un-il₂ (“Menials”) as a Serflike Social Stratum during the Ur III Period

3.7.2.1 Overview

There are five kinds of employment arrangements detailed in Table 3, including conscription, penal labor, and slave labor, which are mandatory, as well as hiring and self-employment, which are voluntary.^[26] Employers were usually institutional, specifically provincial and temple, households and large, especially royal, private households with personnel, which are called administrative households for simplicity’s sake, or smaller private households, which may have included one or more slaves. The details of these employment arrangements are particularly dependent on Umma texts.

3.7.2.2 Mandatory Work: Conscription, Penal Labor, and Slave Labor

Most male citizens with the exception of individuals with certain cultic and high-ranking administrative and managerial occupations were conscripted along with all un -il ₂ .^[27] When citizens were conscripted, they typically received fifteen days off every month, unless they were conscripted for cultivation when they only had these days off for three to four months a year.^[28] It is possible that some of their workdays were actually festival days, which they may have had regularly throughout the year.^[29] Overall, they were conscripted part time and therefore experienced significantly more economic autonomy than un -il ₂ and especially slaves. un -il ₂ were conscripted full time – female un -il ₂ usually received five to six days off each month, if not more in rare circumstances, whereas male un -il ₂ received three days off each month, but only during the same months that citizens received fifteen days off. Like citizens, some of their workdays may have been festival days. It is important to note that while male citizens and un -il ₂ had fewer days off when conscripted for cultivation, they could be assigned different work from year to year, which would have limited their most-demanding work. As mentioned before, usually male citizens and un -il ₂ could be temporarily exempted from conscription to support their parents, which was probably not a possibility for those subjected to penal labor or slave labor (Pottorf 2022: 201–16).

Any individual could be subjected to minimally compensated penal labor for a variety of reasons. For example, though female citizens were not usually conscripted, they could be subjected to penal labor when their male relatives did not fulfill their obligated conscription (n. 14). These citizens would have been deprived of their economic autonomy and stability, at least temporarily. The details of slave labor in private households are not well attested, but female slaves and their children in administrative households were subjected to slave labor alongside conscripted female un -il ₂ . In these cases, they also received the same number of days off (Pottorf 2022: 208–10, 248–51).

When individuals were compensated with allotments by administrative households for mandatory work, they could have been allotted barley monthly and wool or garments annually, in addition to other commodities. Otherwise, they were allotted shares of subsistence land ([gan ₂ ] šuku) that yielded barley annually, in addition to other commodities. The typical monthly allotments of barley (in sila ₃ [∼1 L]) for male and female individuals according to their age brackets in Umma are provided in Table 4.^[30] While these allotments were typical, there were other amounts as well, and individuals subjected to penal labor could be allotted less than usual.

The vast majority of citizen men were allotted shares of subsistence land that sustained themselves and their families, and the rest were allotted barley monthly. Some citizen women who held cultic occupations or were in high-ranking families were also allotted subsistence land.^[31] Interestingly, the Nippur text AuOr 40, p. 259 10 details an ereš-dingir priestess who rented her subsistence land as tenant land. Only about half of un -il ₂ men were also allotted subsistence land for themselves and their families, whereas the rest of the un -il ₂ and all slaves were allotted barley monthly (Pottorf 2022: 180–83, 219–27). The subsistence-land sizes (in iku [∼0.35 ha]) for most citizens (excluding exceptional individuals like the governor) and un -il ₂ in Umma are visualized in Graph 1.^[32]

Graph 1:

Subsistence-land sizes (in iku) for most citizens and un -il ₂ in Umma.

Although these data are reliable to some extent, the vast majority of sizes were for individuals with unknown orders. The middle 50 percent of citizen men and un -il ₂ were allotted 4 to 12 iku and 3 to 6 iku, respectively. It is probably the case that many of the uncertain individuals allotted 3 iku were un -il ₂ and those allotted 6 iku were citizens. While it is difficult to precisely account for this, it can be roughly estimated that the middle 50 percent of citizen men and un -il ₂ were allotted 6 to 12 iku and 3 to 4 iku, respectively. As these data indicate unsurprisingly, un -il ₂ were allotted significantly less than citizens, though it is remarkable that they could be allotted as much as 36 iku. The few un -il ₂ with larger subsistence-land sizes would have had more economic autonomy than most un -il ₂ , at least in terms of managing their possessions. Although un -il ₂ were allotted about twice as less on average than citizens, this was probably because they tended to have low-compensation occupations rather than because they were un -il ₂ .

