Version classiqueVersion mobile

Territoires, architecture et matériel au Levant

Gestion des territoires : agglomérations et échanges

Landscapes of production in Late Antiquity: wineries in the Jebel al-‘Arab and Limestone Massif

Andrea Zerbini

Résumé

L'Antiquité tardive est une période d'essor démographique et économique dans les campagnes du Levant, dont les traces sont particulièrement bien enregistrées dans les vestiges des villages du Massif Calcaire en Syrie du Nord et du Jabal al-'Arab dans le Hawran. La croissance de l'habitat a été accompagnée d'un développement sans précédent des infrastructures de production agricole (olive et raisin). L'article examine l'évolution des capacités de production des pressoirs à vin dans les deux régions. Ces structures sont difficiles à dater, bien que les fouilles et les prospections permettent d'établir au minimum une chronologie relative des évolutions. On a souvent noté que les techniques les plus perfectionnées en usage dans les pressoirs à olive et à raisin (mécanismes à vis et levier, et à vis directe) étaient déjà connues à l'époque de Pline l'Ancien, mais que ces techniques n'avaient eu que peu d'applications concrètes avant l'Antiquité tardive. L'article explore les implications du passage du pressoir à levier au pressoir à levier et vis, et l'évolution des complexes de pressoirs dans les deux régions. On pense généralement que des choix malthusiens ont conduit à la fin des villages. L'implantation de technologies de production plus efficaces laisse penser à un scénario différent dans lequel la diminution des revenus est contrebalancée par des améliorations technologiques, dans un modèle parfaitement Boserupien.

Texte intégral

  • 1 On the hinterland of Si’a see: Gentelle 1985; Villeneuve 1985; Dodinet, Leblanc & Vallat 1992; Eid. (...)
  • 2 The earliest excavations of Dehes were planned by Georges Tchalenko, but only realised after his de (...)
  • 3 Dr Callot’s work on the presses of Dehes remains entirely unpublished. I should like to thank him f (...)

1This paper presents a comparative assessment of wine production in two areas of Syria between the fourth and seventh centuries: the jebel al-‘Arab in the Hawran district of southern Syria and the Limestone Massif of northern Syria. Within these regions, two case studies have been selected: the territory surrounding the hilltop village of Si’a (mostly known because of the sanctuary of Baalshamin and discussed by other contributors of this round table); and the village and hinterland of Dehes in the Jebel Barisha (figs. 1-2). The former has been the object of a number of studies and, most recently, of an intensive survey conducted by the author and by C. Hatoum in 20101. On the other hand, Dehes is the only village in the region of the Limestone Massif for which published excavation reports may be combined with work done on wine and oil presses2 : in the last fifteen years, a team led by O. Callot has excavated twenty out of the twenty-nine wine and oil presses discovered in the vicinity of the village3.

Fig. 1. The Si’a Valley

Fig. 1. The Si’a Valley

Dentzer-Feydy, Dentzer, Blanc 2003: Pl.3 Figure 2

Fig. 2. The village and immediate hinterland of Dehes

Fig. 2. The village and immediate hinterland of Dehes

2In this paper I will tackle three areas: first, I will survey the evidence concerning wine making in the two case study areas; second, I will look at the capacity of the wineries in Dehes and Si’a and explain what the figures obtained may be able to tell us in regard to the use of these installations and patterns of land holding. Finally, I will conclude with a tentative interpretation of the diffusion of technological improvements in these installations, and most notably that of the screw-press in the Limestone Massif. The chronological boundaries of this paper are somewhat difficult to frame for the reasons that will be discussed below. In broad terms discussion can be said to embrace the centuries between the fourth and the eighth CE.

Wine production and its evidence: wineries in the jebel al-‘Arab and the Limestone Massif

  • 4 Amouretti & Brun 1993. Brun 2003; Id. 2004.
  • 5 The best general works on this subject are: Frankel, Avitsur & Ayalon 1994; Frankel 1999; Ayalon, F (...)
  • 6 Callot 1984
  • 7 Callot 1984: 31-70.
  • 8 See, in particular, Brun 2004: 111-9. For presses with rollers and their identification with wine p (...)
  • 9 Mainly Tosef.Ter. 7:15. Frankel 2009: 11.
  • 10 The procedure is well explained by Callot 1984:74 (section B). For a recent detailed discussion of (...)

3Ancient wine and oil presses have been often discussed as part of the wider debate on the ancient economy and ancient technology4. As far as the Eastern provinces of the Roman Empire are concerned, much work has been done by Israeli archaeologists to describe and date the wine and oil production infrastructure of ancient Palestine5. Overall, the different kinds of wine and oil-producing infrastructure shared some common features: all may be divided in at least two functional spaces, the working area and the storage area; they all included vats and receptacles for the collection of the unfinished product, which were in most cases directly hewn in the rock. While some presses were located in underground chambers, the vast majority were open-air or roofed structures. Finally, most (but not all) were provided with a pressing mechanism. Despite these common features, wine and oil presses display a high degree of local variation, something that becomes immediately apparent when comparing Dehes and Si’a. Let us begin with the former. The different types of presses of the Limestone Massif have been thoroughly described by Olivier Callot6. The vast majority was provided with lever-presses, which comprised a timber beam one end of which was anchored to a niche in a wall whilst the other end was pulled down by means of a system of ropes and counterweights (lever-and-windlass) or by a screw mechanism itself anchored to a counterweight (lever-and-screw)7. In Dehes, two types in particular may be distinguished: those equipped with two square decantation basins (P6, P22 and P28 in fig. 2) and those which comprised a wide sloping area connected to a deep rock-hewn vat shaped in the form of a bell (P2-4, P9-12, P21, P26-7). These latter were also sometimes provided with oblong stone rollers which were used as crushing mechanisms (e.g. P4). This second type of installations is by far the most frequently attested in the Limestone Massif. The similarity of its plan and components (especially the bell-shaped vats with sumped bottom) to those of Palestinian wine presses has led several scholars in recent years to regard these as wine presses8. In particular, the large receiving vats prove mostly unpractical for oil separation, but are well suited for wine production: as Rabbinic sources seem to suggest, the grape must was left into these containers for first fermentation, which normally took up to two weeks9. On the other hand, presses with double decantation basins are specifically designed for oil production as they allow oil collection by overflowing10.

  • 11 Grape remains and olive kernels were found in the basins of the large press of Sergilla pictured ab (...)
  • 12 For the W press of Deir Dehes – which combines a bell-shaped vat with a rotary crushing mill – see (...)

4If, on the whole, presses with large treading areas and bell-shaped vats (and, occasionally, stone rollers) were designed for wine production and those with square decantation basins for oil making, it is very likely that they both could be used for either product. Indeed, archaeobotanic evidence suggests that at least some presses were used for both activities11. At the same time, the occasional association of rotary crushing mills – which are only found in oil making – with presses with bell-shaped vats (as, for example, in the case of Deir Dehes) attests to the mixed use of such structures12. Thus, while presses could no doubt be used for both wine and oil production, differences in their plan that made them more suitable for one or the other are likely to suggest a prevalent function. In what follows, therefore, presses furnished with bell-shaped vats are treated as wine presses while presses with double decantation basins are regarded as oil making facilities. While these latter will be occasionally mentioned, it is on the former that we will concentrate (i.e. P2-4, P9-12, P21, P26-7 in fig. 2).

