Content uploaded by Michael Decker
Author content
All content in this area was uploaded by Michael Decker on Jul 17, 2020
Content may be subject to copyright.
187
Journal of World History, Vol. 20, No. 2
© 2009 by University of Hawai‘i Press
Plants and Progress: Rethinking the
Islamic Agricultural Revolution*
michael decker
University of South Florida
In 1974, Andrew Watson published an infl uential article titled “The
Arab Agricultural Revolution” in which he argued that Muslim agri-
culturalists transformed Mediterranean farming beginning shortly after
the seventh-century conquests of most of the Middle East and North
Africa. Watson elaborated his claims in another article, “A Medieval
Green Revolution” (1981), and in a monograph, Agricultural Inno-
vation in the Early Islamic World: The Diffusion of Crops and Farming
Techniques, 700 –1100.1 In these works, Watson proposed a medieval
“Green Revolution” that entailed the spread of intensive methods of
farming and irrigation technology and a rise in crop yields because of
these farming techniques. Accompanied by a demographic upswing,
intensive farming methods elevated labor requirements and yielded
higher crop surpluses. In turn the abundance of food supported the
larger and more numerous cities of the Muslim world.2
In support of his thesis, Watson charted the advance of seventeen
food crops and one fi ber crop that became important over a large area
* I am grateful for Mr. Nick Maroulis who supported this work via my endowed profes-
sorship. I also thank the editor and anonymous reviewer for their insightful criticisms of
this paper.
1 The trio of Andrew Watson’s works, “The Arab Agricultural Revolution and Its Dif-
fusion, 700–1100,” Journal of Economic History 34 (1974): 8–35; “A Medieval Green Revo-
lution,” in The Islamic Middle East, 700 –1900, ed. Abraham Udovitch (Princeton, N.J.:
Darwin Press, 1981), pp. 29–58; and Agricultural Innovation in the Early Islamic World (Cam-
bridge: Cambridge University Press, 1983). All dates, unless otherwise specifi ed, are c.e.
2 Watson, “Arab Agricultural Revolution,” pp. 9–17; and Watson, Agricultural Innova-
tion, pp. 2–3.
188 journal of world history, june 2009
of the Mediterranean world during the fi rst four centuries of Islamic
rule (roughly the seventh through eleventh centuries c.e.). Among
these fl ora we fi nd familiar items whose impact on our diets today is
self-evident: Asiatic rice, sugar cane, banana and plantain, lemon,
lime, hard (durum) wheat, and sorghum. Others are of less importance
but familiar and signifi cant: watermelon, eggplant, spinach, artichoke,
colocasia, sour orange, shaddock, mango, and coconut palm. Eventu-
ally, during the European colonization of the New World, a number of
the crops of the Green Revolution became major components of the
Columbian Exchange and thence passed into global agriculture and
industry. Chief among the crops are ubiquitous and fundamental New
World planter crops: sugar cane, banana, and rice.3
Following the mid seventh-century collapse of Byzantine author-
ity in the eastern Mediterranean and the demise of Sasanian Persia,
the Muslim caliphate unifi ed under a single authority for the fi rst time
lands from Afghanistan to Spain, an unprecedented and unduplicated
success. The colossal embrace of the Islamic polity meant that east and
west were connected as never before. This unity facilitated communi-
cation and trade and created an atmosphere that encouraged the spread
of knowledge and goods. Further, the Arabs’ own familiarity with farm-
ing in arid regions meant that they were both experienced in develop-
ing marginal lands and interested in doing so.
Some of the fl ora of the Green Revolution were adapted to drylands,
such as hard wheat and watermelon, and thus permitted the extension
of the land without irrigation. Others, such as sugar cane and coconut,
were able to fl ourish on soils that had been rendered saline by previous
irrigation schemes that had rendered the land useless for conventional
crops such as wheat and barley. Crops such as rice were best grown
under regimes where heavy water input was available. To supply the
demands of thirsty crops, Islamic farmers turned to numerous meth-
ods of irrigation, including machinery and underground canals (qanats)
that in turn sustained the expansion of plants and the farmed land-
3 An overview of the transfer of major crops is found in Alfred W. Crosby, The Colum-
bian Exchange: Biological and Cultural Consequences of 1492 (Westport, Conn.: Praeger
Publishers, 1972), pp. 64–121; and Alfred W. Crosby, Ecological Imperialism: The Biological
Expansion of Europe, 900 –1900 (Cambridge: Cambridge University Press, 1986). Further
light on the Columbian exchange supra those crops that formed part of the Green Revolu-
tion may be found in Judith A. Carney, “African Rice in the Columbian Exchange,” Journal
of African History 42, no. 3 (2001): 377–396. Finally, a brief overview of the environmental
transformation of the New World through European farming methods in the age of colonial-
ism may be glimpsed in B. L. Turner II and Karl W. Butzer, “The Columbian Encounter and
Land-Use Change,” Environment 34, no. 8 (1992): 16 –20.
Decker: Plants and Progress 189
scape. The massive infl ux of crops that quickly became vital to life,
expanded the farmed landscape, and replaced inferior foodstuffs altered
the Middle East and Mediterranean drastically.
Watson also proposes that the Muslims encroached on the tradi-
tional summer fallow of the ancient world: “the opening of the summer
season was one of several factors—perhaps the principal one—permit-
ting systems of rotation which made much more intensive use of the
land.” This statement ignores evidence that shows that the Romans,
Byzantines, and Sasanian Persians exercised continuous cropping of
the land and a sophisticated annual crop rotation.4
Few changes that Watson traces are more important than the asser-
tion of Islamic introduction of new techniques of irrigated farming.
In the area of irrigation technology, the Muslims are credited with
an array of new devices and management techniques that, while not
new in some regions of their empire, were a rarity before the arrival of
the Muslim conquerors.5 Beyond aiding the acclimatization and dis-
semination of new plants, these hydraulic systems increased yields of
earlier established crops. Because the preindustrial world was agrarian
in it essence, the alteration of farming life was embedded within a cas-
cade of change that touched on labor, technology, trade, industry, and
demography. Grand in its scope and scale, the Green Revolution posits
a classical Islamic agriculture fundamentally different than preconquest
Roman, Byzantine, Persian, and Jewish practices.
Watson’s thesis has been accepted without serious challenge.
Though some criticism of faulty methodology, errors of fact, and false
assumption were raised by early reviewers, major secondary specialist
and general sources now take for granted that his portrayal refl ects real-
ity on the ground in the medieval Mediterranean.6 Given the stability
4 The Geoponica, a Greek farming manual compiled in the tenth century from late
antique sources 2:3 shows that summer sowing was normal in the Roman Mediterranean.
