-~u-ba".~~ If he could deliver 80 male onagers, he must have ruled over people who were adept in capturing, rearing, and working with onagers.
13 Oxen and donkey groups As a rule, plow animal groups consisted of oxen or donkeys, but mixed groups resulted occasionally from substitutions of donkeys for oxen and viceversa. Cf. e.g. the assets of the second group in text 42:
Heimpel
Plow animal inspection records
3 yoke bulls 1* anSe su-ga engar sag! gu4-gS-Sk 1 mare, replaced (by) the cultivator, as substitute for a yoke bull l * dur su-ga sag guq-gig-82 engar 1 stallion replaced, as substitute for a yoke bull, (by) the cultivator 1 guq-giS su-su nu-bhnda-guq 1 yoke bull to be replaced (by) the bull lieutenant gub-ba-iun assets Such replacements are attested in 7 additional cases, all but one62 from texts of type B.63 Presumably the animals of the "wrong" species were soon moved to an appropriate group in the process of intramural redistribution (cf. section 28). The equids among the plow animzls were called aGe "donkeys". The word was used as a general term for equids and as a specific term for donkeys (Equus asinus) so that it is not always clear whether "donkeys" were donkeys. The only other equid that can be expected is the mule. A few mules (an3e-kunga) were listed in text 37 (cf. Fig. 6): 3* guq-gig
Figure 6
household Nanshe Nindara Dumuzi
male
.female
There is no proof that the mules were used for plowing even if they appear among plow animals. They may have been identical with the "donkeys" of the siKKum which pulled chariots. The appearance of mules in text 37 assures us that the equids that were not called mules in that text were donkeys. Since the mules are attested in such small numbers, we can be sure that the "donkeys" of the other texts were donkeys. Overall, there were more oxen than donkey groups. In the extant parts of text 37 the species of plow animal of 311 groups can be identified. 176 consisted of oxen and 135 of donkeys for a ratio of roughly 4:3. In the individual households the ratio was widely different. In that of Ninmar (according to text 32) and Gatumdu it was 6:l or 7:l. In the household of Nanshe in the early reign of Urnammu it must have still been higher in favor of oxen. In the preserved summary of individual animals 125 (and a half) oxen contrast with 14 (and a half plus possibly a few more in the broken section) donkeys. According to text 37 the ratio was approximately 5:2 in [SS 11. In the household of Ningishzida on the other hand it was about 1:2. The case of this household is interesting because the unusually high number of donkey groups is attested in texts from 3 34 to SS 1, demonstrating that it was a characteristic of the household. The text from S 34 is CT 10,34f. 15322, a seed-and-fodder text which recorded 11 calves and 28 foals as recipients of "barley of the calves" (Se-amar-ra). The household of Arnarsin had a ratio of about 1:l with slightly more donkey groups. "House Amarsin" was, as
Heimpel
Plow animal inspection records
Maekawa demonstrated,64 the name of the household of Narnhani between AS 3 and Shusin 1. Is it coincidence that the households that were presumably founded in the period of the 2nd dynasty of Lagash, i.e. Ningishzida by Gudea and Namhani by the last independent governor, had so many donkeys? There exists a noteworthy correlation with plot categories in the household of Ningishzida. According to the numbers of TUT 5 47 (cf. Maekawa's tabulation in BSA 1, 90f.) this household cultivated an unusually high percentage of category A land, that is land where 10 furrows per nindan (6m) were seeded with plow animals. It means that donkey groups must have been used to cultivate category A land. If we ask whether donkeys also plowed category B land, that is land where 12 furrows per nindan were seeded with plow animals, the numbers that are available today fail to give an answer. The oxen groups could have easily plowed all category B land in all households. In the household of Dumuzi they would have been quite busy but, judging on the basis of the numbers in TLJT 5 and text 37, they could have done the work by themselves: 56.3% of the land was of category A and 43.7% of category B; 44.74% of the plow groups were oxen and 55.26% donkeys. If we assume that a team of oxen plowed as much as a team of donkeys we can see that the oxen plows could have plowed all category B land. The guq-apin-gub-ba texts contain no information on the particular type of work done by these animals. It is known from a variety of documents that plow animals were used for plowing, harrowing, and seeding. In texts from Girsu that record labor expenses for tilling and seeding,65 work teams were called invariably guq, that is guq-giS.ur "harrow bull", guqTUG2.KIN "deep-plow bull", guq-ki-durg, "wet land bull", and guq-numun "seed bull". The designation "bull" designated draft animal teams of both species. If we ask whether donkeys were preferred for the presumably lighter task of seeding, the answer must be negative because of the lack of evidence for a division of labor between tilling and seeding within agricultural units. The administration and management of cereal cultivation was based on the one to one connection between cultivator and plow animal group throughout the season of soil preparation and seeding. The lack of mixed plow groups indicates that the donkeys performed the same tasks as the oxen. Donkeys are not as strong as oxen and essentially disappear as plow animals after the Ur I11 period from Babylonian documentation. Why were so many donkeys used as plow animals in the Ur III period? They seem to have reproduced more successfully. According to text 12 the 'donkey household' of Ningishzida had generated a surplus of donkeys so that 8 of them could be transferred to the household of Lugal-URUxGt (see below under redistribution among households) and, judging by the numbers of working animals in text 4 (cf. Table 9) and a transfer to another household, the household of Igalima had developed a surplus of donkeys in S 45.
14
Classification of animals in totals, grand totals, and summaries
Text 12, and most other texts of type A, used a threefold system of-totaling and summarizing animals by which they were reclassified into successively fewer categories. Table 8 shows how it was done. The totals of the category that is encountered first in a text will be called "totals"
Heimpel
Plow animaI inspection records
here. They were written with a ligature of the signs Su and LAGAB. It is commonly read Su@in, but since this transliteration does not account for the ligature I use the transliteration Sunigin instead. Most designations found in the itemizations are also found in the totals. Left out were the often rather lengthy designations for causes of replacement, the identification of the liable persons, and the time elapsed since commencement of liability. For example the "2 year cow yet to be replaced since last year (by) the cultivator (and) bull lieutenant" of the assets section of group 9 became a "2 year cow to be replaced" (Bb-2 su-su) in the total. Also, the notation "nu" was not repeated. Thus the "non(present) siKKum stallion" of the assets section of g o u p 6 became simply a '%iKKumstallion" in the total. The two total (Sunigin) sections which concluded the descriptions of the divisions were combined to a grand total. The sign for grand total is a ligature of the signs Su and N I ~ I N The . sign N I ~ I Nitself is a ligature of two LAGAB signs. The whole complex is commonly transliterated Su-nigin. It is conceivable, but unlikely, that the two kinds of totals which had clearly defined separate functions were expressed with the same word in two different writings. I therefore assume that the ligature of LAGABLAGAB was pronounced kilib and transliterate the whole ligature complex as Sukilib. In grand totals the designations of substitutions were left out, the "siKKum donkeys" became simply "donkeys", and the category of supplements was integrated into the category of assets. In a third step, the grand total was summarized. Summaries were called NI~IN-bawhich is read commonly kilib-ba. At this stage only 3 categories per species were distinguished. They will be discussed next. 15
Plow animal inspection records
2ade
5 diir 1 dur-1
1 ade-mar-ga 1 diir Su-gi4 giS-56 1 dur Su-giq gub-ba-8m Kag-a-tur engar
2 mares 5 stallions 1 one year stallion 1 milk mare foal 1 Su-gi4 stallion (fit) for the yoke 1 Su-giq stallion is standing Katur cultivator
The division of working and not working animals in plow animal groups is tabulated in Fig. 7: Figure 7
working
-
in summaries
in grand totals
Bb-gu4-giS1anSe-giS
Bb(-m6h)lade ,guq-giSldur
I
BblanSe/guq/dur-Su-giq gig-&
not working
mar-guq-@-a
BbIanSe-2, guddiir-2
Working and not working animals
The categories of the summaries were (1) rib-guq-gig "yoke cows (and) bulls" or anSe-giS "yoke donkeys9,66 (2) amar-guq-M-a various calves" or amar-anSe-hi-a "various foals", and (3) guq/Bb/an8e/dur Su-giq "Su-gi4 cows/bullslmares/stallions". By matching the numbers of these categories with the corresponding numbers of the animals of the more numerous categories in the grand totals we can fix their semantic range: any h-giq animal in the grand totals remained a Su-giq animal in the summary. "Milk calflfoals" and "1 year" animals were classified as calveslfoals. The 8 amar-guq-@-aof the summary of text 12 correspond to 1 Bb-1 + 4 guq-1 + 3 Bb-amar-ga of the grand total. The "two year" animals and adults of the grand total were summarized as working oxen, i.e. Bb-guq-gig It is surprising that all "two year" animals were classified as working animals because of the existence of "two year" animals which were qualified as gig-Sk "(fit) for the yoke". This qualification makes only sense if the "two year" animals that were not so qualified were not fit for the yoke and did not work. Comparison between grand total and summary of texts 5, 10, and 31 (cf. Table 9 for the first two texts) shows that, contrary to text 12, the unqualified "2 year" animals counted as expected among the calveslfoals (amar) and those qualified as gi8-~& "for the yoke" among the working animals. Su-gi4 animals were also, but rarely, qualified as gig-R "for the yoke". There are two references in the corpus, namely text 37 VID 14 and the description of group 4 in text 4: 66
Only text 12 deviated from this scheme by counting "two year" animals as working animals.
16
Norm of working animals in a group
The summary of the number of plow animal groups in text 1 reads as follows: the plow bull (groups) are 15 guq-apin-bi 15-im the plow donkey (groups) are 17 anSe-apin-bi 17-8m 7 bulls each, 8 donkeys each. guq 7-ta a 6 e 8-ta In the individual group descriptions of the text the number of working animals was in most cases less than 7 oxen or 8 donkeys. The summary statement can therefore not be a characterization of the plow animal groups as they appeared in the text, but should be a norm. Its existence can be demonstrated with the help of text 31. In this text, the notation 16-i plus number was occasionally inserted at the end of a group description, immediately before the name and title of the cultivator. It did not, as in the other texts, designate deaths. The sum of the assets and supplements of working animals and the 16-i number equal the norm of 7 oxen or 8 donkeys. Cf. e.g. the description of the 9th group (LU 3-10):
Heimpel
Plow animal inspection records
2 mares 3* dur 3 stallions l*dur su-ga engar 1 stallion replaced (by) the cultivator gub-ba-hm assets 1 dur ba-TIL 1 stallion dead libir-hm is old 16-i 2 deficit 2 d~tu-gu10 engar Utumu cultivator The six working donkeys of the assets section plus the 2 unspecified animals of the deficit add up to 8, the norm for donkeys. The numbers of working animals in assets and supplements sections and the deficit number of fully preserved group descriptions are given in Table 10. They add up to the norm without exception. Judged on the basis of closeness to the norm, the guq-apin-gub-ba texts form a rather diverse group. There are some texts in which the majority of groups is below norm. Such a text is # 1. The numbers of working animals per fully preserved group description in the sequence of the text are (asterisks identify donkey groups): 4-6-8*-7-5*-6-4-5*-3-5*-6-7-5*-2*-2-6*-3. Three groups conform to the norm, the others are below norm, several considerably so. Text 2 has groups that correspond to the norm or are slightly below or above it: 8-6-7-7-6-5-7*-6-8. Text 4 has the least consistent numbers. There were 6 groups with numbers above the norm, 3 at norrn, and 9 below norm. 2* erne6
17
The ideal composition of a group of working animals
There probably existed an ideal of working animals in a plow animal group. Such an ideal would have been realized when a group was formed from scratch and in one stroke under conditions of unlimited supply. According to text 4, two newly formed groups consisted of "6 bulls of the house of pens from the NaGABtum" (6* gu4 6-thr na-gib-tum-ta); according to text 12 another pair of newly formed groups consisted of 8 stallions each, 4 from the NaGABtum and 4 from another household. The two donkey groups may well represent an ideal: an all male group of 8. The two oxen groups of 6 males were 1 animal short of the norm and could thus have been only partial realizations of an ideal. Among the great variety of compositions of oxen groups one can perceive a tendency to place one cow into oxen groups. The best data come from text 12. 7 of its 13 oxen groups consisted of six bulls and one cow; the details of supplementation and substitution are as follows (I and I1 refer to the divisions of the household, arabic numbers to the groups within these divisions): I 1: the one cow was a supplement I 8: there was one cow, the supplements were bulls 19:sameasI8 115: sameas1 1 II 7: the one cow was substitute for a bull
Plow animal inspection records
Heimpel
II8:sameasIl 119: same as 117 Other groups show that there was no strict rule. All male oxen groups existed, as I 5 and 114, and these groups were formed partially by supplementation which would have allowed addition of a cow. The seeming preference of groups of 6 bulls and 1 cow is intriguing insofar as it may be connected to the curious fact that the norm of working oxen in a plow animal group was an odd number.
18
Not working animals in the plow groups
The numbers in Fig. 8 for animals that were part of plow animal groups but too young for work come from the summary sections of well preserved texts of type A. The numbers in ( ) refer to the working animals.67 Figure 8
text
donkeys #
%
1 (14) 16 (70) 16 (70) 66 (124)
7.1 22.9 22.9 53.2
4 (15) 2 (63) 3 (78)
26.7 3.1 3.9
On the average, about one fifth of the plow oxen in a household were too young for work. But the numbers are quite disparate and suggest that the population of young animals in the groups was conditioned by a complex set of circumstances. The numbers for donkeys are even more disparate; more than half of the donkeys of the 'donkey household' of Ningishzida (text 10) were too young to work; according to text 12, on the other hand, hardly any young donkeys were kept. In the grand totals of the same texts the animals were distinguished by age and gender. The distribution is tabulated in Fig. 9. 2, 1, ga, designate "two year", "one year", and milk class animals; 37a the section of the household of Nanshe, 37b that of Shulgi).68 The numbers suggest a pronounced attrition rate. Among the female oxen and the donkeys, there were more milk class than "one year" animals, and more "one year" than "two year animals". Especially dramatic was the decline from "one year" to "two year" donkeys of the household of Ningishzida in [ ] according to text 10. Most "two year olds" seem to have entered the work force early. The high numbers for "one year" male oxen is remarkable.
,
-
Heimpel
Plow animal inspection records
calves was recorded. Correlation of the two sets of numbers yield an approximate, but not a fixed, proportion. Cf. the 5 sets of numbers in Fig. 10:
Figure 9
female oxen
male oxen
text 2
2 1
1 3
ga 9
2
5 10 11 31 37a 37b
3 2
3 4
3 2
-
-
4 2
2 8 8
2 4 4 3 12 7
sum
16
32
47
female donkeys
Figure 10
male donkeys
1 7
ga
2
1
ga
2
1
ga
-
-
-
-
-
-
-
3 7 8 1 8
1 1
1 16
4 18
2 1
2 11
6 19
-
-
-
-
-
2
2 1 3
1
4
2 1 4
6
4 5 4 8 10 11
10
[
23
58
44
8
-
2 8
1 2
I 21
30
8
22
39
Bu-gi4 oxen and donkeys occur steadily and in small numbers, many in withdrawal
sections, none in deaths sections. The lack in deaths sections is noteworthy considering the fact that most h-gi4 animals were probably old animals. It indicates that they were not kept in a plow animal group but withdrawn shortly after exhibiting the negative quality described as Bugi4-
19
Plow animal inspection records
l lcimpel
Barley of the calves
It is plausible that "2 year" animals were kept in the plow animal groups: they stood in the 3rd year of their life and would soon join the workforce. But why were milk class and h-gi4 animals kept there? As indicated in the previous section, ru-& animals were quickly withdrawn and so need not have caused an additional expense to the cultivator. Concerning the "one year" and milk class animals, we might speculate that their stay in the plow animal group, their growing familiarity with conspecifics, oxdrivers, and the rhythm of agricultural life generally primed them for later work.69 Yet it is hard to imagine that the benefit of such priming outweighed the considerable expense of feeding and caring for the not inconsiderable numbers of young animals in the group. Of course, they had to be fed and cared for by somebody. It could have been done within the confines and responsibilty of the NaGABtum, and it is likely that many of the adult animals that came from that facility to the plows were indeed fed and cared for there for some time. Yet others, begining from their first year of life, were turned over to cultivators and became part of their plow animal groups. The cultivators presumably fed them and cared for them. They also received barley, called "barley of the calves" (Be amar-ra), probably in partial compensation for the expenses. In the seed-and-fodder texts the barley of the calves was routinely listed after barley expenses for seed, plow animal fodder, and wages of human labor in the formula "n calves of n liter each". In TUT 5 S 47, the number of plow groups per household and the barley of the
household Ningirsu commander Nanshe Ninmar Shulgi
units 62 40 50 60 50
calves 21 10 30 10 18
On the average, about 1 calf corresponded to 3 agricultural units. Since most groups had non working young animals, and many of them several, only a fraction can have been eligible for the "barley of the calves". Text 6 in our corpus is dated to the same year as TUT 5 : it recorded the groups of the household of Shulgi. The text is a large fragment of a tablet with 16 columns but the sections of the totals and grand totals are so fortunately preserved that all subadult animals are accounted for. Among them we can look for the 18 calves of TUT 5. The subadults in the totals are: XI1 2'-8': 3 two year bulls 3 ~ 4 2 1 i b 2 sag guq-glH-lb* 1 two year cow, substitute for a yoke bull
1 guq-gi~sag gu4-2-Sb* 1 yoke bull, substitute for a 2 year bull 2 ab 1
2 one year cows
6 ~ 4 1 9 Bb amar-ga*
6 one year bulls 9 cow milk calves 6 bull milk calves
6 guq amar-ga*
XIV 6'-13': 1a d e 2 4 ade 1 11.5 dur su-su* 1 dur 2 gig-Sb* 1 a d e 2 Su-su* 2 dur 1 4 anSe amar-ga* 2 diir amar-ga*
1two year mare 4 one year mares 11.5 stallions, to be replaced 1 two year stallion for the yoke 1 two year mare, to be replaced 2 one year stallions 4 mare milk foals 2 stallion milk foals
Heimpel
Plow animal inspection records
The list contains 42 subadults. The only obvious group of 18 among them are the unqualified "1" and "2 year" animals (indicated by asterisk). The amount of barley per calf was 120 liters in seed-and-fodder texts of the years 3 337O, 3 3671, and S 4472, and 90 in the texts of 4173, 3 4374, S 4575, 3 4776, and AS 177. Some texts itemized the different rates for calves and foals. Figure 11 text
date
L.965478 SNAT 75
[
1
AS 2
Plow animal inspection records
l lcimpel
deficit 1 mare deficit 1 two year stallion la-i 1 dur-2 Texts 12 and 15 marked deficit constituting deaths by using the designations ba-TIL for the singular and ri-ri-ga for the plural.80 Cf. the following examples from text 12. 17: 16-i 1 guq-giS ba-TIL deficit 1 yoke bull dead 16-i 1 a d e
1114-5: calves
foals
120 120
90 60
There is no evidence that calf and foal fodder was given to the cultivators above and beyond the barley of the calves, so that we can assume that it represented the annual total. I do not have any clear criteria by which to measure the degree to which 120 liter for oxen, and 90 or 6 0 liter for donkeys, covered the expenses. The young, as well as all other, oxen and donkeys had probably suitable pasture in fallow fields and the nearby marsh, and stabling and herding may have been a minimal expense, so that one third or one fourth liter fodder daily throughout the year may have been a fair compensation.
IV Loss and replacement, withdrawal and supplementation 20 Deaths The term "deaths" is used here in a technical sense for animals that died of causes that exempted their caretakers from the liability to replace them. Three aspects are treated in this section: I the notation of deaths, I1 the requirement of restoring the skin of the dead animal to the administration, and III the cause of death that constituted deaths.
I As in other animal inspection records, deaths were recorded in a separate section that was placed after the withdrawal section. The numbers of dead animals were written with the letter stylus. The texts used several designations and notations. Most introduced the deaths section with the formula 16-i "deficit" and ended it with 164-gm.79 Cf. e.g. 1 I 9-13: la-i 1 6b-ma deficit 1 cow 1 guq-gig 1 yoke bull 1 Bb-3 1 three year cow 1 three year bull 1 gu4-3 18-i-gm is deficit If the section consisted of a single entry, the concluding 16-i-gmwas often left out. In text 5 16-i-h was left out but the initial 1B-i was repeated in every entry. Cf. e.g. V 14-15:
I 19-22:
18-i 2 anSe-2
deficit 2 two year mares
ri-ri-ga
dead
18-i 1 hb-me
deficit 1 COW 2 yoke bulls 1 one-year cow for the yoke
2 guq-gig
1 rib-1
dead In texts 31, 32, 34, 37,38, and 47, la-i was not used as notation for deaths. As before, death of one animal was indicated with ba-TIL, death of 2 or more with ri-ri-ga. The three types of notation follow each other chronologically. The first type was used from 3 41 to AS 2, the second in AS 2, the third in AS 5. The use of the formula 1B-i "deficit" in deaths sections betrays their original function: they constituted the place in the form where animals were registered that counted as deficit for the administration. Normally the deficit consisted exclusively of deaths, and accordingly the sections were, in effect, deaths sections. For an example of a combination of deaths and other deficit cf. section 24. 18-i ri-ri-ga
I1 The skin of a dead plow animal that belonged in the category of deaths went to the administration; that of a dead animal that was to be replaced did not. The principle is clear and logical and needs no proof. Yet since the proof is included in the information of two Urnrna texts it shall not be suppressed. A plow animal group of a cultivator named Aakala was registered in (Urnma) text 7.*l It includes the following two entries: dead: 1 yoke stallion, to be replaced (by) the cultivator TIL 1 dur-giS su-su engar dead: 1 two year mare TIL 1 em% mu-2 The second animal counted in the category of deaths, and its skin had to be delivered to the administration. In the comprehensive deaths register of the same year Aakala was indeed registered as the source of 1 donkey hide.82 ~L 86 and MVN 5, 180 from Girsu illustrate that the expense of plow animals by the administration to cultivators "for cause of having died" required the receipt of the hides of the fallen animal, and that the plow bull scribe (dub-sar-guqapin) was responsible for producing the written receipts. The texts seem to be audit reports. &L 86: 1 bull 1 donkey 1 guq 1 a d e AS 8 mu en ~ r i d u ~ 2 bulls 2 donkeys 2 guq 2 a d e Ss 1 mu dSu-d~uenlugal
Heimpel
Plow animal inspection records
16 guq 6 anSe mu mai dara-abzu
16 bulls 6 donkeys Ss 2
ba-ab-dug mu ri-ri-ga-S& engar-ra ba-na-zi kul-bi la<-ba>-gid-da-S& guq anie-bi lul-la dub-sar guq-apin su-su-dam 6d~mar-d~uen
for cause of having died were expended for cultivator(s) because their skins were not received the plow bull scribe has to replace these bulls and donkeys ... house Amarsin
Plow animal inspection records
l leimpel
1 male goat for the "oath taking of bull lieutenant(s) and cultivators" (nam4rim kug-da nu-bandaguq u engar-e-ne).84The oath may be interpreted as an annual ceremony when bull lieutenants and cultivators swore that they did not misrepresent the death of an animal in their charge.
21 Death rates In order to get an idea about death rates among plow animals, I have added the numbers of assets, withdrawals, and deaths from grand totals and summaries of oxen and donkeys of suitable texts and computed the percentage of death from deaths sections and replacement entries in Fig. 12. The death rate appears to be high. Yet the data are thin and may not be representative.
The same type of irregularity was uncovered in the household of Nindara. Again the plow bull scribe was found liable.83
111The principal criterion for deaths was defined as 8-sig = asakku. Cf. MVN 7 , 2 0 3 S 30 (or IS 5 ) : 1 guq ba-TIL 1 bull, dead guq 8-sig-ga ba-TIL-a M a , the son of the gardener, nam-6rim-bi swore 6 d~in-marld-ka in the house of Ninmar Al-la dumu Sandana to the effect in-kug that the bull died of Asakku zi-ga expense Da-da of Dada Asakku was not a specific cause of death or a specific disease (or a specific demon) but a general term for the kind of death that could not be blamed on the caretaker of the dead animal. For example, the deaths among the goats of the palace and the temple households of Girsu in the uzud-gub-ba text DAS 50 IS 3 was characterized in toto as caused by Asakku: kilib-ba 225 uzud-hi-a ri-ri-ga 8-sig "summary 225 various goats, deaths, Asakku". The same animals were qualified as ri-ri-ga or ri-ri-ga 8-sig in the itemization. The same inconsistency in designation is found in the Bb-gub-ba text TCT I 740. Compare e.g. ri-ri-ga 8-sig in the itemization in column VI 19 and ri-ri-ga in the corresponding subtotal in line 30. There was obviously no difference between deaths and deaths caused by Asakku. MVN 7 , 203 demonstrates that there was room for fraud in the determination of the cause of death of livestock, as might be expected. The cultivator or his oxdrivers were normally present when the animal died, and given the choice between having to replace it - which could cost as much as 8 shekel of silver or 24 hektoliter of barley - or returning its skin he must have preferred the latter. Whatever really happened to the animal could easily be obscured by skinning the animal, and the oath remained as only recourse for the administration. A passage from an Umrna text suggests that the administration attempted to shield itself in some measure against fraud by involving the bull lieutenant. SET 130:313-3 15 AS 4 recorded the expense of
Figure 12
22
text
oxen
donkeys
2 4 5 6 12 27
14.2 12.9 14.6
15 19.8 5.9 23.8 14.3 7.9
13.3 18.2
Replacement, including a note on the nu gloss
Replacement was a serious issue which took up much room in the texts. It could impact severely on the fortunes of managers and administrators from cultivator to bishop. We have seen above that an audit revealed irregularities in the documentation of the receipt of skins of fallen plow animals and resulted in plow bull scribes having to replace dozens of animals (section 20 11). The impact on cultivators is possibly best captured in the subscript of the Umma text SNAT 509 SS 6 which listed amounts of barley that were taken out of the allotments of cultivators and withdrawn from the wages for work done by their wives and children: wives and children of cultivators dam dumu engar 6-hul-a-keq-ne of ruined houses for cause of having to replace bulls mu guq su-su-da-S& The designations of replacement specified a time relative to the incidence which triggered the obligation to replace, the liable person, and sometimes the cause of replacement. If the obligation to replace an animal was satisfied, the animal was called "replaced" su-ga. If the obligation to replace was established and the animal was not replaced by the time of recording, it was called su-su "to be replaced". After one year it became su-su nu-su im-ma, literally "to be replaced, not replaced, last year", in effect "one year overdue". This status is frequently attested and seems to have been commonplace. In text a we find su nu-su instead of su-su nu-su and a
Heimpel
Plow animal inspection records
contrast beween su nu-su libir and su nu-su im-ma. There is too little of the text preserved for a meaningful context, so that we are left with the literal meaning which is problematic ("oldAast year's not replaced replacing"?). Animals that were 2 and 3 years overdue are attested rarely.85 Text 26 rev. 5 may even have recorded an animal that was 4 years overdue.86 The managers and administrators of a household, from the cultivator to the divisional head, and occasionally also persons from outside the household (the NaGABtum, repsectively its scribe, a man of the siKKum, and a shepherd) could be found liable for replacement. Their rank and the frequency of incidence of their responsibility is summarized in Table 11. Frequency of incidence correlated with rank in the following way: as a group, cultivators replaced most often, bull lieutenants less, bishops and commanders even less, and plow bull scribes, 10-bull-scribes, and others least. If we adjust for the relative number of individuals in these ranks, and leaving aside cases of shared responsibilty, the situation changes. The corpus includes 107 replacements of cultivators and 35 of bull lieutenants. Assuming sin average of 4 cultivators per bull lieutenant, a conservative estimate, the individual bull lieutenant was more likely to become liable for replacement than the individual cultivator. There were on average 3.8 bull lieutenants to 1 divisional head in our corpus. Divisional heads replaced 18 animals, less than one half the number of bull lieutenants. The individual divisional head was clearly more likely to become liable for replacement than the individual bull lieutenant. If the numbers are not grossly unrepresentative we can conclude: the higher the rank the costlier the responsibility of replacement. Often, cultivators and bull lieutenants shared responsibility, once a "lo-bull-scribe" was a third liable party. It is likely that such arrangements were the cause for the half animals that are recorded occasionally.87 Animals to be replaced were listed in the assets section even if they were not physically present when the text was recorded. They also counted toward fulfillment of the norm of working animals. Cf. e.g. the 2nd group in text 3 1: I* Bb-ma 1 COW I* guq-gig 1 yoke bull 1 yoke bull, replaced (by) l * guq-giS su-ga Sabra the commander 1 yoke bull, replaced (by) I* guq-giS su-ga nu-bhnda-guq the bull lieutenant 1 yoke bull - non(present) I* nu guq-gig su-su engar to be replaced by the cultivator gub-ba-8m is standing
...
...
Ub[ir-&n]
is old 2 yoke bulls from the count of the NaGABtum Ur[ cultivator]
2 guq-[gig ni-kas71
na-lg6b-hlm-taI
Ur-[
engar]
l Ieimpel
Plow animal inspection records
Only if we count the yoke bull that was to be replaced by the cultivator do we obtain 7 working animals. According to the format of the text, 16-i 1 "deficit 1" would have been placed before the name of the cultivator if the bull to be replaced had not counted toward the norm. The physical absence of animals to be replaced - which is inherent in their status - was indicated in this example, as in most texts of type A, but not in text 12, by a small nu "not" that was placed after or below the number sign. Lafont observed that it was written with a smaller character and "aprhs In the texts of type B the nu gloss is not found; instead, the number was written with the letter stylus and thereby associated with dead and withdrawn animals whose numbers were also written with the letter stylus, the common denominator being absence. I assume that the nu gloss designated also absence in the case of the "donkeys of the siKKum" and the other references, which are tabulated in Fig. 13: Figure 13 reference
animal gu4 Su-gi4 guq-ku-ta sale-a guq-su-ga Igi-LUM.LUMg9 guq-gig guq-1 su-ga engar sai guq-2-S& guq-fiS engar su-ga nu-bhnda-gyq Bb-1 su-gal engar?
Replacements were closely monitored as is demonstrated by TUT 93 AS 3, the fragment of a VIII column tablet with the subscript guq a d e su-su C d~in-gir-su"bulls and donkeys, to be replaced, house Ningirsu", and the difficult texts STA 27 AS 1 and ATL 99 AS 3. The latter two texts attest that replacement in kind was accepted. For that case equivalences were established. According to STA 127 I11 3-5, adult oxen and donkeys were equivalent to 6 gin silver or 6 gur barley and calves and foals (amar) to 2 gin silver or 2 gur barley.
23
Causes of replacements
In most cases the cause of replacement was not mentioned, which is unfortunate because it deprives us of the information that would allow determination of the rules of liability. The cases where the cause of liability was spelled out are arranged below according to their nature: deaths, loss, theft, predation, distress, miscellaneous. Deaths: 31 VI 19: 1 guq su-ga Ba-zi sipa mu dumu-ni guq i-gaz-S&"1 bull, replaced (by) the shepherd Bazi because his son killed a bull". CH $255 reckons with the possibility that somebody kills a bull by hitting it. It is not easy to do that without using much strength and a mace or similar weapon.
l lci~t~pcl
Plow animal inspection records
9 1 21': [I] gu4 su-su mu p N ] nu-bhnda-guq-keq [x (x)] i-na-16 ba-TIL-a-s&"1 bull, to be replaced because the bull lieutenant hitched it for him [ ] (and) it died". TUT 93 I 3'-6': 1 dur [sul-su engar mu mar-[x] IbP-I6 ba-TIL [x (x)] nu-gu7-a-@ "I stallion, to be replaced (by) the cultivator because he hitched it [to] a cart(?), it died, (and) ... did not eat". Or: "because he hitched it [to] a cart (and) it died, and not (because) [a beast (ur-re)] ate it". This is one of the rare references from Ur 111known to me which may attest the use of a cart. 12 I 2-3: 1 su-ga engar mu az-a TIL-a45 "1 bull for cause of a bear having killed it7'. This is a bizarre case. Syria and Persia contained until fairly recently a few bear habitats. But it is hard to imagine that these forest animals ever roamed the countryside of Lagash. It may have been the bear of an entertainer ~q-da-tuSthat got away. -a after nz is perhaps a phonetic rendering of the agentive postposition. Loss: Disappearance as cause of replacement (u-gu d6-a "lost7') was fairly common; it is mentioned for 22 animals in our corpus, understandably without any indication of the circumstances. An illustration for a case of lost and found is contained in text 2. In each division of the household of Gatumdu the withdrawal of 1 mare was recorded and described as follows: 1 anSe ph-da im-ma
1 mare found last year @r ~a-tar-d~a-li intermediary Katarbau mu ~ r - ~ ~ dumuu e n because Ursin, son of the king, lugal-keq Su-na brought it back into ba-a-gi4-a-55 his possession I assume that 2 mares had been found and given to the household of Gatumdu in 42, the past year (pl-da im-ma) from the perspective of the text, and the heads of the divisions split them among themselves. It turned out that they belonged to Ursin, a prince, and they were handed over to an intermediary to be returned to their rightful owner. The last, apocopated and badly phrased, law in YOS 1,28 reads: tukum-bi guq-N .g'~n-na guq u-gu ba-an-dC guq guq-gin7 "If a bull that is wandering about, (and that) bull was lost, bull like bull (it shall be replaced)". Theft: 4 14'-5' seems to be the only example for theft as cause of replacement: [I 1 sag a d e 1 [ 1, substitute for a mare (that was) [S& I]u-la-ga Ah-a [ dlur su-su
stolen by a rustler ?
I la-ga x ? A l a p person was a type of out-law, perhaps specifically a cattle rustler.90 According to the messenger text MVN 7,98 a certain Gurruru traveled Iu-la-ga dabs-dabs-dl "in order to seize a rustler". SNAT 210 = E. Sollberger, AOAT 25,447 records the following sworn promise of a certain Urd~bshen:~l [
uq-Hakar iti munuq-gu7-S& Iu-la-ga udu-zm-a LU-d~anna-ka
I will have brought by the beginning of month IV the rustler of the stolen sheep of Lunanna.
Plow animal inspection records
Heimpel
If not, I will be Iu-la-ga-bii-me-a that rustler. NG 169 contains the case of a bridegroom who engaged i11laga lifestyle which seems to have been reason for annulling the betrothal (lines 21-23): nam-lu-la-gaba-a-hul-a-82nam-dam-nita Nin-hi-li-sii-keqtug fb-da-an-ur "because he was mined by living as a rustler, Ninhilisu was released from being his wife". Predation: Loss to wild animals is attested 13 times in the corpus and twice in TLJT 93, The formula for it was ur-re/e gu7-a "eaten by a beast". Most of the plow animals that were declared "eaten by a beast" were adult. The only predator of adult oxen and donkccys in Babylonia was the lion. Among the Urruna texts replacement as result of predation of a Lion was mentioned once (text 6, group 6). In that case the word ur-m& was used.92 Tt is remarkable how common predation by lions was.93 It provides a telling background for Shulgi's praise as lion hunter in hymn R, and it demonstrates that large tracts of the southe1-nBabylonian countryside were wild and uninhabited country, as they are today. Cultivators94, cultivators md bull lieutenants,95 bishops,g6 and thc administration could be responsible for replacement caused by predation. The last case is attested in 35 IV 4-5: a mare and stallion that had been "eaten by a beast" were recorded as deficit (la-i). The Sumerian law AS 16,7 III 9'-13', stated the case of a lion - here called as expected ur-m& - having killed "a bull 'by' its yoke" (gu4 BiSgudun-b6).The preservation of the negative prefix of the verb of the apodosis indicates that the caretaker was not responsible for replacing it. YOS 1, 28 $8, a Sumerian law in questionable Sumerian, seems to say that a bull that was killed by a lion while wandering about (nigin-na) was to be replaced by its keeper. CH $244 on the other hand stated that the renter of a bull that was killed by a lion ina ,Grim ("in open country"?) was not Ijable. Distress: 4 YJII 6: 3* nu guj su-su saraga nu-lbanda-yu.pke4 mu ge-kurg-ra engar-ra nu-nastam-ma46 "3 non (plcsent) bulls, to be replaced (by) the bishop because the bull lieutenant did not give the cultivator the allotmerit barley". The cultivator withheld 3 bulls as distress in lieu of the barley that the administration was obliged to give him as allotment. The bull lieutenant was supposed to hand it over to the cultivator. The fact that the bishop, not he, was responsible for replacing indicates that the allotment barley was issued by the bishop and distributed by the bull lieutenant. Maekawa demonstrated that the allotment barley was l/lOth of the harvest of a domain unit that was worked by a cultivator who held 12 iku as allotment plot, or 1/20th of the harvest if he held a 6 iku plot.g7 According to the standard yield ratio of 30 gur per bur, a cultivator with 6 iku could expect an allotment of 10 gur. Accoriling to the replacement value of 6 shekel silver, or 6 gur barley, for 1 bull that was used in Lagash, 10 gus barley corresponded to 1 and two thirds bulls.98 With three bulls the cultivator held a value of 18 gur which was presumably more than his allotment barley. CH $241 refers to a bull as distress: "If a person takes a bull as distress t ~ eshall pay one third pound silver", if 11: had no valid elairn. In our case the cultivator had a valid claim, he kept 3 bulls, and the bishop the barley. thus becoming liable for the replacement of the three bulls. Miscellaneous: 9 IV 13 and V 6: I* nu dur su-su US-e ttim-ma "1 non (present) stallion, brought to US". It is very tempting to understand this as a case where a stallion was removed from the plow group because it was needed for breeding: gig-e "for the penis", i.e. for stud mu-tum nu-mu-ttim
Heimpel
Plow animal inspection records
service. The interpretation would mean that plow stallions were not necessarily castrated. I have long sought for, but not been able to obtain, reliable information on the existence or nonexistence of uncastrated males as draft animals from written material. Convictions pro and contra are occasionally expressed. The best I can offer is the statement of an old Black Forest farmer concerning oxen: "wenn er gutartig ist, spannt man einen Muni schon einmal an" [a docile bull (i.e. uncastrated adult male bos taurus)99 is occasionally hitched]. 5 I11 19-21 (cf. collation Maeda): l* anSe su-ga engar sag anSe-Sk mu su su-a-S& 6 V 7-8: I* nu gu4 na-grib-tum nu-sum-ma mu-3 su-su ~ i i - d ~ i n - S u b u"1r non (present) bull, not given (by) the NaGABtum since 3 years, to be replaced by Luninshubura". This is not the only time that an animal was "not given (by) the NaGABtum.lo0 17, 22: l * anSe-1 sag anSe-85 mu in-su sag &rum-ma ba-TIL-R 27, 5: l*gu4 su-su engar mu hug-gri ~ r - d x - x - ~
24
Deficit replacements
We have seen that animals that had to be replaced were recorded in the assets sections. Occasionally, however, animals that had to be replaced were recorded as deficits. Cf. e.g. the deaths and deficit section 4 IX 8-10: 16-i 4 anSe su-su nu-su deficit 4 mares yet to be replaced engar ri-gu dC-a (by) the cultivator, lost 1 dur su-su nu-su 1 stallion yet to be replaced engar u-gu dC-a (by) the cultivator, lost 2 dur 2 stallions 16-i-iun is deficit The administration had apparently no expectation that the cultivator would fulfill his obligation to replace 5 lost donkeys and reckoned them as deficit. The few other references101 for deficit replacements contain no clue about the particular circumstances. In the type C text 30 deficit replacements were totaled as 16-1su-su after the total of deaths.102 It is not likely that the administration simply waived the obligation of replacement. Indeed, a case of deficit replacement from Umma shows that a person liable for replacement did not replace in kind but had his family work off his debt (cf. section 46). The debt could also be paid in silver, barley, or wool as is attested in STA 27 and ATL 99.
25
Substitution
Substitutions were expressed with the fomula sag animal-& "substitute for animal". Their notation took up even more space than that of replacements. The data are collected in Table 14. Substitutions were often connected with replacements. Cf. e.g. 1 rib-mribsu-ga engar sag guq-giS-SG "1 cow, replaced (by) the cultivator, as substitute for a yoke bull". This should mean that the cultivator had to replace a yoke bull and offered a cow as replacement, and that the administration accepted the cow as substitute for the yoke bull. The same process was also
Heimpel
Plow animal inspection records
expressed in a different way: 1 dur sag gu4-gig-G su-ga engar "1 stallion as substitute for a yoke bull, replaced (by) the cultivator" (25 rev. 3-4). Another occasion for substitutions was the process of intramural redistribution for which see section 28. Cf. the following case from text 12: a yoke bull was transferred from cultivator Lugishbare to cultivator Lugalsipa; in the assets section of Lugishbare is found the entry 1 6 b - m a sag guq-gig-3 "1 COW as substitute for a yoke bull".lo3 The substituted yoke bull could be the very animal that was transferred to Lugalsipa. There were surely other occasions for substitutions, but the current documentation is too thin to uncover them. Compared to the animals which they substituted, the substitutes were: replaced 3 67 of same age and different gender 29 one year older1o4 7 two years older none three years older 29 one year younger 14 two years younger none three years younger 1 a Su-giq class animal 2 substitute for a Su-gi4 1 a siKKum donkey 14 of the other species
26
Withdrawals
Withdrawals were registered after assets and designated with the label zi-ga "withdrawal". The numbers of withdrawn animals were written with the letter stylus. In the majority of cases no animals were withdrawn from a plow animal group during a given year. When a withdrawal was made it consisted typically of 1 animal. According to text 37,26 animals were withdrawn from all groups of the households of Shulgi, Nindara, Dumuzi, Amarsin, Ninsun, and Gatumdu. According to the numbers collected in Table 15, withdrawals averaged 3% of assets. Oxen and donkeys of both sexes and all ages except the first year were withdrawn. Most were adult, less than 9% subadult. One "two year" and many adult animals were characterized as Su-giq. Most withdrawn animals went to the NaGABtum. This was expressed as "to the NaGABtum" (na-grib-turn-%), or abbreviated "NaGABtum". There, they were received by the "fattener" (kiiruSda) Si-KAK. Transfers to other households were registered as withdrawals in the household from which they were transferred (cf. section 29).
Heimpel
27
Plow animal inspection records
Supplementation
Sections consisting of records of supplernented animals were introduced with the formula libiram "is old". It identified the preceding entries as the state of affairs before supplementation. Supplement sections occur regularly in texts of type A. They are lacking in texts of type B and C ; the former rnay thus have been written before supplements reached the groups, the latter after supplements were integrated among assets. The goai of supplementation was the preservation of efficient plow animal groups, specifically the replacement of deaths, and the new formation of groups in times of expansion of arable land (cf. section 4). According to TUT 27 of AS 2, the 480 unit group employed 3679 oxen and donkeys that were organized into 480 plow animal groups in AS 2. The average rate of deaths was 10.5% according to the numbers collected in Table 15. It meant that the administration had to replace an average of 386 animals per year. The flow of these animals to the households is documented in the supplement sections. The information of these sections is assembled in Table 16. It will be treated according to the different sources of the supplemented animals. A special form of supplementation was redistribution of animals within a household and between households. It will be treated separately in sections 28-30. Table 16 (1): Most supplemented animals came from the NaGABtum. The designations vary: "from the NaGABtum" (4,7) "from the NaGABtum, from Urningishzida" (9) "NaGABtum, from the kurugda Si-KAK, intermediary Urningishzida" (9) "from the house of pens,l05 intermediary bishop of Ningirsu" (10) "from the count (of) NaGAB tum" (12,3 1,32,37) "from the house of pens, from the NaGABtum" (6 and 7) "from the house of pens" (2,6,39) "house of pens" (9) "from Si-KAK (9) "from the fattener (kurugda) Si-KAK" (5) Oxen came from the house of pens and the NaGABtum, donkeys came from the NaGABtum but not the house of pens. It indicates that the house of pens belonged to the NaGABtum but was a separate entity devoted to oxen. It was probably the central administrative unit for oxen where the decisions about the future of each animal were made: wether it was fattened for slaughter, kept for breeding, used for dairying, or for agricultural work. The fragment HLC 260 table 122 contains a passage that recorded the number of oxen of the house of pens of the year 9 47, 1329 cows and 600 bulls, and the animals among - them that were destined for agricultural work: ih-ba 111 guq apin-na si-ga i-g6l "1 11 bulls among - them are incorporated among the plows". The number represents probably the yearly total of oxen that were supplemented in the 480 unit group. 62 bulls were supplemented from the "account of the NaGABtum" in [SS 11 to households representing 244 units according to text 37 (cf. note 1 to Table 16).
