2. Aristotle, Pol. 1253bl–14; cf. Pol. 1260b8–27, Eudemian Ethics 1242a40–b2.

3. E.g., recently Hansen, M. H., The Polis as an Urban Centre and as a Political Community (Acts of the Copenhagen Polis Centre 4: Copenhagen, 1997), 10–12Google Scholar, citing passages from ancient texts to illustrate the conception of the polis as a community of households.

4. On slaves in the oikos see Cox, C. A., Household Interests: Property, Marriage Strategies, and Family Dynamics in Ancient Athens (Princeton, 1998), 190–4Google Scholar.

5. Xen, . Mem. 2Google Scholar.7.2. The exact number of female relatives sheltered by Aristarchos is not given, but it consisted of ‘sisters, nieces, and cousins’.

6. Gallant, T. W., Risk and Survival in Ancient Greece (Stanford, 1991), 21Google Scholar: he does not give details of the study, save that it covered 52 newly-married couples. Sallares, R., The Ecology of the Ancient Greek World (London, 1991), 196–7 statesGoogle Scholar, without citing evidence, that, even when brothers did not divide their father's estate, they did not live together.

7. Jameson, M. in Murray, O. and Price, S. (edd.), The Greek City from Homer to Alexander (Oxford, 1990), 171–95Google Scholar.

8. Cox, , op. cit., 132–5Google Scholar.

9. MacDowell, D. M., CQ 39 (1989), 15CrossRefGoogle Scholar shows that oikos could mean descendants through the male line, citing a passage from [Dem.] 43.48 which names four successive generations of men, sons and fathers, back to Stratios son of Bouselos, and identifies them as the oikos of Stratios. MacDowell (18) also cites [Dem.] 43.49–50, in which the line of descent is traced through the woman Phylomache: MacDowell describes the passage as unparalleled. Likewise the concern, well attested among Athenians, that an oikos might be left empty shows a conception of the oikos lasting beyond a single generation: MacDowell, 16.

10. Op. cit., 16.

11. Patterson, C. B. in Pomeroy, S. (ed.), Women's History and Ancient History (Chapel Hill, 1991), 60Google Scholar. Another recent example is Foxhall, in Spencer, N. (ed.), Time, Tradition and Society in Greek Archaeology (London, 1995), 134Google Scholar; see also Foxhall, in Cartledge, P., Millett, P., and von Reden, S. (edd.), Kosmos: Essays in Order, Conflict, and Community in Classical Athens (Cambridge, 1998), 52—67Google Scholar, tracing interaction both within the household and beyond. Cox, , op. cit., 132–5Google Scholar reviews critically the tendency to equate oikos and nuclear family, and points (134) to recent work, e.g., Foxhall, , CQ 39 (1989), 22–44CrossRefGoogle Scholar, which has noted that ‘oikos boundaries went beyond the nuclear family’. Though evidence for metic families is limited, the oikos no doubt had a similar importance for metics resident in Athens over a long period as for Athenian citizens: see Harrison, A. R. W., The Law of Athens: the Family and Property (Oxford, 1968) i. 195–6Google Scholar for the responsibilities of the polemarch concerning family matters of metics, and also Pomeroy, S. B., Families in Classical and Hellenistic Greece (Oxford, 1997) 37Google Scholar, suggesting that ‘metic oikoi seem to have functioned according to principles similar to those governing citizen oikoi’.

12. See V. J. Hunter, especially 108–10 on women's connection to their natal kin in Halpern, B. and Hobson, D. W. (edd.), Law, Politics and Society in the Ancient Mediterranean World (Sheffield, 1993)Google Scholar.

13. Thompson, W. E., Phoenix 21 (1967), 273–82CrossRefGoogle Scholar (marriage of first cousins) and California Studies in Classical Antiquity 5 (1972), 211–25 (remarriages)CrossRefGoogle Scholar; Isager, S., Classica et Mediaevalia 33 (1981–1982), 81–96Google Scholar; Hunter, , Journal of Family History 14 (1989), 291–311 (widows)CrossRefGoogle Scholar.

14. See Cox, , op. cit, 68–129Google Scholar, noting particularly tensions between father and son (84–8), between stepfathers and stepsons (90–2), between brothers (109–14), and between brothers-in-law (124–5).

