Hostname: page-component-7c8c6479df-7qhmt Total loading time: 0 Render date: 2024-03-28T18:39:19.839Z Has data issue: false hasContentIssue false

Brother-Sister Marriage in Roman Egypt

Published online by Cambridge University Press:  03 June 2009

Keith Hopkins
Affiliation:
Brunei University

Extract

A Favourable Horoscope: ‘If a son is born when the Sun is in the terms of Mercury, he will be successful and have great power … He will be brave and tall and will acquire property and moreover will be married to his own sister and will have children by her.’

Type
Husbands and Wives: Inheritance and Work
Copyright
Copyright © Society for the Comparative Study of Society and History 1980

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

This is a revised version of my inaugural lecture at Brunei University; I am very grateful to my colleagues at Brunei for comments; and to Dr. Graham Burton, Professors P. A. Brunt, Sir Moses Finley, Jack Goody and Sir Edmund Leach, Dr. John North and Mr. Peter Parsons for detailed and searching criticism

1 From an astrological handbook of Graeco-Egyptian origin, surviving in a late Latin translation. See Gundel, W., ed., ‘Neue astrologische Texte des Hermes Trismegistos’, Abhandlungen der Bayerischen Akademie der Wissenschaften (Phil.-hist. Abl.) vol. 12 (1936): 99.Google Scholar

2 The classic study to which I am heavily indebted throughout this article is by Hombert, M. and Préaux, C., Recherches sur le recensement dans I‘Égypte romaine, Papyrologica Lugduno-Batava 5 (Leiden, 1952)Google Scholar. See also the only account in the sociological literature, Middleton, R., ‘Brother-sister and Father-daughter Marriage in Ancient Egypt’, American Sociological Review 27 (1962): 603–11.CrossRefGoogle Scholar

3 Here I follow Goody, J. and Harrison, G.A., ‘Strategies of Heirship’, Comparative Studies in Society and History 15 (1973) 1618; in a self-reproducing population with high mortality, averaging six children ever born per family, and with only a one in three chance of a son surviving his father, then roughly 17 percent of families would have no heirs, and 42 percent would have either only sons or only daughters. Therefore 41 percent of families would have both son(s) and daughter(s); these families would provide about half the next generation, but in some such families marriage with a sibling would be possible for only one pair and in others impracticable because of a large age gap between surviving siblings, especially if the sister were much older. My conclusion ‘one-third and perhaps more’ is thus based on the calculation: 15–21%/ > 50%.CrossRefGoogle Scholar

4 By no means all leading anthropologists have considered the universality of the incest taboo central, nor do the necessarily selective quotations which follow do justice to all those quoted. As introductory reading, I found Fox, R., Kinship and Marriage (London, 1967)Google Scholar and Goody, J., ed., Kinship (London, 1971) particularly useful and readable.Google Scholar

5 Kroeber, A.L., ‘Totem and Taboo in Retrospect’, American Journal of Sociology 45 (1939/1940): 447.Google Scholar

6 Murdock, G.P., Social Structure (New York, 1949), p. 12Google Scholar, based on an analysis of 250 societies. Murdock's view was cited approvingly by Parsons, Talcott, ‘The Incest Taboo in Relation to Social Structure’, British Journal of Sociology 5 (1954): 101; he related the universal incest taboo to the (alleged) universality of the nuclear family.CrossRefGoogle Scholar

7 White, L., The Evolution of Culture (New York, 1959), p. 9.Google Scholar

8 Lévi-Strauss, C., The Elementary Structures of Kinship (London, 1969), p. 8Google Scholar; this translation is from the second revised edition of 1967 in which Lévi-Strauss was partially aware of the practice of brother-sister marriage in Egypt (ibid., 9–10), but argued nonetheless, correctly but tendentiously, that all societies had some incest prohibitions, even if the relationships covered were wider or narrower in some societies.

9 Lévi-Strauss, C. on ‘The Family’ in Shapiro, H.L., ed., Man, Culture and Society (New York, 2nd ed., 1960), p. 278.Google Scholar

10 Fox, R., ed., Biosocial Anthropology (London, 1975), p. 12Google Scholar; Fox owes a lot here to Lévi-Strauss, but also considers the possibility of a social evolution of humanity in which the incest taboo may or may not have been crucial (ibid., pp. 30–31).