  • 13 These are, in chronological order, the famous temple oil press of Kafr Nabo (IGLS II 376, AD 224); (...)
  • 14 Sodini et al. 1980: 155-8

5Dating presses represents an often impossible problem to solve. This is mainly due to their being partly rock-hewn, a fact that makes it difficult to establish when they first came into use. Moreover, only rarely is diagnostic pottery found in association with these structures. This holds true for Dehes, where very little datable material has been found in the twenty presses excavated so far. In some rare instances, presses are dated by epigraphy, though the date may at most indicate the beginning and not the length of occupation. Only three such cases are attested for the Limestone Massif, none of which in Dehes13. However, excavation has shown that several of the presses of Dehes underwent two phases. The best known case is that of P22, an underground oil press which was cleared during the excavation of the dwelling complex 101-8 in the 1970s (fig. 3)14.

Fig. 3. The underground oil press in the courtyard of complex 101-8

Fig. 3. The underground oil press in the courtyard of complex 101-8

Sodini et al. 1980: 157 Figure 200

Fig. 4. Complex 101-8 in Dehes

Fig. 4. Complex 101-8 in Dehes

Sodini et al. 1980 : Figure 5

  • 15 Sodini et al. 1980: 153; 156.
  • 16 Dating relies primarily on the finding of decorated blocks in re-use in the walls of P27, to which (...)

6The excavators noted that this underground press had two phases, the first of which included the digging of a small press (right half of fig. 3). This must be posterior to the building of house 105, itself dated by an inscription to AD 360/1 (see fig. 4)15. The second phase entailed an extension of the press to the west (left half of fig. 3). The excavators did not attempt to date this second phase, but the likelihood that building 106 is to be dated to the 6th century coupled with the fact that, with the establishment of this building, the press could only be accessed through it make it likely that the enlargement of the press should also be dated to the 6th century. In another two cases a similar relative chronology can be established. The first case is that of winery P21, excavated in 2004; this is located beyond the NW corner of building 107 (see fig. 4) and seems to have undergone an occupation consistent with the chronology of this dwelling (from the fourth through the sixth century). Like P22, this press underwent a transformation – which Callot sets in the sixth century – and which entailed not only a change in orientation, but possibly also a change in the pressing technique, with the switch from a lever-and-windlass to lever-and-screw mechanism. Finally, further north of building 107 a double press (P26-7, excavated in 2006 and 2007) – again designed for wine production and provided with a screw mechanism – has been tentatively dated by Callot to the early-Umayyad period16. It is also possible that P26 had originally been equipped (similarly to P21) with a lever-and-windlass press later substituted by a lever-and-screw press. To sum up, the little dating evidence that we possess for the presses of Dehes seems to point to two main phases of building, one set in the 4th century, the other in the 6th and 7th centuries.

7Let us now turn to Si’a. Figure 5 shows the wineries surveyed by the author and C. Hatoum in the immediate vicinity of the hill of Si’a.

Fig. 5. Wineries in the Si’a Valley

Fig. 5. Wineries in the Si’a Valley

  

  • 17 Dentzer et al. 2003.
  • 18 For the technical specifications of these wineries see Zerbini 2010.
  • 19 The definition of the “complex” type is due to Frankel (Frankel 1999). In the surveys conducted by (...)

8Most of these structures receive some mention in Hauran II, but only Si’a 8 and Si’a 353 were described in detail17. The 2010 survey was able to recover information regarding the plan, components and vat capacity of presses K3-6 and Si’a 2118. Unfortunately, recent bulldozing has completely obliterated K7 and K1-2, which had been recorded by Mikael Kalos in the mid-1990s. The ground plans of the wineries of the jebel al-‘Arab radically differ from those of the Limestone Massif, but they closely resemble the “complex wineries” of ancient Palestine and Arabia19. In broad terms, they comprised a central paved treading floor, which sloped toward a large receiving vat dug in the rock. Moreover, the treading area included a morticed stone sunk in its centre used to anchor a direct-screw press (never encountered in the Limestone Massif). This type of press (figs. 6-7) was the most common type in Levantine wineries during the Byzantine period and is frequently depicted in mosaic pavements such as the one of the church of Lot and Procopius at Khirbet al-Mukhayyat (dated AD).

Fig. 6. Detail of the mosaic pavement of the church of Lot and Procopius at Nebo/Kh. al-Mukhayyat

Fig. 6. Detail of the mosaic pavement of the church of Lot and Procopius at Nebo/Kh. al-Mukhayyat

Dunbabin 1999: 198 Figure 211

Fig. 7. A reconstruction of a direct-screw press from Tell Qasile

Fig. 7. A reconstruction of a direct-screw press from Tell Qasile

Ayalon 2009: 202 Figure 36.6

  • 20 For some examples of complex wineries from Israel see Yevin & Finkielsztejn 2009 (H.Castra); Yanai (...)
  • 21 Pliny, NH XIV 11.85; 9.75; Athaen., Deipn. I 30b; II 45e. On this see Decker 2009: 128-9. Vats for (...)

9Surrounding the treading floor, the wineries of the “complex” type – to which those of Si’a belong – were provided with a number of additional compartments. The precise function of these compartments remains somewhat debated, though scholars agree that they were used as either unloading areas or additional treading floors20. In the wineries of the jebel al-‘Arab, such compartments were arranged on two levels, the lower of which was hewn in the rock in the shape of vats (fig. 8). These were of much smaller size than the main receiving vat toward which the treading floor sloped. It is very likely that these vats were used to collect what the sources call the protropon or prodromos must, i.e. the “self-flowing must” that was produced by the natural pressure of a load of grapes onto itself. Such must was particularly prized in Antiquity and was used to produce choice wines21.

Fig. 8. Winery Si’a 21 (looking south)

Fig. 8. Winery Si’a 21 (looking south)

The photograph shows the central treading floor (with morticed stone for the anchorage of the direct-screw press) and the additional loading compartments on three sides

  • 22 Blanc 2003: 35-6.
  • 23 Dentzer et al. 2003:139.

10The excavation of two wineries of Si’a (Si’a 8 and 353) allows us to establish a chronological framework for the building of these structures. Si’a 8, originally a shrine possibly connected with the sanctuary of Baalshamin, was converted into a press in the late Umayyad period. Diagnostic pottery found in stratigraphic context consists, mainly, of fragments of Jerash lamps and white-fabric jugs with painted decoration that may be attributed to the late-Umayyad period. In the filling of vat 5, a coin dated to AD 696/7-750/1 represents a terminus post quem for the abandonment of the installation, which may be set in the Abbasid period22. In the excavation of Si’a 353, instead, a coin of Honorius (AD 402-8) was found sealed in the mortar of the south-west corner of the treading floor. This coin was in very worn condition, suggesting a long circulation and, consequently, a possible dating in the late-5th or early-6th century23.

  • 24 For a discussion of fabrics in Si’a see: Barret et al. 1986; Orssaud, Barret & Blanc 2003.
  • 25 Although the Nabataean inscription contains a lacuna in the patronymic, but the name of the dedican (...)

