Hostname: page-component-7c8c6479df-xxrs7 Total loading time: 0 Render date: 2024-03-19T04:34:22.869Z Has data issue: false hasContentIssue false

The Size of the Roman Population: Beloch and the Meaning of the Augustan Census Figures1

Published online by Cambridge University Press:  24 September 2012

Elio lo Cascio
Affiliation:
University “Federico II” of Naples

Extract

The importance of Beloch's Bevölkerung der griechisch-römischen Welt and its influence on subsequent research in ancient demography can hardly be overstated. This book represents the key-stone of all modern investigation on size, structure, and, to a certain extent, dynamics of ancient populations. It was the first overall scientific treatment of the subject and it is still unparalleled in its scope. An attempt at its critical evaluation is not just an historiographical exercise: we must come to terms with Beloch's Bevölkerung, because its detailed treatment of most of the topics concerning the population of the ancient world is at the root of all modern debates and controversies.

Type
Articles
Copyright
Copyright ©Elio lo Cascio 1994. Exclusive Licence to Publish: The Society for the Promotion of Roman Studies

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

2 Steinberg, S. (ed.), Geschichtswissenschaft der Gegenuart in Selbstdarstellungen 2 (1926), 127, at 2f.Google Scholar; what Beloch says in his autobiography is at the basis of the biographies by Christ, K., From Gibbon to Rostovtzeff (1972), 248–85Google Scholar, and Momigliano, A., Dizionario biografico degli Italiani 8 (1966)Google Scholar, s.v., now in an English translation by Cornell, T. J. in Bowersock, G. W. and Cornell, T. J. (eds), A. Momigliano, Studies in Modern Scholarship (1994)Google Scholar.

3 A. Momigliano, ‘Dopo Max Weber?’, ASNP3 8 (1978), 1315–32, at 1324ff. = Sesto contributo alla storia degli studi classici e del mondo antico I (1980), 294–3, at 303ff.Google Scholar

4 ibid., 1320 = 300. See now N. F. Parise in Polverini, op. cit. (n. 1), 107–11.

5 Or even ‘perhaps the most important branch’: Die Bevölkerung, V.

6 And that explains the famous, acerbic critique by B. Croce of Beloch's conception of history: Intorno alle condizioni presenti della storiografia in Italia, IV, La storiografia sociale e politica’, La critica 27 (1929), 241–63Google Scholar, at 253 = Storia della storiografia italiana nel secolo decimonono (2nd edn, 1964), 11, 246Google Scholar.

7 Die Bevölkerung, 454f.

8 A contradiction stressed by C. Ampolo in Polverini, op. cit. (n. 1), 85f.

9 ‘Sulla popolazione dell'antica Sicilia’, RFIC 11 (1873–1874), 545–62.

10 Steinberg,op. cit.(n.2), 6.

11 ‘Die römische Censusliste’, Rh. Mus. N.F. 32 (1877), 227–48; Der Italische Bund unter Roms Hegemonie (1880), ch. IV.

12 Nuove osservazioni sulla popolazione antica della Sicilia’, Arch. Stor. Sic. N. S. 20 (1895), 6370Google Scholar; Zur Bevölkerungsgeschichte des Altertums’, Jahrbücher für Nationalökonomie und Statistik 68 (3. F., XIII) (1897), 321–43Google Scholar (a reply to Seeck, who had criticized the methods and the conclusions of Beloch's Bevölkerung, in ‘Die Statistik in der alten Geschichte’, ibid., 161–76); ‘Die Bevölkerung Galliens zur Zeit Caesars’, Rhein.Mus. N.F. 54 (1899), 414–45.

13 Una nuova storia della popolazione d'ltalia’, Nuova Antologia 22 (1887), vol. XCV (3 s. XI), 4861Google Scholar; cf. La popolazione d'ltalia nei secoli XVI, XVII e XVIII’, Bulletin de l'Institut international de statistique 3 (1988), 1, 142Google Scholar; ‘Die Bevölkerung Europas im Mittelalter’, ibid., 3 (1900), 405–23; ‘Die Bevölkerung Europas zur Zeit der Renaissance’, ibid., 765–86.

