1. Introduction

In Irisaĝrig, the sesame oil mill belonged to the governor’s palace. Under two overseers, 40 to 60 women worked there; on average, about 20 to 28 females were always occupied with the process of sesame oil pressing. At the palace, they not only received a monthly wage in barley, but also beer and fish, soup and meat as meals. When the governor of the province returned to office after an interregnum of 20 months, central sectors of the palace were meticulously documented, including the sesame oil mill. Officials drew up a list of the equipment in the workshop where sesame oil was pressed. It included grinding stones, wooden mortars and pestles, and vessels. They also documented the total sesame transactions during the previous 20 months: the women processed 145,900 litres of sesame seeds (almost 100,000 kg) into 28,866 litres of sesame oil, a yield of 20%. According to these data, a female worker could produce slightly more than 2 litres of sesame oil per day.

2. The Overseers of the Oil-Milling Women

The sesame oil mill belonged to the governor’s palace of the of Irisaĝrig, and its personnel was supplied directly from the governor’s kitchen. For 20 months, from 30/VI/Amar-Suena 7 to 30/II/Amar-Suena 9, the governor Urmes was not in office, and apparently the crown administered both the palace and the province. When Urmes returned to the palace of Irisaĝrig on the accession of his brother-in-law Šu-Suen, the administration drew up accounts for the previous 20 months.  Evidently, only those sectors that were directly connected to the palace were recorded. The sesame oil mill was among the posts controlled by the scribe Šarakam[individual=Šarakam] and the general Ur-Engaldudu[individual=Ur-Engaldudu].

The fact that several documents concern the production of sesame oil is all the more remarkable because only a few other areas are similarly recorded for the same period : one label comes from a tablet basket with accounts for the „administrator“ (aĝrig) Šu-Mama (CUSAS 40/2 0602), one from a tablet basket with the expenses of the overseer of the blacksmiths (CUSAS 40/2 0437). Besides the oil mill, the scribe and the general also controlled several other groups of people, the merchant’s silver, or the textile workshop (Nisaba 15/2 0148).
The documents record the following transactions in the management of sesame over the 20 months of the interregnum (30/VI/Amar-Suena 7 to 30/II/Amar-Suena 9):

• transfer of ca. 146,000 sila/litres of sesame seeds from the „scribe“ (dub-sar) Ne.ne[individual=Ne.ne] to Būr-Mama[individual=Būr-Mama], „overseer of the oil-milling women“ (ugula geme2 ĝeš-i3-sur-sur-ra) (CUSAS 40/2 0068); the quantity given in the text (only transliteration) is: 486.2?.⸢x⸣ 8 ⅔ sila3, thus 145,928.7 + x sila/litres, whereby 0 ≤ x ≤ 170 litres, thus not more than 146,098 litres.
• transfer of „28,866 sila/litres of sesame oil“ (96.1.0 6 sila3 i3-ĝeš gur) from Būr-Mama, „overseer of the oil-milling women“ to Ne.ne, the „scribe of oil (matters)“ (dub-sar i3) (CUSAS 40/2 0136)
• transfer of „6,083.5 kgs of sesame oil cake“ (202 gun2 47 ma-na tuḫ ĝeš-i3[glossary=tuḫ ĝeš-i3]) and 2,140 „workdays of women“ (geme2 u4 1-še3) from Adalal[individual=Adalal], „overseer of the oil-milling women“, to Būr-Mama, „overseer of the oil.milling women“ (CUSAS 40/2 0680)
• transfer of „27 workdays of men“ (27 ĝuruš u4 1-še3) from Ne.ne, taken over by Adalal, „overseer of the oil-milling women“ (Nisaba 15/2 0167)
• account of 3,970 workdays of men (unclear), perhaps (lines not preserved) under the responsibility or at the disposal of Būr-Mama, „overseer of the oil-milling women“ (CUSAS 40/2 1574)

Būr-Mama[individual=Būr-Mama] and Adalal[individual=Adalal] both also acted as „overseer of the oil-milling women“ (ugula geme2 (še)-ĝeš-i3-sur-sur(-ra)[glossary=ugula geme2 ĝeš-i3-sur-sur]) at the governor’s palace of Irisaĝrig during the interregnum. The state controllers handed over the palace’s oil mill with its 29 saddle querns and wooden mortars to Adalal to continue (CUSAS 40/2 0428). Būr-Mama and Adalal were still in office almost 10 years later in IX/Šu-Suen 9 (CUSAS 40/2 0017).