11During the survey conducted in 2010, the author and C. Hatoum have gathered surface pottery in the area of wineries K6 and K3-5. This ceramic material has yielded few diagnostic sherds and generally reflects the whole gamut of fabrics attested in the area of Si’a24. Sherds of the local fabric A are particularly abundant, but fabric C, dated to the Hellenistic period, was also frequently encountered, and especially in the area of K3-5. The presence of Hellenistic and Roman pottery in the vicinity of the wineries does not mean the presses should also be given an early date. Most of them, in fact, appear to have been built on the site of earlier buildings, be these tombs or rural shrines: in K6, for example, two probably funerary inscriptions (one Greek the other Nabataean) have been found in re-use in the built parts of the winery25. The transformation of these structures into wineries should, consequently, still be placed in Late Antiquity, though excavation is needed to establish a more precise chronology.

Production capacity of wineries and its implications

12We have so far described the evidence and established a broad chronological framework in which to set our discussion. We may now turn to the second part of this paper, which attempts to quantify the production outputs of the wineries in the two case study areas in order to gain insights into the scale of economic development but also on the configuration of patterns of land holding. In what follows, I will look at the capacity of the collecting vats of the wine presses of Dehes and Si’a and extract, from this data, the catchment zones of individual presses. In other words, the following section is devoted to answering the question of how much land had to be cultivated with vineyards in order to work the wineries under examination to their full capacity. These calculations, I will argue, allow us to gain a better understanding not only of how these structures were used but also of how land was likely to have exploited.

  • 26 Frankel 1999; Brun 2003: 63-4.
  • 27 Vintage in the Levant has a duration of between four and seven weeks (usually mid-August/mid-Octobe (...)
  • 28 Since must flowed into the receiving vat through channels/openings hewn into the wall separating th (...)

13Quantification of wine production relies on measuring the size of the receiving vats which, as we have seen above, were likely used as containers where the first fermentation of the must took place26. Since the first fermentation did not usually take more than two weeks, vats were likely to be refilled two or three times over a vintage season27. Some caveats are in order when measuring bell-shaped vats. First, given their irregular geometry, their shape is usually approximated to that of a cylinder or a cone frustum (the former is used for the wineries of Si’a while the latter is employed for those of Dehes). Second, the height of the vat is not measured from its top but from the level of the treading floor: although most containers do extend above this level (either as built compartments, as in Si’a, or because the treading floor is sunken deeper in the bed-rock as in Dehes), this upper portion was no doubt meant to make up for the increase in volume that occurs to during the first fermentation (which can be reckoned at ca. 25-30% of the volume of unfermented must)28. Tables 1 and 2 contain the technical specifications of the wine presses of Dehes and Si’a respectively.

Tables 1. Wine presses of Dehes and Si’a

Press (Callot)

Press Device1

Vat (H)

Vat (R)2

Vat (r)3

Vat (V)4

Notes

P2

W ?/S

150

100

67.5

3.35

P3

-

130

220

220

6.3

Rectangular vat/cistern

P4

W/S

75

110

55

1.66

P9

S

70

85

65

1.24

height is based on top level of plaster found in excavation

P10

W/S*

93

101.5

80

2.42

P11

W/S

?

65

215

The press reuses a rock-cut tomb as the receiving vat.

P12

S

70

75

62.5

1.04

Press occupies the ground floor of a one-room house

P13

?

95

92.5

75

2.1

P26

W ?/S

80

85

70

1.51

P27

S

70

100

77.5

1.74

The walls of this press re-use blocks whose décor is dated by Callot to the 6th c., thus positing a very late date for the press

P21

W ?/S

70

100

70

1.61

Tables 2. Wine presses of Dehes and Si’a

Press

Press Device1

No. of vats (seen)

No. of vats (measured)

Total volume (m3)

K3

DS

4

3

2,55

K4

DS

5

3

8,78

K6

DS

4

2

4,67

Sia 21

DS

7

5

6,53

Sia 353

DS

6

6

14,6

Sia 8

DS

5

5

8.72

  1. W= lever-and-windlass; w= lever-and-weights; S=lever-and-screw; DS= direct screw

  2. Radius at the bottom of the vat (unless otherwise specified)

  3. Radius of the vat at the level of the treading floor (unless otherwise specified)

  4. Volume of vat (m3) All other measurements are in cm

  5. With the aid of this data, we can proceed to estimate not only the production capacity of wineries, but also the size of vineyards. Estimates of production capacity are given in Table 3 below.

Table 3. Capacity of wineries of Dehes, Si’a and comparanda1

Press

Vat/Stack (V)

Wine (1 filling)

Wine (2 fillings)

Wine (3 fillings)

P2

3.35

3350

6700

10000

P3

6.3

6300

12600

18900

P4

1.66

1700

3400

5100

P9

1.24

1250

2500

3800

P10

2.42

2400

4800

7200

P12

1.04

1000

2000

3000

P13

2.1

2100

4200

6300

P26

1.51

1500

3000

4500

P27

1.74

1700

3400

5100

P21

1.61

1600

3200

4800

K3

2.55

2500

5000

7500

K4

8.78

8800

17600

26400

K6

4.67

4700

9400

14100

Si’a 21

6.53

6500

13000

19500

Si’a 353

14.6

14600

29200

43800

Si’a 8

8.72

8700

17400

26100

Dor2

8400

W.Galilee 33

W. Galilee 4

10.6

1.28

10600

1280

Mulabbis 37S4

Mulabbis 37W.1

18.9

4.7

18900

4700

  1. All estimates of production are in kilograms. Volumes are in cubic metres.

  2. Kingsley 1999: 74

  3. Frankel 2009b: 20

  4. Gudovitch 2009:203-7

  • 29 For a plan of Behyo see Tchalenko 1953-8: Pl. 110 (Vol.2).
  • 30 Callot 1984: 125.

14The comparison between the wineries of Dehes with those of Si’a and other Levantine installations shows that the former were on average smaller, though this was partly compensated by their being often arranged in small clusters of two or three presses. The combined production capacity of some such clusters (of which that of P2-4 and that of P21 and P26-7 are the most significant) is on par with that of the biggest Levantine presses (as illustrated by Table 3). The reasons underlying the creation of “press clusters” are difficult to pin down. Amongst the plans published by Georges Tchalenko and his followers, those of Behyo and Dehes best illustrate this particular distribution29. One cluster in Behyo comprised a staggering 14 presses in less than 0.5 ha. On average, such groups included between two and four presses. The concentration of these structures in so limited a space makes it unlikely that they should be explained as a function of the “morcellement des propriétés” – as Callot held30. Against this view goes also the fact that, in the majority of cases, these clusters are located along, and in some cases even obliterate the ancient paths that connected the different villages of the massif (this is the case of the already mentioned cluster P2-4 of Dehes). This peculiar arrangement must surely be significant and may indicate shared ownership and use of these installations.

  • 31 On the hilltop village of Si’a see Villeneuve 1997.

15Press capacity is on average higher in the surveyed installations of the Si’a valley, with that of Si’a 353 being particularly significant. As we shall see below, so high capacities allowed these installations to have catchment zones that far surpassed the size of the land parcels in which they were set. In the Si’a Valley, too, the placement of presses appears to have been significant. Thus, a cluster of smaller presses (K3-5) was arranged near the ancient Roman road that reached Si’a from Suwayda – possibly configuring again a situation of joint village ownership. However, this argument cannot be easily made in the lack of a straightforward connection between areas of settlement and landscape of production. Indeed, after the apparent abandonment of the hilltop village of Si’a in the early-fourth century CE, the foci of settlement of the Si’a valley become more and more difficult to pin down31. Smaller nuclei of settlement existed in the area, amongst which Kh. al-‘Anz, Khirbet Khneyfs and Khirbet Khassin, but their chronology of occupation remains to be fully clarified. The patchy distribution of late antique findings may suggest that the Byzantine period brought about a change of settlement from nucleated sites to dispersed settlement, though this is yet to be convincingly proved.

