Some of the conclusions of this article were presented at the 196th Meetings of the American Oriental Society, held in New Haven, March 1986.

Unpublished texts utilized here are published by permission of the Trustees of the British Museum, London, and by permission of the Visitors of the Ashmolean Museum, Oxford; editions will be found in my forthcoming study of Neo-Babylonian matrimonial property. Source abbreviations follow those of the Assyrian Dictionary of the University of Chicago (hereafter cited as CAD) (Chicago and Glückstadt: The Oriental Institute and J. I. Augustin, 1956–Google Scholar).

In accordance with standard Assyriological conventions, portions of a text placed within square brackets are no longer preserved on the cuneiform tablet, and women's personal names are marked by a supers0cript letterf. Akkadian is set in italics, Sumerian in letter-spaced roman type, and logograms in small roman capitals.

Source citations include the place and year of drafting. For conversions of the Babylonian dating system to B.C. dates, I have relied on Parker, R. A. and Dubberstein, W. H., Babylonian Chronology 626 B.C.-A.D. 75, Brown University Studies, no. 19 (Providence: Brown University Press, 1956Google Scholar). In the Babylonian calendar, the year began in the spring; therefore, for example, when no month is preserved, “217/16 B.C.” is used to indicate the ninety-fifth year of nthe Seleucid Era, which began in the spring of 217 B.C. and ended in the spring of 216 B.C. When no year appears on a tablet, but the name of the ruler is preserved, the dates of his reign are given; for example, “555–22 B.C.” indicates that the tablet was drafted during the reign of Nabonidus. The notations “n.p.” (no place) and “n.d.” (no date) are used when the ancient scribe omitted the pertinent information or when it is no longer preserved; in the latter case, the notations are placed within square brackets.

The following notations are used to express familial relationships: B ═ brother, D ═ daughter, F ═ father, H ═ husband, M ═ mother, S ═ son, W ═ wife, Z ═ sister. Combinations of these notations are to be read in possessive construction, thus HF ═ husband's father, WMM ═ wife's mother's mother, etcetera.

¹ For more detailed discussion of some of these points, see Fortes, M., “Age, Generation, and Social Structure,” in Age and Anthropological Theory, Kertzer, D. I. and Keith, J., eds. (Ithaca: Cornell University Press, 1984), 99–122Google Scholar. The essays in Age and Anthropological Theory all address different questions and approaches; for my purposes, I found particularly germane the articles by Maybury-Lewis, D., “Age and Kinship: A Structural View,” 123–40Google Scholar; Hammel, E., “Age in the Fortesian Coordinates,” 141–58Google Scholar; and Keith, J. and Kertzer, D. I., “Introduction,” 19–61Google Scholar.

The problem of reconciling life stages as defined by observable outward signs with chronological ages appears in discussions in the Babylonian Talmud concerning the age of majority. According to the Rabbis, the observable sign of the onset of puberty is the appearance of two pubic hairs. The Babylonian Talmud records arguments in which an attempt is made to reconcile the observable sign with the chronological ages of majority, twelve years and one day for females and thirteen years and one day for males, and the problems that arise when these signs are not observable by the appropriate ages. (See Bab. Niddah 46a–47b and 48a–49a.) A legend concerning Rabbi Eleazar ben Azariah (who lived in the first and second centuries after Christ), illustratesn the importance of observable signs for fulfilling cultural role expectations. According to the legend (Bab. Berakhot 28a), Rabbi Eleazar was elected patriarch (nasi) of the Academy when he was only eighteen years old. Although he possessed the intellectual and academic credentials, his youthful appearance defied his eminent position. Therefore, in order to reconcile his appearance with his new role, his hair miraculously turned white overnight.

² See the discussions in Driver, G. R. and Miles, J.C., The Assyrian Laws (Oxford: Oxford University Press, 1935), 182ffGoogle Scholar.; and in Cardascia, G., Les lots assyriennes (Paris: Les Editions du Cerf, 1969), 21 IffGoogle Scholar., and bibliography. Note that the right (or, perhaps better, the obligation) of the widow's husband's son by another marriage to marry her is not contingent upon her own marriage not having been consummated or upon the marriage having ended childless. MAL A § 46, in discussing the fate of the widow, indicates that the obligation of her own sons to support her is voided should one of her husband's sons take her in marriage. The minimum age of ten for performance of levirate is also found in the Babylonian Talmud, Bab. Niddah 45a.

³ STT 400 r. 47–48, published in O. R. Gurney and Hulin, P., The Sultantepe Tablets (London: The British Institute of Archaeology at Ankara, 1964), IIGoogle Scholar; see Nougayrol, J., Revue dassyriologie et d’ archéologie orientate, 62:1(1968), 96Google Scholar; and Sjöberg, Å.W., “DerExamenstext A,” Zeitschrift für Assyriologie, 64:2 (1974), 137–76, at 164Google Scholar. The six decades in this compilation refer to longevity aspirations; there is no comment upon the activities or social roles of one who reaches any of these ages. For an Egyptian composition reminiscent of the seven ages of man in Shakespeare's As You Like It, see Insinger, Papyrus, sixteenth instruction, 17, 11–18, 3Google Scholar, inLichtheim, M., Ancient Egyptian Literature (Berkeley: University of California Press, 1980), III, 198fGoogle Scholar.

⁴ Edel, E., Ägyptische Ärtze und dg00E4;yptische Medizin am hethitischen Kōnigshof (Göttingen: Westdeutscher Verlag, 1976), 68ffGoogle Scholar.

⁵ This letter informs us that one Hittite royal woman lived to the age of sixty, and that at that age she needed to produce an heir; we certainly cannot utilize this letter to infer anything about age at marriage or about childbearing age. For a summary of the problems of the namef Matanazi (f Maššananzi-IR), see Edel, Ågyptische Årtie, 32–36. ^f Matanazi had been given in diplomatic marriages twice by her brother Muwatalli, the second marriage having been concluded only by her nephew Muršili III; see Ph.Cate, H. J. Houwink ten, “The Early and Late Phases of Urhi- Tesub's Career,” in Anatolian Studies Presented to Hans Gustav Güterbock, Bittel, K., Ph. Cate, H. J. Houwink ten, and Reiner, E., eds. (Istanbul: Nederlands Historisch-Archaeologisch Instituut, 1974), 123–150Google Scholar, at 127–29, and note 29. I am grateful to Prof. H. G. Güterbock for discussing with me the problem of^f Matanazi's marriages.

⁶ An age at marriage for seventh-century Neo-Assyrian princes has been assumed by Parpola, S., Letters from Assyrian Scholars to the Kings Esarhaddon and Assurbanipal, Part II (Kevelaer: Butzon und Bercker, 1983), 231Google Scholar, note 390: “If it is accepted that Assyrian princes must have been at least 20 years old before they could be introduced into the Succession Palace, and that they married at approximately the same age (cf. ABL 308).… ” Although not implausible, this is a working assumption and is not documentable insofar as I have been able to determine. (ABL 308 informs us that Assurbanipal sired a daughter while he was still crown prince—but not at what age; the pertinent passage of ABL 308 is cited in CAD Š s.v. šarru in mārat šarri usage e, and the entire letter is translated in Oppenheim, A. Leo, Letters from Mesopotamia (Chicago: University of Chicago Press, 1967), 158, No. 97.Google Scholar) Further, the elite classes could certainly display marriage and life-stage patterns different from those of the population at large (note the case of the Hittite princess above).