A third overseer of the oil-milling women was Šu-Šamaš[individual=Šu-Šamaš], who is attested in one single document only; there, he transfers „a replaced remainder“ (la2-i3 su-ga) of 159.5 litres of lard (i3-šaḫa2) to Ne.ne[individual=Ne.ne] „scribe (i.e. accountant) of oil (matters)“, dub-sar i3[glossary=dub-sar i3] (CDLB 2014 1 3; ŠS.08.00.00).

3. The Number of Oil-Milling Women

3.1. The Pool of Oil-Milling Women

How many women (geme2 i3-sur-sur(-ra)[glossary=geme2 i3-sur]) worked in the oil mill of the palace at Irisaĝrig? Since no personnel lists are preserved, direct and indirect evidence has to be evaluated for this question. Moreover, the number of workers can hardly ever be given with a single figure anyway. We know very well that there were also workers with half or two-thirds of the working hours ((Englund 1991: 257)), and women were released from their work for 5 to 6 days per month ((Englund 1991: 275-279)); moreover, workers could always be assigned to other tasks: the oil-milling women, for example, were also employed in the sesame harvest (JAC 24 56 02 and Dossier A.1.1.10, § 7; transaction of working days CUSAS 40/2 0680). All these factors influence the figures given in a substantial way.
Here is a model calculation: If only 10 of 40 women were employed as half-time workers, there were mathematically 35 workers available each day; after deducting the free time of 20%, this leaves only a continuous crew of 28 women. With 20 half-time workers, the calculated crew was reduced to 24 women, stemming from a pool of 40 persons.

Back to the pool of oil-milling women. The documents concerning the harvesting of sesame provide the most important clues on the number of women: a total of 33 to 63 oil-milling women worked there. From Adalal[individual=Adalal] came 18 to 26 female oil millers, from Būr-Mama[individual=Būr-Mama] 15 to 27 women (table in Dossier A.1.1.10, § 7). Ku-elak[individual=Ku-elak], also mentioned in these harvest documents, was responsible for craftsmen; in addition, women who were employed in the sesame harvest worked for him, although no oil-milling women. Since the harvesters listed in the documents worked on a particular day, the highest number must have included women working part-time.
Another piece of information is provided by the number of the 29 handmills in use in the palace of Irisaĝrig (§ 5 below), which Adalal took over (CUSAS 40/2 0428; AS.09.02.30). Did the maximum of 26 oil-milling women supervised by Adalal work on these 29 handmills? Or did Adalal run the whole workshop on behalf of both supervisors? In my opinion, the data do not allow a clear answer.
The quantities of barley allocated monthly to the oil-milling women indicate the total number of women, because in Irisaĝrig women, especially „grain-milling women“ (geme2 kinkin2), almost always received 30 litres of barley monthly. Only young women received 20 litres, children also 10 litres monthly. Such distribution patterns and shortfalls during the respective month explain results such as „36.8 women“, which implies that at least 37 women were employed at the time.