  • 32 Geoponika, 6.11. See also Decker 2009: 122-9 for discussion and 263-71 for the value of the Geoponi (...)
  • 33 According to Goor (1966:63), in 1946 Mandate Palestine produced a total of 9,000 tons of grapes for (...)
  • 34 Data published in the official reports Études sur la Syrie économique (1953-9).

16From the capacity of production of wineries we can also obtain their areas of catchment (i.e. the size that vineyards had to attain in order to work the wineries to full capacity) by turning quantities of grape must back into quantities of grapes and adopting an appropriate grape yield/ha ratio. This can be done as follows. One hundred kilograms of trodden grapes produce ca. 55-65 litres of first must (or mosto fiore). Grape pomace (the residue of treading) was then pressed at least twice, each time producing another 5-8 litres of must. The ancient agronomic literature diverges on whether this must was added to the first must or not. If we accept the Geoponika to reflect a more “eastern” tradition of winemaking, we would be led to believe that must derived from pressing was kept aside32. In this case, the vats of the wineries of Dehes would have contained only first must, thus suggesting a must/grape ratio of 0.55-0.65. A similar ratio was still obtained in the last years of the British Mandate in Palestine33. Figures for yields of grapes/unit of surface rely more heavily on comparative literature. The earliest reliable dataset for Syria was compiled by the Syrian Ministry of Agriculture and covers the years 1955-60; for this period, the average yield was 3000 kg/ha34.

17With this data, we can now plot on a map the catchment zones of the wine presses analysed. Figure 9 displays three possible configurations for the hinterland of Dehes, which vary according to whether we assume vats to have been filled once, twice or three times in the course of a vintage season.

Fig. 9.1-3. Catchment zones of wineries of Dehes. The three scenarios correspond to more intensive uses of the installations (1, 2, 3 fillings)

Fig. 9.1-3. Catchment zones of wineries of Dehes. The three scenarios correspond to more intensive uses of the installations (1, 2, 3 fillings)
  • 35 For Dehes see especially Sodini et al. 1980. More generally for the Limestone Massif: Tate 1992.
  • 36 These figures are obtained by applying D.Mattingly’s approach to the calculation of maximum oil out (...)

18The size of individual vineyards would have varied between 0.5 ha and 10 ha, thus configuring a regime of small and middling ownership that fits well with the evidence for private architecture and use of the domestic space that characterised this region35. It is significant to note that the cluster P2-4 possesses the biggest catchment zone, a fact that may support our argument for the joint village ownership of this group. Overall, the ten presses under examination would have needed between 13 and 77 ha of land to be cultivated with vineyards in order to work to full capacity. But the wine presses analysed represent only a fraction (35%) of the overall number of presses of the hinterland of Dehes. Of the remaining 16 presses, the majority would have also been wine presses judging by the ratio of 1:3 between oil and wine presses that applies to the surveyed installations. Another 10 wine presses would have needed as much land to be cultivated with vineyards. In Antiquity, the overall territory of Dehes would have not exceeded ca. 250 ha since the village was set in a densely populated area and surrounded by other medium and large settlements such as Barisha, Bab Ayan, Bashaku and Khirbet Hassan. The calculations above suggest that as much as 60% (154 ha) of the entire hinterland of the village could have been cultivated with vineyards. When olive orchards, tilled fields, pastures, barren land and built areas are taken into consideration, this upper limit would almost configure a regime of monoculture and is thus likely to represent an exaggeration. Olive orchards, for example, are likely to have taken up another considerable portion of the village territory: the three oil presses P6, P22 and P28 required between 20 and 85 ha planted with olive trees to work to full capacity, though bumper yields – owing to the olive’s biannual cycle of fructification – would at best have occurred every other year36.

  • 37 For the dossier of ostraka from Abu Mina see Litinas 2008. These texts show that the press was used (...)

19If the upper limit of the aggregate size of vineyards (77 ha) may seem excessive for the sole village of Dehes, this Figure shows that the wine presses of the village were designed to cope not only with the yields obtained locally (i.e. in the hinterland of Dehes) but also – in all likelihood – with grape yields brought over from other villages. None of the villages in the immediate vicinity of Dehes features a similarly high number of presses: this, and the fact that some of the most capable installations were placed along the roads connecting Dehes with Kafr Daryan, Bashmishli and Barisha suggests that at least some wineries were used to process the crop of other villages too. In this way, the wineries’ owners would have been able to farm a profit in a similar fashion to that generated by the winery of Abu Mina in Egypt, on which a dossier of ostraka has recently shed light37.

  • 38 Gentelle 1985: 32-9

20Let us now turn to Si’a. Here, the preservation of the ancient field patterns down to the mid-twentieth century allows us to reflect on the relationship between wineries and the parcellation of the countryside. As P.Gentelle first noted, field patterns in this part of the jebel appear to have been established as early as the second century BC38. While mechanised agriculture has led to the destruction of many ancient boundary walls over the last three decades, ancient field systems have left a trace on aerial photographs taken during the French Mandate period as well as in the late 1950s. Figure 10 uses one such photograph of the Si’a valley as the base map on which wineries and field patterns are plotted.

Fig. 10. Catchment zones of wineries of Si’a (1 filling) plotted on the field pattern of the valley

Fig. 10. Catchment zones of wineries of Si’a (1 filling) plotted on the field pattern of the valley
  • 39 Division of property due to inheritance may indeed have caused the fractioning of holdings. Yet, it (...)

21Even when assuming that wineries were only used once in the course of a vintage season, Figure 10 allows us make an important point: all the vineyards analysed had catchment zones that surpassed the size of the parcels of land immediately surrounding them. This becomes obviously even more apparent if two or three successive fillings of the vats are posited. From this observation, two possible arguments may be drawn: on the one hand it could be argued that, by the time the complex wineries of Si’a were built, the parcellation of the Si’a valley had no longer any straightforward relationship with the patterns of landholding. The case of Si’a 353 best illustrates this point: set on a parcel of less than 3 ha, this large winery would have required at least twice as much land to be intensively cultivated with grape vines in order to work to full capacity. Arguably, the owners of this press would have possessed other holdings, be these continuous to the parcel on which Si’a 353 was set or located further away from it39. On the other hand, the same argument made above as regards press clusters in Dehes may hold true for the bigger wineries of Si’a as well as for the valley’s own press cluster K3-5: that presses were likely to be used to process crops of many different middling landowners whose small holdings would still be to some extent reflected in the parcellation of the territory.

Avoiding the Malthusian trap?

22In this final part, we will turn to the oft-debated issue of technological development as observed in the wine presses described above. This may be characterised as follows: in Si’a, technological development may be said to be represented by the appearance of the complex wineries themselves, which featured additional compartments around the treading floor. In Dehes, instead, the most significant transformation regards the appearance of the lever-and-screw press, which came to substitute the earlier lever-and-windlass system in as many as six of the ten wineries analysed (see table 1 above). What I am going to suggest is that, in broad terms, improvements in production infrastructure was induced by a growing pressure on natural resources. The outcomes of pressure-induced technological developments varied widely from region to region, but - I would suggest – they all aimed at a more efficient and intensive use of labour.

  • 40 See n.20
  • 41 An apt example is that of the treading installations of Achziv, in which a vat had a capacity of al (...)