⁷ I know of one Old Babylonian “birth certificate,” YOS 13 192, published in Finkelstein, J. J., Late Old Babylonian Documents and Letters, Yale Oriental Series, Babylonian Texts, no. 13 (New Haven: Yale University Press, 1972), with an edition and discussion on pp. 14–16Google Scholar. The text was drafted upon the occasion of a slave's parturition, and its purpose appears to have been to reconfirm her status as slave of her mistress.

⁸ An exception to the statement that age at death is almost never known concerns ^fAdadguppi’, mother of the Neo-Babylonian ruler Nabonidus; she claimed to have lived 104 years; see Gadd, C. J., “The Harran Inscriptions of Nabonidus,” Anatolian Studies, 8 (1958), 69fGoogle Scholar. A unique Old Babylonian “death certificate” is Dalley Edinburgh 21, published by Dalley, S., A Catalogue of the Akkadian Cuneiform Tablets in the Collections of the Royal Scottish Museum, Edinburgh, Art and Archaeology, no. 2 (Edinburgh: Royal Scottish Museum, 1979Google Scholar). The text simply records that a woman died on the twelfth day of the month and was buried on the thirteenth; no additional information, such as her age, is included.

⁹ For some observations about Near Eastern literary and religious traditions of longevity, see Ackroyd, P. R., “Two Old Testament Historical Problems of the Early Persian Period,” Journal of Near Eastern Studies, 17:1 (1958), 13–27Google Scholar, at 23ff.; Borger, R., “An Additional Remark on P. R. Ackroyd,” Journal of Near Eastern Studies, 18:1(1959), 74Google Scholar; and, most recently, Malamat, A., “Longevity: Biblical Concepts and Some Ancient Near Eastern Parallels,” in Vortrāge Gehalten auf der 28. Rencontre Assyriologique Internationale in Wien 6.–10. Juli 1981, Archiv für Orientforschung Beiheft, no. 19 (Horn, Austria: Berger und Söhne, 1982), 215–24Google Scholar. See also Dandamaev, M. A. (Dandamayev), “About Life Expectancy in Babylonia,” in Death in Mesopotamia, Mesopotamia, no. 8, Aister, A., ed. (Copenhagen: Akademisk Forlag, 1980), 183–86Google Scholar; that essay draws upon data from detailed studies of 2,997 scribes and 1,234 privately owned slaves, which Dandamaev published in two monographs: Pists y Vavilonskie [Babylonian scribes in the first millennium B.C.] (Moscow: NAUKA, 1983) (with English summary, pp. 235–42,also published as “The Social Position of Neo-Babylonian Scribes,” in Gesellschaft undKultur im alien Vorderasien, Klenge, H.l, ed. (Berlin:Academie-Verlag, 1982), 35–39Google Scholar), and Stavery in Babylonia, rev. ed., Powell, V. A., trans., Powell, M. A. and Weisberg, D. B., eds. (DeKalb: Northern Illinois University Press, 1984Google Scholar). In “Life Expectancy,” Dandamaev cites ten scribes with active careers spanning thirty-eight to sixty-eight years, and five slaves who are attested in documents dating over the course of from thirty-four to fifty-one years. Dandamaev assumes that the slaves were all “adults” when first attested, and implies that many more scribes than those mentioned had long active careers. He concludes that life expectancy in first-millennium Babylonia was “relatively high” (p. 185). Certain reservations must be expressed about these conclusions, however. First, we are unable to determine the lowest chronological age at which a scribe generally began his career, or that at which a slave would be considered a fully able-bodied worker (that is, an “adult”). Estimates of ages at death of the persons mentioned by Dandamaev, therefore, can be imprecise by a decade or more. Second and most important, Dandamaev's own qualifications (“many scribes lived, at least 70–90 years” and “at least, some slaves lived 60– 80 years”) must be stressed. My own impression is that the examples adduced represent the exceptions rather than the norms, and that the scribes and slaves mentioned are, if not all, then certainly almost all, of those with careers of any long duration. It is nearly inevitable that persons who lived longer and more active lives would be prominent in our sources, and the appearance of a few notable people living long lives cannot be indicative of a general demographic trend. Another approach, yielding negative results, is taken by Snell, D. 1982 work, “Plagues and Peoples in Mesopotamia,” Journal of the Ancient Near Eastern Society, 14 (1985), 89–96Google Scholar, which looks at the lengths of kings' reigns in an attempt to find an indication of cycles of infectious disease affecting longevity patterns.

¹⁰ The terms family and household are inherently vague, and can be used in many different ways. In this study, I use household to refer to the domestic group with whom one resides, including the people to whom one is related by blood or marriage (the common definition of family) as well as unrelated coresident persons. The simple (or nuclear) family household is one that has one or no conjugal couple, with or without children; the extended family household includes, in addition to the nuclear family, members of the preceding generation (widowed parents) or of the conjugal couple's own generation (unmarried sisters or brothers); multiple family households (stem or joint households) include more than one (usually related) conjugal couple. See Wall, R., “Introduction,” in Family Forms in Historic Europe, Wall, R., Robin, J., and Laslett, P., eds. (Cambridge: Cambridge University Press, 1983)Google Scholar, 6ff., where further criteria are elaborated.

¹¹ In addition to the references to works in historical demography cited in the preceding and following notes, see the articles in Netting, R., Wilk, R., and Amould, E., eds., Households: Comparative and Historical Studies of the Domestic Group (Berkeley: University of California Press, 1984Google Scholar); American Behavioral Scientist, 25:6 (1982)Google Scholar; Kramer, C., ed. Ethnoarchaeology: Implications of Ethnology for Archaeology (New York: Columbia University Press, 1979)Google Scholar; and idem, Village Ethnoarchaeology: Rural Iran in Archaeological Perspective (New York: Academic Press, 1982Google Scholar). For an attempt to correlate textual and archaeological evidence, see Stone, E., “Texts, Architecture, and Ethnographic Analogy: Patterns of Residence in Old Babylonian Nippur,” Iraq, 43 (1981), 19–33Google Scholar. For late-third-millennium Mesopotamia, see Gelb, I. I., “Household and Family in Early Mesopotamia,” in State and Temple Economy in the Ancient Near East, Orientalia Lovaniensia Analecta, no. S, Lipinski, E., ed. (Leuven: Departement Oriëntalistiek, 1979), 1–97Google Scholar.

