¹ Garnsey, P., ‘Introduction’, in P., Garnsey (ed.), Non-slave Labour in the Greco-Roman World (Cambridge, 1980), 3.Google Scholar Cf. Garnsey, ‘Non-slave labour in the Roman world’, ibid., 37; J. E. Skydsgaard, ‘Non-slave labour in rural Italy during the late Republic’, ibid., 67; Dyson, S. L., Community and Society in Roman Italy (Baltimore, 1992), 134–5.Google Scholar

² Already recognized by F. Engels: ‘Competition enables the capitalists to substract from the price of labour what the family produces in its own garden and small plots.’ Quoted from Medick, H., ‘The proto-industrial family economy. The structural function of household and family during the transition from peasant society to industrial capitalism’, Social History 2 (1976), 299.Google ScholarKriedte, P., Spätfeudalismus und Handelskapital. Grundlinien der europäischen Wirtschaftsgeschichte vom 16. bis zum Ausgang des 18. Jahrhunderts (Göttingen, 1980), 96Google Scholar: ‘Das zentrale Moment für das Funktionieren des proto-industriellen Systems war die Externalisierung der Arbeitskosten. Das Handelskapital wälzte sie zu groβen Teilen auf den Agrarsektor ab, indem es nur einen Teil der Reproduktionskosten der Arbeit übernahm.’ The concept did not apply to the growing numbers of rural proletariat, who had an increasing role in rural industry.

³ Garnsey (n. 1), 4.

⁴ Chayanov, A. V., The Theory of Peasant Economy, D., Thorner et al. (edd.), (Manchester, 1996), 189Google Scholar; Spurr, M. S., Arable Cultivation in Roman Italy, c. 200 B.C.-c. A.D. 100 (London, 1986), 67.Google Scholar Cf. Halstead, P. and Jones, G., ‘Agrarian ecology in the Greek islands. Time stress, scale and risk’, JHS 109 (1989), 47CrossRefGoogle Scholar, regarding the time stress of small farmers on the Aegean islands of Karpathos and Amorgos. They point out the dangers of sudden storms, theft, birds, or raiding by livestock (49–50), and one may add the danger of fire.

⁵ Similarly, Columella 3.12.10 advises the cultivation of different kinds of vine, in order to spread the labour demand at vintage time. Cf. Spurr (n. 4), 139, referring to estates: ‘Perhaps only on those estates (rare in my view) which grew only one type of cereal, and where the topography was even throughout, would free labour have been hired on any scale.’ On crop diversification in ancient Greek agriculture see Gallant, T. W., Risk and Survival in Ancient Greece. Reconstructing the Rural Domestic Economy (Cambridge, 1991), 36–7.Google Scholar

⁶ Cf. Garnsey, P., Famine and Food Supply in the Graeco-Roman World (Cambridge, 1988), 49CrossRefGoogle Scholar; Halstead and Jones (n. 4), 50–1; Gallant (n. 5), 42ff.

⁷ E.g. Columella 2.12.9.1 will discuss an alternative option later.

⁸ Also Garnsey (n. 1), 37–8; Rathbone, D. W., ‘The development of agriculture in the ager Cosanus during the Roman republic. Problems of evidence and interpretation’, JRS 71 (1981), 15.Google Scholar Cf. Knotter, A., ‘Problems of the family economy. Peasant economy, domestic production and labour markets in pre-industrial Europe’, Economic and Social History in the Netherlands 6 (1994), 35ff.Google Scholar, regarding the ‘labour cycle’ in early modern Europe.

⁹ Cf. Chayanov (n. 4), 91ff.

¹⁰ For Garnsey (n. 1), 34, this is the defining characteristic of the smallholder. Pliny, N.H. 18.38 is interesting here. In the interpretation of Scheidel, W., ‘Coloni und Pächter in den römischen literarischen Quellen vom 2. Jh. v. Chr. bis zur Severerzeit. Eine kritische Betrachtung. Colonus Studien 1’, Athenaeum 80 (1992), 354–5Google Scholar, the passage is translated thus: ‘Die Land- wirtschaft gut zu betreiben ist notwendig, aber verderblich sie bestmöglich zu betreiben, auβer wenn der Landwirt mit seinen Nachkommen oder jenen, die er ohnedies ernähren muβ, wirtschaftet.’ Scheidel rightly says: ‘Die Stelle nimmt somit auf die Subsistenzwirtschaft von Bauernfamilien bezug, fur die intensivste Bewirtschaftung deshalb möglich sei, da der “labour input” mit keinerlei Kosten verbunden ist.’ In other words, reproductive costs were borne by the farm anyhow. However, the question is also whether there were alternative options of employment for this labour.