23In the jebel al-‘Arab large, multi-compartment wineries using a direct screw press appear to have come into widespread use only in the early-Byzantine period and more likely not earlier I than the beginning of the 6th century. This conforms with the archaeological evidence from Israel and Jordan, where this kind of installation, when dated, is rarely if ever associated with occupation before the sixth century (some of the best examples are indeed dated to the Umayyad period, e.g. the wineries of Umm ar-Rasas and Umm al-Walid)40. Direct-screw presses required significant labour (usually three people working the bars that lower the disc of the press onto the stack of grape skins) but they also produced great pressure. This, in turn, meant that stacks could be processed more quickly and, as a consequence, that more stacks can be processed in the course of a day. If more labour was required to work a direct-screw press than a lever press, the additional compartments of the wineries of Si’a guaranteed labour-saving production of high quality must by exploiting gravity: grapes were unloaded onto the upper levels of the compartments and left to produce the “self-flowing” must, which was extracted through the sole pressure of the grapes onto themselves – with little additional labour needs. Thus, while 3-4 workers were busy treading and then pressing a grape load in the treading floor, another four or more stacks (depending on the number of additional compartments) were unloaded by vine workers into the peripheral compartments and allowed to rest and produce protropon must. Once the pressing of the first load was completed in the treading floor, the next load would be taken out of the compartments, thus reducing idle times to a minimum. In this way, the complex wineries of Si’a would not necessarily have provided higher yields than a simple treading floor with a massive collecting vat (some simple found wineries in Israel have vats each of which capable of containing more than 10,000 l of must)41. What they did was to make a more efficient and intensive use of labour.

  • 42 The two presses of Deir Dehes were also seemingly provided with screw counterweights from the start (...)

24In Dehes, as many as six wine presses (out of ten included in this analysis) appear to have undergone two phases; during the first, the pressing was achieved by means of a windlass mechanism; in the second phase this device was changed to the lever-and-screw type (Figure 11). Another three presses were constructed from the start with the screw counterweight whilst at least one other structure (P28) was of the archaic lever-and-weights type42.

Fig. 11. Different pressing technology in the Massif

Fig. 11. Different pressing technology in the Massif

The drum and windlass (left) and the screw counterweight (right)

Callot 1984: Pl.35 & 56

25It is important to note that, when two phases are detectable, it appears that the second phase only involved a switch to the screw mechanism without, for example, an extension of the receiving vat or the digging of a new one. This may suggest that it was not the intrinsic capacity of installations that needed to be modified to make up for lowering returns. We have already seen how the catchment zones of the wineries of Dehes were large enough to require a major portion of the village territory to be devoted to grapevine growing in order to be worked to capacity. The implementation of the screw mechanism, while sometimes joined by an extension of the pressing facilities, was not – in the majority of cases – meant to increase the size of individual loads of grapes (or olives, for that matter) to be pressed, but to make the pressing more efficient. The screw, Hero of Alexandria already acknowledged, made the pressing safer (and its parts less prone to breaking) and, more importantly, created a much stronger pressure than the windlass system would do. While this latter, granted long periods of operation (20-24 h) may be able to press out more juice (particularly in the case of oil), the screw press could process a stack in less than an hour, thus representing a clear improvement on the previous technique.

  • 43 Cato, RR, 19; Hero, Mech., 3.15
  • 44 Amouretti et al. 1984; Mannoni 1985 (both describe the use of the lever-and-screw press in oil maki (...)

26Why, and when did such developments occur? Both lever-and-windlass and lever-and-screw technologies had been invented long before the presses of Dehes were built: while Cato had already described the former, Hero of Alexandria knew the latter43. Hero was quick to highlight the advantages of the screw mechanism: it reduced the risk of breakage, it was less labour-intensive and it eased the lifting of the counterweight. Ethnographic parallels suggest that a large press of this kind could be operated by as few as one to three workers44. Furthermore, the screw made pressing less time-consuming because it made use of both first- and second-class levers by being alternatively turned clockwise and anticlockwise. Finally, the screw counterweight developed a stronger force which, despite producing only limited yield gains, increased the speed of pressing.

27Despite its clear advantages over the windlass press, the lever-and-screw press only found broad diffusion in the Mediterranean from the fourth century onwards. Even when they appeared in numbers, lever-and-screw presses did not take hold everywhere, with important oil exporting regions such as Africa Proconsularis and Tripolitania maintaining the old lever-and-windlass mechanism down to Late Antiquity. Set in the broader debate over technological progress in Antiquity, the late spread of the screw has been often seen, in Finleyan terms, as a reluctance to adopt new strategies of production, with the effect of downplaying the economic rationalism of the ancient landowner. In fact, the reasons underlying the uneven distribution of lever-and-screw presses appear to be more complex.

  • 45 Lewit 2007: 135.
  • 46 Mattingly 1996.

28In a recent article about technological change in oil- and wine-making, Tamara Lewit has noted that screw presses were normally associated with village dwellings or farms which were inhabited by resident owners45. These latter, Lewit continued, were more likely to be open to technological development for the sake of higher yields and efficiency. On the other hand, the absence of screw mechanisms in Tripolitania, Proconsularis and the Methana in Greece (three of her case studies) would have reflected the absentee landlords’ resistance to technological change. In an earlier article, D.Mattingly had argued that it was precisely a rational decision on the part of the great landowners of Tripolitania that legitimated this resistance to change: while lever-and-screw presses prove more efficient and can be operated more quickly, lever-and-windlass ones can process larger loads granted long cycles of operation46.

29Yet, what distinguished the great landowners of Tripolitania from the middling owners of the Limestone Massif was not so much a different attitude to technological development, but a different availability of land. For the latter, population pressure on limited resources created by the incentive to adopt technological improvements, thus bypassing what behavioural economists call the “status quo bias”, namely the condition that leads human beings to reject change regardless of whether such change may or may not bring benefits.

  • 47 Boserup 1965; Wood et al. 1998 (esp. 111-3).

30In adopting strategies for labour intensification, the peasants of northern Syria were doing, or at least attempting to do, what Esther Boserup lucidly described in her essay on the conditions of agricultural growth: traditional agricultural societies respond to shrinking returns by increasing output, something that may be achieved either by increasing input of labour or by adopting limited technological innovation47.

31Admittedly, the evidence remains limited. But further data may allow us to put into question the traditional view s on the demise of rural settlement in marginal zones, which predict that decline set in in the late-sixth century as a consequence, precisely, of a Malthusian crisis.

Bibliographie

Amit & Baruch 2009
D. Amit, Y. Baruch, “Wine Presses with Stone Rollers – An Ancient Phenomenon Seen in a New Light”, in Ayalon, Frankel & Kloner 2009, p. 429-40.

Amouretti et alii 1984
M.-C. Amouretti, G. Comet, C. Ney, J.-L. Paillet, “A propos du pressoir à huile : de l’archéologie industrielle à l’histoire », MEFRA, 96, p. 379-421.

Amouretti & Brun 1993
M.-C. Amouretti, J.-P. Brun (eds.), La production du vin et de l’huile en Méditerranée = Oil and wine production in the Mediterranean area (BCH Suppl. 26), Paris : De Boccard.

Ayalon 2009
E. Ayalon, “Roman and Byzantine Wine Presses at Tell Qasile”, in Ayalon, Frankel & Kloner 2009, p. 195-202.