¹² The broad distinction between Western and Eastern models was formulated by Hajnal, J., “European Marriage Patterns in Perspective,” in Population in History, Glass, D. V. and Eversley, D. E. C., eds. (London: Edward Arnold, 1965), 101–43Google Scholar; and idem, “Two Kinds of Preindustrial Household Formation System,” in Family Forms, Wall, , Robin, , and Laslett, , eds., 65–104Google Scholar. Laslett, P., “Family and Household as Work Group and Kin Group: Areas of Traditional Europe Compared,” in Family Forms (pp. 513–63Google Scholar), subdivides the Western and Eastern models into two groups each; see especially the table of features at pp. 526–27. The division of the Western model into West and West/Central groups is distinguished primarily by the addition of workers or servants to the household and the correspondingly higher number of adult workers in the household of the West/Central group. The distinction within the Eastern model of Mediterranean and East groups is further elaborated below.

¹³ This older-groom/younger-bride marriage pattern was also the dominant pattern in ancient Rome; see Sailer, R., “Men's Age of Marriage and Its Consequences in the Roman Family,” Classical Philology, 82:1 (1987), 21–34Google Scholar; and Shaw, B. D., “The Age of Roman Girls at Marriage: Some Reconsiderations” (Paper presented to the Annual Meeting of the American Ancient Historians, Pittsburgh, 05 1985Google Scholar). Both studies revise some of the conclusions of Hopkins, M. K., “The Age of Roman Girls at Marriage,” Population Studies, 18 (1965), 309–27Google Scholar. I wish to thank both Shaw and Sailer for sharing their prepublication studies, and particularly Sailer for discussing with me many of the issues raised here. Combined with life-expectancy simulations, this marriage pattern in Rome has at least two important historical consequences. First, less than one fourth of the male population would have had a living father at the time of their first marriages: as a result the practical, as opposed to theoretical, effects of patria potestas are considerably minimized. Second, the proportion of women in noa-manus marriage remaining under the potestas of their fathers must be reconsidered in light of the conclusion that no more than about half of the female population would have had a living father at the time of their marriages, and these women would have been freed in rapidly increasing numbers from this potestas by their fathers' deaths. (See Table 8.)

¹⁴ The material is studied in my forthcoming monograph on Neo-Babylonian matrimonial property.

¹⁵ Two of the grooms in the marriage agreements are former slaves, manumitted and given in marriage to free girls. One bride is a princess, married to a high temple official.

¹⁶ The Assyrian Doomsday Book texts were originally edited by Johns, C. H. W., An Assyrian Doomsday Book or Liber Censualis of the District Round Harran in the Seventh Century B.C. (Leipzig: J. C. Hinrichs'sche Buchhandlung, 1901)Google Scholar; they were recently re-edited by Fales, M., Censimenti e catasti di epoca neo-assira (Rome: Istituto per l’Oriente, 1973Google Scholar). It is certain that these texts were compiled for the Nineveh central authority, but their precise purpose is never explicitly indicated. Postgate, J. N., in a review of Fales's edition, “Some Remarks on Conditions in the Assyrian Countryside,” Journal of Economic and Social History of the Orient, 17:3(1974), 225–43Google Scholar, and in his Taxation and Conscription in the Assyrian Empire, Studia Pohl series maior 3 (Rome: Biblical Institute Press, 1974), 28–39Google Scholar, advances the theory that the lands listed were exempt from taxation. See also the review by S., , “A Note on the Neo- Assyrian Census Lists,” Zeitschrift für Assyriologie, 64:1(1975), 96–115Google Scholar, for further observations about the nature and dating of these documents; and see most recently the brief comments by Fales, M. in “The Assyrian Period,“ in Circulation of Goods in a Non-Palatial Context in the Ancient Near East, Archi, A., ed. (Rome: Edizioni dell’Ateneo, 1984), 207–20, esp. 212fGoogle Scholar.

¹⁷ For age-group terminology utilized in the Middle Babylonian period, see Brinkman, J. A., –Sex, Age, and Physical Condition Designations for Servile Laborers in the Middle Babylonian Period, ” in Zikir Šumim, Assyriological Studies Presented to Kraus, F. R., Driel, G. van et al. , eds. (Leiden: E. J. Brill, 1982), 1–8Google Scholar.

¹⁸ The agent acting for the bride is lost in OECT 10 313 ([n.p., n.d.]), BM 66005 ([n.p., n.d.]), and BM 84127+ (Borsippa, 304 B.C.).

¹⁹ VAS 6 3 (Babylon, 624 B.C.); Liverpool 8 (Opis, 564 B. C ); Ner. 13 (Babylon, 559 B.C.); TuM 2–3 1 (Borsippa, 550 B. C ); Nbn. 243 (Babylon, 549 B. C ); OECT 10 110 (Ufcudu, 544 B. C ); Cyr. 183 (Sippar, 535/4 B. C ); ASS. Bab. A. 129 (Borsippa, 527–22 B.C); BM 82609 (Borsippa, 520 B. C ); Dar. 301 (Babylon, 512–486(?) B. C ); BM 76029 (n.p., 433 or 373 B. C ); CT 49 193 (Babylon, 2 8 1 - B. C ); BM 76968/72 (Borsippa, 203 B. C ); the father and mother were the bride's agents in TBER 78 AO 26775 ([n.p., n.d.]).

²⁰ The bride's mother contracted her daughter's marriage in Nbk. 101 (Babylon, 592 B. C ); VAS 6 61 (Babylon, 555 B. C ); VAS 6 95 (Borsippa, 556–539 B. C ); BM 61984 ([n.p.], 543/2 B.C); OECT 10 130 (Larsa, 523 B. C ); TUM 2–3 2 (Borsippa, 494 B. C ); OECT 9 73 (Kish, 330- B. C ). The bride's mother and sister (or her maternal grandmother and mother?) contracted the marriage in BM 70235 (Sippar, 548 B. C ). The bride's mother and brother contracted the marriage in BM 54158 (Dilbat, 634 B. C ); YOS 6188 (Alu-sa-Lane, 542 B. C ); BM 65149 ([n.p., n.d.]). The bride's brother(s) contracted her marriage in BM 61176 (n.p., 584 B. C ); RA 25 81 No. 23 (Neirab, 555–39 B. C ); Nbn. 990 (Babylon, 540 B. C ); BM 82597 (Kutha, 464–338 B. C ); VAS 6 227 ([n.p.], 217/16 B. C ); CT 49 167 (Babylon, 2 8 1 - B. C ); BM 50106 ([n.p., n.d.]); TBER 93f. ([n.p., n.d.]); CT 49 165 (Babylon, 2 8 1 – B. C ); JCS 1 350 No. 3 ([Kish?], [n.d.]); BM 76202 ([n.p., n.d.]). The bride was given in marriage by her brother but herself promised her dowry in BM 61434 (Babylon, 543 B. C ); she acted completely on her own behalf in BM 50149 (Sippar, Sin-sarislcun yr. 2) and BM 83249 ([n.p., n.d.]).

²¹ If, in all three marriage agreements that do not preserve the identities of the brides' agents (see note 18), the agents were the brides' fathers, then seventeen out of forty-two brides (40.5 percent) would have had living fathers at the time their marriages were contracted; if the agents were the brides' brothers and/or mothers, then fourteen out of forty-two brides (33.3 percent) would have had living fathers.