15. On the composition and the functions of the anchisteia see Harrison, , op. cit., i. 143–9Google Scholar.

16. Harrison, , op. cit., i. 143Google Scholar; MacDowell, , op. cit., 18–19Google Scholar.

17. Or, in the case of a thetic epikleros, who should either marry her or provide her with a dowry: Harrison, , op. cit., i. 132–8Google Scholar.

18. Since ranking within the anchisteia could determine rights of succession, many Athenians who might some day inherit from a given kinsman will no doubt have borne in mind their position within the kinsman's potential anchisteia while he was still alive: but such awareness would not require any activity on the part of the anchisteia during the man's lifetime.

19. Humphreys, S. C., GRBS 27 (1986), 57–91Google Scholar, especially 87–8, from which the quotation comes.

20. The household also of course existed within a network of other social contacts, notably friendships: see Konstan, D., Friendship in the Classical World (Cambridge, 1997)CrossRefGoogle Scholar, and especially Cox, , op. cit., 168–208Google Scholar.

21. Humphreys, in Humphreys, The Family, Women, and Death: Comparative Studies (London, 2nd ed. 1993), 104–18Google Scholar analysed the c. 600 known fourth-century funerary inscriptions of Athenian citizens with deme affiliations. While she was careful to note (82) that inscriptions found alone may originally have been grouped with those of kin, the results are striking. Inscriptions which group people who were not at some stage of their lives members of the same nuclear family amount to well under 5% of the total. 17 three-generation groupings are found (113), 4 four-generation groupings (116), and 1 six-generation grouping (117). The principal occasion for the ritual commemoration of the dead at the tomb was the Genesia: as Humphreys observed (87–8), only those ancestors who were buried together could be commemorated on the day of the Genesia. In law-court speeches collateral heirs show very little interest in carrying out ritual commemoration of those whose property they are claiming (Rubinstein, L., Adoption in Fourth Century Athens [Copenhagen, 1993], 68–76Google Scholar, esp. 74).

22. Humphreys, , Classical Journal 73 (1977–1978), 97–104Google Scholar, reprinted in Humphreys, The Family, Women, and Death: Comparative Studies; and ‘Oikos and Polis’, 1–21 in The Family, Women and Death.

23. Leduc, C.in Pantel, P. S. (ed.), A History of Women in the West, Vol. 1: From Ancient Goddesses to Christian Saints (Cambridge, Mass., 1992), 235–94Google Scholar; and Leader, R. E., AJA 101 (1997), 683–99CrossRefGoogle Scholar. Cf. the comments of Strauss, B. S., Fathers and Sons in Athens: Ideology and Society in the Era of the Peloponnesian War (London, 1993), 33–53CrossRefGoogle Scholar. Foley, H., CPh 77 (1982), 1–21Google Scholar already used the opposition of oikos and polis as a tool of analysis, with important nuances. See also Patterson, C. B. in Boegehold, A. L. and Scafuro, A. (edd.), Athenian Identity and Civic Ideology (Baltimore, 1994), 200–3Google Scholar examining the public role of the oikos, including its female members, and pointing out that the opposition public/private does not entirely match the relation of polis to oikos.

24. Humphreys, , op. cit. (n. 22), 8Google Scholar: Humphreys's main point is important even though Pericles' law was not in fact a ban on marriage between Athenian and foreigners. Patterson, , Pericles' Citizenship Law of 451–50 B.C. (Salem, 1981), 29 n. 3 and 99–100Google Scholar, had already pointed out that there is no reason to suppose that Pericles' citizenship law of 451/0 was directly concerned with such marriages; and the fact that two subsequent pieces of legislation were adopted with the clear intention of banning marriages between Athenians and non-Athenians (Harrison, , op. cit., i. 26–9)Google Scholar suggest that such marriages had continued after Pericles' law.

25. This proposition would be less true if one accepted the arguments of Cohen, E. E. in Thür, G. and Vélissaropoulos-Karakostas, J. (edd.), Symposium 1995. Vorträge zur griechischen und hellenistischen Rechtsgeschichte, Korfu 1.–5. September 1995 (Cologne, Weimar, and Vienna, 1997), 57–95Google Scholar that metics were astoi and that their sons were regularly enrolled as citizens: but Cohen does not meet the arguments already advanced by Whitehead, D., The Ideology of the Athenian Metic (Cambridge, 1977), 60–1Google Scholar against the view that metics were astoi. Cf. Ogden, D., Greek Bastardy in the Classical and Hellinistic Periods (Oxford, 1996), 69Google Scholar, rejecting an earlier attempt by Walters, K., Classical Antiquity 2 (1983), 314–36CrossRefGoogle Scholar to argue that astos meant ‘free resident of Attica’. Cohen's arguments are in fact untenable in the light of the conclusions drawn by Lévy, E., Ktéma 10 (1985), 53–66Google Scholar from a careful analysis of the uses of astos and polites: cf. Mossé, C., Ktéma 10 (1985), 77–9Google Scholar on the comparable feminine forms aste and politis.