14 As it emerges from the picture that Beloch himself gave of his scientific achievements in his autobiography.

15 Bevölkerungsgeschichte Italiens I (1937), ed. De Sanctis, G.Google Scholar; 11 (1939), ed. G.De Sanctis; III (1961), ed. L. Pareti and W.Hagemann. To take just an example, Beloch's work is at the basis of the reconstruction of the history of the Italian population given by Cipolla, C. M., ‘Four centuries of Italian demographic development’, in Glass, D. V. and Eversley, D. E. (eds), Population in History (1965), 570–87.Google Scholar

16 Steinberg, op. cit. (n. 2), 11.

17 Die Bevölkerung, vi.

18 Die Bevölkerung, 502. It must be pointed out that the book was conceived as the first of a series of Historische Beiträge zur Bevölkerungslehre.

19 Bevölkerung, vf.

20 S. Mazzarino, Aspetti sociali del quarto secolo (1951), ch. V, I; A. Chastagnol, ‘Le ravitaillement de Rome en viande au Ve siècle’, Rev. Hist. 210 (1953), 13ff.; for a reappraisal of Mazzarino's calculations, see Hodges, R. and Whitehouse, D., Mohammed, Charlemagne and the Origins of Europe. Archaeology and the Pirenne Thesis (1983), 48ff.Google Scholar; and E. Lo Cascio, Quaderni catanesi di studi classici e medievali 11 (1990) [Studi in memoria di Santo Mazzarino in], 67–92.

21 Mostly the Bevölkerung is seen in a very favourable way, even if there are severe judgements: see the statement by P. D. Warden and R. S. Bagnall, ‘The forty thousand citizens of Ephesus’, Class. Phil. 83 (1988), 220–3, at 220, n.1, according to whom Beloch's Bevölkerung ‘though always cited, is more often a convenient target than a model’.

22 It is not by chance that in the recent book by Parkin, T. G., Demography and Roman Society (1992)Google Scholar, which gives an overview of contemporary research, just a page or so is devoted to the size of ancient populations and the means of establishing it: the few figures that Parkin gives (for Rome, Italy and the whole Empire) are presented as ‘tentative estimates, which seem to represent the broad consensus of modern scholarly opinion… and which have remained largely unchanged since the appearance of Beloch's masterly work in 1886’ (5).

23 See the remarks by Duncan-Jones, R. P., ‘Aqueduct capacity and city population’, Society for Libyan Studies, 9th Annual Report (19771978), 51Google Scholar, criticizing the suggestion of Lloyd, J. A. and Lewis, P. R., ‘Water supply and urban population in Roman Cyrenaica’, Society for Libyan Studies, 8th Annual Report (19761977), 35–40, at 36Google Scholar; idem, The Economy of the Roman Empire (2nd edn, 1985), 261Google Scholar; Leveau, Ph. and Paillet, J. L., L'alimentation en eau de Caesarea de Maurétanie et l'aqueduc de Cherchell (1976), 15ff.Google Scholar; for the evidence provided by theatres (and amphitheatres), see Gallo, L., ‘La capienza dei teatri e il calcolo della popolazione: il caso di Atene’, in Studi Salernitani in memoria di Raffaele Cantarella (1981), 271–89Google Scholar.

24 ‘Die Bevölkerung Galliens zur Zeit Caesars’, op. cit. (n. 12), 443f.

25 ‘Zur Bevölkerungsgeschichte des Altertums’, op. cit. (n. 12), 323f. on Galen, de propriorum animi cuiuslibet affectuum dignotione et curatione V.49 Kühn; see now Duncan-Jones, R. P., The Economy of the Roman Empire (1985), 259ff.Google Scholar; Warden and Bagnall, op. cit. (n.20), 220–3.

26 Hopkins, K., Conquerors and Slaves (1978), 2f.Google Scholar, n. 4.

27 L'impero romano (1956), 35f., n. 1. The most important interventions for or against Beloch's Bevölkerung around the turn of the century were collected in an Italian translation in Pareto, V. (ed.), Biblioteca di storia economica IV (1909)Google Scholar, as an appendix to the Italian translation of the Bevölkerung itself.