²² The role of the groom's father in contracting his son's marriage and in controlling the dowry brought into his household by his daughter-in-law is explored in detail in my forthcoming study of Neo-Babylonian matrimonial property. That the lack of parental consent could void a groom's attempted independent actions in contracting a marriage is demonstrated by Cyr. 312 (Babylon, 531 B. C ), which records the invalidation of the marriage of^f Tablufu, contracted on her behalf by her brother, to Nabü-aḥḥē-bulliṭ without the consent of the latter's father Nargija; further, it threatens ^f Tablutu with slavery should she ignore this invalidation and continue her association with Nabû-aḥḥē-bulliṭ. This same Nabû-aḥḥē-bulliṭ also figures in Cyr. 311 (Babylon, 531 B. C ), drafted three days earlier. Cyr. 311 stipulates that a father and son, who were respectively a witness to and the scribe who recorded yet another document, guarantee the validity of the offer of the bride ^f Kubbutu in marriage to Nabû-aḥḥē-bulliṭ and agree to secure the consent of his father Nargija. (See Dandamaev, , Slavery in BabyloniaGoogle Scholar, 432f., with bibliography, but note that the two guarantors are not witness and scribe to Cyr. 311.) See also Cyr. 307 (Sippar, 531 B. C ), in which a woman named ^fṬabat-IŠŠar j s cautioned under threat of slavery (as was ^fTablutu above in Cyr. 312) against believing the seductive lies of Kulû and warned to insist that his father Kalba be summoned to confirm his approval and consent of the union. (Cyr. 307 is translated in Dandamaev, , Slavery in Babylonia, 105Google Scholar, with bibliography.)

²³ The identity of the party who requested the bride is not preserved in OECT 10 313 ([n.p., n.d.]) or in BM 66005 ([n.p., n.d.]). In six marriage agreements, BM 83249 ([n.p., n.d.]), OECT 9 73 (Kish, 330- B.C.), BM 84127+ (Borsippa, 304 B.C.), VAS 6 227 ([n.p.], 217/16 B. C ), CT 49 193 (Babylon, 2 8 1 – B. C ), and JCS 1 350 No. 3 ([Kish?], [n.d.]), although the tablets are broken at the offer and acceptance of marriage clauses, enough of the rest of the document is preserved to allow us to conclude that the groom acted for himself.

²⁴ BM 61176 (n.p., 584 B. C ); Nbn. 243 (Babylon, 549 B. C ); OECT 10 110 (Uḥudu, 544 B. C ); BM 76029 (n.p., 433 or 373 B. C ); and CT 49 167 (Babylon, 2 8 1 – B. C ). BM 61176 and OECT 10 110 are cases in which the marriage of the groom (a former slave) was arranged by his adoptive father.

²⁵ TBER 78 AO 26775 ([n.p., n.d.]).

²⁶ The husband's father was the first witness in VAS 6 61 (Babylon, 555 B.C.) and BM 82609 (Borsippa, 520 B. C ). Although the absence of the husband's father in the other marriage agreements cannot be taken as absolute proof that he was not alive at the time, follow-up prosopographic study has failed to find any of the fathers of these grooms to be living at or after the date of the marriage agreement.

²⁷ If, in the two marriage agreements in which the identity of the person who requested the bride is not preserved (see note 23), those persons were the grooms' fathers, then nine out of forty-two grooms (21.4 percent) would have had living fathers when they married: if the grooms themselves acted in those two agreements, then seven out of forty-two grooms (16.6 percent) would have had living fathers.

²⁸ In two marriage agreements in which the groom acted for himself it is clear that this was his second marriage: VAS 6 3 (Babylon, 624 B.C.) and Cyr. 183 (Sippar, 535/4 B. C ). In each of these two cases at least, therefore, although the groom's father was apparently not alive at the time of this marriage, we cannot be certain whether he was alive at the groom's first marriage.

²⁹ See text at p. 737 for suggested chronological age ranges for first marriage. Note that, especially in preindustrial societies, the mortality rate rises sharply each year after age fifty; see Howell, N., “Toward a Uniformitarian Theory of Human Paleodemography,” in The Demographic Evolution of Human Populations, Ward, R. H. and Weiss, K. M., eds. (London: Academic Press, 1976), 25–40, esp. 33ffGoogle Scholar.

³⁰ Twenty-five of the forty-two marriage agreements preserve dowry clauses, comparable in style and formulae to the dowry promises and/or receipts. These marriage agreements are not repeated in the following tabulations.

³¹ The wife's father is identified as the donor of the dowry in the following dowry promise/ receipt documents: TuM 2 - 3 48 ([n.p.], 600–562 B. C ); Nbk. 91 (Babylon, 594 B. C ); TCL 12 32 (Babylon, 588(?) B. C ); Nbk. 161 (Babylon, 579 B.C.); Sack Amel-Marduk 57 (Bit-barf, 560 B. C ); Ner. 25 (Babylon, 559 B. C ); BM 33114 (Babylon, 550 or 540 B. C ), and Cyr. 129 (Babylon, 536 B. C ); Nbn. 313 (Sippar, 547 B. C ); Nbn. 348 (Sapija, 547 B. C ); Peiser Verträ 121 ([n.p.], 539 B.C.); Cyr. 111 ([n.p.], 536 B.C.); Cyr. 143 (Babylon, 535 B.C.) and Camb. 214 (Paširi, 526 B.C); YOS 7 59 ([Urak], 532 B.C); Camb. 193 ([Babylon], 527 B.C); Camb. 215 (Paširi, 526 B.C); Camb. 216 (Babylon, 526 B. C ) and OECT 10 161 (Babylon, 521–486 B.C); BM 82609 (Borsippa, 520 B.C); Dar. 522 (Babylon, 501 B.C), Peiser Vertrage 100 (Babylon, 501 B.C) and Dar. 530+ with dupl. (Babylon, 500 B.C); BM 74596 (Uruk, 493 B.C); BM 33932 (Babylon, 490 B.C); BM 84129 ([n.p., n.d.]); VAS 6 275 ([n.p., n.d.]).

The wife's mother and brother were the donors of her dowry in YOS 6 124 (Larsa, 547 B.C).

The wife's brothers) were the donors of her dowry in TCL 12 32 (Babylon, 588 B.C); BM 82629 (Borsippa, 556 B.C.); Nbn. 258 (Babylon, 549 B.C); Nbn. 760 (Babylon, 542 B.C.); Rechtsl. 013f. (Babylon, 535 B.C); Dar. 568 (Babylon, 499 B.C); CT 5153 (Babylon, 497 B.C); BM 33934 (Babylon, 487 B.C.) (this dowry was earlier promised by her father in BM 30441 (Babylon, 497 B.C)); BM 78109 (n.p., n.d.).

The identity of the donor is not preserved in BM 77626 (Babylon, 560 B.C); Dar. 4 (Babylon, 503/2 B.C); BM 74645 (Sippar, 491 B.C).

³² The husband's father was the recipient of the dowry in BM 78109 (n.p., n.d.).