Since the handbook was compiled from late antique sources mainly from Syria, it does not
refl ect the “minor role in some parts of the northern Mediterranean” (italics in the original)
ascribed it by Watson, “Arab Agricultural Revolution,” p. 10; see Geoponica, ed. Heinrich
Beckh (Leipzig, 1886), 2.3. On the Geoponica, see below. On Roman systems of crop rota-
tion, see recently G. Kron, “Roman Ley-Farming,” Journal of Roman Archaeology 13 (2000):
277–287; for the Jewish and Sasanian Mesopotamian case, see Julius Newman, Agricultural
Life of the Jews in Babylonia between 200 C.E. and 500 C.E. (Oxford: Oxford University Press,
1932).
5 Watson, “Arab Agricultural Revolution,” pp. 9 –13.
6 Critical reviewers include J. Johns, “A Green Revolution?” review of Agricultural
Innovation in the Early Islamic World, by Andrew Watson, Journal of African History 25
(1984): 343–344; and C. Cahen, review of Agricultural Innovation in the Early Islamic World,
by Andrew Watson, Journal of the Social and Economic History of the Orient 29, no. 2 (1986).
Those who accept Watson’s view include J. H. Galloway, “The Mediterranean Sugar Indus-
190 journal of world history, june 2009
of farming systems throughout much of recorded history, the possibility
of a major break with the agrarian past, driven by new crops, methods,
and technologies, poses an intriguing prospect.
It is impossible in the space provided to undertake a thorough cri-
tique and analysis of all components of the medieval Green Revolution
thesis. Though there has yet to be a systematic treatment of these in
light of the Green Revolution thesis, some specialist work has been
done on ancient demography and irrigation that rebuts sizeable parts
of the Watson thesis.7 A recent study of Mesopotamian irrigation sys-
tems shows, for example, that intensive hydraulic farming was at its
apogee there under the Sasanians; subsequent Islamic work was largely
restricted to restoration or expansion of older systems. Work in the
Mughan steppe of northwestern Iran demonstrates that the Sasanian
systems there were abandoned in the seventh century; the region’s agri-
culture never recovered and the area became a sparsely populated pas-
toral landscape until the twentieth century.8 In Spain, Glick’s theory of
a radical transformation of the irrigated landscape has been challenged
by thorough archaeological survey work that suggests that the Islamic
hydraulic infrastructure was built up from the prior Roman network,
and that it augmented rather than replaced the latter landscape.9
Nothing has been written, however that attacks the central pillar
of Watson’s thesis, namely the “basket” of plants that is inextricably
try,” Geographical Review 67 no. 2 (1977): 179; D. Hill, “Engineering,” in Encyclopedia of the
History of Arabic Science, 3 vols., ed. Roshdi Rashed (London: Routledge, 1996), 3:751–795;
Ahmad Y. al-Hassan and Donald R. Hill, Islamic Technology: An Illustrated History (Cam-
bridge: Cambridge University Press, 1986), p. 206; A. Dallal, “Science, Medicine, and
Technology: The Making of Scientifi c Culture,” in The Oxford History of Islam, ed. John
L. Esposito (Oxford: Oxford University Press, 2000), pp. 196 –197; and Francis Robinson,
ed., The Cambridge Illustrated History of the Muslim World, with a foreword by Ira M. Lapidus
(Cambridge: Cambridge University Press, 1996), pp. 149 –154.
7 Örjan Wikander, Handbook of Ancient Water Technology (Leiden: Brill, 2000); P. Briant,
Irrigation et drainage dans l’antiquité, Qanâts et canalisations souterraines en Iran (Paris, 2000);
Ariel S. Lewin and Pietrina Pellegrini, eds., Settlements and Demography in the Near East
in Late Antiquity: Proceedings of the Colloquium, Matera, 27–29 October 2005 (Pisa: Istituti
editoriali e poligrafi ci internazionali, 2006).
8 Peter Christiansen, The Decline of Iranshahr: Irrigation and Environments in the History
of the Middle East 500 B.C. to A.D. 1500 (Copenhagen: Museum Tusculanum Press, 1993),
pp. 73–116. For the Mughan Steppe, see Karim Alizadeh and Jason A. Ur, “Formation and
Destruction of Pastoral and Irrigation Landscapes on the Mughan Steppe, North-Western
Iran,” Antiquity 81, no. 311 (2007): 148–160 for recent work on Sasanian era agricultural
landscapes and their medieval transformation.
9 Thomas F. Glick, Irrigation and Society in Medieval Valencia (Cambridge, Mass.: Har-
vard University Press, 1970). A more nuanced view is held by Karl W. Butzer et al., “Irriga-
tion Agrosystems in Eastern Spain: Roman or Islamic Origins?” Annals of the Association of
American Geographers 75, no. 4 (1985): 479 – 509.
Decker: Plants and Progress 191
linked to all other elements of his analysis. This work will therefore
assess the place and importance of four crops of the “Islamic Agri-
cultural Revolution” for which there is considerable pre-Islamic evi-
dence in the Mediterranean world. Plants and the allied technologies
needed to grow them traveled as a bundle into semiarid zones of the
Islamic Mediterranean, and their exploitation accompanied the demo-
graphic growth and prosperity of medieval Islam. A case study of the
plants themselves thus offers a sound initial view of the major tenets of
Watson’s work, since the spread of these plants offers insight into the
complexities that attended the widespread adoption of certain plants
over others, and ultimately allows us to test the notion that Muslim era
agriculture was essentially different than what came before. All four
crops were selected because of the evidence for their pre-Islamic use
in the Mediterranean. Three of these crops remain of great economic
signifi cance today: durum wheat, rice, and sorghum. The fourth exam-
ple, artichoke, is of only modest signifi cance today, but serves as a fi ne
proxy for the bevy of medieval “minor crops” that fi ll out Watson’s own
group.10
It must be stated from the outset that the aim is not to overturn the
whole of Watson’s thesis by showing that all of the plants he discusses
were present in the Mediterranean world before the coming of Islam;
clearly several were not. While it is conceded that Muslims made an
important contribution to world farming through the westward diffu-
sion of some crops, the Islamic introduction of agronomic techniques
and materials was not as widespread, as consistent, nor as deeply applied
as the Green Revolution proposes. Watson failed to account for a size-
able body of evidence that demonstrates the presence of new plants,
whose place in the landscape demonstrates an interest in experimen-
tal crops, intensive farming methods, and the widespread application
of irrigation technology that he views as belonging to a much later
period. The failure to account for these data reveals a profound lack of
interest in the pre-Islamic landscape and a host of fl awed assumptions.
The resultant Green Revolution thesis is therefore a simplistic, linear
model of the movement of ideas and goods that fails to acknowledge
the complexities of these transmissions, the correct range of their dif-
fusion, and the real limits of their signifi cance.
10 Jeremy Johns already noted that banana, coconut, mango, shaddock, and others of
Watson’s examples had no importance in Mediterranean agriculture; see Johns, “A Green
Revolution?” p. 343.