According to the two accounts of monthly barley allocations (Table 1), about 40 oil-milling women were employed towards the end of the Irisaĝrig archive period; at least one account is dated to Ibbi-Suen 3 (Nisaba 15/2 0951), while the other date is not preserved. A reduction in staff had previously occurred in Ibbi-Suen 1, as that was when 11 full-time oil-milling women were no longer taken over for work under the overseer Adalal (Nisaba 15/2 0657).
According to these figures, one may conclude that there was, at a certain time, a pool of up to 63 oil-milling women working under the two overseers Adalal and Būr-Mama. After a reduction by 11 full-time workers in year Ibbi-Suen 1, there were still at least 37 women working in the oil mill in year Ibbi-Suen 3.

3.2. Beer and Food for the Oil-Milling Women at Work

It must have been particularly lucrative for Irisaĝrig women to work in the palace’s sesame oil mill, because they not only received the usual monthly allotments of barley there, but also meals and beer from the palace kitchen. Like other women employed at the palace, namely „women in the brewery“ (geme2 babir2), „grain-milling women“ (geme2-kinkin2) and „women at the animal fatteners“ (geme2 e2-gurušta), the female oil millers also received meat, soup, fish, beer and bread. Of course, among the oil millers, this special provision only applied to those who were producing sesame oil on that particular day.
As Tables 2 and 3 below show, the oil-milling women usually worked „at“ (ki) Ne.ne[individual=Ne.ne], the „scribe of oil (matters)“ (dub-sar i3), i.e. the person in charge of all aspects of oil production and distribution in Irisaĝrig; on one occasion their place of work is with the well-known overseers Būr-Mama[individual=Būr-Mama] and Adalal[individual=Adalal] (CUSAS 40/2 1755). The delivery of the food was confirmed by sealing, usually by Māšum[individual=Māšum], son of Ne.ne, who always sealed for his father; one time, another son of Ne.ne sealed (CUSAS 40/2 0622).
The monthly accounts on foodstuffs reflect the actual working days. If women were hired out for other duties outside the oil mill, then they could not receive a meal in the palace . The amount of food deliveries implies a certain number of women actually working; a calculated value of, for example, 24 women on each day of a month, implies a pool of 35 to 40 women on rotating duty (see above § 3.1); furthermore, an average value of 24 women in the oil mill says nothing about daily fluctuations, so that on one day 29 women, another day only 20 women might have been present.

The oil-milling women received between 570 and 720 sila/litres of beer per month (see Table 2). The beer was drunk by the female oil millers working in the palace, as some texts explicitly state („they drank it“, ib2-naĝ). But how much beer did an oil-milling woman drink per day on duty? Brunke had inferred the following allotments per day for the „yard cleaner“ (kisal-luḫ): 1 litre of soup, 1 fish, 2 litres of beer and 2 litres of bread (level III-2 after Brunke 2011: 305-6). However, although the court cleaners were given a meal in the palace, they did not receive monthly barley allotments like the oil millers. The oil-milling women, on the other hand, never received bread with their meal, probably because they could bake it themselves from their barley allotment. Therefore, a female oil miller received most probably one litre of beer per day from the palace. Table 2 shows the receipts for beer over a month to the working oil-milling women; there were therefore on average 19, 20 or 22 women working every day.

The oil-milling women also received meat, soup and fish from the palace, which they „ate“ (ib2-gu7), as it is often said in the tablets. The soup was made of meat, pulses, cereals, salt and spices, so it was a very rich kind of paste or cream (Brunke 2011: 165-196). The women’s diet of meat seems impressively varied and extensive: from cattle and sheep to game, pigs and the „reed rats“ popular as a delicacy, plus larger birds, mostly pigeons, as well as extensive quantities of small birds. Can we deduce that the leftovers from the feasts in the governor’s palace were distributed to the women working there? In any case, the quantities of meat cannot be easily quantified, so they offer no usable indication of the number of oil-milling women.
The quantities consumed in a given month are shown in the following Table 3. It should be noted that most of the tablets are only „published“ in transliteration, so that errors are to be expected, especially in the case of figures. Errors in publication are rarer in the case of counted units, in this case fish, which are therefore particularly used in the analysis.