11 Goody, J., ‘A comparative approach to incest and adultery’, British Journal of Sociology 1 (1956): 285ff., also in Goody 1971: 64ff.Google Scholar

12 Westermarck, E., The History of Human Marriage (London3rd ed., 1901), p. 293, lists these and several others, such as the Baghirmi, Sandwich Islands, Siam, Burma, Ceylon, etc.Google Scholar

13 Slater, M.K., ‘Ecological Factors in the Origin of Incest’, American Anthropologist 61 (1959): 1042.CrossRefGoogle Scholar

14 Fox, R., ‘Sibling Incest’, British Journal of Sociology 13 (1962): 128ff.CrossRefGoogle Scholar; Westermarck, E., 1901, p. 320.Google Scholar

15 Spiro, M., Children of the Kibbutz (Harvard, 1958), p. 347Google Scholar; Talmon, Y., ‘Mate Selection in Collective Settlements’, American Sociological Review 29 (1964): 492–93.CrossRefGoogle Scholar

16 Bischof, N., ‘Comparative Ethology of Incest Avoidance’, in Fox, R., ed., Biosocial Anthropology (London, 1975), pp. 37ffGoogle Scholar; Bischof describes practices of mammals which encourage outbreeding, but his conclusions (a) that inbreeding among mammals is rare in their natural habitat and (b) that such incest as does occur, occurs only among lower animals, overreaches the data. See also, Kortmulder, K., ‘An ethological theory of incest taboo and exogamy’, Current Anthropology 9 (1968): 437ff.CrossRefGoogle Scholar

17 Weinberg, S.K., Incest Behaviour (New York, 1955) reported, inter alia, 42Google Scholar cases of brother-sister incest in Illinois, U.S.A. Of these, only six couples contemplated marriage, and in all six cases and only those six cases had the brother and sister been separated from childhood; also Wolf, A., ‘Childhood Association, Sexual Attraction and the Incest Taboo: A Chinese Case’, American Anthropologist 68 (1966): 894.CrossRefGoogle Scholar

18 Freud, S., Totem and Taboo (London, 1938) especially, pp. 189ffGoogle Scholar; Malinowski, B., The Sexual Life of Savages (London3rd ed., 1932), p. 30Google Scholar; Sex and Repression in Savage Society (London, 1927), p. 79.Google Scholar

19 Maisch, H., Incest (London, 1973), reviews recent literature; his history is wide-ranging but error prone.Google Scholar

20 The sim-pua women also had more known extramarital affairs than did women married to outsiders, but the evidence for either group was slight: see Wolf, A., 1966:863ff.Google Scholar, and Adopt a Daughter-in-law, Marry a Sister’, American Anthropologist 70 (1968): 864ff.CrossRefGoogle Scholar

21 From the much quoted and reprinted article by Aberle, D.F. et al. , ‘The Incest Taboo and the Mating Patterns of Animals’, American Anthropologist 65 (1963): 260CrossRefGoogle Scholar. But now see Fox, R., ed. (1975), especially 9ff. and the literature cited there. On genetic damage from inbreeding, see below p. 326.Google Scholar

22 According to one American study (Weinberg 1955: 41), the average age of daughters committing incest with fathers was 15 years (n=164); of sisters committing incest with brothers, 19 and 24 years respectively (n=42).

23 Parsons, T., ‘The Incest Taboo in Relation to Social Structure’, British Journal of Sociology 5 (1954): 104ffCrossRefGoogle Scholar; but see Levy, M., ‘Some Questions about Parsons’ Treatment of the Incest Problem‘, British Journal of Sociology 6 (1955): 277ff.CrossRefGoogle Scholar

24 Mentuhopte II married his full sister Nefuru III, see Hayes, W.C. in Cambridge Ancient History vol. 1.2 (Cambridge3rd ed., 1971), pp.478,481;Google Scholar and about 1600BC, Seqenenre Tao II married his full sister Ahhotpe, ibid., vol.2.1.73.

25 Černy, J., ‘Consanguineous Marriages in Pharaonic Egypt‘, Journal of Egyptian Archaeology 40 (1954): 23ff. The custom of calling one's wife “sister” seems to date from the eighteenth dynasty in the middle of the second millennium BC.CrossRefGoogle Scholar

26 Several inscriptions set up in theislands of the Aegean record the title Philadelphos given Arsinoe from the time of her marriage, see Wilcken, U., Archiv für Papyrusforschung 3 (1906): 318.Google Scholar

27 Memnon, Similarly, in Jacoby, F., ed., Die Fragmente der griechischen Hisloriker IIIB (Leiden, 1950) 434.8.7.Google Scholar But Kornemann, E., ‘Die Geschwisterehe im Altertum’, Mitteilungen der schlesischen Gesellschaft für Volkskunde 24 (1923): 17ff, thought the custom of Persian origin.Google Scholar

28 I have not included Ptolemy VII who on the day of his marriage to his mother was murdered by his uncle who then married his sister (Justin 38.8). Two other kings married a niece and a cousin/step-mother. See Seibert, J., Historische Beiträge zu den dynastischen Verbindungen in hellenistischer Zeit (Wiesbaden, 1967) pp. 81Google Scholar ff. On brother-sister marriages in the parallel ruling house of the Seleucids in Syria (there one sister married three brothers) in the early second century BC, ibid., p. 68.