192 journal of world history, june 2009
Durum Wheat
The most important crop considered by Watson is durum wheat. Today
the various varieties of wheat account for as much as 20 percent of
daily calories consumed worldwide. Durum wheat (hard wheat, Triti-
cum durum) is a major variety, most familiar to world consumers as the
main ingredient of most pastas. Watson argues that durum wheat was
unknown or scarcely grown in the pre-Islamic world, when in fact it
was widely cultivated and consumed in the classical Mediterranean.
Durum wheat possesses several advantages over other varieties, one
of which is its lack of the tough outer husk common to most ancient cul-
tivars. In the case of husked grains, before bread can be made, the outer
culm must be removed, usually by pounding in a mortar and pestle or
parching. Durum, however, is free-threshing; it does not require signifi -
cant preparation before milling, and this saves the producer signifi cant
effort during processing. Hard wheat is also drought resistant, a quality
that makes it especially attractive in the Mediterranean, with its sparse
and often unreliable rainfall. Durum also keeps for an especially long
time, a key advantage over other varieties in a world before modern
storage methods prevailed.
The pre-Islamic textual and material evidence for the cultivation of
durum wheat is substantial. As early as the seventh millennium b.c.e.,
the family to which durum belonged is attested archaeologically at sites
in Anatolia and Syria. These free-threshing tetraploid (having four
chromasome sets) wheat varieties appear in Iran by the sixth millen-
nium b.c.e. and in Greece by the fi fth millennium b.c.e.11 Durum wheat
was probably already present in the Neolithic period in the valley of
the Euphrates, where tetraploid wheat remains have been identifi ed in
archaeological excavation.12 The group of free-threshing, durum-like
wheats were domesticated by the fi fth millennium b.c.e. In the Balikh
River valley of northern Syria, a hard wheat species that is probably
durum or a closely related species was recovered in Bronze Age levels.
At Tell Keisan, hard wheat was found in archaeobotanical investiga-
tions in contexts dating to the eleventh century b.c.e.13
11 Daniel Zohary and Maria Hopf, Domestication of Plants in the Old World, 3rd ed.
(Oxford: Oxford University Press, 2000), p. 46.
12 George Willcox and Valérie Roitel, “Rapport archaeobotanique préliminaire de trois
sites précéramiques du Moyen-Euphrate (Syrie),” Cahiers de l’Euphrate 8 (1998): 75.
13 Willem van Zeist, “Evidence for Agricultural Change in the Balikh Basin, Northern
Syria,” in The Prehistory of Food: Appetites for Change, ed. Chris Gosden and Jon Hather
(London: Routledge, 1999), pp. 350–374; Mordechai Kislev, “Contenu d’un silo a blé de
l’époque du fer ancient,” in Tell Keisan 1971–1976: Une cité phénicienne en Galilée, ed.
Jacques Friend and Jean-Baptiste Humbert (Paris, 1980), pp. 361–380.
Decker: Plants and Progress 193
A good deal of evidence exists then to suggest that durum or its
close relatives had been an important feature in the eastern Mediterra-
nean landscape for centuries before the Macedonian Ptolemaic dynasty
fostered its cultivation in Egypt between the fourth century b.c.e. and
the fi rst century c.e. Durum featured in an account dated 248–245
b.c.e. preserved in the Zenon Papyri.14 It apparently supplanted the
old emmer varieties as the dominant wheat in Egypt and continued to
gain ground among the cultivators of the Roman Empire from the fi rst
through sixth centuries c.e.
By the second century, durum wheat was already widespread across
the Mediterranean. The fi rst-century medical author Dioscorides, who
hailed from the Cilician city of Anazarbus in what is now southeast-
ern Turkey, was familiar with the grain and described it in his com-
pendium of medicines.15 The early Roman agronomist Columella, who
also wrote in the fi rst century, advised how durum should be grown, as
did the later agricultural writer Palladius (fourth century). The sec-
ond-century physician Galen compared durum wheat to barley, and he
considered them similar in their nourishing qualities.16 The mention of
durum by the early third-century polymath Athenaeus of Alexandria in
Egypt further suggests it became a familiar part of the diet of antiquity.17
The fourth-century physician Oribasius noted that durum then grew in
Anatolia, and the doctor considered this wheat variety to be superior
to most other kinds of cereals.18 Based on its mention in the Geoponica,
a medieval compilation of predominantly late antique material (fourth
through seventh centuries), durum was extensively grown around the
ancient Egyptian metropolis of Alexandria, a city that at that time lent
its name to a particular subvariety of the wheat.19
14 PZen 77, ed. William Westermann, Clinton Keyes, and Herbert Liebesny, Columbia
Papyri: Greek Series no. 4: Zenon Papyri, 2.52–53.
15 Dioscorides, Materia Medica Libra Quinque, ed. Max Wellman (Berlin, 1906–1914),
2.285.
16 Galen, De victu attenuante, ed. Karl Kalbfl eisch, Galeni de victu attenuante (Corpus
medicorum Graecorum 5.4.2. Leipzig: Teubner, 1923): 433–451, section 34.1.
17 Athenaeus, Deipnosophistae, trans. Charles Burton Gulick, 7 vols. (Cambridge,
Mass.: Harvard University Press, 1927–1941), 1:120, 2:26, 40, 88, 280.
18 Mark Grant, Dieting for an Emperor: A Translation of Books 1 and 4 of Oribasius’ Medi-
cal Compilations with an Introduction and Commentary (Leiden: Brill Academic Publishers,
1997), pp. 32–33 for durum (semidalite) wheat; M. S. Spurr, Arable Cultivation in Roman Italy
(c. 200 B.C.– c. A.D. 100) (London: Society for the Promotion of Roman Studies, 1986), pp.
15–17. Columella, De Re Rustica, ed. W. Lundström (Uppsala, 1897), 2.6.1, 2.6.4, 6.5.2;
Palladius, Opus Agriculturae, ed. R. H. Rodgers (Leipzig, 1975), bk. 2.9, p. 53, l. 2; bk. 10.2,
p. 187, l. 5.
19 Heinrich Beckh, ed., Geoponica Sive Cassiani Bassi Scholastici De Re Rustica Eclogae
(Stuttgart, 1994), bk. 3.3.