Besides determining the sizes of subsistence-land shares that citizens and un -il ₂ were usually allotted, it is important to consider their yields (in sila ₃ /iku). At Ur III Girsu-Lagaš, the average yield for subsistence land was under ∼333 sila ₃ /iku (Maekawa 1986: 116). Noemi Borrelli (2013: 27) states the following about this yield average in comparison to the higher yield average for domain land:

The cultivation of sustenance plot[s] was conducted with the equipment granted by the institutions, which provided the plowing team and the seeds. The productivity level of the šuku plots of the household of Namhani was lower than the one retrieved for the domain land and it seems to be usually lower than 20 gur per bur₃ [∼333 sila ₃ /iku]. Whether or not there was an implicit practice of assigning land with low productivity level to temple personnel is still under debate.

Though there may have been several reasons for this difference in yield averages, perhaps a significant reason is that the subsistence-land yields were probably artificial if they were calculated after factoring in cultivation costs, such as animal fodder, equipment, and seeds, among others. This means that they would not have been the actual yields of the fields but administratively reduced yields. It may have also been the case that these artificial yields were altered based on individuals’ orders, occupations, or other circumstances. For simplicity’s sake, these possibly artificial yields are hereafter referred to as simply yields. Otherwise, it would also make sense to assign more-productive land to be domain land so that some of its yield could be used to cover the cultivation costs of subsistence land. Either way, it is assumed here that the recipients of subsistence land were allotted the entirety of their reported yields without any deducted costs.

As for citizens and un -il ₂ in Umma with known subsistence-land sizes, it is possible to compare their yields, as seen in Graph 2.^[33] While the data are unevenly spread, the median yields for everyone but citizen women are 300 sila ₃ /iku or a little less, which fits well with the average yield of under ∼333 sila ₃ /iku at Ur III Girsu-Lagaš. It is difficult to detect reliable trends based on these uneven data, but it seems that citizens generally had better yields than un -il ₂ . The few citizen women with subsistence land apparently had higher yields on average, probably due to their privileged positions described above. It is possible that occupations played a significant role in whether yields were above or below the median. For example, the median yield for scribes or individuals with scribal training (Steinkeller 2017b: 53–54) was 500 sila ₃ /iku, and the highest yield was ∼774 sila ₃ /iku.^[34] They and other individuals with administrative and managerial occupations probably tended to have the highest yields overall. If the middle 50 percent of citizen and un -il ₂ men had conservative yields of 300 to 450 sila ₃ /iku and 200 to 350 sila ₃ /iku, respectively, then these citizen men could have received 1,800 to 5,400 sila ₃ of barley annually, whereas these un -il ₂ men could have received 600 to 1,400 sila ₃ of barley annually.

Graph 2:

Subsistence-land yields (in sila ₃ /iku) for most citizens and un -il ₂ in Umma.

3.7.2.3 Voluntary Work: Hiring and Self-Employment

Citizens, especially male individuals, were able to hire themselves out when they were not conscripted, but perhaps un -il ₂ and especially slaves could not do so. When citizen men were hired, they were usually paid daily wages of 6 sila ₃ in barley, though they could probably have been paid more when demand was higher for harvests or if they were hired for highly skilled work, such as craftworking (Pottorf 2022: 251–56). If they were paid 6 sila ₃ daily, it was three times the daily amount of the monthly allotment of 60 sila ₃ many men received while conscripted. Although it is difficult to know how often citizen men would have hired themselves out a year, it could have been between 30 and 120 days a year. This estimate assumes that they could have hired themselves out half a thirty-day month minus five days for festival days and days off for as few as three months or as much as a whole twelve-month year.^[35] If so, they could have been paid between about 200 and 750 sila ₃ annually, which are both rounded up to account for possibly higher wages in some cases. It is not certain how many male citizens per family hired themselves out and what wages younger individuals may have earned, so this is probably a conservative estimate. It is possible that female citizens could hire themselves out, but this needs further study.

Self-employment is significantly less attested than the other employment arrangements, given the nature of the evidence, though it was ubiquitous. Female citizens were usually self-employed when they engaged in crucial domestic work, including childcare, cooking, and textile work, among other duties. Although it was not often profitable, it did result in uncompensated benefits. While individuals from other orders could engage in domestic work for themselves, they would have been far more restricted by their limited time and resources. Female un -il ₂ and slaves could have used some of their days off for domestic work, especially since they had more days off than male un -il ₂ . Female citizens could also be self-employed in various occupations, such as midwives, physicians, priestesses, sex workers, and tavern keepers, among others (Steinkeller 2022).^[36] Male citizens could be self-employed as merchants, for example. The profits from any of these occupations are difficult to ascertain, however.