¹¹ Chayanov (n. 4), 56ff. His hypotheses have been rightly criticized for their rigidity, but in principle the cyclical development of the consumer-worker ratio remains a valuable concept. Cf. Medick (n. 2), 298–9. Chayanov's ideas are applied to ancient Greece by Gallant (n. 5), 60ff. A shortcoming of his analysis of the ratio between the cyclically changing number of workers and the labour requirement. however, is that he uses labour intensity as a rigid variable, which it is not.

¹² Skydsgaard (n. 1), 70; Dyson (n. 1), 187. Regarding ancient Greece, see Gallant (n. 5), 133ff.

¹³ Sailer, R. P. and Kertzer, D. I., ‘Historical and anthropological perspectives on Italian family life’, in Kertzer, D. I. and Sailer, R. P. (edd.), The Family in Italy from Antiquity to the Present (S.I., 1991), 9–10Google Scholar, point out that values of honour and shame precluded an important role for such service. Cf. Evans, J. K., War, Women and Children in Ancient Rome (London, 1991), 117–18Google Scholar; Reher, D. S., Cambridge Studies in Population, Economy and Society in Past Time, 12 Town and Country in Pre-industrial Spain. Cuenca 1550–1870, (Cambridge, 1990), 201ff.Google Scholar; Barbagli, M. and Kertzer, D., ‘An introduction to the history of Italian family life’, Journal of Family History 15 (1990), 381.CrossRefGoogle Scholar

¹⁴ Cf. Knotter's criticism (n. 8), 20–1, of the concept of the ‘family economy’ in early modern Europe as used, for instance, by Medick (n. 2). Knotter: ‘People appear to allocate and coordinate their labour within the family in much more varied ways than originally assumed.’ It has to be stressed, however, that Knotter has the pro-industrial family rather than the peasant family in mind. The fewer the alternative economic strategies available (e.g. as a result of a hardly developed wage economy), the more the peasant family has to operate as a production unit. The concept of the family economy goes back to Chayanov. Central to Chayanov's theories was the peasant family, which was firmly attached to its basic source of livelihood in working the land and for whom external labour was at most secondary. Central to Knotter's analysis, however, seem to have been the marginal, rural groups, who had little access to land and for whom economic strategies apart from subsistence farming had become of primary importance (‘sub-peasants’). Nevertheless, Knotter's emphasis on social differentation among the peasantry provides important refinement of the family economy concept. Cf. Knotter (ibid.), 22ff., who also refers to Mitterauer's distinction between the peasant and sub-peasant strata. Medick (ibid.), 295, had already emphasized the importance of social context when stressing the differences in function between the extended household among the rural proletariat and among the peasantry.

¹⁵ Cf. the joint property of separate households in late medieval Macedonia. Laiou, A., Peasant Society in the Late Byzantine Empire. A Social and Demographic Study (Princeton, 1977), 73ff.Google Scholar, emphasizes that, while household and family changed cyclically in accordance with the succession of the generations, they retained joint property and continued economic cooperation. According to early medieval law in Italy, where partible inheritance predominated, it was common for heirs to hold property jointly and undivided. Ring, R. R., ‘Early medieval peasant households in Central Italy’, Journal of Family History 4 (1979), 16.CrossRefGoogle Scholar

¹⁶ Thus Barbagli and Kertzer (n. 13), 375, regarding nineteenth- and twentieth-century Italy. The same considerations apply in the Roman world. Cf. D. I. Kertzer and Brettell, C., ‘Advances in Italian and Iberian family history’, Journal of Family History 12 (1987), 92 and 113Google Scholar: ‘Both southern Italy and southern Portugal were characterized by agrotowns, and these are the areas where large, complex family households were, in general, least frequent.’