194 journal of world history, june 2009
Archaeological evidence also supports the view that, by the Roman
era at the latest, durum had replaced emmer wheat as the dominant
grain in Egypt. At Mons Claudianus, in the Eastern Desert, the early
Roman mining settlement relied heavily on durum imported from the
Nile valley.20 The excavations at Karanis, in the Fayyum of central
Egypt, produced fi nds of durum wheat from the early Roman period.21
Along the Red Sea coast of Egypt at Quseir al-Qadim and at Berenike,
Roman levels have yielded fi nds of cultivated hard wheat that is also
probably durum. As at Mons Claudianus, the durum found at these Red
Sea ports was likely imported from the Nile valley, which further indi-
cates the prevalence of the grain within the classical period Egyptian
agricultural regime.22
In North Africa, durum wheat remains were recovered from Roman
levels at Ghirza in Libya, and the UNESCO Libyan Valley surveys dis-
covered evidence of its production in Cyrenaica (modern Libya), while
the farming of durum in Africa Proconsularis, one of the granaries of
the empire, is supported by textual evidence.23
Asiatic Rice
Asiatic rice (Oryza sativa) is currently the second most important grain
in the world diet.24 Watson believed the plant to have been widely
diffused in western Eurasia only after the Islamic conquests. Although
20 Marjike van der Veen, “The Plant Remains from Mons Claudianus, a Roman Quarry
Settlement in the Eastern Desert of Egypt—an Interim Report,” Vegetation History and
Archaeobotany 1–2 (1996): 137–141; Marjike van der Veen, “A Life of Luxury in the Desert?
The Food and Fodder Supply to Mons Claudianus,” Journal of Roman Archaeology 11 (1998):
101–116.
21 C. E. Leighty, “Botanical and Zoölogical Reports,” in Karanis: The Temples, Coin
Hoards, Botanical and Zoölogical Reports Seasons 1924–1931, ed. A. Boak (Ann Arbor,
Mich., 1933), pp. 87–88.
22 Marjike van der Veen, “Trade and Diet at Roman and Medieval Quseir al-Qadim,
Egypt: A Preliminary Report,” in Food, Fuel, and Fields: Progress in African Archaeobotany,
ed. Katharina Neumann, Ann Butler, and Stefanie Kahlheber (Köln, 2003), pp. 207–212;
Rene T. J. Cappers, “Archaeobotanical Remains,” in Berenike 1995: Preliminary Report of
the 1995 Excavations at Berenike (Egyptian Red Sea Coast) and Survey of the Eastern Desert,
ed. Steven Sidebotham and Willemina Wendrich (Leiden, 1996), pp. 332–335; and Spurr,
Arable Cultivation in Roman Italy, pp. 15–17.
23 G. Barker, “A Tale of Two Deserts: Contrasting Desertifi cation Histories on Rome’s
Desert Frontiers,” World Archaeology 33, no. 3 (2002): 494; Marjike van der Veen, “Botani-
cal Remains,” in Ghirza: A Libyan Settlement in the Roman Period, ed. Olwen Brogan and D. J.
Smith (Rome, 1984), pp. 308 – 313.
24 R. Prescott-Allen and C. Prescott-Allen, “How Many Plants Feed the World,” Con-
servation Biology 4, no. 4 (1990): 368.
Decker: Plants and Progress 195
wheat is planted on more acres, rice is without peer as a subsistence
crop. It is the staple food crop for more than half of the world’s 6.6 bil-
lion inhabitants.25 Although not as nutritious as wheat, rice is a major
provider of carbohydrates and thus offers ready energy. Like wheat, the
high yield and disease resistance of rice make it an attractive staple
grain. Similarly, ample evidence shows that rice was a crop of some
importance in the ancient Mediterranean and Mesopotamia before the
Islamic conquests. Rice originated in China and spread south and east
several millennia before the ancient Greeks learned of it, around the
time of Alexander the Great’s conquests. Hellenistic writers accurately
described the plant as growing in fl ooded paddies, and the fourth-cen-
tury b.c.e. Greek botanical writer Theophrastus knew rice as a prolifi c
producer.26 By the fi rst century b.c.e., the Greek writer Diodorus Sicu-
lus viewed rice as primarily an Indian crop that formed an important
part of the intensive agricultural year on the subcontinent. His con-
temporary Strabo knew rice through his source Aristobulus as a crop
that grew in the eastern regions of Bactriana, Babylon, and Susis, but
also in Lower Syria. When seventh-century Muslim armies fi rst reached
Basra in southern Iraq, they saw rice for the fi rst time where it had been
established for centuries in the marshy lands of southern Mesopota-
mia.27 The Babylonian Talmud (redacted ca. 500 c.e. but containing
earlier material) records that rice ranked behind only barley and wheat
as a staple grain among the ancient Jewish communities of Mesopo-
tamia.28 These documentary data thus indicate that rice had become
signifi cant for Jewish farmers under Sasanian Persian political control
(third through seventh centuries c.e.).
By the second century c.e., the crop was established in Roman
Egypt: it is mentioned as growing there by the rhetorician Julius Pol-
lux of Naucratis, and it is noted by the Alexandrian author Athenaeus
(late second/early third century c.e.), who drew much of his material
from earlier sources. Several papyri also preserve record of rice produc-
tion and trade throughout the early Roman period. By the third cen-
tury c.e. at the latest, rice was a minor but well-known plant exploited
25 FAO World Rice Situation, http://www.fao.org/docrep/004/Y0906T/y0906t01.htm
(accessed 6 August 2008).
26 Theophrastus, Historia de Plantis, 4.4.10.
27 M. Canard, “La riz dans le Proche-Orient aux premiers siècles de l’Islam,” Arabica 6
(1959): 113–131.
28 J. Newman, The Agricultural Life of the Jews in Babylonia between the Years 200 C.E. and
500 C.E. (London, 1932), pp. 91– 93.
196 journal of world history, june 2009
in the Nile Valley.29 Rice may have arrived in Egypt via Mesopotamia
and through the fertile corridor of the Palestinian coast, where the crop
was certainly an important part of Jewish agricultural life. The Jerusa-
lem Talmud (redacted ca. 400 c.e.) records centers of rice production
near Antioch in northern Syria as well as around the upper waters of
the Jordan by Lake Tiberias, at Banias (Paneas/Caesarea Philippi) in
the Lower Golan, and on the coast at Caesarea Maritima in northern
Israel.30 The rabbinic literature is further supported by a sixth-century
mention of rice among the local crops in a Hebrew inscription from a
synagogue at Rehob, near Bet She’an (Scythopolis) in northern Israel.
The grain was thus a common local crop around the well-watered
regions of Bet She’an and Banias in the Golan.31
Although rice may have arrived in Egypt via the Fertile Crescent, it
may well have come to the Nile Valley from India. By the Hellenistic
period, there were direct contacts between the Red Sea ports of the
Egyptian coast and the Indian subcontinent. By the second century c.e.
rice was a trade item for Roman merchants plying the Red Sea route
that led down the African coast to Arabia and India. The Periplus Maris
Erythraei, a fi rst-century guidebook for merchants traveling the Red
Sea and Indian Ocean, mentions that the grain was found along the
coast of northwestern India.32 Rice recovered in excavation of Roman-
period levels at Berenike, on the southern coast of Egypt, confi rms that
it continued to be a traded commodity into the late Roman period.