Heimpel
Plow animal inspection records
The institution NaGABtum deserves a study of its own. For the time being cf. CAD. S.V.nagabtum, where many references are collected, principally from Drehem texts. M. Sigrist mentions the "royal NaGABtum" (na-gsib-tum lugal) in Uruk which is attested in BIN 3, 281 and, generalizing the reference, says: "Le roi avait dans le diffkrentes villes du royaume ses propres parcs h bCtail desquels il pouvait tirer les animaux qu'il voulait distribuer aux autoritks des temples du lieuV.l06The NaGABtum of the texts discussed here obviously had a much broader function. An "account of the NaGABtum" of the KA-la-ti type is preserved in MVN 6, 294. Table 16 (4): A considerable number of animals was "bought with silver" (ku-ta sale-a), especially in the year 3 47 according to texts 6 and 7. Text 3 of 43 identified the source of the purchase, a merchant by the name of Iluma. He also sold a "rear ox" in 3 42 to the household of URUxGt according to NATN 408: 1 guq-6b-ur 1 rear ox from Iluma merchant ki Ilu-ma dam-gar!-ta apin d ~ u g a l - ~ ~ ~ ! x ~ t k i - g & for the plows of Lugal-URUxGt received by Lugahpa cultivator Lugal-sipa engar i-dabs supervisor Abbakala commander ugula A[b?-b]a-kal-laga[bra] gir LU-d~in-gubur intermediary Luninshubur There is a small group of similar texts which augments the infomiation from the gu4apin-gub-ba texts and demonstrates that purchase of plow animals was always an important aspect of supplementation. 57 donkeys that had been bought for silver and brought from Uruk, with a royal messenger serving as intermediary, were received by Adda, the bishop of Inana, for the household of Igalima at the end of the year AS 3 according to TUT 49; Berens 36 AS 4 I11 10 recorded the receipt of 35 donkeys "from merchants for the commanders (mu gabra-nenesic-g&) for the "plows of Ningirsu;" and 68 oxen and 13 donkeys were bought for wool (sig-ta sale-a) "(for) the yokes of the house of Amarsin" (gig C d ~ m a r - d ~ u eaccording n) to HLC 164 table 103 (without year date). With these animals each of the 50 plow animal groups of the household (cf. Table 3) could have received 1, and more than half of them a second, animal as supplement. According to MVN 7, 111 IS 1 IX the sum of almost 18 pound silver was used for the purchase of the oxen and donkeys for the district Kinunir-Nina. The price for one ox averaged 7 shekel, for a donkey 5 shekel; that is 18 pound silver bought about 180 animals. It translates into about 1 and one third animals per plow animal group if we localize in the district the households of Nanshe, Dumuzi, Nindara, Gatumdu, and Lugal-URUxGt, representing 125 agricultural units in S 47 (cf. Table 3).1M A small group of texts of the year mu Shashru ba-bul (5 42 or AS 6) recorded the purchase of "rear oxen" gu4-6b-hr-ra.log Cf. e.g. SNAT 16 (envelope): 1 g. for the plow 1 guq-aib-hr apin-a from the merchant Luningirsu ki LU-d~in-gir-su-ta dam-ghr-ta tablet of Urnumushda dub ur-d~u-mu8-da bishop of URUxGt sanga U R U X G ~ ~ ~
Heimpel
Plow animal inspection records Heimpel
Table 16 (5): It escapes me why animals of the category "for cause of fallen cattle" mu guq ri-ri-ga-gbshould have been supplements. Their description resembles entries of replaced and to be replaced animals in asset sections. They were replaced by the commander Luzagesil09, a bishop and a bull lieutenant,llo and unidentified persons.111 The main purpose of supplementation must have been maintenance of groups with full norms of working animals, and thus the great majority of supplemented animals were adult and fit for work. Only rarely do we find other kinds of animals among them. Text 1 recorded supplements from gifts (a-ru-a) that had presumably been given to Ningirsu andlor other divinities of his household. Since one may not have been fussy about accepting votive gifts of questionable value,l12and since there may have been a scarcety of sources of supplementable animals at the time, we find as an exception a "2 year" stallion which was not yet fit for the plow (dir-2) and even a stallion and mare of the Gu-giq class. A hapax is the milk cow calf (ibamar-pa) which came from the house of pens and was supplemented in a group of the household of Shulgi (6 I1 6). Among the adult animals that were fit for plowing, the males outweighed the females by a larger margin than that of the average in assets sections. According to the preserved sections of text 37 for example, 63 bulls and no cow, 14 stallions and 3 mares, were supplemented.
V Other administrative procedures of plow animal management 28
Intramural redistribution
Many animals in supplement sections were qualified with the phrase "from PN cultivator". Cf. e.g. text 12 18-12: libir-im is old 2* guq-gig 2 yoke bulls ki ur-d~umu-ziengar-ta
I* ib-m& ni-kas7 na-gib-tum-ta
from Urdumuzi cultivator 1 COW
from the count of the NaGABtum Urdumuzi was the cultivator who managed group 8 (I11 14-31). In fact, every cultivator who was mentioned as source of animals in supplement sections managed one of the groups that were registered in the text. The movement of animals constituted apparently intramural redistribution. The pattern is shown in Table 12. It will be seen that redistributions involved the groups of one and the same bull lieutenant, or those of the two bull lieutenants of one division, or two bull lieutenants of different divisions. In one exceptional case a cultivator, the aforementioned Urdumuzi of group 8, received a supplement from himself via a bull lieutenant zi "1 yoke bull, from the other division: 1 guq-j@ s u - y G u - p nu-bhnda-guq W ~ r - d ~ u m u -engar-ta replaced (by) bull lieutenant Gugu, from cultivator Urdumuzi". The redistributed animals were not totaled as supplements even if they were registered in the supplement sections of the itemization, but disappear instead in the total of assets.
29
Plow animal inspection records
Redistribution and borrowing between households
Group 10 in text 12 was managed by the cultivator Lugalmud*; it was described as follows: deficit 2 yoke bulls 18-i 2 guq-gig ri-ri-ga libir-hm 4 dir ni-kas7 na-gib-tum-ta 4 dir apin d~in-gig-zi-da-ta
deaths is old 4 stallions from the count of the NaGABtum 4 stallions from the plows of Ningishzida
As can be seen from Table 12, three oxen had been withdrawn from the group of Lugalrnudah to serve as supplements for as many cultivators. The three oxen had constituted the assets of his original group. Two dead yoke bulls were left on his account. With all oxen gone, a new group, this time of donkeys, was formed from supplements and assigned to him. Four stallions came from the household of Ningishzida. Another batch of 4 stallions from the household of Ningishzida was used for the constitution of group 16. Group 10 belonged to the division of Urnumushda, group 16 to that of Narnrnah: the transfer from Ningishzida was evenly split between the two divisions. The transfer of animals between households was registered as a supplement in the receiving household, as here, and as a withdrawal in the household where the animals originated. The latter case is attested in text 4 which registered the plow animals of the household of Igalima (IX 4-7): 4 diir 4 stallions for the plows of Lugal-URUxGt apin d ~ u g a l - ~ ~ ~ x ~ t - g & Nam-mab i-dabs zi-ga
Narnmah received it withdrawal.
One household borrowed occasionally plow animals from another. Cf. the memorandum Gpigraphie 1,56 of 3s 2 IV 4: 4 harrow bulls, to harrow 1.62 h each 4 g ~ q - i i 4.5 r ~ iku ~ ~ gig-ir-ta 4 bulls to harrow 1.44 h each 4 guq 4 iku gig-iir-ta the bulls are to be handed back gu4 gu giq-giq-dam bulls of Urbau, commander of the guq ~ r - ~ ~ agabra - l i 6 d~inmarki house(ho1d) of Ninmar place of the bishop of Nanshe ki sanga d ~ a n ~ e plow land (managed by) bull lieutenant Luutu gina-gu4 LU-dutu nu-bhnda-guq field Gigara a-gi Gi-gira
30
Redistribution within the household of Ningirsu
The household of Ningirsu was the largest of its kind and it was probably also the most complex. In the context of redistribution of plow animals the divisions could behave like two
Heimpel
Plow animal inspection records
separate households. Texts 1 and 5 recorded the plow animal groups in the division of bishop Urbau. The administrator of the other division at the time was Utumu, bishop of Bagara and commander. In text I we find the following entry (IX 23'-24'): 1 dur 1 stallion ki ~ 6 - ~ ~ i n - g i r - s u - t a from Luningirsu apin d~tu-gulO-ta from the plows of Utumu
In text 5 VII 23-29: 1 stallion, replaced [(by) ] ki ~ f i - ~ ~ i n - g i rengar-ta -su from the cultivator Luningirsu 1 dur-2 gig-R sag guq-Si? 1 two year stallion for the yoke as substitute for a bull ki Ur-sukkal engar-ta frorn the cultivator Ursukkal 1 dur su-ga sag guq-62 1 replaced stallion as substitute for a bull ki Lugal-ni-zu engar-ta from the cultivator Lugalnizu apin d~tu-gu10 gab[ra-ta] [from] the plows of the commander Utumu As in intramural redistribution the cultivators from whom the supplements originated were mentioned by name and title, and as in redistribution between households the animals were designated as belonging to the "plows" of another administrative unit, and listed among the totals of supplements. 1 dur su[-ga
31
]
Dissolution of groups and change of managers
Plow animal groups were subject to constant change as animals died, were lost, withdrawn, redistributed, replaced, substituted, and supplemented. Normally these changes were balanced so that the group composition stayed close to the ideal. Not infrequently, however, group composition deteriorated so far that drastic action was needed. Unfortunately, the text corpus does not yet provide the density of documentation that would allow the reconstruction of the development that led to the dissolution of groups, or for that matter any developments over the years. So far, only short episodes can be grasped. Some are quoted below: (1) According to text 4, a group of donkeys managed by Henati, cultivator in the household of Igali~na,had ceased being viable. Its assets were two donkeys that had not yet been replaced by the bishop, 4 donkeys were withdrawn and transferred to another household, 5 lost donkeys were reckoned as unrecoverable and counted as deficit, and the rest constituted deaths. As a non functioning group it had no supervisor anymore and was placed as appendix at the end of the itemizations in the division of Adda (IX 1-12). According to VII 16-19 a cultivator by the name of Shagube "took charge of 6 bulls of the house of pens from the NaGABtum; they are the bulls of Henati" (6* gu4 6-tur na-gab-tum-tagu4 IjC-na-ti-kam Sh-gti-b6 idabs). It appears that Henati, after dissolution of his donkey group, had received of a group of oxen which had been newly formed from animals of the house of pens. But by the time the text was recorded, this group had been transferred from Henati to Shagube.
Heimpel
Plow animal inspection records
(2) The story of Urzikuma of text 4 was similar. He too had originally a donkey group. It had 4 adult donkeys - of which 1 was one year overdue from the NaGABtum - and 1 gu-gij animal as assets and deaths of 2 animals, and was attached in this state as a supplement to the group of Lugirnun who managed group 6 of the second division of the household (VII 6-14). Just as Henati, Urzikuma received a newly formed group of 6 bulls of the house of pens frorn the NaGABtum, but they were slated to be transferred to another cultivator. The latter had not been determined by the time the text was written so that the place where his name would have appeared was left uninscribed (VI 21). In the end, Henati and Urzikuma did not manage any group in the household of Igalima. Urzikuma appears as cultivator in the household of Ningishzida according to text 10. He must have been transferred from one household to another; the date of text 10 is unfortunatley broken, and it remains unclear in which direction the transfer was made. The cases of Henati and Urzikuma constitute the longest episodes in the corpus. They include three developments: dissolution of groups, assignment of newly formed groups, and change of managers. Dissolution and assignment of a newly formed group is attested in the case of the 10th group of the first division and the 6th group of the 2nd division in text 12 (see above in section 29). A nonfunctional group which had not yet been completely dissolved and was not reassigned occurs as first group in text 47: 1 two year mare 1 emeg-2 1 one year mare 1 emeg-1 deaths ri-ri-ga libir-hm is old 1 non (present) stallion bought for silver I* nu dur ku-ta-sale-a Ur-Sag-ga engar cultivator Urshaga The group consisted of deaths and a nonpresent supplement. Urshaga's was a ghost group which can only have had an administrative function. Change of cultivators of a group was often recorded.l14 It was designated with the formula guqlange PN1-kam PNz (engar) i-dabs "the oxerddonkeys of PN1 were taken in charge by (the cultivator) PN2". Once, in text 16, the formula i-dabs was left out. In text 19, the change was expressed in a curious way by designating the group as supplement coming from the former cultivator to join, as it were, a zero group of the current cultivator: libir-hm is old 6 yoke bulls 6 guq-gig gub-ba-hm are assets they were the bulls of Ursheila guq Ur-Le-ll-la-kam Kulikala took them in charge Ku-li-kal-la i-dab5 The reason for change of cultivators was never recorded. It may have been death or financial ruin of the original cultivator, or transfer between households. The only information we can wrest from the texts is to determine whether the cultivator who gave up a plow animal group was kept on the roster of the household. Even that is possible only in the case texts 4, 10,
H
Heimpel
Plow animal inspection records
and 12, which are well enough preserved for the complete roster of cultivators to be known. Henati and Unikuma of text 4 were not on it, neither was Urashnan whose animals were taken over. by A N . ( ? I R ~ - ~according [~ to text 10, and neither were Lugaldalla and Lupabilsag according to text 12 (cf. Table 5). On the other hand, Urmes and Urdumuzi of this text were on the roster. The donkeys of a cultivator Urmes changed to Luurasari, and a cultivator Urmes took the oxen of Lupabilsag; the oxen of a cultivator Urdumuzi changed to Lukunun, and a cultivator Urdumuzi managed the 8th plow animal group in the first division. Even if Urmes and Urdumuzi are very common names, we can be reasonably sure that these were, infact, the same persons because the scribes distinguished two persons of the same rank with the same name appearing in one and the same text by adding the father's name of one or both. Change of bull lieutenants in the supervision of a group is attested in four texts of type B. It was designated by the same formula and formed part of the subscript in texts 22 and 24: guq PN1-kam PN2 i-dabs. In text 28 ugula PN2 is found instead of PN2 i-dabs; and in text 47 ugula P N l nu-bhnda-guq instead of guq PN1-kam. Another way to designate the previous supervisor of a group or a block of groups seems to have been the placement of a ki PN-ta formula at the end of the itemization and before the subscript and, in cases where the subscript was separated from the itemization by a blank, before that blank. The placement of the formula suggests at first sight that it qualified the last group of the itemization, but the fact that it is not found in any other position indicates that it referred to all groups of the itemization. An exception is text 27 where the formula is found after each itemized group. The names of the persons in the formula are listed in Fig. 14. One of them was titled bull lieutenant. Ld-IB and Igishasha occur in two texts each, Ld-IB in two texts of the same date. If the transfer of the groups took place in the same year, Ld-IB would have been in charge of at least 10 groups at one time. He could not have been a bull lieutenant in that case, and may have been a 10-bull-scribe. Figure 14 text
group
1 19 1 all 120
1 all
name o f source
I LB-IB
I LB-IB Lugal-x-x
1
1
5
1
~ r - ~ nu-bhnda-guq ~ t u ~ r - ~ udumu t u ~r-d~amma Igi-Sag-Sag Igi-Sag-Sag
Heimpel
Plow animal inspection records
It is unfortunate that the realities behind these dry pieces of information cannot yet be perceived. I suspect that they will only become clear when the system of land distribution for cultivation is explained.
32
Inspections
The Umma texts characterized the process which resulted in their recording as "inspections" (giirum); since these texts are very close in subject matter and format to the guqapin-gub-ba texts from Girsu, it is hardly plausible that the latter were products of a different process. The term gururn appears in text 1 XI1 20-21 where a mare was qualified as egers drumta "from after the inspection". There was apparently enough time beween inspection and recording to fit in late-comers. HLC 346 table 133, a tag of a tablet box, mentions an inspection in connection with plow animals: tablet box pisan-dub-ba d r u m aka done inspections plow bulls having passed guq-apin dib-ba balanced accounts of wages ni-kas7 aka ai C d~umu-zi house Dumuzi C d~in-dar-ra house Nindara C d~in-mar~j house Ninmar house Shulgi cdSu~-~i C d~mar-d~uen house Amarsin and house Gatumdu a C dgsi-tam-du10 i-gril are in it. The terse document did not express the syntactic relationships between the phrases and thus leaves us guessing whether 1,2, or 3 types of texts were in the box, "done inspections of plow bulls having passed (and of) balanced accounts of wages" or "done inspections of plow bulls having passed" and "balanced accounts of wages", or "done inspections", "plow bulls having passed", and "balanced accounts of wages (of agricultural workers)". A check of the texts of tablet box tags that were collected by R.C.Nelsonlls indicates that the assumption of 2 text types is the most likely because the object of an inspection was nearly always mentioned after the phrase "done inspection",l16 and inspections of accounts, balanced or not, are not attested, while "balanced accounts of wages" are attested as contents of tablet boxes.117 The "inspections of bulls having passed" were presumably texts with the subscript guq-apin-dib-ba "bulls having passed". 5 such texts are known to me. One of them is # 23 of our corpus. It differs only in this phrase from the guq-apin-dib-ba texts of type B. The other four texts differ substantially in format.
Heimpel
33
Plow animal inspection records
Plow animal inspection records
gu4-apin-dib-batexts
Lunindara supervisor bull lieutenant Uludj
~u-d~in-dar-a
CT 7,32 BM 18394 AS 2 consists of short lists of oxen and donkeys followed by the name of a cultivator. The first group looks very much like a plow animal group from a guq-apin-gob-ba text: 3 ab-m& 3 COWS 4 guq-gig 4 yoke bulls
ugula Ulu3-df nu-blnda-gud
The subscript is: guq su-ga engar nu-bhnda-guq
2 guq-1
2 one year bulls 1 COW milk calf 1 nu guq-gig ki-gig-i 1 non (present) oilplant land yoke bull 1 anSe su-ga engar 1 mare replaced (by) the cultivator LU-d~in-dar-a Lunindara There is a working group of 7 adult oxen, 3 young oxen, a non present bull connected with oilplant cultivation, and a replaced mare. The remaining 6 groups were much smaller. They included a fair number of animals connected with oilplant cultivation, none of them non present. There was no division into sections recording assets, withdrawals, deaths, and supplements. A supervisor for the first 5 groups and another for the remaining two groups was mentioned. Then follows the subscript:
guq dib-ba SB ~ a ~ a g ~ ~
1 aib-amar-ga
ugula sanga d~in-dar-a
bulls, replaced (by) cultivators (and) bull lieutenants bulls having passed, of/in Lagash supervisor bishop of Nindara
The three texts demonstrate that the guq-dib-ba was connected with a place, and that it was an occasion when assets and deaths of animals of plow animal groups were established and replacements were executed and registered. According to CT 7,29 18383, the occasion fell late in the year as the text is dated to the 1lth month. The fourth text, HLC 251 table 126 AS 5 is the fragment of a 4 column tablet. 'The subscript does not include a geographical name. It consists of a list of plow animal groups without division into sections of assets, deaths etc., includes several unique features, and is essentially unintelligible thus reminding us that the mechanisms of monitoring plow animals remain essentially ununderstood.
guq-apin dib-ba
plow bulls having passed house of Nindara ugula SeS-kal-la supervisor Sheshkala SB a-SB La-za-PI oflon the field La-za-PI The second text is CT 7,29 BM 18383 AS 3. It recorded 4 donkey groups. There are sections of assets and deaths, but not of withdrawals and supplements. The numbers in the assets sections decline from 10 working animals in the first group, to 8, 6, and 4. The fourth group includes "1 yoke stallion, lost, to be replaced (by) the cultivator" which was listed ----- between the assets and deaths sections. Blocks of two groups each were subscribed by anb nlgurll PN "donkeys, ... of PN". The first PN was Urbau, the divisional head and bishop in the household of Ningirsu.l18 The second was his colleague Dudu. Then follows the subscript: C d~in-dar-a
ugula Ur-gar nu-blnda-gu4 guq dib-ba dumu Cnsi-ke4-ne
31Ur-sag-pa-&
supervisor Urgar bull lieutenant bulls having passed (of the) children of the governor oflin Ursagpae
@rUr-mes Cnsi
intermediary Urmes governor h g"lr Ur-nigin sag-dug lugal and intermediary U. land registrar of the king iti Se-gurlo-kug month 11 The third text, HSS 2,32 AS 2, records distribution of oxen that had been returned by cultivators and bull lieutenants and were distributed to cultivators under their bull lieutenants. A typical entry is obv. 7-9: 1 guq su-ga nu-bBnda-gu4
1 bull, returned (by) thela bull lieutenant
34
The term dib-ba in gucapin-gub-ba texts
In guq-apin-gub-ba texts, the phrases 6 1 dib-ba, egerg dib-ba, and nu-dib-ba occur as characterizations of animals. The totals of text 1 include 1 yoke bull, 1 "two year" cow, and 1 "two year" mare, all characterized as egers dib-ba-ta "from after the passing" (XI 37; 40; XI1 15). Next to it (XI1 20-21) we find 1 mare egerg gurumta "from after the inspection". An entry from a supplements section of the itemization (1 IX 20') reads: l* dur a-ru-a-taegers dib-ba-ta "1 stallion from gifts from after the passi1lg".~~9 An example for Sh dib-ba "of the passing" is 33 "IV" 17-20: 5 yoke bulls of the passing 5 guq-gig Sa dib-ba na-grib-tum-S2 gir Si-KAK kurugda zi-ga
to the NaGABtum intermediary fattener Si-KAK withdrawal
The withdrawal left just 2 fallen animals in the group. It seems that the passing was the occasion on which it was determined to move the 5 yoke bulls to the NaGABtum.120 According to 2 I 3 a bull "of the passing, cultivator Urshaga" was part of the assets of the group of Urenki. The passing was here apparently the occasion on which a bull was transferred from the group of Urshaga to that of Urenki. A similar entry is found in 40,9: I* dOr Sh dib-ba ~b-ba-gu~~.'~~
The phrase nu-dib-ba occurs in 43, 3: 1 guq-2 su-su engar nu-dib-ba im-ma "1 two year bull, to be replaced (by) the cultivator, not having passed last year". Here, nu-dib-ba fills in for nu-su and designates the occasion when its status changed from "to be replaced" to "one year overdue".
Heimpel
Plow animal inspection records
Combining the clues of this and the last section indicates that dib-ba designated an inspection: it took place in a specific location, it served as occasion for transfer of animals from one group to another and back to the administration, it was the time when the replacement status changed and when the composition of plow animal groups was determined. In section 46 a text from Umma is quoted which demonstrates that withdrawals were made at the occasion of a field inspection. The verb dib was written with the sign L A G A B X P A ~which ~ ~ was fused with the sign LAGABXASin post Ur 111 script. In lexical tradition dib = etequ = to pass. But it was often identified (i.e. confused) with dabs = jabatu = to grasp. Many scholars follow this identification (i.e. confusion) and assume that already in Ur I11 texts dib was simply a scribal variant of dabs. This cannot be so because the contexts of dib and dabs fall into distinct groups and because dabs occurs consistently with the prefix i- while dib occurs consistently with the prefix ba-. Not a single dib is found in thousands of references for the formula i-dabs "helshe received". dib-ba occurs as a noun for which a translation with the gerund "passing" with infinitive function fits best, thus egers dib-ba "after the passing" and Sh dib-ba "of the passing". It also occurs as a participle, as in nu-dib-ba. Being the hamtu form, it should mean "not having passed". The literal meaning of the term indicates probably the particular form of the inspection so designated. In Cylinder B of Gudea, the inauguration of the newly built Eninnu is described. Its central part was a procession in which the servant gods passed before their master god Ningirsu with their "divine powers" (me-ni-damu-na-da-dib-e).Described was apparently a form of inspection in which the head of a household surveyed his servants as they passed before him. According to this parallel I suggest that in dib-ba inspections plow animals were lead in a file before the owner or his representative and the recording scribe.123 The results were the gu4apin-dib-ba texts. They were close to the gu4-apin-gub-batexts of type B, and the fact that text 23 was characterized as guq-apin-dib-ba in the subscript while being a guq-apin-gub-ba text in format may be understood as confirmation of that closeness. Yet they were not identical. The gu4-apin-gub-batexts of type B were not linked to a particular place, and they regularly recorded assets, withdrawals, and deaths. Unfortunately, there does not seem to exist for the time being guq-apin-dib-baand guq-apin-gub-ba texts registering the plow animal groups of one and the same bull lieutenant from one and the same year.
Heimpel
1
1I I
I
Plow animal inspection records
comparison of the two groups, but it is often difficult to decide whether differences resulted from different processes of field administration in the two provinces - which looked back on a long history of independent development by Ur 111 times - or from different scribal conventions. The subscripts are a case in point. In Girsu, they consisted of the identification of the household as administrative unit and the phrase gu4-apin-gub-ba "standing plow bulls" which expressed the verification of the number and kind of assets of plow animals as essential feature of the process that led to the recording of the texts. In Umrna, that phrase is not found. Instead, the process that led to the recording of the texts was designated as "done inspection7'(g"rum aka), and the inspected groups were not identified by household but by region. Was the process by which the gu4-apin-gub-ba texts were produced really essentially the same as that which produced the aka of plow animals? And was the field administration in Lagash not divided into regions and that of Umrna not into households?
rum
I
36
Inspections of administrative units and field inspections
The numbers of cultivators and bull lieutenants in texts 7, 9, and 11, correspond to the hierarchical relationship of 5 to 1 between these ranks that was standard in Lagash and Umma agricultural administration.124 In texts 1, 2, 4, and 5, the number of cultivators and bull lieutenants was equal. This ratio of 1:l reflected the distribution of cultivators on the field. Contrary to what we might expect, the administrative organization was invertly proportional to distribution on the field: cultivators who worked next to each other were supervised by different Figure 15
cultivator Dug d
~
r
3
-
~
~
~
+
~
~
~
VI Umma 35 Umma texts The Umma plow animal inspection texts are listed in Table 17. They are fewer and shorter; their scribes did not use the number stylus, and the texts differ in substance and details of notation and terminology from the Girsu texts and among each other. Yet their essential feature is the same, namely a description of the assets, withdrawals, deaths, replacements, and supplements of groups of plow oxen or donkeys that were managed by cultivators. The texts from Urnrna and Girsu complement each other in several aspects so that we can gain information from
13-sum Lugal Zi600
Gi-apin-kug-r6 Zi600
bull lieutenants, and these bull lieutenants were in turn supervised by different superiors. The principle can be demonstrated with the help of BM 110116.125The text listed the yield of the
Heimpel
Plow animal inspection records
district Da U m a , detailing the yield of the individual domain units and their cultivators. The sequence of units and cultivators followed the administrative organization: after groups of five cultivators the supervising bull lieutenant was mentioned, and after two groups of five cultivators and their bull lieutenants the untitled supervisor (uguls), elsewhere called the 10-bullscribe (dub-sar-gu4-lo).Cf. the first group of 5 cultivators in Fig. 15. There is very little overlap; the 5 cultivators worked plots in 9 different field tracts. If we choose a field and observe who is working there, we obtain the following result (the field Uku3nu-ti is chosen in Fig. 16).
Mushbiana. Accordingly the public agricultural land of the province increased slightly between AS 2 and IS 3. Figure 17
district
bull lieutenant
!h-kii-ge
22
area in h 2197.02
#of units 68
-
100
Anegara Mushbiana
3
-
-
all
-
4849.6"
7,
I-Sar-ru-um Ab-ba-gi-na Seg-kal-la
# offiedrs
1 Da Umma I
Figure 16 cultivator
Plow animal inspection records
Heimpel
I
-
I
-
1
105
I
text BM 110116
99
date
IIS 3 77
BM 105334
1 AAS 83
AS 2 )[
1
I
* This is the total of the text, 5184.4, minus oil-plant land and garden plots. The unit numbers show that even the comparatively large plow animal inspection texts recorded only small fractions of the plow animal groups that worked in the districts.
lrll-dSrira
38
7 cultivators worked on the field Ukunuti; all had different bull lieutenants, and all bull lieutenants had different superjors.l26 It follows that inspection texts with the same number of cultivators and bull lieutenants were field inspections and that texts with 5 cultivators per bull lieutenant were not. The ratios between the two ranks in texts 6 'and 8 remain unexplained. 37
The districts
The subscripts include geographical nanles. Some were placed after Sa "in", others not. Text 5 combined both: "done inspection, gu4-LCb, Guedena (and) Mushbima, in Edur Shulpae". The wording suggests that plow animals working in Guedena and Mushbiana were inspected in Edur Shulpae. Guedena and Mushbiana were districts of the province, Edur Shulpae presumably a village jn the vicinity of the Shulpae field. Text 6 relates the "donkeys" - if Gomi's restoration is correct - to Guedcna, and text 8 its 13 plow animal groups to Da Umma, another district. The districts of Urnma have not yet been treated systematically, and this cannot be done here. Instead I tabulate the data that are helpful as background for understanding the plow animal texts in Fig. 17. The references were assernbled and partly treated already by Maekawa in Zinbun 22. Fish 539 did not record the numbers of units of Akasal and Guedena. The ones given here are based on the assumption that the relationship between area and units was equal to that of BM 110116.127Both texts recorded the yield of cereals harvested by so many cultivators from so much land. The total number of units in IS 3 was 102 plus a few units from
The inspection sites
Dukusig "Gold Hill", presumably a tell where an Ummaite struck it rich, was also the name of a field. The latter was located in the district of Akasal according to RA 79, 3 1 # 23; according to the same text the field Shulpae was located there too, and this in turn localizes Edur Shulpae "Shulpae village". It is unclear why plow animals that were connected with the districts of Guedena and Mushbiana should have been inspected in the district of Akasal. Ukunuti was a field in the district Da Umma. Enkar, or Menkar,l28 was treated as field or as district. In the phrase "bulls (of) Men[kar]" in the comprehensive deaths list (cf. Table 18) it was placed after the deaths of the combined districts of Guedena and Mushbiana and was thus seemingly equaled with districts. Yet there exist other examples for uniting fields and districts in one descriptive category.129 In AnOr 7, 313 Menkar was called a field among other fields which were summarized in the subscript as a-SB Gu-dBna u Amar-GIR2-gun3[ ] x-na. Accordingly, Menkar was a field and located in or near the district Guedena. According to AAS 83 Menkar was the location of gu4-iuhub land (see section 40); DCS 271 AS 7 V listed five cultivators who were responsible for harrowing 6 iku in the "field Menkar". The actual locality of the inspections was the threshing floor according to MVN 4,75 AS 6 IV 6 which recorded the results of an inspection of plow animals and agricultural workers. Its subscript reads: glirum aka 4-6-kam ki-su, a-SBd~in-ur4-ra"done inspection of the 6th day (on) the threshing floor of the Ninura field". Closer examination of the text shows that it recorded all plow animals of the district of Akasal where the field was located (cf. Fig. 18):
Heimpel
Plow animal inspection records
Figure 18
a-Sh A-ki-lhh
tu-ra
Lu-kal-la engar
ugula I Da-DU-MU Lugal-6-ma Lugal-ra-ni Lugal-kii-zu Lugal-nisag-e Ur-ma ~r-~~n-lfl-ld 1.1-pa-&
If we divide the number of oxen and donkeys by the average group strength 8.4,13O we arrive at 28 or 29 groups. The result is close to the 26 units which were postulated above for the district in IS 3 according to Fish 539. 39
The bull of Gula
Text 3 recorded the inspection of 1 unit which was designated as "bull of Gula". The term is reminiscent of the small households which were part of large households in Lagash, especially that of Ninhursag which also had just one unit (cf section 2). The "bull of Gula" appears also in BM 105334, a balanced account of the total surveyed public agricultural land (ni-kas7 aka a-Sh @d-da)of the province in AS 2. After the description of the domain land of 100 units, including the allotments of its managers and the fallow, the following passage is found: 6 (bar) ghn-guq
38.88 h domain 1 engar 1 Me 1 cultivator 2.16 h 3 BB-guq 3 iku-ta 3 oxdrivers 1.08 h each @r-sG-gaguq d~u-la-me they are personnel of the bull of Gula Accordingly, the "bull of Gula" described an administrative unit of the public land consisting of 1 agricultural unit including the allotment plots of its cultivator and his 3 oxdrivers. The cultivator of the unit was a certain Lukala in SS 5 acc. to the comprehensive deaths list IV 20-22 (cf. Table 18): 1 kul guq
1 ox hide Lukala cultivator guq d ~ u - l a bull of Gula Lukala managed the unit as cultivator still in IS 3 according to BM 110116 XVII 11-16 where he and his unit were recorded as 68th, and last, unit of the district Da Umrna: Lu-[kall-la engar
3 (bur) 7.5 (iku) ghn 10-ta
le-bi 90 gur
Plow animal inspection records
Heimpel
22.14 h land of 10 (furrows per 6 m) each the barley (yield) 27000 liter
nu-bhnda-guq Lugal-6-zi-da
field Akishah Lukala cultivator bull lieutenant Lugalazida
blank bull of Gula The plow animal inspection of the bull of Gula, our text 3, is from AS 5. The cultivator at that time was a certain Sipaanshe.132 There existed two Gula figures in the province, Gula of KI.AN and Gula of Urnrna. It is not clear which Gula was associated with the agricultural unit. Gula of Umma employed 2 shepherds; one had a flock of Sumerian sheep, another one of goats according to YOS 4,237 SS 7 11, a count of the small cattle of the p r 0 ~ i n c e . l ~ ~ guq d ~ u - l a
Three of the 12 plow animal inspection texts from Umma concern guq-Suhub. 17 of the 19 groups consisted of oxen, the remaining 2 of donkeys. The group compositions varied from group to group, but they do not stand out in any way from the others or from those of the groups in the texts from Girsu. They were apparently unaffected by the special attributes of that which is designated with the term guq-Suhub. Maekawa has repeatedly addressed the problem of the g ~ ~ - S u h u and b l ~developed ~ a hypothesis based on several lines of evidence, namely (1) its occurrence in Ur I11 texts of agricultural context from Girsu and Umma and (2) a passage in the OB Georgica, and (3) its alleged literal meaning. He showed that it designated land consisting mainly of C plots and expressed his conviction that they were specifically C1 plots. His plot classification is as follows: C refers to plots for which 360 sila barley were expended for seeding 1biir of land. If plow animals were used, 240 sila were used for seeding 1 b b of land in 8 rows per nindan and the remaining 120 sila for plow animal fodder. Such plots are called specifically C2. If plow animals were not used for seeding, the whole amount of 360 sila was expended for seeding 12 rows per nindan. Such plots are called C1. Normally, the context of the references for C plots did not express whether C1 or C2 plots were meant. Maekawa, BSA 5, 127, states with respect to the large tracts of guq-iuhub land of C plots in TUT 5: "I assume that C1 plots were much more extensive than C2 plots in Girsu in Shulgi 47". Thus guq-luhub land would not be seeded with plow animals but instead with the implement called @;-gaba-tab which was operated by workers. He shows further that this land was newly converted from waste land, especially reed thickets, producing relatively low yields. Being new land, it was not yet integrated into the "temple households" managed by bishops and commanders but stood under the direct control of the governor. Lines 6-7 of the Georgica indicated to Maekawa that the Babylonians introduced luhub oxen into fields that they had flooded early in the year. Following a suggestion of Civil, he connected the practice with "puddling" in which oxen or buffaloes are let into fields that are to be planted with rice in South East Asia.
Heimpel
Plow animal inspection records
The word Saub may mean "boot" and is, in that meaning, written with the determinative kuS "leather". The Umma field name a-Sg guq-MUL was indeed spelled once, in BIN 5,266, with the determinative, suggesting that guhub in gu4-guhub represented the word "boot". Literally, gu4-guhub would then mean "boot bull". These boot bulls "were probably brought into the fields so that they could break up clods and make the bottom ground smooth. This is also an effective method for stamping out weeds and plants, as is described in Georgica 8. Indeed, oxen seem to have had their hooves and legs covered primarily for protection from the thorny plants and roots of reeds cut out of the field. Another benefit of this method was that salts, which would otherwise remain on the ground surface, were scattered deeply and thoroughly into the ground (BSA 5, 129)". In ASJ 14, Maekawa made some modifications: (1) He voiced doubts about the relevance of the spelling in BIN 5,266 and quoted the equation MUL = l u b b = suguNdnu = member of a herd which Civil had suggested. Indeed, since iuhub represents two words - sugulldnu and suhuppatu - the scribe of BIN 5,266 may have used the determinative to merely indicate the pronounciation Suhub as opposed to mul, without intending the word "boot". (2) He embraced the fact that the verbal form in line 7 of the Georgica has prohibitive function, and now understands it correspondingly as "an advice to avoid using gu4-suhub oxen's tramplinglpuddling in a field of average soil conditions after the artificial flooding in early summer". (3) He quoted evidence for small areas of gu4-suhub land managed by temple households. Maekawa did not consider the occurence of the term in the plow animal inspection texts from Umma. The evidence of seemingly normal plow oxen and donkey groups that were designated as gu4-suhiib does not seem compatible with essential elements of his hypothesis. If trampling the flooded fields that were to be planted in C l plots was the sole objective, any herd of oxen would have sufficed, and the more the better. Why would groups be needed that were managed by cultivators? Would donkeys be suitable for puddling? In order to resolve this difficulty while adhering to Maekawa's conclusions that gu4suhub was a term for land that had been recently converted to agricultural use and was cultivated in a special way, and adopting Civil's identification of gulplb with sugulldnu, 1suggest to use two meanings for guq-suh6b: (1) herd oxen and (2) (former) herd oxen (land). In Georgica the first meaning applies, in the Ur III texts the second. Georgica 5-8 has been translated often. I add my own version: uq a-ta im-mu-e-8d8-a
a-gii ki-durug-bi en-nu-un ak-ab ki-L-er a-ra-ab-tuk guq-luhub @r na-ra-ab-zukum-e
u KIN-bi u-bi-zi5
...
"When it (the ground) emerges from the water watch the field and its wet spots. When it develops kiSer, do not let herd oxen trample it.
Heimpel
Plow animal inspection records
When you have weeded it ..." Presumably this refers to the fields that had been lying fallow the previous year had been used as oxen pasture. In preparation for cultivation the fields were flooded early in the year. When the soil dried up it was time for the oxen to move on. Otherwise the ridges and depressions produced by its tracks would prevent even drying. In the Ur I11 sources discussed by Maekawa, the term gu4-suhub clearly characterized a type of land. Originally, it may have been "herd oxen" land, especially reed thickets (gig-pi) that may well have been used by oxen herds as pasture. Oxen, next to buffaloes, are kept in the marshes still today. In converting such land to fields, the water had to be drained and reeds and other grasses and sedges had to be eradicated. Enki converted marsh (sup) to fields for Enrnerkar - in Old Babylonian view - as follows: d ~ n - klugal i ~ r i d u ~ * - ~ agi- kTIL-bi e ~ lp-ma-an-zb a-bi ha-maan-ti] "Enki, the king of Eridu, eradicated its ... reed for me, brought its water to an end for me".l35 Most work-intensive must have been the removal of the troublesome rhizomes of grasses, including reeds, and sedges.136 It was probably a lengthy progress that took several years137 during which the cultivation methods may have changed and the yields improved in steps. Important are the observations of Maekawa on the fields of the household of Ninmar in the district of Guabba (BSA 5, 132-134) which document slow improvement of wide areas with very low initial productivity. The fact that this was newly converted land is - in addition to Maekawa's arguments - indicated by the nature of field names.l38 Returning to the inspection of plow animal groups on gu4-suhub land, it still remains unclear what the plow animals did on that land. If Maekawa is correct in his assumption that gu4-suhub land consisted mainly of C1 plots and thus was seeded without the use of plow animals and if that was also the case in Umrna, it follows that the plow animal groups were used exclusively for the preparation of the ground prior to seeding. The early date of text 5 may confirm this conclusion, but it is counterweighted by text 8 which is dated late in the planting season. This text may prove to be important for the question of the nature of gu4-suhub. As is argued in section 43 below, the gu4-suhhb groups were formed from normal, that is non gu4suhub groups, after they had finished their seeding work on other land. I believe that it is premature to attempt a complete interpretation because of several intangible factors: (1) We do not yet know what kind of work was meant with the designation gu4 ki-durg. It is mentioned in MVN 6, 305 beside harrowing and seeding; it required many plow teams and a remarkable number of 8 workers pro animal team; the literal meaning "wet spot bulls" indicates that it had something to do with wet conditions on the field. Perhaps the translation of Georgica 7-8 as given above is wrong after all, and the gu4-suhub did have something to do with ki-durg. (2) The process by which newly developed gu4-suhub land was integrated with old farmland is not clear, and neither is the point of time when gu4-suhub land was fully converted and stopped being so designated. The designation may still have been used in Umma during a period when original gu4-suhub land was already so far improved that cultivation fits into the seeding could be done with plow animals. (3) It is not clear how oil~~lant picture of plow animal use. A connection of gu4-suhhb land and oilplant cultivation is attested in
Heimpel
Inana field @r Ha-ba-an-zi-zi intermediary Habanzizi There are 7 group descriptions. All end in the formula PN engar i-dabs "cultivator PN took in charge". Apparently, the groups came without the cultivators who had managed them under Nigarkidu and were received by cultivators under Abbamu. The year date is 5 33 (so Owen), 5 45 or IS 4. The fact that the persons do not seem to be attested elsewhere argues against S 45. The assignment of the text to Umma seems inevitable because of the month name RI and is confirmed by the personal names Ur-Ell-e and Nig-
ki-dulo. The terminology points to Lagash (anle instead of emeg, amar-ga instead of riblguq-ga, Su-giq instead of Su before the animal designation, and 164"deficit" for deaths). a-Si d~nana
a fragmentary text that is now destroyed: ITT IV 7331 (cf. MVN 6, page 208). Clearly, more source material is needed to advance the understanding of this important topic.
41
Month dates
One obstacle to interpretation of the texts from Girsu is the almost complete lack of month dates. The only exception is the guq-apin-dib-batext CT 7, 29 18383 which is dated to the 1lth month. In the Urnma texts five month dates are preserved. Three, the 5th, 6th, and 7th, fall squarely into the season of seed plowing. Using Maekawa's Table 1 in BSA 5, we can see that in Lagash most expenses in the 5th month went for deep plowing and harrowing, and those for seed plowing commenced. In the 6th and 7th month the expenses for seed plowing dominated and those for deep plowing and harrowing trailed. The inspection record MVN 4, 75 AS 6 (quoted above in section 38) was dated to the early 4th month when expenses in Lagash went only for soil preparation. Texts 5 and 8, both referring to gu4-suhub land, date to the 3rd and the 9th. Maekawa found a single text from a 9th month with expenses for seeding, and none with expenses for soil preparation or seeding for the 3rd month. The wide spread of month dates indicates the lack of a seasonally fixed date for plow animal inspections in Urnma. This finding agrees with the fact that the gu4-apin-gub-ba texts of type A from Girsu show much difference in the degree to which the norm of plow animal groups was reached. If the existence of a norm indicates that there was a season during which all animals of the norm were needed it follows that the degree of divergence from the norm was proportional to the closeness to that season.
42
Text formats
The only features that are shared by all Umma plow animal inspection texts are lists of assets and deaths of plow animals and the record of their managing cultivator and his supervising bull lieutenant. All other features are strewn among the texts seemingly at random so that there is no convincing division into types. Using the criterion of the ratio of cultivators to bull lieutenants we have separated field inspections from "administrative" inspections. Looking at the criterion of the place of inspection, we expect field inspections to identify the place of inspection. Indeed texts 1,4, and 5 do so, yet 2 does not, and 7 does without being a field inspection. Looking at the criterion of absence or presence of sections, the small scope of most texts obscures possible typological differences. The absence of withdrawals and supplements among the 5 groups text 9 for example may be due to the absence of such sections from the format of the text, or to the fact that in the recorded year no animals were withdrawn from, and supplemented to, the recorded groups. Only in the case of text 4 with its 9 groups does the latter possibility become unlikely. Text 12 shares with the other texts descriptions of plow animal groups, but deviates in almost every other aspect. The full subscript is: guq anie si-la engar oxen (and) donkeys, entrusted (to) cultivators ki Nig-ki-dulo-ta from Ni
Plow animal inspection records
Heimpel
Plow animal inspection records
43
1
The persons of text 8
Text 8 recorded a bewildering number of persons as result of numerous transfers on different levels of the administration. They are presented in Table 19. Cf. for example the subscript of group 1: from Luutu, son of Ludigira ki L B - ~ U ~dumu U Lii-digir-ra-ta L B - ~ U engar ~U cultivator Luutu, dumu Gu-u-g[u-a] son of Guugua ki Lugal-[x-x] from bull lieutenant nu-bhnda-guq-ta Lugal[ 1 [nu-binda-gu4] Lu-dijjir-ra bull lieutenant Ludigira [ugula LIB-nir supervisor Lunir The group was received by Luutu, son of Guugua, whose bull lieutenant was Ludigira, from Luutu, son of Ludigira, whose bull lieutenant was Lugal[ 1. In 8 of the 13 groups of the text transfers of the same kind were indicated. The text also includes two cases of the formula PN i-dabs. They are found after the descriptions of group 5 and 12, and were separated from the descriptions by a blank space, indicating that they referred to all previous groups. The group descriptions include the names of several supervisors above the rank of bull lieutenant. The persons who received these groups either ranked yet higher than the supervisors, taking charge of the latter, or the responsibility for the groups was transferred from them to these two. The second alternative seems more plausible because the two, Lunir and Shaninga, were the supervisors of groups 1 and 7. Combining this feature with the fact that the groups were inspected late in the seeding season (month IX) and connected with the term gu4-suhub, it appears that supervisors had supervised the plow animal groups during the regular seeding, and that they relinquished that function to two of their own for the late work on gu4-suhub land. 44
The norm of working animals in Umma
The Girsu texts contain sufficient information to distinguish working and not working animals and to determine norms of working animals in the plow animal groups. For oxen groups the
Heimpel
Plow animal inspection records
norm was 7, for donkeys it was 8. The crucial pieces of information that are needed to demonstrate norms are lacking in the Umma texts. Working and not working animals of all ages and both genders were summarized without distinction with the terms guq-rib-?@-a and anSe-?@-a; there is no text that recorded the difference between actual assets and norm as did Girsu text 31; there is no outright mention of a norm as in Girsu text 1. In Table 20 the data from fully preserved group descriptions in Umma are collected. Umma terminology did not use the designation "(fit) for the yoke". The cultivators and oxdrivers in Umma surely hitched some "two year" animals and Su-gi4 animals to the yoke as their colleagues in Lagash, if the animals were fit for it. Looking at the numbers for adults, "two year", and Su-gi4 animals one is tempted to choose from among the last 2 groups in such a way that a norm results. Take for example the 4 groups of text 1: in 1.3 and 1.4 the total of adults plus supplements (which were all adult) equals 8. If we admit the 2 "two year" animals as working animals, the resulting groups would have 10 working animals which is too much according to Lagash standards. In 1.1 we could reach 8 by counting the one 2 year old. The second group would have without the one 1 year old an exact half of 8, possibly a half group. A more compelling argument is found in the fact that 8.8 and 8.13 were newly formed with 8 adults. The number 8 is also repeatedly found for donkey groups. It seems to have been the norm for oxen and donkey groups. As noted in section 17, oxen groups in Girsu often consisted of 1 cow and 6 bulls, and the pattern of supplementation and substitution indicate that there was a tendency to produce such groups. The group of text 10 presents a parallel case. It had 4 bulls as assets and received supplementation of 2 bulls and 1 cow. It is, however, the only cow that served as supplement in Umma according to our corpus.