Ayalon, Frankel & Kloner 2009
E. Ayalon, R. Frankel & A. Kloner (eds.),
Oil and Wine Presses in Israel from the Hellenistic, Roman and Byzantine Periods (BAR International Series 1972), Oxford : Archaeopress.

Bavant & Orssaud 2001
B. Bavant & D. Orssaud, “Stratigraphie et typologie. Problèmes posés par l’utilisation de la céramique comme critère de datation : l’exemple de Déhès”, in E. Villeneuve & P. Watson (eds.), La Céramique Byzantine et Proto-Islamique en Syrie-Jordanie (IVe-VIIIe siécles apr. J.-C.), Beyrouth : IFPO, p. 33-48.

Barret et alii 1986
M. Barret, L. Courtois, D. Orssaud, F. Villeneuve, “Le matériau céramique”, in Dentzer 1986, p. 223-234.

Bavant, Sodini & Tate 1983
B. Bavant, J.-P. Sodini, & G. Tate, “La fouille de Dehes” : 1976-1979’, AAAS, 33, p. 83-97.

Biscop 1997

J.-L. Biscop, Deir Dehes. Monastère d’Antiochène : étude architecturale, Beyrouth : IFPO.

Blanc 2003
P.-M. Blanc, “Conclusions : l’histoire du site”, in Dentzer-Feydy, Dentzer & Blanc 2003, p. 33-7.

Boserup 1965
E. Boserup,
The Conditions of Agricultural Growth. The Economics of Agrarian Change under Population Pressure, Chicago : Aldine Pub. Co.

Brun 2003
J.-P. Brun, Le vin et l’huile dans la Méditerranée antique. Viticulture, oléiculture et procédés de fabrication, Paris : Errance.

Brun 2004
J.-P. Brun, Archéologie du vin et de l’huile dans l’Empire romain, Paris : Errance.

Callot 1984
O. Callot, Huileries antiques de Syrie du Nord, Paris : Geuthner.

Callot 2002-3
O. Callot, “Les broyeurs à rouleaux de Syrie du Nord”, AAAS, 45/6, p. 341-4.

Clauss-Balty 2008
P. Clauss-Balty (ed.),
Hauran III. L’habitat dans les campagnes de Syrie du Sud, Beyrouth : IFPO.

Dar 1986
S. Dar, Landscape and Pattern: an Archaeological Survey of Samaria (800 BCE-636 CE) (BAR International Series 308), Oxford: Archaeopress.

Decker 2009
M. Decker,
Tilling the Hateful Earth. Agricultural Production and Trade in the Late Antique East (Oxford Studies in Byzantium), Oxford: Oxford University Press.

Dentzer 1985
J.-M. Dentzer (ed.), Hauran I. Recherches archéologiques sur la Syrie du Sud à l’époque hellénistique et romaine. Première partie, Paris : Geuthner.

Dentzer 1986
J.-M. Dentzer (ed.), Hauran I. Recherches archéologiques sur la Syrie du Sud à l’époque hellénistique et romaine. Deuxième partie, Paris : Geuthner.

Dentzer et alii 2003
J.-M. Dentzer, F. Villeneuve, R. Donceel, M. Kalos, “Le pressoir”, in Dentzer-Feydy, Dentzer & Blanc 2003, p. 112-90.

Dentzer-Feydy, Dentzer & Blanc 2003
J. Dentzer-Feydy, J.-M. Dentzer, P.-M. Blanc, Hauran II. Les installations de Si’ 8. Du sanctuaire à l’établissement viticole, Beyrouth : IFPO.

Dodinet, Leblanc & Vallat 1992
M. Dodinet, J. Leblanc & J.-P. Vallat, Essai sur l’organisation des paysages du Hauran. Rapport de prospection (Novembre 1992), Unpublished survey report.

Dodinet, Leblanc & Vallat 1994
M. Dodinet, J. Leblanc & J.-P. Vallat, “Étude morphologique des paysages antiques de Syrie”, in P.N. Doukellis & L.G. Mendoni (eds.), Structures rurales et Sociétés antiques. Actes du colloque de Corfou (14-16 mai 1992), Paris : Belles Lettres, p. 425-42.

Donceel 1998
R. Donceel, “Note sur la viticulture dans la région de Si-Qanaouat (Syrie méridionale) à la fin de l’Antiquité”, Revue des archéologues et historiens d’art de Louvain, 31, p. 45-62.

Dunbabin 1999
K.M.D. Dunbabin, Mosaics of the Greek and Roman World, Cambridge : Cambridge University Press.

Frankel 1999
R. Frankel, Wine and oil production in antiquity in Israel and other Mediterranean countries, Sheffield : Sheffield Academic.

Frankel, Avitsur & Ayalon 1994
R. Frankel, S. Avitsur & E. Ayalon (eds.),
History and technology of olive oil in the Holy Land, Kemblesville, PA : Olearius Editions.

Genequand 2001
D. Genequand, “Wadi al-Qantir (Jordanie) : un exemple de mise en valeur des terres sous les Omeyyades”, Studies in the History and Archaeology of Jordan, 7, p. 647-54.

Gentelle 1985
P. Gentelle, 1985. “Éléments pour une histoire des paysages et du peuplement du Djebel Hauran septentrional”, in Dentzer 1985, p. 19-62.

Goor 1966
A. Goor,
The History of the Grape Vine in the Holy Land, Economic Botany, 20, p. 46-64.

Gudovitch 2009
S. Gudovitch,
Wine Presses at Mulabbis (Petah Tiqwa)”, in Ayalon, Frankel, Kloner 2009, p. 203-11.

Kalos 1996
M. Kalos, “Un pressoir à vin byzantin en Syrie du Sud”, Le Monde de la Bible, 101 (Dec 1996), p. 46.

Khalil, al-Nammari 2000
L.A. Khalil,
F.M. al-Nammari, Two Large Wine Presses at Khirbet Yajuz, Jordan, BASOR, 318, p. 41-57.

Kingsley 1999
S. Kingsley,
Specialized production and long-distance trade in Byzantine Palestine, Oxford : DPhil thesis.

Lewit 2007
T. Lewit,
Absent-minded landlords and innovating peasants? The press in Africa and the Eastern Mediterranean, in L. Lavan, E. Zanini, A. Sarantis (eds.), Technology in transition: A.D. 300-650, Leiden : Brill, p. 119-39.

Leblanc & Vallat 1997
J. Leblanc & J.-P. Vallat, “L’organisation de l’espace antique dans la zone de Suweida et de Qanawat (Syrie du Sud) ” in J. Burnouf, J.-P. Bravard & G. Chouquer, La dynamique des paysages protohistoriques, antiques, médiévaux et modernes : actes des rencontres 19-20-21 octobre 1996, Sophia Antipolis : APDCA, p. 35-67.

Litinas 2008
N. Litinas (ed.), Greek ostraca from Abu Mina: (O.AbuMina), Berlin : De Gruyter.

Mannoni 1985
T. Mannoni, Come ho visto funzionare un torchio a leva e vite”, in A. Carandini (ed.), Settefinestre. Una villa schiavistica nell’Etruria romana. Volume 2, Modena : Panini, p. 251-2.

Mannoni 2007
T. Mannoni, Appendice B. Il funzionamento dei torchi, in C. Vismara (ed.), Uchi Maius 3. I frantoi, Sassari : EDES, p. 497-503.

Mattingly 1988
D. Mattingly, Megalithic madness and measurement. or how many olives could an olive press press?, Oxford Journal of Archaeology, 7, p. 177-95.