28 Beloch himself speaks in his autobiography of a ‘geschlossenes System’: Steinberg, op. cit. (n. 2), 12.

29 Meyer, Ed., ‘Die Zahl der römischen Bürger unter Augustus’, Jahrbücher für Nationalökonomie und Statistik 70 (3. F. XV) (1898), 5965Google Scholar, commenting on E. Kornemann, ‘Die römischen Censuszahlen als statistisches Material: zum Strei t Seeck-Beloch’, ibid., 69 (3. F XIV) (1897), 291–6. The estimate of the whole population of the Roman Empire as 54 million in the Augustan age, which appears at the end of the volume, is based on the estimate of the population of Italy; it is worth noticing that Beloch himself in a subsequent essay posited the highest level reached by the population of the Empire in Caracalla's times at 100 million: Die Bevölkerung im Altertum’, Zeitschrift für Sozialwissenschaft 2 (1899), 618ff.Google Scholar (Italian trans, in Biblioteca dell'economista 5 XIX (Scritti di statistica teorica e applicata) (1908), 464 ss.Google Scholar).

30 Bevölkerung, 254ff., on Diod. 1.31; cf. 1.80; Jos., Bell. 11.385; Philo, in Flaccum 43. The reason given by Beloch for his refusal to give credit to Philo is the alleged desire by Philo to ‘present the community of the Jews as important as possible’: Bevölkerung, 258. On the antisemitism of Beloch and its probable motives, see Momigliano, op. cit. (n. 2), Dizionario biografico, 42f. The plausibility of the piece of information given by Josephus is stressed by, e.g., A. H. M. Jones, Ancient Economic History (1948), 10; Finley, M. I., The Ancient Economy (2nd edn, 1985), 31, 215 n.49Google Scholar; and Bowman, A. K., Egypt after the Pharaohs (1986), 17f.Google Scholar, 90, 238; see also D. Delia, ‘The population of Roman Alexandria’, TAPhA 118 (1989), 275–92, at 282f. I am not convinced by the arguments put forward by D. W. Rathbone, ‘Villages, land and population in Graeco-Roman Egypt’, PCPS n.s. 36 (1990), 103–42, at 105ff. (followed by Bagnall, R. S. and Frier, B. W., The Demography of Roman Egypt (1994), 53f.CrossRefGoogle Scholar), in order to reject the value of Josephus' testimony: see E. Lo Cascio, ‘Civium capita’. Le cifre dei censimenti e l'evoluzione demografica dell'eta repubblicana, (forthcoming), ch. 1.2.

31 That is why Beloch reached an estimate of Egyptian population in Philo's times that has no basis whatsoever in the ancient evidence, namely 5 million inhabitants, ‘that is 180 per square km’: Bevölkerung, 258; he seems, how ever, to have forgotten this conclusion when he wrote p. 499, for in a discussion of the increase in the Greek population of the East after Alexander, he apparently accepted the validity of Josephus' datum: ‘Egyptian population rose, between 300 B.C. to A.D. 70, if our information is correct, from 3 million to 8 million or so’.

32 So Drinkwater, J. F., Roman Gaul. The Three Provinces (1983), 169f.Google Scholar

33 Bevölkerung, 400f.

34 Bevölkerung, 476, with n. 2.

35 Bevölkerung, 477f. The estimates given now by Engels, D., Roman Corinth. An Alternative Model for the Classical City (1990), 84Google Scholar, are not far from these: 80,000 urban and 20,000 rural population; see, however, the critical remarks on the way in which Engels arrives at his estimates by Whittaker, C. R., Land, City and Trade in the Roman Empire (1993), ix, 5Google Scholar, and by R. P. Sailer, CPh 86 (1991), 351–7, at 352f.

36 Unless there is a marked increase in the levels of agricultural productivity, a genuine urban growth and a high degree of urbanization can be achieved only if rural population increases as well: see Boserup, E., Population and Technology (1981), ch. 6.Google Scholar It is certainly a mistake to see Greek and Roman ‘normal’ towns as simply ‘agro towns’: see, for Roman Italy, the conclusion reached by P. D. A. Garnsey, ‘Where did the Italian peasants live?’, PCPS n.s. 25 (1979), 1–25.