29 This point is well made by Thierfelder, H., Die Geschwisterehe im hellenistisch-römischen Agypten, Fontes et Commentationes 1 (Münster, 1960) passimGoogle Scholar; his review of the evidence is incomplete, and the criteria for acceptance seem occasionally too severe. See also Modrzejewski, J., ‘Die Geschwisterehe in der hellenistischen Praxis’, Zeitschrift der Savigny Stiftung 81 (1964): 52ff.Google Scholar

30 See also Seneca, , Apocolocyntosis 8Google Scholar; he says that at Athens it was possible to marry a half-sibling, at Alexandria, a full sibling. The article by Weiss, E., ‘Endogamie Und Exogamie im römischen Kaiserreich’, Zeitschrift der Savigny Stiftung 29 (1908): 340ff, contains an interesting discussion of the evidence.Google Scholar

31 Since Hombert, and Préaux, , RecherchesGoogle Scholar, several additional census papyri have been published. I used the list of census documents given by Nachtergael, G., Papyri Bruxellenses Graecae (Brussels, 1974) pp. 5158Google Scholar, adding only P.Mich.Inv. 4716c = SB 9555, P.Vindob. Tandem 20, P.Med.Inv. 13 and 69.59 published in Aegyptus 54 (1974): 1721, P.Haun. 24, and BGU 2220–6. My initial cast differed in some respects significantly from Hombert and Préaux's; I wondered whether this was my fault. Fortunately, Helen Cockpe of University College, London came to my rescue and kindly agreed to read the census documents also. The figures I present here are the product of our joint understanding. I am very grateful to her for her caution and care.Google Scholar

32 The earliest certainly attested census in Roman Egypt dates from A D 33–4; but I accept the conclusions of G.M. Browne ad P.Mich. 578 that there was probably a census in A D 19–20. See also Montevecchi, O., ‘II censimento romano d'Egitto’, Aevum 50 (1976): 72–3.Google Scholar

33 Breasted, J.H., Ancient Records of Egypt (Chicago, 1906), vol. 1, p. 59 (#106)Google Scholar; Pirenne, J., Histoire des institutions et du droitprive de I'ancienne Egypte (Brussels, 1934) vol.2, pp. 178-79.Google Scholar

34 See Griffith, F.L., ed., Hieratic Papyri from Kahun andGurob, ThePetrie Papyri (London, 1898), pp. 1929Google Scholar; and for returns from the third century BC, see Wilcken, U., Chrestomathie (Leipzig, 1912), vol. 1.2, nos. 198–99Google Scholar; P.Petrie 59 B and D; and Tebt, P.. 814, 45–7. The term laographia, meaning census, was used in Greek tax lists from Egypt in the early first century BC (P.Tebt. 103 and 189); later, laographia came to mean poll-tax. Diodorus Siculus, writing about 50 BC, gave a figure for the total population of Egypt (1.31), based in part on priestly records but revised after the Greek conquest of Egypt; unfortunately the text is corrupt at the critical point. Thus we have evidence for previous population counts in Egypt, but not for periodic censuses.Google Scholar

35 On the Roman census, see Brunt, P.A., Italian Manpower (Oxford, 1971) pp. 3ffGoogle Scholar; on the census in Judaea see the excellent summary in Schürer, E., The History of the Jewish People in the Age of Christ, revised by Vermes, G. and Millar, F. (Edinburgh, 1973), vol.1, pp. 399ff.Google Scholar

36 For example, the long tax-list from a town-ward in Arsinoe dated AD 72/3 is based on the census of AD 61–2 (11.30–32); it comprises lists of those liable to and exempt from poll-tax in different categories, such as slaves, potters, Jews, the young, the old; for example (11.449–54): “of minors one year old in the third year [of Vespasian's reign], but three years old in the fifth year [of t he reign] and liable to the t ax on Jews, males, Seuthes son of Theodoros son of Ptolemaios, mother Philous, who was examined for privileged status in the fourth year [of the reign] when he was two years old; sub-total 1; (running) total 15.” This extract illustrates t he mass of detail collected by Egyptian administrators. See Wessely, C., Studien zur Paldographie und Päpyruskunde 4 (Leipzig, 1905): 58ff.Google Scholar