¹⁷ On household patterns, see Sailer, R. P., ‘Familia, domus, and the Roman conception of the family’, Phoenix 38 (1984), 336–55CrossRefGoogle Scholar; Sailer, R. P. and Shaw, B., ‘Tombstones and Roman family relations in the principate. Civilians, soldiers and slaves’, JRS 74 (1984), 124–56.Google Scholar Cf. Gallant (n. 5), 12ff.; and Garnsey, P. and Sailer, R. P., The Roman Empire. Economy, Society and Culture (London, 1987), 126ff.Google Scholar, who postulate the predominance of the nuclear family in the ancient world. The methods of Sailer and Shaw are rightly critized by Martin, D. B., ‘The construction of the ancient family. Methodological considerations’, JRS 86 (1996), 40–60.Google Scholar Saller and Shaw had analysed relationships in funerary inscriptions between the dead and the ones commemorating them. Parent-child relationships counted as nuclear. One of Martin's main arguments is that ‘their study demonstrates only that most people depended on members of their immediate family for commemoration; it does not demonstrate, and should not be taken to imply, that other “non-nuclear” relationships were absent’ (45). Furthermore, ‘their procedure is methodologically biased to emphasize the nuclear family and de-emphasize the extended family from the outset’(47).

¹⁸ Whether daughters had strong rights or an equal part to their brothers is unclear. In general, see R. P. Sailer, ‘Roman heirship strategies in principle and in practice’, in Kertzer and Sailer (n. 13), 26–47.

¹⁹ Cf. Laiou (n. 15), 197 regarding late-medieval Macedonia, where partible inheritance predominated. ‘The low rate of reproduction of this population… would preclude a high incidence of fragmentation of holdings…. The lands of families that died out could flow to other families.’ However, peasant holdings increasingly came in the hands of wealthy landowners, in particular the many monasteries (208). One may also point to the relationship between partible inheritance and the lack of alternative economic strategies among the share-croppers in early modern Tuscany and the high proportion of extended households. Cf. McArdle, F., Altopascio. A Study in Tuscan Rural Society, 1587–1784 (Cambridge, 1978), 137–8CrossRefGoogle Scholar; Ring (n. 15), 19; and Sallares, R., The Ecology of the Ancient Greek World (London, 1991), 208.Google Scholar

²⁰ Thus Evans, J. K., ‘Plebs rustica. The peasantry of classical Italy’, AJAH 5 (1980), 19ff.Google Scholar; Garnsey (n. 1), 35–6; Skydsgaard (n. 1), 68; Dyson (n. 1), 44; Bergqvist, S., ‘Considerations on yields, the distribution of crops and the size of estates. Three Roman agricultural units’, Classica et Mediaevalia 43 (1993), 112–13.Google Scholar The continued existence of small and medium-sized farms in many regions of Italy has been confirmed by archaeological surveys. For example: Wightman, E. M., ‘The lower Liri valley. Problems, trends and pecularities’, in Barker, G. W. W. and Hodges, R. (edd.), Papers in Italian Archaeology II. Archaeology and Italian Society (Oxford, 1981), 278Google Scholar; Dyson, S. L., ‘The villa of Buccino and the consumer model of Roman rural development’, in C., Malone and S., Stoddart (edd.), Papers in Italian Archaeology IV. Classical and Medieval Archaeology (Oxford, 1985), 76Google Scholar; Ward-Perkins, B. et al., ‘Luni and the Ager Lunensis. The rise and fall of a Roman town and its territory’, PBSR 54 (1986), 106ff.Google Scholar; Coccia, S. and Mattingly, D. (edd.), ‘Settlement history, environment and human exploitation of an intermontane basin in the central Apennines. The Rieti survey 1988–1991, part I’, PBSR 60 (1992), 271–2Google Scholar; Curti, E., Dench, E., and Patterson, J. R., ‘The archaeology of central and southern Roman Italy. Recent trends and approaches’, JRS 86 (1996), 175 and 177Google Scholar (cf. pp. 186f). On the emergence of large estates, see Barker, G., ‘The archaeology of the Italian shepherd’, PCPhS 35 (1989), 13Google Scholar; Arthur, P., Romans in Northern Campania (London, 1991), 66 and 100–1Google Scholar; Dyson (n. 1), 31.

