1. This article is a version of a paper read to the Cambridge Philological Society on 22 February 1979. I am grateful to Professor P.A. Brunt, M.H. Crawford, R.P. Duncan-Jones, M.W. Frederiksen and to members present at the meeting for helpful comments, to Professor G.B.D. Jones for discussing with me the Apulian evidence, and to G. Barker, J. Lloyd and D. Webley for allowing me to see their report on the Molise survey in advance of publication. The debt I owe to earlier writing on the subject of Roman agrarian history is obvious.

2. Full documentation cannot be attempted here. Heitland, W.E. (Agricola (1921) 182Google Scholar refers once to the survival of peasant-farmers, in a gloss on Varro 1.17.2. In Toynbee, A.J., Hannibal's legacy (1965)Google Scholar the massive argumentation for the eclipse of peasant proprietors quite overshadows the qualifications introduced (251-2; ch.VI: Annex V 563-7, showing familiarity with much of the archaeological evidence which might have persuaded him to modify his conclusions). The review of Toynbee by Gabba, E., Riv. di Filol. 90 (1968) 68–75Google Scholar (reprinted in Republican Rome: The army and the allies (1976) 154–61Google Scholar) endorses Toynbee's views, at least with reference to Central and Southern Italy. No change of basic position is detectable in Gabba, 's important recent paper ‘Considerazioni sulla decadenza della piccola proprietà contadina nell' Italia centro-meridionale del II sec. A.C.’, Ktema 2 (1977) 269–84Google Scholar, espec. 271, 273-4, 283. See also the different approach but similar conclusions of Staerman, E.M., Die Blütezeit der Sklavenwirtschaft in der römischen Republik (1969)Google Scholar, e.g. 5.

3. For Heitland (n.2) the terms ‘small farmer’, ‘peasant farmer’, ‘peasant proprietor’ are synonymous. On p. 157 he correlates private tenancies with a shortage of free labour, which he appears to see as an early imperial phenomenon (cf. 161). He notes references to tenancy in pro Caecina (190) and in Varro (183), and other references in various speeches of Cicero to tenants of state land in Campania (177; 198) and Sicily (194), but concludes too conservatively that tenant farmers ‘were no exception at this time, though perhaps not a numerous class’ (195). One is left with the strong impression that Heitland is unwilling or unable to rise above the texts of Varro and the others, which, as he regularly complains, are marked by serious omissions. Thus, valuable insights (such as the blending of free and slave labour on the slave-estate, 171) are left undeveloped. On individual points his discussion has been improved upon but only relatively recently. See e.g. on private tenancies work by Brunt, P.A. (initially in ‘The Army and the Land in the Roman Revolution’, JRS 52 (1962) 68–86, at 71Google Scholar) and Finley, M.I., ‘Private farm tenancy in Italy before Diocletian’, in Finley, (ed.), Studies in Roman property (1976) 103–22CrossRefGoogle Scholar. On tenancies of ager publicus see a recent paper of Gabba (n.2 ‘Considerazione …’), building on major articles by Tibiletti, G. published in Athenaeum between 1948 and 1950, and ‘Il sviluppo del latifondo in Italia dall'epoca graccana al principio del impero’, Atti del X Congr. Int. Sc. Stor. (1955), II 237–92Google Scholar. For slave-coloni, who are well attested under the Empire but might also have to be reckoned with in the Republican period, see Staerman, E.M., Trofimova, M.K., La schiavitú nell' Italia imperiale I-III secolo (1975) 43–50Google Scholar.

4. These are best discussed by Frederiksen, M.W., ‘The Contribution of Archaeology to the Agrarian Problem in the Gracchan Period’, DdA. 4–5 (1970–1971) 330–67Google Scholar.

5. P.A. Brunt (n.3); id., Italian manpower 225 B.C. – A.D. 14 (1971); Hopkins, Keith, Conquerors and slaves (1978)Google Scholar; adumbrated by Gabba, E., in Ath. 27 (1949) 173–209Google Scholar (reprinted in Republican Rome (n.2) ch. 1, at 24).