³³ The donor of the dowry in the second category of texts was the wife's father in BM 52925 (Babylon, 668–27 B.C.); YOS 17 322 (Uruk, 596 B.C.); Nbk. 265 (Babylon, 571 B.C.); Ner. 7 (with the wife's mother) (Babylon, 559 B.C.); VAS 6 95 (with the wife's mother) (Borsippa, 556–39 B. C ); VAS 5 25 (Borsippa, 550 B.C.); TuM 2 – 3 1 (Borsippa, 550 B.C.) with TCL 12 85 (Sūr-amīlāti, 551 B. C ); VAS 4 46 (Babylon, 545 B. C ); Bold Leiden Coll. 3 874 (Borsippa, 545 B. C ); TCL 13 179 (with the wife's mother) (Uruk, 527–22 B. C ).

The wife's mother alone was the donor in Nbk. 283 (Babylon, 570 B. C ); Nbn. 1113 (Bīt-šar- Bābili, 548–40 B. C ).

The donor is not preserved or indicated in Dalley Edinburgh 69 (Babylon, 559 B.C.) (two dowries); VAS 5 25 (Borsippa, 550 B. C ); BE 8 64 (Til-Galala, 534 B. C ); Dar. 379 (Babylon, 508 B.C.) (two dowries); BM 62687 ([n.p., n.d.]).

³⁴ The husband was apparently the recipient in BM 52925 (Babylon, 668–27 B. C ); YOS 17 322 (Uruk, 596 B. C ); Nbk. 265 (Babylon, 571 B. C ); Nbk. 283 (Babylon, 570 B. C ); Ner. 7 (Babylon, 559 B. C ); Dalley Edinburgh 69 (Babylon, 559 B. C ); TUM 2 – 3 1 (Borsippa, 550 B. C ) with TCL 12 85 (Sūr-amīlāti, 551 B. C ); VAS 4 46 (Babylon, 545 B. C ); Böhl Leiden Coll. 3 No. 874 (Borsippa, 545 B. C ).

³⁵ The husband's father was apparently the recipient in Dalley Edinburgh 69 (Babylon, 559 B. C ); VAS 5 25 (Borsippa, 550 B.C.) (two dowries); Dar. 379 (Babylon, 508 B.C.) (two dowries).

The identity of the recipient is not preserved in BM 62687 ([n.p., n.d.]).

³⁶ In addition to the husband and the husband's father as recipients, note that the wife herself claimed to have been the recipient of her own dowry in VAS 6 95 (Borsippa, 556–39 B. C ); Nbn. 1113 (Bīt-šar-Bābili, 548–40 B. C ); BE 8 64 (Til-Galala, 534 B. C ); TCL 13 179 (Urak, 527–22 B. C ). If these claims are to be accepted as indications of donations directly to the bride herself, the dowries might have been awarded in the context of inheritance directives (see text immediately below).

³⁷ The wife's father assigned a pre-marriage dowry award to his daughter in Nbk. 251 (Babylon, 572 B.C.) (two dowries); Nbk. 368 (Babylon, 565 B.C.) and Nbk. 403 (Babylon, 563 B. C ); Strasbourg 130 (Babylon, 548 B.C.); Nbn. 356 (Babylon, 547 B. C ); VAS 6 108 (Babylon, 538 B. C ); AnOr 8 18 (Borsippa, 527–22 B. C )

The wife's mother assigned a pre-marriage dowry award to her daughter in Nbk. 198 (Babylon, 576 B. C ); Rechtsl. 0 20f. (Babylon, 541 B.C.) (two dowries); BRM 2 5 (Uruk, 281 B. C ).

The wife's father charged his son to provide a dowry with his daughter to the latter's future husband in BM 33092 ([n.p.], 564 B. C ); VAS 5 129 (Sippar, 535 B. C ).

The wife's mother charged her son to provide a dowry with her daughter to the latter's future husband in TCL 13 174 (Borsippa, 523 B. C ).

³⁸ Fales, , Censimenti e catasti, 134–35Google Scholar, Figure 2.

³⁹ Parpola, S., “Note on Neo-Assyrian Census Lists,” 96–115Google Scholar.

⁴⁰ It is of course possible that Scribe B included daughters among the DUMU. MEŠ, “sons” (or “children”), however.

⁴¹ Contrast the ADB pattern of all males followed by all females with the pattern found in the Old Babylonian Kish document, Ki 1056, in Donbaz, V. and Yoffee, N., Old Babylonian Texts from Kish Conserved in the Istanbul Archaeological Museums, Bibliotheca Mesopotamia, no. 17 (Malibu, Calif: Undena Publications, 1986)Google Scholar, 57ff. Recording the rations due to or received by each member of a number of households, Ki 1056 lists individuals by the quantity of their rations in order of decreasing amounts. Each head-of-household received the largest ration share and is listed first, followed in turn by his wife, sons, daughters, and female slaves. Several households also then include additional members: the head's sister or brother, the head's suckling children, his unmarried (betrothed?) daughters (see note 54), or his mother. Yoffee (p. 63, and see his translations at p. 62) understands the sons listed in sequence after the female slaves to be the sons of that slave and not of the head-of-household. The difficulty of that position becomes evident in Household K, where we find the sequence g e m e2 - n i, d u m u - n i, n i n - n i, which is taken as “his (the head-of-household's) female slave, her (the slave's) son, his (the head-of-household's) sister.” Yoffee's interpretation of d u m u - n i as the slave's child is based on his reading the signs in Household A col. i 1´; as d u m u - g e m e2, “son of the female slave.” Based on the traces of the line in the copy and photograph, two alternative readings are possible here. First, theb group of wedges could be read gaba, and thus line i ´ would read [ … ] [ xdumu ] - g a b a “suckling child.” Secdnd, there is not the usual blank space followed by an incised horizontal line between i 1´ and 2´ and preceding the summary é PN, “household of PN,” and the traces at the end of i 1´ are not very different from those at the end of i 8´. This suggests that i l ´ - 2 ´ records the name of the head-of-household and his patronym (… - x - b u/m u (?) d u m u ´ - dn i n - g a 1), and therefore that Ki 1056 i 1 ' - 8 ' lists only one household, Household A/B. In either case, there would be no reason to assume that any member of the household was identified by an association other than that to the head-of-household. (I thank Miguel Civil for discussing Ki 1056 with me.)

⁴² Brothers are found in No. 2 iii 15–19; No. 3 vi 4–11 (brothers understood here on the basis\ of placement); and No. 9 i 18–23.

⁴³ No. 1 iii 16–23.

⁴⁴ For the reading of the logogram SAL in NA as issu, with the meaning “ife,” see Postgate, J. N., “On Some Assyrian Ladies,” Iraq, 41 (1979), 89–138Google Scholar, at 95 note 9; and see also CAD S 292 s.v. sinniŠtu, discussion section.