Mattingly 1996
D. Mattingly, “Olive Presses in Roman Africa: Technical Evolution or Stagnation?”, in M. Khanoussi, P. Ruggeri & C. Vismara, L’Africa Romana XI.2. Atti dell’XI convegno di studio. Cartagine 15-18 dicembre 1994, Ozieri : Editrice il Torchietto, p. 577.-95.

Orssaud 1992
D. Orssaud, “Le passage de la céramique byzantine a la céramique islamique. Quelques hypothèses à partir du mobilier trouvé à Déhès”, in P. Canivet & J.-P. Rey-Coquais (eds.), La Syrie de Byzance à l’Islam. Actes du colloque international Lyon - Maison de l’Orient Méditerranéen. Paris - Institut du Monde Arabe 11-15 Septembre 1990. Damas : Institut français de Damas, p. 219-28.

Orssaud, Barret & Blanc 2003
D. Orssaud, M. Barret & P.-M. Blanc, “La céramique de Si’ en pâte A. Essai de formalisation des éléments descriptifs”, in Dentzer-Feydy, Dentzer & Blanc 2003, p. 199-222.

Orssaud & Sodini 2003
D. Orssaud & J.-P. Sodini, “Le “brittle ware” dans le massif calcaire (Syrie du Nord) ” in C. Bakirtzis (ed.), VIIe Congrès International sur la Céramique médiévale en Méditerranée, Thessaloniki, 11-16 Octobre 1999 : actes, Athènes : Caisse des Recettes Archéologiques, p. 491-504.

Piccirillo 1997
M. Piccirillo, Umm al-Rasas, Church of St Paul : Southestern flank, Liber Annuus, 47 (1997), p. 542-6.

Piccirillo 1998
M. Piccirillo, Umm al-Rasas Kastron Mef’a. Excavation Campaign 1997. Church of St. Paul : northern and sourthern flanks, Liber Annuus, 48, p. 484-8.

Schachner 2006
L.A. Schachner, Economic production in the monasteries of Egypt and Oriens, AD 320-800. Oxford : DPhil Thesis.

Sodini et alii 1980
J.-P. Sodini, G. Tate, B. Bavant, S. Bavant, J.-L. Biscop, D. Orssaud, C. Morrisson & F. Poplin, “Déhès (Syrie du nord) Campagnes I-III (1976-1978) recherches sur l’habitat rural”, Syria, 57, p. 1-304.

Syon 2009
D. Syon, “A Wine Press from Achziv”, in Ayalon, Frankel, Kloner 2009, p. 35-40.

Tate 1992
G. Tate, Les campagnes de la Syrie du Nord du IIe au VIIe siècle : un exemple d’expansion démographique et économique à la fine de l’antiquité, Paris : Geuthner.

Tchalenko 1953-8
G. Tchalenko, 1953-8. Villages antiques de la Syrie du Nord : le massif du Bélus à l’époque romaine. 3 vols, Paris : Geuthner.

Tchalenko 1971
G. Tchalenko, “Traits originaux du peuplement de la Haute-Syrie du 1er au 7e siècle, tels que les révèle l’architecture”, AAAS, 21, p. 289-92.

Vallat & Leblanc 2008
J.-P. Vallat & J. Leblanc, “Archéologie du paysage et prospections : habitat rural et parcellaires du djebel al-‘Arab (Si’/Qulib)”, in Clauss-Balty 2008, p. 19-40.

Villeneuve 1985
F. Villeneuve, “L’économie rurale et la vie des campagnes dans le Hauran antique (ier siècle avant J.-C.- vie siècle après J.-C.) ”, in Dentzer 1985, p. 63-136.

Villeneuve 1997
F. Villeneuve, “Économie et société des villages de la montagne Hauranaise à l’époque romaine”, AAAS, 41, p. 31-8.

Wood et alii 1998
J. Wood, G.L. Cowgill, R.E. Dewar, N. Howell, L.W. Konigsberg, J.H. Littleton, R.D. Attenborough & A.C. Swedlund, “A Theory of Preindustrial Population Dynamics : Demography, Economy, and Well-Being in Malthusian Systems, Current Anthropology, 39, p. 99-135.

Wuthnow 1930
H. Wuthnow,
Die semitischen Menschennamen in griechischen Inschriften und Papyri des vorderen Orients, Leipzig : Dieterich.

Yanai 2009
E. Yanai, A Byzantine Wine Press at Tel Hefe, in Ayalon, Frankel & Kloner 2009, p. 149-52.

Yevin & Finkielsztejn 2009
Z. Yevin & G. Finkielsztejn, Wine and Oil Presses at H.Castra, in Ayalon, Frankel & Kloner 2009, p. 105-117.

Zerbini 2010
A. Zerbini, The Si’a Valley Survey, Palestine Exploration Quarterly, 142.3, p. 217-21.

Notes

1 On the hinterland of Si’a see: Gentelle 1985; Villeneuve 1985; Dodinet, Leblanc & Vallat 1992; Eid. 1994; Leblanc & Vallat 1997; Vallat & Leblanc 2008. On wine production and wineries in this area see: Kalos 1996; Dentzer et al. 2003: 112-90; Zerbini 2010.

2 The earliest excavations of Dehes were planned by Georges Tchalenko, but only realised after his death: Tchalenko 1971; Sodini et al. 1980; Bavant, Sodini, Tate 1983. More recently, excavation has resumed in the area of the so-called “villa” of Dehes, but a full excavation report is yet to be published. However, see: Orssaud 1992; Bavant & Orssaud 2001; Orssaud & Sodini 2003.

3 Dr Callot’s work on the presses of Dehes remains entirely unpublished. I should like to thank him for allowing me access to his unpublished notes, which have supplied vital evidence for the preparation of this paper.

4 Amouretti & Brun 1993. Brun 2003; Id. 2004.

5 The best general works on this subject are: Frankel, Avitsur & Ayalon 1994; Frankel 1999; Ayalon, Frankel, Kloner 2009.

6 Callot 1984

7 Callot 1984: 31-70.

8 See, in particular, Brun 2004: 111-9. For presses with rollers and their identification with wine presses see Frankel 1999; Amit & Baruch 2009. O. Callot, who has studied these structures in detail continues to regard them as oil presses: Callot 1984; on presses with stone rollers: Callot 2002-2003.

9 Mainly Tosef.Ter. 7:15. Frankel 2009: 11.

10 The procedure is well explained by Callot 1984:74 (section B). For a recent detailed discussion of the decantation process in oil making see Mannoni 2007. Some presses with bell-shaped vats were, however, provided with decantation basins placed just before the main vat and which could be used to collect the oil with the overflowing technique as explained in Callot 1984: 72-3.

11 Grape remains and olive kernels were found in the basins of the large press of Sergilla pictured above (G.Charpentier, pers.comm.).

12 For the W press of Deir Dehes – which combines a bell-shaped vat with a rotary crushing mill – see Biscop 1997: 23-6; Pl.62-3.

13 These are, in chronological order, the famous temple oil press of Kafr Nabo (IGLS II 376, AD 224); a simple treading floor from Kh. Sheikh Barakat (Callot 1984: Pl.136b); and a large oil press from Hass in the jebel Zawiyé (IGLS IV 1509, AD 372).