37 So also Thierfelder, 1960: 9091Google Scholar. It seems probable that filing birth and death certificates was compulsory; how else can we explain the existence of birth certificates for children of (poor) tenants and death certificates of old men no longer liable to poll-tax. Self-interest would not be sufficient. Yet t he number of surviving certificates, about 100 currently, is small, so perhaps compliance was intermittent. See, Mertens, P., Les services de I'étal civil et la contrôle de la population á Oxyrhynchus, Mémoires de I'Académie Royale de Belgique 53 (Brussels, 1958): 48ff. and 130–1.Google Scholar

38 The size of the sample proportionate to the universe from which it is drawn does not matter much. For example, in modern opinion, the sample is often only 1,000 drawn from an electorate (in the United Kingdom) of about 40 million (1/40,000). The strict randomness of the sample, its absolute size, and the number of cells into which the results are split (e.g. Conservative/Labour/Liberal; age, sex) are more important.

39 Grenfell, B.P. et al. , Fayum Towns and their Papyri (London, 1900), pp. 1 ff.Google Scholar

40 See Hombert, and Préaux, , Recherches, pp. 84ff. but note in their table on pp. 93ffthat a single roll (tomos) sometimes contained several usable census returns. In some districts, declarants initially delivered multiple copies; for example, we have six copies surviving of one census return (see BGU 537 and P.Grenf. 55). Of course, it is usually impossible to distinguish copies made for private and for public use.Google Scholar

41 See Hombert, and Préaux, , Recherches, pp. 84, 99 and 128–29.Google Scholar They reckoned that the authentication was written by the declarant himself in 15 out of 46 originals. On scribal fees, at just under 1 percent of the poll-tax (2½ obols per 40 dr), see the excellent article by Shelton, J., “The Extra Charges on Poll-Tax in Roman Egypt’, Chronique d'Egyple 51 (1976): 182. But there is no explicit evidence for scribal fees for writing census returns.Google Scholar

42 It is clear that Alis was Hilarion's wife, but the greeting ‘sister’ is not proof that she was his sister. Similar sentiments about female babies were expressed in Apuleius, The Golden Ass 10.23; Hierocles in Stobaeus, , Anthologion, eds. Wachsmuth, C. and Hense, O. (Berlin, 18841912), vol. 4, p. 603Google Scholar; Musonius Rufus, ibid., p. 665; Poseidippus, ibid., p. 614: ‘Everyone, even if he happens to be a beggar raises a son, but even if he is rich exposes a daughter.’

43 Duncan-Jones, R.P., ‘Age-rounding, Illiteracy and Social Differentiation’, Chiron 7 (1977): 333ff. has constructed an index designed to measure the tendency to round ages to numbers ending in 5 a nd 0. It is sensibly based on age-groups 23–62 years, and scores from 1 (low tendency to round) to 100 (high tendency). Roman tombstone inscriptions score mostly in the range 40–65 (low scores from Italian town-councillors, high scores from the least Romanized provinces); censuses of Italian cities in the fifteenth and sixteenth centuries produce similar scores. Duncan-Jones thinks that this index score may correlate with literacy in antiquity as it does in modern underdeveloped nations. But there are difficulties: first, on this index, the Roman Egyptian census score is only 9 for males and 13 for females; for the tax-list (P.Lond. 257–9 of A D 94), it is only 2. Alas, this does not mean that the Egyptians were precise, very numerate or very literate, even though bureaucratic demands may have heightened age-sensitivity; besides, precision is not the same as accuracy. Finally, one should take account of the social context in which a n age was declared; surely, declaring one's own age in a census is different from recording the age of someone else who is dead. And I feel uncomfortable about separating male and female ages for this purpose, since a husband might record his wife's age inaccurately or imprecisely, but whom then do we count as innumerate?Google Scholar

44 On model life-tables, see Coale, A.J. and Demeny, P., Regional Model Life Tables and Stable Populations (Princeton, 1966), pp. 5ffGoogle Scholar; Carrier, N. and Hobcraft, J., Demographic Estimation for Developing Societies (London, 1971), p. 2Google Scholar; Ledermann, S., Nouvelles tablestypes de mortalité (Paris, 1969).Google Scholar

45 Hombert, and Préaux, , Recherches, p. 102.Google Scholar

46 Le Roy Ladurie, E., Montaillou (London, 1978), pp. 36 and 52, reported the musings of a leading heretic in a village in the Pyranees. ‘When the world began brothers knew their sisters carnally, but when many brothers had one or two pretty sisters, each brother wanted to have her or them. Hence many murders. That is why the sexual act between brother and sister had to be forbidden. [Shades of Talcott Parsons, see note 28 above.] If my brothers G. and B. had married our sisters E. and G., our house would not have been ruined by the capital carried away by those sisters as dowry; our ostal [house] would have remained intact, …’Google Scholar

