¹ Unfortunately, although many such tablet archives exist, most were excavated before modern archaeological techniques allowed the recording of the details of their contexts. In spite of the inadequacies of some of its strategraphic and findspot information, the recent publication of the Old Babylonian levels at Ur (UE VII) may enable such a study, as will the tablets from the Old Babylonian structures at Isin when they have been fully published ( Hrouda, B., ed., Isin-Išan Bahrīyāt I (1977)Google Scholar).

² With the exception of the archive partially published by Goetze, A. (“The Archive of Atta from Nippur”, JCS 20 (1966))Google Scholar, all of the contracts from this more recent series of excavations remain unpublished. A catalogue is included in Stone, E., The Social and Economic Organization of Old Babylonian Nippur (University of Chicago unpublished Ph.D. Dissertation, 1979)Google Scholar, and a full publication is currently in preparation. The archaeological remains from TA have been published in OIP 78 (1967)Google Scholar.

³ Great discrepancies exist between the stratigraphy of TA as it was recorded in the field and the stratigraphy as it was published by McCown, and Haines, in OIP 78 (1967)Google Scholar. A careful examination of both the published and unpublished stratigraphy suggests that the records made in the field may be more reliable than those published. My forthcoming publication of the TA and TB contracts will be accompanied by a re-evaluation of the TA stratigraphy. Figs. 1–3 are therefore based on the more correct field records, rather than on the OIP 78 publication. The reader is referred to E. Stone, Social and Economic Organization of Old Babylonian Nippur for a more detailed discussion of the stratigraphic problems encountered in working with the TA material.

⁴ A later pit destroyed much of this wall, but its stubs still remained.

⁵ Thureau-Dangin, F., “Numération et métrologie sumériennes”, RA 18 (1921)Google Scholar, and “Le ‘Grain’ mesure de surface”, RA 35 (1937)Google Scholar.

⁶ For consistency, I have used the numbers given to the tablets in the field. The tablets themselves are housed in the Oriental Institute in Chicago, and in the University Museum in Philadelphia. I must express my gratitude to the Director of the Oriental Institute, John A. Brinkman, and to the Curator of the Babylonian Collection at the University Museum, Åke Sjöberg, for kindly allowing me to examine these tablets. The museum numbers of the tablets mentioned in this paper are as follows: Oriental Institute: 3N-T 85 = A30138, 3N-T 83 = A30139, 3N-T 92 = A30140, 3N-T 93 = A30141, 3N-T 94 = A30142; University Museum: 3N-T 86 = UM 55-21-238, 3N-T 87 = UM 55-21-239, 3N-T 91 = UM 55-21-243.

⁷ The plan of House K seen in Fig. 1 is taken from the earlier levels of the house because at the time of the occupation of House I, House K seems to have been largely abandoned as a residential structure.

⁸ The room next to the front door served as an entrance chamber and acted, we assume, as the locus of external interactions. Since the eldest son was the head of the family, we might expect him to have received the entrance chamber as his preference portion to enable him to represent the rest of the family in all such interactions with outsiders.

⁹ Dates used in this article follow the Middle Chronology as presented by Brinkman, J. in Oppenheim, A. L., Ancient Mesopotamia (1964)Google Scholar.

¹⁰ See Stone, E. in Levine, L. and Young, T. C. Jr, eds., Mountains and Lowlands (1977)Google Scholar.

¹¹ The text records the plots as equal in size.

¹² An examination of the field notes for TA discloses that several bricks and stones were found in the vicinity of the doorway between loci 157 and 173. These were interpreted by the excavators as the possible remains of a step, but since they were clearly found in a disturbed condition, they may represent the remains of a blocking to that door. The field notes record the north-west jamb of a door between loci 152b and 179b, but suggest that this “might be where later construction abutted earlier walls”. In both instances, although the evidence is ambiguous, the postulated door blocking is not ruled out.

¹³ By “extended family” I refer to a situation where two or more families share a single structure. In other archives (for instance BE 6/2 31, 32, 33, 34) we find all but one of the brothers selling their shares immediately upon inheritance, suggesting that in those cases, nuclear family residence was being practised.

¹⁴ The Nippur contracts make it clear that sons only gained independence at their father's death, when they inherited his real property; until then the father exercised absolute economic control.

¹⁵ It is difficult to tell the age of the owner of this house; it is possible that he had adult children who may, or may not, have been sharing the house with him. 3N-T 88 shows that some 45 years earlier his father, Ur-Bau, was allocated a share of a temple office. Unfortunately, since this was a unique text for Nippur and since we neither know Ur-Bau's age at the time of the text, nor the usual age of marriage, this information is difficult to assess (see footnote 34).

¹⁶ The houses in question are E, F, G, H and I (see Fig. 1).

¹⁷ See E. Stone, Social and Economic Organization of Old Babylonian Nippur.

¹⁸ The houses in question are C, D and Z (see Fig. 4).

¹⁹ The plans of the TB houses show great similarity to the houses found in TA (e.g. the incompletely excavated House K) and also to the house plans found drawn on tablets (see OIP 78, plate 52).

²⁰ For the houses from TA and TB, only those areas thought to have been roofed were measured.

²¹ TA loci 180, 181, 182/205, 154, 173, 178 and 152a were designated as living rooms (see Fig. 1), as were TB loci 109, 158, 123, 105, 57, 136, 500, 175 (see Fig. 4). Only TB locus 175 seems by its location to have been perhaps no living room. Perhaps one of the other loci in this house, loci which were almost as large, actually served this purpose.