14 Sodini et al. 1980: 155-8

15 Sodini et al. 1980: 153; 156.

16 Dating relies primarily on the finding of decorated blocks in re-use in the walls of P27, to which Callot attributes a late-sixth or early-seventh century date (pers. comm.).

17 Dentzer et al. 2003.

18 For the technical specifications of these wineries see Zerbini 2010.

19 The definition of the “complex” type is due to Frankel (Frankel 1999). In the surveys conducted by P. Gentelle, M. Kalos, J.-P. Vallat and J. Leblanc between the 1970s and 2000s simpler wineries were also discovered such as Si’a 222 and Si’a 291. Dentzer et al. 2003: 128. Vallat & Leblanc 2008: 34 Pl.4

20 For some examples of complex wineries from Israel see Yevin & Finkielsztejn 2009 (H.Castra); Yanai 2009 (Tell Hefer). From Jordan: Piccirillo 1997; Id. 1998 (Umm ar-Rasas). Khalil & al-Nammari 2000 (Kh. Yajuz); Genequand 2001 (Wadi al-Qantir).

21 Pliny, NH XIV 11.85; 9.75; Athaen., Deipn. I 30b; II 45e. On this see Decker 2009: 128-9. Vats for the collection of protropon are also found in the large winery 37S of Mulabbis (Gudovitch 2009: 204-7). On the uses of the additional compartments in the wineries of the jebel al-‘Arab see Donceel 1998: 52-7; Dentzer et al. 2003: 158-9.

22 Blanc 2003: 35-6.

23 Dentzer et al. 2003:139.

24 For a discussion of fabrics in Si’a see: Barret et al. 1986; Orssaud, Barret & Blanc 2003.

25 Although the Nabataean inscription contains a lacuna in the patronymic, but the name of the dedicant is well visible and reads {C}WYDW, ‘Uwayd, which corresponds to the Αουεδιος of the Greek text (for attestations of this name see Wuthnow 1930:24). The patronymic in the Greek text is Motaimos/Notaimos, an otherwise unattested name.

26 Frankel 1999; Brun 2003: 63-4.

27 Vintage in the Levant has a duration of between four and seven weeks (usually mid-August/mid-October). See Dar 1986: 154.

28 Since must flowed into the receiving vat through channels/openings hewn into the wall separating the vat from the treading floor, vintage had to be stopped once the level of must in the vat reached the bottom of the channel, or else the must would have spilled back onto the treading floor. Most wine presses of the Limestone Massif show signs that wooden stoppers or some other equipment was employed to block the channels/openings whilst fermentation was underway. See e.g. Callot 1984: Pl.92. In the wineries of the jebel al-‘Arab, the additional vats had no direct connection with the operations carried out in the treading floor from which they were actually separated by means of a shallow wedge (see Figure 8).

29 For a plan of Behyo see Tchalenko 1953-8: Pl. 110 (Vol.2).

30 Callot 1984: 125.

31 On the hilltop village of Si’a see Villeneuve 1997.

32 Geoponika, 6.11. See also Decker 2009: 122-9 for discussion and 263-71 for the value of the Geoponika in illustrating agricultural practices in the Eastern provinces.

33 According to Goor (1966:63), in 1946 Mandate Palestine produced a total of 9,000 tons of grapes for wine production. From these, 6 million litres of must were produced equivalent to a ratio of .67 kg grapes/l must. Since the specific gravity of must is close to 1 (1.05-1.07 depending on the amount of solid residue left in it), we could say that 1 kg of grapes produced between 550 g and 650 g of must.

34 Data published in the official reports Études sur la Syrie économique (1953-9).

35 For Dehes see especially Sodini et al. 1980. More generally for the Limestone Massif: Tate 1992.

36 These figures are obtained by applying D.Mattingly’s approach to the calculation of maximum oil outputs (as outlined in Mattingly 1988). The ratio of olive trees/ha used (145) and oil yield/tree (3.85 kg) are taken from Schachner 2006:158-62.

37 For the dossier of ostraka from Abu Mina see Litinas 2008. These texts show that the press was used by as many as 500 vine growers over a vintage season (which lasted 12-14 days) and that it was capable of processing some 250 baskets (or 7,500 l of must) per day for a total of over 60,000 litres of wine overall. See Schachner 2006: 188-9.

38 Gentelle 1985: 32-9

39 Division of property due to inheritance may indeed have caused the fractioning of holdings. Yet, it seems unlikely that the vineyards making use of this press could be more than a couple of kilometres away from it.

40 See n.20

41 An apt example is that of the treading installations of Achziv, in which a vat had a capacity of almost 25000 litres (Syon 2009:35).

42 The two presses of Deir Dehes were also seemingly provided with screw counterweights from the start. Biscop 1997: 21-6.

43 Cato, RR, 19; Hero, Mech., 3.15

44 Amouretti et al. 1984; Mannoni 1985 (both describe the use of the lever-and-screw press in oil making).

45 Lewit 2007: 135.

46 Mattingly 1996.

47 Boserup 1965; Wood et al. 1998 (esp. 111-3).

Table des illustrations

Titre Fig. 1. The Si’a Valley
Légende Dentzer-Feydy, Dentzer, Blanc 2003: Pl.3 Figure 2
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-1.png
Fichier image/png, 243k
Titre Fig. 2. The village and immediate hinterland of Dehes
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-2.png
Fichier image/png, 294k
Titre Fig. 3. The underground oil press in the courtyard of complex 101-8
Crédits Sodini et al. 1980: 157 Figure 200
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-3.png
Fichier image/png, 8,2k
Titre Fig. 4. Complex 101-8 in Dehes
Crédits Sodini et al. 1980 : Figure 5
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-4.png
Fichier image/png, 37k
Titre Fig. 5. Wineries in the Si’a Valley
Crédits   
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-5.jpg
Fichier image/jpeg, 288k
Titre Fig. 6. Detail of the mosaic pavement of the church of Lot and Procopius at Nebo/Kh. al-Mukhayyat
Crédits Dunbabin 1999: 198 Figure 211
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-6.jpg
Fichier image/jpeg, 300k
Titre Fig. 7. A reconstruction of a direct-screw press from Tell Qasile
Crédits Ayalon 2009: 202 Figure 36.6
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-7.jpg
Fichier image/jpeg, 288k
Titre Fig. 8. Winery Si’a 21 (looking south)
Légende The photograph shows the central treading floor (with morticed stone for the anchorage of the direct-screw press) and the additional loading compartments on three sides
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-8.jpg
Fichier image/jpeg, 188k
Titre Fig. 9.1-3. Catchment zones of wineries of Dehes. The three scenarios correspond to more intensive uses of the installations (1, 2, 3 fillings)
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-9.jpg
Fichier image/jpeg, 284k
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-10.jpg
Fichier image/jpeg, 276k
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-11.png
Fichier image/png, 148k
Titre Fig. 10. Catchment zones of wineries of Si’a (1 filling) plotted on the field pattern of the valley
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-12.jpg
Fichier image/jpeg, 256k
Titre Fig. 11. Different pressing technology in the Massif
Légende The drum and windlass (left) and the screw counterweight (right)
Crédits Callot 1984: Pl.35 & 56
URL http://books.openedition.org/ifpo/docannexe/image/2886/img-13.png
Fichier image/png, 18k

Auteur

PhD student in Classics at Royal Holloway, University of London

Le texte et les autres éléments (illustrations, fichiers annexes importés) sont sous Licence OpenEdition Books, sauf mention contraire.

Acheter

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search