45
Heimpel
Plow animal inspection records
animals". Until that point they had been amar-guq-DU.DU. 2 or 3 of the young animals each were assigned to cultivators and their immediate superiors. The transfer was not an isolated case. AnOr 7,264140registered also milk class animals, called guq-anSe-DU.DU, that were turned over to plow groups (apin-S&!KA kCS-rri); and AAS 62 recorded 13 milk class oxen and donkeys in 3 48 and 8 in AS 1 that were "bound (by agreement)" as "milk calfs (of) gu4-suhub". Figure 19 plow animal price in barley
cultivator
2100 (liter) 2200 1800
1 gu4-giS
1 emeg-m&
4.5 gin x
1 dur
2000 2130
1 guq-gig 1 guq-gig
46
Withdrawals and a deficit replacement
Withdrawals of plow animals went to the person and institution whence supplements came. In Girsu it was Si-KAK and the NaGABtum, in Urnma the person was KAS4, and then US-mu, and the institution was the "bull house" (6-gu4).Withdrawn animals were adult, among them several Su-gi4, or in Umma terminology Su, animals.
Supplements
As in Girsu, supplements were designated indirectly by setting them off from the rest of the group description with the formula libir-irm "is old". Once, in text 11 rev. I 7', the formula dahhu-irm "addition", literally "it is (an) 'add!"', was added after listing the supplemented animals. Supplemented animals were adults and mostly male (cf. Table 21). They came from KAS4 in the texts from AS 2 and the undated text 10, and from US-mu in those from SS 5 and later, including text 11 where the date is broken. TENUS 346 recorded the supplements of cows and bulls that KAS4 transferred to the "bull commanders" (labra-gu4)139in 46. It was 11 cows and a little over 30 bulls. Others were to be bought for grain (Se-ta salO-salO)"from Ir" (hil), the chief of magazines (ka-gur,), that is to say Ir would provide the grain for buying plow animals. TENUS 351 (undated) lists the actual purchase price in grain and silver for plow animals of the groups under the bull lieutenant Da-DU-mu (cf. Fig. 19). Group 2.1 was managed by the cultivator P&-d who was supervised in that text also by Da-DU-mu. The description of the group includes indeed "1 yoke bull to be bought with barley". It cost 2200 liter. The youngest supplemented animal was the "1 year" mare of group 6. There were several milk class animals in the plow animal groups. AAS 63 recorded transfer of such animals to the cultivators. The text instructed an unnamed addressee to transfer milk class oxen and donkeys to the account of KAS4 as guq-giS.apin KA-kCS-rsi "bound (by) agreement as plow
in silver
text
Figure 20 withdrawal [n S]u guq 6-guq-S&KAS4 i-dabs 1 emeg-m& 1dbr-gig KAS4 i-dabs 1 Su diir "[Ir]ll"-mu i-dabs 1 emeg-msih US-mu]-dabs 1 SU diir US-mu i-dabs 1 dhr-gig US-mu i-dabs 1 dur-gig US-mui-dabs 1 rib-mu3 (deficit replacement) 1 guq-gig 1 Su emeg US-mui-dabs 1 Su guq US-mu 1-dabs 1 guq-gis 6-guq-S& 2 guq-gis 6-guq-R 9'
Heimpel
Plow animal inspection records
Group 8.4 listed one "3 year" cow and one yoke bull "yet to be replaced by Lugalgigire; in its stead his wife Ninabbana and child Amagina 'were passed' to weaving" (su-su LugalgiSgigir-renu-su mu-bi-S&Nin-ab-ba-na dam-ni Ama-gi-na dumu-ni US-bar-;&ba-ab-dib). The two animals were counted in the summary of the text as part of the withdrawal (cf. Table 22). They correspond to the deficit replacements in the Lagash texts. The details of withdrawals are tabulated in Fig. 20. SNAT 314 AS 1 V listed oxen and donkeys, with one exception all Su animals, that were received by KAS4 from cultivators at the occasion of an inspection on the Lamah field (giirum aka SB a-SB LA-mah). It appears from this text that withdrawals were made at the occasion of a plow animal inspection. The receipt of a single Su cow was recorded in UDU 38 $34. The recipient was Rasha, a predecessor of KASq.
47
Fall and death rate
Dead animals were listed as deaths with the formula "TIL 1" before the designation of a single animal and ri-ri-ga behind the designation of 2 or more animals. The designation of the singular is characteristic for the plow animal inspection texts from Umma. The scant data on deaths are assembled in Table 23. An approximate estimate of the death rate can be computed by totaling assets, replacements, deaths, and withdrawals, and determining the percentage of replacements and deaths. It was 15.79% among oxen and 12.42% among donkeys. The numbers are displayed in Table 24. Considering the small corpus of texts they cannot be regarded as reliable. They agree nevertheless in a general way with the death rate that can be extracted from the Lagash texts. The cause of death was not recorded in the deaths sections and rarely in the replacement sections.
48
Replacement
The notations for the various aspects of replacement in Umma equal essentially those in Lagash. Animals that had been replaced by the time of text recording were designated as su-ga "replaced", those that were absent at that time were designated as su-su "to be replaced", and if they were not replaced at the next inspection they were then designated as su-su nu-su "yet to be replaced". The length of the overdue period was recorded in Girsu by adding im-ma "last year", "two years", etc. In Umma this was not done. Text 12 used a different notation: Sunigin 16-15 dur-@;iS18-i engar-kam "sum of deficits: 5 stallions deficit of cultivators". The titles or names of the liable persons were added after (su-ga/su-su titlePN) or within the replacement formula (su-su titlePN nu-su). In 6: 1 it was actually spelled out that the person was the one who would have to replace the animal: 1 yoke stallion yet to be replaced 1 dar-gig su-su Lu-digir-ra nu su by Ludigira Ludigira has to replace it. Lu-digir-ra-ke4su-su-dam
Heimpel
Plow animal inspection records
Instructive is also the case of 5:2 where the name and title of the cultivator who had replaced the animal was recorded. It was not the cultivator of the group, and thus confirms indirectly that the formula su-ga engar referred to the cultivator of the group in which the entry with the formula is found. The replacing persons were generally identified by title when they were cultivators and by name when they were bull lieutenants or higher ranks. Many persons of text 8 can be identified from that text (cf. Table 19): Urgipar and Uramma were the bull lieutenants of groups 4 and 5, and Luutu and Lushulgira were supervisors. Urani appears in the formula "from" before the name of a bull lieutenant in group 6 and must therefore have been a bull lieutenant himself (cf. the quote in section 43); he is in fact attested as bull lieutenant under the supervisor Luutu in BM 110116 IS 3.141 Gu-TAR is attested as bull lieutenant in the comprehensive deaths list of SS 5 (cf. Table 18) and elsewhere. A man with the very colnmon name Urgigir was identified as bull lieutenant in SNAT 399 AS 8.142 I cannot identify the remaining nontitled persons. The distribution of persons who had to replace an animal was accordingly as follows:
liable persons cultivators bull lieutenants supervisors unidentified
#
41 8 2 5
The ratio between the number of animals that were replaced, or had to be replaced, by cultivators and bull lieutenants was 5: 1 and corresponds to the number of cultivators that were supervised by one bull lieutenant. In Lagash, the ratio was 3:l while 4 cultivators were supervised by one bull lieutenant (cf. section 22). Thus, the Lagashite bull lieutenant became more often liable for replacement than his colleague in Umma, if the scanty numbers from Umma are representative. Death as cause for replacement was routinely recorded in texts 2, 4, and 7. Cf. e.g. the entry from group 1 in text 2: TIL 1gu4-@Ssu-su engar "dead, 1 yoke bull, to be replaced (by) the cultivator". In 6:6 the cause of death was given: a stallion was "killed by a lion at the place of .." (ki nam-btir-a-kaur-mah-e TIL-a). Text 7 included a note about a returned animal in the group of the cultivator Aakala: 1dhr-giS su-ga engar 1 yoke stallion replaced (by) the cultivator uen inim ~ r - ~ ~engar on the word of the cultivator Ursin dumu Ka-ka-ta son of Kaka. Ursin was the cultivator of the first group in this text. He seems to have served as witness to the fact that his colleague Aakala did in fact return a stallion to his group. Replacement took place at the occasion of the dib-ba inspection in Lagash (sections 33 and 34). If that was the normal procedure, the present case can be understood as an exceptional procedure when not even the bull lieutenant was present and the only witness another cultivator who may have worked side by side with Aakala.
Heimpel
49
Plow animal inspection records
Registers of outstanding replacements
A small group of receipts143 (cf. Table 26) recorded plow animals that were to be replaced (susu) or yet to be replaced (su-su nu-su). 3 of the 4 texts have the subscript guq ange su-su engar-e-ne "oxen (and) donkeys to be replaced (by) cultivators". Texts 1 and 4 were sealed; the legend Lugalkuzu and Lugalemahe are well reads: Lugal-ku-zu / dub-sar / dumu Lugal-ni-[ Lugalugkine was a cultivator who served as supervisor of other attested as bull 1ie~tenants.l~~ cultivators.146 The persons in the formula "receipt of P N were accordingly the direct supervisors of the cultivators who were associated with the plow animals that had to be replaced. The formula was most often used to record receipt. Here, it must have a different function because animals that had yet to be replaced could hardly have been received earlier. I suggest that it identified the person who kept the tablet until its information was processed. In this case, the bull lieutenant kept a record of plow animals that had to be replaced by cultivators under their supervision. Texts 2 and 4 registered such animals for all 5 cultivators, text 3 for 4. The obligation to replace these animals hardly commenced at the same time, and tablets 2-4 were presumably written at the end of an annual cycle. Text 1 registered only one obligation to replace an animal. Perhaps texts 2-4 were summary statements that were produced by transferring the information from single records such as text 1. 3.3 and 3.4 were subdivided into 3 entries under the name of a cultivator. In two of these entries further persons were mentioned as ultimately liable for replacement. Their title was not recorded and their names are common; I cannot identify them. 4.5a attests that liability could be shared between cultivators and bull lieutenants. The presently available documentation suggests that this was more common in Lagash.
VII Conclusions The most important conclusions are: (1) Plow animal groups consisted of oxen or donkeys. Mixed groups occur occasionally as result of substitutions but the animals of the "wrong" species were quickly transferred into a group of the right species. (2) The ratio of oxen to donkeys was about 4:3 in Lagash and 3:l in Umma. Considering the fact that the use of donkeys in agriculture had become negligible by the Old Babylonian period, the ratio in Umma was progressive, that in Lagash conservative. (3) The groups included working and not working animals, the former being animals in their fourth year and older - the upper age limit was never indicated - and some of the animals in their third year. It agrees with Starkey's dictum (cf. note 69) on the age of oxen when they start work: "Oxen should be 3-4 years old". Text 12 from Girsu seems to have included all animals in their third year among the working animals. The texts from Umma do not allow the distinction of working and not working animals.
Heimpel
Plow animal inspection records
(4) There existed a norm of working animals per group. It was 7 for oxen and 8 for donkeys in Lagash. Circumstantial evidence indicates that it was 8 oxen and 8 donkeys in Urnma. There must have been a reason for the number of animals in these norms, but it is not indicated in the texts. Most intriguing is the odd number of oxen in Lagash. In Lagash texts a tendency was observed to form groups with 6 bulls and 1 cow. (5) It is known that the working animals were used for deep plowing,147harrowing, and seeding, but it is not known how they were used. We expect deep plowing to have been the task which required most strength, and it is conceivable that one hitched 6, or even 8, oxen and 8 donkeys before the plow. It would not be without paralle1.14* (6) Cultivators in Lagash were given "barley of the calves" for calves and foals in their second and third year. (7) There existed plow animal inspections in Lagash and in Umma. They were not conducted in one clearly defined season. In Lagash the plow animals, and presumably their managers, passed in a file, presumably before the owner and the plow bull scribe. In Umma existed field inspections and administrative inspections. (8) At the occasion of inspections assets and deaths were established and replacements executed (section 33), withdrawals were made (section 46), transfers from one group to another were effected and the replacement status was changed (section 34) (9) Text 8 from Umma shows that guq-luhub groups could be constituted from plow animals that had finished their regular work of seeding (cf. section 40). (10) The higher the rank the costlier was the responsibility of replacement in Lagash (section 22).
Plow animal inspection records
Heimpel
NOTES K. Maekawa, "Cultivation Methods in the Ur III Period, BSA 5 (1990) 115-145. W. Heimpel, "Toward an understanding of the siKKum institution", RA 88 (1994) 5-31. M. A. Littauer and J. H. Crouwel, "Ceremonial Threshing in the Ancient Near East", Iraq 52 (1990), 1519. P. Steinkeller, Ceremonial Threshing 19. If Steinkeller's identification of W R - b - d a as threshing sledge is correct there is also an Early Dynastic reference for a, presumably ceremonial (cf. Littauer and Crouwel p. 18), sledge that was pulled by mules. Maekawa, ASJ 14 (1992) 208f. with note 43. See section 23 below. I. Gelb, "Household and Family i? Early Mesopotamia", in E. Lipinski ed., State and Temple Economy in the Ancient Near East, Orientalia Lovanensia Analecta 5 and 6 (1979), 20. Maekawa, ASJ 9 (1987) 97f. When used in this sense the word apin was written without the determinative gi.That it was in fact read apin, and not engar, is shown by the formulas apin-na si-ga and apin-na ba-an-si (HLC 290; MVN 2, 109 etc.). Lafont's publication confirms Maekawa's (1.c.) correction of the transliteration of the passage in I l T 2. The expenses were assessed at 18 gur per agricultural unit. L.c. 98. Cf. Maekawa's tabulation in BSA 1,90f. The number of units of the division of Akala in the household of Narnhani is not preserved in the text. It was 15 in text 9, dated to an unknown year in the reign of Shulgi. Cf. e.g. Gelb, Household 12. Cf. Heimpel, JCS 33 (1981) 98. Cf. A. Falkenstein, AnOr 30 (1966) 161f. Falkenstein, 1.c. 130f. Heimpel, 1.c. lines 85 and 175. D.O. Edzard and C. Wilcke, "Die Hendursanga-Hymne", AOAT 25 (1976) 145f.:26-30. Falkenstein, 1.c. 130f. and 162f. It seems strange that it belonged to the household of Nanshe while the Eninegara, the temple of Ningirsu in Nina, did not belong to the household of Ningirsu but also to that of Nanshe. Zinbun 24 (1989) 171. Falkenstein. 1.c. 166. Text 1 IX 1'-10' (cultivator Gina) and text 5, group 13 (cultivator still Gina). Cf. e.g. Ad-da-mu sanga dInana MVN 12, 27 S 46 V; Ad-da sanga dInana MVN 12, 176 S 47. Note that MVN 5, 272 (undated), subscribed as gu4 si-la-a gu4 d~nana,registered 5 old and 1 newly formed group of this sub-household. ~ 6. ~ MVN 11,51 SS 4 VI; T U T SS Cf. Maekawa, Zinbun 21 (1986) 97f. Cf. Maekawa, ASJ 9 (1987) 91ff. The household of Ningirsu was divided into 4 parts in some texts including CT 9, 33 which is used as source for S 47 (1) in table 4. Maekawa, Zinbun 24 (1989) 171, characterizes the evidence of the seed and fodder texts as follows: "Each "temple" land ... was divided into two parts in Sulgi 41, each part being managed by an administrator of high rank. The two parts of land are almost equal to each other in area. This is mainly due to there being an equal division into two of each field within each "temple" land." Small differences in total area
Heimpel
Plow animal inspection records
between divisions are proportionate to small differences in the area of individual agricultural units. Divisions of a household divided the total of agricultural units but often shared the fields on which they were located. Restoration derived from the grand total of 16 units. Cf. Maekawa, Zinbun 24, 140. IX 33; the sign Sabra was collated by T. Maeda (ASJ 2 (1980) 203). The reason for restoring the date for text 37 will be given below in section 7. Household 18f. The text concerning the Bagara, I'TT 211 907, was known to Gelb from an envelope. It has been opened in the meantime, and Lafont gives the text of the tablet in TCT I: it was dated SS 1, two years after that of the envelope. It reminds us how little we understand these deceptively simple looking texts. The text was tabulated Gelb, Household. It is now transliterated as MVN 17, 3. CT 1, 23; 26; 27; 47. Maekawa, Zinbun 22 (1987) 38. The syntactic relationship between the phrase gu4-apin gub-ba and the designation of the household was in fact a genitive as is demonstrated at times graphically. Cf. e.g. the subscript.of text 3: gu4-apin gub-ba Bnsi-ka "standing plow bull (record) of the (household of) the governor;" text 5: ..d~in-jjir-su-ka. III 25-26 and VIII 4-5. JCS 21 (1967) 64-69. Cf. the contribution of van Driel in this volume. For gig-dii as qualification of small cattle cf. BSA 7, 130f. MSL VIII/l, 44:298 and 49:339 o. Su-gi4 donkeys served as dogfood according to texts from Ur (UET 3, 1236 and 1253-55), a Su-gi4 cow as offering acc. to SNAT 46 G. Su-gi4 oxen received fodder acc. to, e.g., SNAT 4; 8; 33; 35 (all G). Sigrist, JCS 33 (1981) 147, note 18, noted the expression kuS-amar-Su-gi4 and suggested as translation "returned." I hesitate following this suggestion because it would require a form ending in -a which is never found. The correspondance was demonstrated by Lafont in the index to TCT 1 S.V. Su-giq. M. Molina, who was not aware of it, collected references of Su before animal designations (it was also used before designations of persons) back to the archaic texts in Aula Orientalis 8 (1990) 74ff. Cf. note 2 above. I follow Foxvog's reading and interpretation of the genitive, unable to suggest a better alternative. This would be the only reference for writing the genitive of URUxGt and doubly unusual because the determinative would have been written against the convention after the genitive. It confirms the observation of E. Sollberger that the name ends in fbl. Civil, AS 16 (1965) 4ff. MSL 14, 42f.: 286, and cf. M. Stol, Epilepsy in Babylonia (1993) 19. Cf. for the same sequences in the case of small cattle BSA 7, 118f. 1lth edition XII 673f. Baba Meqi'a 6.3. Maekawa, ASJ 13, ... and cf. text 1 in F.Yildiz and Maekawa, "Animal and Human Castration 111", Zinbun 18 (1982) 96. The term means probably "holding." He is attested e.g. in BIN 5, 336 S 43; MVN 14, 231 referring to events between AS 5 and 8; TCS 164 AS 5 VII. Gorni pointed out the connection with Buffalo 3. PIOL 19,225 D AS 5 I 24.
Heimpel
Plow animal inspection records
ASJ 12,41 text 10 AS 5 IV 30. JANES 9, text 5 D AS 7 VIII 12. MVN 13,507 D S 47 VII5. 31 VIII 12. 25 rev. 3; 44 first and fifth group; 45 fourth group; 47 third group. Zinbun 21 (1986) 97f. They were collected by Maekawa and tabulated in BSA 5 (1990) 136-137. The parallel use of these designations indicates that Bb-gu4-gig was compounded from Bb-gu4 (corresponding to anSe) and gig, not from Bb and gu4-gig. The numbers come only from the assets sections. The summaries of withdrawals and fall did not distinguish age. The numbers are taken from the assets, withdrawal, and fall sections. Such priming may not have been a consideration. P. Starkey says: "Oxen, or castrated bulls, are considered most suitable for training ... which takes just three weeks" (Farming with work oxen in Sierra Leone, Njala University College 1981). CT 7, 15 BM 15324; STA 28; MVN 11,90; ATL 4. ATL 3; TUT 2; TUT 3. ATL 2. CT 7 , 9 BM 12927; ATL 8; 10. ATL 74; TUT 3. OBRT 258. TUT 5; DCS 68. TUT 4. Maekawa, Zinbun 18 (1982), text 4. Date and subscript of the text are broken. The signs 18-1 have the tendency to form a ligature in which i was placed inside the angle formed by the 16. As Steinkeller noted, the word was pronounced rather Ilaul than flail. /u/ after a base without preceding affixes seems to typical for imperatives, perhaps imperatives of ventive forms. Cf. e.g. gB-nu = alkam, and, from Ur I11 texts, dah-hu. Etymologically, la-i and dab-bu is a pair of opposing meanings: "deduct!" and "add!" The distribution 8S (singular) and ug7 (plural) that was suggested by J. Bauer in RA 64 (1970) and elaborated by P. Steinkeller, Orientalia 48 (1979) 55, does not seem to have been used in the Ur I11 texts from Girsu and Umma. ri-ri-ga means literally "picked up (things)." Kraus suggested the possibility of reading de-de-ga instead of of ri-ri-ga in Ein Edikt des Konigs Ammi-Saduqa von Babylon (1958) 114' because of SaVoc. F3'a [di-el [RII mi-qit-t8, but transliterated Id/ and frl alternate often without change of meaning. The Umma texts are identified in table 14. Gomi, Orient 20 p. 23 V 30-31, and table 18 (third cultivator of the bulls of Menkar). L SS 3. MVN 5, 180: a certain Namhani was found liable; he was a plow bull scribe according to T ~ 89 Cf. the collation in ASJ 15 (1993) 209. 24 rev.10 and 15 (2 years); 19,4-5; 38, 13-14; 46,40 (3 years). The copy of the line gives 1 gu4-giS SU-su mu-4. Expected is nu-su after su-su. The copy of this text does not make an impression of accuracy. Remarkably often in text a; further in 5 XU1 5'; 6 XV 6'; TLB 3,42. The sign nu could be very small and may have been overlooked occasionally by copyists. Lafont, in his comments to DAS 52, says about the nu signs: "quand on a la tablette entre les mains, on a I'impression que les signes NU en debut le ligne, plus petits, ont Btd ajoutks aprks coup." He goes on to suggest that it meant cancellation of the entry which does not fit the contexts known to me.
Plow animal inspection records
According to Maeda's collation in ASJ 2 (1980) 203. In OB lexical tradition equated with bbbktum; perhaps an old loan from Akkadian laqium "taker." SNAT 21 1 records the fact that a city elder guaranteed this very promise. The formula ur-re gu7-a was also used in Umma: Molina 5 (Aula Orientalis 8). Overlooked by me in the article "Liiwe" in RLA. E.g. 22, 21. 9 IV 20; 14, 18-19. 4 I 19; 34 11' 5'. ASJ 8 (1986) 90ff. STA 27 I11 4. Cf. my remarks on the subject of designations for castration in section 11. I have been unable to find an opinion on the etymology of the word Muni which is used in the dialect Alemannisch and guess that it derives from municipium which would make the Muni a centrally kept stud bull. Cf. MVN 17, 76 rev. 9': 1an[Se na]-gib-tum nu-sum-ma. G. Pettinato's transliteration here (I-da-tam) and elsewhere (su-su nu-su "gB", i.e. engar obv. 10') is erroneous. 1 X 6'-7'; 4 III 15; 5 IV 5-6; 26 rev. 10. The corresponding entries in the itemization list animals of the status su-su nu-su (III5 and 16). The restoration is deduced from the total and certain. Since "3 year old" was not consistently used, an adult animal is counted as one year older than a two year old etc. 6-thr seems to be a genitive construction because of gu4 6-thr-ka "bull of the house of pens" in 39 11' 3-4 where the phrase constitutes an absolute case and must therefore be a double genitive. It was usually not expressed graphically. Drehem (1992) 40. For the district division of the province see RA 88 (1994) 5-31. UDT 18; NA1W 408; SNAT 16, 17, and 19. 33 "W" = I' 22'-26' and "HI" = V' 20-22. 5 IX 25-27. 6 XI11 18; 9 XI 6'-7'. Einem geschenkten Gaul schaut man nicht ins Maul. The traces as copied by Limet suggest as possibility gig or hr or apin. giS would mean that 4 individual yoke b ~ l l swere meant, and further that each of them harrowed, which is unlikely. A reading apin is possible, yet most likely is the reading br. Cf. the similar text SNAT 14 5 40 IV where the sign is fairly PN2 i-dabs "6 harrow bull well preserved in Gomi's copy: 6 gu4-hr ki PN1 mu sanga *~~-alim-ka-l& (teams) has PN2 received from PN1 on behalf of the bishop of Igalima." 1 V 6-17; 4: groups 1, 4, and 7 in the second division; 5 VI 7-12; 10: group 8 in the first division; 12: group 5 in the first division and groups 3, 5, 6, and 8 in the second; 13: group 2; 16: group 3; 19 group 5; 21: group 1; 24: group 1; 34 I1 11'-16'. PISAN-DUB-BA Textsfrom the Ur III Dynasty, University Microfilms, Ann Arbor, # 76-27,824 (1976). 63-85. 110, 112, 113, 115, 116. MVN 17, 7 is a list of the possessions of a certain Urbau, among them "15 various yoke bulls, 30 various yoke donkeys, the plow animal groups being 5." The other references are: 1 111 12; 11 I 25. Cf. also "V" 15-19. The other references are: 1 1119; VI 12.
Heimpel
Plow animal inspection records
1 I
I disregard here the relationship to the sign LAGABxGU4 over GU4 which occurs in different contexts and is often transliterated dib or even dab5. dib-ba inspections of personnel are attested in the tablet box tags 79ff. Note that Nelson transliterates the sign dib erroneously with dib. For Umma cf. Maekawa, Zinbun 22 (1987) 37f., for Lagash cf. section 5 and table 2. Published by Maekawa in Zinbun 22. The distribution may have grown out of the experience that, as R.C. Hunt put it, "water theft is extremely common in imgation systems ... the desire to take water out of turn is nearly irresistible" (BSA 4 (1988) 197). If a cultivator stole water in order to raise the yield on his unit he would gain a profit because his allotment barley was a fraction of that yield. Yet since he would in effect favor one of his plots over another his gain would be canceled out by the loss in the other parcel. And even if his parcels were so located that they all would benefit from water theft, his bull lieutenant whose supervision extended over larger tracts would probably not etc. The discrepancy between the ratio in BM 110116, appr. 32, and BM 105334, appr. 49, is due to the fact that in the first textonly the area of the domain (grin-guq) was given, whereas in BM 105334 the allotment land of the agricultural managers was included. En-krir in text 7; ENxME-khr in AnOr 7, 313:46 and the "Non-sealed labor assignment" ASJ 8, 68f.; Me-en-[khr] in the comprehensive fall list and in DCS 27 1. A field was placed on the same level with districts in the format of BIN 5, 108 S 44, a list of silver payments of field tenants to the administration. The public tenant land in this text was identified as follows: Adm. unit silver for rent intermediary (Sr) Akasal 2.2.15 Uree Mushbiana Daagi Guedena Aabba a-Sii dSaira Lugalemahe Da Umma Urnamnunaka %i didli Lugalnisage As based on the assets and supplements of the 44 fully, or nearly fully, preserved group descriptions of the Urnma texts. Maekawa read 56-gu-la. The photo shows AN, not SES. "Donkey herder," which is also a title. Here it must be a name. Waetzoldt restores title and name of his supervisor as [ugulla Frl-DUN. Cf. BSA 7 152, table 15. BSA 1 (1984) 74; 5 (1990)129ff; ASJ 14 (1992) 210f. C. Wilcke, Das Lugalbanda Epos, lines 300-301. I cannot imagine how the trampling of oxen could have eradicated these rhizomes. If they are severed they sprout with redoubled vigor unless they are pulled out of the ground. Notice that this pulling of weeds is mentioned in line 8 of Georgica, and clearly separated from the action of the gu4-Subub contrary to Maekawa's interpretation. BM 105334 listed 3 types of land as available for cultivation, "good" (sig5), "medium" (murub4), and gig-gi a-ta-DU-a mu-4-kam-ka Sii KIN-KIN-dam which means possibly "drained reed thicket being worked (?) since (?) 4 years." Because they make a ready-made impression, starting en masse with the word nin "lady" which refers to Ninmar. Cf. Fish 881 = MVN 12 CXI-CW + BM 13650 = ASJ 10,86-87: Nin-tbr-e-si Ambar-sur-ra, Nin-zi-uru-na "Good lady of her city", i.e. Ninmar, Nin-gal-zu-gu-la, Nin-galzu-tur, Nin-ni-brim, Nin-e-gb-ba-an-db, Nin-Su-gi4-gi4(ATL 13).
Heimpel
Plow animal inspection records
For this term cf. Maekawa, Zinbun 22 (1987) 39f. The distinction between bull lieutenants and bull commanders was not always clearly drawn. The bull commanders of TENUS 346, and AAS 83, were called bull lieutenants elsewhere: cf. AnOr 1,78 AS 2 (A-gu-gu); TENUS 83 AS 4 (Da-a-ga); TENUS 361 AS 1 (Lfi-dulo-ga, Lugal-kh-zu, and ~r-d~n-lfi-lai); TENUS 374 AS 2 (1 of the 2 Lugal-giS.gigirre); TENUS 82 AS 6 (A-ab-ba); AnOr 7, 200 AS 2 among many other texts (I7-pa-&). The text was a fragment that was shaped by the excavator or an antiquities dealer so as to look like a complete tablet. The collator L. Cagni exclaims: non esistono rotture! Cf. Maekawa, Zinbun 22 table 1 [3]. Many prosopographical data on bull lieutenants in Umma, including Sheshkala and Uramma, were collected by M. Baldacci and F. Pomponio in SEL 6 (1989) 1lff. dub, mostly read KGib "seal". The meaning must be "sealed tablet", and cannot be "seal (of ...)" because of the existence of the formula dub-n(-hm). Cf. YOS 4, 207:183-185 ri-ri-ga dub-bi 7-hm dub Da-da-ga "fall constituted by 7 receipts, receipts of Dadaga". I therefore retain the reading dub. The reading kiiib, and meaning "seal", applies to the phrase kiSib ib-ra, as proven by, e.g., MVN 13, 825 and 826 where the recipient sealed the envelope with his father's seal and stated on the tablet: "they impressed the seal of his father". For this translation and more examples see M. Yoshikawa, ASJ 9 (1987) 293-302. Gomi restores the name of the father as Lugal-ni-[LAGAR-el. Lugalkuzu: supervisor of Udaura in text 1 (SACT 2,250) and BIN 5, 111 (undated); mentioned parallel to the well -known bull lieutenants Uramma, Ipae, and Sheshkala in the inspection text of SS 6 published by Baldacci and Pomponio, 1.c.; titled bull lieutenant in YOS 4, 164 AS 1 and the comprehensive deaths list of SS 5 (cf. plate 18). Lugalemahe: titled bull lieutenant in the same list. Cf. BIN 5, 111 (undated) 10-11: 1 gu4-jjii Lugal-ezen I ugula Lugal-ugkin-nb engar; and AAS 83 V 21 for which cf. Maekawa, Zinbun 22, 55 note 2 to table 8. His emendation is unnecessary. According to gu4-apin gub-ba text 4 111 6 the cultivator Lubau served as a supervisor. It is not known whether the term "deep plowing" for TUG2.KIN or T U G ~ . ~ E . K I Nis appropriate. It designates the task of seed bed preparation that preceded harrowing. Isaac Adams, Persia by a Persian (1900): When they use a pair of oxen for plowing they make a very shallow furrow. If they use two buffaloes (sic) it can be made much better but even then the furrow is not deep enough to be really good. .... The best and in every way most satisfactory plowing they do is when 4 farmers, each owning a pair of buffaloes, club together and do their plowing. Then they get a very large plow with a big plowshare and hitch the eight buffaloes to it." Joseph Knanishu, About Persia and its People Rock Island, Ill. (1899), p.111: One man drives each team, sitting on it to do so, while a fifth man guides the plow." On page 112 is a picture of 4x2 buffaloes before the plow. Brehrns ~ierlebenl Saugetiere, vol IV plate XVIII opposite p.344, shows 6 very large Hungarian oxen before the plow.
Heimpel
Plow animal inspection records
1 Gu,-apin gub-ba texts from Girsu text RTC 265 TUT26 Biga SEL 3 Arnherst 38 CT 5,29 BM 19024 HLC 134,80 Amherst 52 Gomi Orient 16,136 HLC 120,370 BE 212,124 CT 5,21 BM 12913 CT 5,27 BM 18933 M. Lambert RA 67 CT 7,49 BM 19774 CT 7.49 BM 19775 Watson 2,291 Gomi Orient 16, 144 MVN 9,32 MVN 9 33 CT 7,30 BM 18387 CT 9,26 BM 19769 CT 9,22 BM 19054 CT 7,44 BM 17766 CT 10,43 BM 19778 CT 10,42 BM 19064 CT 10,43 BM 19053 MVN 2,84 Scheil RT 17,29 UDT 69 UDT 71 CT 9,40 BM 14617
BE 2/2, 126 TCT 1,988 TUT 43 DAS 52 TUT 25 HLC 111111,243 TCT 1,877 TCT 1,899 MVN 13,433 MVN 13,583 OBRT 178 CT 10,46 BM 19772 CT 7,18 BM 12944 CT 7,17 BM 12938 TUT 33 MVN 9,111
date U6 3 41 S 43 S 43 S 45 S 46 S 47 S 47 S 47 $[I S[ I
S[ 1
AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 2 AS 3 AS 5 [AIS [ I [ I [ I [ I [- 1 [SS I?] [ I [ I [ I [ I
I
type A A A A A A A A B A A A A B B B B B B B
B B B B B B B B B B C A A A A
B ? C A A B B B
B
-
B B B B
oxen [1 15 14 6 15 15 34 n 5 18 10 8 13 2 3 2 2 2 3 5 4 3 3 5 4 5 3 4 4 3 51 38 19 71
n n n n 180+ n n n lor2 3 3 5 3 3 1
governor Igalima Ningirsu Shulgi
I
Nanshe [Namhalni Ningishzida Nindara Lugal-URUxGt Ninmar Ninmar Ninmar Ninmar Ninmar Ninmar Nanshe Nanshe Namhani commander Ningirsu Ningirsu Dumuzi Dumuzi Shulgi Shulgi Shulgi Ningirsu Ninmar Nanshe Ninmar [ I Gishbare, Shulpae, [ [ I [ I entire province [ I [ I [ I
2 2
-
2 2 1
I
~umuzi Shulgi Shulgi Ningishzida Ningishzida
-
Heimpel
Plow animal inspection records
Heimpel
Plow animal inspection records
3 Overview of unit numbers per household and plow animal groups in Lagash household
( Ningirsu
I
gaGa Nanshe Ninrnar
I
I(,j
TUT 5 S 47 HSS 4,4 AS 2 62
I
98
I
70
I
37 [SS 11 others [ ] I 15/10 S 46 (5); 51/27 AS 3 (30)
Gatumdu Dumuzi Nindara Shulgi Namhani Arnarsin Ningishzida Igalima Lugal-URUxGt Ninsun Numbers to the left of the slanted line /refer to oxen, to the right of the line to donkeys. TUT 5 did not distinguish the species of plow animals. Numbers in brackets ( ) correspond to those of the texts in the corpus. * MVN 12,428.
1
Heimpel
Plow animal inspection records
Heimpel
Plow animal inspection records
Plow animal inspection records
Heimpel
6 Sample descriptions of three groups in text 12 G
5 Structure of text 12 G Itemization of first division of household: bull lieutenant cultivator dUtu-kam Ba-zi Nin-a-na
supervisor Ur-dNu-muH-da
Ur-Sa6-ga Lugal-mu-dah* totals of animals (SuniD) count of plow animal groups: gu4-apin-bi7 ade-apin-bi 3 identification of supervisor: ugula Ur-dNu-mul-da Itemization of second division of household: bull lieutenant cultivator IUr-dBa-6 1Ur-dL6-liil*
3
I I
2* anSe 5* dhr l * anle su-ga engar 1* (nu) d h z i - g h l* dhr-amar-ga gub-ba-hn 16-12 anle-2 ri-ri-ga Ma-an-sum engar
2 mares 5 stallions 1 mare replaced (by) the cultivator 1 non(present) sikkum stallion 1 stallion foal is standing deficit 2 two year stallions deaths cultivator Mansum
4* gu4-gig I* 6b-mah su-ga engar mu 6-gu d6-a-Sb l * 6b-2 su-su nu-su im-ma engar nu-bbda-gu, gub-ba-iim 16-1gu,-gil ba-TIL libir-2un 1* gu,-gil ni-kas, na-g6b-tum-ta Ur-Sa6-ga engar
4 yoke bulls 1 cow replaced (by) the cultivator for cause of having been lost 1 two year cow yet to be replaced since last year (by) the cultivator and bull lieutenant is standing deficit a dead yoke bull is old 1 yoke bull from the account of the NaGABtum cultivator Urshaga
4* gu4-gif l * gu, su-ga engar I* gu, Su-gi, gub-ba-iim 1 gu, Su-gi, zi-ga libir-hm l* gu4-gif su-ga sanga ki L6-dGig-bar-b engar-ta I* 6b-m6h ni-kas, na-g6b-tum-ta Lugal-sipa i- [dab,]
4 yoke bulls 1 bull returned (by) the cultivator 1 shugi bull is standing 1 shugi bull expense is old 1yoke bull returned (by) the bishop from cultivator Lugisbare 1 adult cow from the account of the NaGABtum Lugalsipa was in charge
'
-
I
I
I Lugal-zs-ge-si totals of animals 6 m E . n ) count of plow animal groups: gu4:apiGbi 6 ade-apin-bi 4 identification of supervisor: ugula Nam-mah grand total (Sukilib) of animals summary (kilib-ba) of animals summary of count of plow animal groups subscript: gu,-apin gub-ba / dLugal-URUxGt/ Ur-dLamma/ Cnsi / AS 2
The stars identify donkey groups.
I
z
assets
deaths
.
assets
deaths supplement
.
Asterisks after numbers identify numbers written with the number stylus.
assets
withdrawal supplement
-
I
1
Heimpel
Plow animal inspection records
7 Designations and quantities of plow animals in Girsu and Umrna texts Urnma &-mi& gu,-@I 6b-mu-3 gu,-mu3
6b-mu-2 gu4-mu-2 &mu- 1 gu,-mu-1 6b-ga gu4-ga Hu 6b Hu gu4 erne6-m6h dk-@I erne6-muS dhr-mu3 emebmu-2 dk-mu-2 emecmu-1 dk-mu- 1 erne6-ga dhr-ga
.,su eme, Hu d h
gu,-6b-hi-a anle-hi-a
Girsu 65 6b-m6h 253 gu4-@I gu4-%Idiri 9 6b-3 7 gu4-3diri ib-2 @&I& ga4-2@I-:& 4 fib-2 11 gu.-2 12 kb-1 46 gu,-I 8 6b-amar-ga 8 gu4-amar-ga amar-ga 1 6b-Iu-gi, 2 g~,-Hu-gi, 51 anIe eme, anHe amar-nil-a 64 d b 2 anIe-3 6 anIe-2 @:-It db-2 @I-It 2 anHe-2 anIe-2 Hu-gi, 2 db-2 db-2 Iu-gi, 5 anHe-1 3 dk-1 8 anIe-amar-ga 9 dhr-amar-ga amar-ga dilr-GABA ans'e-Hu-gi, gs-st 2 anIe-Iu-gi, d b - 1-Hu-gi, 3 dilr-Hu-gi, dh-zi-gb anEe-kung6 anle-kung8-2 dh-kung6 6b-gu4-hi-a anHe-hi-a 6b-gu,-gig anIe-BiI #
-.v
Y
translation cow yoke bull spare (?) yoke bull 3 year cow 3 year bull spare (?) 3 year bull 2 year cow (fit) for the yoke 2 year bull (fit) for the yoke 2 year cow 2 year bull 1 year cow 1year bull milk cow calf milk bull calf calf Hugi cow Iugi bull donkey (i.e. mare) mare mare without foal stallion 3 year mare 3 year stallion 2 year mare (fit) for the yoke 2 year stallion (fit) for the yoke 2 year mare fugi 2 year mare 2 year stallion Iugi 2 year stallion 1 year mare 1 year stallion milk mare foal milk stallion foal foal text in order? Iugi mare (fit) for the yoke Iugi mare hgi 1year stallion Hugi stallion siKKum stallion mule mare 2 year mule mare mule stallion various cows and bulls various donkeys yoke cows and bulls yoke donkeys
148
# or reference
171 641 31 VII 15' 11 1 VIII 11 8 6 13 14 18 34 23 28 text 2 2 18 254 31 passim 9IV4 379 2 7 8 7 1 XII 34 12 1 11 11 30 28 34 39 text 2 26'10 4II8 5 1 7 2 7 2 1 in totals 99 99
"
Heimpel
Plow animal inspection records
Plow animal inspection records
Heimpel
Heimn-l
Plow animal inspection records
10 Sums of a(ssets), ~(upplements)and deficits (15'1) of working animals in text 31 G
9 Grand totals and summaries in texts 5 and 10 G text 5
Sukilib 30 anse 41 diir 3 diir su-ga 2 d~r-2 1 ange-2 su-ga 1 diir-2 gig-3 1 an6e-1 2 diir-1 4 an&-amar-ga 6 diir-amar-ga
kilib-ba
1 75 anHe-EiH
group 1 2 5* 9* lo* 13* 14* 15* 16 17 18 19* 21
first line I1 I7 I31 m3 In 11 IV 1 IV 6 IV 11 IV 22 V2 v 12 V 23 VI 17 VII 11'
27 28 29 30
VIII 16 VIII 22 IX 2 IX 10
a 7 5 8 6 8 8 8 7 3 5 4 5 5
s
-
2
3 2 2
-
2
16-'i s u m 7 7 8 2 8 8 8 8 1 8 1 7 7 1 7 8 3 7
I
text 10
2
7 7
Asterisks denote donkey groups.
11 Working animals in texts 12 G (left) and 4 G (right) group first line
The restoration [47 gu,-gig] derives from the number of the summary; the sum of the totals - as they are copied - is 46.
5* 6* 7 8 9* 10 12* 13 15 16 17 18* 19 20 21* 22 23*
I1 4 I1 12 I1 28 m2 1117 111 19 IV 10 V 26 VI 11 VI 19 VI 24 VI 31 W 16 VII 20 VII 30 VIII 5 VIII 10
sum 8 9 11 6 9 5 7 5 5 6 5 8 6 7 10 8 10
24
VTTT 19
6
1
3+ = "3 year" animals and adults; 2 g i g = "2 year (fit) for the yoke; 2 = "2 year" animals; s = supplements; a = assets.
Plow animal inspection records
Heimpel
Plow animal inspection records
Heimpel
12 Redistribution in text 12 G
13 Persons responsible for replacements in Girsu texts 8 PN 106 1 1 1 PN 5 21 1 1 2 PN 32 1 PN 1 PN 2 PN 2 1 PN 5 1 2 PN 10 1 1
engar engar 6-gu d 6 a engar engar engar engar engar engar engar nu-bh-da-gu, nu-bhda-gu, nu-bhda-gu, nu-bhda-gu4-ta dub-sar gu4-10 dub-sar gu,-apin dub-sar gu4-apin Habra Habra dub-sar na-g6b-tum sanga sanga na-g6b-tum 16-zi-giim
ii PN ti PN, ii PN engar nu-bhda-gu, ii nu-bhda-gu, ii dub-sar-gu4-10
19 rev. 5-6 121 31 14'10 9 VI 1 5 VII 3-4 9I l l ' 12 IV 2-3 26 rev. 6-7 18,22 25'7 1214 6 X 11'-14' 31 V 25-26 4 1 14 6 VII 1 31 IX 10 43 rev. 13 14 rev.1-2 41X 1 10 V 37 4 VII 9 46'40
The first column gives the number of references in all texts. Each integer corresponds to one replacement animal. The 5th column gives one reference as example.
Asterisks denote donkey groups.