37 See the complete list of the figures and the almost complete list of the sources for them in Toynbee, A., Hannibal's Legacy (1965), 1, 438ff.Google Scholar

38 R.G. 8; a fragment of the Fasti Ostienses seems to give a different figure for the census of A.D. 14, but the most plausible explanation of the seeming diversity is that the figure of the F.O. is the result of a mistranscription of the right figure by the stonecutter: see now Nicolet, C., ‘Les Fastes d'Ostie et les recensements Augustéens’, in Epigrafia. Actes du colloque en mémoire de Attilio Degrassi (1991), 119–31Google Scholar; for the different figures given for the censuses of 28 B.C. and A.D. 14, in the Armenian version of Eusebius, by Jerome, by Georgius Syncellus, and by Prosper Aquitanus, see Bevölkerung, 371, n. 1.

39 A detailed treatment of the whole issue will be found in the first chapter of my forthcoming book ‘Civium capita’. Le cifre dei censimenti e l'evoluzione demografica dell'età repubblicana; see also Lo Cascio, E., ‘La dinamica della popolazione in Italia da Augusto al III secolo’, in L'Italie d'Auguste à Dioclétien, Proceedings of the International Conference, Ecole française de Rome, 25–25 March 1992 (forthcoming)Google Scholar.

40 Italian Manpower (1971), esp. the introduction and Part 1, ch. IX, see also Schulz, O. Th., ‘Die Zensus des ersten Prinzeps (Augustus)’, Mnemosyne3 5 (1937), 161–92Google Scholar; T. Frank, ‘Roman census statistics from 225 tO 28 B.C.’, CPh 19 (1924), 329–41; Jones, A. H. M., Ancient Economic History (1948), 4K.Google Scholar; see also Wiseman, T. P., ‘The census in the first century B.C.’, JRS 59 (1969), 5975Google Scholar, at 71ff.

41 As evidenced by the Tabula Heracleensis, FIRA I2 13, ll. 142ff. (on which see Cascio, E. Lo, ‘Le professiones della Tabula Heracleensis e le procedure del census in età cesariana’, Athenaeum 78 (1990), 287–317, at 308ff.)Google Scholar.

42 See above, n. 11.

43 See Mommsen, , ‘Das Verzeichniss der italischen Wehrfähigen aus dem Jahre 529 der Stadt’, Hermes 11 (1876), 4960Google Scholar (repr. in Röm. Forschungen 11, 382–406); Beloch, op. cit. (n. 11), ‘Die römische Censusliste’, 245ff., put an ‘Anhang’ at the end of his paper, in order to discuss Mommsen's article, which appeared, as Beloch says, ‘when the present work was already written in its final form’.

44 This is the solution advocated in the Italische Bund, op. cit. (n. 11): the exclusion of proletarii and seniores from the enumeration would date from the middle of the third century, and from this same date cives sine suffragio would have been included. In enumerating all the adult males, Augustus, therefore, would have come back to the older practice.

45 To suppose that the Republican figures referred, always, just to the iuniores assidui would have implied much too high a density of population of the ager Romanus in certain periods of Republican history; on the other hand, Beloch gives up the hypothesis put forward in the Italische Bund, op. cit. (n. 11), that only from the middle of the third century onwards would the figures have referred to the iuniores assidui, probably because for this hypothesized change no plausible reason could be offered. The other possible motive for his change of mind could have been that Beloch was, by now, convinced that the free population of Italy was declining in the last two centuries of the Republic.

46 Bevölkerung, 438ff.

47 In vol. 11 (1891), 443–56. Meyer had already declared his approval of Beloch's new interpretation of the Republican census figures as representing all adult males in his letter to Beloch of 23 October 1886, in which he thanked him for sending a copy of the Bevölkerung. In his reply of 30 October, Beloch wrote that he was pleased by Meyer's agreement on the ‘Auffassung der römischen Censuszahlen als Summe sämmtlicher Burger’; it is certainly interesting that, in fact, there is no word in Meyer's letter about Beloch's interpretation of the Augustan census figures (I should like to thank Leandro Polverini for letting me read these two letters).

48 ‘Ich habe mich lange gegen diese von Beloch gegebene Lösung gestraübt’: Bevölkerung, 453 n. 1.

49 The argument was reaffirmed, e.g., by G. Tibiletti, ‘The “comitia” during the decline of the Roman Republic’, SDHI 25 (1959), 126, n. 129; and repeated by Jongman, W., The Economy and Society of Pompeii (1988), 66Google Scholar.