47 Following Egyptian practice, these marriage settlements were often drawn up, or redrawn, sometime after the marriage. That is why I have called them marriage settlements, not marriage contracts. See Pestman, P.W., Marriage and Matrimonial Property in Ancient Egypt, Papyrologica Lugduno-Batava 9 (Leiden, 1961): 652Google Scholar; cf., for example, Mich, P.. 340341; BGU 252.Google Scholar

48 Thirteen wedding invitations survive; five are invitations to the wedding of a daughter (SB 7745; P.Apoll. 72; P.Fay. 132; P.Fouad Univ. 7; P.Oxy. 927); three are invitations to one's own wedding or to t h e wedding of a son or sister (P.Oxy. 1486–7,2678); two just say wedding. The two quoted are to the weddings of children (teknori) not of sons (huion) nor daughters (thugateron). I owe this list to Helen Cockle.

49 From a brief series of letters (P.Mich. 214–221) concerning a husband and wife, whom he also called sister, H. Zilliacus argued that they did not have the same mother, since one wrote of ‘my mother’ (217.7) and ‘your sister’ (214). Even this is not proof positive; see the comparable usage in t he letter cited above, pp. 324–25: ‘my father’. And besides, the couple could have been half-siblings of the same father. I go into such detail because Zilliacus is sometimes cited as having justified a disassociation between wife and sister; see ‘Zur Sprache griechischer Familienbriefe des III Jahrhunderts nach Chr’, Commentationes humanarum litterarum 13.3 (Helsinki, 1943): 9 and 31.Google Scholar

50 For this opinion, see originally, Wilcken, U., Die Bremer Papyri (Berlin, 1936), p. 13Google Scholar; and now see the corroborating discussion by Fuks, A., Aegyptus 33 (1953), pp. 132–5Google Scholar, who also lists the papyri in this archive, most of which are in Giss, P.. vol. 1, ed. Kornemann, E. and Meyer, P. M. (Leipzig, 19101912). But needless to say, some scholars have disputed that Aline was Apollonius’ sister.Google Scholar

51 The noted palaeopathologist, Sir Marc Ruffer, investigated the mummified remains of the ancient Egyptian pharaohs to see if he could find traces of pathology in their inbred offspring, inter alia by measuring the size of their crania. Ruffer concluded that ‘although they were not tall men, they were by no means undersized’. Cf. Ruffer, M.A., ‘On the Physical Effects of Consanguineous Marriages in the Royal Families of Ancient Egypt’, Proc. Roy. Soc. Med. 12 (1919): 166.Google ScholarPubMed

52 Mather, K., Genetical Structure of Populations (London, 1973), p. 33.1 thank Dr D. C. Coleman of University College, London for guidance in this section.Google Scholar

53 Bodmer, W. F. and Cavalli-Sforza, L. L., Genetics, Evolution and Man (San Francisco, 1976), pp. 369–76.Google Scholar

54 Cavalli-Sforza, L. L. and Bodmer, W. F., The Genetics of Human Populations (San Francisco, 1971), pp. 369ffGoogle Scholar; Adams, M. S. and Neel, J. V. noted the much greater pathology of a sample of children born of brother-sister incest compared with a matched sample of unwed mothers in the United States, in ‘Children of Incest’, Pediatrics 40 (1967): 5562. But there is a lot of difference between illegitimate incest in the U.S.A. and favoured first-cousin marriage in postwar Japan.Google ScholarPubMed

55 Schull, W. J., ‘Genetic Implications of Population Breeding Structure’, in Harrison, G. A. and Boyce, A. J., eds., The Structure of Human Populations (Oxford, 1972), pp. 146ff.Google Scholar

56 Cf. McKusick, V. A., Human Genetics (Englewood Cliffs, 1969), p. 173.Google Scholar

57 Coale, A. J., ‘Estimates of Average Size of Household’, in Levy, M., ed., Aspects of the Analysis of Family Structure (Princeton, 1965), pp. 64ff.CrossRefGoogle Scholar; his brilliantly simple argument is developed by Burch, T. K., ‘Some Demographic Determinants of Average Household Size’, in Laslett, P., ed., Household and Family in Past Time (Cambridge, 1972), pp. 91ff.CrossRefGoogle Scholar

58 Laslett, P., Household and Family in Past Time, especially pp. 28ffGoogle Scholar; I found this work extremely useful, even though the criticisms by Berkner, L., ‘The Use and Misuse of Census Data’, Journal of Interdisciplinary History 5 (1975): 72 Iff. seem well-founded.CrossRefGoogle Scholar

59 There were 100 slaves out of a total population of 880 recorded in t he census returns (11 percent), but the status of some of these 100 slaves was deduced from their position in the census return after other slaves, not explicitly stated. I may therefore have overstated the number of slaves in households with more than one slave. That said, the estimate, about one tenth, coincides with the proportion of slaves in other listings.