Heimpel
Plow animal inspection records
Heimpel
Plow animal inspection recordh
14 Substitutions in Girsu texts substitute B-mah Ab-mah 6b-mah fib-ma su-ga nbg Ab-mah su-ga engar Ab su-ga engar 6b su-ga
for gu4-gil
B-~plah 6b su-ga engar gu4(gis')
gu4-gif su-ga engar g"4 gu4-gii gu4-gif gu4-gif gu4-gf gu4-gif Ab-2 gu4-2 Ab-mAh
gu4-gig su-ga gu4-giZsu-ga engar gu, su-ga engar gu4 gu4-gig gu4-gif
6b-mAh Ab-m6h Ab-m6h Ab 6b-3 gu4-2
gu,-gig su-ga engare ... gu4-giH gu4-giH 6b-3 Ab-2 gig-Zb 6b-2 gig-fk 6b-2 gig-Zb 6b-2 6b-2 6b-2 su-ga gu4-2 gu4-2 gu4-2 gu4-2su-ga gu4-2su-ga gu4-2su-ga [x] 6b-1 Ab- 1 gu4-1 gu4-1su-ga engar gu4-1su-ga gu4-1su-ga engar gu4-1 gu4-1su-ga engar gu4- 1 gu4-1su-ga engar gu4-amar-ga su-ga engar gu4 fu-gi,
gu,-2 fib-1 gu4-1 Ab-m6h Ab-m6h gu4-gif Ab- 1 Ab-m6h gu4-gif diir Ab-miih gu4-gig gu4-3 gu,-3 Ab-mAh 6b-2 gu4-gif gu4- 1 6bm& gu4-gig gu,-gif gu4-gif gu4-gif gu4-2 6b-1 diir Ab-1 gu4-2
ab
references 12 VIII 30 (2); 22 rev. 4; 23'14; 24 rev. 11; 25'4; 32 VII 14'; 34 11118 (2); 39 V' 5' 46'25 26'18 20'6; 44'8 20.24; 44 rev. 5 (2) 20'39 34 111 19 12 IV 21 32 VII 15' 34118 10 II 38; 11 IV 6; 12 VIlI 32 (2); 23'6; 24.4; 19; 31 I 3 (2); 11 end; 32 VII 17'; 38 I1 5' 19'8; 31 X 8 (2) 14'16; 18'3 31 I 4 331112 11 IV 12 6 XI1 5; 10 11136; 11I 2; III 20; 32 VII 18'; 38 1 7 18'22 24'3 23-15 1 VII 11; VIII 9 3 1 I1 end 8 rev. 5'; 12 IV 24 23 rev. 3 16'6 6 XI1 3; 8 rev. 6' 32 VII 21 ' 3 111 13 10 VI 24 1 XI 43 4 IV 26 31 X 12 34 I1 21 2 111 9 7V4' 38 IV 2' 31 V 14 4 IV 29; IX 22; 31 X 7 (4) 27 rev. 4 219;115:31VI20 24'18 17'21 31 VIII 12 20'3 10 IX 7
14 Substitutions in Girsu texts substitute
for
references
anfe anfe su-ga engar anf e su-ga engar anZe su-ga engar ange su-ga nbg anZe su-ga nbg h e anfe anf e ans'e dh d h su-ga d h su-ga dhr su-ga d h su-ga Zabra d h su-ga dhr su-ga d h su-ga sanga dhr
dhr dhr gu,-giS dhr-1 gu,-gig gU4 dar anf e-2 dhr-2 anHe-1 gu4-gig-li?su-ga engar gU4 gu4-gig gu4-giS-g&!engar gu,-gig anHe h e dhr-2 anfe
d h su-ga dhr d h su-ga engar dhr dk-"gaba" d h zi-@m dhr dhr su-ga engar nbg dh dfir anZe-3 anfe-3 su-ga anZe-2 giEfb anfe-2 gig-:& anfe-2 giS-fb anZe-2 su-ga anHe-2 dhr-2 gig-fb dhr-2 gig-fb dhr-2 gi;i&s'& dhr-2 dar-"a" (= 21) dh-2 su-ga dh-2 dhr-2 anfe- 1 anle- 1 dhr-1 dhr-1
dhr anSe-2 anfe-2 dhr-2 anfe dhr-1 dhr-1 dar-1 anIe Zu-gi, d h fu-gi, su-ga nbg anSe ange anHe dhr anie-1 anfe anle- 1 gU4 dh-3 anZe-1 h e dhr dhr d6r-3 dhr-1 anf e dhr- 1 anHe-1 dh-2
10 III42 (2); 17'2; 18'15 22'18 42'7 18'16 44'2; rev. 13; 45 rev. 2; 46, 16 45 rev. 3 38 III9' 10 IX 37; 29'15; 42 rev. 10 10 IX 39 27'13 25 rev. 3 5 IX 32 32 VIII 4' 42'9 44 rev. 16 33 I1 7 31 X end 24 rev. 8 4IV35;5127;6XII21; 1OX 1; 111 22; 16'18; 33 119 (3); IV 7; 43'20 4 IV 36; 32 VIII 5' (2) 22'4; 42 rev. 6 46'3 5128;613; 101114;32~1112' 26'10 11 I27 10 III 7 39 VI' 10' 3 III 3 3 III 5 1 XI1 6 1 XI1 8 6XII16; 1 0 x 1 1 6 XI1 18 10x3 5 VIII 17 1 V 22; XI1 17; 5 V 7; 15, 13 5 VII 25 39 V' 11' 10 IV 3; VIII 14 32 VIII 7' 10 VI 30 4x2 1 VI22; 1 XII 13 4 IV 38 17'22 6 XI1 23; 10 I11 27; 18'12 (2) 10 VIII 29; 39 V' 13' 33 V 10
4
Plow animal inspection records
Heimpel
Plow animal inspection records
Heimpel
15 Numbers for assets, withdrawals and deaths in perfectly preserved or restorable summary sections in type A and C texts from Girsu. The numbers of the category "to be replaced" are excerpted from the assets. text 4
assets withdrawals deaths to be replaced 5 31 17
1 211
* Numbers of the household of Ninmar.
u
S2 Q1
f
t
4 'nr
a
C1
Y
n
Q)
E
Q)
U
I
?i
a$-n%ua!z-nurnap-~nq al-ie%ua~@-:s-~a%nl q ( ~ a s u a v~-.xa%-ys-p%nq ) q a~-~o%ua O1@-nlnpq at-a%uauaza-~a%qq m-n%uaw q z - ~ q n
[I)
% 0
E c
.-0 4-a
U
4)
.-3
;@-*n% z JQP I JQP Z ;r%-*dI st%-*& -.I aaua I
6 I ZI 62 III 6 ST16 OZ X 9 PZ A1 S 91 IIIA P
s%-*n% I JQP Z
*8 A1 E *9 A1 Z
P E ague I a!uv [1 *n% [ I
*61 III €&
4
m
.-E
s
Z
I
uognqyspai p m u w y war^ :L al-p?%-n;-~~%n? q [s-]~a9u.l[ 1-aq-qv PI NNd woq :9
..
I
a;-e%-p-u *n%nw
aiqa; a%-ns [i] ns-ns a;-a%-p-u a%nw [i] ns-ns nl-v el-io%uaoy-auau~,;y nwnp ey-auaqp aw-wy a-I$ ns-ns %qu al-~eSuaqn;l-:~-n~ 17 a%-IS-aqq-inaw-wy a-;@ ns-ns [ ] ~ I - I ~ Y -ns-ns V NV-a%-xnwnp wns-uo-aw ns-ns ?;-a%-u-u-'n% nw ns-ns 3-a-9p-n% q - ~ a s u an s - q ~ - q ~q, - ~ -? ~ nw %quo p - ~ z - ; g - q ~ ~ns-ns -~n ?;-a-9p-&-y ol-na%uaOIn&rrq-qv ~q nw it&-opqq-nu aWWt?T,-ln l'ls-IN [%quap]-!z-;@-[y~,-ifl ns]-ns '&-apqq-nu ns-ns %qua%-ns a( a;-a%-U-u nw a%uasa%-ns a;-a%-u-u-*n%nw %qusaw-in a%-ns ,,nq 30 asnaa io~,,:S a-O1asal-qy LL
66
+ e,
$ 0
2
-8 c
.-0
R
-5 .-C .-
a-O1as ttl-~y
a-O1as01-rn BI-QT a-O1asa l - ~ y E-O1ass-QJ a-O1as01-M
I
9A6
JQP(nu) JQP(nu) *&(nu) *n%
5
S
£I A1 6 I€ 1116 61 III 6 *8IIIM9
JQP(nu)
I
ZI 1x9
JQP(nu) *n%(nu) 'd (nu) *& *n%
I I I I Z I
I I x9 ,LXI 9 01 A1 9 *SZ XI S 9Z A1 S
JQP (nu)
I
S 'LP
a?ua
[1 [I I
*&€Ix~ 61 IA 6
*
* '&
*
bL
:
I
r
I I Z I I
a;ua JQP Z ;g-*n% €1 ague 81.
t?-O1I?sB1-QT
5
*
n
E-O'BS
6L
*&9IX6
JQP(nu)
n$*&
I-~'~S [el-~y]
I
LIA6 PI III6 IZII6 IIII6 LII6 *PM L
,-..
L
LL LL LL
an-*n%gas a%-ns az-*n%%as a%-ns a;-*& gas a%-ns
LL
0~-ml-qy3-auC[IS-P .
JQP JQP a;-;@ z - q p JQP JQP
P I I
I I
OZX I
*8IAZI *O€XI S LZIIAS
A.
I-qP
LL
8
.-F
z0
I I I
9awa s s@-*n% zg twrssad LE
a(
-
2-1PP ??-?I% Z-1PP JQP
soxal n s q 9
squawa~ddns91
I
Z
0\ V1 3
4
16 Supplements in Girsu texts
12 I 2 4 12 1122 12 m 23 12 m 25 12 III 27 12V33 12VI11 12VI36 12VIII4 12VIII 14 12 VIII 16 31IV29-31
1 1 1 1 1 4 1 1 1 1 1 1
gu4-gig &-gig gu,-gii. gu4-gig @,-gig dh db gu, gu4-giS gu, gu4-gig gu4-1
31 VI 10 31 VII 14' 32 I 2 1
1 1 11 1 1
gu,-g[iS] db gIud4 anfe dh
ki Lugal-mu-dah engar-ta
ki Lugal-mu-dah engar-ta
su-ga nbg su-ga sanga su-ga Gu-gu nbg su-ga engar su-ga sanga su-ga engar su-ga im-ma Ur-6-ninnu sanga sag gu,-gig-Sb su-ga engar sag gu4-gigdiri apin-St su-ga sanga [iml-ma sag dh-Sb
ki Lugal-mu-dah engar-ta ki Lugal-sipa-ta ki Ur-dDurnu-ziengar-ta ki Ma-an-sum engar-ta ki Lugal-kd-zu engar-ta ki Lugal-mu-dah engar-ta ki ~b-4%-bar-b engar-ta ki Ur-dDumu-zi-da-ta ki Lb-dUr-dAsari-ta ki Ur-6-gibil-ta ki Ur-dBa-b-ta ki Ur-dNin-pirig-ta [ki Sigl-su,, engar-ta [ki] Lb-dInana engar-ta 9,
-.w
ki Lugal-sipa engar-ta
'Asterisks mark data from totals. Preserved for the households of NieiSzida, NanSe, Sulgi, Nindara, Dumuzi, Amarsin, representing 134 gu,-apin and 108 anSeapin. 3The dativ refers probably to the cultivator of the group to which this entry belongs, a certain [Ur-mles. 4Lafont transliterates [la]-1.
8
s.
E
17 Plow animal inspection texts from Umma
w
s
text
date
a
1Wengler49 2 Watson2,7 3 MVN 14,556 4 Watson 2,11 5 UTI 2093 6 UTI 2281 7 MVN 4,76 8 TIM 6,55 9 MVN 8,210 10AnOr1,288 11TIM 6.56
AS 2 VII AS 2 AS 5 AS 8 VI SS 3 III SS 5 V SS 5 IS 2 IX
112 NATN 376
IS 33? V )+
[
w d s
+ - - + + + + + + + + - - + + + - + + + - + + + + + + + + + + + + + - + + + +
1
r
-
+
-
Dubu-sig,,
gunim aka gu, %u-la gun5m aka gu, zgapin &nun aka gu, Sbhub Gb-eden-na Mug-bi-an-na "[ang]e!"Gb-dbna gunim aka! gu,APIN 5-kam g d m aka gu,-guhbb APIN 13-kam Da Umma"
Ukb-nu-ti FA 6duq dSul-pa-b En-k&r
-
+ + + + + + + + [
I- I+ I+ I-
rra
subscript gu, Huhiib
1
1
I* Igu, a n b si-la engar
I
11?17
1
a = assets; r = replacement; w = withdrawal; d = deaths; s = supplement; SA = "in" referring to the place of inspection; n = nu-bhdagu,; e = engar, * = ki engar-ta gur-ra.
Heimpel
Plow animal inspection records
Plow animal inspection records
18 Comprehensive deaths list of Umma: BM 106041 + 106045 SS 5
18 Comprehensive deaths list of Umma: BM 106041 + 106045 SS 5
Asterisks in the first column refer to donkey skins, the others are of oxen.
Heimpel
Plow animal inspection records
20 Working animals per group in Umrna
To1 = Total of all categories; To2 = Total of adults of assets, replacement, and supplement. (7) replacements in text 8 are the total for all groups.
Heimpel
Plow animal inspection records
Plow animal inspection records
Heimpel
22 Totals and sumtnaries in text 8 21 Supplements in Urnrna texts text
supplement
total (gun in) section 12 6b-m6h gub-ba-hm assets
summary (killib-ba) section 95 gu4-6b-hi-a gub-ba-am
I
1 gu4-gibki KAS4-ta 2 gu4-gibki KAS4-ta 2 gu4-gi6ki KAS4-ta 5 gu4-gigki KAS4-ta I gu,-gib He-"?-?" ki Ur-Ell-e-ta 1 gu4-gibbe-ta sale-salo ki I&,-ta 2 gu4-gi6ki KAS4-ta 1 gu4-glsse-ta sale-sa,, ki Iq,-ta 1 dhr-gig be-ta sale-salo ki Ir,,-ta 1 gu4-Bbki US-mu-ta r."
Y
1 eme, mu-1 ki US-mu-ta 1 diu-gi6 ki US-mu-ta 1 dhr-gib ki US-mu-ta [8] gu4-gi6ki US-mu-ta 1 6b-mah ki US-mu-ta 1 gu4-gigki US-mu-ta 8 gu4-gibki US-mu-ta 3 gu4-gibki US-mu-ta 8 gu4-gi6ki US-mu-ta 2 gu4-gibki KAS4-ta 1 6b-m6h s'e-ta sale-sa,, ki Ir,,-ta 1 eu,-Zib ki US-mu-ta
1. su-su regular replacement 2. su-su deficit replacement zi-ga-bn withdrawal ri-ri-ga-h deaths
libir-bn supplement]
3 gu4-6b-hi-a 1 dhr-gig 1 bu erne, 2 6b-m& 3 gu4giS 1 6b mu-2 1 Ab mu-1 I 1 bu gu, "is old"
2 ans'e-hi-a
zi-ga-h withdrawal
I
! 8 gu,-6b--hi-a
ri-ri-ga-h deaths
I
I
Heimpel
Stars outside the text column refer to the designation TIL 1.4.5: TIL 1 gu4-giE[su-su engar] or TIL 1 gu, m[u-n]. 5.2: TIL 1 gu,-@;is'AN X. 9.1: 2 [hb] or [gu,] amar ga. f = female; m = male; f3 = female "3 year" etc.; fO = fe,ale milk class; rnS = male Eu animal.
Plow animal inspection records
Heimpel
Plow animal inspection records
Plow animal inspection records
Heimpel
26 Registers of outstanding replacements in Umma
25 Replacement in Umma texts
text I1 SACT 2,250 2 MVN 14,252 3 PIOL 19,257 4 UTI 2127 text 1 2.1 2.2 2.3 2.4 2.5 3.1 3.2 3.3 a 3.3 b 3.3 c 3.4 a 3.4 b 3.4 c 4.1 4.2 4.3 4.4 4.5 a 4.5 b
Stars in the text column and row 8 A(ppenix) designate donkeys; the other stars designate the formula TIL 1.
date IAS2 SS 3 SS 3 SS 3
subscript
1-
gu4ange su-su engar-e-ne ,, ,,
dub
ILugal-M-zu Lugal-ugkin-n6 Lugal-bmah-e Lugal-kh-zu
status animal 1 fib-mu-2 sag fib-mi%-gb SU-SU SU-SU SU-su SU-SU SU-SU
nu-su
SU-SU
su-su engar su-su engar su-su Lugal-bmah-e nu-su su-su L6-Ib-gal nu-su 6-gu-d6a su-su Lb-Ib-gal su-su dUtu-bar-ra su-sol engar su-su Ur-AB-zu SU-su nu-su su-su nu-su SU-su nu-su SU-su nu-su su-su nu-bbda-gu4ii engar nu-su su-su engar nu-su
cultivator U-da-ur4-ra dSar6-zi-da
I
OLD BABYLONIAN CATTLE Marten St01 (Vrije Universiteit,Amsterdam)
We will use the word "cattle" for bulls, cows, and calves; in other modern languages "large cattle" (gros be'tail, Groflvieh).l A. THE ANIMALS
Collective "cattle" Almost always the combination of the Sumerian signs hb.gud.N.a, "Cow-Bull-Plurality", stands for cattle. Rarely we find them in the reverse order, gud.hb.~.a.~In one of these texts, the totals offer [email protected] as subtotals and gud.6b.u.a as grand total (YOS 14 311). When calves appear in a list of cattle, the scribe sometimes feels free to add amar "calf' to the signs combination: 6b.gud.amar.M.a (Riftin 58: 19). Similarly, one list, beginning with 6b.mafi "fullgrown cow", has the summation 6b.mahgud.M.a (YOS 8 164:8). These are idiosyncrasies of particular scribes and we need not look for parallel words in Akkadian. What is the Akkadian equivalent of [email protected]? B.Landsberger and A. Goetze saw that the Akkadian word is feminine plural (as in 6b.gud.u.a damqiitum uggiii, AbB 10 32: 13), and they suspected that the word is li'iitum (ldtu), the plural of littum F COW".^ The Chicago Assyrian Dictionary indeed gives a number of references for the Sumerograms 6b.b.a and 6b.gud.M.a under littu: "cow"; (in the pl.) "bovines" (of both sexes and all ages). Now, it is a problem that the word littum is hardly attested in the Old Babylonian texts. There is only one reference, in a letter: "[ ...I and the pregnant ones are well; the littum is well (li-it-tum $a-alma-at); the herd of PN, pregnant ones, are resting in GN (su-ku-ul PN a-ri-a-tum ina GN rabgii)" (AbB 10 180:3-5). Here, littum does not mean "the cow" but something like "the cattle". The wording of the entire phrase is unusual: littum, sukullum and rabagum are not attested in the everyday language of letters and there is perhaps a literary flavour to the whole passage. I think that the singular form littum is behind the Sumerogram 6b.gud.M.a in a letter: 6b.gud.M.a dii1i3tum[sing.!] "threshing ~attle".~ The plural li'iitum is attested in only one Old Babylonian text, an entry in the lexical series Proto-Diri: u'-tu-ul Ja li-a-ti "herdsman of cattle". We translate "cattle" because here it is followed by "the herdsman of sheep and goats (g2num)".5 At the end of the section "Breeding" we will say that "the majority of animals in the herds may have been female". This accounts to some extent for "cows" meaning "cattle". Tricky are those passages where gud.M.a does not mean "bulls", or "oxen", as usual, but "cattle", including the females (as in YOS 14 311, CT 8 30a:l). One Old Babylonian letter, after have spoken of gud.M.a "cattle", recommends to separate the bulls (alpu) from the cows (burtum) (AbB 9 83). The general meaning "cattle" for gud.M.a is in two other passages Bull. Sumerian Agric. 8 (1995)
173
St01
Stol
Old Babylonian cattle
Full-grown cow The Sumerogram for a full-grown cow, the animal often heading the cattle lists, usually is hb.AL. We know that this AL was pronounced as mab.12 Thus, the reading is bb.mhh. This is confirmed by the two cases where we find in the first line of lists "simple" aib.mah.13 It has not yet been seen that in the sequence 2 rib am-ma-bu 2 aibpu-ru-se-e-tum, the first qualification must be ammahu = rib.m& (cf. gumQhu = gud.m&) (CT 8 28c:ll-12). We will retain here the writing rib.~L,just for the sake of clarity. We add that a full-grown goat was named uz.AL (VAS 13 86: 1).
indicated by its predicate which appears in a feminine plural form, typical of the collective "cattle" in Akkadian. Those passages are: - "The 'bulls', these and those, are well" ([email protected] anniatu u anniatu Zalma) (AbB 5 218:7); - "PN, (walking) behind the threshing 'bulls' " (PN arki gud.N.a mu-di-?a(!)-tim) (YOS 12 421:2).
Ages Lists of cattle from Larsa and Ur have been studied by B. Landsberger, followed by F.R. Kraus. They did not see the roster of cattle from Sippar published by A. Goetze.6 Here, we find the word salatum correctly interpreted by F.R. Kraus as "Viehzahlung". This text names itself in col. I, line 6 sh-la-at mu NN, "the counting of cattle over Abi-eSub year n".7 Other important lists are YOS 13 348, 350-1 (from Kish), Riftin 90 (from Larsa or Bad-Tibira). Several of these texts confirm Kraus' thesis that this stock-taking took place in month VIII (cf. AbB 2 15). Landsberger and Kraus did not point out that the terminology in their Ur and Larsa texts (in the South) partly deviates from the Sippar and Kish texts (in the North). The terminology for the animals up to four years old is different, as we will see below. These lists always start with the full-grown cows and descend to the suckling calves. We will adopt this same order. It is a problem what the specification mu.112/3/4 "year 1121314" after an animal means: "in its first, etc., year", or: "one, etc., year old" ? In the preceding Pre-Sargonic and Ur I11 periods the latter is meant. In the Ur I11 text "The Growth of a Herd of Cattle", gud.amar.ga (year 1) is followed by gud.mu.1 (year 2).8 The OB roster published by Goetze lists the "weaned calf' (amar.ga.KUD) among "bull-calves in their second year" (amar.mu.2), and the "suckling calf' (ramar.ga1) among those in their first year (amar.mu.1) (JCS 2 105 no. 9: 14-5). As our translation already shows, we assume that amar.mu.1 means "bull-calf in its first year", because they appear to include the suckling animals. And this is why the roster has no separate entry for suckling calves. A text from the South offers aib.ga.mu.1, "female calf in her first year, (still) suckling", so we s ~ p p o s eIn . ~Nuzi, the Akkadian specifications "third", etc. mean "in the third, etc., year"; so actually two years old. The same has been said for the MB and NB periods and it is also true for MA. '0
Qualities of mature animals "good" (damqum), passim "beautiful" (baniim) (ARM 1 118:6) "reliable" (taklum) (AbB 2 86: 10, 31; UET 5 31:21) "ripe" (baZlum) (VAS 16 93:32) We will not study the qualifications given in the Nippur archive studied by Kraus.11
Old Babylonian cattle
Two examples show that these cows can be pregnant (peg4 "pregnant", or Lkba "with foetus"): - 14 6b.mahN.a S[h].bi 10 sib.Sh.ba "14 full-grown cows, among which 10 pregnant cows"; - 59 hb.mah@.[a]Lh.ba 38 Ab.rpeL4l ''59 cows, among which 38 pregnant cows".14 Elsewhere, we find [..I rib.AL [LIB 4 hb.AL.Lh "[XI full-grown cows, among which 4 pregnant cows" (PRAK 2 C 73: 1-2). A text from Ur lists them separately: 24 hb.AL, 56 aib.Lh.ba (YOS 14 31 1:25-26, cf. 21-22). Another Ur text speaks of 56 aib.AL e-ri-tum next to 9 aib.AL re-qd-tum "56 pregnant full-grown cows, 9 'empty' full-grown cows".15 The Surnerogram 6b.AL.SB.ba seems to stand for "pregnant full-grown cow" in a line recurring in one text, n hb.AL.Lh.ba 1 gud.hb d 1 gud.AL ma-al-lu (YOS 13 350:10, 12, 15, 17, 19; total in 20). It is unlikely that Lh.ba here is a separate word meaning "among which", as in the following list of calves, lines 23, 27, etc. Then, we have to interpret these lines as "n full-grown cows, pregnant, 1 breeding bull, and 1 full-grown cow, mallu". Below, we will give a full transliteration of this text. Another word for "pregnant" could be the unique Sumerogram Lagan (?) in: hb Sagan (?): "18 full-grown cows, apart from 1 cow ittanapras' and 1 cow Lagan (?)" (YOS 13 97:3; cf. note 46). This will give new fuel for the discussions on Sumerian AMA.GAN.SA.~~
I
i
The later lexical tradition still was aware of the meaning of rib.AL in explaining this Sumerogram as: "[aib].AL mab = burtu : the big cow (arhu) that has given birth" (Rm 307:8, cited MSL 812 63, below). It now comes hardly as a surprise that in sales of these full-grown cows occasionally their calf is added: 1rib.~Lqadu amar.ga "one full-grown cow with its suckling bull-calf '.I7 A full-grown cow can have a "personal" name which may attest to her 'high' status (YOS 12 185:6). A yield (biltum) of milk is expected from this cow (UET 5 821). She was part of an inheritance share (van Lerberghe, Philadelphia 71:35). The Ur texts give these cows some qualifications most of which are completely obscure; they certainly belong to "the language of ox-drivers, the language of herdsmen" of which "Examenstext A" speaks.18 They were summarily discussed by Landsberger and Butz. We already cited the "pregnant" and "empty" cows, the only adjectives that are meaningful to us. In sa-ma-tum (839: 12) one could recognize the word "red". The adjectives are in one text:
Old Babylonian cattle
Stol
e-ri-turn, re-qd-tum, u'-pa-turn (823: 1-6); others: za-hi-turn (835: 13), nu-ri-turn (836:5), pi-iturn (?) (838: 13), ti-zu-pa-turn (?) (839:26).19 What is the Akkadian equivalent of 8b.AL 1 8b.mhh ? Not littum, as became clear. The other Akkadian words for "cow'' seem to be arhum and burtum. It is difficult to make the choice and we also should be aware of dialectical differences. The lexical text Rm 307 said: burtu. One text from the Diyala region speaks of both categories of cows. We first read of "barley, fodder of the burtum's when they were threshing" (;&.gal bu-ra-tim i-nu-ma i-di-Sa).20 One expects that docile full-grown cows are apt to do this work; note the remark in a letter "that ox, being young, cannot thresW.21 An unpublished letter suggests that "mothers of calves" (um-ma-at [email protected]) are the first choice for doing the threshing work: "They did not give us the mothers of the calves that are standing (here), (so) I seized the cows and (...) I threshed 10 kor of barley".22 So we see that a burtum threshes and "the mother of a calf' can only be the 8b.AL. This pleads for 8b.AL = burtum. The predicate "trustworthy" (taklat) for burtum also suggests a ripe age (AbB 2 86: 10, 3 1). The (obscure) association of this cow with "reins" also gives the impression of a trained mature anima1.23 The same Diyala text with burtum in it also speaks of "fodder of the arhum of PN when she had no work". Landsberger translated: "the young Above, we have suggested that ammahu could be the pronunciation of 8b.AL = ab.mah (6b am-ma-& in CT 8 28c: 11). But we do not consider ammahu to be an everyday word.
St01
Prices: In her third year, with "its (?) ... calf ': 7 213 shekels of silver (Scheil, RA14 [I9171 153-4) (see note 28). In her third year: 5 112 sh. (YOS 13 245); 8 sh. (YOS 13 244); 18 sh. (YOS 13 262) In her second year: 3 sh. (YOS 5 13:16-17); 3 112 sh. (YOS 13 243); 6 112 sh. (YOS 13 371) In her first year: 1 112 sh. (YOS 13 263); 4 sh. (BIN 7 209) Cow Sa tamSari : 4 112 sh. (CBS 1314, unp.) Prices for "cow" without further specification: 516 sh. (Arnaud, ARV 44); 5 (Ib burtum ; AbB 2 86:12, 19); 10 sh. (AbB 5 175:21-22;9 1745); 11 (lu burtam ... lu 11gin, AbB 9 242) 4 1
Prices: 8 shekels of silver (BIN 7 208) "Two cows (8b.lj.a) whose price is 8 sh." (YOS 12 322:8) With suckling bull-calf: 14, and 66 sh. (see note 17). The 66 sh. is exceptionally high. Two buyers purchase the animals from two owners; time of ~amsu-ditana.~5 "Cow, calf': 20? + 4 sh. (BM 80389, cited H.Farber, JESHO 21 43 and R.Harris, Ancient Sippar 252).
Cows, in their years 1-3 After the full-grown cow (6b.mrih) the fullest lists give Bb. mu.3, ab.mu.2, 6b.mu.1~~; in others we find only Ib.mu.3 and Bb.mu.1 or just ~ b . m u . 2 .We ~ ~ assume that this does not mean "cow, three / two / one year(s) old" but "cow in her third I second I first year (as cow)"; see the end of the section on Age. A cow of in these years never has calves.28 It looks possible that the Akkadian word for the cow in this age was arhum, "young COW";see above. The cow in her third year was the most valuable of the three. A letter speaks of "two cows, either in their third year, or draught-animals" as alternatives (Summa mu.3 Summa Saddidatim, AbB 11 124:10). This cow is part of a dowry (CT 48 84: 1). Cows of all ages can be part of an inheritance.29 We believe that we can distinguish the following phases in the life of a cow. In the first year of its life, it begins as "suckling she-calf' (commonly Ib.ga) and after it is weaned (KUI)) it is "cow in its first year" (6b.mu.l) (cf. RFH 30; note 9). It enters the second and third years of its life as Ib.mu.213. In its fourth year she is "full-grown", indicated by 8b.AL. The cow can now become pregnant.
Old Babylonian cattle
j
Bulls, in their years 1-3 The terminology for this category is different in the "North" (Sippar, Kish) and the "South" (Ur, Larsa). No attention has been paid to this difference by earlier authors, mainly because they did not discuss the texts from the North. In the North these animals were named "(bull-)calves" in their first I second I third year, arnar.mu.lI2I3. In the South they were named "bulls" in these years: gud.mu.11213. The "bull-calf' in its third year is often attested, amar.mu.3. The Palace sold these animals whole-sale from its large stocks; see D. Charpin, JA 270 (1982) 39-45: only thirdyear "calves". One gets the impression that the Palace had breeding stations including large herds and capitalized on them by selling the animals as soon as they were apt for practical use. We suspect that at this moment, in the third year, most of the bull-calves had been castrated and were subsequently trained as draught-animals. After this, the anjmal becomes a gud.gi25 "ox yoked to a wooden implement (gig)", the yoke. And indeed, in the lists from Ur and Larsa the bull in its third year (gud.mu.3) often is preceded by the gud.gig as the older animal, presumably the bull in its fourth year. A roster of animals gives in columns 1-111 the gud.gi9, gud.mu.3, g~d.mu.2.~~ The bull in its third year, together with the full-grown cow, is able to draw the plough: "three full-grown cows, two three-years old bulls which 'came up' to the plough of PN9'.31 Another such bull is taken over by a fattener "for the sacrifice of the nabril festival" (Szlechter, TJAUB 26). A man was ordered to buy "good" (damqum) bulls of this age (Haldar, BiOr 14 [I9531 14). It is striking that we have only a few attestations for this bull in its second year; typically, only in lists of herds. So only in the administration of breeders this age-group was relevant. In other lists, this category is ~kipped.3~ Few are the references for amar.mu.1, "bull-calf in its first yeaf.33 A list from Mari offers this sequence: "Two bull-calves in their first year (amar.mu.1) - five she-calves in their first year (6b.amar.mu.l) - five suckling calves (amar.ga)" (ARMT 23 58:7-9). Prices: Bull in its third year: 6 sh. (YOS 13 354); 8 sh. (Birot, TEBA 3 1: 11-12); 10 sh. (BM 97147; Arn~5; courtesy Els Woestenburg); 30 sh. (CT 8 lb) Bull in its second year: 6 sh. (BM 97043; Els W.); 7 sh. (BIN 7 207) Bull in its first year: 2 sh. (VAS 7 148), 2 112 sh. (YOS 13 379)
Sto1
Old Babylonian cattle
Fourth year and older A bull in its fourth year is attested in only three texts. The sign "4" is written with horizontals (limmu = 4)s4; once with verticals (gud."Z~", YOS 12 334:2). In a survey of a herd, the gud.mu.limmu is found in the position where we normally see the breeding bull and I or the trained ox.(Riftin 90 I 15, 24). A bull or ox of this age normally has entered another status, that of breeding bull (gud.aib) or of trained ox (gud.giS), and it is named after its new activities.35 Old (Jibum) animals are rarely attested. Only old cows are found; in Ur: sib.Su.gi4 "old cow" (YOS 5 21:3; 14 311: 13); following "full-grown cow" (UET 5 825: 17); or Bb.AL.Su.gi4 "old full-grown cow" (8405). Six "old cows" are delivered for fattening (6 sib Ji-ba-tum a-na nam-ri-tim, Kienast, Kisurra 98:6).
Calves We define the animal in its first year as "calf '. The Sumerian sign for "calf ', amar, looks very much like gud, "ox". The scribes were aware of this and always wrote the distinguishing Winkelhaken at its beginning.36 The Akkadian word for "calf' must be burum in view of the remark "Let the cows ([email protected]) be assembled (...) and give fodder to their calves" ([ilp-ra-am a-na bu-ri-s'i-na i-diin).37 Calves must be meant here. In later periods, Akkadian miru stands for "calf' (Deller, Baghd. Mitt. 16 [I9851 323). Most of the calves that we meet are bull-calves if we take amar in this restricted meaning. As we have seen, the Northern texts distinguish five categories of bull-calves; in descending order: a calf in its third, second and first year; weaned; suckling calf. We have one intriguing text on the herding of c a l ~ e s . ~ The 8 first part is a list of small numbers of calves followed by their owners (?). They are totalled as "30 calves, part of 200 calves, of the provisioning (?) of Sippar-Amnanum. Of the herdsman (na.gada) Ina-paleSu, son of Ilum-kinam-ide. The silver of the fodder (kissatum) of the 30 calves, of two months, InapaleSu, herder (sipa), son of Ilum-kinam-ide, has received". Date: ll.X. Most owners (?) are in the food catering business: brewers (1-4), a laundry-man (6), maltsters (7-9), ... (lo), a slaughterer of pigs (11). The herdsman tends their calves in a large herd of 200 and is refunded for feeding them. Feeding calves in months IX-XI is known from other texts (Riftin 58, 124).
Old Babylonian cattle
St01
offer gud.amar.ga for the suckling male calf and sib.amar.ga for the female. In Ur we find the same distinction in some texts.39 amar.ga "bull-calf' - hb.ga "she-calf' In other areas we find another combination of signs for "suckling she-calf ': sib.ga (amar "calf' omitted!). This is the more common Sumerogram. As Landsberger has seen, sib.ga is "suckling she-calf' and amar.ga "suckling bull-calf' (MSL 811 64 and note 4). In Larsa, Sippar and Kish they followed this terminology; thus, a list ends in 7 [email protected], 6 sib.ga (YOS 13 97: 10-11). In Ur, with the other system, we come across this distinction within one group of texts: for example in UET 5 8 19: sib.ga (5) and amar.ga (1 l)POThis system is in Ur more frequent than the other. In these areas sib.amar.ga is the collective for "suckling she- and bull-calves" (YOS 5 200:4 with Kraus, Viehhaltung 45). A similar collective was used for weaned (KUD)calves of both sexes: collective sib.amar.ga.KUD, taking together the male amar.ga.KUD and the female sib.ga.KUD (YOS 13 350:29-31, etc.). The reader should be aware of the fact that in the other system sib.amar.ga stands for "suckling she-calf '!41 The signs combination sib.ga "cow-of-milk" is logical when viewed as the first stage in the phases of the life of a cow: sib.ga "suckling she-calf' - sib.mu.1/2/3 "cow in her first I second I third year" - sib.AL "full-grown cow". The Akkadian word for "suckling she-calf' is feminine, of course (sib.ga Siniti, AbB 9 89:17). One text speaks of "One burtum-calf of milk has been entrusted" (1 amar burti Jizbi ... paqdat); a she-calf is meant and amar "calf' looks like a determinativeP2 In another instance sib b b ~ oprecedes: ~" in 2 sib bu-ra-tum 1 gud bu-ru-um (R.D. Biggs, Bibl. Mesop. 3 [I9761 no. 38: 1-2). We admit that the price 5 shekels for one sib burtum better fits a calf than a cow (AbB 2 86:12, 19). No sex distinction in suckling calves is made in Mari: "Two bull-calves in their first year (amar.mu.1) - five she-calves in their first year (6b.amar.mu.l) - five suckling calves (amar.ga)" (ARMT 23 58:7-9).43 66
Weaned calf' is in Sumerian amar.ga.KUD where KUD means "to separate" (Akkadian parasum; cf. pirsum "weaned animal").44 In Akkadian, the general word for separating animals is parcisu and it does not necessarily refer to weaning; actually, it does not in any of the OB references known to me.45 The unique 2 sib pu-ru-se-e-tum in CT 8 28c: 12 could refer to mother-cows that have been separated from their calves (the other unique am-ma-hu = sib.msih pre~edes).~6
Suckling calves gud.amar.ga "bull-calf' - sib.amar.ga "she-calf' In the sign combinations to be studied now, Sumerian ga means "milk" and the inclusion of ga indicates that the animal is suckling. The lists from Kisurra differentiate according to sex and
Prices: Suckling bull-calf, weaned (??): 4 sh. (Waterman, BDHP 33) No age given: 3 sh. (AbB 10 117:5')
Old Babylonian cattle
Breeding It has been suggested already that the Palace had breeding stations and sold the bulls when they were in their third year. This is clear from the Late Old Babylonian texts from Sippar studied by D. Charpin, and the Larsa lists treated by F.R. Kraus show the same. We could add a group of lists from Ur; see above, under "Bulls, in their years 1-3". The Sippar texts are complemented by four Late Old Babylonian lists of cattle which give a survey of the animals of the breeding stations in Kish and Sippar.47 In Larsa and Ur texts similar lists have emerged. What unites them is the completeness of the surveys: ideally, all ages are represented. The specialized business of breeding cattle was an enterprise also undertaken by princesses. In the bride-price (terhatum) for a princess, the daughter of Yarim-Lim, at the end of the list, we find "3 breeding bulls, 100 cows, [x]400 ~heep".~8 She certainly was supposed to make profits with the bulls and cows. Similarly, princess Iltani had herds of large and small cattle in the Sippar area. We learn from Riftin 90 that the State supervised herds of large and small cattle; the oxen were put at the disposal of land stewards (ensi). This text is the stocktaking (6u.sum.ma) of the herds over two years, "ceased" by Ali-talimi (Sa A. i-zi-bu, 11.8, rev. 12). The governor Mar-er~etimis involved (I1 14); a man with this name was governor (6agina) of Bad-tibira at about the same time. At the end we read the names of [Sin-m]uStil, overseer of merchants at Ur and Sep-sin, merchant of Larsa.49 We have numerous contracts entrusting (paqadum) sheep and goats to herdsmen. Far less frequent are those on cattle. We give them and add between square brackets the numbers of full-grown cows: YOS 8 164 [14]; 12 103 [12]; 13 97 [18]). The totals differ widely (50, 56, 28). One text shows that extra animals, of outsiders, could be added to the herd.50 A difficult but interesting contract is YOS 13 348; we give a transliteration:
(5)
(10)
(15)
i- 'su '-x x 99 hb.mu.2 93 hb.mu.2 u-bar-rum 80 $b.mu.2 tu-un-nu-hu 106 amar.mu.2 dingir-ha-bil 'TUR1 103amar.mu.2 qd-du 6 s[ipad.mel] (?) dumu.me9 [I]u.SA.GUD.~~S a-bi-e- 'ra '-ah 5 hb.'AL1 (?) ki ib-na-tum 277 hb.mu.2 209 amar.mu.2 6u.nigin 8 Zu-Si 6 hb.gud Sa &gal fa na.gada di6kur-na-~i-ir a-nu ib-nikdmardukegir [email protected] dumu dmarduk-e-mu-qi ip-qi-du-ma i-na itu kimdinanna uq.28.kam Pib-ni-dmarduk a-na pi-qi-it-tim ub-lam (one line completely destroyed)
Old Rahylonii~ncrrttlc
St01
(19)
[x XI X X ha [ ..... ] x ab.gud Sa ur.m@ i-d[u-uk-ku] x [x] x x x Li-ul ub-ba-lam (!)-ma a-na x x x x x x (10 witnesses; date 10(+).VI Ammi-ditana 3 1)
The text begins with a list of two-years old female and male calves. The names that follow are also found in the cattle survey YOS 13 350: 1-9 (edition below).51 We suppose that they are low-ranking cowboys, just as "the h[erdsmen] (?), sons of ox-drivers, of Abi-erah (5-6). The animals total "486 head of cattle of the Palace" (10) and are "entrusted" (paqadurn) by the chief herdsman Adad-nasir to Ibni-Marduk, "(the man) behind his cattle". This is the herdsman who presented the animals at the "entrusting" / "mustering" (piqittum) on 28.VI. The text itself is dated to 10(+).VI; is this the same day? The lost and fragmentary lines must have given some contractual conditions. We can only read "The cattle that a lion kills, .... If he does not bring (them), he will ... to ...".52 The first witness is ISme-Sin, "the scribe of the three heads of temples (sanga) in Kish and a number of "priests" personnel" (dub.sar.er~n),~~ follow. For breeding, one needs a breeding bull. The Sumerogram is gud.hb. The lists often mention this animal; in Ur after the section of cows. In this position, the breeding bull and the "yoked bull" (gud.giS) head the section of the bulls. They indeed are bulls in their fourth year (or older). In the lists only a few breeding bulls are mentioned after many cows. One in Meek, AJSL 33 (1916-17) 239 RFH 30:13. One "large breeding bull" (gud.hb.ga1) is followed by 38 "large cows" (hb.ga1) in ARMT 23 505: 1. In the bride-price of the princess we had 3 of such bulls on 100 cows. More data: 76 cows, 2 bulls (YOS 5 150:26). The list that gives us some information on breeding is YOS 13 350, from Kish. It has close connectjons with 348, studied above. We will give a transliteration. 90 ab.mu.3 86 hb.mu.3 71 hb.mu.3
'i '-su-x [XI u-bar-r[um] tu-un-nu-hu 248 hb.mu.3
(5)
42 amar.mu.2
dingir-ha-biz TUR 105 amar.mu.2 a-bi-e-ra-a[h] 147(!) (text: 87) amar.mu.2 394 [email protected] li-tum ib-na-tum
(10)
27 hb.AL x x [..I gi-mil- [...I
Old Babylonian cattle
Stol
(15)
38 6b.AL.Lh.ba 1gu[d.rib (?)I 1gu[d] x x d 16b.KUD.KUD Sa i-na @ur.sag-kalam.maki mu-ti-AN 55 6b.AL.Lh.ba 1gud.6b d 1gud.AL ma.al.1~ dingirb-biz GAL 57 6b.AL.Sh.ba 1gu[d].hb d 1gud.AL ma.al.1~ dumu-ug30.kam GAL 51 6b.AL.Sh.ba 1gud.6b id-da-tum dumu e-Se-i-dingir
(20)
218 6b.AL.Lh.ba 4 gud.6b [O] 5 gud.AL ma.al.1~ 90 x x x [x] ga.KUD L&ba50 [..I hb..gaKUD
lo(+) [...a]mar.ga.KUD L d (??) x x [x x] x dingir
(25) 86 [
...g]a.KUD &b[a ..I '5l 6b.ga.KUD
'lo1 [..I
x amar.ga.KUD
mu-ti-AN
86 6b.amar.ga.KUD
(30)
lh.ba 54 6b.ga.KUD 32 amarga.KUD dingir-b-biz GAL 84 6b.amar.ga.KUD Lh.ba 49 6b.ga.KUD 35 amar.ga.KUD dumu-u4.20.kam GAL 44 6b.amar.ga.KUD Lkba 23 6b.ga.KUD 21 amar.ga.KUD id-da-tum dumu e-[Se-il-dingir 390 hb.amar.g[a].KUD Lh.ba 238 &b.[g]a.KUD 212 amar (!).ga.KUD li-turn dingir-3u-ib-ni-Su (date: 14+. [VIIII (?) Arnrni-ditana 33)
St01
Old Babylonian cattle
The text starts with groups of cows in their third year and bull-calves in their second year, under the authority of Ibnatum (litum Ibnatum; 1-9). We suspect that they were given out for pasturing by the Palace, as the parallel list, studied above, shows (YOS 13 348:l-10). There, we found the same persons; they are cowboys taking care of the young animals. Next comes as second section a survey of the full-grown cows, pregnant (6b.AL.Lh.ba) (10-20). To each group of cows (comprising each 27, 38,55, 57, 51 cows) "one breeding bull" (1 gud.6b) is added. Clearly, these are herds. As extra addition we find once "one ... cow (6b.KUD. K U D ? )which ~ ~ is in Hursag-kalarnrna" (13), and two times "1 gud.AL ma.al.1~" (15, 17). The summation seems to take them together as "5 gud.AL ma.al.1~" (21). We do not understand this unique qualification ma.al.lu.55 The third section (22-44) gives a long list of weaned bullcalves and she-calves (amar.ga.KUD and hb.ga.KUD), under the authority of IlSu-ibniSu. This list deserves further study. Size of herds: 50 and 57 full-grown cows and [x] bulls in two subsequent years (Goetze, JCS 2 105 no. 9:9); compare also the text YOS 13 350, just studied. In Mari one bull and 38 cows (ARMT 23 505). In the MB period 30-40 cows and one bull (K.Kessler, ZA 82 [I9921 95); note the 34 full-grown cows in MA; K.Deller, Baghd. Mitt. 16 (1985) 318. We have seen that the Palace sells the bulls in their third year. That is why we find no older bulls in the breeding stations - with the notable exception of the few breeding bulls which they had to keep. Thus, we can explain the problem that Goetze had in interpreting the growth of the herd of princess Iltani. After having shown how the scribe arrived at 33 bullcalves in their third year, he wrote: "It is surprising to note that the amar.mu.3 do not seem to advance in age; they are not transferred to the 'cows' either" (JCS 2 82). The reason is, that they were removed from the herd and the text itself seems to indicate how: column (g) has the heading amar ngmettum "calves - impost". This impost was levied by the State who sold the bull-calves in their third year. This column has only one figure, at its beginning: 30. Did they deliver 30 out of the 33 bulls and did they retain three bulls for breeding or other purposes ? As to the "other purposes", many lists also mention the "yoked bull" (gud.giL), certainly a castrated bull (= ox), at least four years old. On their usefulness in the breeding station one can only speculate. One text speaks of "3 full-grown cows, 2 bulls in their third year / one cow in her third year, four bulls in their third year, which came up to the plough of PN" (Sa a-na gud.apin PN i-lu-zi), VAS 18 93:3, 7 (23.V). "To come up to the plough" (elam) means that the animals arrived from the herd in order to be trained for a ploughing-team. This situation fits their being in their third year. The text is dated to the fifth month. Training oxen seems to be meant by the entry "when we instructed the plough", in an unpublished text.56 A Sargonic text gives various destinations for animals leaving the herd: for consumption, branding, and the plough.57 "Herdsman of cattle" is sipad 6b.gud.M.a. Herdsmen sometimes are named "herdsman of cows" - not oxen.58 The majority of animals in the herds may have been female; this could be the reason why the Edict of Ammi-~aduqa912 only speaks of the skins of dead cows, ewes, she-goats - all female.