50 On the proposal by Nicolet, op. cit. (n. 38), 128, to understand what Augustus says in the first edict from Cyrene as an implicit argument in favour of Beloch's solution see my comment s in Lo Cascio, op. cit. (n. 39), ‘La dinamica’.

51 See especially the critical evaluation of modern views given by Brunt, P. A., Italian Manpower (1971), 1525Google Scholar.

52 Note the use of the Republican formula for a census which is conceived as a resumption of an old tradition (as is made clear by the sentence which closes ch. 8 of the Res Gestae: ‘multa exempla maiorum exolescentia iam ex nostro saeculo’). Wiseman, op. cit. (n.40), 71, stresses against Pieri, G., L'histoire du cens jusqu'à la fin de la république romaine (1968), 192ff.Google Scholar, (rightly, in my opinion), the absolutely traditional character of at least the census of 28, in terms of aims and methods.

53 Eus., Chron., p. 146 Schoene (= Sync. 602, 17).

54 See Mosshammer, A. A., The Chronicle of Eusebius and Greek Chronographic Tradition (1979)Google Scholar, ch. 1. Neither the Armenian version nor Jerome seem to be of any use in this particular case.

55 N.H. III.3.28.

56 Bevölkerung, 374ff. Beloch's followers did not accept the connection beween the census of citizens and the provincial censuses: see Brunt, op. cit. (n.51), 113f. The difference between the two kinds of census is vigorously argued by, e.g., Braunert, H., ‘Der römische Provinzial-census und der Schätzungsbericht des Lukas-Evangelium’, Historia 6 (1957), 192ff.Google Scholar

57 To suppose that there could have been a shift in the aims of the census, with a purely demographic concern coming to the forefront, seems to me to attribute anachronistically modern preoccupations to the Roman government: certainly one cannot read the statement by Claudius about the function of his census as revealing this sort of purely statistical interest: ILS 212, ll. 65ff.: see Lo Cascio, op. cit. (n. 39), ‘Civium capita’, 1. 2.

58 Dio LIII.22.5.

59 Plin., N.H. XXXIII.16. Clason, Already O., Römische Geschichte seit der Verwüstung Roms durch die Gallier 1 (1873), 54Google Scholar, maintained that the use of capita libera in this passage would show that the earliest census figures referred to the whole population, and not just to the adult males; this hypothesis is restated by T. Frank, ‘Roman census statistics from 509 to 225 B.C.’, AJPh 51 (1930), 313–24, at 314, n. 5 (without any quotation of Clason); in this same sense, see now Coarelli, F., ‘Demografia e territorio’, in Storia di Roma Einaudi, I, Roma in Italia (1988), 317–39, at 320Google Scholar. Other theories about the use of the expression in this passage of Pliny and the reason for it have been put forward by Momigliano, A., ‘Timeo, Fabio Pittore e il primo censimento di Servio Tullio’, Miscellanea di studi alessandrini in memoria di Augusta Rostagni (1963), 180–7Google Scholar = Terzo contributo alia storia degli studi classici e del mondo antico (1966), 649–56, at 652f.Google Scholar; Brunt, P., Italian Manpower (1971), 113Google Scholar; see also den Boer, W., ‘Demography and Roman history: facts and impressions’, Mnemosyne4 26 (1973), 2946CrossRefGoogle Scholar, at 42; Alfisi, E., ‘Le fonti dei censimenti romani in Plutarco e Plinio’, Atti Cesdir 6 (19741975), 929Google Scholar, at 20ff.: all of them imply that capita libera must have included women and children. A different explanation of the occurrence of the expression in this passage of Pliny is advanced in Lo Cascio, op. cit. (n. 39), ‘Civium capita’, 2.3.1.

60 Livy, V.30.8; VI.12.5; XXII.57.11; XXIII.19.5; XXVI.47.1; XXVII.19. 2; XXIX.29.3; XXXI.21.18; 40.3; XL.38.6; XLII.41.11; XLV.24.11; Gaius 1.166 a; Ulpian XI.5; Digest XIV.2.2.2 (Paul.); XXVI.I.I pr. (Paul. [Serv.]).