26. Perikeiromene 435–6 (Körte). See Harrison, , op. cit., i. 18, 50Google Scholar.

27. Todd, S. C., The Shape of Athenian Law (Oxford, 1993), 177Google Scholar.

28. The purpose of the law of 451/0 is still debated, and it is not clear to what extent it imposed a new pattern of behaviour, to what extent it confirmed what was already normal practice: see Ogden, , op. cit., 59–69Google Scholar and Osborne, R., Past and Present 155 (05 1997), 4–11CrossRefGoogle Scholar for recent reviews of the question.

29. On the evidence suggesting that in the later fifth century a law was passed that allowed a male Athenian citizen to marry one woman and also have legitimate children by a second see Harrison, , op. cit., i. 15–17Google Scholar; cf. Todd, , op. cit., 117Google Scholar suggesting that the law of 451/0 may at least have ceased to be applied in the later fifth century. Ogden, , op. cit., 72–7Google Scholar accepts that legislation was passed during the later Peloponnesian War to allow Athenian men to have legitimate children by two wives. The limited evidence on whatever change to laws on hereditary citizenship may have occurred in the later fifth century does not suggest how an Athenian with two wives would have managed his domestic circumstances – whether, for instance, both wives would have lived in the same household. See also Cox, , op. cit., 172–7Google Scholar.

30. Harrison, , op. cit., i. 26–8Google Scholar.

31. Scafuro, A., Athenian Identity and Civic Ideology, 158–65Google Scholar.

32. See Osborne, op. cit., arguing that the much greater prominence of women on sculpted funerary monuments from (probably) the third quarter of the fifth century B.C. is due to the changed status of the wife and mother in Athenian society. Other recent studies of these representations of women take different approaches: Stears, K. in Spencer, N. (ed.), Time, Tradition and Society in Greek Archaeology (London, 1995), 109–31Google Scholar also acknowledges the impact of Pericles' law and women's importance to the familial group, but with nuances about the exposure of women to public view, while Leader, op. cit., stresses the projection of an idealized conception of the oikos in the way women are shown. Cf. Shapiro, H. A., AJA 95 (1991), 629–56CrossRefGoogle Scholar, on the iconography of mourning in Athenian art, particularly vases showing funerary scenes.

33. Kallet-Marx, L. in Rosen, R. M. and Farrell, J. (edd.), Nomodeiktes: Greek Studies in Honor of Martin Ostwald (Ann Arbor, 1993), 140Google Scholar.

34. Morris, I., Death-ritual and Social Structure in Classical Antiquity (Cambridge, 1992), 141–4CrossRefGoogle Scholar.

35. Kallet-Marx, op. cit.

36. On newly married couples living with the husband's family see Gallant, , op. cit., 21Google Scholar.

37. There are recent discussion by Gallant, , op. cit., 17–19Google Scholar; Sallares, , op. cit., 148–50Google Scholar; Pomeroy, , op. cit., 23Google Scholar; Cox, , op. cit., 120–2Google Scholar.

38. Hesiod, , Works and Days 696–8Google Scholar; Plato, , Laws 721bGoogle Scholar, 785b, cf. Rep. 460e; Aristotle, , Pol. 1335a28–30Google Scholar.

39. Xen, . Oik. 7Google Scholar.5; Aristotle, , Pol. 1335a29–30Google Scholar.

40. Hunter, , Policing Athens (Princeton, 1994), 29–33Google Scholar. To defend her interests at law, if need arose, the widow would need a male kyrios, though not necessarily from within her household: see Foxhall, in Foxhall, L. and Lewis, A. D. E. (edd.), Greek Law in its Political Setting: Justifications not Justice (Oxford, 1996), 149–50Google Scholar on the range of potential kyrioi available to a woman.

41. Hunter, , op. cit. (n. 13), 298Google Scholar.

42. Except in so far as childbirth posed a particular danger for women: on the dangers of childbirth see Demand, N., Birth, Death, and Motherhood in Classical Greece (Baltimore, 1994), 71–86Google Scholar.