St01 Stol
Old Babylonian cattle
Herding The normal word is re'iim "to pasture". Pastureland is ritum. "Perhaps the 6 KIN ["Weidesucher"] Attl will tell you 'There is pastureland (ritum) of the oxen' - (well), the entire Euphrates is pasture land !" (AbB 4 150:33-38). This land was not open to everybody59 and one had to pay "silver for pasture land" (kasap ritim).60 Month I1 is a good time for herding because there is enough grass in that season; we have a text on hiring a man for herding a trained ox in this month.61 "To put animals out to pasture" is nadiim lit. "to throw" in Akkadian (CAD N/1 79). The word indicates that animals are let into the fields in order to feed on the young shoots. Examples for "putting out" cattle are letters referring to hired plough-teams (initum) and calves.62 I do not know what gararum 5, said of cattle, exactly means; "to let roam freelyW?63
B. WORK OTHER THAN PLOUGHING In cattle not so much their meat or milk were appreciated but rather their strength. This strength was used in the heavy work done by them. Oxen "stand (ready)" to do the work (izuzzum).64 A number of unpublished dockets dated to month V documents their presence: one, two, or five oxen, followed by the name (?) Sb-bu-3um.65 One of the fears was that the oxen should remain idle (regum). The texts distinguish between ploughing oxen (2ri3utum) (a distinct category in Mari and Chagar Bazar) (ARM 3 33:9; 22 320:8; 23 505:6; Chagar Bazar 26:2). threshing oxen (di1i3zitum)(AbB 7 3:28; 11 86:24; Ellis, JCS 24 61 no. 53 rev. 5), once mudiS3iitum (YOS 12 421:2,9); Sumerian gud.Se.6s.sa (MSL 1 55 IV i 69). the draught-animal (3addidum) (see CAD).
Castration The animals involved must be full-grown and can be both male and female. Only some animals drawing the plough have to be male; this is the exclusively male "rear ox". Undoubtedly, all males had been castrated. Maekawa wrote that castration took place "in spring just before or after they became two years old"; "young male animals must be castrated before their first sexual excitement" (Zinbun 15 [I9791 109). A modern manual on cattle in Africa: "Castration should take place between the age of ten months and two years, the best time being between 18 months and two years. Training may start at 2 112 years if feeding is sufficient. However, as the best age for training - in sub-Saharan Africa - the time between 3 112 and 4 years is recommended. Since it is the general experience that animals learn quicker the younger they are, making them ready for early training depends much on good feeding9'.66 If all this is true, the castrated bulls were kept in the herd for at least one
Old Babylonian cattle
more year until they were strong enough. Dr. Uerpmann at our meeting: "castrated before they enter their third year; they mature further and training follows later".67 We should be aware of the fact that these animals had been trained and fetched a relatively high price. Here, we will first make some remarks on the threshing and drawing oxen; the ploughing oxen deserve a separate section.
Threshing, wagons Whenever the texts are a bit more explicit, we see that the cows or oxen have been hired (id: "hire") when threshing. An ox that is too young (jebrum) is not able to thresh.68 The point is not its weight - an ass and a goat can thresh, as CH $269 f. show - but that the young animal is unruly. A herdsman (sipad) is asked to give the animals to this purpose (AbB 7 170). Cows of the Palace or oxen of the common citizens were required to do the threshing (ARM 14 48). "Drawing" animals are drawing the yoke (nirum). A cow was hired "in order to draw the yoke" during two months (YOS 13 17; dated 5.VII), a rear-ox for the same job (ana niri Zadiidim) during three months (YOS 13 370; date 3.111). "Ox of the yoke" was a fixed expression in the Diyala region (UCP 10 27:9) and became common in the MB period.69 It is highly probable that Ur III and OB (South) gud.giS "ox of the wood" is an abbreviation of "Ox of the wooden y0ke".~0 Now, we have to ask ourselves: what yoke is meant, that of the plough or of the wagon ? The Farmer's Almanac shows that a plough has a yoke (15, with bil. version 24, in ed. M. Civil, p. 42) but there is ample evidence that wagons have yokes, too. We assume that an "ox of the yoke" can draw both. And that "ox of the yoke / of the wood" is one expression for "trained ox". We will now make some remarks on the wagons. The Farmer's Almanac says: "Your wagons should be in working order; feed (well) the wagon's oxen" (92 f., M.Civi1). There, they are used after the harvest, probably for transporting the sheaves to the threshing-floor. The series Ana ittiiu also mentions the wagon in the context of the harvest and a letter from QaQunin says, "They harvested the field of the Palace and the citizens 'brought up' (the crop) and threshed it; moreover, they transported (zabilum) the barley in cleaned state (zakiitum) in the wagons of the citizens from GN1 and GN2 to Qa?funin".71 The wagon is a neglected aspect in the study of Babylonian agriculture. The word "wagon" (giS.mar.gid.da, ereqqum) is frequently attested in agricultural texts from the Diyala region. In one group of texts they are issued by an authority, including the Palace, in the harvest season.72 Letters from the same area speak of these wagons in the same way texts from other places speak of "the plough", i.e., the plough together with animals and personnel.73 The personnel is what CH names the murteddiim "driver(s)" ($271, hiring). A text from Sippar tells us that each wagon requires two men (D.O. Edzard, Tell ed-D2r [I9701 nos. 221-2). Two other texts show that a man "received" a wagon ana sadiirim on 10.11 and "returned" it on 14.lII - the harvest months.74 Oxen were used to draw these wagons at harvest time (AbB 7 3:17, 21; S.Dalley, OBTR 137). The wagon is associated with the yoke: a lion can kill "the ox while it was under the yoke of a wagon" (gud gii.Sudun.mar gid.da.bi).75 In the northerly sites Chagar Bazar and Shemshara we see the oxen associated with the majjaltum a word translated as "(a kind of wagon or sledge)" by the CAD. New texts
Old Babylonian cattle
St01
from Mari clearly show that a "riding" wagon is meant. Note this passage in a letter: "I want them [the babies] to ride on a majjaltum and their nurses should ride (rakibum) with them. Five oxen shall draw (Sadadum) that one majjalt~rn".~6 This wagon, used on the fields, must have functioned like the ereqqum in Babylonia. Another trained ox or cow is associated with the tamSarum. We do not know what is meant by it. "Whip" is von Soden's translation; in one text the word is perhaps preceded by the determinative "cord", suggesting some line.77 A final remark on trained oxen. To indicate this category, various expressions existed. Some have been studied, others will be studied below, under "ox-driver". The expressions are : gud.gig "ox of the wood" (Ur, Larsa; the South) gud Sa nirim "ox of the yoke" (Diyala region; MB) gud SP.gud "ox-driver's ox" (Sippar; the North) gud.da.ri.a "driven ox" (Sumerian law) Prices of trained oxen: Draught-ox (gud iaddidum): 12 sh. (YOS 13 349) ...-ox (gud putarum): 6 112 sh. (Riftin 26) ...-ox (gud tamiarim): 8 112 sh. (YOS 13 380); 10 sh. (BM 80408). A ..-cow:4 112 sh. (CBS 1314, unpub.; grim 1sib ta-am-ria-ri7) Rear plough-ox (gud.8.hr.ra): 10 sh. (YOS 13 259; BM 97131 [courtesy K.R. Veenhofl;gud ar-ki-a, AbB 12 519); 13 112 sh. (VAS 16 206); 14 sh. (Waterman, BDHP no. 62; gud.hr.ra) "Ox-driver's ox" (gud.gB.gud): ca. 12 sh. (CT 8 30~9-11) gud.gig: 5 sh. (VAS 18 100:34) Fattened ox ( g ~ d . S ~8) 2 :13 sh. (OBTI 94:l-2); (three)fattened oxen ([email protected])"of 15 sh." [each ?] (AbB 11 4 rev.) No specifications (gud): 4 112 sh. (Birot, TEBA 34:16); 6 sh. (BAP 2); 7 112 sh. (CT 45 215, 15 sh. for two oxen); 7 112 sh. (CT 4 27b); 8 112 sh. (YOS 12 71); 8 114 sh. (TCL 1 233; weight-stone of Sama~); 15 sh. (AbB 9 64); 9 112 sh. (Szlechter, TJAUB [I9631 48 UMM H 26:4); 12 sh. (Charpin, RA 74 [1980] 119 no. 68:3); 13 sh. (TLB I 193:l-2);21 sh., price asked (AbB 8 78:33). Mari: 6 or 8 112 sh. (Kupper, Studia P. Naster I1 [I9821 117); 9 sh. (ARM 27 111:4). Rental of oxen: see M.Sto1, RIA 813-4 (1994) 169 f., $ 2 "Tiermiete". ~
C. PLOUGHING
The concept "plough" "Plough" is apin in Sumerian, epinnum in Akkadian. It often is preceded by the determinative gig "wooden". Now, it has become clear from the Mari texts that epinnum also stands for the plough with the oxen and personnel, a complete "team", a meaning not registered in the dictionaries of Akkadian.78 Similarly, in the Old Babylonian period outside Mari, the Sumerogram gud.apin can stand for such a team, but some instances require its literal meaning, "ox of the plough". So the Sumerogram gud.apin means either "ox yoked to a
Old Babylonian cattle
plough" (alap epinnim ?), or "plough-team consisting of a number of oxen with attendants" (epinnum). We will first sort this out. Animals: The following passage shows that a gud.apin can be the animal: "Plough-ox, plough, and what is needed for the seed-ploughing" (gudapin gig.apin u biSehti erCSim, AbB 6 129:17). Oxen are meant in "29 [email protected]'for the land-stewards (Riftin 90 I 9, I1 9; contrast the team gud-apin 6.ta in line 1). In Mari, [email protected] stands for the animals (ARM 4 7523). Sometimes grammar helps: "oxen" are masculine plural in Akkadian, whereas "ploughs" is feminine plural (epinnitum). "The plough-oxen of my lord are well" is masculine and oxen are meant ([email protected] ...Salmu [masc.], AbB 12 9:4). The plural of epinnum, "team", however, is feminine and [email protected] i-re-qd [fern.] should mean "my plough-teams are idle" (AbB 11 66:20).79 Contrast with this: 4 [email protected] re-qli [masc.] "my four plough-oxen are idle" (AbB 3 29:7). Plough-oxen are meant in lists of family property or gifts: "644 [email protected] 2 Su-Si 6b.M.a" (AbB 5 244: 11); "3 [email protected](CT 45 27: 1); "6 [email protected] apin 20 [email protected](Scheil, SFS 10:24); [email protected] apin 1Su-Si 5e.gur.ta.hm (AbB 10 6: 10). Team: When a letter says that a gud.apin is "scattered" because the oxen have been taken away, the team is meant (gud.apin ... PN uptarrir, AbB 11 116). "They will take over half of your gud.apin9' implies a plurality (AbB 7 415). When texts speak of the "task to be performed" (iS.gPr) of one or more gud.apin's, a team is meant (&.gar 2 [email protected], AbB 12 16:7; note [email protected] in 30, 34).g0 In one text one can find both gud.apin and gud, in the singular or plural. Perhaps "team(s)" and "ox(en)" are meant, respectively, but there often is no certainty.81. g2 In most cases, gud.apin stands for a team. We add that this team is State owned, as far as we see.
Excursus - The "plough of the task to be performed" gis.apin Cg.gPr, seems to have been a special instrument. The attestations are few and point to a plough of more animals: (a) 1 gis.apin aiS.ghr, CT 48 43: 1. Contrast "2 initums" in line 6. (b) Sum-ma 1 gis.apin.Cl.gk [so !] Sum-ma gig ha-ar-bi [email protected] (...) Sibilam, "Have brought to me either one 6S.gPr -plough, or a deep-going plough of oxen" (K. van Lerberghe, MHET I no. 71:16). The barbum-plough is the deep-going plough and seems to be as useful as our plough. Now, deep-going ploughs require more animals and this could be true for our plough, too. (c) 2 gig.zu.apin.Cg.giir, AbB 10 30 rev. 4,7. The "teeth" (zu) are especially typical of the deep-going plough. We assume that this was a plough with preset measures, especially of the amount of seed.
Old Babylonian cattle
One tends to associate this team (gud.apin) primarily with the activity of seedploughing (APIN = u ru4 = eregum). However, the gud.apin does not only do the seedploughing but also the deep-ploughing (majjiiri rnaha~um,AbB 10 15:15-18, cf. no. 10) and the pariikum (Riftin 53:ll-12). There is plenty evidence that deep-ploughing was done by oxen.83 Three teams needed parts of a deep-going plough, a p i n . ~ u ~ . I(AbB U ~ 5 155). In the disputation between the Hoe and the Plough, the hoe says that the plough works only during four months per year (108). When working, "your oxen are six, your men are four, you yourself being the eleventh" (91).84 The numbers apply to a plough with six oxen and four men. Let us see how this worked in practice. Four OB texts detailing the expenses in working large tracts of fields refer to the animals and people involved: CT 8 30a:l-7 - "6 'oxen' of 1 gu.la among which 2 cows; Awil-Adad, plough-man (engar), slave of the house; 2 hired workmen - ox-drivers, free-born men (erCn hun.gh &gud dumu.mes' 16): (total) 3, plough-man - workmen - ox-driver; 1 female ox-driver, slavegirl of the house" (2.V Amrni-~aduqa18). Two remarks: (a) Instead of the "1 gu.laWwe expect 1 gig.apin. It is normal that six animals are going with one plough. The 'oxen' include cows. At the beginning of this article, we have seen more occasions where [email protected] means "cattle" in general. (b) Female ox-drivers always are slaves. She deserves a separate entry here because she does not require expenses like hire. - Four people accompany one plough. This, and the six oxen, fits "The Hoe and the Plough". Riftin 6O:l-5 - "8 bur85 of irrigated field (mikrum); 6 oxen; 4 ox-drivers (ih.gud.me8); 46, 2 kor [=I3920 liters] of barley (as) seed and fodder (&.gal) of the oxen - task of IpquIdiglat" (19.VI Rim-Sin 3 1). TIM 5 50:ll-15 - "14 oxen; 2 plough-men; 6 ox-drivers; 14 sickles; 2 x x [hardly gud.apin] SAL du-di-ib-bu (?)" (-.IV Rim-Sin 47). Two teams seem to be working here. The fields to be cultivated are 14 bur "irrigated" fields (mikrum, 5) and the furrows are to be 'laid' in months 6 and 7 (itu.6.kam Li itu.7 (!).kam ab.sCn gai.gh, rev. 9- 10). Kienast, Kisurra no. 97 - a roster: "Field - 6 bhr [= 108 iku]; Oxen - 7; Ox-drivers - 6; Its barley - 120 [or: 180 ?] kor; Name - PN". (No date). We will see now that 108 iku is one of the standards for an area worked by one team. The amount of land worked by one "plough" (iZkirum) varies86: 50 iku in Mari, ARMT 13 39 rev. 20. 70 iku in Mari, AEM 111 217 no. 76:9, but the king said: 100. 80 iku in Mari, ARMT 13 375, with M-Birot, Syria 41 (1964) 45 f. "We will tear out the sesame" follows, an independent activity. Cf. 160 iku for more than one plough, 38: 11.
St01
Old Babylonian cattle
90 iku in the Sargonic Mesag archive, BIN 8 203. Note AbB 3 29: "My four [email protected] are idle; look for 90 iku of field". This suggests a team-of-four. 108 iku in a group of OB texts, YOS 14 300-305. 135 iku in OB Larsa, worked by a team of seven oxen, BSA 5 (1990) 197. 150 iku in Mari, when combining ARM 27 2 and 36-37 with 100; Birot, ibidem, p. 11 $ 7 (450 iku for 3 @Lapin,600 for 4,900 for 6). 166,66 iku in Mari, calculation based on 6 gikapin for 1000 iku in ARM 27:38-40 (cf. 36: 11-13). Seven gikapin would mean: 142,85 iku . 180 iku in the Farmer's Almanac 26 and in YOS 5 164. The figures given above serve as standards set by the authorities. The variation is due to many unknown factors, like the work, the number of oxen, the distance between the furrows. Thus, we can see that 45 or 50 iku is the norm for opening up new land (petiim).g7
Number of animals yoked to a plough The lexical tradition knows of ploughs drawn by 2, 3, 4, 6, and 8 oxen. The Sumerian expressions are giS.apin.gud.2/3/4/6/8.lh "plough, yoked to 213141618 oxen". They are not attested in daily life texts. The Akkadian equivalents are: s'u-nu-'-tu, la-lu-ul-tu, e-pi-in er-bi-it, e-piin Si-Sa-at, su-mu-un-tu (MSL 6 [I9581 17 Hh 129-133). Daily life texts only have erbettum and, perhaps, sumuntum. Not only oxen, also cows can draw the plough; cf. VAS 18 93, discussed at the end of "Breeding". I do not know what lines 59-60 in The Song of The Plowing Oxen mean: "0 wooden KU, your fathers are four, o wooden KU, your mothers are eight". Oxen in teams of four, cows in teams of eight ?88 From the MB texts we can deduce We believe that ploughs with that a harbu plough was drawn by 4 to 8 oxen (GUD.SA.GUD).~~ more than 2 oxen were primarily used on the large estates of "public" institutions. This is the Old Babylonian evidence: Two oxen. Perhaps meant by gud.apin ta-US-ni-imin AUAM 73.3233 = HTS 43 (see note 82). The problem is that while this must have been the most common team we never see an explicit remark on it. The one rear-ox and the one ox in his fourth year used "for cultivating sesame" of YOS 12 334 may have been a team of tw0.~0A rear-ox and a cow (?): YOS 12 381. A share given to a brother consists of: 1 ox, 1 cow, 2 wagons, 1 plough, 1 deepgoing plough, 1 harrow (CT 6 28a: 10-12). When seed-ploughing, 2 animals is the maximum: more is unnecessary and cumbersome and makes it impossible to draw neat parallel furrows.91 Four oxen. Best known are teams with four oxen. The texts give erbettum as a special word for this "Viergespann" and a cow with calf is once qualified as "part of a team-of-four" (erbetitum, YOS 13 355:l with C.Wilcke, WdO 8 [I9761 273 note 36). The plural of this word is erbenbum only attested in texts in the Diyala region.92 Just the four animals (plough la & excluded) are meant in these lines in a letter: 1 gig.apin.ta.h er-be'-tam qadum t ~ - ~ k a -l-tim
St01
Old Babylonian cattle
gh.gud.meg lilqfinim "Each one of the ploughs [of the three gi8.apinfsmentioned before], a teamof-four, together with the ...s and the ox-drivers they should take to me9'.93 A. Salonen names this team of four the "Normalgespann", probably because "four" fits the four Sumerian qualifications of the different oxen going in front of a plough (Agric. 384) (to be studied below). This is confirmed by a letter where the writer says: "I put up three bovines (3 6b.gud.w.a us-zi-iz) but I do not have a rear ox" which suggests a team of four (AbB 2 159:14). Pre-Sargonic texts refer to four asses yoked to a plough; see G.J. Selz, "Ein Notiz zum Tiergespann aus vier Arbeitstieren", ArOr 61 (1993) 11-2. A Sargonic text speaks of "three oxen, one ...-ass, belonging to one plough" (Su 1gikapin, MAD 1 47.11.5-7); another of two teams: "four oxen, four she-asses, belonging to 2 ploughs" (Su 2 gig.[apin], MAD 1 136:7, cf. 3). The Neo-Babylonian Edict of Belshazzar speaks of "one plough, four oxen, two,cows, four plough-men", and there is much more evidence for four oxen.s4 Van Driel wrote in BSA 5 (1990) that (a) the cows are included in the Edict because they have to "keep up numbers" by calving (p. 225 sub 4); (b) that not necessarily four oxen were yoked to the plough: rather two alternating teams of two oxen (p. 224, 227). As to (b), he sees four oxen as exceptional; teams with two are normal in private archives. Indeed, the four oxen are not necessarily one team yoked to one plough. Van Driel does not see what the profit of four oxen could be and he says that heavy work like deep-ploughing only requires a heavier plough, not extra animals. The four oxen have different qualifications that will be studied below ("rear-ox" etc.). Six oxen: Hoe and Plough, 91, quoted above. Another interesting literary reference may be hidden in Enkidu's report about the deceased in the Nether World? "Did you see him who had four sons?"'I saw". How does he fare?' "Like a man who yokes four asses his heart rejoices". "Did you see him who had six sons?"" saw". How does he fare?' "Like a man who yokes the plough his heart rejoices". We surmise that the ploughman yokes six oxen, clearly a higher state of bliss than working with four asses. 6 gud.w.a apin, perhaps "six oxen, a plough", Scheil, SFS no. 10:24 (a gift; cows and sheep follow). "Ploughs, six each 1 seven [oxen] each" is attested in one text.96 Cf. "12 oxen, one cow, of 2 ploughs: plough-team of PN" (12 gud.w.a 1 rib Sa 2 pikapin apin PN, ARMT 24 135-6). In a Sargonic text: 6 gud Su 1gii.apin PN, MAD 1 47.11.2-4,111.2-9. We read in a letter "As to the report on the five oxen (gud.M.a), about which you have sent me to PN, I have investigated the matter and one ox is not there. I spoke to him as follows: 'The oxen that you and the gentleman have promised, show (them to me) (...)' " (ABIM 9). Had a team of 6 oxen been promised and is one lacking ? Seven oxen: "Seven oxen of one plough (7 gud.w.a Ja 1 gud.apin) have been entrusted to PNl and PN2", PBS 812 189; "he has received seven yoked bulls ;or his plough", 7 gud.giS a-na giS..apin-Su ma-hi-ir, Stol, BSA 5 (1990) 197:16 f. "Ploughs, six each I seven [oxen] each" (see above, under "Six oxen"). 7 [email protected] apin PN, ARMT 24 13:1-2. The remarkable 3 112 [email protected] Sh.gud "three-and-half ox-driver's oxen" suggests half a team (BE 611 68:4). Two teams of seven: 14 gud 'a-na e-pi-ni-iml a-nu Sa-ab-re-e, Kienast, Kisurra 98:4.
Sto1
Old Babylonian cattle
Eight oxen: 8 [email protected] apin PN, ARMT 24 13:3-4 and 7-8; possibly sumuntum in a text from the Diyala region.97 Even numbers are expected and whenever we find an odd number, an extra ox has been added to the pairs. Note that in the group of texts published by M. Birot and M. deJong Ellis we find these numbers of oxen: 3,6 (often), 14 [email protected] I gud.gi5 (and: 1or 5 gud.apin).98
I
Functional qualifications of the oxen The rear ox, often Sumerian gud.iir.ra or gud.egir.99Less often gud.i.iir.ra, Akkadian alpum arkium, alpum Sa warka. 100 He has to be the strongest animal and the laws prescribe for him a higher amount of barley as hire, 4 kor per year, whereas the other animals cost 3 kor.101 The only hiring agreement that has been published gives as hire (idii) 5 kor of barley plus 115 kor of sesame per year (LB 771; now TLB 1 213; Ammiditana 4). We know that this ox "during two years cultivated (epiSum) barley and sesame for PN the Iu.gis'.Su.grin" (Stol, JCS 34 (1982) 171 no. 30). He was also used "to draw the yoke [of a wagon]" (YOS 13 370). Another contract remains unpublished (CBS 1318). One text speaks of 1 shekel per month which would mean 12 shekels = 12 kor per year. Clearly, this is a temporary high price that one could demand in the high season (VAS 7 92 = VAB 5 148). Above, we have given the prices of these animals when they are bought (under "Prices of trained oxen"). A letter tells us that a rear-ox whose foot was "broken" was moved to the "middle" position.102 The combination of signs gud.rib.iir.ra "ox-cow-of-the-rear" in a Sumerian law (M.Civi1; note 101) shows that, theoretically, a cow could be in the rear position. A rear ox was an important asset. We find him in inheritance shares.103 The animals in the middle and the front are alpum Sa qabla (of the middle) and $a mahra (of the front). In the terminology of the laws they were taken together as gud.rib.murub.sag (CH) or gud.6b.sag.murub (M.Civil), "ox-cow-of-the-middle-front". This is now confirmed by a text from Nippur where in a list of inherited goods one reads: "1 rear-ox, his name is PN; [I] ox-cow-of-the-front-middle, his name is PN" (YOS 14 321.1.14-17; cf. 11.19-20). One specific animal is meant and the Sumerogram shows that neither gender nor middle I front position matter. Another Sumerogram is perhaps g u d . s a g . d a ~ ~ . ~ ~ ~ . l o 4 . The extra ox. The Farmer's Almanac recommends to take an extra ox (gud diri.ga) with the plough oxen (23). The lexical texts indeed speak of an extra ox. We suspect that Sumerian gud.da.ri.a identified by those texts as extra ox, gimlu, is an error for the similarsounding gud.diri.ga (see discussion Salonen, Agric. 385-6).1°5 The only context reference for gimlum is extremely obscure.106
Personnel A famous seal impression from Nippur shows three men serving on the plough: a man behind the plough, steering its handles; a man handling the seed-funnel; a man behind the two oxen with a stick in his hand.107 A lexical text listing professions gives these occupations: nu.banda gud, SEgud, ih.gud.zi.ga, B.gud, gud.dbus, gud.da.ri.a, gud.du.a.na.an.DU, 1u.numun.gar. log We will not comment on these lines but go to the Old Babylonian texts.
Slol
Old Babylonian cattle
The plough-team is headed by the sag.apin, so we learn from the Ur 111 texts. The OB texts hardly ever speak of this or a similar function.109 An isolated text lists: "ox-drivers, chiefs, father-of-a-plough"; herdsmen follow: (14) Sh.gud.me Sabra.me (15) a-bi e-pi-ni-im (Kraus, JEOL 29 [1985-861 36 Ni. 2387: 14-15). The "chief' (Sabra) can be found back in the entry: "fourteen oxen for the plough, for the chiefs" (14 gud 'a-na e-pi-ni-iml a-na $a-ab-re-e, Kienast, Kisurra 98:4).
ikkiirum "plough-man" We have seen that the personnel comprised of "plough-men" and "ox-drivers". The Sumerian sign for plough-man (engar) is identical with that for plough (apin), although the Hammurabi Code makes an (artificial ?) distinction between the two signs (8 257 and 259; collated by Frans Versteeg). In Mari it is said that these men are not available for the ploughs that had been constructed (ARM 1 44). Are the plough-men (ikkirum) only needed when the seedploughing is done? There must be more to the job of the ikkrum. Gimillum rents himself out to be plough-man (ana ikkarutim) and part of his responsibility is: "He is accountable for the eye, horn, hoof and tail [of the oxen]; if he allows the oxen to be idle, he has to pay the silver for an initum " (PBS 812 196). His responsibilities include those of the ox-driver: the oxen's health. He earns 1 shekel of silver per month plus rations: 2 liters of bread and 3 liters of beer. This is much more than the 8 kor (= 8 shekels) per year prescribed by the Hammurabi Code $257. Another ikkzrutum contract is BM 97543 (5.111) (courtesy Els Woestenburg): one year, at 8 shekels of silver, plus 2 and 2 liters as daily food and drink. The clausula on eyes, etc., f01lows.l~~ These plough-men are always mentioned before the ox-drivers and indeed, the Hammurabi Code ranks them higher in allowing them a hire of 8 kor of barley per year; ox-drivers receive 6 kor ($257-8). A hidden hire of a plough-man, styled as a loan of 8 kor barley and 8+ minas of wool, is YOS 8 105. The 8 kor fits the standard wages and an extra clause says that he "will bear any fault for one plough" (bitit 1gikapin ittanazii). Barley and wool are also given to a man who is to work with two initums (Limet, Mdlanges J.-R. Kupper [I9901 40 no. 4). A plough-man can hire seven people (VAS 9 103:lO) Perhaps one of those hirelings was a man "holding a deepgoing plough", an occupation known from an unpublished text (idi mukil barbi, BM 971 18:6). We know of a chief of plough-men (PA engar.meS, Waterman, BDHP 80 rev. 4; UET 5 236:24). On the other hand, the plough-man is in two texts a house-slave.111 Agricultural implements were received by a plough-man. l 2 Looking at the clausulas in the hires of both persons, we observe that a thin line divides the responsibilities of the ploughman and the ox-driver. But their wages are different. Ox-drivers Far more frequent are the ox-drivers. Several words or expressions indicating them existed. To start with the less frequent: Sumerian gud.da.ri.a. Well attested in the Ur I11 period, notably as hirelings. Rare in Old Baby1onian.l l3 The Sumerian verb da.ri means "to drive (animals),
St01
Old Babylonian cattle
to deliver, etc.".l14 We assume that gud.da.ri.a in a Sumerian law means "driven OX"; see note 105. warki alpi. In: PN ar-ki gud.N.a mu-di-Sa-tim (!) (YOS 12 421:2), PN egir 6b.gud.wa (YOS 13 348:12), also PN egir gud.115 Von Soden's dictionary has this occupation under (w)arki A 1 d, "Gespannfiihrer". Its literal meaning is "behind the oxen", elliptic for alik warki alpi "(the man) walking behind the oxen", a form actually attested in a Sumerian literary text (speaking of a boy) and an unpublished OB text.l Another elliptic expression is attested, a-li-ik wa-ar-ki-i-$u "walking behind him", so perhaps "ox-driver" (TIM 4 36:23). "To walk behind the oxen [once: sheep]" is an expression rather frequently attested in the texts but only a few refs. will be found in the dictionaries under alakum warki.l17 Most frequent is Sumerian SA.GUD[pronunciation in Sumerian unknown], Akkadian kullizum; a few times ( l ~ GUD.SA.GUD.~~~ ) The Sumerian signs combination was no more used after the Old Babylonian period and the element GUD, misread as GA, survived as ga in the later lexical tradition: (gud) Sh.ga (fJb). This later tradition seems to have mixed up oxen and ox-driversll9: gud. 3h.ga = kul-li-zu (see below: "trained ox") gud.ud.uS = re-du al-pi (cf. AHw 968b, 5. "Begleitrind" ?) gud.ud.da.ri = gim-li (see note 106) gud.ud.diri.ga = at-ta-ri (see note 106) gud.ud.da.na.gad = SU-u As a matter of fact, in the Old and Middle Babylonian periods an ox is meant by gud.Sh.gud; see simply CAD K 507b. As to the OB period, we suggest that this was the general word for "trained ox", used in the North (Sippar) whereas gud.giS was used in the South (Larsa, etc.). The best reading in Akkadian would be alap kullizi. This seems to be confirmed by the writing for its plural in Sumerian: 2 gud.w.a Sh.gud "two oxen of the ox-driver" (CT 48 33:7, after slaves; similarly 3 112 [sic; half a team of seven] [email protected] Sh.gud, in BE 611 68:4). Ox-drivers must have been low-class people. Parents rent out their children as oxdrivers.120 We read in a letter "The sons of Apil-kubi, whom you had given to an ox-driver, have gone away at your instigation", and two texts speak of "sons of ox-drivers".l21 In a list with expenses, the section on the plough-oxen is followed by "rationed food and drink of fJuzalum and sons".l22 Female ox-drivers are slaves.123 Hirelings served as ox-drivers.124 An account says that 4 ox-drivers received 24 kor of barley as allowance and adds: "that (is), per man a barley allowance of 6 kor" (24 gur Se.ba 4 erCn.Sh.gud, Sa 1u.l.e Se.ba 6 ~ e . ~ u r )It. lis~ not ~ said that these men were hirelings. Still, this 6 kor is in line with barley allowed to an ox-driver for one year's work in the Hammurabi Code, $ 258: "If a man has hired (agirum) an ox-driver, he shall give him 6 kor of barley per year". They are hired ana kullizutim.126 The hire (6.bi) of 6 shekels per year in YOS 13 20 (3.VII) is again in line with the Hamrnurabi Code (6 kor of barley [so 5 liters per day]) but this contract allows him an extra 130 liters barley as "his ration (Suku-su)". A youngster is hired from his father during one month and receives 112 shekel as monthly wages (6) plus 60 liters of barley
St01
Old Babylonian cattle
as his ration. The hire (4 shekels per year) is here less than the norm; the ration is far more (Goetze, JCS 11 [I9571 26 no. 14; 11.VI). Why are the rations also given ? We noticed that the hired plough-men also received an extra rationing of 2 liters of food and 2 I 3 liters of beer [per day] (PBS 812 196:9-10; BM 97543). To preclude pilfering with the barley that is to be used as seed? But allotting rations is more common in hirings of persons. Perhaps to our surprise, those poor ox-drivers could have a field assigned to them (Larsa).l27 They, and their replacements (diri), are given in a roster of personnel.l28 They had a "house" (= family?) in a village in the land of Larsa.129 We have records of ox-drivers male and female - killed by the enemy at the time of king Arnmi-~aduqa.130 It is an interesting question how many ox-drivers served in a team; we already have made some remarks. Survey: CT 8 30a, 1 plough-man (engar),2 ox-drivers, 1 female ox-driver; total 4 persons. Oxen: 6. Riftin 60: 4 ox-drivers. Oxen: 6. Hoe and Plough 91: 4 men. Oxen: 6. Kisurra 97: 6 ox-drivers. Oxen: 7. TIM 5 50: 2 plough-men, 6 ox-drivers; total 8 persons. Oxen: 14.
In Riftin 53:16-17, seven men appear to serve on two teams. Do three-and-half men work with one team ? Yes, this more or less fits CT 8 30a, TIM 5 50, Riftin 60, where a team has 6 oxen.131 Passages where the number of oxen is not given: a letter speaks of three ox-drivers; according to an account, "three hirelings went behind the oxen". 132 Twelve men are supposed to do service.133 Barley is given to "6 hirelings who go behind the oxen" (OECT 13 263 rev. 5). Pettinato and Waetzoldt wrote: "Die Aussaat erfolgte auf dem vorbereiteten Feld mit Hilfe eines Sapfluges, der von zwei Tieren gezogen wird. Zu seiner Bedienung wurden 2-3 Manner benotigt: einer fuhrt den Pflug, der andere treibt die Tiere an und der dritte wirft durch einen daran angebrachten Satrichter das Saatkorn in die Furchen". In Note 18 they add: "Fur ein Bedienungspersonal von 3 Mannern spricht auch, dass diese Leute [gud.da.ri.a] in den Personallisten der Tempe1 immer in Zahlen vorkommen, die durch 3 teilbar sinfl.134 Our figures are also often 3, 6, 12. Informative is now what M.W. Stolper wrote on the NB period: "Leases from Uruk supply one farmhand for each ox, four men for each team. Cyr. 26, from Sippar, shows eight workers with twelve oxen, and other Sippar texts indicate plowing teams of three or four workers with each four oxen".l35 The "plough" in the Mari texts seems to have included many more men than oxdrivers alone; there were even women among them. A personnel list from Mari shows that one "plough" consisted of 14 men and 2 women (millers); parallel texts name them "(people) serving on the field" (dik eqlim). They include persons not working with the plough (kullizu, fisimu, samihu, muZaqqd, guzalfi, and the two female millers).l36 An important letter from Mari complains that 12 men had been assigned to one "plough": one needs at least 15 men for the cutting of reed (kismum).l37 Can we say that 12 is the norm?
St01
Old Babylonian cnttlc
One more remark on the teams working with a plough. We have seen that those men could be put to other work, like cutting reeds, and not in the ploughing season. Other letters from Mari show that "the ploughs" were active in harvesting and threshing, too. Here, animals were used. In about the same months they prepared the soil for a summer crop like sesame.138 In our discussion on hired oxen (initum) we will see that they were involved in irrigation, and more.
Fodder Cattle need a lot of food, %.gal, ukullfim "(animal) fodder"; also imrfi.l39 We distinguish water, grass,l40 barley, draff, straw and reed. "There is no grass and the water is bitter".141 "Look where there is grass", for the calves and lambs (AbB 10 117). "Let them eat grass and straw; let one ox eat 20 liters of straw" (AbB 3 11:18-19). "The oxen eat moist draff (...); let thirty oxen eat grass" (TCL 17 409). "Since yesterday I have no barley and straw; they are starving" (AbB 11 27:27-8). "There is no straw as oxen fodder" (ana li-ku-le-e gud.u.a, AbB 8 7: 15). "That straw has been used up and what shall your oxen eat ?' (AbB 11 160). "Let them give me straw and barley (as) fodder of (!) your oxen" (AbB 9 259:21-24). Grass is cheapest but was only in spring abundantly available. Extra feeding could be necessary for sheer survival or "reinforcement" (tadnintum, tablittum) (more below). "Do the oxen that have not eaten fodder in months IV-VI stay alive ?'(AbB 1 118:8-9). This means that the amounts of fodder given by the texts are variable; they depend on the season.142 "To take food", said of the oxen, is jabaturn in Akkadian.143 Barley must have been an expensive kind of food. The amounts given to oxen (gud) vary wildly, but 10 liters seems to have been one norm. In our discussion on hired cattle (initum) we will see that the hire was 112 shekel of silver for 15 days. In barley, this would be 112 kor, 150 liters. This means: 10 liters per day as hire. Explicit about individual animals is a text on expenditures for months IX-XI: a bull in its second year receives 3 liters, an ox or a cow 3 liters, a calf 2 liters (Riftin 58: 2,6, 10, 14). From another text, on month X, we can compute this: an ox (gud) receives 10 liters daily, a calf 3 liters daily.144 For month VIII, to 30 head of cattle 4 liters per day is allotted.145 For Mari, we can perhaps compute 5 liters per day for ploughing oxen,l46 but another text explicitly gives 10 liters,l47 and we find 30 liters.148 Elsewhere, an ox gets 10 liters.149 In the Diyala region: 15 liters for (trained) oxen.150 C.J. Gadd wrote in his first discussion of the Chagar Bazar texts "at 10 sila [= liters] daily apie~e".l5~ Far lower amounts are also recorded, notably 2 or 3 liters.152 In the Middle Assyrian period a "Pflugrind" received 1 112 or 1 114 liter per day.ls3 It is thinkable that in the North the animals could get additional food by grazing "on the side". Sometimes we can try to compute how much an ox in a team received. 10 liters: "40 liters of barley, fodder of one plough-team of P N . If this is a team of four, each animal gets 10 liters.154 Does the recurring "they ate 60 liters each" in another text presuppose a team of six oxen?l55 13, 333.. liters: 80 liters per day reserved for "6 plough-oxen", which would mean 13, 33 ... liters daily per ox (for months 111-V).l56
St01
Old Babylonian cattle
-
unknown: per day 12 liters, "fodder of oxen", as can be deduced from "264 liters for 22 days" in the text.157 An intriguing text from Ur begins with these two entries (UET 5 582): 1260 liters of barley, fodder of the oxen, 30 days; per day 42 liters. 640 liters, fodder of the oxen, 20 days. The first entry gives 42 liters per day. We suggest a team of 6 oxen; every ox takes 7 liters per day. The second entry means 32 liters per day. We suggest a team of 4 oxen; every ox takes 8 liters per day. Their work must have been different; deep-ploughing followed by seed-ploughing? Other texts on fodder given to a gud.apin do not yield usable figures.158 M.A. Powell wrote on this, for the Ur I11 period: "The daily feeding rate per ox is calculated on the basis of 30 sila (= liters) per day for a team. This means that a team of four oxen would get 7.5 sila each per day, in cases where there is an extra ox, six sila each". "Six sila per day, perhaps with some grazing in the evening or fodder like reeds, seems about right. Much less than this it cannot be".l59 During our meeting, it was remarked that four liters per day is much; ten is far too much. The maximum that I have seen is in an Ur I11 text: 16 liters for an ox, 9 for a calf.160 There is an answer. It is possible that animals got extra fodder for "reinforcement" (tablittum or tadnintum). The word tablittum can be translated as "physical improvement". It derives from the causative of the verb balatum "to make healthy". Another word for the same activity is tadnintum "reinforcement", from the verb daniinum "to be I make strong", often said of oxen.161 Both words are particularly used for strengthening the "trained ox" (gud.giri): examples are YOS 5 181:3, 9, 15 (tablittum); 175:4-5, 184:2, 9 (tadnintum).'62 We learn from some letters that, on the other hand, they tended to feed the animals too much which resulted in disease.163 D r a a a residue of beer-brewing, is given for fattening cattle or sheep.164 "Your fodder is draff', a fable says of the cow (im-ru-ka tub-bu, Lambert, BWL [I9601 180 A rev. 29). There was dry and moist draff; moist draff is more often asked for; dry draff 0ccasionally.~65The daily amounts used for working animals are: 10 liters of barley and 30 liters of draff (AbB 4 156 rev. 8); 20 or 30 liters of draff (TLB 1 45, cf. 44). 20 liters of moist draff (AbB 7 3: 15); 30 liters of moist draff (CBS 528:3). Straw. In a land lease, "barley-seed, fodder, straw, draff, expenses" are given to the cultivator.166 20 liters per day (AbB 3 11:19; perhaps also in YOS 12 219:2). Reed. Two letters show that oxen can be found in the reed lands (ap~m).16~ A spare ox (?) entering the reeds (gikgi) was in danger of being killed by a lion (FLP 1287 VI 16-22, diss. M.T. Roth, 229). From a letter it appears that they brought the oxen reed (@[email protected])as food for lack of barley in month XIII. The Ur I11 texts from Umma show that reed was given in months X-XI11 (MHET I no. 89:15-26, with van Driel, p. 134). Groats are exceptional; each ox 6 213 liters (Chagar Bazar nos. 33,38).
