THE LAND 107ten- fold (8.8–10.2:1). This is by no means a low yield, and in fact recalls the
returns cited by Cicero from Leontini.
In short, returns on seed sown seem relatively high, whether in Sicily
following Cicero, or in Etruria following Varro, or on Evans’ own fi gures, in
Campania under a biennial fallow regime. Columella is the odd man out,
and the various attempts to save him – by urging that he must have been
referring to intercultivated grain, that he deliberately omitted Etruria and
Campania, that he was furnishing not a crop yield but ‘the productive
capacity of a given property’ – only emphasize the shakiness of his testimony.
There is nothing we can do about Columella except distrust him.
The upshot is that the ancient evidence, such as it is, does not support an
argument for ‘perilously low’ average yields in wheat or other cereals in
Italy, Sicily or any other part of the empire. Even in the case of the smallholder,
it is improper to deduce or simply assume that he necessarily expected and
received a low return. A re- evaluation of the peasant economy of the Roman
period of classical antiquity is overdue, one which escapes the stifl ing effect
of the initial premise that it laboured under a chronic weakness which
guaranteed its collapse, except insofar as the farmer could supplement his
income and food resources off the estate. One might start by exploring the
implications of the fact, already appreciated by prehistoric archaeologists
and beginning to infl uence the writing of ancient history, that smallholders,
especially where settlement was dispersed and farmers lived and worked on
or near their properties, were in a position to obtain good returns from their
crops by intensive methods of production.^30
ADDENDUM
If cities were the hallmark of Roman imperial culture, the countryside continued to
generate the bulk of economic production. An educated guess is that 80–90 per cent
of the population worked in agriculture and accounted for 70–80 per cent of the
value of production. The consequence of this preponderance of rural production was
that growth, or limits to growth, in the rural sector overshadowed any developments
in urban production (Zelener 2006).
Over the past twenty- fi ve years survey archaeology has come to the fore and added
much to our knowledge of the countryside, but it has also raised challenges about how
to interpret material fi nds as evidence for property size, organization of production,
and types of labour employed. On site classifi cation and property size based on
surveys, see Barker and Lloyd (1991), Patterson (2006), de Ligt and Northwood
(2008), Attema and de Haas (2011), Mattingly (2011), Launaro (2011), Witcher
(2011, 2012), and de Ligt (2012). Rathbone (2008) is a salutary caution about the
diffi culty of identifying diagnostic material remains for impoverished peasants.
Peasants and tenancy have been the subjects of continuing research: de Neeve
(1984), Foxhall (1990), Garnsey (1998), Erdkamp (2005), Kehoe (2007). For
women’s labour, see Scheidel (1995, 1996b). Rosenstein (2004) considers peasant