Notes to Table 3:
Birds: first number = doves, pigeons etc.; second number = „small“ birds
CUSAS 40/2 0276 (AS.07.12.00): The transliteration reads „8“ instead of „8,00“ = 480.
Nisaba 15/2 0120 o.5 (AS.08.10.00): The photo allows a correction to ki [Ne.ne … ] (instead of ĝi[ri3 …]).
Nisaba 15/2 0201 (ŠS.01.13.0) is published in transliteration only; the number of „3.4.1 4 sila3“ is incorrect (the word gur is missing) and it does not fit into the series.
CUSAS 40/2 1897 (xx.xx.xx.xx) is published in transliteration only; the publication writes „72“, which must have been written as 1×1,00 + 1×10 + 2×1, or 60+10+2; I interpret this writing as 1×10,00 + 2×1,00, or 600+120.

* * *

The number of fish in Table 3 gives a first indication of the number of women. Between Amar-Suena 8 and Šu-Suen 7, i.e. for about nine years, all values lie in a narrow corridor of 22 to 27.3 fish per day; it is obvious that one fish was the daily meal of an oil-milling woman. How much soup she received is not quite certain, possibly ⅔ or ¾ litres. The figure of an average of 22 to 28 women on duty corresponds very well to the figure for beer deliveries (Table 2), which suggested an average presence of 20 to 22 women for the years Amar-Suena 8 to Ibbi-Suen 1. The high numbers for meat, soup and fish in the last two texts from the year Ibbi-Suen 1 remain unexplained (Table 3: CUSAS 40/2 0476; Nisaba 15/2 0603).
A lower number of women worked in the oil mill according to the oldest text, dating to the end of the year Amar-Suena 7 (Table 3: CUSAS 40/2 0276), namely only 16 women according to the number of fish. Perhaps the rising number is related to an increase in the production of sesame seeds after Amar-Suena 8 (cf. Dossier A.1.1.10)?

4. The Production of the Oil Mill in the Palace of Irisaĝrig

Food was allocated to oil-pressing women in different months (attested are months I, III, V to VII, IX to XIII in Tables 2 and 3 above), and this attests to the production of sesame oil throughout the year.

The accounts of the administration over the 20 months of the interregnum at the palace of Irisaĝrig (AS.07.06.30 to AS.09.02.30) have a great advantage for us: the transactions were documented over a relatively long period of time, so that monthly fluctuations do not matter much. Two texts in particular are very significant: Ne.ne[individual=Ne.ne] transferred slightly more than 146,000 sila/litres of sesame seeds (which is about 100,000 kg) to Būr-Mama[individual=Būr-Mama] (CUSAS 40/2 0068); Būr-Mama, on the other hand, transferred 28,866 sila/litres of sesame oil to Ne.ne (CUSAS 40/2 0136). The volume of sesame oil, 28,866 sila/litres, comprised just under 20%_vol of sesame seeds, about 146,000 sila/litres, slightly below the usual yield of sesame oil of 22-27%_vol in this period (see Dossier XXX). So it stands to reason that the women under the overseer Būr-Mama processed exactly this volume during the 20 months. In each month, 1,443 sila/litres of oil were pressed from 7,300 sila/litres of sesame seeds.
As the food deliveries discussed in paragraph 4 show, about 16 to 28 women worked during this period (see above), an average of about 21 women. The data are summarized again in Table 4.