Sto1
Old Babylonian cattle
D. THE initum This is a difficult topic. At the meeting of the Sumerian Agriculture Group, all references were presented and discussed in a long but inconclusive paper, by this author. It could be established that the value of one initum was 112 shekel of silver and that this covered work performed during half a month. Furthermore, standards were discovered for the surface worked by the various types of initum. These conclusions were published in Studies U o n de Meyer (1994).168 I must confess that the initum still eludes me. Here, only a few aspects of the problem will be discussed.
initum = oxen Landsberger gave two translations for the word initum. In his own words (MSL 1 234; italics mine): 1. "Saisonsweise zu variablem Kurs gernietetes Rind (nebst Treiber)"; 2. Mietspreis dafiir, pro Monat berechnet". These remain the best definitions.169 The Sumerogram is i.gud, literally "wages for an ox". The OB refs. where it clearly stands for initum and does not mean just "wages for an ox" are few; PBS 812 179:2 and AbB 6 191:14. They were not studied by Landsberger. That initum stands for the oxen, can be shown by new examples not known to Landsberger: in hires of persons working with oxen we have a clausula where oxen ([email protected]) and initum are variants: 170 i-ni-tam 6-ra-aq-ma ku.babbar i-ni-tim i.1i.e (a) [email protected] 6-ra-aq-ma khbabbar i-ni-tim i-?a-qd-la (b) "If he makes the i. I oxen remain idle, he shall pay the silver of initum". Another clausula, attested in corrupted form in a late version of the handbook of scribes Ana ittis'u - corrupted in offering eliitu for initum - is now known from the contracts, in slightly different forms, where we read again [email protected] "oxen" instead of elitum = initum: amilu kima amili [e-l]a-ta ki-ma e-la-ti [ze-ra ki-m]a ze-ri [is'-te-nis' i-s'a-kan]u (Ana ittifu, MSL 1 55 Tf. 4, I:62-65) a-we-lum ma-la a-we-lim [email protected] rie.numun 3 ma-nu-ab-tam i-Sa-ak-ka-nu (VAS 22 26: 19-22) [email protected] ma-li [email protected] Se.numun ma-li Ee.numun i-s'a-ka-nu-ma iz-te-ni-iS i-zu-uzzu (BIN 7 191:ll-13) [email protected] Se.numun 2 ma-nu-ab-tam a-we-lum ma-la a-we-lim ma-na-ab-tam i3a-ak--nu-mu aJB i-ir-re-Su (YOS 13 332: 10-13). Letters and a few contracts use verbs for initums normally used for handling cattle: they are "driven" (initam redam, Scheil, SFS 242); "put out to pasture" (nudam) into the fields. l7l The identity of initum with "oxen" (plural) in the clause just studied could give us the idea that it actually is a team of more than one animal. Some passages suggest this but in
St01
Old Babylonian cattle
every case there is a problem. The 60 litres of barley as food of one initum (?) in TLB I 46 could be used as an argument for a team of six animals each eating 10 litres per day (see note 155). However, another explanation has been offered: it refers to the standard period of 15 days for an initum and the ox eats 4 liters per day.172 The combination "ox of initum(s)" is attested twice: (a) In: 6 gud i-ni-tim (YOS 13 424). Six of them looks at first sight like a team of "6 oxen of an initum" , but "six hired oxen" is as good. (b) The same may be said of "their ox of the initums" (gud i-ni-a-ti-Su-nu), in: "The plough-men (...) should gather the barley [apparently as fodder] of their initumoxen". The meaning is: hired oxen.173 A few texts speak of "half' of an initum (YOS 13 149 [8 112 and 12 1/21; BM 97006, rev. [ l 112 and 32 1/21). Here, we ought to be cautious. At first sight, this suggest a plurality, a team, but the fractions may have to do with calculations based on a standard duration of the work performed. This norm is 15 days, as we have shown in Studies L60.n de Meyer.174 There, it appeared that the initum hire is 112 shekel for 15 days, so 1 shekel per month.175 This gives us an additional argument for thinking of only one animal: the hire of a rear ox or a cow is in the ploughing season 1 shekel per month - exactly the initum.
initum as hire The greatest discovery by Landsberger was that he saw the parallellism between hiring oxen and hiring harvesters. The point is, that the demand of harvesters I oxen will be very high at harvest I ploughing time, so the labour force was contracted long before, as evidenced by two kinds of contract. The so-called "harvest labour contracts" were and are available in great numbers and follow these two formulas: in the one, an amount of silver (or barley) is borrowed (Sumerian Su.ba.an.ti) by a middle-man with the promise to deliver harvesters (a "loan"); in the other, the debt incurred by such a transaction is recorded (eli PN iSfi; the "Verpflichtungsschein"). A point of discussion is, whether the amount of money borrowed was the full price ("Entgelt"), or only a first down-payment ("Angeld"). In my Een Babylonier maakt schulden (1983) 6-7 I have proposed that the "Verpflichtungsschein" concerns a part of a debt that still is to be paid ("Restschuld"). If this also applies to our texts, the Su.ba.an.ti was the original contract giving the full amount, and the "Verpflichtungsschein" was only secondary. 176 Let us now have a look at the oxen, especially needed during the ploughing season, months V-VIII. Landsberger had four "Verpflichtungsscheine" when showing the parallellism with the harvesters and he listed them under group I, "Saison-Vorausmieten von Rindern", and he had one "loanW.l77We now have more texts. Two more "Verpflichtungsscheine" and no less than nine more "loans".178 The "loan" in the Archive of the ensi U~riyaruns perfectly parallel to the harvest labour contracts.179 The same middle-men act in the contracts on oxen and harvest labour of this archive (Stol, Studies on Old Babylonian History 90 ff.). The remark about 12 112 initums "which did work on the field Bdi-SamSi" is phrased like that on a given number of harvesters "who will harvest on the field of PN.180 Another common point is that some initum -contracts give the clausula on the application of the "regulation of the king" so well known from harvest labour contracts.181 This replaces "he shall pay the
Stol
Old Babylonian cattle
silver of an initum" in other contracts (given in note 170). We assume that this royal regulation was the official tariff for hiring oxen. Note that the gu.ba.an.ti "loans" for oxen differ widely in their stipulations. There was no standard. In contrast, the harvest labour contracts follow one rather rigid "loan" formulary. Some texts mention both "price of an initum" and "hire of oxen". What the difference is, is a matter of further research. lg2
Joint ownership In contrast to the epinnum, the "plough" of a large organisation, the initum could be hired. What complicates matters is that the initum could have more than one owner. There is evidence that the hired oxen were owned by more than one person. The best example is VAS 7 92 (VAB 5 148). A rear ox was rented out by Ibni-S?n, "at the word of Q."; the rent is 1 shekel. The text says that Q. had already received half a shekel. We assume that they were partners. More complicated is the sale of an ox (with a proper name !), TCL 1 233.183 TWOpersons buy this ox at "8 114 shekel silver, weight-stone of Sama~".This weightstone points to the temple of SamaS, where partnerships were concluded and dissolved. The price is low and I hold it possible that the seller remained one of the partners.184 The conditions are as follows: "The one, as much as the other, will receive its initums (i-ni-a-ti-Su a-hu-um ki-ma a-hi-im i-le-qk). On the day they wish (to do this) (lit. "on the day of their heart"), they may sell it and its price they will divide". The initums is here "the income derived from its hire" (CAD); the plural shows: expected many times. If one of the two partners wishes to sell the ox, this will be done. "Their heart" referring to the buyers suggests an opposition between the two buyers and another party - the seller, as third partner, I think. 185 The soldier Ubarrum seems to share an initum with his colleague.lg6 TWOtexts from the same city show that initums are owned by a general and another person (YOS 13 239, 241).187 A "round tablet" from Lagaba speaks of "150 liters, initum, from his share".188 In Sippar, the manager of land (ensi) BElijatum worked fields for many high-placed persons and he received a "share" of the yield (TCL 1 168, 230 rev.).The first person he worked with was the herdsman ElmESum. They settle their accounts in TCL 1 229. Here, we see that B. has a share in the "price" of 3 initums "with E." which was 360 litres; "200 litres with E., from the price of the initums; 90 litres of barley, hire of 6 oxen 1 cows - 290 litres of barley, of B .".I89 Partners (be1 tappzitim) are involved in the discussion on initums in AbB 1 17:3234.190 A long list of expenses groups together: "60 litres of barley (for) the man going behind the oxen; 10 litres (for) Mir-er~etimand his partner (tappfim); 20 litres, 'party' of the initums of a price; 5 litres (for) SamaS-ni~uand his partner".lgl Remarkably similar is this passage not mentioning the word initum but epinnum: "And a letter of mine has gone to PN the judge and PN2 the general about giving the plough-teams and working with you in partnershipW.l92
Sto1
Old Babylonian cattle
Work other than ploughing The oxen had several distinct tasks to perform and the initum is often qualified as such. Best known is their preparing the soil for cultivation: "deep-ploughing", "loosening", "harrowing", "breaking", "seed-ploughing"; see the survey in Studies U o n de Meyer (1994) 232. When studying the epinnum-crews, we were a little surprised to see them working at harvest time. There is evidence that the initum, too, had work to do other than ploughing. The initum can also be active during the harvest season, months 1-111. This is surprising, because we have tacitly assumed that they only did the ploughing (in months VVIII). At threshing time "the other initums" are put out to pasture (AbB 7 170:17). They are transferred "before the harvest" (AbB 9 85:11). Working with 3 initums and working with ten harvesters are mentioned in one breath (AbB 10 181:6). A man is hired to "go in and out" during one month and is entitled to 2 initums as monthly wages - the text is dated to 11.11, harvest time (VAS 7 144). Irrigation is another example. Not using the initum implies the danger that the field "dries up": "He shall bear the responsibility for the field drying up" (pibat eqlim Sa ib-ba-lu inas's'i, BIN 7 204: 11-12). In a letter on initums: "You shall not say: My field had been cleared before S., and he made my field dry up" (eqli ana pani $. z[a]-k[u]-ma eqli u'-Sa-bi-il ,AbB 3 31:ll-14). In no. 29 the writer complains that the oxen are idle because there is no water in the canal. In another letter, "plough-teams" are asked for and this remark follows: "They will call you to account for the field that dries up" (ana eqlim s'a i-ba-lu i,Falluka, AbB 6 199:2223). Excursus -The words used for this irrigation could be dilitum and talpittum. The first word is perhaps attested in Gautier, Dilbat 50:l (VAB 5 111). The CAD suggests, with hesitation, 1 i-ni-it di-3u-tim ("threshing"). But di-ku-tim ("forced labour") fits the copy best. I suggest di-lu-tim ("hoisting irrigation water"). The delivery of the initum "for hoisting water" is expected in month 111. Exceptional is iniit talpitti attested in two unpublished texts.193 An Old Babylonian text has these lines: i-na Si-pi-ir ta-al-pi-t[im] s'a 2 i-ni-a-tim is'-te-at (??) i-nu-di-in, "at (the time of) the work of t. of [= for] two initums, he will give one" (YOS 12 221). Perhaps the irrigation technique me^lapatum is meant, lit. "to touch with water" (CAD L 86a; add AbB 10 42:34). A Neo-Assyrian text qualifies a field as A.SA.GA tal-pi-tu' (Postgate, CTN I1 no. 64:4, with Postgate, SAAB 7 [I9931 6). The presumed "hoisting irrigation water" and "touching" the field with water suggest that oxen were involved in irrigating fields. Indeed, in earlier Assyriology GIS.APIN = naqtabu was taken to be a hoisting machine. In Egypt, oxen are seen to draw. similar machines; in Iraq the saqia; see Charles BSA 4 (1980) 20. However, there is no Babylonian evidence for oxendrawn irrigation machinery, I believe (cf. Laessae, JCS 7 [I9531 22 f.).lg4
A ceremony A remarkable expression is kirri iniatim. In lists of expenses for agricultural work we come across "15 liters, kirrum of an initum" (TLB 1 184:4); "150 liters, kirrum of initums" (BM 97133:8); "20 liters kirrum of initums of a price" (BM 96967 rev. I 6, 2 bin ki-ir-ri i-ni-a-at
St01
Old Babylonian cattle
s'i-mi; see note 191). Kraus has shown that this word kirrum refers to a solemn ''partyW.195 One can imagine that this "party of the ploughing oxen" refers to the beginning of the ploughing season. Confirmation comes from a school letter where we expect to find normal requests: "You wrote: Find me a field of 36 iku for cultivating so that I may cultivate a field. Now, I have found you a field of 36 iku for cultivating; come quickly here and pour out the kirrum of the field and make out a document".l96 The "party" inaugurates the cultivation sealed by the contract ("document") (erres'utum). A passage overlooked by Kraus is this: "Write me (the name of) the field that I have to 'open' so that I may pour out the kirrum" (AbB 2 157:12-14). This must be the occasion when the "deep-ploughing" initums start working; the very beginning of this season.lg7 This "party" also seals the agreement in marriage, or takes place when a boat is going to make a trip (hire?). So it is a ceremonial happening of legal importance.198
NOTES
I thank Bram Jagersma, Els Woestenburg, K.R.Veenhof and M.Sigrist for giving me access to unpublished texts transliterated or copied by them. UET 5 806:10, 810:1, 6.21; 811 rev. 4; 842:l; 849:1,3, etc.; YOS 14 311:36, all from Ur; S.D. Walters, Water for Larsa (1970) Text 25:15. Landsberger, MSL 811 (1960) 61; Goetze, JCS 17 (1963) 85a. - For Goetze's error "4 la-at' instead of sh-la-at, see note 7. AbB 7 170:6, 10. A new example is the OB omen 6b.gud.w.a ana bitim irub [sing.], F.Joann&s,Studies Uon de Meyer (1994) 306: 18. OECT 4 152 V 40 f., li-tu-ul Sa li-a-ti, with Landsberger, MSL 2 (1951) 106, and CAD L 217a (ProtoDiri 278 f.). The redundant Sa in this title is a grammatical problem. The epithet of the goddess of cattle, Nindara, is ri'i utullciti; Sumerian sipa 6b.lu.lu.~.a(AHw 1445b, utullu A 3.c; M.Civi1, Studies A.W.sjoberg [I9891 51b). This is another problem: according to the Sumerian, the plural utullciti means "cows". The word looks like a frozen epithet and A.Cavigneaux has explained it as utul licti, presumably "herdsman of cows I cattle"; Aula Orientalis 5 (1987) 254 ad 22. Landsberger, MSL 811 76; F.R. Kraus, Staatliche Viehhaltung im altbabylonischen Lande Larsa (1966) 43. The roster is A.Goetze, JCS 2 (1948) 80-83, 105-106no. 9 (dated 3O.VIII). Misread by Goetze who saw in la-at an attestation of the expected li'citum "cattle". Other refs. for salatum (of cattle or sheep and goats): AbB 2 15:23, cf. 6; AbB 10 23: 18; AbB 13 33: 10; Arnaud, ARV 159:11. Correct CAD $88, ~altu3.b. K. Maekawa, Zinbun 15 [I9791 98. "Growth of a Herd of Cattle": I.J.Gelb, JCS 21 (1967) 64-69. Th. Meek, AJSL 33 (1916-17) 239 RFH 30:4,20; contrast 6b.mu.l in line 11. Certainly not "milchende Kuh von 1 Jahre" as in HG 6 1930. K. Butz, WZKM 65-66 (1973-4) 44 note 111: "Kuhkidber werden ab
Old Babylonian cattle
der 4.-5. Woche, Bullenkdber ab der 6. Woche langsam entwohnt". - The scheme given by K. Butz for Ur, p. 44, is not in line with Sumerian terminology, nor with practice. Ages in Nuzi: Landsberger, MSL 811 47; MB and NB: ibidem, 70,77; G. van Driel, BSA 7 (1993) 228, 230. MA: K. Deller, Baghd. Mitt. 16 [I9851 320 ff. For Human words, see note 35. - See also A.Salonen, Agricultura Mesopotamica (1968) 388-9. G.Dalman, Arbeit und Sitte in Palustina VI (1939) 163, "Altersstufen in Aleppo": einjtiluig - zweijtiluig - dreijtiluig - vierjahrig - tor (bull) 1 bqara (cow). In the Bible: Dalman, p. 172: "Dreijahrige Kalbe [fem] in Gen. 15:9; einjahriges Kalb in Lev. 9:2, 4; dreijahriger Stier in 1 Sam. 1:24 [emendation]". Cf. MSL 811 47 note, end. F.R.Kraus, ZA 53 (1959) 140 f., "B. Rinder". There is now an unpublished dissertation on this archive; J.F.Robertson, Redistributive economies in ancient Mesopotamian society. A case-studyfrom Isin-Larsa period Nippur (University of Pennsylvania, 1981). B. Landsberger, MSL 811 63 note 1; 76 sub V: mahx; now in R. Borger, ABZ: m6h. See below; the other refs. are CT 45 27:2 and B.Kienast, Kisurra 124:2. YOS 8 164:l-2, resp. TCL 10 99:l-2, as collated by D. Arnaud, RA 70 (1976) 87a. UET 5 823:l-4; cf. 827:l-2 [75 6b.AL e-ri-turn],830:2 [13 6b.AL re-q6-turn],also 839:26. Lastly, J.Bauer, AfO 36-37 (1989-90) 82b ad 13 IV 5K. E. Szlechter, TJAUB [I9631 37 UMM G 29 [price 14 shekels of silver], Th.G.Pinches, PSBA 39 [I9171 Plate XII no. 24 [price 66 shekels of silver, in the weight-stone of $amas]. Note the sequence "one bull, one cow, one suckling bull-calf' in a list (Walters, Water for Larsa, Text 25:12). There is one ref. for simple "cow" with suckling bull-calf: 1rib qa-du-urn amar.ga dumu.ni (BM. 96956:22,96980: 10; see L.DeKiere, NABU 19911110). A cow belonging to a team-of-four of a plough (rib erbettitum) has a suckling bull-calf (YOS 13 355). Line 26: eme.gud.SB.gud.ra I li-$6-an kul-li-zi; eme AB.KU I li-Sd-an li-tul-li, A.w.sj6berg, ZA 64 (1975) 144. Landsberger, MSL 811 76, with note; Butz, WZKM 65-66 (1973-74) 44-5 note 113. More words, hard to read, in UET 5 839:14, 19,20. UCP 10 no. 3:3-4; collated by S. Greengus, Studies 98. M.deJong Ellis, JCS 24 [I9721 67 no. 69: 10-12; see note 68. A. 3524:18-22; see MSL 811 63 and CAD B 341a, [email protected] iz-za-az-za li-ul id-di-nu-ni-aSi-im 6b.w.a li-$a-ab-bi-it-ma10 Se gur KI.UD Sa a-hi-tam-ma ad-du-li ad-di-iS. A.Goetze, Sumer 14 (1958) no. 9:4, bu-lir-ta-am Ja up-pa-ti hu pu ti ina pcinika l i t h i m ; again from the Diyala region. UCP 10 3:11, cf. no. 41, SB.gal ar-hi-im Sa PN i-nu-ma re-qa-at. Landsberger, MSL 811 64 note 1. Is the 66 two-thirds of 100 in a commercial society? Cf. notes 184, 188. YOS 5 150:14-16,27-29; 13 97:5,6,9. YOS 8 164:3,7; 12 103:2-3;just 6b.mu.2 in TCLlO 99:5. V.Schei1, RA 14 (1917) 153-4 (HG 6 1650) is not clear; sale of:1 Bb.mu.3 q6-du amar ta(?) bi ir Su. A cow in her thlrd year: van Lerberghe, Philadelphia (= OLA 21) 71:11,23. In her second year: YOS 13 29:4; BM. 96956:39, see DeKiere, NABU 19911110. In her first year; Pinches, Peek 14:9. Lists: UET 5 819:7,820:7, etc; YOS 5 150:19,32. The roster is UET 5 846. 3 6b.AL 2 gud.mu.3 Sa ana gud.apin PN i-lu-li, VAS 18 93:1.
Old Babylonian cutllc
YOS 13 348:4-5, 8; 350:5-7; 351. Skipped in YOS 13 97 and O.Loretz, Texte aus Chagar Bazar und Tell Brak (1969) no. 35. Meek, AJSL 33 242 RFH 36,38 (= HG 6 1927,1929); AbB 9 89:7; CT 48 100 (as a funeral offering). Examples: 1gud limmu.ba (BM 96956:10, inheritance, courtesy K.R. Veenhof; cf. DeKiere, NABU 19911110); and: gud mu.limmu (PRAK 2 C 73:5). Higher ages in Human Akkadian from Nuzi: tumnarpu "in its fourth year" and narijarpu "in its fifth year"; see for the latter G.Wilhelm, SMEA 24 (1984) 223-4. Akkadian: Deller, Baghd. Mitt. 16 323 f. Note, however, the unusual signs for amar in the Chagar Bazar texts (ed. O.Loretz, nos. 12 and 35; Gadd, Iraq 7 [I9401 Plate I1 A. 976) and in Ur, looking like gri in UET 5 831:5-6. Goetze, Sumer 14 no. 6: 14. - Note the obscure "16b Sa bu ra turn " in UET 5 839:20. Van Lerberghe, Philadelphia 23. We read in 14 'Sa ji-di7-it. The word kissatum for "fodder" is interesting; why not SB.gal = ukulllim? CAD K 427: "chopped straw". For the double function na.gada and sipa of Ina-paleHu, see M.Stol, Mklanges M.Birot (1985) 274 f. (not this example). His father Ilumkinam-ide and cattle in AbB 7 37:6; see D.Charpin, JA 270 (1982) 39. Kisurra: Kienast, Kisurra 2 nos. 100:3-4; 124:3; 125-6; cf. p. 123. Ur: UET 5 806:5 and 9; cf. 6b.amar.ga in 811:3 and gud.amar.ga in YOS 5 39:4. 6b.ga and amar.ga also paired in 831:2,6; 836:18,20; 843:20,25,26; amar.ga alone in 820:11,825:24, 840:19. We cannot decide what these signs stand for in a text from the Diyala region: after 1gud Sa ni-ri-im we read: 3 6b.amar.ga lu.nigin 3 [error for 4?] gud.w.a: "3 sucking calves", or "3 sucking she-calves" "total 3 [error for 4 ?] 'oxen' " (UCP 10 no. 27:9- 11). 1 'amarl bu-ur-ti ti (?)-iz-bi,YOS 12 447:l. If burtum means "cow" here, "calf of a cow" would be the literal translation of amar burti . But in view of the form paqdat the animal is female, so burtum is the subject and amar is just determinative. Does amar precede in order to avoid confusion with burtum "full-grown cow" ? 4 6b.amar.u.a in CT 48 3 3 9 is a problem. It occurs in a list: named [male and] female slaves - 2 gud.u.a SB.gud (ox-drivers?) - 4 6b.amar.w.a - 30 [email protected] (sheep and goats). Error for db.gud(!).M.a"cattle"? The attestations for amar.ga.KUD are few: Goetze, JCS 2 (1948) 105 no. 9:14 (among bulls in their second year); YOS 13 350:21-44. On weaning in the animal's first year, see above, under "Ages". Maekawa has written that in earlier periods amar.KUD means "castrated calf'. According to J. Bauer, however, the word KUD means "to separate" young animal and mother and he gives examples for this practice (AfO 36-37 [1989-901 88-9). D. Arnaud, ARV 98:7; AbB'9 83:24; Friedrich, BA V 51 1 no. 46:6, 11. See also J.-M.Durand, "Enclos purrusaturn", NABU 1993155 ("l'endroit ob on rBserve les meilleurs animaux"). Note the obscure remark 18 rib.AL e-zu-ub 1rib it-ta-nu-up-ra-a9li 16b Sagan (?), YOS 13 97:l-3. Is ittanapraS/s the cow's name? Kish: YOS 13 350,351, cf. 348 (the Palace); Sippar: Goetze, JCS 2 105 no. 9 (F'rincess Iltani). ARMT 25 616:59-60, see now J.-M.Durand, AEM 111 (1988) 101. The governor Mir-er~etim:AbB 4 5:8, cf. VAS 13 30 rev. 2. For Sin-mustid: M.Sto1, SOBH 51 f. YOS 13 97, translated by J.N.Postgate, JSS 20 (1975) 17. We have seen that 30 calves of citizens were part of a herd of 200; van Lerberghe, Philadelphia 23.
St01
Old Babylonian cattle
Here, we find "little" Ilum-babil (TUR;4); he and "big" Hum-babil are attested in 3505, 16. Mr. Big has a higher position. One is inclined to relate the ul ubbalamma in line 19 to the ublamm in line 15 and to emend the latter into ubbalamma "he shall bring (them) on 28.VI" - "if he does not bring them.. ." Line 20 and seal impression. Also in YOS 13 17:14 (use of a cow; first witness). Could this sib.KUD.KUD be the sib pu-ru-se-e-tum(plur.) of CT 8 28c: 12? Two observations on ma.al.1~: (a) in the Sumerian Emesal dialect the word em.ma.al "cow" exists; (b) each section in The Song of the Ploughing Oxen opens with the words e.el.lu ma.al.1~(var. ma.a.1~); M. Civil, Kramer Anniversary Volume (1976) 83-95. BM 96967 I1 20-22 (transcription by Bram Jagersma): (20) 0.0.4 Ili.;h.gud (21) [0.0].1, 6 sila i-nu-ma gi;.apin nu-la-mi-du (?) (22) [...I 4 sila ma-li gi;.apin. MAD 1 305. For the plough: 1 gud.sib gud.apin.G PN sag.zug7.ra i.na.sum, 6-9 (sag.zug7 =
SAG.^). "Herdsman of cattle" is frequent; AbB 7 139:2; 9 1954; etc. "Herdsman of cows": Goetze, Sumer 14 no. 18:9 (sipad 6b.M.a Sa &gal-lim); TCL 1 1:17 (= VAB 6 37) (Sh sipad 6b.w.a). J.-M. Durand, Etudes Paul Garelli (1991) 52, a. Kasap ritim: AbB 3 8:4, 59:14, with R.Frankena, SLB IV (1978) 29 f. Among the investments in agriculture: 66 2/3 liters (of barley), kb.babbar ri-tim,TLB 1 128:4 (collated); the 66 213 must represent two shares out of three (see note 188). Cf. perhaps "(silver) as price for grass of the pastures of sheep" (ana 5 m Sammi ra-a-at [email protected]), CT 33 27:3. - Note the obscure first lines of a contract published by A.Ungnad, ZA 36 (1925) 92 BJ 87:l-5,4 Se.gur Su-ta-up-tigud ri-i-tim [so?] S.u.ti.a PN. TCL 1 166, VAB 5 162. This text is still misunderstood in Assyriology. Our translation: "From 14.11, Sumum-lib~i,son of Ardu, was pasturing (i-ri-i ,from re'fim) the trained ox (gud.lih.gud) of Sh-iqiSam, the head of the assembly. Belijatum had assigned for him 10 liters of barley as his wages; "does he quit (and) go away, he will lose his wages". Until 27.n he was pasturing the trained ox, but he has gone away (u i-ta-li-ik). Via BElijatum and Sh-Semi. 27.11, Ammi-~aduqayear 13". This is a document establishing the disappearance of the herdsman, written on that very day. Belijatum, his chief, is a well known manager of a landed estate. What was so misleading is that the clausula about disappearance "does he quit ..." is quoted from the original herding contract, without introduction. Nadfim of hired plough-teams: AbB 7 107:17; 10 64:13 [ina hamn']. Calves: AbB 8 139:14. AbB 3 4: 12, with Frankena, SLB IV 18-9, "frei umher laufen lassen"; accepted by Kraus in AbB 10 101:15. Verb izuzzum: YOS 13 5025; AbB 11 27:24; ARM 4 758; cows: UET 5 839:6, 15; A. 3524:18 (see note 22). Causative 3: YOS 12 381:13-14; AbB 2 159:14. Cf. [mu-an]-za-zu-urn pa-nu-urn, AbB 8 70:15, and see W. von Soden, Stud. Or. 46 (1975) 331, on CT 13 37:30. Found in the British Museum by Bram Jagersma. One ox: BM 97963 (12.V). Two oxen: BM 97954 (16.V), 97993 (21.V), 97997 (8.V). Five oxen: BM 98004 ([x].V). No number: BM 98032 (7.V). Cited by Renger, BSA 5 [I9901 271. Quite different is K.Deller, Baghd. Mitt. 16 (1985) 323, "Als Zeitpunkt der Kastration kame wohl die Entwohnung in Frage". Is this eine reine Vermutung '1 On the other hand, Jean-Pierre Digard, Techniques des nomdes baxtiyhri d'lran (198 1) 84: in Persia "les boeufs (sont) castris vers 4-5 ans". The situation here is different: those animals were not used for ploughs by these nomads, I think.
St01
Old Babylonian cattlc
We do not know for sure the Akkadian word for castration. Landsberger thought of kaparum, kazuzum (The Date Palm 34b). For the NB period, see G. van Driel, BSA 7 (1993) 238,250 note 121. M.deJong Ellis, JCS 24 [I9721 67 no. 69:lO-12, gud Su-li ~e-he-erde-5-a-am li-ul i-le-i, cf. 15, gud Su-li de-Sa-am li-ul i-le-i. CAD N12 261-2; add GUD SUDUN, Arnaud, RA 66 (1972) 173:64. Cf. [gud.giS] = a-lap ni-i-[ri], Hh XI11 292 (cf. Salonen, Agricultura Mesopotamica 376 f., "Jochochse"). In Assyrian royal inscriptions, GIS.GIS is a Sumerogram for "yoke"; W.Mayer, Or NS 56 (1987) 208. MSL 1 55 IV i 68 (Ana ittiSu); ARM 27 37:33-36. M. deJong Ellis, JCS 24 (1972) 48-50, nos. 13-20. Note 61 no. 53, with the groups guru;.;e.KIN "harvesters" (4), gi9.mar.gid.da "wagon" (7). guru; (!) da-i-Su "threshers" (lo), gud da-i-Su-tum "threshing oxen" (rev. 5). Goetze, Sumer 14 no. 21:2,37:5; cf. Greengus, OBTI 33; YOS 14 66. Goetze, JCS 11 (1957) 39 no. 31 and BE 611 55. The verb sadirum here must have a special meaning. It is also attested in Waterman, BDHP no. 1 7 5 (person a-na sa-da-ri-im i-gur-Su). M.Civi1, Studies Landsberger (1965) 7 III 10; simple "under the yoke" as variant in Roth, JCS 32 (1980) 130 $7. A. 3892, cited AEM 111 123 note 20. "Litikre B roues", Durand, ARMT 21 258:16,20. CT 48 43:3,1 ' 6 1 ta-am-Sa-ru [o], after 1gil.apin.S.ghr (a plough) and 2 gig ni-ru "yoke". Cf. 3 ta-amSa-ra-[tulm, OBTI 248:3. The other refs. for tamSarum are: YOS 13 380:1, OBTI 123:2 [see my note 106, end], BM 80408 (all qualifying an ox); CBS 1314:2 (4 112 gin ku.babbar ;Am 1sib ta-am-rSal-r[i] Sa PN1 ugu PN;? i-Su-4 (15.IV). For gi8.apin.w.a "les Cquipes rurales qui travaillent sur ces terres", see M.Birot, Syria 41 (1964) 45 f., "(personnel) d'une charrue", Ph. Talon, Iraq 45 (1983) 47 ff. One Mari text differentiates between giS.apin "the plough", and apin "plough-team", ARMT 24 13:5-6. Note in line 14 8h.gal [email protected], "fodder of the oxen"; "oxen" clearly distinguished from "plough teams" ([email protected]) in line 20. 65.ghr followed by gud.apin also AbB 1 33:11, 14; 8 74:4, 10; by giS.apin in YOS 14 300-305; ARMT 13 38:11, and more often in Mari. AbB 3 29; 94; 6 131:18-20; 10 16:15 f., TCL 1 37:7,23. Cf. "2 gud.giLapinWnext to alpum "ox" in AbB 8 84:33-38. An unpublished letter speaks of a gud.apin ta-aPni-im, loosely translated as "double plough". Does this mean a plough drawn by two animals ? (AUAM 73.3233 = HTS 43; courtesy Marcel Sigrist). Note the fern. plural in lillikamma epinmitim (gud.apin-tim) ipqidu in this text, 22. Deep-ploughing by oxen: AbB 4 121:12-13; 9 98:9; 10 7: 13, 15:17; Riftin 53:ll; Etudes Garelli 22 A. 4347:12-16; OBTR 137. After H.J.L. Vanstiphout, Aula Or. 2 (1984) 245 note 25 (line 108), 244 note 22 (line 91). The eleventh is the plough. The copy in Riftin 60:l suggests 8 iku. Wim Burggraaff convinced me that 8 bur is better in view of the yield (compare also TIM 5 50). This means that the signs are Winkelhaken. For the amount of land worked by one "plough" (igkarum), see B.HruSka, BSA 5 (1990) 106, 112 note 17; M.Sto1, ibidem, 198; M.Civi1, The Farmer's Instructions. A Sumerian Agricultural Manual (1994) 75. Mari: F.Joannb, ARMT 23 (1984) 111; M.Birot, ARMT 27 (1993) 11 note 30; S.Mau1, Flon'legium
Old Babylonian cattle
marianum I1 (1994) 40. It is not certain that the iku in Mari is the same surface as the iku elsewhere! Note that Birot and Maul repeat the error made by Simmons in YOS 14 (1978) p. 19: not "60 acres", but 6 b a r = 108 acres in YOS 14 300-305. A definition of idkiirum: J.-M. Durand apud Bonechi and Catagnoti, Florilegium marianum I1 73, note d. Another one by D.Charpin, Le clerge' d'Ur (1986) 476; cf. M.Civi1, The Farmer's Instructions 181. Norms in opening new land (teptitum): AbB 13 4 rev. 16 (4 bbr, 3 eie, 40 sar [= 90 iku, 40 sar!] a.Sh pe'te-e-em for two gud.apin's); Florilegium marianum I1 (1994) 38 no. 11:19-20, cf. 25 (50 iku a . 3 ba-litim te-ep-te-et PN). Note that 80 iku for 4 ploughs follows! M.Civi1, Kramer Anniversary Volume (1976) 88,93. G.R.Driver and J.C.Miles, The Babylonian Laws (1955) I1 269, and Salonen, Agricultura 63. We do not understand here Se-pi-it [email protected] B a l m u n w e ir-te-nk-ed-de,YOS 12 334:7-8. Expanding on Christiansen-Weniger as cited by Pettinato and Waetzoldt, Studia Orient. 46 (1976) 271. UCP 10 163 no. 94:20,22; Goetze, Surner 14 no. 30:9 f. Not clear is: 16gud.?$.a rqd7-du-u'-um17[e]r-biit-ma mah-ri-ja in the letter TCL 17 6:17 (coll. K.R. Veenhof): "the oxen ... being a team-of-four (?), are with me". AbB 6 199:16-18. I do not know what tiikaliitum - if read correctly (the copy also allows ta-[alb-la-tim, cf. tablittum ??) - means; cf. Farmer's Almanac, 24 I 15 bil., ed. M.Civi1 (1994) 42: ni-ir-ka i-na ta-ka[al-ti ...I. 1-en GIS.APIN4 GUD.MES2 GUD.AB.MES4 LO.[ENGAR.MES], YOS 6 103:19-20. More evidence for four oxen: M.W.Stolper, Entrepreneurs and Empire (1985) 141 f.; G. van Driel, JESHO 32 (1989) 219221. A.Shaffer, Sumerian sources of Tablet XII of the Epic of Gilgamesh (1963) 89 f., 117, lines 261-2 (Iu.ange.4.18.gim), 265-6 (l6.giS.apin.li.gim). lu.sum.ma 15 gud.apin 6.ta, Riftin 90 I 1, I1 1. Seven: [...I gud.apin 7.ta, Riftin 90 rev. 2; (total) 16 gud.apin 7.ta, rev. 9. In the plural: a-na pa-nu a-bi-im su-m~-~na~-tum i-li-ka, UCP 10 163 no. 94:6, collated by S.Greengus, Studies in Ishchali documents (1986) 161, 231 who tranlates "(For) six days the eight-spanned plows went in front of the seeder". Or is the word sumunfi "eight-year-old" (cows)? M.Birot, TEBA (1969) nos. 1-11, with p. 44 f.; Ellis, JCS 29 (1977) 127-150. Fourteen [email protected] in Ellis, 142 no. 4:24. According to M. Stol, Epilepsy in Babylonia (1993) 19, iir and egir are one and the same word. gud.h.ur.ra in CH 5242 and FLP 1287 V 45 (diss. Martha T.Roth p. 43);YOS 13 29:3, 21; 259:l; 14 321.1.14, 19; VAS 16 206. Akkadian: new refs. are: 0.0.3.3 &gal gud wa-ar-ki-i, BM 97133:4, and in the related docket BM 97939; see note 193. CH 5242-3; M.Civi1, Studies B.Landsberger (1965) 4 UM 55-21-71 col. 11.2-7. gud.egir da dipdu iSbiru gud.egir ana qabalti erdi [?I, AbB 12 177:10-11. K. van Lerberghe, Philadelphia 71: 11; YOS 13 29:3,21; YOS 14 321.1.14,11.19. We suspect that the gud.sag.da."UD.SAR" in S.D. Walters, Water for Larsa (1970) Text 25:2 (copy Plate 111 8) is a writing for "ox-of-the-middle" (cf. the discussions on Sum. iti = UDgunu = murubq ?). We cannot believe that gud.da.ri.a in a law published by M.Roth can be the "spare ox". This is a trained ox in danger of being killed by a lion; contrast the "wandering ox" in another law (JCS 32 136-8; diss. p. 229). The Sumerogram means "driven ox" (see below, under Ox-drivers, and notes 113, 114).
The word gimlum is attested in a letter where it rather should be related to the vcrh ~untcrlrtn~ ut ttr beginning; it probably has nothing to do with our ox (AbB 6 9:27 with Salonen, Agric. 385 f.; diss. Hoth 82-84; JCS 32 [I9801 137 f.). - I do not understand wa-ta-ra-am "extra" in OBTI 123:2,4. Transliteration: (1) itu ger-re-tim uq.l.kam (2) 3 gud ta-am-rda-ra7-tirnfa gud wa-x [x] x (3) i-di-n[u] (4) 4 (PI), 1 bhn ;&.gal g[ud) (?) w[a]-tara(!)-am ipba-tum. Year name. Salonen, Agric. Tafel VI; now D.M. Matthews, The Kassite Glyptic of Nippur (1992) no. 164. MSL 12 48 Proto-Lu 436-442. For gud.dbus, see W.von Soden, AHw 965b, redfi 4.b. - The man at the seed-funnel on the Nippur seal could be the 1u.numun.gar mentioned in a lexical text after the men working with oxen: "the man who lays the seed" (Proto-Lu 442). E.Sollberger, RA 74 (1980) 54 no. 120:41, sag."engarY'(= sagapin). Read in 40: e-zu-ub. A third contract is fragmentary (TLB 1 254). More: BM 97278,97466 (Els Woestenburg). CT 8 30a:3 (discussed above); AbB 8 98:35, 39. Greengus, OBTI 2619, YOS 12 64:10, cf. 179:7 (?). Ur 111: in h [= hire] 1li.hun.gh [= hireling] gud.da.ri (NATN 30.11.4, etc.; Sauren, WMAH 90:12). Old Babylonian: h [= hire] 113gud.da.ri.a (YOS 14 244:17-18); gud.an.da.ri (TLB 1 65:9). E.I. Gordon, Sumerian Proverbs (1959) 213, 245. - Also Sumerian ra, lit. "to beat"; see Sjoberg, ZA 64 (1975) 162, on line 26. S. Sanati-Muller, Baghd. Mitt. 21 (1990) 180 no. 126:7, with Charpin and Durand, MAR1 7 (1993) 373 (egir gud). A.w.sjoberg, JCS 25 (1973) 111, 116 "Der Vater und sein missratener Sohn", 77-80: "In keinem Falle habe ich dir gesagt, hinter dem Ochsen zu gehen (egir gud.da.ka du.ii), hinter meinem Ochsen herzugehen (gud.mu.S& 6s.ii.dit), habe ich dir nie als Arbeit zugewiesen, mein Feld zu pflugen, habe ich dir nie als Arbeit zugewiesen, mein Feld mit der Hacke zu bearbeiten, habe ich dir nie als Arbeit zugewiesen". - The OB passage is BM 96967 rev. 4 (0.1.0 l e a-li-ik wa-ar-ki [email protected]) (courtesy Bram Jagersma). Complete survey for aliikum warki: Friedrich, BA V/4 (1906) 511 no. 46:9 and OECT 13 263 rev. 5 (hirelings); AbB 11 7:8 ("PN was walking behind [the oxen], his ox died and they came to me for a verdict and I questioned the ox-drivers [[email protected]], his colleagues" - d i k ididu lit. "going at his side"); Arnaud, Larsa et Oueili (1983) 261 no. 1:26; Gautier, Dilbat 44 (rental of: 1gud.iir.ra liP N iilik warkidu); YOS 11 96:6 with W. Farber, Schlaf ... 42 (cf. 36). Sheep: Riftin 37:6-7 with note (rental ana kaparriitim). 18.gud(!).S&gud (TCL 1 54:18), and note ana gud.Sh.gud-tim = ana kullizbim, BM 81424 (unp.). Note in Examenstext A 26 gud.ih.gud.ra I kullizu; A.w. Sjoberg explains the element r a as "to hit"; "to drive"; ZA 64 (1975) 162. Hh II325-329 (MSL 5 76). A similar confusion exists in Izi H 251-255, MSL 13 203. The confusion with kullizu can also be detected in Ijj[lylXI11 297 (MSL 811 43): among oxen we suddenly find the person gud.Sh.ga = kul-li-zu. The matter is complicated by Hh XI11 305, gud.Sh.ga = kar-dd-nu-u "ox with a large belly [Sh]". W. von Soden, OLZ 70 460, was wrong in reading gud.Sh.gud in BBSt 7 I 20 as kardiinii. By the father: Goetze, JCS 11 (1957) 26 no. 14; BM 81424; by the mother: BM 97463. By himself: YOS 13 442. In general: M.Sto1, RlA 813-4 (1994) 172b, art. "Miete", 53.9, "Ochsenknecht".
Old Babylonian cattle AbB 6 111:6-9; YOS 13 348:6; unpub. A. 3559:4. - Children (?) in earlier times: Salonen, Agric. 324. TLB 1 251:4-6, 20 liters per day. Another text speaks of 17 liters of beer for ox-drivers (kaS maitit lii-gud, CT 2 18:22-25 with van Lerberghe, MHET I 50-1). An ox-driver received a ration of 2 liters per day (TLB 1 42: 14) and in another instance we can compute somewhat more than 4 liters of barley per day (yearly 30 gur Se.ba 6 Ili.l&gud, YOS 5 175:6). Female ox-drivers as slaves: van Lerberghe, Philadelphia 58:2,7 (10 shekels of silver, borrowed "for the price of a sag.geme ll.gudW);BE 611 68:ll ("30 shekels of silver, price of half of the price of two female ox-drivers"; strange; replacing the dead ox-driver of line 6 ? cf. the 3 1/2 oxen of line 4); CT 8 30a:6 (geme 6); cf. J.-M.Durand in Florilegium marianum IL Recueil d'ktudes ci la mkmoire de Maurice Birot (1994) 87 no. 49:24 (9 SAL 9 gud.l?;i.a e-ri-[Su-tumfi. Hirelings as ox-drivers: 2 erCn.eun(!).gi ll.gud, CT 8 30a:4; barley a-na i-di er6n.hun.gsi Iu.gud.Sii.gud, TCL 1 54: 18. Forthcoming in D.I.Owen, M.Sigrist, The Cornell Cuneiform Tablets, as no. 42. - Note 30 gur le.ba 6 lu.lii.gud, YOS 5 175:6 (= 5 kor per man as "barley allowancei'). Cf. 30 gur le.ba Iu.ll.gud, in the letter of the same group, TCL 17 1 % The Sumerogram with this reading kulliziitum was not recognised by Finkelstein in YOS 13 20,442,500. Fields of ox-drivers: Arnaud, Larsa et Oueili (1983) 286 no. 2:18., all er6n lLgud.mG n i g h PN; AbB 1 59:9, 12 iku a.ll jibit Kunzulum ll.gud (...) Sa ,fama.-bazir iStu mu.7.kam iddinuSurn; AbB 4 641213, a.ll bit abini ana Bl.gud-ma ugdammer. M. Bonechi, MARI 7 (1993) 130-131,124,27,29 (Larsa). AbB 11 189:7 (9 er6n Sa C Sl.gud uru ~ b - n i t a ~ ~ ) . VAS 7 128, 161, YOS 13 141. Discovered by S.G.Kushumikov; see Sulmu ZV. Everyday life in Ancient Near East (1993) 176. It is a problem that two other texts in the same group as Riftin 60 do not fit this picture: in YOS 5 184:3 (and in 175:6, probably) only six men work on three teams; they receive 30 kor as barley ration (le.ba) (= somewhat more that 4 liters per day). Six men is not enough for three teams, unless only two oxen go with the plough. But this is not the rule in this group of texts. It is possible that in a letter the "four men, three women" serving as ox-drivers were attached to the one plough (gud.apin) mentioned at the beginning (Goetze, Sumer 14 no. 1:16-17). My theory would expect two ploughs. According to AUAM Sigrist no. 80,4 ox-drivers work with 5 + 7 plough-teams (initum ). AbB 6 50:14, resp. Friedrich, BA V/4 (1906) 511 no. 46:8-9. In rev. 5 "six hirelings" (wages: each 6 and 7 liters daily). CT 48 32:4, Sa 12 er6n.ll.gud.md fa PN (...) la illiku iqbfi. Studia Orient. 46 [I9761 270 f., 283. Note K.Maekawa, BSA 5 (1990) 119: "According to Noor I, Obv. I 1, a team of three men plowed and harrowed a plot of 13 iku"; p. 124, "the rule of Ur I11 Umma that a 'cultivator' would direct three subordinates (ll.gud)". M.W.Stolper, Entrepreneurs and Empire (1985) 141 f. "Eight workers with twelve oxen"; according to the text 12 GUD.MES 8 LU.ENGAR.MES. ARM 9 26:19-21, "Total: 14 men, 2 women, of 1 plough, under Ibni-Amurrum". The parallel lists are nos. 25:l-17 and 27:l-34. See M.Birot, ibid., p. 332-4 8132; Ph.Talon, Iraq 45 (1983) 47, 50 f.; F.Joann&s,ARMT 23 (1984) 109-113.