²² No ethno-archaeological studies on mud-brick architecture have been conducted in southern Mesopotamia, and the few available ethnographies do not include the architectural and metric data necessary for this study. The villages studied in Iran, however, share significant traits with both ancient and modern settlements in Mesopotamia; traits which make comparison worthwhile. In the first place, significant cultural distinctions do not exist between modern Iran and Iraq; in spite of language differences they share a common social organization. Secondly, although the materials available for roofing may vary slightly, mud-brick is the major building material in both areas.

²³ Kramer, C. in Kramer, C., ed., Ethnoarchaeology (1979)Google Scholar, and Watson, P. in Redman, C. et al. eds., Social Archaeology (1979)Google Scholar.

²⁴ See Berry, B. et al., “Urban Population Densities: Structure and Change”, Geog. Rev. 53 (1963)CrossRefGoogle Scholar. It may be, however, that ancient Mesopotamian cities experienced density patterns different from their modern counterparts since their city walls would have limited expansion.

²⁵ Although textual reference to second storeys is attested for northern Babylonia, closer to sources of beams for construction ( Harris, R., Ancient Sippar (1976)Google Scholar), none exist for southern Babylonia. In addition, those structures at Nippur which contain stairways (e.g. TA House K, see Fig. 1) have walls which are no thicker than the walls of the structures without staircases, a mere 60 cm, much less than the one metre thick walls usually used today for second-storey construction. Since we know that access to the flat roofs was as important in the past as it is today, it seems reasonable to interpret the internal staircases as providing access to the roof, rather than to a second storey. These comments are probably as valid in the interpretation of the Old Babylonian house structures at Ur (see UE VII) as of those from Nippur.

²⁶ See E. Stone, Social and Economic Organization of Old Babylonian Nippur; and Greengus, S., “Old Babylonian Marriage Ceremonies and Rites”, JCS 20 (1966)Google Scholar, and “The Old Babylonian Marriage Contract”, JAOS 89 (1969)Google Scholar.

²⁷ Not only have few Old Babylonian burials been excavated, but there is evidence to suggest that differential treatment was given to adults as opposed to infants.

²⁸ The material from Hasanabad (P. Watson, in Social Archaeology) was collected in 1960, before Iran's thrust towards modernization. Kramer chose Shahabad (Kramer, in Ethnoarchaeology) in part because it was sufficiently remote and inaccessible that modernization had yet to arrive.

²⁹ Watson's data (Watson, in Social Archaeology) was based on actual on-the-spot room measurements, while Kramer's was based on her field plan, drawn at 1:250 (Kramer, in Ethnoarchaeology). In spite of the fact that we can determine the average width of the walls at Shahabad from the detailed plans of two of the houses ( Kramer, , in Ethnoarchaeology, Figs. 5:3, 5:4 Google Scholar), we cannot use this information to subtract the area of the walls since the larger and squarer the room, the less area taken up by the walls. For this reason, we must simply accept the fact that some portion of the greater size of the Shahabad houses is due to the inclusion of the wall thicknesses.

³⁰ Such a statement is in direct contrast to the assertions of Naroll (“Floor Area and Settlement Population”, American Antiquity 27 (1962)Google Scholar) and LeBlank, (“An addition to Naroll's suggested Floor Area and Settlement Population Relationship”, American Antiquity 36 (1971)Google Scholar). It must be stressed, however, that what may be considered insignificant variation in cross-cultural studies, may be significant variation within the context of a single culture area.

³¹ The area of the house, 52·21 m² divided by mean space per person, 5·31 m².

³² The number of living rooms (3) multiplied by the mean number of people per family (4·35).

³³ See footnote 14.

³⁴ He would be of an age to have had all of his children. If modern marriage patterns of the Middle East pertain to ancient Mesopotamia, he would probably have married in his late twenties to a woman in her teens ( Gulick, J., The Middle East: An Anthropological Perspective (1976), 183 Google Scholar). At forty-five, then, both he and his wife would have had about fifteen child-bearing years.

³⁵ By family, I mean nuclear family, the unit that shares a single living room.

³⁶ These figures are obtained from Watson, , in Social Archaeology, Table 7:11 Google Scholar.

³⁷ See Watson, , in Social Archaeology, 132 Google Scholar.

³⁸ To determine the number of people expected to occupy the 60 rooms (46 rooms excluding stables) recorded in Watson, , in Social Archaeology, Table 7:1 Google Scholar, I took the amount of living space recorded (455 m²) and divided it by the figure of 7·3 m² of living space per person that Watson had calculated (p. 137). This gave me an estimated population of 62 people occupying the 60 rooms which she had measured.

³⁹ Information derived from Kramer, , in Ethnoarchaeology, Table 5:1 Google Scholar. Note that the measured rooms come from 62 houses, not the total of 67 (see Table 5:2, note g; Table 5:1, note a; and p. 160 note 6).

⁴⁰ The population of the 62 structures recorded in Kramer, , in Ethnoarchaeology, Table 5:1 Google Scholar, can be calculated by taking the mean household size for landed families, 6·8 (Table 5:2, note c), multiplied by the number of landed families concerned, 29 (Table 5:2, note g), added to the mean household size for landless families, 5·7 (Table 5:2, note d) multiplied by the number of landless families concerned, 33 (Table 5:2, note g). This provides a total of 385 persons. Since Kramer indicates that each nuclear family occupies a single living room (pp. 147, 155), I divided the population, 385, by the number of living rooms, 90, to determine the number of people per family.

⁴¹ In this instance, I took the total area of the rooms (excluding stables) provided by Kramer, , in Ethnoarchaeology, Table 5:1 Google Scholar, and divided it by the population, 385 (see note 40).

⁴² The estimated family size used here is the mean of those derived from Kramer's and Watson's data (see notes 37 and 40). In order to estimate total population, I assume one nuclear family per living room (see note 40).