6. Hopkins (n.5) 67-9.

7. Modern interpreters of the agronomists have similarly tended to neglect peasants and the peasant community, whether or not they have believed that the Catonian ideal of autarky was actually realized in the slave estate. No challenge to this way of thinking was provided by the traditional archaeologist, with his preference for uncovering fine public buildings on urban sites and luxury villas in suburban or rural areas.

8. Garnsey, P., ‘Non-slave Labour in the Roman World’, in Garnsey, (ed.), Non-Slave Labour in the Graeco-Roman World, PCPS suppl. 6 (1980) ch. 6Google Scholar.

9. See the summary by Potter, T.W., The changing landscape of South Etruria (1979) 1–18Google Scholar.

10. Salmon, E.T., Samnium and the Samnites (1967) 77Google Scholar; Livy 9.13.7.

11. Barker, Graeme, Lloyd, John, Webley, Derrick, ‘A Classical Landscape in Molise’, PBSR (forthcoming)Google Scholar; cf. Barker, Graeme, ‘The archaeology of Samnite settlement in Molise’, Antiquity 51 (1977) 20–24CrossRefGoogle Scholar.

12. See e.g. Kornemann, E., ‘Polis und Urbs’, Klio 5 (1905) 72CrossRefGoogle Scholar; RE xviii, 2, 2318–37Google Scholar, s.v. ‘pagus’, with bibliog.

13. Nicolet, C., Rome et la Conquête du monde Méditerranéen 264-77 av. J.C. I: les Structures de l'Italie romaine (1977) 106 cfGoogle Scholar. 109.

14. See, briefly, Brunt (n.5) 352-3 on S. Etruria.

15. Jones, G.B.D., ‘Capena and the Ager Capenas. Part II’, PBSR 31 (1963) 147Google Scholar. Comments such as those of Duncan, G., in ‘Sutri (Sutrium)’, PBSR 26 (1958) 97Google Scholar n.74 are, however, of a similar order: ‘The density of Roman sites is approximately 3 per sq. km.’

16. Bradford, J., ‘The Apulia Expedition: An Interim Report’, Antiquity 1950, 84–95Google Scholar; ‘“Buried Landscape” in Southern Italy’, Antiquity 1949, 58–72Google Scholar.

17. Bradford, J., Ancient landscapes (1957) 177Google Scholar; cf. 168-9 with Plate 39 (Padova, Cesena, etc.)

18. This charge can certainly not be levelled at Trousset, P., ‘Nouvelles observations sur la centuriation romaine à l'est d'El Jem’, Ant. Afr. 11 (1977) 175–207CrossRefGoogle Scholar. This is the most successful attempt I have seen to fit together the evidence for centuriation and habitation. See further n.78.

19. For the general picture see e.g. Dovring, F., Land and labour in Europe 1900-1950: A comparative survey of recent agrarian history ed. 2 (1960)CrossRefGoogle Scholar. On the agro-town see e.g. Blok, A., ‘South Italian Agrotowns’, Comp. Stud. Soc. Hist. 10 (1968) 121–35Google Scholar, with bibliography.

20. Silverman, S.F., Three bells of civilization: The life of an Italian hill town (1975)Google Scholar; Davis, J., Land and family in a South Italian town (1973)Google Scholar. Cf. Carlyle, M., The awakening of Southern Italy (1962)Google Scholar.

21. Silverman, S.F., ‘Agricultural organisation, social structure, and values in Italy. Amoral familism reconsidered’, Amer. Anthr. 70 (1968) 1–20CrossRefGoogle Scholar.

22. The great prosperity of the North in modern times in relation to the South and even Central Italy has no analogue in the ancient world, being largely a consequence of the Industrial Revolution, and to a lesser degree, medieval developments.

23. Hopkins, Keith, ‘Economic Growth and Towns in Classical Antiquity’, Abrams, and Wrigley, (ed.), Towns in societies (1978) 35–77Google Scholar, at 68: ‘Village populations were sometimes synoecized, perhaps forcibly into towns … Turning villages into agro-towns probably contributed little of itself to economic growth’.