43. See Sallares, , op. cit., 107–29Google Scholar: he constructs (109–16), with considerable emphasis on the problems and uncertainties involved, a life table for men (110) based on the skeletal material from Greece in the period c.650–350 aged by Angel: the table gives a man at age 30 a further lifeexpectancy of 15.9 years. Cf. Gallant, , op. cit, 19–20Google Scholar.

44. Gallant, op. cit., 29–30Google Scholar supposes a life-cycle of 24 years for an ancient household, and that within part at least of the last 6 years of the 24 the household would consist of a widow and her sons, besides any slaves. Exactly how Gallant calculates the figure of 24 is not clear: he supposes an age at marriage for men of 25–30 (18) and an average life-expectancy for men of 40 (20), though he does not say what further life-expectancy he supposes for a man who reached 30. He supposes that the life-cycle of the household would repeat itself in a new generation after 24 years, but it is not clear how that is to be reconciled with marriage for men at age 25–30. See also Cox, , op. cit., 141–3Google Scholar.

45. Thompson, , California Studies in Classical Antiquity 5 (1972), 211–25CrossRefGoogle Scholar; Isager, op. cit; Hunter, op. cit. (n. 13); Cox, , op. cit., 152–5Google Scholar.

46. Coresidence was an important feature of the typical Athenian household, but there must have been numerous exceptions where members of the household (e.g., soldiers, sailors, traders, and itinerant craftsmen) would often be absent for long periods: see Cox, , op. cit., 155–66Google Scholar. Coresidence therefore could not be legally significant.

47. See, e.g., Knox, B., Word and Action: Essays on the Ancient Theater (Baltimore, 1979), 20–3Google Scholar: but note the arguments of Redfield, J. in Vernant, J.-P. (ed.), The Greeks (Chicago, 1995), 163Google Scholar that ‘Athenian drama only permits the representation of domestic life as triply separated from immediate experience’ by being represented in public by men for men, as if it occurred in public, and set in heroic time or, in comedy, in a ‘fanciful suspension of time, space, cause, and effect’.

48. Strauss, op. cit.; also Cox, , op. cit., 84–8Google Scholar. Cf., e.g., Gardner, J. F., Greece & Rome 36 (1989), 51–62CrossRefGoogle Scholar on husbands' anxieties over their wives' behaviour; and, generally on tensions among kinsfolk, including cooperation and discord between siblings, Cox, , op. cit., 68–129Google Scholar. Sissa, G. in Andreau, J. and Bruhns, H. (edd.), Parenté et Stratégies dans L'Antiquité Romaine. Actes de la Table Ronde des 2–4 Octobre 1986 (Paris and Rome, 1990), 199–223Google Scholar sees kinship as a source of dispute (203–5) and marriage between kin as a way of healing disputes (208–9).

49. Humphreys, , op. cit (n. 22), 5–7Google Scholar; Hunter, , op. cit. (n. 40), 43–69Google Scholar.

50. Humphreys, , op. cit. (n. 22), 5 (from which the quotation is taken)Google Scholar. Cf. Dover, K., Greek Popular Morality in the Time of Plato and Aristotle (Oxford, 1974), 273–6Google Scholar on the Athenian's duty to parents first, other kin second, and friends third; and the view (275) that ‘litigation within the family was a matter of great embarrassment’. Humphreys may be right (7) that written law did as much to create disputes as to prevent them. However, when she suggests that at an earlier period family disputes were settled by (7) ‘well-informed public opinion in a village community’ (cf. 9, ‘local communities which had once mediated between oikos and polis’), this supposed earlier stage of Athenian society is not clearly defined or located in time, and may be questioned. Heads of household who took disputes to court may have had various motives for doing so: see Foxhall, , op. cit. (n. 40), 137Google Scholar on lawcourts as ‘one of a number of arenas in which males competed with each other’.

51. Dover, , op. cit., 273–6Google Scholar, from which the translation of Isaios is taken.

52. Hunter, , op. cit. (n. 40), 29–33Google Scholar. See Günther, L.-M., Historia 42 (1993), 308–25Google Scholar for a general review of the position of widows.

53. See Cohen, E. E., Athenian Economy and Society: a Banking Perspective (Princeton, 1992), 82–90Google Scholar on the ‘banking household’ (though one might question his view [87] that ‘by the fourth century, however, the traditional oikos had been significantly transmuted' if it is intended as generalization rather than a comment on the special case of the bankers). See also Trevett, J., Apollodoros the Son of Pasion (Oxford, 1992), 1–49Google Scholar on the family of the banker Pasion.