139
Numbers of oxen in these texts: six in ARMT 23 nos. 113-115, ARM 24 16; eight in ARMT 23 I (M,lo7 The letter was published by C.F. Jean in 1948; now Birot, ARM 27 no. 1. - The statement "unc chnrnlc de 10 a 15 hornmes" (ARMT 23 p. 111) seems to be based on this letter and ARM 9 25-26. Cutting reeds: ARM 27 1. Threshing time: ARM 27 39, Florilegium marianum I1 no. 10. Harvesting and sowing sesame: ARM 27 38. Sesame: ARMT 13 37,39. Imrfi (Sum. mur), in Farmer's Almanac, bil. version IV 1 (= 93); see ed. M.Civil (1994) 43; and see
140
below, under "Draff'. The simple trans1ation"grass" for Sammzi seems to be a problem. CAD S/1 317b gives the Wanslation
137 138
141 142 143
144 145 146
147 148 149
"hay, fodder"; not grass. Von Soden, AHw never gives the translation "grass", nor "hay" ("Futter"; Frankena in SLB IV: "Futterkraut"). UET 5 16:19-20, with D.Charpin, Le clerge' d'Ur (1986) 461. Durand and Charpin wrote on these aspects in M.-Th. Barrelet (ed.) L'Archkologie de l'lraq (1980) 150152. Verb jabaturn: Jean, Sumer et Akkad CCII:193:7 ([email protected] Sii.gal $a-ab-tu); OBTI 120:4; ARM 27 110:22; Friedrich, BA V/4 (1906) 51 1 no. 46:2 ([email protected] duhdurug is-ba-tu). This verb is also used when human beings receive rationed food; a few examples in M.Sto1, BiOr 33 (1976) 147 note 3. Riftin 124:2-3. Read this: (2) '2 (PI)' (!) 4 b i n (= 160 liters) l e %.gal 16 gud.[N.a], (3) '2 (P1)l 1 bhn 8 sila (= 138 liters) 46 [email protected] (4) (total) 4 (PI) 5 b i n 8 sila (= 298) (5) i-na uq. 1.kam. Szlechter, TJDB (1963) 89-92 IES 336 (Samsu-iluna 34). - The price of barley in silver is here double the normal price. ARM 7 263.11.14 f., with coll. Charpin, Durand, MARI 2 (1983) 96: (14) 12 a.gitr Sa a-na ll.gal 24 [email protected] e-re-Si (15) Sa itu.4.kam i-nu-ma [email protected] [...I. Computations: 1 ox receives 112 a.ghr = 5 kor (of 120 liters each) = 600 liters. The time is four months = 120 days, and 120 x 5 liters = 600. ARMT 23 106 no. 106:1, with comm. ARMT 23 p. 62 f., no. 7 1: 15, 18. - See note 150. H.Waetzoldt, AOF 13 (1986) 4-5 no. 2 ("Eine Futtermenge von 10 sila taglich ist durchaus moglich"; "je
Ochse meist 5 oder 12 sila", as Ur I11 texts show) (ref. Wim Burggraaff). 150 UCP 10 62 with Greengus Studies 136. The "30 liters each" in UCP 10 33 with Greengus, Studies 115 has a parallel in Mari (see note 148) and cannot refer to one ox. G.R.Driver, Iraq 7 (1940) 48,51. The copies by O.Loretz, Chagar Bazar (1969), indeed show "10 liters" 151 for a breeding bull in no. 12:2 (A. 930) (1 bin, with slanting horizontal) and again for an ox, no. 22:3 (A. 95 1). 152 Survey: oxen 2 liters (Iraq 7 51 A. 952), 3 liters (Loretz, 22:4), 3 liters for "ploughing oxen" (Loretz, 26); calves in their first year (2 Iiters), second year (3 liters), third year (4 Iiters) (Loretz, 12:3-5). Cf. Bottkro, ARMT 7 p. 247 note 3. 153 H. Freydank, AOF 21 (1994) 26 ad 3, with Rollig in note 34. 154 TLB 1 285: 1 (4 b i n l e ll.gal 1gil.apin PN). 155 TLB 1 46:3,5, 14; with full discussion by M.Sto1, Studies Uon de Meyer (1994) 232-5. His explanation at the end of that article is quite different. See also below, in D, "initum = oxen". 156 TLB 1 25 1:1-3 (lit.gal6 gud.apin 1 PI 2 bin1 le.ta.h). The oxen work from 1.III until 15.V, 75 days, and 75 x 80 liters indeed is the 20 kor (6000 liters) of line 1.
'
Old Babylonian cattle
Gautier, Dilbat no. 47:3 (over month IV, days 8-30). Line 2: 5 b i n , en> sila each; line 6: 2 b i n each. Obscure. - 110 (line 6) or 100 (line 8) liters as "fodder of one plough-ox" (Sii.ga1 1 gud.apin 1 PI 514 ban(.ta), Ellis, Studies A. Sachs [I9881 137 no. 10:6, 8). Line 12 offers "fodder of oxen, 100 litres each", gii.gal [email protected] 1 PI 2 bin.ta (here .ta added) - CBS 483:2' [... SiiI.gal2 gud.apin.Mi.a ..I - moist bran ana Sl.gal [email protected], TCL 1 158:4 - barley for oxen, TLB 1 87 Individual oxen: - 50 litres, in: 5 b i n S.ii.gal gud.apin(?), YOS 14 56: 1,3 (=JCS 14 5 1 no. 75) This fodder is not to be confused with the 50 liters, in: 12 gud.apin.na 5 bin.ta, J.F.Robertson, Studies A. W. Sjoberg (1989) 460 CBS 7423:l. 4. Another text from this group offers: (1) 4 gud.apin.na x x (2) Se.numun apin.1.e. (3) 5 bln.ta. The extra line 2 shows that not fodder but the allotment seed for an ox is meant (F.R.Kraus, ZA 53 (1959) 146 Ni. 1705). gud.apin.bi 514 bin.iim, NATN 30 I1 1 I 4 Ur 111: ange.apin.bi 1 (PI).h, I1 8 gud.apin 4 b i n 5 sil8.h I 3 bin.ta, WMAH 90: 1, 5 M.A. Powell, AfO 31 (1984) 51, and note 64. R.L. Zettler, The Ur III Temple of lnanna at Nippur (1992) 277b, 6NT 225 rev. I 5-8: (5) 2 gud &gal Sii.gal uq. 1.a.bi 0.0.1.6 sila Se.ta (6) ge.bi 1.3 gur itu.l.Si! (7) 3 amar &gal uq.l.a.bi 9 sila le.ta (8) Se.bi 0.4.'3' itu.l.BR R. Frankena, SLB IV (1978) 252. - Note the letter on giving fodder to oxen that are "... and weak" (haul-lu in-Su), ARM 27 110. For eneum, "to be weak, see also DCS 100:l l : 14 (u i-ni-is'-ma PN pa-ha-at [email protected] it-ta-nu-pal). These texts belong to one group that we will not study here (YOS 5 175, 181, 184; Riftin 53, cf. 60). The following remark may suffice. The standard amount is in Riftin 53:6 and YOS 5 181:15 three kor tablittum for one team. Riftin 53 specifies that the physical improvement in this standard lasted one month; YOS 5 184:2 speaks of two months requiring 18 kor tadnintum for three teams. This means again 3 kor per team per month. In YOS 5 175:7 of 27 kor tadnintum for three months, certainly for three teams (this text has much in common with YOS 5 184 where lines 1-12 speak of three teams). - Another ref. for tablittum is TLB 1 46:19. - I do not know what tadnintum in Riftin 90 means. ARM 2 82 = ARM 27 75:28-38; ARM 14 5-6. YOS 14 214-216; D. Arnaud, Me'langes M. Birot (1985) 35-38. For this draff, see M.Sto1, BiOr 28 (1971)169-171, "Die Treber". Moist draff: AbB 2 92:22-23; 3 92 rev. 7; TCL 1 158:4; OECT 13 117, 152; Friedrich, BA V 511 no. 46:2. Dry draff: AbB 11 66:13:14. - Note "Barley for draff for fodder of the calves" (YOS 12 80). C1'. BIN 7 192:ll (duh Se-a-am u ku.babbar Sa 2 i-ni-a-tim). UET 5 103:22-23 (Se.numun ;&.gal in.nu.da duh a.kug.a.ni). - For straw (in.nu.da, tibnum), sec H, Frankena, SLB IV (1978) 37-8. CAD K 427 takes kissatum to mean "chopped straw" (together with normal straw in TCL 1 142 = VAB 5 129:16). ABIM 9:23, TCL 17 40:23.
Old Babylonian cattle
M.Sto1, "Constant factors in Old Babylonian texts on ploughing with the initum", in Cinquante-deux reflexions sur le Proche-Orient ancien offertes en hommage it Uon de Meyer (1994) 229-235. The Dictionaries summarize his results in giving these meanings for initum: 1. services of an ox team, for agricultural work; 2. rate of hire of an ox, amount to be paid (CAD IIJ [I9601 148-150); "Mietrindergespann"; also "(jeweils gangiger) Mietpreis fiir M." (AHw I, Lief. 5 [I9631 382b). - The German word "Gespann", "Rindergespann", can create a problem: are one or two animals under the yoke? The identity of initum and "oxen" (plural) (see below) could suggest more than one ox. Landsberger thinks of one animal (242 f.) and Lautner, in his review, stressed that he, too, had meant "(ein) geschirrtes Rind" by it in his book (Or. NS 7 [I9381 388 n. 2). (a): BM 97278; BM 97463; (b): PBS 812 196:14-15; BM 97543. See CAD N11 79 (nadO); AbB 7 170:17 (iniitirn ahiitim Sa ina libbi eqlim innadii); AbB 10 64: 13 (ina hamri Summa digilki initam iStiat idi). M.Stol, Studies L. de Meyer (1994) 235, end. AbB 8 98:41 (ge-am Sa gud i-ni(sic)-a-ti-Su-nu li-ik-mi-su). With the copy -ni- ; the -nu- in AbB is a misprint. A plural like alap (gud) iniatim is good Akkadian. Fractions: a parallel can be found in harvest labour contracts (that have so much in common with initum contracts; see Landsberger): some speak of 112 or even 113 harvesters; YOS 13 357:2 (1 113 er6n.Se.KIN.KUD); YOS 12 283: (7 112 e-ji-du ... mahir); Goetze, JCS 11 (1957) 28 no. 17:7 (3 erdn.Se.KIN.KUD.mei d mi-Si-il-Su). This must be related to fractional days that the harvesters are calculated to work, YOS 12 213:3 and Pinches, Berens Collection 96:3 (13 213 days). Neo-Assyrian: 9 sutu and meS-li e-ji-du in ND 2083 (communicated by R.M.Jas). M.Sto1, Studies L. de Meyer 230 f.; RIA 8 169b "Miete" $ 2 ~ . But we tend to agree with Landsberger in his interpretation of VAS 7 23 (p. 238 sub a; 244 sub c, 1): in this "Verpflichtungsschein" a man, right after having seed-ploughed (10.X), establishes his claim for the same initum for 10.VIII of the next year. This document has no witnesses and does not give a price but bears the sealing of the man who has to deliver. This is a claim; the formal contract still is to be written. The only "loan" contract (Gautier, Dilbat 55 = VAB 5 109) made Landsberger think that the low amount there (114 shekel of silver) was only a first down-payment. Moreover, the following clausula "as soon as he asks for it, he [= the borrower] shall give the one initum according to the current rate" made him think that profits could be made. We wish to point out that this is the only text with the phrase "at the current rate". Survey of both types of initum contracts: - gu.ba.an.ti "loan": BIN 7 201, OECT 13 197, Gautier, Dilbat 55, Szlechter, TJDB 118-9 MAH 16.305, Anbar, RA 72 122 no. 9, Sigrist, nos. "2, 10, 13,27, 146". - eli PN iSi2 "Verpflichtungsschein": Gautier, Dilbat 49,50, SFS 242, PBS 812 179, VAS 7 23, YOS 12 221. Szlechter 118 f. and YOS 13 48,56,59; cf. Yoffee, The economic role of the crown (1977) 96 ff. YOS 13 149, compared with YOS 12 399, etc. Text of YOS 13 149: (1) 8 112 Ba-ag-gu (2) 4 Be-el-Su-nu (3, indented) 12 112 i-ni-a-turn (4) Sa i-na a.Sh ~ e - l i - ~ u t u - S(5) i ~Si-ip-ra-am ~ i-pu-S[u] (ruling) (date). [note the masc. plural ippuSu: oxen are meant]. Compare YOS 12 399, (harvesters) Sa i-na a.Sh gu.un PN i-je-du. Szlechter, TJDB 118 f. MAH 16.305; VAS 7 87.
,
Old Babylonian cattle
These are the texts where we find both idu and iniium: - OECT 13 208. In line 1 "160 liters [of barley], price of an initum"; in lines 4-5 "30 liters PN, hire of his cow" (6.bi aibdu). - TCL 1 229. In line 6 "360 liters of barley, price [= value] of 1 112 shekel of silver, price of 3 initums of B.". His partner, the herdsman [TCL 1 168:2] E., contributed 200 liters to the price of the i. and the text adds (10): "90 liters of barley, hire of six oxen/cows" (Se i-di 6 6b.gud.w.a). - CT 45 104. In lines 18-22: "240 liters of barley, price of an initum, received by PN from PN2; 750 liters of barley, hire of the oxen" (Se Si-im i-ni-tim; Se.gur i-di al-pi). We assume that the "price" is the prepaid "Saisonvorausmiete" and that the "hire" is paid for extra animals needed at the last moment, so to say. - A problem is YOS 14 43:7 (with JCS 14 30 no. 63): "100 liters of barley, hire of an initum of [= for] sesame" (i-di i-ni-tim Sa Se.i.giS). Landsberger ,MSL 1 (1937) 233,242 (with wrong number "223"); in CAD I/J 150b (cited as "Cros Tello 195"); also VAB 5 97. Then, the (trained) ox would be worth 3 x 4 118 = 12 318 shekels; not a round figure! Similar calculations in note 188, on TLB 1 128. The main person is Beli-rim-ili. Other texts of this group are: TCL 1 11, 12, 129; Cros, NFT 192 A 0 4260, Gig, Belleten 26 (1962) 37-39,44 Ni. 13235. From Telloh. JCS 5 (1951) 79 MAH 15885:14-16; 85 MAH 15914:31-33. Cf. B.Landsberger, JCS 9 (1955) 126. Another example: Bal.a and IlSu-ibbiSu are both associated with fields worked by deep-ploughs and harrows (Riftin 128); elsewhere they jointly rent a field for cultivating sesame (YOS 12 543); the word initum is not used. TLB 1 128:13 (Sh @t.la-Su). The figure "66 213" litres, silver for pasturing (ku.babbar ri-tim) in line 4, must be a procentual share in a partnership: clearly, this is two-thirds of a round standard amount in silver (the kasap ritim referred to in two Lagaba letters; see note 60). Note "7 kor, the share" in the related round tablet, TLB 1 1215. TCL 1 229: 9-12; we cannot read in line 8 the crucial signs introducing this section: bar Su tab nig uz iS. For the problem of the "hire", see note 182. One wonders what purSumum in connection with initum in two letters means; a powerful man who is thc main partner? (AbB 6 79:20; 7 112:19). BM 96967 rev. 14-7: (4) 0.1.0 Se a-li-ik wa-ar-ki [email protected] (5) 0.0.1 dumu-ki u tap-pi-Su (6) 0.0.2 ki-ir-ri i-ni-a-at Si-mi (7) 5 sila dutu-ni-Su 2 tap-pi-Su. Cited by courtesy of Bram Jagersma. AbB 12 8, cf. no. 72 (aSSum gud.apin.hi.a nadinimma tapputika alikim). Another example of partnership is perhaps "Seven oxen of one plough (7 [email protected] Sa 1 gud.apin)" are entrusted to two persons (PBS 812 189). Are they just ox-drivers, or partners ? BM 97133: 1 and lines 6-7 in the related docket BM 97939. Full transcription of the docket BM 97939 (by Bram Jagersma): (1) 0.0.3 3 sila Se (2) :&.gal gud (3) wcrar-ki-i (4) 0.0.1 5 sila Se (5) a-na m6S.u.a (6) 18 i-ni-a-tum (7) ta-a~-~i-it.'(text: IK; collated)-ti (8) 0.2.0 ge mu-ki-il(9) ha-ar-bi (10) 0.0.2 4 sila zid.da (1 1) ku-ru-ma-sli-nu. First lines of BM 97133 (by K.R.Veenhof): (1) 18 i-ni-a-at ta-al-pi-it-ti (2) 0.2 Se mu-ki-il ha-ar-bi-im (3) 0.0.2 4 sila Suku 1u.meS. See now M.Civi1, The Farmer's Instructions (1994) 69 and 100 note 9.
Old Babylonian cattle
195
196 197
198
On kirrum, see F.R.Kraus, JEOL 16 (1959-62) 24 f.; cf. B.Landsberger, Symbolae M.David I1 (1968) 7678; K.Butz, OLA 5 (1979) 381 note 339; CAD K 410a, kirm A. -New refs. are AbB 2 157:4 (opening a field), Birot, TEBA 17:12, 19:6 (excavating earth), 20:4-5 (reed and straw), TLB 1 160:17 (ship); also AbB 12 148:20 (ki-ra-am Sa a3 Sa ti), van Lerberghe, Philadelphia 22:13 (it-ti 30 er6n ki-ir-'ri'). AbB 5 205 with Kraus, 23 f.; also no. 94 and AbB 8 118. It is striking that contracts state four times that the initums have to start working on the tenth day of months VI, VII and VIII (10.VI: Sigrist "13":8 and "2":9; 10.VII: BIN 7 192:l; 10.VIII: VAS 7 23:6). Does the tenth day traditionally inaugurate a new agricultural activity or is this just coincidence ? In all cases seed-ploughing (eriium) is involved. Does the unusual Sapikum in the line "let the barley be given and let the initums be poured out" (AbB 12 16:20) have to do with this ceremony?
,
-.
CATTLE IN THE NEO-BABYLONIAN PERIOD* G. van Driel (Leiden)
'
1 Introduction As always when dealing with Neo-Babylonian animal husbandry, the administrations of the temples in Uruk and Sippar provide the most evidence. Some additional information can be gleaned from the MuraSu archive, but otherwise, the contribution of the family archives is negligible. In general, the information about cattle is less comprehensive than that concerning sheep, though much of the relevant material still awaits study. The most important difference is that for sheep husbandry production norms are known, but not for cattle breeding. This seems to be the case for all periods. Archaeozoology has hardly contributed to our understanding of animal husbandry in this period and only textual evidence is available. As in other periods, cattle are of prime economic importance for two reasons. They provide traction for the plough and are bred for meat, in particular for the offerings to the gods. From the texts milk would not seem to be an important product and there are no clear indications that oxen were used for transport in this period, though this is certainly a result of the one sidedness of our sources. This contribution aims to collect the available information on the two main topics relating to cattle in the Neo-Babylonian period. There are a number of previous studies of importance, especially concerning terminology. For Uruk San Nicolb's contribution dealing with YOS 6,1301 is basic. Much information can also be gleaned from Landsberger's studies on animal terminology,2 which are not only important for the Neo-Babylonian period, but cf. also Gehlken, Uruk, Wirtschaftstexte Z(1990) p. 42f, especially note 41 for remarks of basic importance. Yet many problems remain. In general the (local) Neo-Babylonian readings of the logograms used are not always known. Different sections of the economy each have their own (written) terminology, adapted to their own needs, but, certainly, also overlapping with that used by others. The terminology used for the administration of cattle herds differs from that used for the administration of offerings, in Sippar more markedly so than in Uruk. It is perhaps useful to differentiate between three fields of activity: a) cattle herding, b) cattle in the sphere of arable agriculture and c) offerings of cattle. Terminology deals with gender, age and training, to which the complication caused by castration must be added, especially in its religious ramifications.
*This study was made possible by the liberal attitude of the Trustees of the British Museum, represented by the Department of Western Asiatic Antiquities which has allowed the quotations from the unpublished texts in the Museum. NWO, the Netherlands Organisation for Scientific Research, granted the means for a visit of a month to the British Museum. The Vakgroep Assyriologie, RU Leiden, contributed towards an earlier stay. Bull. Sumerian Agric. 8 (1995)
215
,
van Driel
Cattle in the Neo-Babylonian period
2 Sources: differences between Sippar and Uruk a. Sippar The difference in emphasis between the information from Sippar and that from Uruk that was apparent when dealing with sheep is also in evidence where cattle are concerned. In Sippar the day to day administration is better documented, while in Uruk there are more texts dealing with "management". This is no doubt the result of the arbitrary way in which the evidence has come down to us, in the present case it is perhaps also the result of the manner in which the texts have been selected for publication. In Sippar we have very many fragments of tabulated documents representing the results of temple herd accounting and only very few of the small one-herd inspection texts. As far as can be ascertained in the present state of publication the tabulated texts represent several types of texts, as is the case with sheep.3 Of great importance for the understanding of animal husbandry are those texts which give the results of inspections, amirtu, of cattle herds in which the various categories of animals are differentiated for bureaucratic purposes. These can be one-herd lists or tabulated documents listing several herds. High numbers in such a text, no doubt, indicate cattle kept in a herd for breeding purposes, low numbers suggest, in general, plough teams, unless they sum up smaller herds towards the end of lists beginning with more important units. As the plough teams had also to keep up their own numbers to the required level, they also contained cows and young animals. The cows were, like the oxen, used for pl~ughing.~ Another important aspect of Sippar cattle husbandry seems to be that the institution had not enough animals fit for the purposes for which they were required. Especially not, it would seem for the constant drain on the resources caused by the daily offerings, on which there is a particularly rich, but incompletely available, file. Cattle appear to have been bought by the Sippar temple for silver through specialised traders. A small number of texts deals with the attempts to keep plough teams up to strength.
b. Uruk The same types of texts probably existed in Uruk. For the moment there are more one-herd inspection texts from Eanna than from Sippar, but there are only very few tabulated documents concerning cattle from Uruk. Published commercial transactions involving cattle are equally rare, but, conforming with the prevalence of the management aspect in the published Uruk material, there is much concerning the disputes caused by accounting and (presumed) embezzlement involving cattle. The Uruk "general farm" documents are of special importance for the organisation of plough teams. This last type of evidence is somewhat supplemented by the Muragu texts, which deal with the Nippur neighbourhood.
3 Organisation After what has been said about the organisation of sheep husbandry, cf. the contribution to BSA 7, little can be added about that of cattle breeding. There are no herding contracts, only a few documents dealing with single animals, which are of little help to us, especially as there are additional financial arrangements in these texts. We can only assume that conditions similar to what can be found for shepherds prevailed and that herdsmen had an interest in the herds they cared for. If the institutions needed the male animals they will have been obliged to cede the females to the herdsmen, as was the case with the sheep. But circumstances do differ in several important respects. Cattle breeding did not take place at great distances from the seat of the institutions. The few localities that are mentioned in the available texts are well-known places in the neighbourhood where the institutions also had
van Driel
Cattle in the Neo-Babylonian period
other agricultural interests. The opportunities for closer control were much better for cattle than for sheep. There is no reason to assume that herds of cattle were kept in the far North, as was the case with the sheep. Perhaps ecological circumstances there were far less suitable for cattle, though the presence of wild cattle there is well known in Assyrian times at the latest. Cattle herding on the Habur in the Middle Assyrian period is relatively well-documented now. The institutions obtained cattle primarily from their own herdsmen, whether entrepreneur or simple employee, niiqidu or re'ii, grouped under a rub biili. He supervised both flocks of sheep and herds of cattle. Either type of animal was used for sacrifice, sometimes sold or, more often, taken to the bit urii, where the B.GUD, "cattle-house", constituted in both Sippar and Uruk a special department. Occasionally an B.GUD of the king is mentioned. In this type of establishment cattle, in the main male animals, were fed on grain and prepared for sacrifice, which in the end meant human consumption. On breeding for the plough teams see below the separate section dealing with this question. Gifts of cattle to the temples, especially by the king, were insufficient to supplement a yield which was unable to fill the gap between production and requirements. It should be noted, that, though there are occasional sales of cattle by the organisations, their needs obliged them to buy in animals, both sheep and cattle. Especially in Sippar there seems to have been a constant shortage, which may help to explain certain differences between Sippar and Uruk herd inspection texts: culling took place earlier in Sippar than in Uruk. There was a careful administration of dead animals and of hides and tendons. In Sippar the occasional sale of dead animals was a source of income, but most carcasses were probably fed to the lower echelons of the personnel. It must be assumed that herdsmen were obliged to deliver at least the hide and tendons of dead animals, as was the case with sheep. But there are no texts which spell out the obligations of the herdsmen towards the owners of the herds as are available for sheep herding. This makes it much more difficult to reconstruct the aims and policies from the texts available. 4 Terminology: general The administrative texts from both Sippar and Uruk use in the main a standardised, in some respects local, terminology written in logograms, the Akkadian reading of which is often problematic. Even if the Akkadian word is known its exact meaning is not always obvious. It is more than likely that the written administrative terminology is only a bleak reflection of the vocabulary in common use. The terminology is not the same in the various departments of the same administration and in Sippar the differences seem to be more marked than in Uruk. The relation between written administrative terminology and the spoken word is lamentably weak. Nobody would call a young bull calf a NINDA2, but is it really a birlij ? And is a birli in practice (in Uruk) up to one year old, or are (in Sippar) other male animals in their second or third year also called birii because they are regarded as not fully grown? The written terminology can be understood in this case, but spoken language will have been far richer. The word for cattle in general in the Neo-Babylonian period is a problem. AB.GUD.~I.A is easily understood and generally used in Uruk, but less so in Sippar. The reading litu, CAD L p. 217 and AHw p. 557R, cf. 558L, is not confinned by Neo-Babylonian practical texts, which always use logograms.6 The practical vocabulary from Assur, AfO 18 (1957-8) p. 331 gives sagullu, which would be sug/kullu in the South, for AB.GUD.~I.A, but CAD S p. 345 can only quote YOS 3, 9:36, context not clear. Nbn 477:ll does not relate to cattle. In BM 67462
,
"
van Driel
Cattle in the Neo-Babylonian period
(27.IV.Nbn.8) we find in line 1 ~%IPAD.MES $6 su-kul-lum $h a-na G I S SUM-na, ~ ~ ~ "herdsmen of the herds (of Sama~)destined for service as bowmen". The word primarily means "herd of animals", it therefore could be used for cattle. We will return to the question when we discuss Uruk and Sippar terminology in detail. Cows in calf are not, it would seem, listed separately from other cows in this period.7 Nor is there clear evidence about special breeds but for the drawing of the humped bovine on the tablet published by Pinches in AJA 8 (1893) p.190f. Especially the private documents often use the adjective samdsiindu "red", CAD S p.l28R, in describing cattle, especially cows. This could, perhaps indicate a favoured breed. 5 An early text: terminology and meaning There is one early Neo-Babylonian text, quoted by both dictionaries, which, though without known background, deserves mention, not only because it can be used to demonstrate the problems of further interpretation, but also because it seeas to indicate that the differences between the administrative terminology of herding and offering developed later on in the NeoBabylonian period. The text is BRA4 I 3, 19.VIII.Nabu-gum-iFkun.13. The collective term for ma-li-da-ti, but remarkably enough, "cattle" used is AB.GUD.HI.A. There are 30 AB.GAL.MES one of them (ina libbi) is a GUD pu-@-lum, a bull. This will be a case of awkward formulation male animals in by the scribe. There are 5 GUD.NINDA2.MESMU.3.Khi and 5 ditto MU.~.KAM, ~ ~ . 2 . K hfemales i, in their second year, in their third and second years, and 6 AB.NIGIN.MES which should refer to all animals mentioned previously, including the total 46 AB.GUD.FJ[I.A, MU.[AN.NA],ina libbi 9 NITA~.M[E~], 17 calves, bull. There follow 17 GUD.AMAR.MES.TUR.MES among which nine males, in total 63 AB.GuD.M.A. The text is remarkable, and not only because it comes from an otherwise badly documented period. Bulls are never counted with cows in later texts, and certainly not with those of proven fertility, as malittu must imply. But of even greater importance is the fact that the text is, as far as I am aware, the only Neo-Babylonian text which lists the male animals in their third year separately, but not the females. Having calved, heifers are added to the cows, and, perhaps, those that have not calved are removed. In Sippar the three-year-olds of both sexes are normally listed separately in inspection texts. But we will see that in Uruk three-year old females, like their male counterparts, will be registered separately only as an administrative innovation during the reign of Nabonidus. Evidence is insufficient to accept that the innovation represents a conscious attempt in Uruk at preventing third year females from breeding, in order to gain more strength, whereas this was ordinary practice in Sippar. The possibility that second year animals were kept separate for that purpose remains, however, as they are always listed separately. An experiment with third year animals is not completely excluded therefore. There is only one bull, puhilu, but five third- and five second-year male animals are kept. They were almost certainly eliminated on becoming older. That third-year animals were capable of breeding would seem certain, though we know little about the circumstances in which they were kept, which are obviously of importance in this case. We will see that there are clear differences between Uruk and Sippar regarding the age at which young bulls are withdrawn from the breeding herds. Another problem is that of the productivity of the herds. As we have no herding contracts for cattle, we can at most speculate. The casual observation that, with 28, the number of animals in their second and first year together almost equals the number of cows (29) and
van Driel
Cattle in the Neo-Babylonian period
that when counted over the two years, there are 14 male and 14 female calves might suggest a productivity of one calf per COW in two years, were it not for the fact that losses and culling would be excluded in this manner, which is unlikely. As there is only one full-grown bull in the herd, it seems reasonable to suppose that young males were culled at the latest in their third or early fourth year. This makes sense, in that it is the most economic manner of providing meat. But is there also an (unformulated) policy which meant that, ultimately, each year one quarter of the number of cows were retained in female calves? Even if this is the case we still have a poor idea about productivity, but it would tell something about replacement policy. Using an isolated text for the extrapolation of norms is unacceptable, but it makes some sense to consider a number of alternatives. No norms for cattle seem to be present in the available texts. It can only be noted, cf. below, that among several texts from Uruk dealing with this matter, only one seems to regard a number of cows equalling 25% of that of the oxen as sufficient to keep the plough teams in full strength, all other texts mention 50%. Not that this solves our problem, as we do not know how long the working life of a plough ox lasted. We are equally uninformed on the total number of calves expected from a cow. The difference between 25 and 50% indicates that in the same period of Babylonian history expectations regarding productivity varied considerably in the same managerial circle. As the texts, which will be referred to in more detail, belong to the system of the "general farms", we can expect a certain degree of inbuilt lining of certain pockets through the contracts, which makes them less suitable for our purposes. Yet the texts remain the only basis for speculation and we must continue to question them. A pile of nicely edited texts does not mean in itself understanding ancient animal husbandry.
6 Uruk terminology The Uruk material on cattle will be increased immeasurably by the two next instalments of E. Gehlken's texts from the 12th and 13fi campaigns at Uruk, which will probably render many of the observations here obsolete. As with the sheep and goats the most important texts are those dealing with the amirtu, inspection, of the breeding herds. Cattle in general are AB.GUD.M.A, with, as we have seen, an uncertain Akkadian reading. The texts available possibly suggest a certain change in policy, marked by the introduction of separate categories for animals in their third year, but the material is insufficient for a precise date during the reign of Nabonidus. As is to be expected, the texts differentiate between male and female, but age categories vary. Full-grown are pubalu "bull" and AB.GAL, COW".^ In their second year are the GUD 2.MEl-il-u (male) and the AB 2.m-ti (female). The male and the female calf are the GUD.NINDA2 and the AB.NIGIN,possibly biru and siibirtu, though these wo~dsare hardly used in the available (practical) texts. NINDA2 in the combination GUD.NINDA2 is, in Uruk, exclusively the male calf. Somewhere in the reign of Nabonidus, cf. YOS 6, 243 (3.VIII.Nbn.8, a text dealing with dead animals), a new category is introduced: "in the third year", GUD 3-i/u and AB M ME/turn/-ti, though already an earlier text like GC I 195 (7.II.Nbk 41) knows in Uruk of third-year animals as a separate category. Texts with and without the third-year category clearly overlap in time, and the category is known both in earlier periods as well as in contemporary Sippar. This makes it difficult to suggest a reason for the change.9 But it must be noted that inclusion of a third-year category makes good sense from an administrative point of view. Withdrawal in the third year, at least
,
van Driel
Cattle in the Nw-Babylonian period
for male animals, is perfectly understandable if meat yield is considered or if the moment for transfer to a plough team is of importance. A third-year category for females must be the result of systematisation, unless it became the moment when the fate of animals that had remained barren was decided. Another possibility, preventing breeding in order to allow further growth, has already been considered, but is entirely speculative. Several of the texts summarised in appendices I and I1 include statements on the presence among the males counted of animals qualified as KO, ellu, literally "pure".lo This KO can, in all likelihood be compared with the Sippar Suklulu, "unblemished", which is apparently used in a similar manner, but which only rarely appears in the Uruk material.11 If some of the male animals are designated KOin certain texts (YOS 6,118 and YOS 6, 130, for instance), that inevitably means that the others are not. We cannot be absolutely certain that the terms KO, used in Uruk, and Su.Du7 or Suklulu occurring in Sippar, are completely identical. As SU.DU~ is not totally unknown in Uruk, s'uklulu is therefore probably not the Uruk reading for KO. But KO and Zuklulu both probably indicate animals suitable for offerings. The most important problem is whether KO and/or Zuklulu preclude castration. We will see that castration does not automatically imply complete unsuitability for sacrifice. But it is inherently unlikely that someone would offer a non-KOanimal to the gods. Only in Uruk does the epithet KO seem to be used for full-grown male animals. It never occurs for females, which are neither suitable for sacrifice, nor spayed, normally, during their productive life. In YOS 6,118 only 2 out of 8 pu&iiluare KO and in the two first herds of YOS 6, 130 respectively 5 out of 10 and x out of 16. The number of cows in the three cases is 144, 37 and 270, which makes conclusions based on the bull to cow ratio hazardous, especially as the young bulls were certainly capable of mating. The question is whether or non-KO stands in some relation to the problem of castration. From the administrative point of view it would make sense to keep track of the number of animals potentially available for sacrifice. It is perhaps likely that herds in earlier herd inspection texts from Uruk which do not differentiate between KO and non-KO still contained both categories. The matter is perhaps clarified in an unexpected manner by Gehlken I 48, a text listing cattle withdrawn towards the end of month VIII or at the beginning (?) of month IX Nbk 35 from herds managed by niiqidu's. The text differentiates between bulls (puklu), young bulls in their second year, and, numerically unimportant, GUD.NINDA2, which here will mean "bull calf'. The bulls and young bulls in their second year are divided in groups called K0.mti ta-mi-me-e (cf. line 10) and Zen-di-tu. This suggests that ellu and being branded with a mark are mutually exclusive. It could mean that branded animals had become unsuitable for sacrifice. tamimu is, if Gehlken I 48 is correctly interpreted, only an amplification of KO. This is in a certain sense confirmed by YOS 17,55 (9.I.Nbk.19) which summarises 1 GUD ummdnu, 7 &.GAL.MES, 3 G U D . ~ and S 2 AB.~.MES as PAP 13 ~ . G U D . @ . ta-mi-ma-[ta], A as "in total 13 (head of) cattle in perfect condition". The qualification perhaps turns up again, for a cow, in FLP 1482, Stolper, BaM 21 (1990) p. 588-9, a private text from Sippar, line 2, read probably ta-mi-im!-ma-rum. But as suggested by the ummhu, "trained plough ox" an animal which was probably not suitable for sacrifice, tamimu must indicate not cultic but ordinary healthy perfection. That is probably the reason why it can be used for female animals, which are to the best of our knowledge in general unfit for sacrifice. We can thus eliminate this word from the discussion.
van Driel
Cattle in the Nw-Babylonian period
CAD E (1958) sub ellu ,translates ~ 0 . m when said ofpuhalu, (e.g. YOS 3,41:23) as "pure ones", and avoids the question of castrated or not. B (1965), sub biru p. 266R, adds "(i.e. castrated)". This is after Landsberger's discussion in MSL 811 p. 66ff. Though it was, as we will see, not impossible to use (in Sippar) castrated animals for sacrifice, this was only done for lower deities: Suklulu animals went to the more important gods. There is, it would seem no basis at all for the equation of ellu or Suklulu with "castrated". KO can of course indicate other aspects of perfection than non-castration or even the lack of branding. Branding, damage to horns, hoofs or even tail need not impair the function of a proper bull. Yet Old-Babylonian contracts and Neo-Babylonian rituals still attach significance to completeness on these points.12, If KO primarily refers to completeness in this respect we can stick to the notion that puhiilu indicates the full-grown breeding bull. One might expect that this Akkadian word had an unambiguous meaning. This is exactly the reason why the word is not used in the plough team texts: the castrated plough ox is unfit for the prime task of the breeding bull. Their number in the herding texts is usually low, especially in Sippar, even though the ratio to the cows is strikingly variable. There is, it would seem no direct proof that castrated, i.e. taptiru animals were used as offerings for the gods in Uruk. The word occurs in Uruk in GC I 166:4, in YOS 17 48 and in BIN I 144,'3 cf. also GC I1 62. In the first text three of these animals are paralleled by 33 GUD.MES.The text deals with fodder barley, which could suggest that the animals were prepared for sacrifice, and not for the plough, where castration would be required. We will return to the question in the context of the Sippar Lists of offerings. The meaning of all words depends on the context in which they are used. A GUD is "pure" when used for sacrifice, but that could be the fate of a castrated animal, which therefore is not non-KO. Placed in front of the plough the GUD had to be castrated, and it had to be an ummdnu. That this animal is a taptiru is either self-evident or irrelevant and such animals are not qualified as KO or Suklulu. In a breeding herd a GUD had to be a pu&iilu,and though not all of these were KZI, some were. In YOS 17,346 (20.12.Nbk.10) we find NINDA2.K0 as a sacrificial animal, listed before similarly used simple NINDA2. If the omission of KO is intentional it was not strictly essential that an animal is qualified as KZI in order to serve as an offering, strange though that may sound, though one of the problems with administrative documents is that repetitive epithets can be left out. Yet the omission could be intentional. In that case KO is optional, perhaps desirable, but non-KZI animals are not strictly forbidden or tolerated in practice. The Uruk cattle husbandry texts provide us with a few more terms. Old animals are called s'ibu, cf. GC I1 15 (12.V.Nbp.7) 3 AB.MES s'i-bi-e-ti, but they are not a separate category in inspection texts. Cf., perhaps, SAL.MAH (=AL), discussed in the contribution on NeoBabylonian sheep, which also occurs in Uruk for cows in YOS 6, 58 (Nbn yrs 1-6), lines 5 and 7, silver SAM&.GAL. SAL.AL(!).This could confirm what was observed regarding NeoBabylonian sheep: SAL.AL females are sold or disposed of, they are "too old", "barren". If this is correct the 34 AB.GAL.sAL.AL in Gehlken I 48:5 are old animals withdrawn from the herds. The herd inspection texts differentiate as to age between calves, two- and three-year olds and full-growns. But the factual age of individual head of cattle was known. UCP 911 no.9 lists GUD 8-i.ME, a GUD 5-ti urn-man-nu occurs in YOS 17,51 and simple GUD 5-i .MES are found in GC I1 328, Mich. 35 knows of GUD 4-i.ME. NINDAz seems to be reserved for calves in Uruk, they are occasionally classified as KUD.DA or GA, "weaned" or "suckling": UCP 911, 12. The
"
van Driel
Cattle in the Neo-Babylonian period
use of G u D . for ~ male calves in NBC 4883 instead of GUD.NINDA2, cf. the texts in Appendix I, is aberrant. Is it really in the text?