Perhaps the best-attested form of self-employment was renting tenant land ([gan ₂ ] apin-la ₂ ), which required tenants to invest their own resources in order to gain profits that generally amounted to about maybe a third or nearly half their tenant-land yields. Tenant land producing various kinds of crops or for pasturage could be rented, but barley was the most significant and assumed in this discussion.^[37] Most individuals who rented tenant land were citizen men, but some citizen women could rent tenant land. There is also one possible instance of an un -il ₂ renting tenant land. In Orient 21, p. 2 obv. iv 19′–20′ and rev. iv 10–11, there is an animal fattener named Bida who rented 1 iku of tenant land yielding 200 sila ₃ /iku before deducting cultivation costs, rent, and taxes and who was allotted 9 iku of subsistence land yielding 200 sila ₃ /iku. Due to the text's damage, it is possible that he rented more tenant land. This Bida appears to have been an un -il ₂ based on OrSP 47–49: 483 and probably CUSAS 39: 129. While this singular instance of a possible un -il ₂ renting tenant land indicates that other un -il ₂ may have been able to rent tenant land, this individual was allotted more subsistence land than most un -il ₂ , and he rented a small amount of tenant land with a below-average yield according to the extant text.

Although it is difficult to know the full sizes of tenant-land plots because individuals could rent tenant land from multiple areas, their known minimum sizes with regard to barley land for citizens and un -il ₂ are given in Graph 3.^[38] Whereas the order of most tenants is not certain, they were overwhelmingly citizens. The minimum sizes of 0.25 iku were especially small and were likely pieces of larger rented amounts. While the middle 50 percent for the total amount is 1.5 to ∼13 iku, it can be rounded up to at least 2 to 14 iku since these amounts were based on known minimums. Unsurprisingly, individuals with occupations that were better compensated during conscription tended to rent more, but even individuals with low-compensation occupations could rent tenant land (Pottorf 2022: 262).

Graph 3:

Minimum tenant-land sizes (in iku) of barley land for citizens and un -il ₂ in Umma.

As for tenant-land yields, they are difficult to calculate given the variety of texts. A few texts detail tenant-land yields prior to the deduction of cultivation costs presumably as well as any rent paid in barley and irrigation tax paid in silver, which are given in Graph 4.^[39] Though there are fewer data for these yields than for subsistence-land yields, tenant-land yields were typically larger. The middle 50 percent of the total yields ranges from 240 to 485 sila ₃ /iku, whereas the middle 50 percent for all subsistence land is 200 to 400 sila ₃ /iku. The average yield for citizen women is similarly larger than for all others, again perhaps due to their privileged positions. The yield for the one known un -il ₂ is unsurprisingly low. While it is possible to speculate about the motives of landlords and tenants in improving tenant-land productivity, this requires further study to account for other factors, such as annual variances. More importantly, these yields did not factor in cultivation costs, but it is not known who cultivated tenant land and how its costs were paid. Although it is admittedly simplistic, perhaps tenant-land yields were virtually equal to or just a little higher than subsistence-land yields after deducting cultivation costs. For example, the cost of seeds and animal fodder may have been about 25 sila ₃ /iku, which were not the only costs, of course.^[40]

Graph 4:

Tenant-land yields (in sila ₃ /iku) of barley land for citizens and un -il ₂ in Umma.

In addition to the few texts that document tenant-land yields before the deduction of the barley rent and silver tax, there are a few texts that record the barley rents of tenant land according to whether various plots had silver taxes or not. Most texts involving tenant land from Umma do not specify whether silver taxes were paid or not unfortunately, but the barley amounts appear to be rents. For the sake of brevity, the middle 50 percents of the barley-rent rates in all these texts are provided in Table 5.^[41]

The middle 50 percent of barley-rent rates for tenant land with silver taxes was much higher than those for tenant land without silver taxes because the latter was less productive. When there were silver taxes, they ranged from 10 to 20 še (∼0.46 to ∼0.92 g) per iku. The middle 50 percent for the barley-rent rates of all tenant land is 76.25 to ∼230 sila ₃ /iku, which fit well with the yields of 240 to 485 sila ₃ /iku since the barley-rent rates at Ur III Girsu-Lagaš tended to range between one third and one half the yields (Steinkeller 1981: 126–27). If the yields of the middle 50 percent were about 200 to 400 sila ₃ /iku after cultivation costs, the profits before tax for the middle 50 percent could range from roughly 125 to 175 sila ₃ /iku. If the silver-tax range is averaged to 15 še/iku, it could be equivalent to about 25 sila ₃ /iku based on the typical equivalency of 1 gin ₂ of silver to 300 sila ₃ of barley (Cripps 2017). While this silver tax would only apply for some of the tenant land, it can be conservatively applied to the entire profit range established above, resulting in a range from 100 to 150 sila ₃ /iku. Moreover, it is important to note that the rent may have varied beyond the estimates here in various circumstances regarding the relationship between the landlord and tenant. Given the adjusted middle 50 percent of 2 to 14 iku, the middle 50 percent of tenants perhaps earned profits of 200 to 2,100 sila ₃ of barley annually. The larger end of this range may have required further cultivation costs to account for hired workers, however. Given the conservative estimates here, this amount is still plausible perhaps. Overall, the nearly exclusive access that citizens had to voluntary work granted them significantly more economic autonomy and stability than un -il ₂ and especially slaves.