The early Roman culinary writer Apicius included a recipe that used
the water in which rice has been cooked, and in the second century
Galen (and later his fourth-century follower Oribasius) prescribed the
grain as a stomach medicine.33
In the fourth century, rice was a crop of commercial signifi cance
as far west as the city of Ephesus on the coast of Asia Minor (mod-
ern Turkey).34 By this time, the grain was apparently well established
29 Heinrich Konen, “Reis im Imperium Romanum: Bemerkungen zu seinem Anbau und
seiner Stellung als Bedarfs- und Handelsartikel in der Römischen Kaiserzeit,” Münstersche
Beiträge zur antiken Handelsgeschichte 18 (1999): 29–35.
30 Z. Safrai, The Economy of Roman Palestine (London: Routledge, 1994), pp. 117–118.
31 J. Sussman, “The Rehob Inscription: A Translation,” in Ancient Synagogues Revealed,
ed. L. I. Levine (Jerusalem: Israel Exploration Society, 1981), pp. 152–154. I am grateful to
Dr. David Milson for bringing this reference to my attention.
32 Lionel Casson, ed. and trans., The Periplus Maris Erythraei (Princeton, N.J.: Princ-
eton University Press, 1989), p. 76.
33 Apicius, ii.2.8, ed. and trans., B. Flower and E. Rosenbaum (London, 1958); and
Grant, Dieting, p. 53.
34 Expositio totius mundi et gentium, ed. J. Rougé (Paris, 1966), XLVII.8.
Decker: Plants and Progress 197
throughout the Mediterranean world: the Edict on Maximum Prices of
the emperor Diocletian (c.e. 284–305) that regulated maximum prices
empire-wide set the price for cleaned rice at 200 denari per modius.35
The inclusion of the grain in an edict issued empire-wide can only
indicate that it was a commodity commonly traded in the marketplaces
throughout the Roman world.
Although the advent of Pax Romana and the intensifi ed trade that
accompanied it led to the wide diffusion of Asiatic rice, the decline
of the Roman world did not spell the end of its production. In the
fi fth-century West we fi nd the grain being traded as a foodstuff in post-
Roman Gaul.36 Similarly, the late Roman physician Cassius Felix, who
wrote at Carthage in 447 c.e., recommended its use as a treatment for
headache and dysentery, indicating perhaps that it was grown locally in
marshy areas around Africa Proconsularis.37 The continued production
of rice and its general use, particularly in medicine, is attested in later
Greek writers such as the physicians Aetius of Amida (fl . 530 –560 c.e.)
and Alexander of Tralles (ca. 525–605 c.e.).38 Rice is mentioned in the
Anazarbus Tariff inscription (sixth century c.e.) that spelled out cus-
toms duties on goods entering a city that lay in the well-watered plains
of Cilicia (modern southeastern Turkey), where one would expect such
a crop to fl ourish.
Cotton
Cotton is a tropical plant that originated in southern Africa. In prehis-
tory, early cotton varieties were spread by people traveling the Indian
Ocean routes to the subcontinent. According to fi nds from the Indus
valley the plant was integrated into agriculture, probably by the fourth
millennium. From its wild forms, two major species of cotton devel-
oped: Gossypium arboreum L. and G. herbaceum L. Both are perennial
shrubs that typically grow about two meters tall and yield a fi ber-cov-
ered seed whose utility was recognized by early humans. The fi rst major
type of Old World cotton, G. arboreum L., sometimes referred to as
35 1.23, ed. Lauffer, pp. 100–101.
36 M. Grant, ed. and trans., Anthimus, On the Observance of Foods (Blackawton, United
Kingdom, 1996).
37 Cassius Felix, De la Médecine, ed. and trans. A. Fraisse (Paris, 2002), i.11; xlviii.17.
38 Aetius of Amida, Aetii Amideni Libri Medicinales i–viii, ed. A. Olivieri, Corpus Medi-
corum Graecorum (Leipzig, 1935–1950) i.116–117 (bk. 1.298); Alexander of Tralles, ed.
Theodore Puschmann (Berlin, 1878–1879), ii.61.19; ii.403.14.
198 journal of world history, june 2009
“tree cotton,” is found mostly in Asia, with limited distribution in
eastern Africa and Arabia. The second, G. herbaceum L., belonged to
Africa and Arabia. G. herbaceum L. became more widely diffused and
agriculturally developed, with many new varieties resulting from culti-
vation. The most important change was the shift in growth cycle with
varieties of G. herbaceum L. altered from perennial to annual cultivars,
which adapted them to the cooler environments found in Mesopota-
mia, Egypt, and India. This evolution occurred at a later stage than
the initial diffusion of G. arboreum, which had also entered cultivation
and continued to be grown in Africa.39 It is possible that cotton was
introduced to the Roman world from the empire’s eastern neighbors
and Africa via Egypt.
Most of the evidence for cotton in the pre-Islamic Mediterranean
world has recently been collected by John Peter Wild, but a brief survey
of the data is in order.40 Indian cotton itself was imported into Neo-
Assyrian Mesopotamia, as confi rmed by fi nds from Iraq.41 The Greek
historian Herodotus (484–425 b.c.e.) knew of the production of cotton
from trees in northern India and noted that soldiers of the Persian king
Xerxes (485–465 b.c.e.) wore cotton clothing.42 During Alexander the
Great’s naval expedition in the Indian Ocean and Persian Gulf the
Macedonians encountered cotton, and by the fourth century b.c.e., at
the latest, the plant was grown along the Gulf, where Theophrastus
noted its presence. Cotton production seemingly endured there for
centuries, as Pliny wrote that cotton was worked into fi ne cloth on
the island of Tylos (Bahrain).43 By the time the plant had reached the
Persian Gulf, it was likely also growing in Mesopotamia and elsewhere
in the Hellenistic world.
During Late Antiquity (fourth through seventh centuries c.e.), cot-
ton had become an important crop beyond the Araxes River in present-
day Azerbaijan, which belonged to the Sasanian Persian empire. Far-
ther east, excavations in the oasis of Merv (in modern Turkmenistan)
has yielded fi nds of cultivated cotton from the Sasanian period (third
39 C. L. Brubaker, F. M. Bourland, and J. F. Wendel, “The Origin and Domestication of
Cotton,” in Cotton: Origin, History, Technology, and Production, ed. C. W. Smith and J. T.
Cothren (New York: Wiley, 1999), pp. 20–21.
40 Watson, Agricultural Innovation, pp. 34–35; J.-P. Wild, “Cotton in Roman Egypt:
Some Problems of Origin,” Al-Räfi dän 18 (1997): 287–298.
41 F. Hideo, K. Sakamoto, and M. Ichihashi, “Textiles from at-Tar Caves, Part II(4)
Cave 16, Hill C,” Al-Räfi dän 17 (1996): 160.