24. Dovring (n.19) 23: ‘The agro-town of Medieval Europe probably has its roots in the city of antiquity.’

25. See e.g. Heitland (n.2) 51; Toynbee (n.2) 246; Scullard, H.H., The Etruscan cities and Rome (1967) 63Google Scholar; Duncan-Jones, R.P., The economy of the Roman Empire: Quantitative studies (1974) 260Google Scholar; Ward-Perkins, J.B., Landscape and history in central Italy (n.d.) 14Google Scholar; Kahane, Anne, Threipland, Leslie Murray, Ward-Perkins, John, ‘The Ager Veientanus, North and East of Rome’, PBSR 36 (1968) 70–1Google Scholar; etc.

26. J. Bradford (n.17) 176-7.

27. For this development in the Ager Veientanus, see Anne Kahane et al. (n.25) 164-79. At p. 165 they write: ‘The pattern is clear and consistent. It was no longer safe to live in the old villas and farms of the open countryside, and one by one these were abandoned in favour of the nearest easily fortifiable site.’

28. Frederiksen, M.W., ‘Changes in the Patterns of Settlement’, in Zanker, P. (ed.), Hellenismus in Mittelitalien. Abh. Akad. Wiss. Gött. Phil. – Hist. Kl., Dr. Folge 97 (1976) 341–55, at 342–3Google Scholar, referring to the last centuries of the republic and ‘well into the empire’. Cf. Gabba, E. (n.2, ‘Considerazioni’) 273Google Scholar, contrasting Etruria, Umbria and the Greek colonies with the rest of Central and Southern Italy. For a summary account of pre-Roman urbanization in Italy, see Pallottino, M., ‘La Città Etrusco-Italica come premessa alla città Romana, varietà di sostrati formativi e tendenze di sviluppo unitario’, Atti CeSDIR. 3 (1970–1971) 11–14Google Scholar. For the under-urbanized or pre-urban area, see La Regina, A., ‘Ricerche sugli insediamenti Vestini’, Mem. Acc. Lincei Cl.Sc.Mor.Ser. VIII, XIII 5 (1968)Google Scholar; ‘I territori Sabellici e Sannitici’, DdA 4–5 (1970–1971) 443–59Google Scholar; Gabba, E., ‘Urbanizzazione e rinnovamenti urbanistici nell' Italia Centro-Meridionale del I sec. A.C.’, SCO 1972, 78–112Google Scholar; Laffi, U., ‘Problemi dell'organizzazione paganico-vicana nelle aree abruzzesi e molisane’, Ath. 52 (1974) 336–9Google Scholar.

29. For N. Etruria see the summary discussions of Torelli, M., ‘Contributo dell'archaeologia alla storia sociale: 1 – l'Etruria e l'Apulia’, DdA 4–5 (1970–1971) 431–42Google Scholar; ‘La situazione in Etruria’, in Zanker, (ed.) (n.28) 97–110Google Scholar. For a recent report on the ager Cosanus see Dyson, S.L., ‘Settlement Patterns in the Ager Cosanus’, Jl. Field Arch. 5 (1978) 251–68Google Scholar.

30. H. Scullard (n.25).

31. Anne Kahane, et al. (n.25) 70-71.

32. Frederiksen, M.W. and Ward-Perkins, J.B., ‘The Ancient Road Systems of the Central and Northern Ager Faliscus’, PSBR 25 (1957) 67–208Google Scholar.

33. Brief discussion in Ward-Perkins, J.B., ‘Città e Pagus; Considerazioni sull' organizzazione primitiva della Città nell' Italia centrale’, Atti del convegno di studi sulla città etrusca e italica preromana (1970) 293–7Google Scholar.

34. Ward-Perkins (n.33) 295.

35. A. Kahane el al. (n.25) 146, 148. In Veian inscriptions of imperial date mention is made of both municipes intramurani (CIL XI 3797Google Scholar, A.D. 1) and municipes extramurani (CIL XI 3798Google Scholar = ILS 6581, Augustan).

36. On Augustan developments see, briefly, Potter (n.9) 111-15.

37. The evidence is summarized in Brunt (n.5) 351-2.

38. For the civil war period see Jones, G.B.D., ‘Civil War and Society in Southern Etruria’, in Foot, M.R.D. (ed.), War and society (1973) 277–87Google Scholar. Jones stresses the impermanence of the veteran settlements in this period.