60 See, for example, J. Goody in Laslett, Household and Family, p. 122, C. Klapisch, idem., p. 276.

61 The tax-list is Lond, P.. 2.257–9. The tax-list and the census returns are not completely comparable, since the tax-list includes only adult males; the census returns include females.Google Scholar

62 Hajnal, J., ‘European marriage patterns in perspective’, in Glass, D.V. and Eversley, D.E.C., Population in History (London, 1965), pp. 101 fT.Google Scholar

63 Hopkins, K., ‘On the age of Roman girls at marriage’, Population Studies 18 (1965): 309ff.CrossRefGoogle Scholar

64 The two fundamental works on which I have depended for understanding demotic marriage settlements are by Pestman, P. W., Marriage and Matrimonial Property in Ancient Egypt, Papyrologica Lugduno-Batava 9 (Leiden, 1961)Google Scholar, a model of its kind, and Lüddeckens, E., Ägyptische Eheverträge (Wiesbaden, 1960), which gives the demotic text and German translation of 65 marriage contracts, plus a commentary on each successive clause. I must stress that I do not read demotic and so depend on these works; I only wish that non-Greek readers had available secondary works as well-ordered as these.Google Scholar

65 I have translated the demotic term Ŝp n s.hm.t and s'n, left in their original by Pestman, into bride-price and dowry. Something may be lost in translation, and even more in the original.

66 See P.Mich. 121 R 2.2.

67 In citing demotic papyri, I have given the papyrus collection and number, and the number given it by Luddeckens (L 7), who reprinted the text and a German translation. My English translation is usually from his German, but sometimes from Pestman's English translation, or a mixture of both.

68 The basic work is still Kreller, H., Erbrechtliche Untersuchmgen auf Grund der gräkoägyptischen Papyrusurkunden (Leipzig, 1919)Google Scholar; but see also, Seidl, E., Ptolemäische Rechtsgeschichte, Ägyptologische Forschungen 22 (Gluckstadt, 1962), pp. 179ffGoogle Scholar; Rechtsgeschichte Ägyptens als römischer Provinz (Sankt Augustin, 1973), pp. 213ffGoogle Scholar; La preminente posizione successoria del figlio maggiore nel diritto dei papiri’, Rendiconti, Istitulo Lombardo 99 (1965): 185ff.Google Scholar Seidl bravely attempted to find the underlying principles behind the confusion of individual instances, but testators obviously had considerable freedom in choosing how to distribute their property between spouse and children. The law operated in cases of intestacy and to protect children's minimum rights. But formal disherison was possible (see the eloquent though late P.Cairo Masp. 67353-sixth century AD, cf. P.Mil.Vogl.84-AD138). Only rarely were gifts made outside the family. Daughters sometimes, but by no means always, received a smaller share than sons (see P.Oxy 907 and 1034). Dowries, , which pace Modrzejewski, sometimes included land and slaves (P.Oxy 265, late first century AD; CPR 24–6 of AD 136) counted as part of a daughter's inheritance.Google Scholar Cf. Modrzejewski, J., ‘Zum hellenistischen Ehegiiterrecht im griechischen und römischen Ägypten‘, Zeitschrift der Savigny Stiftung 87 (1970).Google Scholar

69 It is impossible to illustrate all the contingencies covered in the surviving wills or to trace developments over time; but I can cite some examples: a pregnant widow received only her dowry back, released all claims on her husband‘s estate, and in return secured the right to expose her baby (BGU 1104—Augustan period; cf.P.Oxy. 268). For a widow‘s life-interest in her house, see Oxy, P.. 105Google Scholar; similarly for a widower, see Oxy, P.. 104Google Scholar; for a widow‘s maintenance, comprising food and pocket-money, providing she stayed ‘irreproachably‘ at home, Grenf, P.. 21Google Scholar. For a widow holding property in trust for her children, with lifetime usufruct, see Oxy, P.. 494Google Scholar; and for a married couple, each leaving his or her property to the survivor, in trust for the children, see Eleph, P.. of 284 BCGoogle Scholar; cf. Lond, P.. 1727Google Scholar; Oxy, P.. 493.Google Scholar For a list of surviving wills, see Montevecchi, O., Lapapirologia (Turin, 1973), pp. 207–08. These wills and other transfers of property in Egypt deserve more study.Google Scholar