²¹ Also Foxhall, L., ‘The dependent tenant. Land leasing and labour in Italy and Greece’, JRS 80 (1990), 104.Google Scholar Scheidel (n. 10), passim, has shown that, despite many assumptions often made, the literary sources shed little light on the social position of tenants or on the nature of tenancy contracts in the Roman world. The seeming predominance of wealthy tenants renting estates might reflect social bias, while small-scale tenants remain largely invisible. In any case, Scheidel points out, the data do not point to increasing small-scale tenancy before Severan times. Cf. Scheidel, W., Grundpacht und Lohnarbeit in der Landwirtschaft des römischen Italien (Frankfurt am Main, 1994).Google Scholar According to De Neeve, P. W., Colonus. Private Farm-tenancy in Roman Italy during the Republic and the Early Principate (Amsterdam, 1984)Google Scholar, tenancy developed mainly from the first century B.C. onwards. Share-cropping only occurs in the sources for the principate (15ff.). Cf. Dyson (n. 1), 79ff. and 132ff.

²² Cf. Garnsey (n. 1), 38.

²³ Cf. Foxhall (n. 21), 107–8, who points out that the landowner may have provided the use of oxen.

²⁴ Thus, regarding share-cropping in early modern Tuscany, see Kertzer, D. I., ‘The joint family household revisited. Demographic constraints and household complexity in the European past’, Journal of Family History 14 (1989), 4.CrossRefGoogle Scholar As a result, multiple households predominated among share-croppers. In one nineteenth century community in Tuscany, three-quarters of the sharecroppers lived in households containing two or more simple family units. Cf. Caiati, V., ‘The peasant household under Tuscan Mezzadria. A socioeconomic analysis of some Sienese Mezzadri households, 1591–1640’, Journal of Family History 9 (1984), 113–14.CrossRefGoogle Scholar As Kertzer (ibid.), 10, observes, similar considerations applied to the Russian serfs. Cf. Kertzer and Brettell (n. 16), 93. Soil productivity, therefore, could be higher than on plots cultivated by wage labourers or slaves, as remarked by Foxhall (n. 21), 102. She also points out that small plots provided a supply of wage labour for landowners.

²⁵ Cf. the differences in household structure between the short-term share-croppers (mezzadri) and the tenants who had contracts over three or four generations (livellarl). McArdle (n. 19), 177–8.

²⁶ Cf. Caiati (n. 24), 122–3, regarding early modern Tuscany: ‘In general, however, neither the familial relations of the mezzadri nor the frequent reordering of leaseholds succeeded in eliminating the tendency for an imbalance to develop between production and consumption on leaseholds.’ This imbalance led to an increasing frequency of demands for loans and assistance in the form of corn from the landowner (125).

²⁷ Cf. Chayanov (n. 4), 106fF. Evans (n. 20), 137, also concluded that ‘many of the peninsula’s smallholders were free to seek supplemental employment, even at a considerable distance from their homes’. He assumed that many would spend part of this time hunting or food-gathering. See also De Neeve, P. W., De boeren bedreigd (Amsterdam, 1993), 25.Google Scholar

²⁸ Halstead, P., ‘Traditional and ancient rural economy in mediterranean Europe: plus ça change?’, JHS 107 (1987), 81ff.CrossRefGoogle Scholar

²⁹ On the marketing of fruits, flowers, and similar items, see Frayn, J. M., Markets and Fairs in Roman Italy. Their Social and Economic Importance from the Second Century B.C. to the Third Century A.D. (Oxford, 1993), 76.Google Scholar She emphasizes the consideration that the most perishable goods would come from the immediate neighbourhood of the town. Cf. Chayanov (n. 4), 113ff.

³⁰ Gregory of Nazianzus, Or. 43.34, quoted from Evans, J. K., ‘Wheat production and its social consequences in the Roman world’, CQ 31 (1981), 429.CrossRefGoogle Scholar See e.g. Frayn's analysis (n. 29), 91ff., of commerce in mountainous regions in Italy. Interesting also is De Neeve's application (n. 27), lOff., of Von Thünen's location-principles in central Italy. The redistributive economy of ancient Egypt, where the circulation of corn was controlled by the central authorities, may also have offered more stability and therefore more economic differentiation than most other provinces. The seeming importance of flax and the frequent mention of village weavers in Egypt does not necessarily reflect the situation in the rest of the Roman empire. Cf. Jones, A. H. M., The Roman Economy (Oxford, 1974), 355ff.Google Scholar

³¹ Xenophon, Oecon. 5.17. The situation was different for wealthy producers, who could fall back on financial reserves. Chayanov (n. 4), 293, however, points out that, due to the labour intensity of cash crops such as flax, these were less attractive to commercial agriculture than to peasant fanners.