54. On incest see Karabélias, E. in Thür, G. (ed.), Symposion 1985. Vorträge zur griechischen und hellenistischen Rechtsgeschichte, Ringberg 24.–26. Juli 1985 (Cologne and Vienna, 1989), 233–51Google Scholar; and Mülke, C., ZPE 114 (1996), 37–55Google Scholar.

55. There is a considerable modern literature on the subject of infanticide in Classical Greece, and in particular in Athens. The arguments of Engels, D., CPh 75 (1980), 112–20Google Scholar against the possibility that exposure was practised were refuted by Golden, M., Phoenix 35 (1981), 316–31CrossRefGoogle Scholar and Harris, W. V., CQ 32 (1982), 114–16CrossRefGoogle Scholar. See also Patterson, , TAPA 115 (1985), 103–23Google Scholar; Golden, , Children and Childhood in Classical Athens (Baltimore, 1990), 86–9Google Scholar; Sallares, , op. cit., 151–7Google Scholar; Gallant, , op. cit., 21, 133Google Scholar; Riddle, J. M., Contraception and Abortion from the Ancient World to the Renaissance (Cambridge, Mass., 1992), 10–15Google Scholar; Demand, , op. cit., 6–8Google Scholar; Pomeroy, , op. cit., 120Google Scholar. Cf. Huys, M., Ancient Society 27 (1996), 47–74CrossRefGoogle Scholar, who considers the evidence for infanticide at Sparta (where it is generally supposed that exposure was certainly practised), and shows that such evidence is closely related to ancient utopianist thinking. Oldenziel, R. in Blok, J. and Mason, P. (edd.), Sexual Assymetry (Amsterdam, 1987), 87–107Google Scholar provides the historiographical context of modern discussion of ancient infanticide.

56. Golden, op. cit.

57. Riddle, , op. cit, 15–19Google Scholar.

58. On contraception and abortifacients, including pennyroyal, see Riddle, , op. cit., 57–61Google Scholar and Eve's Herbs: a History of Contraception and Abortion in the West (Cambridge, Mass., 1997), 35–63, 75–7Google Scholar. There are references to pennyroyal in Lysistrata 85–9, Achamians 860–77, and Peace 706–12. See Riddle, , op. cit., 53–4Google Scholar and op. cit., 1997, 46–7, and Scarborough, J. in Faraone, C. A. and Obbink, D. (edd.), Magika Hiera: Ancient Greek Magic and Religion (Oxford, 1991), 145Google Scholar on the properties of pennyroyal as an emmenagogue and abortdfacient, and the dangers of an overdose. Scarborough, ibid., 144–5, and Riddle, , op. cit., 1992, 59Google Scholar and op. cit., 1997, 46–7 see references to these properties in Lysistrata 85–9 and Peace 706–12: but pennyroyal is not taken as a herb in Lysistrata 85—9 and is recommended to be taken by a man in Peace 706—12. Fisher, N. R. E., Classics Ireland 3 (1996), 82 n. 19CrossRefGoogle Scholar questions whether Riddle is right to see allusions to the use of pennyroyal as a contraceptive in Lysistrata 85–9 and Peace 706–12. The Aristophanic passages show that pennyroyal was well-known in Classical Athens, and, if any of the three passages referred to pennyroyal as a means of avoiding unwanted pregnancy (as Scarborough and Riddle suppose), we could safely assume that the herb's abortifacient quality was widely known to Athenians: but further argument is needed if the view of Scarborough and Riddle is to be sustained.

59. Riddle, , op. cit. (n. 55), 74–6Google Scholar.

60. Riddle, , op. cit. (n. 55), 23–4Google Scholar (early abortion), 76–82 (abortifacients).

61. Harrison, , op. cit., i. 72–3Google Scholar.

62. It is worth remembering that infant mortality must have been high: no figures are available, but Golden, , Greece & Rome 35 (1988), 155CrossRefGoogle Scholar suggests that between 30% and 40% of children in Greco-Roman society may have died in the first year of their life.

63. In certain circumstances, an epikleros who was already married could be required to divorce and enter into a second marriage with a kinsman: Harrison, , op. cit., i. 11–12Google Scholar.

