The Tropical Turn: Agricultural Innovation in the Ancient Middle East and the Mediterranean

The following review offers a survey of Sureshkumar Muthukumaran’s monograph The Tropical Turn: Agricultural Innovation in the Ancient Middle East and the Mediterranean.1 Since the reviewer is a cuneiformist, the review will concentrate on the aspects relevant to the ancient Near East, add some of the reviewer’s own reflections, and refer to further literature on the subject.Muthukumaran is interested in the anthropogenic long-distance dispersal of cultivated plants in premodern Afro-Eurasia from the mid-fourth millennium BCE to the early centuries CE. The monograph under review focuses on seven cultivated plants from tropical-subtropical South Asia and the question of when and by which routes they reached Southwest Asia and the Mediterranean region: rice, cotton, citrus fruits, eastern cucurbits, the lotus plant, taro, and the sissoo tree. Each plant identification is based on an interdisciplinary synthesis of previous archaeobotanical, archaeological, ethnobotanical, and historical-philological studies.The emphasis is on philological-ethnographic comparisons and very often relies on phonological and semantic similarities to identify ancient plants, as in the case of Sanskrit kārpāsa, Prākrit kappāsa, and (neo-)Babylonian karpassu (76), Hebrew karpas, or Greek kárpasos (“cotton,” 84, 91); Persian gurinǧ and Akkadian kurangu (“rice,” 106–7); and Indo-Iranian *vrīziš and Neo-Elamite miriziš (“rice,” 107–8). The individual sections of the book under review are discussed below.1. Historical context (15–63). The aim of this chapter is to sketch, as accurately as necessary and as briefly as possible, the history of the ancient long-distance networks linking the Mediterranean, Southwest Asia, and South Asia from the fourth millennium BCE to the late centuries BCE. Muthukumaran outlines the routes and the modes of transport (long-distance versus short-distance exchange, pack animals such as the donkey, the zebu, and the Bactrian camel versus boats and ships), as well as the goods transported, especially valuable and prestigious goods such as lapis lazuli or rare woods. It is plausible that the movement of food crops, such as rice, often occurred when traders carried them as food for their own consumption. Regarding the discussion of land routes (16–17), the reviewer would have liked Muthukumaran to refer to previous research on the exchange of crops along an alternative route across the Inner Asian Mountain Corridor in Eurasia (see, for example, Frachetti 2012 and Miller, Spengler, and Frachetti 2016: 1568–69, in the context of millet migration).Muthukumaran focuses on water transport, particularly via the Persian/Arabian Gulf and the Arabian Sea, between southern Babylonia, Magan (Oman), and Meluḫḫa (Indus Valley/Gujarat), especially from the late third to the mid-second millennium BCE (22–26). In this context, the reviewer would also mention some further work on the role of Dilmun (Bahrain), Magan (Oman), and the Arabian Sea in the context of the exchange of goods and crops (e.g., Heimpel 1987, 1993; Hammer et al. 2009; and Dornauer 2018: 79–84). Muthukumaran does include a discussion of the maritime routes between the Mediterranean and South Asia via the Red Sea, especially in the second half of the first millennium BC (60–62).2. Cotton (64–98). The first volume of the RlA contains a lemma on cotton trees, albeit a short one (Ebeling 1932). Somewhat more detailed lemmas are found in RlA 12 (Waetzoldt 2012; Völling 2012). It is therefore gratifying to find that Muthukumaran’s chapter on Old World cotton plants and their spread from South Asia is much more comprehensive and detailed. We learn that sub-Saharan cotton (Gossypium herbaceum L.) and South Asian tree cotton (Gossypium arboreum) were used for textile production, although on a small scale. In the context of G. arboreum, the author also refers to Sanskrit kārpāsa and Prākrit kappāsa, both of which mean “cotton” (71). He explains that South Asian tree cotton had been known as iṣṣū nāš šīpāti (trees bearing wool) in Assyria at least since Tiglatpileser III (744–727 BCE) and cultivated at least since Sennacherib (704–681 BCE). In Babylonia, tree cotton was known as kidinnû and in Hellenistic times by the Sanskrit loanword karpassu (see also Hebrew karpas). While archaeobotanical and philological evidence suggests that cotton was grown and processed in Assyria only for royal-palatial use, the author suggests that cotton was more widespread in Babylonia. This idea is supported by an increasing number of cuneiform inscriptions and by evidence of falling cotton prices and rising demand that apparently led to cotton imports (80–82).According to archaeobotanical evidence, cotton has been