³² Also, Halstead (n. 28), 86; Jongman, W. and Dekker, R., ‘Public intervention in the food supply in pre-industrial Europe’, in P., Halstead and J., O'shea (edd.), Bad Year Economics. Cultural Responses to Risk and Uncertainty (Cambridge, 1989), 116Google Scholar; De Ligt, L., Fairs and Markets in the Roman Empire. Economic and Social Aspects of Periodic Trade in a Pre-industrial Society (Amsterdam, 1993), 130.Google Scholar Cf. Chayanov (n. 4), 121ff., on the degree of market orientation of the peasant agriculture. An important criticism to be made of his hypotheses is that they fail to consider the element of minimalization of risk.

³³ Tchernia, A., ‘Italian wine in Gaul at the end of the Republic’, in P., Garnsey et al. (edd.), Trade in the Ancient Economy (London, 1983), 87–104Google Scholar; Middleton, P, ‘The Roman army and long distance trade’, in P., Garnsey and Whittaker, C. R. (edd.), Trade and Famine in Classical Antiquity (Cambridge, 1983), 75–83.Google Scholar This is not to say that Burford's, A. article, ‘Heavy transport in classical antiquity’, Economic History Review 13 (1960), 1–18CrossRefGoogle Scholar, did not have an important point to make in pointing out that ancient transport technology did have the capacity to move heavy goods. Cf. Snodgrass, A. M., ‘Heavy freight in Archaic Greece’, in P., Garnsey et al. (edd.), Trade in the Ancient Economy (London, 1983)Google Scholar; Osborne, R. G., Classical Landscape with Figures. The Ancient Greek City and its Countryside (London, 1987), 84 and 91.Google Scholar

³⁴ On transportation in antiquity: Westermann, W. L., ‘On inland transportation and communication in antiquity’, Political Science Quarterly 43 (1928), 364–87.CrossRefGoogle ScholarYeo, C. A., ‘Land and sea transportation in Imperial Italy’, TAPA 77 (1946), 224Google Scholar: The cost of transporting corn 100 miles by ox-team ‘was equivalent to about 62% of the Roman price’. ‘It is not to be wondered at that the large-scale production of wheat was unprofitable and that Italy was unable to support herself in cereals.’ Yeo derived the price of transporting corn from a passage in Cato on the transportation of an oil-press. The arguments that will be given anon will hopefully explain why the transportation of an oil-press is not a fair comparison to the distribution of corn. Yeo's figures are criticized by Laurence, R., ‘Land transport in Roman Italy. Costs, practice and the economy’, in H., Parkins and C., Smith (edd.), Trade, Traders and the Ancient City (London, 1998), 130ff.Google ScholarJones, A. H. M., The Later Roman Empire, 284–602. A Social, Economic and Administrative Survey (Oxford, 1964), 821ff.Google Scholar, was influential. Consequently, Landels, J. G., Engineering in the Ancient World (London, 1978), 170Google Scholar, starts his chapter on land transport with remarks on its unimportance. Similarly, Rickman, G., The Corn Supply of Ancient Rome (Oxford, 1980), 13Google Scholar: movement by land ‘was both extremely slow and intolerably expensive’. See also White, K. D., Greek and Roman Technology (London, 1984), 131–2Google Scholar; Schneider, H. C., ‘Die Bedeutung der StraBen für den Handel’, Münstersche Beiträge zur antiken Handelsgeschichte 1 (1982), 85–95.Google ScholarSippel, D. V., ‘Some observations on the means and cost of the transport of bulk commodities in the late Republic and early Empire’, Ancient World 16 (1987), 35–45Google Scholar, argues that cities like Rome would be willing to pay for high costs of transport overland up to 200 miles, especially in winter, when the sea was closed and farmers profited from a seller's market. Recently, Dyson (n. 1), 34, according to whom transport problems limited the ability of Italian fanners to supply Roman markets; this discouraged the development of any major Italian domestic grain market. A structural approach to transport in antiquity was introduced by Hopkins, K., ‘Models, ships and staples’, in P., Garnsey and Whittaker, C. R. (edd.), Trade and Famine in Classical Antiquity (Cambridge, 1983).Google Scholar He pointed out that the high cost of land transport was only one factor among others in shaping trade and economy. See also Hopkins, K., ‘Economic growth and towns in classical antiquity’, in P., Abrams and Wrigley, E. A. (edd.), Towns and Societies (Cambridge, 1978), 43ff.Google Scholar