61 See Livy VI.12.5;XXII.57.II; XXVII. 19.2; XXXI.21.18.

62 ‘Age and sex patterns of mortality: Model Life Tables for underdeveloped countries’, United Nations, Department of Social and Economic Affairs, Population Studies 22 (1955); ‘Methods for population projections by sex and age’, ibid. 25 (1956); ‘Model Life Tables for developing countries’, ibid. 77 (1982); Coale, A. J. and Demeny, P., Regional Model Life Tables and Stable Populations (2nd edn, 1983)Google Scholar.

63 On the theory of the stable population developed by Lotka at the beginning of this century, see Livi-Bacci, M., Introduzione alla demografia (1981), 372ff.Google Scholar; Newell, C., Methods and Models in Demography (1988), Part 11.Google Scholar

64 See the admirably lucid account of the use of life tables in historical demography by M. Livi-Bacci, ‘Fonti e metodi per lo studio delta demografia’, in Comitato ita liano per lo studio della demografia storica. Le fonti della demografia storica in Italia (1972), 1, 2, 955–98, at 973ff. ( = Nuovi metodi della ricerca storica (1975), 311–39, at 324ff.); and idem, ‘Sull'applicazione delle tecniche di analisi basate sulla teoria della popolazione stabile agli studi di demografia storica’, in Atti della XXV riunione della Società italiana di statistica (1969), 917–32; idem, ‘Una disciplina in rapido sviluppo: la demografia storica’, Quademi storici 17 (1971), 279–98; see also Hollingsworth, T. H., Historical Demography (1969), 339–53Google Scholar; de Walle, E. Van, ‘De l'emploi des modèles en demographie historique’, Annales de démographie historique (1972), 153–77.Google Scholar

65 Or immune even from wide fluctuations in birth-rate and death-rate and in pretransitional populations from ‘crisis mortality’. But this does not seem to legitimate a sceptical attitude towards the models based on the theory of the stable population, provided that the population under study is big enough and the analysis is a long-term one.

66 Except for some of the data relating to European populations, which go back to the nineteenth century.

67 The legitimacy of the use of the model tables in reconstructing pre-industrial populations has been questioned, precisely on the grounds of the possible differences between the age-specific mortality schedules assumed in the tables and the ones of pre-industrial populations: see, e.g., R. S. Schofield, in his review of Hollingsworth, op. cit. (n. 64), Historical Method Newsletter 4 (1970), 1416CrossRefGoogle Scholar.

68 It must be stressed that structure and dynamics of a population depend first of all on biological features, which do not present themselves differently in the different populations, historical and contemporary, and which are as such measurable.

69 See in particular On the probable age structure of the Roman population’, Population Studies 20 (19661967), 245–64CrossRefGoogle Scholar.

70 See now Parkin, op. cit. (n. 22), ch. 1, for the references and an analysis of the comparative value of the different types of evidence, which seems to be perhaps too pessimistic. The central role that the model tables can have in studying the demographic structure, and dynamics, of the ancient populations is revealed by the space that Parkin has devoted to a presentation of them in his ch. 2.

71 As measured, in the model tables, by the Gross Reproduction Rate, GRR, that is, the number of daughters that a women can bear during her fertile age, in absence of mortality.

72 Hansen, M. H., Demography and Democracy. The Number of Athenian Citizens in the Fourth Century B.C., (1986), 12f.Google Scholar, on E. Ruschenbusch, ZPE 35 (1979), 173–6, 177–80; 49 (1982), 267–81; 54 (1984), 253–69.

73 Rates of increase of the order of 2–3 per cent are implausibly high in a pretransitional population, even for a limited span of time and a small group. It is hard to accept the conclusion drawn by Sallares, R., The Ecology of the Ancient Greek World (1991), 90Google Scholar, that at the start of the Late Geometric II period the Athenian population could have exhibited for a few years a rate of natural increase of about 3 per cent per annum. A similar objection can be addressed towards N. V. Sekunda, ‘Athenian demography and military strength 338–322’, ABSA 87 (1992), 311–55.

74 Bevolkerungsgeschichte Italiens, 1 and 11, passim.

75 G. Delille, ‘Un problème de dèmographie historique. Hommes et femmes face à la mort’, MEFRM 86 (1974), 419–43.

76 D. Herlihy and Klapisch-Zuber, Ch., Tuscans and their Families. A Study of the Florentine Catasto of 1427 (English trans., 1985), ch. 5Google Scholar.