known in Iran since at least the first half of the first millennium BCE (83–86). However, it is not until the Achaemenid period that its cultivation and large-scale use in textile production are recorded. Carbonized seed finds from the sixth and fifth centuries BCE and Hellenistic texts also attest to extensive cotton cultivation on Bahrain and Muharraq islands. However, according to Muthukumaran, cotton must have been introduced into Bahrain and possibly eastern Arabia before it appears in Babylonian cuneiform records of the ninth century BCE (86–88). There is slight evidence for the use of cotton in the Levant and Egypt. A cotton textile attributed by Pliny to the Pharaoh Amasis in the sixth century BCE must, according to the author, have been imported from Mesopotamia or the Gulf region (89–90). Cotton was also extremely rare in the classical Greco-Roman cultural area and largely reserved for the elite. It was not grown in the Mediterranean west of the Levant (90–94). At the end of the chapter, the author discusses the appeal and properties of cotton and why it could not prevail against linen and wool (94–98).3. Asiatic rice (99–122). As late as 2005, M. A. Powell wrote in RlA 10 that “[r]ice seems to have come from India after Alexander the Great” (Powell 2005). Three years later, M. Stol corrected this assumption in RlA 11 (Stol 2008). In his chapter on rice, Muthukumaran discusses recent research that shows that Asian rice (Oryza sativa) migrated to Mesopotamia and then to the Mediterranean. We learn that the South Asian subspecies Oryza sativa ssp. aus and indica, having been domesticated along the middle and lower Ganges, migrated to northwestern India between 2430 and 2140 BCE and finally to Harrapan Pirak in the eastern Iranian plateau in the early second millennium BCE (99–103). The author synthesizes the scant archaeological and cuneiform evidence to show that Akkadian kurângu is related to Persian gurinǧ and other Turko-Persian terms for rice. He links these terms to Dravidian words since the “formative suffixes -gu, -ṅgu and -kku are strongly reminiscent of a South Asian, especially Dravidian dialect” (107). Based on these considerations, and since at least two of the rice subspecies were domesticated in India, Muthukumaran concludes that rice came from India to the Iranian plateau and from there to Mesopotamia as a foodstuff for merchants (114–15). Referring to the few—exactly three—nonlexical cuneiform records of kurangu, he shows that rice was already cultivated in Assyria in the eleventh and eighth centuries BCE (116). In the opinion of the reviewer, it should be emphasized that all these texts come from official provincial archives. This fact, together with the scant textual and archaeological evidence, suggests that irrigation-intensive rice cultivation in Assyria was exclusively a state enterprise, perhaps even in the form of agricultural experiments.Given the paucity of written evidence, even in the more recent Babylonian and Achaemenid periods (sixth–fourth centuries BCE), Muthukumaran suggests that rice, as a newcomer to Mesopotamia, was not taxed and would not have appeared in administrative records. He refers to medieval Italy, where rice was grown by the poor in low-lying marshes unsuitable for wheat, and its yields rarely appeared in tax records (117–18).The earliest evidence for Neo-Elamite miriziš, “rice,” a cognate of Indo-Iranian *vrīziš, “rice,” is associated with Achaemenid rice cultivation in the fifth century BCE (107–8), while the earliest archaeobotanical evidence from Persia dates between 25 BCE and 250 CE (114). The philological and archaeobotanical evidence suggests an increase in rice consumption in the late first millennium BCE, including in Rome starting in the first century BCE (114). Muthukumaran argues that its consumers originally were South Asian merchants who took rice with them as food for the overseas trade between South Asia and the Mediterranean via the Red Sea. The importance of first-millennium BCE overseas transport is also reflected by the straightforward Indic loanwords for rice in Elamite, West Semitic, Greek, and Latin (115).Wondering why there was no large-scale cultivation of nutritious, heat- and pest-resistant Asian rice in ancient Mesopotamia, Muthukumaran concludes that rice could only be grown on soaked, damp soils, such as natural marshland in southern Babylonia, or near large irrigation systems. But irrigated land was used almost exclusively for winter-grown barley, which needs less water than rice (118–19), and, as the reviewer would like to stress, there were other endemic crops such as the textile-crop flax and the more nutritious wheat occupying irrigated land. In addition, Mesopotamian farmers were practicing a pure fallow system (Dornauer 2016: 