³⁵ In the Spain of c. 1800, 90% of transport was undertaken by pack animals. Ringrose, D. R., Transportation and Economic Stagnation in Spain, 1750–1850 (Durham, NC, 1990), xxii.Google Scholar

³⁶ Braudel, F., The Identity of France. Vol. 2. People and Production (New York, 1990), 461ff.Google Scholar

³⁷ Braudel (n. 36), 479–80: ‘Waterways were used for medium- and particularly for long-distance traffic’ and ‘should be compared to the main highways’. Braudel, F., The Wheels of Commerce (London, 1982), 350ff.Google Scholar, refers to the study of W. Sombart, who concluded that in Germany at the end of the eighteenth century five times as much goods was transported overland as by river. Another example is given by an estimate of the transportation in France undertaken in 1828:4.8 million tons transported by waterways, 41.3 million by land; the latter can be divided in 30.9 million tons local and 10.4 long-distance transportation. Also Braudel (n. 36), 464, and 488 on the volume of agricultural surplus production.

³⁸ Braudel (n. 36), 464ff. In 1836 there were in France some 34,500 km of first-class highways and 36,500 km of second-class roads. The smallest or third-class roads, ‘which were essential supply-routes for the villages, as well as carrying harvest-waggons, haywains, fertilizer, timber, stone, lime and sand’, accounted for 771,000 km. Highway building in France, Braudel (ibid.), 466ff.; in Spain, Ringrose (n. 35), 14ff. Regarding the Roman empire, Schneider (n. 34), 92–3, concludes that the economic role of the main highways was very limited. Local and small-scale selling of produce constituted the largest part of trade; this trade was conducted by peasants and farmers using pack animals for transport. ‘Straβen wären dabei in vielen Fällen nicht notwendig gewesen, Pfaden hätten wie im antiken Griechenland ausgereicht.’

³⁹ Cf. Knotter (n. 8), 35: ‘The members of the family cannot choose their jobs at random by measuring earning differentials or opportunity costs only, as they would do according to neo-classical economic theory. They have to attune their labour among themselves and to seasonal variations in labour demand in specific economic and ecological settings.’

⁴⁰ Ringrose (n. 35), 48ff. ‘The conversion of such people to specialized transporters would have robbed farming of a large portion of its scarce animal power, destroyed the cost advantages inherent in the peasants’ position as agriculturalists with periods of seasonal idleness, disrupted the subsistence mechanisms of the countryside’ (ibid. 122). Cf. Phillips, C. R., Ciudad Real, 1500–1750. Growth, Crisis, and Readjustment in the Spanish Economy (Cambridge, MA, 1979), 54.Google Scholar Some of the muleteers would make an annual trip to Madrid or a large seaport. Most of them operated, if on a medium distance, within a range of 80–120 km (ibid. 73). Examples are given by Ringrose (n. 35), 50–1 and Braudel (n. 37), 327–8.

⁴¹ On professional transporters, Ringrose (n. 35), 58ff.; Phillips (n. 40), 57–8.

⁴² Ringrose (n. 35), 5, 37ff., and 103ff. Madrid's dependence on governmental support not only involved the state's participation in the corn supply to Madrid; less than half of the corn consumed by Madrid was bought with state involvement. Part of the supply, however, had to be subsidized. Moreover, for the transportation of food and fuel Madrid depended on the professional transporters. These in turn depended on the government for vital access to common grazing and other privileges. Madrid concentrated so much oxen around the city that grazing had become a serious problem. Without government support the professional carreteers could not have existed on the scale they did exist; without the carreteers, Madrid could not have been supplied with food and fuel (until the railway came, that is).

⁴³ According to the census data, Spain counted approximately 450,000 mules and donkeys at the end of the eighteenth century, about half of which were employed in transport. Ringrose (n. 35), 90.

⁴⁴ Hopkins (n. 34 [1983]), 85. Cf. Spurr (n. 4), 145, who emphasizes that most towns in Italy lacked access to river or sea and therefore depended on road transport. Furthermore, the agricultural produce that was distributed throughout the Mediterranean by ship had to be transported to the sea first. Varro, R. R. 2.6.5 gives an example of pack animals transporting oil, wine and corn from Apulia and Calabria to the sea.