42 Herodotus, History iii.106; vii.65.
43 Theophrastus, Historia de Plantis IV.7.7–8; Pliny, Natural History, XII, 21.
Decker: Plants and Progress 199
through seventh centuries c.e.).44 The Babylonian Talmud (compiled
ca. 500 c.e.) mentions cotton-seed oil used by the Jewish communities
of Persian Mesopotamia, cotton dealers, and cotton wool.45
Cotton textiles from the fi rst through seventh centuries c.e. are
found in an arc from the Sudan to Syria. In the Sudan, Nubia, and
Egypt, cloth fragments have been recovered at Meroë, Karanog, and
Qasr Ibrim in Nubia, at Doush in the Kharga Oasis in the Western
Desert of Egypt, and at Berenike and Quseir al-Qadim on the Red Sea
Coast.46 At Karanis, in the Fayyum of central Egypt, a skein of cot-
ton thread, dyed red, and dated to the fourth or fi fth century c.e., was
recovered.47
Cotton textiles are manufactured with both Z-spun (right-hand
spin) yarns and S-spun (left-hand spin) yarns. The distinction is an
important one, since weavers working in the traditions of Egypt and
the Levant normally worked in S-spun yarns.48 The majority of cot-
ton fi nds at Berenike were Z-spun cotton yarns, indicating they were
imports from India; only a small number of the Berenike fi nds were
S-spun. In Palestine, S-spun Roman-Byzantine period cotton textiles
have been found at ‘En Boqeq, 15 kilometers south of the Dead Sea,
and at Nessana in the Negev. At Halabiyah (ancient Zenobia) on the
middle Euphrates in Syria, an S-spun tunic belonging to the sixth cen-
tury was found in excavation, while farther downriver, in the Roman
city of Dura Europus, S-spun cotton was woven into a wool textile.49
44 Robert Hewsen, ed. and trans., The Geography of Ananias of Širak (AŠXARHAC’OYC’):
The Long and the Short Recensions (Wiesbaden, 1992), 65A; Georgina Herrmann, K. Kurban-
sakhatov, and St. John Simpson, “The International Merv Project: Preliminary Report on
the Fourth Season (1995)” Iran 34 (1996): 20; and Georgina Herrmann, K. Kurbansakha-
tov, St. John Simpson et al., “The International Merv Project: Preliminary Report on the
Fifth Season (1996),” Iran 35 (1997): 9.
45 Babylonian Talmud, ed. I. Epstein (London, 1935–1948), Megillah, 105; Sanhedrin,
463; ‘Abodah Zarah, 140; Newman, Agricultural Life, p. 103.
46 Meroë, Karanog, Qasr Ibrim, Quseir al-Qadim: Watson, Agricultural Innovation, p.
34; Doush: Fr. Dunand, J.-L. Heim, N. Henein, and R. Lichtenberg, La nécropole de Douch:
Exploration archéologique II: Monographie des tombes 73 à 92: Structures sociales, économiques,
religieuses de l’Égypte romaine (Cairo, 1992), p. 232; Berenike: J. P. Wild and C. F. Wild, “The
Textiles,” in Sidebotham and Wendrich, Berenike, pp. 245–256.
47 L. M. Wilson, Ancient Textiles from Egypt in the University of Michigan Collection (Nor-
wood, Mass., 1933), no. 133, p. 50.
48 Wild and Wild, “Textiles,” p. 246.
49 Berenike: Wild and Wild, “Textiles,” pp. 251–253; ‘En Boqeq: A. Sheffer and
A. Tidhar, “The Textiles from the ‘En-Boqeq Excavation in Israel,” Textile History 22 (1991):
22–23; Nessana: L. Bellinger, “Textiles,” in Excavations at Nessana I, ed. H. D. Colt (Lon-
don, 1962), 99, nos. 26, 27; Halabiyah: M. Nockert, “Vid Sidenvägens ände. Textilier från
Palmyra till Birka,” in Palmyra: Öknens Drottning, ed. P. Hellström (Stockholm), pp. 81–82,
91–92; and R. Pfi ster, Textiles de Halabiyeh (Zenobia) (Paris, 1951), p. 55 and pl. iii no. 5.
200 journal of world history, june 2009
The sizeable majority of the cotton textiles discovered in the
Roman-Byzantine east thus far are of Z-spun cotton and probably rep-
resent imports from Nubia, India, or the Sasanian empire. The trade in
Indian cottons was frequent and important at least by the fi rst century
c.e., according to the Periplus Maris Erythraei, which shows that cotton
was a common item of exchange in this long-distance trade, evidence
supported in the papyri.50 But the S-spun cotton yarns used in the tex-
tiles noted above were locally produced.
That cotton cloth garments were made in Roman Egypt fi nds some
support in a letter of the second century c.e. from the middle Nile Valley
city of Karanis in which a certain Julius Apollinarius requested white
cottons from his brother, while a second-century letter from Oxyrhyn-
chus relates that the mother of the addressee had made a cotton tunic
for him.51 A contemporary Egyptian papyrus document of unknown
provenance recorded a request for cotton thread so that work garments
could be made.52 One additional scrap of information indicating local
cotton weaving comes from Doush (ancient Kysis) in the Kharga Oasis
in the Western Desert of Egypt, where ostraca of the fourth /fi fth cen-
tury record weights of cotton supplied to fi ve women, who apparently
worked it into yarn or fi nished textiles.53 Local manufacture of cotton
garments does not, of course, prove with certainty that the fi ber itself
was produced there, but textual data indicate that the crop was, in fact,
grown in Egypt.
In the fi rst century c.e. Pliny noted that cotton grew in Upper
Egypt, and in the following century the Egyptian rhetorician Pollux
described cotton (a “tree”) and that its thread was mixed with linen in
cloth production.54 A second-century papyrus (P. Iand. VII.142) from
the Kharga Oasis preserves the record of cotton planting there, as does
50 Casson, Periplus, pp. 6, 41; H. Harrauer, P. Sijpesteijn, “Ein neues Dokument zu
Roms Indienhandel, P. Vindob. G 40822,” Anzeiger der Österreichischen Akademie der Wis-
senschaften, phil.-hist. Kl. 122 (1985): 124–155.
51 PMich VIII.500.7 = H. C. Youtie and J. G. Winter, eds., Papyri and Ostraca from
Kara nis Volume VIII (Norwood, Mass., 1951), pp. 117–118; also F. Preisigke, F. Bilabel,
and E. Kiessling, Sammelbuch Grieschischer Urkunden aus Agypten (hereafter SB), 26 vol-
umes 6.9025 (Wiesbaden, 1963), 38; P. Oxy LIX, 3991 = E. W. Handley, H. G. Ioannidou,
P. J. Parsons, J. E. G. Whitehorne, The Oxyrhynchus Papyri vol. 59 (London, 1992), pp.
127–128.
52 SB vi.9026; J. G. Winter and H. C. Youtie, “Cotton in Graeco-Roman Egypt,” Amer-
ican Journal of Philology 65 (1944): 256 –258.