⁴⁵ The ADB texts do not specify the identity of the husband of any adult woman, and we therefore cannot determine whether an adult woman listed in any household is the wife of the head-of-household or of another adult male. For example, in a household in which there is one head, one adult woman, and one ṣaḥurtu son, it is possible that the head is a widower and the adult woman is the wife of his son (thus yielding an extended rather than a simple family household). However, in the absence of explicit evidence, I have assumed one wife for each head-of-household, and assume “extra” women are the wives of other adult males in the household. In all households there are sufficient males to account for all the adult women. (In No. 9 i 3–9, revise Fales's table to include one head, two fahurtu sons, one unweaned son (GA), and two SAL; there is, therefore, no excess in the number of adult women.) Scribe B, who was not as careful as Scribe A in recording household data and whose information therefore has not been utilized here, did, however, indicate households with more adult females than males: No. 4 iv 3 – 4, No. 11 iv 12–13, and No. 11 iv 17–18 each indicates one head, one son, and 3 SAL; NO. 4 viii 16 and No. 5 vi 8–9 both indicate one head and two SAL. Scribe B did not include unrelated attached persons (the ZI.MES of Scribe A), and only once did he include a head's brother (No. 4 iii 8–9; note that in No. 7:3, the two males listed in Fales's table under the heading “altri” and marked in the text edition as Š[BŠ.MEŠ] are DU[MU.MEŠ], “sons”). These extra women, therefore, might have been the wives of these other unlisted adult males, although we might not expect Scribe B to include women other than the head's wife. Alternatively, these extra women might have been the widowed wives of the head's sons, or may indicate polygamous marriages.

⁴⁶ No. 1 i 1–3 and No. 1 i 16–17.

⁴⁷ No. 2 i 10–13.

⁴⁸ The weaned daughter listed in the household in No. 1 ii 16–23 (which includes a son's son, see text at note 43) might equally be the daughter of the head's son rather than the daughter of the head.

⁴⁹ No. 3 iv 11–12 and No. 10 i 11–12 each includes a head-of-household and three or five ZI.MEŠ; NO. 2 viii 9–10 includes a head-of-household and one named person identified as a LU2. NAR; see text at note 79. In addition to these three households without related coresidents, No. 3 i 12–15 and No. 3 vi 4–11 each includes one unidentified person in addition to family members.

⁵⁰ The population of the forty-five households is 39.77 percent female.The expected the oboretically equal male-female population of the conjugal couple (head-of-household and one wife) is only 37.77 percent female. The total-offspring population finds 33.33 percent female. (And note that in this study I omit the numerous single males that the ABD lists outside of the family households; the inclusion of these male workers would further reduce the percentage of females in the entire ADB population.)

⁵¹ If we assume all offspring whose classifications are broken to belong to the youngest categories, we would find twenty-one out of forty-nine, or 42.86 percent, are female. The sample is so small, however, that the male-female imbalance is certainly not significant. If the proportion of the female-offspring population were to remain lower than 51 percent in a larger sample, the discrepancy could be accounted for by a minimal trend toward poorer care of infant girls or even by infrequent cases of infant exposure.

⁵² If we assume all offspring whose classifications are broken to belong to the oldest categories, we would find six out of twenty-seven, or 22.22 percent, are female.

⁵³ Two brothers of a head-of-household are also described as ṣaḥurtu (No. 9 i 18–23)Google Scholar.

⁵⁴ Compare the two females identified as é - g i - a in Ki 1056, Households D and O (see note 41). Yoffee points out that it is unlikely that the é - g i - a (kallatu) was a daughter-in-law in these two households, neither of which included the requisite sons. Rather, she may have been “a girl betrothed to a man but still living in her father's house. She is given 30 liters rations, more than a child, and is not classified as a ‘daughter’ ” (p. 63). Might these females be in a position comparable to the batussu of the ADB households?

⁵⁵ In preindustrial societies, mortality of mothers at childbirth was probably not excessive. Roger Schofield, of the Cambridge Group for the History of Population and Social Structure, in a private communication of March 1986, kindly informs me that his own investigation of maternal mortality in sixteenth- through nineteenth-century England finds a maternal mortality of between 1.0 and 1.5 percent of birth events. See Schofield, R., “Did the Mothers Really Die? Three Centuries of Maternal Mortality in ‘The World We Have Lost,’ ” in The World We Have Gained, Bonfield, L., Smith, R. M., and Wrightson, K., eds. (Blackwell, 1986), 231–60Google Scholar. For maternal mortality rates in modern industrial societies, see Preston, S., Mortality Patterns in National Populations (New York: Academic Press, 1976), 53ff., Tables 3.1 and 3.2Google Scholar; and Danforth, D. N., ed., Obstetrics and Gynecology, 4th ed. (Philadelphia: Harper and Row, 1982), 286, Figure 15–5Google Scholar. I have not systematically collected data on maternal mortality for ancient Mesopotamia. There are, to be sure, references in our sources to wet nurses (see CAD M/2 265f. s.v. mušēniqtu), at least some of whom no doubt nursed infants whose mothers died in childbirth. And the medical and magical literature devotes considerable attention to the dangers and mysteries of pregnancy and childbirth (see Stol, M., Zwangerschap en Geboorte bij de Babyloniërs en in de Bijbel, Mededelingen en Verhandelingen van het Vooraziatisch-Egyptisch Genootschap “Ex Oriente Lux” 23 (Leiden: Ex Oriente Lux, 1983)Google Scholar; and Reiner, E., “Babylonian Birth Prognoses,” Zeitschrift für Assyriologie, 72:1 (1982, 124–38)Google Scholar. But there are relatively few documentable cases of widowed males remarrying (even allowing for the model of older age at first marriage for males presented here), and widows are relatively common in our economic and legal sources. My preliminary impression, therefore, is that maternal mortality was not exceptionally high in ancient Mesopotamia.

⁵⁶ But see note 45.

⁵⁷ In two other households with more than one adult woman, there are brothers as well as sons of the head-of-household to whom the extra women could be assigned as wives.

⁵⁸ See Gulick, J., The Middle East: An Anthropological Perspective (Pacific Palisades, Calif.: Goodyear Publishing Co., 1976), 183Google Scholar. See also the comments on different groups in the several papers in Mediterranean Family Structures, Peristiany, J. G., ed. (Cambridge: Cambridge University Press, 1976)Google Scholar: D. Seddon, studying a Moroccan rural community, finds age at marriage to be “between 14 and 17 for most girls; between 16 and 30 for most men” (p. 177); P. Benedict notes that men marry between ages 24 and 28 in a Turkish town (pp. 233f.); J. Rosenfeld examines an Arab village in Israel where “[i]n 196S the median marriage age for Muslim men was 23.8, for Christian men 27.3; for Muslim women 19.9, for Christian women 21.2; however, even the median figures do not fully take into account the high rate of remarriage among Arabs, and therefore we do not have correct figures for first marriage” (p. 122); and E. Friedl observes that in a Greek village, economic and social class certainly influences age at marriage: “The higher the economic levels, the greater was the probability that men and women would marry late” (p. 380). See also note 6 above for a similar warning as to Neo-Assyrian elites.

⁵⁹ Recall that in at least one household listed in the ADB (No. 9 i 18–23), which consisted of one head, one four-rūṭu male offspring, two adult women, and two brothers of the head, the two brothers are designated ṣaḥurtu. The designation need not, therefore, be restricted only to sons of the head-of-household.