Approximately every six months, a value is documented; as an average value, a figure of 22 women results. Accordingly, a woman would have produced about 66 litres of sesame oil per month from 332 litres of sesame seeds (figures rounded). A woman thus produced 2.2 litres of sesame oil per day; for the production of 1 litre of sesame oil, one can therefore assume half a day.
In making this estimate, I assume that Būr-Mama[individual=Būr-Mama] accounted for the entire production of sesame oil with Ne.ne[individual=Ne.ne] and that there were not also parallel documents from Adalal[individual=Adalal]. The work process, of course, included all the work from soaking to filtering.
The testimony of the account of sesame press cake (tuḫ ĝeš-i3, CUSAS 40/2 0680) over the 20 months of the interregnum is not easy to interpret. Adalal handed over 6,083.5 kgs of sesame press cake and 2,140 working days of women to his colleague Būr-Mama. Can these figures be combined, for example, so that Adalal booked the leftover work along with the women responsible for it? In 2,140 working days, 4,708 sila/litres of sesame oil could be produced (2.2 litres per day), obtained from 23,540 sila/litres of sesame seeds (yield of 20%_vol). 4,708 sila/litres of sesame oil weigh c. 4,330 kgs (weight of sesame oil is 0.915–0.923 kg/l at 20° C according to Wikipedia), 23,540 litres of sesame seeds weigh 16,007 kgs (sesame seeds: 680 kg/m³ according to http://www.click-chef.com); thus, after pressing, a maximum of 11,676 kgs waste was left over, almost double the weight of the 6083.5 kgs of oil cake in the transaction from Adalal to Būr-Mama. A relationship between the workdays and the sesame oil cake thus cannot be excluded, but no simple explanation seems possible for these numbers at the moment. Ur III sources do not provide any evidence for a ratio between sesame oil cake and sesame seeds and/or sesame oil.

5. The Technology of Oil Pressing

The tablet published as CUSAS 40/2 0428 (30/II/Amar-Suena 9) documents the transfer of the sesame oil mill to Adalal as overseer of the oil-milling women in the governor’s palace. The scribe Šarakam [individual=Šarakam] and the general Ur-Engaldudu[individual=Ur-Engaldudu] listed different sets of property in this way when the governor Urmes[individual=Urmes] returned to his office in 30/II/Amar-Suena 9 after an absence of 20 months (see above §2 and Archive Irisaĝrig).

When the sesame oil mill was handed over, the scribe and general recorded the amount of sesame seeds (2,290 sila/litres) and sesame oil cake (540 kgs) in stock at the time of the inspection, as well as the following equipment: 29 saddle querns with mullers in use (^na4ḫar[glossary=^na4ḫar] a2-da-bar[glossary=ad-bar] šu si3-ga), plus one new unused saddle mill and six old basalt saddle mills. In addition, there is an unknown number of mortars (^ĝešnaĝa4[glossary=^ĝešnaĝa4]) and pestles made of wood (ĝeš-gan〈(-na)〉[glossary=ĝeš-gan-na]), and vessels of 15 litres ([…] 0.0.1 5 sila3-ta). Two or three entries for other tools are unfortunately not preserved on the tablet (CUSAS 40/2 0428, 30/II/Amar-Suena 9).

Another tablet, which, however, does not come from Irisaĝrig but from the Umma region, records the handing over of zibi millstones with a handstone for oil pressing (CUSAS 40/2 0292, Šu-Suen 6).

The production process of sesame seeds can be deduced from the equipment of the sesame oil mill in the governor’s palace at Irisaĝrig: on the 29 saddle querns of stone (from basalt or zibi-stone), the sesame seeds are ground using the mullers. The oily mass was then pounded, with the pestles in the wooden mortars, so that the oil could be scooped out and poured into the containers provided. At the time of the inspection, there were still 2,290 litres of sesame seeds and 480 kg of sesame press cake in the workshop (for more details see the edition of CUSAS 40/2 0428). Regarding the fact that basically identical sets for oil pressing are known from the Old Babylonian period (YOS 12 342; BSA 2 180), the technology of oil pressing apparently remained the same from the 21st to the 18th centuries BCE in southern Mesopotamia. It thus can be identified as the Mesopotamian technology of oil-pressing that combines already well-known techniques: the grinding on saddle querns was practiced for grinding barley and other grains to flour, whereas mortars and pestles were used to crush onions, garlic(?) or other spices. The Mesopotamian way of oil-pressing thus combined effectively two tradtional techniques used in food production.