64. Cohn-Haft, L., JHS 115 (1995), 1–14CrossRefGoogle Scholar: see also Cox, , op. cit, 71–3Google Scholar.

65. Isaios 2.7–12: Cohen-Haft, , op. cit, 10Google Scholar.

66. Harrison, , op. cit., i. 74–7Google Scholar.

67. The evidence is in Dem. 39.39.

68. Harrison, , op. cit, i. 151–3Google Scholar. On the estate of Konon, Lysias 19.39–40: the succession may have been affected by the fact that Konon had property in Cyprus.

69. Harrison, , op. cit., i. 67Google Scholar; Ogden, , op. cit, 38–9Google Scholar.

70. Rubinstein, , op. cit, 76–86Google Scholar, especially 80: see also Cox, , op. cit, 125–8, 148–51Google Scholar.

71. Harrison, , op. cit, i. 79–80Google Scholar on the graphê argias and graphê (or possibly dikê) paranoias.

72. Harrison, , op. cit, i. 77Google Scholar (father), 112 with n. 4 (epikleros), 117–19 (orphans and epikleroi).

73. Harrison, , op. cit, i. 104–5Google Scholar.

74. Humphreys, , ‘Oikos and Polis’, 5Google Scholar.

75. Davies, J.K., Wealth and the Power of Wealth in Classical Athens (New York, 1981), 73–87Google Scholar; cf. Davidson, J. N., Courtesans and Fishcakes: the Consuming Passions of Classical Athens (London, 1997), 183–6Google Scholar.

76. Harrison, , op. cit, i. 73Google Scholar: the law is cited by Plutarch, , Solon 23Google Scholar.

77. Fantham, E., Foley, H. P., Kampen, N. B., Pomeroy, S. B., and Shapiro, H. A. (edd.), Women in the Classical World: Image and Text (Oxford, 1994), 114Google Scholar.

78. Harrison, , op. cit, i. 35–6Google Scholar: see also Roy, , Greece & Rome 44 (1997), 11–22CrossRefGoogle Scholar, esp. 13–15.

79. Harrison, , op. cit, i. 37–8Google Scholar: the law is cited by Aischines 1.13.

80. Isaios 7.30 paraphrases what may be the same law as follows: ‘All those who are about to die make their own precautions so that their oikoi will not become empty, and so that there will be someone to offer sacrifices at their tombs and perform all the customary rites for them. Wherefore, even if they are to die childless, they do in any case leave adopted children. And it is not only private individuals who make this decision; it is also a decision of public concern made by the polis as a community [to koinon tês poleôs]. For it (the polis) orders by law the Archon to take care of the oikoi so that they do not become empty’ (translation by Rubinstein, , op. cit, 105)Google Scholar.

81. Rubinstein, op. cit.

82. Rubinstein, , op. cit., 25–8, 121–3Google Scholar: see also Cox, , op. cit., 148–51Google Scholar.

83. Rubinstein, , op. cit., 22–5Google Scholar (testamentary adoption), 35 (posthumous adoption). The procedures of posthumous adoption are obscure, and it is not clear whether it had to be ratified by the archon or by a court before the adoptee was allowed to succeed to the inheritance of his adoptive father (Rubinstein, 41).

84. In the case of epikleroi Rubinstein, , op. cit., 87–104Google Scholar concludes that ‘adoption was not a necessary supplement to the epikleros-institution, and that the decision to continue the oikos of the epikleros’ father rested entirely with his heirs'. In the case of intestate heirs Rubinstein concludes (105–12) that there was no legal obligation to be posthumously adopted or, if there was, it was not enforced in the fourth century. She also concludes that any moral obligation to be adopted was not strong enough to compel the heir to be adopted.

85. It also appears that posthumous adoption, rather than being an orderly process supervised by the state, was frequently contested and difficult to control. Rubinstein, , op. cit., 25–9Google Scholar.

86. Since this article was written there has appeared Hansen, M. H., Polls and City-state: an Ancient Concept and its Modem Equivalent (Acts of the Copenhagen Polis Centre 5, Copenhagen, 1998)Google Scholar, of which two sections – though written from a very different viewpoint – are particularly relevant to the arguments presented here: pages 86–91 entitled ‘The opposition between the public and the private’ (including at 86: ‘The Athenians practised a separation between a public and a private sphere of life’) and pages 135–7 entitled ‘Was the oikos a civic or a private institution?’ (including at 136–7: ‘In Athens … the oikos belonged basically in the private sphere’).