53 H. Cuvigny and G. Wagner, eds., Les Ostraca Grecs de Douch (O. Douch) (hereafter
O. Douch) (Cairo, 1986), p. 51.
54 Natural History xix.14; Pollux, Onomasticon, ed. E. Bethe (Leipzig, 1967), ii.vii.75.
Decker: Plants and Progress 201
a fourth-century ostracon (O. Douch 51).55 An undated ostracon from
Doush (O. Douch 537 R5) documents an account recording that wine
and cotton were local products. O. Douch 634 5– 6, belonging to the
fourth century, is a letter demanding payment in cotton, and O. Douch
381 required payment of the annona tax in cotton. Farther north, in the
oasis of Dakleh, two ostraca from Kellis (O. Kellis 68, O. Kellis 69), the
second dated 276/277 c.e., record shares of cotton received from indi-
vidual producers.56 A fourth-century papyrus also from Kellis (P. Kell
I Gr. 61) details payment owed in cotton. Farm records of the fourth
century (P. Kell. IV Gr. 96), along with cotton seeds recovered in exca-
vation there, prove that cotton was a locally grown commodity in the
late Roman period of Dakleh Oasis.57
Cotton was also cultivated in Palestine, according to the evidence
of the Mishnah (redacted ca. 200 c.e.), in a passage that is repeated in
the Jerusalem Talmud (redacted ca. 400 c.e.).58 Gregory of Tours (d.
594) wrote that in his day, trees that produced “wool” grew around
Jericho.59
Artichoke
While it is not possible in this brief overview to detail the evidence
of the diffusion of all the crops proposed as Islamic introductions by
Watson, one minor crop is particularly noteworthy because a fair body
of evidence suggests its long-standing production in the Mediterranean
world prior to the arrival of the Muslims. The artichoke (Cynara scoly-
mus) is argued to have been a late medieval arrival in Europe after
centuries of use in the Islamic world. Watson claims that “There is no
reference in Classical literature to a plant of this family with edible
fl esh on the bracts,” allowing only that the probable ancestor of the
artichoke, the cardoon (Cynara cardunculus), was known in the Greco-
55 P. Iand. = C. Kalbfl eisch, Papyri Iandanae (Leipzig-Berlin, 1912–1938), p. 323.
56 O. Kellis = K. A. Worp, ed., Greek Ostraka from Kellis: O. Kellis, nos. 1–293 (Oxford,
2004).
57 R. Bagnall, The Kellis Agricultural Account Book (P. Kell. IV Gr. 96) (Oxbow, 1997),
pp. 39–40, 114, 122, 164.
58 Mishnah Kilayim VII.2, trans. I. Mandelbaum, in The Mishnah: A New Translation,
ed. J. Neusner (New Haven, Conn., 1988); Jerusalem Talmud Kilayim: I. Mandlebaum,
trans., The Talmud of the Land of Israel: A Preliminary Translation and Explanation, 35 vols.
(Chicago, 1990), 4.VII.2.
59 Gregory of Tours, De Gloria beatorum martyrum, ed. J. P. Migne, Patrologia Latina
(1844–1849), pp. 71, 721.
202 journal of world history, june 2009
Roman world.60 The cardoon is today considered a weedy species and
is highly invasive, particularly in dry landscapes. The artichoke, by
contrast, requires regular watering and fertilizing, which fi ts more with
the evidence provided by Pliny discussed below. Morphologically, the
cardoon and artichoke share many similarities: both grow to a height
of about 1.5 meters, have spiny leaves, and fl ourish in dry lands. While
both the cardoon and the artichoke produce a fl ower bud, the artichoke
bud is more developed and provides the fl eshy edible hearts. However,
the fl ower of the cardoon is not eaten, but its stem is consumed.
The major basis of Watson’s argument regarding the artichoke is
linguistic, as he contends that “all the European languages” derived
their names for the plant from the Arabic. The Arabic names used for
the plant, kharshuf, and its related forms possibly derive from the Latin
name carduus, which was used to designate the artichoke. Another
common Arabic name for the artichoke, qinärïya, is clearly derived from
the Greek word for artichoke, kinara, a name that Watson knows, but
disregards. By the fi rst century c.e., the cardus (carduus) plant was grow-
ing in North Africa and Spain. Pliny noted that Carthage and Cordoba
were major suppliers and their artichokes fetched exceptional prices
(6,000 sesterces from small plots).61 The cardus that Pliny discussed is a
plant that required fertilizing, something that would be unnecessary for
the hardier cardoon. While this is far from conclusive, the description
provided by Galen is more helpful. The physician describes the kinara,
which he calls “over-valued,” and notes that people eat the heads,
“which they call whorls” (sphonduloi).62 Since the heads of the cardoon
are not edible, Galen has to be speaking here of the artichoke.
The agricultural handbooks from the fi rst through sixth centuries
suggest that the artichoke was in general cultivation throughout the
Mediterranean world during the whole of the Roman imperial period.
Columella advised that artichoke seeds be sown in spring and autumn,
similarly advised in the Geoponica, while Palladius recommended that
artichokes be planted in the garden, an area typically reserved for
higher value and more intensively worked plants.63 This also implies
60 Watson, Agricultural Innovation, pp. 64–65.
61 An unlabeled Roman North African mosaic in the Bardo Museum, among other
examples I have seen there, clearly depicts an artichoke.
62 O. Powell, trans., Galen on the Properties of Foodstuffs (De alimentorum facultatibus)
(Cambridge, 2003), pp. 104–105, 178n.637.
63 Columella, De Re Rustica X.235; X.iii.14; XI.iii.28; Palladius, Opus Agriculturae,
III.24, IV.8, VI.5, XI.11.
Decker: Plants and Progress 203
that the cardus / kinara of the ancient authors should generally be taken
to be the artichoke, and not the cardoon.
Kai Ruffi ng has collected the evidence for the production of arti-
chokes in Roman Egypt. There artichokes were grown in the midst of
vineyards, probably because many vines were irrigated, and the kinara
also needed signifi cant quantities of water.64 That the vegetable was
widely available throughout Egypt is apparent from the accounts of
the fourth-century c.e. traveller Theophanes, who purchased them at
Babylon (modern Cairo).65 It is apparently with Egypt that the Mus-
lims of Spain most identifi ed the plant: the Andalusian Muslim agri-
cultural writer, Ibn al-‘Awwäm noted that the qinärïya came from Egypt
to Spain, where he wrote in the twelfth century c.e.66 The Edict on
Maximum Prices of Diocletian provides proof of the artichoke as a com-
mon food in the cities of the Roman provinces. The law fi xed the price
of large artichokes at fi ve for ten denarii and the heads (sphonduloi) of
ten for six denarii. Although the list is unreliable as a real indicator of
price, it can nevertheless be used to gauge comparative price levels. For
instance, a sextarius (about half a liter) of ordinary wine was fi xed at
eight denarii. This suggests that artichokes were not cheap, but neither
were they particularly expensive. Once more, the considerable quantity
of textual references and the plant’s inclusion in the Price Edict indicate
that the plant was widely cultivated, known, and consumed.67
Conclusion
Although the archaeological and textual evidence for the crops exam-
ined above are scattered and incomplete, their presence among the
extremely limited body of materials that touches on ancient farming
is signifi cant. This material shows that durum wheat, rice, and cotton
were crops of varying importance over much of the Roman and Per-
sian worlds. Durum wheat was spread over the whole of the Mediter-
ranean by Roman farmers throughout the centuries of their imperial
64 K. Ruffi ng, “KINAPA: Anbau und Vertrieb im römischen Ägypten,” Münstersche
Beiträge zur Antiken Handelsgeschichte 14 (1995): 61–70.