39. Frederiksen and Ward-Perkins (n.32) 183.

40. Frederiksen, M.W., ‘Republican Capua: A social and economic Study’, PBSR 27 (1959) 80–130Google Scholar remains basic, but does not concern itself except in passing with the ager Campanus. Johannowsky, W., ‘La situazione in Campania’, DdA 4–5 (1970–1971) 267–89Google Scholar is summary and impressionistic. Of earlier works, Heurgon, J., Recherches sur l'histoire, la réligion et la civilisation de Capoue préromaine (2nd. ed. 1970)Google Scholar, is not much concerned with agrarian history.–

41. See the doubts expressed by M.W. Frederiksen (n.28) 351.

42. See Frederiksen (n.28) 342 n.7; add Strabo 5.4.2 (C 250).

43. Frederiksen (n.40) 123.

44. In the previous sentence Cicero might be paraphrasing a s.c.: ‘Itaque hoc perscriptum in monumentis veteribus reperietis, ut esset urbs, quae res eas quibus ager Campanus coleretur, suppeditare posset, ut esset locus comportandis condendisque fructibus, ut aratores cultu agrorum defessi urbis domiciliis uterentur, idcirco illa aedificia non esse deleta’ (2.88).

45. Holmes, T. Rice, The Roman Republic and the founder of the Empire I (1923) 249 n.2Google Scholar.

46. Frederiksen (n.40) at 122.

47. The best known literary reference is Cic. de lege agr. 2.84 (63 B.C.). These small farmers can be assumed to have been tenants and labourers on both ager privatus and ager publicus. See Brunt (n.5) 312-19 for the ager Campanus in the late Republic.

48. Frederiksen, , ‘The Contribution of Archaeology to the Agrarian Problem in the Gracchan Period’, DdA 4–5 (1970–1971) 330–67, at 351–5Google Scholar.

49. Jashemski, W., ‘The discovery of a large vineyard at Pompeii’, AJA 1973, 27–42Google Scholar.

50. Bradford, J., ‘The Ancient City of Arpi in Apulia’, Antiquity 1957 167–9Google Scholar. For Apulia in the fourth and third centuries Torelli writes of ‘centri periferici’ in which ‘vediamo emergere gruppi di tipo intermedio, forse piccoli proprietari’. He finds some grounds for comparing these communities with the vici of the tribal areas of Central Italy, and continues: ‘e infatti, alcuni di questi centri minori, anche se presentano dimensioni e tratti monumentali, paragonabili alle città, restano pur sempre strutturalmente una pura e semplice aggregazione di case’. See ‘Contributo dell' archaeologia alla storia sociale: l'Etruria e l'Apulia’, DdA 4–5 (1970–1971) 431–42, at 439–40Google Scholar.

For Tarentum, see e.g. Stazio, A., ‘La documentazione archaeologica in Puglia’, La città e il suo territorio – Atti d. 7 Conv.d.Studi sulla Magna Graecia (1968) 267–79Google Scholar.

51. Sartori, F., ‘Città e Amministrazione locale in Italia Meridionale: Magna Graecia’, Atti CeSDIR 3 (1970–1971) 43–60, at 47Google Scholar. The parallel with the Black Sea region is brought up in the discussion of that paper by both D. Condurachi and A. Wasowicz. See for more detail Wasowicz, A., ‘La campagne et les villes du littoral septentrionale du Pont Euxin’, Dacia 13 (1969) 73–100Google Scholar; ‘Le problème du rapport et l'aménagement du territoire et du plan de cité’, in La città e il territorio (n.50) 195–202Google Scholar.

52. The extent and the character of decline in the South has been much discussed and disputed. See e.g. Kahrstedt, U., Die wirtschaftliche Lage Grossgriechenlands in der Kaiserzeit, Historia Einzelschr. 4 (1960)Google Scholar; Brunt (n.5) 353-75; Small, A.M., in Small, (ed.), Monte Irsi, Southern Italy, BAR Suppl. Series 20 (1977) 97–101Google Scholar.