⁴⁵ Varro, R. R. 1.17.2f. Cf. Rathbone (n. 8), 12ff.; Spurr (n. 4), 139.

⁴⁶ Columella 1.4.6 refers to the hiring out of oxen, but not to advise such action. He warns that without sufficient supervision slaves may cause harm, e.g. by hiring out oxen. It is not stated for what purpose the oxen are hired out.

⁴⁷ Von Bolla-Kotek, S., Untersuchungen zur Tiermiete und Viehpacht im Altertum (München, 1969), 7ff.Google Scholar; Martin, S., ‘Servum meum mulionem conduxisti. Mules, muleteers and transportation in classical Roman law’, TAPA 120 (1990), 301–314.Google Scholar Cf. Buck, R. J., Agriculture and Agricultural Practice in Roman Law, Historia Einzelschriften, 45 (Wiesbaden, 1993), 7ff.Google Scholar, on the biases of juridical literature on agriculture.

⁴⁸ Codex Theod. 8.5.53; Codex Just. 12.50.15; Buck (n. 47), 49. The existence of professional transporters may be concluded from the collegium mulionum et asinariorum in Potentia, CIL X 143. Cf. Buck (ibid.), 48, for corporations of pack animal drivers in late Imperial law.

⁴⁹ Transportation of marble: Burford (n. 33), 16. See also Osborne (n. 33), 14; S. Isager and Skydsgaard, J. E., Ancient Greek Agriculture. An Introduction (London, 1992), 104ff.Google Scholar Hesiod's sailors: Snider, G. L., ‘Hesiod's sailing season (Works and Days 663–665)’, AJAH 3 (1978), 130Google Scholar; Wallinga, H. T., ‘Hesiod's farmer as a sailor’, in H., Sancisi-Weerdenburg (ed.), De agricultura. In memoriam Pieter Willem de Neeve (1945–1990) (Amsterdam, 1993), 1–12.Google Scholar

⁵⁰ Some authors accord transportation cost and effort an important role in establishing trade patterns and production zones around central markets. See e.g. De Neeve's application (n. 27), 5ff., of Von Thünen's principles on a regional and on a local level. The longer the distance, the higher the cost of transport and therefore the more important the factor of transportation cost against other factors. However, regarding short distances, other factors such as soil fertility and availability of natural resources played a relatively larger role. The limitation of the role of cost and effort of transportation, as argued here, assigns a more important role on the local level to other factors in the determination of economic patterns. Accordingly, De Neeve is not very convincing on the local level, i.e. in his case-study of the immediate hinterland of Veii. Cf. Frayn (n. 29), 75–6.

⁵¹ Garnsey (n. 1), 36 and 41–2. Cf. Evans (n. 20), 136; Skydsgaard (n. 1), 66ff.; Rathbone (n. 8), 12ff.; De Neeve (n. 21), 21; Scheidel, W., ‘Zur Lohnarbeit bei Columella’, Tyche 4 (1989), 140Google Scholar; Rosafio, P., ‘Slaves and coloni in the villa system’, in J., Carlsen et al. (edd.), Land-use in the Roman Empire (Rome, 1994), 147 and 152.Google Scholar Cf. Scheidel (ibid.), 144, on the employment of seasonal labourers during the vintage.

⁵² Also Scheidel (n. 51), 141, who refers to Columella 2.2.12.

⁵³ Cf. Garnsey (n. 1), 42; Skydsgaard (n. I), 69; Spurr (n. 4), 66; Dyson(n. 1), 135. Spurr points out that the passage does not necessarily refer to harvesting.

⁵⁴ Cato 1.1.3 advises potential buyers of estates to pay attention to the available labour in the area. See also Scheidel (n. 51), 139.

⁵⁵ Cf. Garnsey (n. 1), 39 and 42; Garnsey, P. and Woolf, G., ‘Patronage of the rural poor in the Roman world’, in A., Wallace-Hadrill (ed.), Patronage in Ancient Society (S.I., 1989), 159ff.Google Scholar; Foxhall (n. 21), 112. On ancient Greece, Osborne (n. 33), 94. See Phillips (n. 40) for problems in finding harvesters in early modern, inland Spain.

⁵⁶ Columella 3.21.9f. Cf. Rosafio (n. 51), 149.