70 Cf. Kreller, , Erbrechtliche, 307.Google Scholar

71 There is considerable scholarly disagreement about the extent and nature of Egyptian influence on marriage practices recorded in Greek and surviving from Egypt. For bibliography and a strong statement of Greek settler autonomy, see Vatin, C., Recherches sur le mariage et la condition de la femme mariée à I'époque hellénistique (Paris, 1970), pp. 163ff. lam sceptical; for example, in Tebtunis in the second century BC, Senesis daughter of Menelaus was married to Didymus son of Peteimouthes, according to an Egyptian form of marriage contract, written in Greek (P.Tebt.776). Were they and their parents half-hellenized Egyptians or Egyptianised Greeks, or a mixture? Those who believe that Greek settlers were impervious to Egyptian influence presumably know the answer.Google Scholar

72 Austin, M. M., Greek and Egypt in the Archaic Age (Cambridge, 1970), p. 19ff.Google Scholar supports the evidence for early Greek settlements in Egypt. For the Aramaic papyri, see especially Kraeling, E.G., The Brooklyn Museum Aramaic Papyri (New Haven, 1953), p. 53 and n° 2 (449 BC): ‘If to-morrow or another day Ananiah [the husband] shall die, Tamut [the wife] shall have power over all the goods which ever may be between Ananiah and Tamut …’Google Scholar

73 If wheat cost 8 drachmae per artaba of 26 kilograms (48 choenices), and a family of four persons need 1000 kilograms of wheat equivalent per year for minimum subsistence. For prices a n d measures, see Duncan-Jones, R.P., ‘The Price of Wheat in Roman Egypt’, Chiron 6 (1976): 246, 258.Google Scholar

74 In the long village tax-list, which we used t o trace the timing of household fission (Table 2), 253 occupations of males were listed; of these 9 percent were wage-labourers, and by chance concentration 15 percent were weavers.

75 On stereotypes of peasants and their typical attachment to particular ancestral farms, see Macfarlane, A., The Origins of English Individualism (Oxford, 1978) 23 and the literature cited there.Google Scholar

76 The area of land is known from 120 of these 147 leases; the total (671 arourai = 168 ha of arable land) represents perhaps just under 10 percent of the village's cultivable land; I reckon this rather speculatively: the traceable built-up area of Tebtunis was 27 ha compared with another Fayum village, Kerkeosiris, whose built-up area was 17 ha. Kerkeosiris is the only Fayum village for which we know the total territory and the cultivable area. If we apply the ratio of built-up areas (27:17) to cultivable area, then the cultivable area of Tebtunis was about 1800 ha; thus all leases and sales, of unknown and known size, may well have covered about 10 percent of the village's cultivated land in the period of one year. Cf. Crawford, D.J., Kerkeosiris (Cambridge, 1971) pp. 4445.Google Scholar

77 For complex and apparently good reasons, the editors of the two alphabetised landlists (P.Landlisten) date the two lists to after AD 307–8 and perhaps after 311–13 and before 325, perhaps before 316. By my count about 23 percent of the names in the first list are not in the second list, and there are a similar number of new names in the second list. I conclude therefore that death was the main cause of drop-out. If all landowners were adult, with an average age of 45 years (that was the average age of heads of households in the census returns, N = 122), and if the average expectation of life at birth was 25 years, then a 23 percent drop-out would take about eight years. In this period of perhaps about eight years, by my count 28 percent of landowners common to both lists changed the size of their landholdings by more than 5 percent. For texts and dating, see Sijpesteijn, P.J. and Worp, K.A., eds., Zwei Landlisten aus dem Hermupolite (Zutphen, 1978). The above estimates are obviously speculative but I hope helpful.Google Scholar

78 This hymn to Isis is one of several surviving; they a re a mixture of old Egyptian a n d Greek ideas. See the very interesting comparisons in Müller, D., Ägypten und die griechischen Isis Aretalogien, Abhandlungen der Sächsischen Akademie der Wissenschaften, Phil-hist. Kl. 53.1 (Berlin, 1961).Google Scholar

79 Moore, S.F., “Descent and Symbolic Filiation,” American Anthropologist 66 (1964), 1308ff, shows how common brother-sister incest is in tribal creation myths.CrossRefGoogle Scholar

80 See the edition by Griffiths, J.G. (Cardiff, 1970)Google Scholar and his Origins of Osiris (Berlin, 1966); he is (132ff) more sceptical about brother-sister marriage among humans than the evidence warrants.Google Scholar