53. In writing this section I have benefited greatly from information provided by Professor Barri Jones. Published work on the Tavoliere includes Bradford (n.16) and Smith, Catherine Delano, ‘Ancient Landscapes of the Tavoliere, Apulia’, Trans. Instit. Brit. Geog. 40–41 (1966) 203–8Google Scholar.

On the practical side of colony-foundation the gromatici are singularly uninformative. ‘As far as I know, no author tells us to what extent it was either expected or customary for colonists to reside on the land allotted to them. Hyginus Gromaticus talks about escorting them to the lands they had received by lot’ (Professor O.A.W. Dilke, personal communication). See in general Hinrichs, F.T., Die Geschichte der gromatischen Institutionen (1974)Google Scholar.

54. This system was laid out ‘per decumanos solos’, a practice applied in other early Latin colonies, e.g. Cales (334), Alba Fucens (303), Cosa (273).

55. Frederiksen (n.28) 344; Toynbee (n.2) II 563-5.

56. Davis (n.20) 99 and App. IV; Duncan-Jones (n.25) 327-33 (Columella).

57. Despite the possible implications of App. BC 1.18 it is highly improbable that many of the (? Gracchan) colonists in Apulia moved into fully equipped and functioning farms.

58. Information from Michael Crawford.

59. Date conjectured from pottery finds: see refs. in n.55; AE 1973 322 (cippus). See now Pani, M., ‘Su un nuovo Cippo Graccano Dauno’, Rend. Ist. Lomb. 111 (1977) 389–400Google Scholar.

60. I have in mind, for example, the colonies established on Campanian and in 194 B.C. (Livy 34.45), and such redistribution of land as officials of the Roman state were able to achieve in 173 B.C. and 165 B.C. See Toynbee (n.2) II 232-3.

61. See Taylor, L.R., The voting districts of the Roman Republic (1960) 35–100Google Scholar. Taylor regards the institution of new tribes in rural areas – made up of viritane allotments without walled towns – as the first solution to conquered land (47-8).

62. La Regina, A., ‘Contributo dell' archeologia alla storia sociale: Territori Sabellici e Sannitici’, DdA 4–5 (1970–1971) 443–59, at 451–2Google Scholar.

63. Inscription published by La Regina (n.62) 452.

64. Tozzi, P., Storia Padana antica: il territorio fra Adda e Mincio (1972) 16–17Google Scholar.

65. P. Tozzi (n.64) 18-21.

66. See e.g. Brunt (n.5) 56.

67. Davis, J., People of the Mediterranean: An essay in comparative social anthropology (1977) at 45Google Scholar states that the relation between settlement and land system is underresearched, and indicates that intensive farming is not invariably correlated with rural residence. In other words, there are always exceptions. Similarly, M. Carlyle (n.20) 40 writes: ‘All over the South he (sc. the peasant) lives in the agricultural towns and goes out from them to cultivate his land, which may be anything from 5 to 10 miles away’. But the various observers do not furnish identical figures; and mules, where possessed, presumably make a difference.

68. Tozzi (n.64) 30-33. Some evidence for vici in North Italy is collected by Ruggini, L., Economia e società nell' “Italia Annonaria” (1961) 527–30Google Scholar.

69. For N. Africa, see Lézine, A., ‘Sur la population des villes africaines’, Ant. Afr. 3 (1969) 69–82CrossRefGoogle Scholar, who arrives at figures of less than 150 people per hectare both for Carthage and for lesser towns with smaller areas of streets and monuments. See also Duncan-Jones (n.25) 259-77.

70. The most obvious and radical form of deviation was the abandonment of a colony. See e.g. Livy 39.23 (Buxentum and Sipontum). It was difficult to get volunteers for some colonies. See Livy 9.26.4; 10.21.10. Many other colonies needed an additional draft of settlers.

71. For Pompeii, see p.10; for the Molise, see Barker, Lloyd and Webley (n.11).

72. Sed populi proventus erat, nonisque diebus

Venalis umero fasces portabat in urbem:

Inde domum cervice levis, gravis aere redibat,

Vix unquam urbani comitatus merce macelli (ll. 78-81).