⁵⁷ Millet, M., ‘Town and country. A review of some material evidence’, in D., Miles (ed.), The Romano-British Countryside, BAR 103 (Oxford, 1982), 428.Google Scholar On the location of such industries, also De Ligt, L., ‘Demand, supply, distribution. The Roman peasantry between town and countryside’, Münstersche Beiträge zur antiken Handelsgeschichte 10 (1991), 35ff.Google Scholar

⁵⁸ We have already seen the example of peasants producing and selling charcoal in early modern Spain.

⁵⁹ Jones (n. 30), 360. On early modern textile industry, e.g. Kriedte (n. 2), 46ff. and 93ff.

⁶⁰ Jones (n. 30), passim. Implements used by women for wool processing are reckoned by Columella 12.3.1ff. among the items regularly used on a farm. Furthermore, on days when the weather does not allow women to work on the land, they should make clothes for themselves, the supervisors or other slaves.

⁶¹ Cato's ideal farmer would buy as little as possible; nevertheless he would purchase items such as clothes for the slaves, mill-stones, iron tools, and ropes on the market, which indicates a demand for such rurally produced goods (Cato 22 and 135). Columella 12.3.6 advises that women make clothes for some of the slaves, in order to save money. Hopkins (n. 34 [1978]), 52ff., argues, firstly, that, despite their poverty, the large number of peasants constituted a significant market. Secondly, some social diversification of the ‘masses’ is required; the buying power and demand of the middle groups should not be underestimated. Cf. Jones (n. 30), 352; De Ligt (n. 32), 140, on the extent of a rural market for shoes, plain clothing, etc. Cf. De Ligt (ibid.), 146.

⁶² Cf. Jones (n. 30), 360; S. B. Pomeroy, Goddesses, Whores, Wives, and Slaves. Women in Classical Antiquity (New York, 1975), 199–200; Treggiari, S., ‘Lower class women in the Roman economy’, Florilegium 1 (1979), 67ff.Google Scholar

⁶³ Cf. Sallares (n. 19), 83. On ancient attitudes, Scheidel, W., ‘Feldarbeit von Frauen in der antiken Landwirtschaft’, Gymnasium 97 (1990), 424ff.Google Scholar; ‘The most silent women of Greece and Rome. Rural labour and women's life in the ancient world (II)’, Greece and Rome 43 (1996), 5ff. Most references to female day-labourers are Greek. Columella 12.3.6—women should make clothes when the weather does not allow them to work on the land—indicates that slave women regularly worked in the field. Cf. Scheidel (ibid. [1990]), 421; (ibid. [1996]), 3. Regarding ancient Greece, Osborne (n. 33), 70. While in modern southern Italy it was improper for women to work outside the house, in northern Italy women played a crucial role in the labour force: Kertzer and Brettell (n. 16), 95. Caiati (n. 24), 120, however, observes that cereal cultivation was the responsibility of adult males in early modern Tuscany. Cf. the role of women in agriculture mainly as day-labourers in early modern Languedoc, Ladurie, E. Le Roy, Die Bauern des Languedoc, Übers, v. Les paysans de Languedoc, cop. 1969 (Darmstadt, 1985), 125–6.Google Scholar On the gender-specific division of peasant-labour, also Knotter (n. 8), 34–5.

⁶⁴ Jones (n. 30), 353. Evans (n. 13), 121ff., argues that, though the clothing industry offered employment for peasant women, these were increasingly confronted with competition from slaves and freedmen. A differentiation between cheap clothing and luxury products may be required. Cf. Kriedte (n. 2), 97.

⁶⁵ Cf. Garnsey (n. 1), 4: ‘Peasant displacement in the Republic and early Empire had not proceeded far enough to make possible the creation of a stable wage labour force.’ Garnsey is, of course, right as far as full-time wage labour, which was independent of agriculture and therefore could be concentrated in large urban centres, is concerned.

⁶⁶ The supply of labour to the labour market by peasants and small farmers is not stable, which provided structural problems for the early modern European rural industry. At times of increased demand for labour, for instance due to increased demand for manufactured goods, wages rose; peasants reacted by reducing working hours. Medick (n. 2), 301; Kriedte (n. 2), 166.

⁶⁷ Finley, M. I., The Ancient Economy (London, 1985), 80–1.Google Scholar Cf. Knotter (n. 8), 39, regarding early modern Europe: ‘The behaviour of worker-peasantries demonstrates that enduring dependence on wage earning and on the forces of the labour market can exist without the creation of a working class.’