65 Pap. Rylands = Papyrus Rylands, ed. C. H. Roberts and E. G. Turner (Manchester,
1952), 627.85, 627.149, 627.197.
66 Ibn al-‘Awwäm, Kitab al-fi lähah. Translated as Le livre de l’agriculture, trans. J.-J. Clé-
ment-Mullet, 2nd ed. (Tunis, 1977), ii.291.
67 Ruffi ng, KINAPA, 65; Ed.Diocl. VI.1–2; II.1; IV.2.
204 journal of world history, june 2009
rule. This grain was a staple in the Roman diet from Syria to Italy and
North Africa, where it remains fundamental to the diet today. Watson’s
assertion that the Muslims had any appreciable role in expanding the
cultivation of Triticum durum in the Mediterranean has no basis in the
evidence, but is an example of an uncritical method that ignores the
available data.
Rice cultivation in Mesopotamia was long-established before Islam,
and its transmission eastward was accelerated by the Greek interest in
exotics during the Hellenistic era. This westward movement continued
during the period of their Parthian and Sasanian Persian successors.
During late antiquity, rice was well established in western Asia and
had spread into Mediterranean Europe. The importance and general
availability of rice from Syria to Gaul is attested by numerous texts and
underscored by its regulation in the fourth-century Edict on Maximum
Prices.
If Watson’s thesis were correct, we would surely expect to fi nd rice
production widespread in Mesopotamia, the heartland of the classical
Islamic empire and the home of the great rivers Tigris and Euphrates.
This region was close to the original points of diffusion proposed for
rice by Watson. The lands of the Tigris and Euphrates also met the
plant’s environmental criteria: there was fertile earth, abundant water,
and extensive irrigation, and a large population of agriculturists who
had grown the plant for centuries. Yet even in the hinterland of the
greatest city in the Islamic world, Baghdad, we fail to fi nd rice in abun-
dance in the medieval period. In fact, among the evidence from the
Abbasid period (eighth through thirteenth centuries c.e.), the peak
of classical Islamic civilizations, barley and wheat dominated the diet,
as they had for millennia previously.68 Rice maintained a similar place
in Iraq at the height of the “Green Revolution” as it had under the
Sasanians: wheat and barley only are mentioned as taxed crops (and
therefore clear staples) in the tax list of Qudäma ibn Ja‘far of the tenth
century c.e.69 In late medieval Turkey, the Pontus and Cilicia were the
only two areas known to have grown the crop, a situation that corre-
sponds more closely to conditions of the fourth century than one would
expect had the Arab conquests and subsequent Muslim takeover of
68 Demonstrated by wheat and barley being collected as tax in Islamic Mesopotamia:
R. McC. Adam, Land Behind Baghdad (Chicago, 1965), 101, tbl. 21.
69 H. el-Sämaräie, Agriculture in Iraq During the 3rd Century A.H. (Beirut, 1972), pp.
99 –103.
Decker: Plants and Progress 205
Anatolia propelled its movement.70 Though the advent of rice cultiva-
tion is heralded as a major Muslim advancement in Spain, its farthest
westward expansion under Islam, cultivation there was likewise of sec-
ondary importance and lagged far behind the old staples of wheat and
barley in the agricultural regime.71
Both the Romans and Persians alike had a signifi cant share in the
expansion of the production of cotton. As demonstrated above, Old
World cotton was of economic signifi cance before the rise of Islam,
especially in Persia, where the evidence from cloth production, archae-
ological recovery of cotton, and textual sources indicate disparate
places of considerable production and use. In the Roman world, cot-
ton cropping and use in cloth making appears to have been relatively
modest and confi ned to Egypt.
Despite Watson’s claim, cotton production in Egypt remained
underdeveloped during classical Islam: the major cloth plant remained
fl ax, as it had been for centuries. The picture of an early Islamic land-
scape where cotton was restricted mainly to Persia and of only minor
importance in Egypt is known by Watson, who quotes the eleventh-
century writer al-Tha‘älabï—“people know that cotton belongs to Khu-
rasän [the region of Merv] and linen to Egypt”—but explains this away
on the strength of anecdotal evidence.72 In the eleventh century, after
a full four centuries of Arab Islamic rule, Egypt remained proverbial
for its linen, a product for which it was famed in classical antiquity.73
The earliest clear evidence cited by Watson that supports widespread
Egyptian cotton growing belongs to the thirteenth century, six centu-
ries after the Muslim conquest and two centuries after the end of the
so-called Green Revolution.74
The case of the artichoke provides an example of a crop that has
remained of scant global signifi cance since its introduction. Again,
Watson is mistaken in his presumption that the plant’s fi rst introduc-
tion in the Mediterranean belonged to the Islamic era. Roman cultiva-
tion is widely attested, and the role of the plant in the Mediterranean
and European diet, albeit minor, is embedded in the Roman, not the
Islamic, past.
70 Watson, Agricultural Innovation, map 2.
71 E. García Sánchez, “Agriculture in Muslim Spain,” in The Legacy of Muslim Spain, ed.
S. K. Jayyusi (New York, 1992), p. 994.
72 Watson, Agricultural Innovation, p. 40.
73 P. Mayerson, “The Role of Flax in Roman and Fatimid Egypt,” Journal of Near Eastern
Studies 56, no. 3 (1997): 201–207.
74 Watson, Agricultural Innovation.
206 journal of world history, june 2009
Though Watson is certainly correct to link crops to agricultural
methods, and link these methods to wider cultural changes, the chro-
nology and scope of his proposed changes are problematic. For decades,
in large part based on Watson’s provocative work, scholars have uncrit-
ically pointed to the Islamic Green Revolution as one of the major gifts
the Islamic world bequeathed to Europe and thence to much of the
world. The contributions of the medieval Islamic agriculturists are cer-
tainly impressive. But a growing body of evidence for pre-Islamic dif-
fusion of key agrarian techniques, tools, and crops challenges the basic
assumption of rapid and deep changes in Muslim agricultural practice.
Rather, it appears that the pre- and post-Islamic Middle East and Medi-
terranean landscape were far more similar to one another than is often
recognized. Islamic farming structures were built atop earlier Roman
and Persian landscapes: these were usurped rather than swept away.
