188 EVI MARgARITIS
Discussion has focused on the nature of ancient agriculture, in an attempt
to evaluate the relationship between animal husbandry, the intensity of culti-
vation, agricultural management, occupation patterns and land use (Foxhall
2007 ; Osborne 1992 ; Lohmann 1992 ; Forbes 1995 ; Halstead 2002 ; Alcock
2007 ; Cloke 2012 ). Two models summarise more or less current research on
ancient agriculture and take into consideration all previous research. The first
‘traditional’ model of settlement and land use pattern relies on the evidence
deriving from the ethnographic record. According to this data, the standard
practice in arable agriculture was the alternation of cereal crops with fallow,
ploughed bare in the spring. This practice reduced the availability of lowland
summer grazing, thereby enforcing the seasonal transhumance of sheep and
goats. A fundamental division between stock husbandry and arable farming
was thus taken for granted (Semple 1932 ; Koster and Koster 1976 ). In this
extensive system of agriculture, only limited manuring of the land would be
possible, and agricultural yields are presumed to have been low due to a lack
of fertilisation (Gamble 1982 ; Amouretti 1986 :2).
Halstead ( 1981 ; 2002 ) introduced a second ‘alternative’ model, where the
predominant pattern would be small-scale, stable gardening systems with crop
rotation and regular manuring. In the agriculture of prehistoric Greece, pulses
seem to be as important as cereals, and arguably such a departure from the tra-
ditional picture for later periods would only be practicable under a small-scale
regime. The rotation of cereals and legumes in arable cultivation resulted in
the increased availability of summer fodder in the lowlands, thus evening out
seasonal imbalances and diminishing the need for transhumance. In return,
livestock consumed weeds and provided manure, thereby maintaining soil fer-
tility and crop productivity (Halstead 2002 ). These models differ essentially
in that one is based on mixed farming whereas the other assumes a coexis-
tence of arable farming and specialised pastoralism without direct interaction
(Hodkinson 1988 ).
The written sources do not support a model where agricultural estates
would have been operations focusing only on a single crop and/or product,
and both Greek and Roman writers suggest a mix of agricultural processes
(Forbes 1995 ). Foxhall ( 1990 ) and Osborne ( 1992 ) also suggest a mixed farm-
ing regime against a single crop. Furthermore, until recently discussions on the
ancient economy tended to downplay trade and market forces, emphasizing
above all the need to satisfy local subsistence requirements; current research,
however, focuses on the role of markets in antiquity in the development of the
economic and social environment of a site or of an entire region (Scheidel and
von Reden 2002 ).
Until recently very few rural buildings have been excavated in Greece. Our
understanding of the character of farmhouses situated in the Greek chora has
largely been based on the excavations of the Vari and the Dema houses in