77 For a balanced account of the problem of sex ratio in the Roman world see Parkin, op. cit. (n. 22), 98ff.

78 See in particular Burn, A. R., ‘Hic Breve Vivitur. A study of the expectation of life in the Roman Empire’, Past & Present 4 (1953), 231Google Scholar, at 10–13. The last attempt has been made by Suder, W., A Study of the Age and Sex Structure of Population in the Western Provinces of the Roman Empire (1990)Google Scholar.

79 Bagnall, R. S. and Frier, B. W., The Demography of Roman Egypt (1994)CrossRefGoogle Scholar. I should like to thank Roger Bagnall and Bruce Frier for having allowed me to read their book in advance of publication.

80 ibid., 108 n. 49.

81 See the foreword of A. J. Coale to P. M. Visaria, The Sex Ratio of the Population of India (1971).

82 Cass. Dio LIV.16.2.

83 Italian Manpower (1971), 151, 155, 558, 561.

84 There are just two cases in which the word might mean ingenuus, but only because in the same passages there is a reference to freedmen: Cass. Dio LV.22.5 and 31.1. That εὐγενεῖς are not the ‘nobility’ is obvious, but that does not necessarily imply that the term means the free-born.

85 As is pointed out by Brunt himself: Italian Man power(1971), 558–66. That is even truer if one accepts the ingenious thesis put forward by A. Wallace-Hadrill, ‘Family and inheritance in the Augustan marriage laws’, PCPS 27 (1981), 58–80, according to which ‘Augustus aimed to encourage the family in order to stabilise the transmission of property, and consequently of status, from generation to generation’; see also Hopkins, K., Death and Renewal (1983), 97f.CrossRefGoogle Scholar On Augustus' marriage, laws L. F. Raditsa, ‘Augustus' Legislation concerning Marriage, Procreation, Love Affairs and Adultery’, in ANRW 11.13 (1980), 178ff.; Mette-Dittmann, A., Die Ehegesetze des Augustus (1991)Google Scholar.

86 See above, on the Egyptian data.

87 Italian Manpower (1971), 97, table VIIIGoogle Scholar.

88 ibid., 117, 202, 242.

89 Such a rate of decrease would imply, however, as said, an incredibly low total fertility rate.

90 If one assumes that the number of adult males was less than 30 per cent, because the citizen population was not declining, the proportion of the adult male citizens who were not in the peninsula becomes paradoxically higher. But, as shown, one cannot accept both the notion of an increasing citizen population and Beloch's explanation of the Augustan census figures.

91 Conquerors and Slaves (1978), 68–9, table 1.2.Google Scholar

92 Even this average is, however, impossibly low, in view of the general conclusions that we are allowed to draw, for the Italian cities of the first centuries of the Empire, from the epigraphic evidence of gifts for feasts or distributions: see Duncan-Jones, op. cit. (n. 25), 259–77.

93 Boserup, op. cit. (n. 36).

94 Bairoch, P., ‘Urbanization and the economy in pre industrial societies: the findings of two decades of research’, The Journal of European Economic History 18 (1989), 239–90Google Scholar, at 247.

95 D. D. Dahmann and L. T. Dacquel, Residents of Farms and Rural Areas: 1990, Current Population Reports, Series P-20: Population Characteristics No. 457 (1992).

96 Bairoch, op. cit. (n.94), 266.

97 Italian Manpower (1971), 97, table VIII, 117;and the Postscript to the reprint of 1987, 717.

98 Nicolet, C., L'inventaire du monde (1988), ch. 6Google Scholar (English trans. Space, Geography, and Politics in the Early Roman Empire (1991)).

99 On Brunt's other assumption, that infants below one year were not registered at the Augustan censuses, see Lo Cascio, op. cit. (n. 39).

100 Sallares, op. cit. (n. 73), 48.

101 See above, n. 22, on the space devoted by Parkin to the data on the size of Roman population.

102 ‘Demographic impressions of the Roman world’is the title given by Parkin to Part Three of his book, op. cit. (n. 22).

103 Italian Manpower (1971), 3Google Scholar.