81 Griffith, F.L., Stories of the High Priests of Memphis (Oxford, 1900), pp. 112Google Scholar dates this version cautiously to the last century BC, but Budge, E.A.W., Egyptian Tales and Romances (London, 1931), p. 149 dates it to 233 BC. I am extremely grateful to Dr. J. Ray for telling me about this story and much else Egyptian.Google Scholar

82 Plutarch, , Roman Questions 289 D, tackles this traditional problem of insider vs. outsider marriage: “Why d o they [the Romans] not marry women who are close relatives? Is it because they wish to increase the number of relationships by marriage and so acquire many extra kinsmen by giving wives to a nd receiving wives from others? Or do they fear the disagreements which arise in the marriages of relatives?”Google Scholar

83 This is the conventional wisdom. See, for example, Seidl, E., Einführung in die Ägyptische Rechtsgeschichte, Ägyptologische Forchungen 10 (1951): 43Google Scholar; Ägyptische Rechtsgeschichte der Saiten- und Persenzeit, Ägyptologische Forschungen 20 (1956): 50.Google Scholar See also Pirenne, J., “Le statut de la femme dans l'ancienne Égypte,” Recueils de la société Jean Bodin 11 (1959) 6377 who calls attention to the changes which took place in the last two millennia BC. I should stress that I have wandered into ancient Egyptian culture with an innocent and ignorant wonder. I am completely dependant on secondary sources. That said, I have two doubts: legal rights may not accurately mirror practice; secondly, in which social strata did women have high status? The love poems derive from the prosperous; the demotic marriage contracts come from lower social strata.Google Scholar

84 Foster, J.L., Love Songs of the New Kingdom (New York, 1974), pp. 18, 17, 5657, 5051Google Scholar. See also Schott, S., Altägyptische Liebeslieder (Zürich, 1950)Google Scholar translated into French by Krieger, L. (Paris, 1956), especially pp. 6365Google Scholar, for a similar translation of the last poem cited here. Obviously much depends on the translation, but to judge from other translations, Foster is faithful to the text. The poems are discussed at length by Hermann, A., Altägyptische Liebesdichtung (Wiesbaden, 1959)Google Scholar, who gives a full bibliography. The full text of the first two poems is given by Posener, G., Catalogue des ostraca hiératiques littéraires de Deir El Medineh (Cairo, 1972), vol. 2, No. 1266.Google Scholar

85 Parlasca, K., Repertorio d'arte dell-Egitto greco-romano (Palermo, 19691977), 2 vols., reproduces 496Google Scholar mummy portraits, mostly in black and white, and gives a brief description of each, a magnificent achievement. In Mumienporträts und verwandte Denkmäler (Wiesbaden, 1966)Google Scholar, Parlasca discussed the origins of mummy portraits, their stylistic development and many other relation problems. The best introduction in English is by Shore, A. F., Portrait Painting from Roman Egypt (London, 1972)Google Scholar, but the description by Flinders, W.F. Petrie of the original finds is still gripping, Hawara, Biahmu andArsinoe (London, 1889)Google Scholar, Roman Portraits and Memphis IV (London, 1911)Google Scholar and Portfolio of Hawara Portraits (London, 1913), a set of lovely reproductions in colour.Google Scholar

86 By the fourth century AD, the mummy portraits had become more stylised and less individualistic, but still recognisably belonged to the same school.

87 The following paragraphs depend heavily on Petrie, , Hawara and Roman Portraits, and on Parlasca, Mumienporträts.Google Scholar

88 Petrie, , Hawara, p. 10.Google Scholar

89 Felix, Minucius, Octavius 31Google Scholar, mentioned that Egyptians like Athenians could marry their sisters and that Persians had sexual relations with their mothers. The Persians were repeatedly accused of incestuous behaviour in pre-Christian and Christian tracts. A Syriac text, attributed to Pseudo-Bardesanes, (Müller, C., Fragmenta Historicorum Graecorum (Paris, 1883) vol. 5.83Google Scholar) mentioned Persian laws which allowed them to marry their sisters and daughters and called attention to Persian emigrants in Egypt and Asia Minor, who followed customary practices. Eusebius in the early fourth century (Praeparatio evangelica 6.10.16) paraphrased this and Theodoret (On the Treatment of Greek Diseases, 3.96–7)Google Scholar a fifth century bishop in Syria, attacked Persians for incest: “Only the Persians do such things, in obedience to some ancient and loathsome law”, cf. Anon., Expositio totiusmundi 19, ed. Rougé, J. (Paris, 1966).Google Scholar See the excellent article by Weiss, E., Endogamie und Exogamie im romischen Kaiserreich Zeitschrift der Savigny Stiftung 29 (1908) 340ff. Was marriage within the family widespread among the Persians, or confined to a religious sect, or was this only a rumour about foreigners?Google Scholar