⁶⁰ The imprecision of the ADB texts in identifying the relationships of the parties allows for the possibility that any of the male or female offspring listed could be the children of any conjugal couple of the household. In the one household to include a sister (No. 2 10´–13´), there is a break following the reference to the sister (1 NIN- su [ … ]), and she might therefore have been further identified as GA, UD, X rūṭu, oi batussu; we cannot, therefore, exclude the possibility that a sister (as opposed to only a daughter) could be classified as batussu.

⁶¹ No age-related terminology is used to describe the grooms in these NB marriage agreements. In the few marriage agreements in which the groom's father or mother contracted the marriage on his behalf, the groom is simply referred to as the “son” (māru) of his parent.

⁶² BM 76202 ([n.p., n.d.]), ba-tul-tum ṣu-ḥir-tum; BM 76968/72 (Borsippa, 203 B.C.), batul-tum SAL-ú-tum.

⁶³ The marriage agreements that omit any descriptive terminology are BM 50149 (Sippar, ca. 625–23 B.C); Nbk. 101 (Babylon, 592 B.C.); TuM 2–3 1 (Borsippa, 550 B. C); BM 70235 (Sippar, 548 B. C); BM 61434 (Babylon 543 B. C); YOS 6 188 (Ālu-ša-Lanē, 542 B. C); Nbn. 990 (Babylon, 540 B. C); BM 82609 (Borsippa, 520 B. C); BM 50106 ([n.p., n.d.]); BM 83249 ([n.p., n.d.]); CT 49 165 (Babylon, 281–B. C). Note that three of these agreements, BM 50106, BM 61434, and BM 83249, are the three marriage agreements in which the bride herself acted in contracting her marriage or awarding her dowry to her husband; see further below.

⁶⁴ The tablets recording marriage agreements that are broken at the pertinent points are BM 61984 ([n.p.], 543/2 B. C); OECT 10 313 ([n.p., n.d.]); BM 66005 ([n.p., n.d.]); OECT 9 73 (Kish, 330–B. C); BM 84127+ (Borsippa, 304 B. C); CT 49 193 (Babylon, 281– B. C); JCS 1 350 No. 3 ([Kish(?)], [n.d.]).

⁶⁵ *Nu’artu is also found in difficult contexts in two NB texts, both of which almost certainly involve the musician's craft. MDP 36 84 No. 4 (edited by M. Rutten, in Ghirshman, R. et al. , Village Perse-Achéménide, Mémoires de la Mission Archéologique en Iran, no. 36 (Paris: Presses Universitaires de France, 1954), 83–85Google Scholar, was drafted in Susa and, from the term of the contract, dates to the third month of the eighteenth year of one of the three rulers named Artaxerxes (thus, 447, 387, or 341 B. C); the text was referred to by Stolper, M. W., “The Neo-Babylonian Text from the Persepolis Fortification,” Journal of Near Eastern Studies, 43:4 (1984), 299–310, at 309 n. 36Google Scholar. The text records the “sale” of a daughter for the sum of fifty shekels for a term of six months ana SAL nu-mar-ra-ú-tu (line 5). The second text is BM 64026 (82–9–18,3995), a copy of which by G. Benin is to appear in a forthcoming volume of cuneiform texts from the British Museum. BM 64026, drafted in Sippar and dating to 7–X–6 Cyrus (533 B.C), records the request by a woman who was herself an oblate and the wife of a farm laborer of the temple of Šamaš to have her daughter listed on the temple rolls. The daughter is identified as a SAL nar-tum (line 3), and appears to have been trained in the musician's craft (ina na-ru-tum, line 4).

⁶⁶ Landsberger, B., “Bemerkungen zu San Nicolò und Ungnad, Neubabylonische Rechts-und Verwaltungsurkunden, Bd. I 1.2,” Zeitschrift für Assyriologie, 39:2 (1930), 277–90Google Scholar, at 290, in reference specifically to the translations of the term in VAS 6 61 and VAS 6 95 by M. San Nicolò and A. Ungnad in NRV Nos. 2 and 3 as “Sängerin.”

⁶⁷ See Soden, W. von, Akkadisches Handwörterbuch (Wiesbaden: Otto Hatrassowitz, 1959– 1982, hereafter AHw.)Google Scholar and the Assyrian Dictionary of the University of Chicago (CAD), at the entries to batultu, nārtu II (AHw.), and nârtu (CAD). Also cited s.v. nârtu among the NB references was UET 4 27:2, which is rather to be read as a personal name; see Nicolò, M. San, Babylonische Rechtsurkunden des ausgehenden 8. und deṣ 7. Jarhhunderts v. Chr., ABAW Neue Folge Heft 34 (Munich: Verlag der Bayerischen Akademie der Wissenschaften, 1951), no. 37Google Scholar. The only NB reference to nârtu, “singer,” that might remain of those cited in the dictionaries is ABL 511 (and see the two texts cited above at note 65). Note also that RA 25 81 No. 23 (read here as nu-maš-ti) was taken as a logogram (NU.BAR-ti) for kulmašītu, “(a woman devotee of a deity),” in the dictionaries (AHw. in 1963 s.v. kulmašītu, CAD in 1964 s.v. aḥātu A mng. la–1´ and in 1971 s.v. kulmašītu), and as a logogram for ištarītu, “hierodule,” by Dandamaev, , Slavery in Babylonia, 133Google Scholar.

⁶⁸ BM 65149 is the only one of this group without a preserved date; it could easily, however, be dated as early as the reign of Nebuchadnezzar on the basis of script, prosopography, and especially the curse formula in lines 23–26.

⁶⁹ See Rabinowitz, J. J., “Neo-Babylonian Legal Documents and Jewish Law,” Journal of Juristic Papyrology, 13 (1961), 131–75Google Scholar. Rabinowitz argues there (pp. 136f.) that SAL.NAR in VAS 6 61 and VAS 6 95 is “the equivalent of” Hebrew na‘ărāh.