7 Sippar terminology Sippar usage differs from that of Uruk,but this is probably more a question of terminology than of substance, though it must be conceded that Sippar herds, as far as is known, all are of medium or small size compared to the Uruk ones. In Sippar cattle in general are indicated by a whole range of terms, which in essence must mean the same. In the fragments of tabulated cattle texts we find in the first line of the headings G ~ . M E SBM , 50159 (15.II.Nbp.12) or GUD.M.A, BM 50136, (M.Nbp.[x]), AB.GUD.~I.A,BM 54055, (x.IX.Nbp 12) and also GUD.AB.MES is used: BM 61899, (21.VIII.Dar.6) and BM 61941, (9.VIII.Nbp.12). AB.GAL.MESis used in the same sense in BM 6021 1 (Nbp.7-8). CT 55,662 (2.[x].Nbn.3) seems to be the only text to use an Akkadian word: GUD bura-a-ni. As the text also includes full-grown and three- and two-year olds and females this u )used in practical texts in this period. cannot mean "young male animals". A M A R ( = ~ ~is~not The Akkadian word for cattle in general is not with certainty known in Sippar. But also in Dutch and English "cows" often has the meaning "cattle". In the Uruk tabulated texts the animals are primarily arranged according to age groups, in Sippar the sexes are arranged together. This difference helps in establishing the origins of the texts. In Sippar males are pu-bal, 3-li,2-li and A.MU, females are a-lit-tu, 3-ti/tum, 2-ti/tum and TUR.SAL.MU, full grown, in their third or second year and calf. These terms are used in herd inspection texts, tabulated or not. alittu for cows is, however, apparently not used, outside these texts, it is not used as a word for "cow" in general14 and the age qualifications for second and third year animals are not used without further specification in other texts. In Sippar the herding texts use a terminology which differs from that of the B.GUD and the offerings. Not only alittu but also p u k l u is not used outside the herding texts, which in turn avoid NINDA2 and NIGIN. pubalu understandably enough, does not occur normally in texts dealing with plough teams, which use ummhnu or simply GUD. The only term which connects herding and plough team texts on the one hand and offering texts on the other is s'uklulu. For terminology BM 55446 is of particular interest as it constitutes an exception. GAL a-lit-tum, "full-grown cow (that has calved)", GUD.NINDA~.ME$ bu-mu-s'e-e, "young s'ul-lu-s'e-e-ti and s'u-nu-a-ta, "three- and two-year old bulls in their fifth year", AB.NIGIN.MES id-lu-s'u-li and iu-nu-'u-ti, "three- and two-year old bulls", heifers", GUD.NINDA~(.ME~) AB.NIGIN.MES.TUR.SAL ;at-ti, "female calf in its first year" and TUR s'a-a-ti "male calf in its first year", are listed in a fairly disorganised manner. There are, it would seem, no full-grown breeding bulls, but the unparalleled five-year olds would have been able to serve. There is also no four-year old bull, only one three-year old besides several of the two younger categories.15 In Sippar NINDA2 is not only used for bull calves but with a further age qualification also for other young bulls. The NINDA2 in his fith year is clearly an exception, three- and twoyear olds are, however, common. In Nbn 202 the temple buys a NINDA2 4-24 a 3-u' and a 2-li. It seems clear that in Sippar NINDA2 is primarily an age category. It seems to alternate with GUD plus indication of the age, cf. the list in section 12, below. Landsberger, MSL 8/1 p.71, has suggested that NINDA2 (and tap?iru) could be opposites of s'uklulu, and (therefore) could mean "castrated". This interpretation is primarily based on the regular occurrence of GUD s'uklulu as the preferred sacrificial animal for the main gods, paralleled by NINDA2 for less important
van Driel
Cattle in the Neo-Babylonian period
immortals in a long file of Sippar texts listing for every day of the month the animals slaughtered, or occasionally dead animals assigned. SarnaS and Ajja are regularly receiving a GUD Suklulu, whereas Bunene gets a NINDA2 without further specification. But Nbk 213 (20.IV.Nbk.30) differs in that Bunene is assigned a GUD tap-ti-ri. In CT 44,7153 the word is shortened to Gmtap, which is replaced in the similarly constructed YOS 17,313, VS 6,26 and 32 by NINDA2. There are sound reasons for regarding this parallelism of tapgiru and NINDA2 as fortuitous. CT 55, 685:l mentions a NINDA2 ;uk-lu-lu TUR.MU.AN.NA. CT 55,697 (3.[x].Nbn.6) lists l 6 GUD s'uk-lu-lu 4-i 2 GUD Suk-lu-lu 3-i 21 GUD Suk-lu-lu 2-ti 3~~~ 6 GUD iuk-lu-lu,4 ,, GUD.NINDA2. Though this, or the copy, is certainly wrong as to the addition, it can be interpreted together with the next passage : 4x (=2?) GUD tap-ti-ri 5 [ ~ 12(?) ~ ~GUD.MES ] etc. Altogether there are 10 animals of the Suk-lu-lu category, but the young ones appear in the summation as NINDA2, they are not grouped with the additional taptiri. In BRM I 91 (Nbn 1314) we find up to line 10 GUD.A.MU, after that, up to line 22 GUD.SU.DU~ A.MU and in line 23 GUD.NINDA~.MES.A.MU.MES which suggests that not all NINDA2 are Suklulu. CT 57,439 may have, the text is damaged, a [GuD].NINDA~tap-ti-ri. NINDA2 is primarily an age category, it would seem. Of the Suklulu animals the categories 3-24 2-ti and A.MU are mentioned regularly, 4-ti is rarer: CT 55,669 (7.IV.Nbn.3), Cyr 330 (20.[x].Cyr.8) and the texts in the list in section 12, below. For the taptiru's no ages seem to be given, but they occur only sporadically in the texts.16 A possible exception is CT 57,439, which seems to provide us with a GUD.NINDA2 tap-ti-[ruf and so could confirm that NINDA2 and taptiru are not identical. But there is more. In the offering text BM 6541 1 we find two columns parallel to the left edge of the tablet. The first is for Suklulu, the second for NINDA2. The Suklulu for Sarna~ and Ajja are duly marked, but for Bunene we find, written out completely, 1 tap-ti-ru, suggesting that NINDA2 and tapfiru are not identical. In BM 64728 there are GUD Suklulu for SamaS and Ajja, a GUD NINDA2 for Bunene, s'uklulu A.MU9s for the ziqqurrat and a NINDA2.A.MU for IStar of Akkad. In 99808 we have again the type of text with two columns along the left edge, but now with headings for Suklulu's and for tapfiru's. The text is broken and we cannot see where the animals went. These lists of expenditure (a better name than offering lists), vary as to the number of columns from day to day. Poultry is often included. Therefore we can safely conclude that terms like GUD Suklulu, NINDA2, taptiru and others occurring in these texts each have a separate meaning, and that all these animals could be used for offerings, perhaps with certain restrictions as to the deity concerned. Castrated animals, taptiru's, could be used and NINDA2's not qualified as ~ukluluas well. There are, however to my knowledge no lists with separate parallel columns for NWDA2 and for [email protected] is not completely solved. Uruk and Sippar usage of NINDA2 must be kept separate. In the herd inspection texts from Sippar, tabulated or not, there are not many full-grown bulls. In those cases in which only of the A.MU category a certain number of Suklulu's is given we must assume that the others were consciously classified as non-s'uklulu, i.e. possibly castrated. But as castrated animals could be used, perhaps under special circumstances, for offerings, other disqualifying factors may well have obtained. There is no firm reason why Suklulu refers to (non-)castration, as the word will have included completeness as to hoofs, horns and the like. But on the other hand there is no reason why young animals should be damaged in these respects at an early age. The possibility that the categorisation is the result of
van Driel
Cattle in the Neo-Babylonian period
branding certainly remains. As tapfiru animals do not seem to have been held in Sippar breeding herds we can perhaps exclude the idea that castration is the main factor determining the classification as Suklulu of animals in these herds. There is nothing to exclude colouring or markings from playing a role. Remains the fact that no female equivalent of Suklulu seems to occur. The qualification is clearly gender related. The problem of Suklulu versus ellu noted in the Uruk texts is not raised by the Sippar texts, which do not use KO,but it is slightly complicated by BM 77549, which to all probability comes from Sippar. The text deals with [x] ME 17 GUD [za-ku-ti] bought and transferred. zakri probably belongs to the same field of meaning as ellu and Suklulu indicating cultic suitability. BM 55446, quoted above, calls a full-grown cow an AB.GAL a-lit-tu, indicating that proven fertility is essential for inclusion in the category. Younger female cattle were indicated by AB.NIGIN followed by the year. The Akkadian for AB.NIGIN is rarely written in full in Sippar: sabirtu. RT 19, p.110-111, no VIII, not belonging to the temple administration, but from Sippar, uses s e i r t u for a newly born animal: 1-it GUD sa-bir-turn hina bit alpi maldata ....... GUD sa-bir-turn bi-si-il-ti i-narn-din, "a female calf born in the cow shed ...... (in month VII,) he will give a weaned female calf". The text is dated to month 11, cf. also CT 55, 665 (23.XI.Nbn. 10) lines 4 and 11.I7 That female animals up to their third year were counted as young animals is remarkable. A certain systematism, the need to parallel the male three-year olds, might be the reason. It is also possible, beside what has been suggested above as a possible explanation that the animals were kept up to this age, and only then, when barrenness was beyond doubt, eliminated. It must, however, be conceded that we know little about Babylonian expectations in this respect. Feeding certainly influences fertility. In Nbk 398 2-ta 's and 3-ta 's are counted together as one category. In texts dealing with plough teams cattle with the additional remark hniri always seem to indicate female animals past their third year. I find no direct evidence for separate herds of castrated animals kept for fattening, but the re'u gin2 in Sippar may have been responsible, as was the case in Uruk, for such animals. If my reading of line 5 of Nbk 74 (BM 60152; 15.11.Nbk.9) is correct both Suklulu and tapfiru were kept in the E 2-ru-J which was the "internal" fattening establishment of the organisation.l 8
8 Plough animals A trained plough-ox is a GUD umrndnu in both Uruk and Sippar. Their numbers were supplemented by trained cows.19 BM 74482 (13.V.Camb.5) mentions x bu-ur-turn Sd-ad-diturn alongside the GUD urn-ma-nu of a certain person. In the institutions these animals were subject to special inspections, of which BM 74464 is a good, though damaged example. A had two urnrndnu oxen and two $4-di-turn.~~S cows, and ploughman called Sama~-~ate-$abit at least an AB.NIGIN in its second year and a Suklulu in its first year. Among the six head of cattle at the disposal of another ploughman were two Sadditu. The tabulated text Cyr 117 (26.IV.Cyr.3) is probably representative. The ploughmen in this text have, in general, no more than two urnrndnu 's, even fewer occur, but these are then supplemented by Sadditurn, 3-i-turn, 2-i-turn and TUR.SAL.MU7s.The younger categories, of course, did not draw ploughs. The teams are completed by the use of trained cows, and among the younger animals of up to three years old there were only females. A complete team of four
van Driel
Cattle in the Neo-Babylonian period
full-grown oxen, as the general farm documents (mainly from U r ~ k ) 2cause ~ us to expect for institutional properties, is rare in the texts. Only one of the 13 ploughmen, the I%AL ~~SAPIN.MES, the leader of a group of ploughmen, has a full team of four, or rather, in this case, even a spare animal, and only two others have enough trained "male" and female animals to make up a team of four, one of them being a leader of another group of teams. The others have no more than half a team, occasionally with a spare animal, in general female. CT 55,691 (no date) lists a team consisting of 3 GUD urn-ma-nu-a-te, 1-it GUD bu-ur-ti (trained or not?), 1 GUD 3 4 and 1 GUD 2-24 a total of six animals. Though generalisation is not completely warranted, the four ox team seems to be rare. VS 6, 274 (Sippar yr 3 of an unknown reign) mentions the receipt of a three-year old ox taken by a ploughman from a herdsman, an indication that training only began at this age. Trained cows would seem always to be older than three years. The main purpose of the cows in the plough teams is explained by the Uruk general farm documents. YOS 6, 11 (29.I.Nbn.l) lines 9-10 assigns 100 AB.GAL.ME,"full-grown cows", and 400 GUD.ME,in view of the general context without doubt "plough-oxen", to 100 plough teams. A ratio of one cow to four oxen is regarded as sufficient to keep up the numbers of the plough animals: ana batqa Sa GuD.ME. But there is a different norm in the so-called Edict of Belshazzar, which is actually, in the form in which it is preserved, no more than a kind of memorandum. There one "plough" contains four GUD.MES,two GUD.AB.MES(and four men). The number of cows is doubled and that is what we find in YOS 6, 150 (28.[x].Nbn.11): 100 GUD urn-[ma-nu] 2 50 AB.GAL.MES.It is possible that the 2:l norm was more realistic, but it certainly was also more generous to the entrepreneur. Nothing is said in these documents about bulls serving the cows, but it is highly unlikely that breeding bulls were used for ploughing. No information exists about the duration of the working life of a plough ox, or about norms in this respect which were expected to be fulfilled. Training probably began (later) in the third year, real labour at four, but for how long an animal could continue is unknown. Texts about dead plough animals are uninformative, as no ages are given. The practical evidence about plough teams at inspection indicates that the "institutional norm" of four oxen was reached hardly ever, and that only the heads of groups of ploughmen might have a spare animal. Yet Ner 57 (27.IV.Ner.3) indicates that occasionally complete four animal teams could be assigned. The question of the work norms for plough teams is of some importance for the understanding of the backgrounds of the MuraSu archive, which shows, it would seem, these entrepreneurs exploiting leases of fief land with means at their own disposal. As I have tried to indicate elsewhere, a change of policy becomes visible in the leases mentioning plough teams: there is a shift from teams with two oxen, which I presume to be the rule in ordinary farming, to teams with four oxen, which seem to be a kind of institutional ideal. In the MuraSu archive this change seems to come rather abruptly in the year Artaxerxes I 37, thus suggesting a specific purpose.21 Teams with two oxen would obviously work and were probably even usual in noninstitutional agriculture. What the advantage of doubling the number of oxen was, is not clear to me: adaptation to institutional work norms may have been a reason. Perhaps one could speculate that an institutional team could do the more intensive preparatory ploughing with two ploughs with two oxen and two men, while the more extensive seeding needed more men, with the ox teams being alternated. The net result would be more land cultivated, but this is no more than a guess.
van Driel
Cattle in the Neo-Babylonian period
Ideally a normal institutional plough team in the Neo-Babylonian period consisted of four men and four oxen, with two cows. But BM 64180 (unpublished22 and sadly damaged), records 6,5 plough teams, the half team of which has a strength of two ploughmen, though it had three GUD, one AB and one AB.NIGIN,i.e. an almost full complement of oxen, suggests again that the actual work was done by two oxen at a time and that the animals were changed in the course of the ploughing. The text, however, does not specify whether the cows were trained or not. But if this text could be used to suggest that not animals, but men were in short supply, BM 75537, probably undated, with eight animals and eleven men indicates rather an even greater shortage of animals. This all the more so because the one team which is described in full, consists of one trained ox, two trained cows and one two-year old heifer: not a complete team. It is difficult to generalise from this type of material, but it seems that the number of animals was more often under-strength than the number of men. Plough teams were often family based, which may help to explain the surplus in men. The Uruk general farm text TCL 13, 182 certainly suggests that manpower was less of a problem than traction animals. If two cows were needed to keep the teams of four oxen up to standard, that could indicate that one young male animal per cow was needed each year, that is if cows were expected to produce a calf on a regular basis. A complete team would have consisted of four oxen, two cows, and three young males in their third, second and first year respectively. Female calves could be removed, unless one of the cows had to be replaced in the near future. Nineanimal teams of this type are not attested in the texts, and it must be concluded that one calf per cow per year was hardly a realistic prospect, especially perhaps, where the cows were also used for ploughing. There are no texts indicating that donkeys were used as plough animals in this period, unlikely though the fact might seem. But all our texts come from an institutional or entrepreneurialcontext and do not inform us about the private sector.
9 Policies? The available material is not very suitable for generalisations as to policy. The are no statements about fixed norms in cattle breeding, though these must have existed and, therefore, should be reconstructible. Cattle breeding by the two institutions from which our documentation derives occurred close to home, which probably meant that the carrying capacity of the land was a factor of some importance. The numbers of cattle are much smaller than those of sheep, which also made administration much easier. We can only hope that publication of the Gehlken material will put speculation on a firmer basis. as yet only a few sets of figures concerning a few herds are available and the (relative) chronology of many fragments in dubious. In some respects NBC 4883 (5.IV.Ner.2) is the most interesting text, especially on account of the first herd.23 It should be noted that the text dates to the period in which numbers of third-year animals are not given in Uruk. There are 293 male and 77 female animals in their second year, in total 370, which could never have been produced by the 240 cows mentioned. Of the 293 males only 38 are stated to be KU.What made the other 255 non-KO is not clear to me, but it seems certain that the herdsman ZEr-iqiSa s. Nabfi-&a-iddina had a special herd in which these animals were concentrated. His 508 animals will anyhow have been divided among a number of herdsmen. Special pasturing of young castrated cattle for fattening purposes would, in itself, make sense, but non-KU animals would be of limited use for sacrifice. At the
van Driel
Cattle in the Neo-Babylonian porlod
age of two transfer to the plough teams was still possible. As the early reign of Nabonidus saw, if not the creation of the general farm system, at least its rapid expansion, shortages of plough animals may have been felt more acutely. The circumstances of the moment could be an explanation in this case. Of the other herdsmen mentioned by the text Ibni-Inanna s. Sum-ukin also occurs in YOS 6, 130 (13.IX.Nbn.9), but as most numbers giving the details about his much more modest herd are lost, we can only observe a general increase in 10 years from 24 to 60 animals. , This could be well in line with general expectations for this period: a considerable increase in numbers in every respect. The animals intended for slaughter stay, after withdrawal, for a certain amount of time in the fattening establishment. A group of texts dealing with withdrawals of sheep from these establishments at Uruk indicate, as is to be expected, that animals are disposed of on a daily basis. Lists of "offerings" from Sippar seem to confirm this, also for cattle. It is therefore not unlikely that animals stayed in the fattening establishments for rigorously fixed terms. They were, probably, constantly replenished by animals entering the system from other sources, termed in a very general way irbu. This probably means that there was a certain flexibility in the moment of withdrawal from the breeding herds. Availability of fodder barley may have been significant. I assume that in a text like Gehlken I 48, which clearly removes 38 second year bullocks suitable for sacrifice and only 16 qualified as "marked" together with 34 old or barren cows (if my interpretation of the meaning of S A L . A L in the Neo-Babylonian period is correct), most of the 57 pu-bal.ME KU.ME and the 56 pu-ha1 Senditu belonged to the third-year category of male animals, which was not yet listed separately when the text was written. The small number of calves removed, one male, one female, also deserves notice. Suppositions regarding the age of male animals at culling are confmed by the changing ratio between the numbers for male and female animals in the texts. Numbers of male animals decline rapidly during the second and third year, those of females far less so. Natural losses and removal of barren animals easily explain the reduction in the number of females. In the evidence gleaned from the few tabulated texts from Uruk the decline in the numbers of male animals in their second and third year is very obvious. In the single-herd texts the situation would seem to be more complicated, as the texts from the reign of Kandalznu and Nabopolassar do not indicate any reduction in the second year. Reductions take place as soon as, in this period, animals are counted with the full-growns, that is on entering their third year. Comparison with the Sippar material, Appendices 11and N ,shows a striking difference in policy. A drastic reduction of the male animals takes place already in their first year, only very few are kept, probably as replacements for the breeding bulls. The number of available texts containing useful information is, however, limited. It is not possible to get a reasonable idea about the productivity of the herds. The combined male and female calves listed in the texts, often do not reach half the number of the cows, or do so with difficulty. Yet it can not be said that one calf per cow in two years was a kind of rule, for there are clear instances where animals in the first and second year together are more numerous than the cows, if we look at the Uruk evidence. Sippar with its intense early culling of male animals is of little assistance in the question. All the numbers suggesting low productivity could be very misleading, as we have no firm information about a share of the herdsmen in the newly born animals.
,d
van Driel
Cattle in the Neo-Babylonian period
If we assume the cattle herding was organised along lines similar to sheep herding, herdsmen contractually received a share of the young animals. As stated there are no herding contracts for cattle and the important difference with sheep is, as stressed above, that cattle herding was done at not too great a distance from the (institutional) owners. Contracts like Dar 257 (29.VI.Dar.9) and 348 (2.V.Dar.13), both from the Egibi archive, and closely related, deal only with one cow (GUD bu-uitum). Parties will share the tamlittu, "offspring", on a fifty-fifty basis, but as the contract contains elements of joint ownership the status of the institutional herds will not have been the same, which makes a fifty-fifty arrangement unlikely. Perhaps we could speculate on a 33% share for the herdsman, as is the case with sheep, though this would seem very generous with regards to cattle. This would notably change our perception of herd productivity, though no proof of such an arrangement can be offered. But, as in both arable agriculture and sheep and goat herding entrepreneurial interests are taken care of, it would be unlikely if cattle herding had escaped this type of attention. It would affect the numbers of female animals in their first year in herd inspection texts. The Neo-Babylonian-Achaemenid period is one of economic expansion. A case of modest growth has been pointed out above, but the entrepreneurial approach visible elsewhere is not really demonstrable. The best example in Uruk appears to be the second herd in YOS 6, 130, that of Nanna-ere8 s. Marduk-eriba, managed by his son IqiSa. Nine years earlier the same herd is listed in UCP 9/1,58 ([x].lX.Nbn.O), a text not giving separate numbers for third year animals. For comparative purposes the third-year animals in YOS 6, 130 must be added to the full-growns. At the beginning of the reign of Nabonidus there are 309 animals, of which 203 are full-grown cows, 13 bulls, in total 56 second-year animals and only, male and female together, 37 calves. But there is also a backlog over the accounting year and the previous one, which is broken down into 8 cows, 24 second year bullocks, 7 heifers, 51 male and 37 female calves. As these numbers refer to two accounting years we can not conclude that according b the - unknown - norms the herdsman had to present 62 male and 63 female calves on 21 1 cows, not only as some of these may belong to the account of the previous year, but also because withdrawals remain unspecified. Neither the seemingly low yield nor the backlog seems to have affected the position of the herdsman. After nine years the number of his cows has risen by about 60, or, rather if we add the third-year group to the full-growns, roughly by one hundred. There are 110 registered animals in their second year and 151 in their first. The number of bulls has virtually remained the same, 16 instead of 15, but there are additionally 17 third-year ones for culling. This means that during the nine years about ten cows were added to the herd on a yearly basis after replacement of losses. Practically all young males and a considerable number of young cows were available for culling. That culling of young female animals occurred is confirmed by YOS 6, 130: there were 81 in their first year, 62 in their second and 43 in their third, of these slightly more than ten were kept as increases of the number of animals in the herd, and at a replacement rate of 10% of the full-grown cows, say some twenty per annum, many were available for removal. If the herdsmen, or at any rate their bosses had a share in the increase, that is if cattle herding was organised along lines comparable to sheep husbandry, that share would have been paid preferably in female animals, since the institutions needed the male cattle for their offerings. The numbers of the females in the inspection texts do not show this possible share of the herdsmen. Anyhow the numbers seem to suggest that all-out growth was not the aim. Male animals were needed for sacrificial purposes, female cattle were occasionally sold for cash. The third line of the box devoted to Nanna-ere8 in YOS 6, 130 tells us that 9 puhiilu and 3 third-
van Driel
Cattle in the Neo-Babylonian period
year males were "for taxation", ana Sibti. pubalu in this case will mean animals which had completed their third year and which were removed on the very moment that they could be classified as full-grown, as the number of bulls in this herd seems to have been stable at 15 or 16. As is the case with sheep all ~ i b t uanimals are male. Breeding requirements and cultic requirements mn parallel. The translation "taxation" is probably too specific in this case, "removal by the owner" is perhaps better. The two texts suggest a gradual, well-planned expansion of the hexd. Iqiga, like his brother Zerija, became a rub bzili, a kind of (collegial?) supervisor of herdsmen, interested in both caprids and cattle. The considerable herd of cattle managed by the father certainly contributed to the career of the sons. That the father managed to make good the backlog of UCP 9/1,33, if that was ever more than a matter of bookkeeping, could indicate private resources. Herds of several hundreds will have been split up for the actual herding. The naqidu's, the administrative herdsmen, will have been managers. We can only assume that part of their herds was their personal property. These animals will not have figured in the inspection texts. The material from Sippar is even less informative in its present state. The many tabulated documents are very fragmentary, and small one-herd texts are rare for cattle. Sippar texts, as stated above, first list all male categories, then all female. If this is kept in mind fragments can be used and some conclusions are possible. It seems evident that there are far fewer young bulls in the generally smaller Sippar herds than is the case in Umk. There are hardly ever more than two breeding bulls, the five of BM 6021 1, with 35 cows and 25 thirdyear ones, and the four in BM 49666 with 66 cows and 16 third-years are exceptions. There are several small herds without full-grown bulls. The second herd of BM 61431 with 47 females has no bulls and there are more examples in Appendices 3 and 4. This seems a kind of policy in Sippar. Bulls for serving the cows had to come from outside the herds, as had the bulls for the cows of the plough teams. Sippar cattle herding seems to have been conducted on a much smaller scale than was the case at Uruk. Information about the calving season is difficult to obtain from the Uruk and Sippar texts. Cows in calf are not listed separately. But perhaps bookkeeping was organised in a manner which made the matter irrelevant.24
10 The bit uri2 :fattening, especially in Sippar The institutions had, as was the case with sheep, an establishment where cattle were prepared for sacrifice. The animals were removed from the external herds, received as irbu, in the sense of "gift", or bought, thus also becoming part of the irbu.The amount of fodder used can be calculated in a number of cases. In Sippar Nbn 998 (7.X.Nbn.16) tells us that 10 GUD each received 0.0.2.0, or 12 litres of barley each per day, and 17 0.0.1.3, or nine litres and 8 a mere three litres. The same amounts occur in Nbn 9 15 (30.XI.Nbn. 15), lines 10-15. No reasons are given for the differences, but we can compare CT 55,660 and 674, which specify the categories of animals: in CT 55, 660 (ll.[x].Cyr.3) 13 GUD s'uklulu 0.0.1.0 or 6 litres per day and 4 NINDA2 3 litres per day, in CT 55,674 (5.XI.yr ?) a GUD s'uklulu gets 0.0.2.0, 12 litres per day, the same as a taptiru, whilst animals qualified as tardennt?, which may refer either to age or to quality got 4 litres each. Castrated or not does not seem to make a difference, which stands to reason, cf. also CT 55,711: 14, but, as is to be expected, age does. No reasons why the amounts vary within the same age group are given, but also in other periods there are differences in quantity related to the stage in the fattening process the
van Driel
Cattle in the Neo-Babylonian period
animal had reached. Careful grading of the amounts of fodder, as seems the case with sheep, was practised. Special circumstances may obtain, cf. CT 55, 660, of which the arithmetic is difficult to understand. Cows rarely occur in the fattening pens, the four in CT 55, 71 1:18 being an exception. The number of cattle in the cattle shed of the bit urii cannot have been much higher than about 20, unless the numbers in the fodder texts refer only to the animals in charge of one person, which may well be the case. Offering lists after all suggest a steady expenditure. That occasionally a calf was produced in the E.GUD has been referred to in the terminology section. CT 56,581 and 558 mention 11 men working in the E.GUD, but CT 56,712 mentions only 7 and CT 55,486 reduces the number to 6. A limited insight into the scale of the activities of a institutional cattle fattener is provided by BM 72543, which covers the period 1l.III.Nbn.[l]3 to 30.III.Nbn.15. The text is the damaged account of the cattle (GUD.MES) in the E ri-ru-ri of Zedtu. At the beginning of the period 18 GUD.MES were at his disposal. During the almost two years of the account 180 were bought "for silver", 6 entered as (unspecified) irbu. The next line mentions ENGAR'S, therefore should have listed the number of the animals withdrawn from the plough teams, and it seems reasonable that other cattle derived from the breeding herds, but the text is incomplete. The need to buy animals is anyhow clear. Actual requirements can only be assessed after the considerable number of offering lists has been studied. Barley used as fodder for cattle is mentioned occasionally in private texts: UET 4, 139.
11 Products: mil4 and hides Very little can be said about the "secondary" products of cattle husbandry. In Eanna in Uruk the SIP AD-^^-^^' Si-zib was in the hands of a special functionary. According to YOS 7, 79 these "milk herdsmen" had to provide milk for cultic purposes, but this hardly constitutes a separate sector of the economy. The three "milk-herds", who together had an eight months share in the function, got on a yearly basis two cows to fulfil their obligations. The real work was done by a iirku, also provided by the administration. The cows came with their calves, a detail indicating that they were required for the release of the It is worthwhile to note that the function of which the obligations were regulated by the document had previously been transferred to the king's share in temple functions. If we compare texts dealing with skins of sheep and of cattle it seems that in Uruk skins of sheep are simple KUS and that cattle yield KUS.TAB.BA: GC I 3, YOS 17,62 and YOS 17, 63 which specify the particular category of animal from which the hides derives. Weights of KUS.TAB.BA for Sindu, "glue" are given in YOS 17,65. In TCL 13, 224 we find numbers of KUS.TAB.BAgraded according to three sizes, and valued at 113, 1/4 and 1/5 8 each respectively. From Sippar there is hardly any comparable material. Only Nbn 617, which tells us that the hides of one-year olds are called KUS.TAB.BA, can be mentioned. We can only assume that herdsmen were under similar obligations as those who were in charge of sheep and goats regarding hides, tendons and carcasses of animals dying from natural causes, but contracts specifying these responsibilities are not available.
12 Prices Prices of cattle depend greatly on the quality of the individual animals. We may expect that season, and availability of fodder are of considerable influence. What is clear is that especially
Cattle in the Neo-Babylonian period
van Driel
male animals were bought and sold, the expendable SAL.AL cows being the exception among females. Only prices that can be calculated for animals belonging to specific age or quality groups are listed here, unspecified GUD'S and cases where combined prices are given for different types of animals are disregarded. For comparative purposes prices for dead animals are included in the list. a.
Uruk
GUD ummdnu
YOS 6,132 (22:X.Nbk.lO) 13.5 9, GC I320 (7.XI.Nbn.10) 12 9,
GUD taptiru
BIN I 144 (5.VIII.Nbk.4) 13 9,
GUD 3-u
GC I 363 (29.IV.Nbn.5) 15 9,
NINDA2.TUR AB .GAL SAL.AL
VS NF IV, 70 (27.XI.Nbp.16) 20 9, GC 1 164 (15.VI.Nbk.43) 4.5 8,
GUD m a r ~ u
Doty JCS 35 p.218 (4.1X.Nbk.36) 3.25 9,
p a g r u $a GUD
GC I 315 (7.VIII.Nbn.5) 5 9.
b.
Sippar
GUD ummsnu
BM 61117 (29.[x].Dar. 29) 30 9,
GUD 6-i
CT 57, 70 (5.V.Nbn.9) 11 1, BM 67021 (Nbn) 12.33 9,
GUD 4-u
CT 55, 669 (7.IV.Ner.l) 18 9, CT 55, 688 (10.IX.Nbn.9) 16 9,
GUD 3-u
Nbn 127 (23.XI.Nbn.3) 25 9, Dar 222 (6.IV..7, cf. CT 55, 666) 10 9, Cyr 105 (10.II.Cyr.3) 18.33 9, CT 57, 289 (22.II.Nbn.9) 6 1, CT 57, 223 (1 l.VI.Nbn.10) 5 9, Carnb. 381 (9.VII.Camb.7) 22 8 (sic),
GUD 2-u GUD NINDA2 GUD NINDA2 2-24 GUD TURIA .KT GUD h k l u l u
BM 83698 (5.II.Nbn.4) 12 1, Dar 185 (19.IX.Dar.5) 10 9, Nbk 213 (20.III.Nbk 30) 8 9, Nbn 639 (25.VIII.Nbn.12) 7.5 9, CT 55, 705 (2.I.Nbn6) 5 1, CT 57, 451 (17.X.Nbn.9) 5 1, Nbk 140 (28.II.Nbk.24) 13 9, Nbk 304 (22.VIII.Nbk.36) 20 9, CT 55, 681 (8.XI.Nbk.36) 10 9, CT 55, 703 (19.XI.Nbn.10) 4 mana of wool, Nbn 923 (20.X.Nbn.15) 15 9, BM 72794 (Dar 22) 28.1 1, CT 55, 694 (ll.VIII.Nbn.[x]) 2 3 3 9, Cyr 218 (13.XII.Cyr 5) 20 9, CT 57, 153 (Nbk.111-Dar 0) 20 1, Nbk 204 (22.VII.Nbk.36) 20 9, Nbk 399 (IV.Nbk.42) 30 9, BM 83698 (5.II.Nbn.4) 20 9, CT 55, 683 (8.IV.Nbn.7) 18 1 BM 83259, (Nbn.8) 15 9, CT 55, 580 (7.VIII.Nbn.10) 8.5 9 , CT 57, 153 (Nbk.111-Dar 0) 27.5 9, CT 55, 681 (8.XI.Nbk.36) 10 9, BM 83261, (23.IV.Nbn.7) 6 8, CT 57, 289 (22.II.Nbn.9) (6 9)?
GUD s'uklulu A.MU
Nbk 304 (22.VIII.Nbk.36) 7 9, Nbn 127 (23.XI.Nbn.3) 18 9, CT 57, 451 (17.X.Nbn.9) 5 9, Cyr, 218 (13.XII.Cyr.5) 6 I, EM 62052 (15.XII.yr4) 5 9
GUD AB
CT 55, 686 (Nbn.7) 10 1,
GUD taptiru
Nbk 132 (Nbk.1819) 25 9, BM 64913 (lO.II.Nbk.(late)) 13 9, BM 83698 (5.II.Nbn.4) 26.8 9, CT 57, 223 (1 1.VI.Nbn.lO) 8 9, CT 55, 693
van Driel
Cattle in the Neo-Babylonian periad
GUD, dead
CT 55, 707 (IX yr.6) 4 S (note 0.5 GUD also 4 S), cf. CT 55, 680 (X.yr.18)4.5 I (dead?), BM 59255 (15.VIII)5 I (dead?)
ummdnu, dead
CT 55, 700 (28.X.Nbn.0) 0.5 kor dates,
Though unspecified GUD's are not included an exception should be made for CT 55, 671 which gives averages for five years of an unspecified reign, either that of Nebuchadnezzar IT or that of Darius I. In yr 19 61 head of cattle were priced at 24.5 S on average, in yr 20 37 at 21.6 S each, yr 21+22 113 at 22.5 S and yr 23 44 at 14.11 1. Years 21 and 22 cannot be calculated separately, as the text is damaged. The total over five years gives 255 at an average price of 21.4 S. Though the lower average for year 23 could be the result of the purchase of animals of a lower age category, it could also be the result of unfavourable circumstances for the herdsmen. If the cattle terminology deals in the main with gender, age and training, we can expect that these are the factors which influence prices, but we must keep in mind that cattle prices are much more based on the qualities of the individual animal than is the case with sheep. The evidence is not very ample for a period of over a century. It could be supplemented by a judicious selection of facts culled'from passages where a total amount is given for a group of animals for which the age is individually specified, or from passages giving prices for unspecified GUD, but this is not attempted here. As yet, we know little about scarcity, exceptional losses and other factors influencing prices as a whole. We can only look in a very general way at the facts collected. Gender obviously plays an important role. For female animals specified prices are only given for one old cow sold by Eanna, and in Sippar there is only one price for a GUD.AB,or rather GUDKB,in this case, which is certainly a cow on account of the 1-it preceding it in the text. Otherwise all evidence concerns "male" cattle, generally required by the institutions for offerings and as draught animals. We are not dealing here with cows in private archives. The importance of age is not always clear. The evidence from Uruk is insufficient in all respects, we must concentrate on Sippar. Sales documents make only a limited use of the term NINDA2 after the second year, and we probably can regard the group consisting of GUD plus age group and all the GUD.NINDA2 as a single general category. From calves, TUR or A.MU in the list to 2-u and NINDA2 2-u to GUD 3-u and 4-24 there is the expected steady increase in price, the high price of 25 8 for a 3-u perhaps obtained at the moment, month XI, when the animal was at its best, but such prices were not always obtained (CT 55,688). The high price for a 2-u in Camb 381 seems suspicious. Prices for GUD 6-i seem lower. But it should be noted that the prices for this type of cattle remain generally well below the high averages of the first four years of CT 55,67 1. There is no quotation for a breeding bull. With the Suklulu's the same observation of prices rising with age seems to hold good, with some of the 3-u's obtaining the high prices of 30 S or 27.5 S, with 20 S occurring regularly.26 This reflects the fact that these animals were sought after for offerings. There are only a few prices for tap~iru'sand ummdnu's. For this last category the 30 1 in Sippar is amongst the highest prices paid for cattle by the institutions, but tapfiru's, excepting those from Nbn yr 10, are also expensive, skill could play a role. These are possibly trained animals. The lower prices of an ummcinu and a tapjiru early in the reign of Nebuchadnezzar I1 in Uruk could perhaps indicate lower overall prices in this period, but the evidence is insufficient.
van Driel
Cattle in the Neo-Babylonian period
Prices for dead cattle seem to be very constant, the SAL.AL cow in Uruk falls, as expected, in the same price bracket, but the sick animal in the same place realises even less: old or dead with some meat is of greater value than younger, but with little that is edible. It should be stressed that the evidence is scanty and that all conclusions are specuiative. Collecting material and avoiding conclusions has, however, little sense. There is as yet no scope for speculation about general long term tendencies.27
13 Final remarks . The information on Neo-Babylonian cattle is limited in scope. The limitations are underlinedby the fact that nothing can be said about the use of donkeys as draught animals. It is unlikely that they were not used. Their purchase and sale is attested by a number of texts, published and unpublished, from private archives. It is, perhaps, less surprising that nothing can be said about pigs, as they really do not seem to have played an economic role in the institutions in this period. But they are not found in the private archives either. Archaeozoology can possibly help us here. still It is, perhaps, of some interest to note that the wild donkey, AN$E.EDIN.NA, occurred in the Uruk area in early Achaemenid times. In YOS 7, 156 (4.XI.Camb.3) six men are engaged to protect the harvest from damage by wild donkeys and gazelles. The overall conclusion must be, that, as with sheep, the institutions needed the male animals as offerings and for traction. If we understand terminology in the correct way, uncastrated, s'uklulu, animals went to the gods of the highest rank, younger animals, possibly even of less quality, to minor gods, who clearly occasionally received castrated animals. Such animals therefore were not totally unsuitable for offerings. As these animals were shared out amongst those who had a right, the bones should turn up in the bone record. In Uruk more male animals stayed for longer with the herds than in Sippar, but, not unexpectedly, most disappear at the age of three. Especially in Sippar much cattle had to be bought for offerings. This is a clear indication of the existence of cattle breeding outside the institutions, producing for their needs. This strongly suggests the existence of a private sector. In turn the fact increases the likelihood that institutional herding of cattle, similar to that of sheep, was, at least partially, in the hands of entrepreneurs. These people will have had a share in the yield of the herds, and that share will have come in the shape of female animals. Only a large scale herding contract can confirm the hypothesis, but we must draw the supposition into account when speculating about the productivity of the herds. The texts available could easily cause us to underestimate productivity. If the herdsmen had a share in the female calves, one calf per cow, or slightly less per year could well be in the range of the possible. Plough teams in the institutions consisted of four oxen and two cows, intended to assure the completeness of the teams. In reality teams were under strength, even if cows were used as traction. The training of the plough oxen started around the completion of their third year. Prices indicate the obvious. There are signs of sharp fluctuations, no doubt related to the special circurnstances of the year. But in general prices rise in each category in a steady manner until the third year, when optimal conditions for slaughter are reached. This holds good for male animals, for females there is no evidence from the institutions. Though the evidence is not
e
van Driel
Cattle in the Neo-Babylonian period
ample, cattle breeding seems to show the signs of steady long term growth we have come to expect of the Neo-Babylonian economy in general.
NOTES
11 12 13
14
OrNS 18 (1949) p. 302-306. Especially MSL 811 (1960) p. 61-78, Excursus 11. The most important difference between texts dealing with cattle and those listing sheep and goats seems to be that cattle texts give only one total, but that most sheep and goat texts provide subtotals and a grand total. There are, however, also texts that deal only with sheep or goats. In these cases there is also only a single total. Fragments can in many cases be classified even though headlines are lost. The tabulated documents are used not only for herd inspections, but also for the registration of shortfalls, rnittitu, for the sale of animals from the herds and for accounts covering more than .one year. The format allows, of course, all kinds of adaptations. In a few cases additional columns list withdrawals, comparable to the sibtu columns in the sheep texts. With slight adaptation of the headings this type of text can be used for the registration of the plough teams. The Sippar material is very fragmentary and reconstruction of the texts will require much time. The difference between texts listing breeding herds and plough teams should be found in the headings. These are situated on the upper edge of the texts and are often badly preserved. Terminologically the texts differ only in the word used for the full-grown "male" animal. BM 50058 rev. 3 has 1-en bi-i-ru s'a PN, a rare case of the actual use of the word. CAD L p.217-219 quotes BIN I 95:12 1-et IB-turn, which could be read lit-turn but also biirturn. In the other text quoted we must read a-lit-turn as the line in question is the heading of a small tabulated section. The sign before LIT must be emended into A, as numbers cannot occur in the heading of the table. A possible reference for littu could be EvM 12:2 given by AHw p. 558, but this too could be read GUDIB-~U,biirtu. arbu, "cow", cf. CAD A12 p.263, is probably obsolete or literary in the NeoBabylonian period. The final passage of VS 6, 274 requires collation. An exception is suggested by Freedman, Sacramento no. 9 ( I . X I I I I . C ~ ~ from . ~ ) , Uruk, JANES 8 (1976), p.42 and p.47 (copy no. 8) line 12: 12 IB.PES~.This is, however, not shown by the copy which, perhaps, could be read BxA, but which probably represents '~B.NIGIN.This would turn the text into a simple Uruk single-herd inspection text. Though not strictly meaning "in calf' a reference must be made to EvM 20:3, C.Wunsch, Iddin-Marduk no. 60,4 GUDbu-ra-turn s'u-le-e-turn. The translation is perhaps based on CAD E p.l30R, e); the passage is quoted CAD B p. 334L: "four bred cows" and Landsberger, MSL 811 p. 66: "four mated cows". A full grown cow is perhaps, with CAD B p.334 a bus'lrtu, but the only practical text giving an equivalent seems to be the Sippar text CT 55, 686, which tells us that 1-it GUD.03, bought for 10 S was placed in the li-ru-li as a GUD bu-us'-turn, in itself an unusual phenomenon. Cf. note 14. The idea that possibly some seasonal factor could be involved cannot be sustained. Though it is true that most texts not mentioning a third year category date from the months VIII-XI, YOS 6, 118, which does include the category also dates from month VIII. Other texts including the group date from months 1I.V and VI, whereas Durand-Joannks TBERITEBR, pl 321110 19, without the third-year group, dates to month 111. An explanation on the basis of supposed seasonality leads nowhere. This is indicated by a K in the columns concerned. Uruk texts with s'uklulu are YOS 3, 138:14, YOS 6, 26:19 1-en GUD S U . D U ~and , UCP 911 no.66, a dead GUD.SU.DU~. It will be clear that Landsberger's interpretation is not followed here, cf. Gehlken, Uruk, Wirtschaftstexte I, 1990 p.42f., note 41. Cf. for the OB contracts PBS 812, 196. The ritual is Rit.Acc. 3, 2. On this text cf. Landsberger MSL 811, p.72: obscure. SAL.AL is otherwise used only for female animals, and tapflru of "males". Cf. note 8. bhrtu seems well established as the word for cow, cf. CAD B p.334. CT 55, 691: 1-it GUDbu-ur-ti, BM 74482 bu-ud-turn $6-ad-di-turn and BM 50028 bu-ra-turn which the shepherds .......
van Driel
Cattle in the Nw-Babylonian period
In CT 55, 686 1-it GUD.PB is bought, but it enters the 6 li-ru-li as a biis'turn, in itself an exceptional occurrence, as female animals are rare in the fattening establishment. Cf. also CT 55,691:2 1-it GUD buur-ti. In BM 55920 we find a GUD bu-d-ti BE-turn, "a dead cow". biirtu for heifer or calf, CAD B p.334, MSL 811 p.65f., is not, in my opinion, attested in NB texts, but for the apparent exception Dar 257:l from the Egibi archive, where biiiturn is explained by a further ma-ri s'a-turn, in its first year, probably because it is intended for breeding it is called a "cow in its first year", possible also with the intention of making certain that for the moment no offspring may be expected.. The text probably does not list a herd. No reason or purpose is given for the list. The final line is PAP 68 TUR s'd-a-ti, "in total 68 males in their first year", which cannot be the offspring of 27 cows in one year. The total of the preceding lines is 69. Something is probably wrong. The unique feature of the text " is that it combines the word alittu from the herding sphere with GUD.NINDA2 and &.NIGIN, of which the first otherwise only occurs in the offering sphere. Note Nbk 132:lO and 12 (IV.Nbk.l8), CT 55, 661 (date ?), 57,223 (ll.VI.Nbn.lO), 55,693 (9 ~ 1 b i ~ . ~ b n . 57,439 l 0 ) , ([x]). Note in CT 56,503 (16.IV.yr.?) line 3 1 GUD I(!) tap(!)-ti-ra, together two GUD at the disposal of ploughmen. Male plough animals must be castrated, the difference between GUD and tapfiru could, in this case, be one of age: the GUD is trained, the taptiru "suitable". In Nbn 646 (13.IX.Nbn.12)line 6 we find 1-en Gull.@ a-lit-tli 1-en 1B.NIGIN a-lit-turn side by side. Though this could suggest that the sahirtu category could contain animals that had calved, Landsberger, MSL 811 p.78 rightly considers the possibility that a negation has been omitted by the scribe before the second alittu. The case is exceptional anyhow as NIGIN is supposed to exclude automatically the possibility of having calved. Is this an attempt to indicate that the heifer was in calf! In this context we must consider a small number of Sippar texts in which the word s'ukulturn occurs, often, but not always, in combination with herdsmen. AHw 1265R translates "Speisung, Futterung" and seems to suggest in connection with Cyr 57:l and Nbn 490:l that the animals are destined for the herdsmen. The texts concern clearly deliveries (by the herdsmen), as is the case in Nbn 948:7. Not clear is CT 55,618:6. s'iikulturn is a qualification of a group of animals delivered by herdsmen, who obviously kept them "to make them eat". s'iikulu is occasionally used for one of the activities of herdsmen. BM Bab-Nar60211 is a tabulated document summarising the herds of herdsmen at Masemi(?). In rev. 1-2 it adds in the remark e-lat 55 TUR.MU s'd LUSIPAD.MES I-s'd-ki-lu, "excluding 55 bull calves which the herdsmen pastured". This suggests that keeping the young male calves with the herd was an additional task of the herdsmen. As with sheep, the institutions were particularly interested in obtaining the male animals for the offerings. This type of information is rare in the cattle texts. BM 83559:l GUD s'u-kul-turn s'd LUSIPAD.MES, followed by a list of animals qualified as A.MU and PN's, could represent a list of fattened calves delivered by herdsmen. Month 11, however, regularly occurs as the month in which calves (and second year olds) are delivered, perhaps shortly after birth, and this is hardly proof that fattening of cattle other than through feeding with barley occurred. The Egibi text Dar 351 (25.V.Dar.13), line 11 seems to know of a bu-us'-turn urn-ma-[nu], but there is no context. And the so-called Edict of Belshazzar, YOS 6, 103. Cf. the table in JESHO 32, p.220. An edition will be included in the dissertation on Sippar agriculture prepared by M. Jursa, Vienna. Published by R.H.Sack in the T.Jones Festschrift, AOAT 203, p.2llff. A certain amount of speculation seems warranted. From Sippar there is a number of texts listing in tabulated form the removal of young male cattle from the herds. Though second and third year animals are included, what concerns us here are the first year animals. One might expect that there should be some connection with the herd inspection texts which have been included in Appendix I11 but only of some of these it is certain that they date to month VIII. (BM 49282, 49666, 54004 and CT 55, 455, date from this month in the years Nbp.l4(?) or Nbk.9.), whereas the culling texts predominantly date from months XI1 to 11: XI1 53982, BRM 1 91; 50598, 53753, VS VI, 203, BRM I 91, 52602, 52602, 52650; I 83717,59616, 83559,51819,74666; I[ with smaller numbers in VI and VII, represented by BRM I 91 and BM 49456. There are of course occasional texts mentioning small numbers of animals in other months. I have not consciously collected
Cattle in the Neo-Babylonian period van Driel
25
26
27
Cattle in the Neo-Babylonian period
this type of text, but the peaking at around the beginning of the year must have some significance. Is this (immediately after) calving time? These are at any rate the young male animals we do not find in the herd inspection texts. If continuation of the speculation is allowed the smaller peak in months VI-VIII might represent the consequences of pressure on the available pasture at the end of summer. The dates of the available tabulated herd inspection texts do not, as yet, betray a pattern, but VIII, IX and X seem to occur, certainly as a group, more often than the other months, though I11 also occurs a few times. This could indicate a system in which accounting was prepared before the stalt of the calving season, possibly with an additional round after it. Much more study is required. It is regrettable that the (tabulated) accounts of the individual herdsmen are (as yet?) too damaged for use. More or less confirmed by some Sippar texts. BM 6021 1 rev 6, and possibly 54348 rev.5, damaged. Furthermore 61941 rev. 6, which does not, it would seem, include calves, but which mentions Nabumugezib, rg'u ia iizib. The 3-u in CT 55, 580 at 8.5 3 and especially the 4-14 at four mana of wool, both end Nbn 10, are striking exceptions. Both tapfiru's of this year are also low priced. This could point to special circumstances. The price for the GUD 2-u also could be below normal. CT 55,655 (23.IX.Nbn.10) gives a combined price of 23 3 for four GUD, including a cow and a calf, another animal is added for 8 3, and especially the 8 3 for 3 (or even 13?) 2-u's in line 1 is a clear indication of problems in this year. In Uruk the urnmanu of this year is only normally priced when compared with the similar animal early in the reign of Nebuchadnezzar LI. If Nbn yr 10 was a bad year for cattle breeding, that illustrates that the Uruk texts discussed in section 9 can only be used with great diffidence: they stop in year 9 and do not include 10. On a possible general deterioration of the economic situation about this time cf. P.-A. Beaulieu, The Reign of Nabonidus, YNER 10 (1989) p. 201-203. Compare the following "institutional" prices for donkeys and horses: Uruk: BIN I 160 10 3 for a shedonkey (14.VI.Nbk.3) and GC I267 (4.vi.Nbn.8) 230 3 (out of which 120 3 is paid) for a selected horse. In Sippar: CT 55, 747 (11.I.Nbn.7) 4 3 for a she-donkey and Camb. 169 (4.vIbis.camb.3) 13 3 for a donkey.
van Driel
APPENDIX I: URUK: COMBINED INSPEmONS NBC 4883 (5 IX Ner 2) = R.H.Sack AOAT 203 v. 112-4
K = KU3: always ina libbi x KU3.
van Driel
Backlog of two ycars. Freedman, Sacramcnlo no 9, JANES 8 (1976) p.42 and p.47 (no 8).
Cattle in the Neo-Babylonian period
Cattle in the Neo-Babylonian period van Driel
van Driel
Cattle in the Neo-Babylonian period