73. App., BC 1. 13–14Google Scholar shows that Tiberius Gracchus had crucial support among the rural population. The archaeological evidence is presented by the various surveys conducted in the area North of Rome (see nn.15, 25, 32). The promised report on the Grottarossa allotments, beside the Via Flaminia between Rome and the Prima Porta (A. Kahane et al. (n.25) 148) might provide information relevant to the point in question; it is apparently seen as an aspect of the establishment of open farming in Roman ager Veientanus. Of the settlements in the area of Sutrium, much further from Rome (i.e. about 50 km.), Duncan (n.15) 97 writes: ‘There are only two of the larger sites which could have been the centres of estates of any size … The number of sites to the square kilometre was very close to the number of buildings that there are today … and it is quite likely that the type of farming practised was also similar in nature. This consists, in the main, of groups of small vineyards, alternating with fields of corn and plots of vegetables with occasional flocks of goats or sheep and a few cattle’.

74. Blok (n.19) 46 adds another motive for maintaining urban rather than rural domicile: he observes that in the modern context the scattering of holdings, the variety of supplementary jobs, the instability of contracts all make it sensible for peasants to live in a city and travel to work on the land. The relevance of this observation to the ancient world is to me dubious.

75. Directly attested at Aquileia, Livy 40.34.2 (181 B.C.).

76. Brunt (n.5) 191. Bononia measured about 60 hectares within the walls and received 3,000 colonists with families, perhaps 12,000 people in all. A density of 200 people per hectare is not impossible, though high by N. African standards. See Lézine (n.69).

77. Brunt (n.5) 297, with reference to the first century B.C.

78. In a different context Trousset has analysed a somewhat similar phenomenon. In the area east of El Jem he identified two forms of ‘habitat rural antique’. The first, Republican in date and reviving at the end of antiquity, ‘est représentée par des petites bourgades’, the second is imperial and dispersed ‘soit en nébuleuse autour des sites précédents, soit sous la forme d'un habitat intercalaire à espacement assez constant pour ponctuer régulièrement le damier des centuries. Ces établissements sont fréquemment situés au centre ou à l'angle d'une centurie, les plus importants à l'intersection des diverses maitresses’. See Trousset (n.18) 203.

79. e.g. Betriacum (Calvatone), Acerrae (Pizzighettone). One must assume a measure of assimilation of Gallic with immigrant families in both country and town. For the Gallic character of the population of Placentia, see Cic., in Pis. 53Google Scholar. The toponomastic survey, accompanied by an interesting map of Celtic place-names, is by Bernardi, A., ‘I Celti nel Veneto’, L'Italia settentrionale nell'età antica; Convegno in memoria di Plinio Fraccaro. Ath. Fasc. spec. (1976) 71–82Google Scholar.

80. Gabba (n.28); Laffi (n.28).

81. Gabba (n.28).

82. The thesis should not be pressed too hard. To some extent there was a natural development of city centres, based among other factors on emulation of Rome. In La Regina, A., ‘Note sulla formazione dei centri urbani in area Sabellica’, Studi sulla città antica (1970) 197–207Google Scholar, it is stressed that municipal status was normally awarded to communities where ‘la condizione urbana’ had already manifested itself.

83. Barker, Lloyd, Webley (n.11); A. Kahane et al. (n.25).

84. Col. 1.7.3. The colonus urbanus is characterized as one ‘qui per familiam mavult agrum quam per se colere’. Columella apparently does not visualize a case where a tenant-farmer does not utilize slaves (familia) and is nevertheless urbanus.

85. CIL IX 2689Google Scholar, utilized by MacMullen, R., ‘Market-Days in the Roman Empire’, Phoenix 24 (1970) 333–41, at 338–9 (with Figure 1)CrossRefGoogle Scholar. I do not know why he assigns the inscription to Pompeii.

86. It is nevertheless perhaps an extravagance to speak of the ‘spiritual urbanization’ of the peasantry of the Middle Ages. The remark is cited by White, Lynn, Medieval technology and social change (1962) 67Google Scholar.