⁷⁰ The syllabically written masculine and feminine terms for “singer” cited in CAD and AHw. also show a weak middle radical: na-’-ra-ti, na-’-ru-ú (MB); nu-a-ra-ti/du, nu-a-ru, nu-’-a-ri (Nuzi); nu-’-a-ri (NA); nu-a-ri-im (OA). The NB terms designating the bride written syllabically nu-mar-tum, nu-man-áš-tum (or nu-man^áš-tum), and nu-maš-ti display the following features: (1) The final consonant before the feminine marker /t/ is represented by either /r/ or /š/, a common late variation; see Soden, W. von, Grundriss der Akkadischen Grammatik (Rome: Pontifical Institute, 1969), § 35cGoogle Scholar; Reiner, E., A Linguistic Analysis of Akkadian (The Hague: Mouton, 1965), 115Google Scholar; Kaufman, S. A., The Akkadian Influences on Aramaic, Assyriological Studies 19 (Chicago: University of Chicago Press, 1974), 144fGoogle Scholar. (2) The /m/ between the two dissimilar vowels /u/ and /a/ is a glide, and may stand for /w/, /j/, or /’/; for intervocalic /m/ representing /w/, see Kaufman, , Akkadian Influences, 143Google Scholar, and von Soden, Grundriss der akkadischen Grammatik, § 31a; for the sequence of dissimilar vowels with intervening /w/, /j/, or /’/, see Reiner, Linguistic Analysis, 46f. (3) That the forms with /nu-ma-/ render West Semitic /n(V)‘(V)/ is not provable. The rendering into cuneiform Akkadian of a West Semitic loan word with a middle ‘ayin radical is not well documented (particularly at the moment of borrowing, thus discounting internal cuneiform Akkadian changes that can occur over the millennia); as Kaufman, Akkadian Influences, 142Google Scholar, notes “there is absolutely no evidence for the preservation of ‘ayin in first-millennium Akkadian.’ Nonetheless, the sequence of dissimilar vowels (/u/ and /a/) with an intervening glide (/m/) is not incompatible with an original ‘ayin in the word borrowed. The remaining difficulty is the quality of the first vowel, and, if my interpretation is correct, this newly posited form *nu’artu suggests that the West Semitic word borrowed may have been *nu‘ar(āh), rather than na‘ar(āh) (→ na‘ăr(āh) in the Masoretic vocalization). I wish to thank Jonas C. Greenfield and Dennis Pardee, who commented on various points raised in this note.

⁷¹ The position of the CAD apparently was influenced by Landsberger, B., “Jungfräulichkeit: Ein Beitrag zum Thema ‘Beilager und Eheschliessung,’” in Symbolae luridicae et Historicae Martino David Dedicatae, Ankum, J. A. et al. , eds. (Leiden: Brill, 1968), II, 41–105Google Scholar. See also J. J., , “Sex Offenses in Sumerian Laws,” Journal of the American Oriental Society, 86:4 (1968), 355–72Google Scholar, at 356f: “batultu, ‘virgin’ [is] an age distinction defining her as ‘prenubile’ and only implicitly, therefore, untouched.” In the Old Testament as well, the translation of betulah as “virgin” has been the object of debate. For a synopsis of the evidence, see Botterweck, G. J. and Ringgren, H., eds., Theological Dictionary of the Old Testament, rev. ed. (Grand Rapids, Mich.: Eerdmans, 1977), II, 338ff. s.v. b‘thûlāhGoogle Scholar. See also Wenham, G. J., “Betulah, ‘A Girl of Marriageable Age,’ ” Vetus Testamentum, 22:4 (1972), 326–48Google Scholar. Contra Wenham, however. see worth, T. Wads, “Is There a Hebrew Word for Virgin? Bethulah in the Old Testament,” Restoration Quarterly, 23 (1980), 161–71Google Scholar. (I owe this last reference to D. Pardee.) See now also Locher, Clemens, Die Ehre einer Frau in Israel: Exegetische und rechtsvergleichende Studien zu Deuteronomium 22, 13–21, Oslo Biblicus et Orientalis 70 (Gottingen: Universitätsverlag Freiburg Schweiz, 1986), 121ff.Google Scholar, for a summary of the biblical and some of the cuneiform problems. Locher's book came to my attention too late to incorporate his positions into my argument. He correctly concludes that SAL.NAR in the NB marriage agreements does not refer to a singer, but follows Landsberger in taking it as a logogram for batultu.

⁷² Oppenheim, A. L., Ancient Mesopotamia, rev. ed. (1964; Chicago: University of Chicago Press, 1977), 77Google Scholar. Note that Landsberger, , “Jungfraulichkeit,” 43Google Scholar, was aware of Oppenheim's position.

⁷³ BM 50149 (Sippar, ca. 625–23 B.C.) lines 4–9: GIR₃^II LU₂.N[ITA] ‘la-pa’-ni-ka pu-ru-si^? lu-ú áš-šá-tum at-[ti] W taš-m[e-e-ma] GIR₃^II LU₂.N[ITA] la-pa-ni-šú] taš^?-pa^?-ru-su. The verb in line 9 is obviously corrupt; see the comments to my edition of the marriage agreement in my forthcoming study of NB matrimonial property. The other two marriage agreements in which the bride acts for herself are otherwise unexceptional; they are BM 61434 (Babylon, 543 B.C.) and BM 83249 ([n.p., n.d.]).

⁷⁴ CT 49 165 (Babylon, 281–B.C.) lacks batultu or *nu’artu; OECT 10 313 ([n.p., n.d.]), BM 66005 ([n.p., n.d.]), and JCS 1 350 No. 3 ([Kish?], [n.d.]) are broken in the offer and/or acceptance clauses where the designation would appear.

⁷⁵ BM 66005:1´-19´, DUMU.SAL LU₂.TUR šá W šá MU 7.KAM₂a-di-‘an-na’ IM.DUB l[a kangu (a-na)] H tu-li-di DUMU.[SAL] H ši-i. Cf. also CT 49 165:4–6, PN PN₂(?) ù PN₃ LU₂.A.MEŠ al-du-’ u a-di-an-na NA₄.K1Š1B nu-duri-ni-šú la kan-gu A.MES “PN, PN₂, and PN₃, the children whom she bore before her dowry tablet had been sealed, are the children < of H> ”; OECT 10 313:5´–6´: [ … ] x a-di i-na-an-ni <la(?)> kan-[gu …] [ … ]-ta ^? ku-nu-uk ma i-X-[ ‖ ]; JCS 1 350 No. 3:6–8, DUMU.MEŠ u DUMU.SAL.M[EŠ] [ … a-d]i i-na-an-nu IM.DUB [nu-dun-ni]-e-šú la kan-gu [ … ]. Note that the tablet referred to in these four clauses is specifically the “dowry tablet” in two cases, and perhaps refers to the dowry in the other two as well (all four marriage agreements include dowry clauses).

⁷⁶ Compare the general essay, drawing on selected sources from all periods of Mesopotamia!) history and from the Old Testament, by Cassin, E., “Virginité et stratégie du sexe,” in La premiére fois, ou le roman de la virginité perdue à trovers les siècles et les continents, Bardet, J. P. et al. , eds. (Paris: Ramsay, 1981), 241–58Google Scholar.

⁷⁷ Babylonian Talmud, tractate Bab. Ketubbot 39a. Of particular importance for our discussion of *nu’artu and batultu is one frequently cited passage in the Old Testament; in Genesis 24:16, when Abraham's servant first sees Rebecca at the well, she is described as “a na‘ărāh, fair to look upon, a betulah, whom no man had known.”

⁷⁸ Labat TOP pi. 49:85 (Hunger Koloprine No. 454), and Gordon Smith College 110:18, respectively. Both are cited CAD s.v. nâru usage c ═ 10´.

⁷⁹ No. 2 viii 9–10, where we find PN A [PN2] LU2.NAR. See note 49.

⁸⁰ 80 The na’ar in the Old Testament as well is often to be understood as a youth in the stage prior to becoming the head of his own household. See Stager, L. E., “The Archaeology of the Family in Ancient Israel,” Bulletin of the American Schools of Oriental Research, 260 (1985), 1–35, esp. 25ffGoogle Scholar.