65–66) with no multicropping. Thus, a field cultivated in the winter season remained fallow in the summer and the following winter season. As a result, rice cultivation would have been limited to the marshlands and the alluvial plain next to the river. However, the land next to the river was occupied by other summer crops, especially sesame. Muthukumaran also notes that while rice migrated to Southwest Asia and the Mediterranean, typical South Asian boiling techniques did not (120). At this point, the reviewer would like to mention that boiling was rather unusual in Mesopotamia, due to the relative scarcity of fuel. This may have been another reason why rice did not catch on in ancient Mesopotamia.4. Citruses (123–44). The authors of Reallexikon der Assyriologie (RlA) 10 deny the existence of citrus fruits in ancient Mesopotamia and claim that the citron migrated to the ancient Near East only in the Hellenistic period (Powell 2005; Nesbitt 2005; and Frantz-Szabo 2005). Muthukumaran tries to convince us otherwise. He begins by introducing his readers to the complex taxonomy of citrus fruits, particularly citron (Citrus medica) and lemon (Citrus × limon) in early South Asia (123–27). The earliest finds of citrus, presumably citron, date from the last quarter of the third- and early second-millennium BCE Harappan sites. By the mid-second millennium, citrus fruits, presumably citron, dispersed throughout the Indian subcontinent (124–27). The author lists eleven Sanskrit terms for citrus fruits reflecting their cultural importance (127). There is no archaeobotanical evidence of citrus from Iran until the late Sasanian period and from Mesopotamia apart from a dubious and not seriously datable find of citron seeds from Nippur (142–43). Muthukumaran, however, attributes this to the chance nature of archaeological finds and follows Emanuel Bonavia’s 1894 suggestion that the famous Assyrian reliefs depicting a cone-like object held by a winged protective spirit standing in front of a tree illustrates a citron fruit. However, this idea was later rejected by Bonavia himself (Franklin 2021: 79), and recent research instead identifies the tree depicted on the relief as a date palm and the cone-like fruit as the male inflorescence of the date palm (Brown 2008; Ziffer 2019; Franklin 2021).Outside South Asia, the oldest archaeobotanical evidence for citrus fruits besides the dubious finds from Nippur comes from 1200 BCE at Hala Sultan Tekke in southeastern Cyprus, though these have not been directly dated to confirm their age (127). A completely preserved desiccated citron, dating between 750–480 BCE, also comes from Cyprus. The earliest archaeobotanical evidence for citron cultivation comes from the seventh-century BCE Phoenician colony of Kyme in the Bay of Naples, the mid-fourth century BCE Phoenician colony of Carthage, and from the fifth- or fourth-century BCE Persian satrapal palace in Jerusalem. Organic citron residues from a wine jug from the sixth-century BCE Phoenician-colony site of Monte Sirai (Sardinia) proves that citron was used for flavoring wine. Based on the chronogeographical distribution of the finds, the author concludes that citron (Citrus medica) was spread throughout the Mediterranean by Greek and Phoenician colonists and traders (127–30).Muthukumaran points out that Greek μῆλον, “apple,” and Roman malum, “apple,” can refer to any roundish fleshy fruit (12, 140) and that early Greek texts refer to citron as μῆλον περσικόν (hereafter melon persikon), “Persian apple.” In Muthukumaran’s opinion, this fact points very clearly to the existence of citrus fruits in Persia, especially since citron also plays a role in the Zoroastrian cult (143–44). However, only one of the cited Greek texts seems to be of any use as evidence. That is Theophrastus’s (fourth–third century BCE) precise description of the characteristics of the citron tree, its flower, and its fruit, including the apt attribution that the fruit is not eaten. Theophrastus tells us that the fruit was known as melon medikon (μῆλον μεδικόν) and melon persikon. It is noteworthy that Theophrastus’s ambiguous nomenclature is unique, as later in the Hellenistic period melon medikon, “Median apple,” seems to have referred exclusively to the citron, while melon persikon, “Persian apple,” referred exclusively to the peach (Prunus persica) (Dalby 2003: 88, 252; see also Schmitt 2002). Even in the archaeobotanical literature, melon persikon is usually identified as peach, not citron (e.g., Dal Martello et al. 2023). It would have been of great benefit for the reader had the author addressed this confusing contradiction in the terminology.Muthukumaran rejects older suggestions, based on vague similarities with Hebrew and Sanskrit that the citron could be identified with Akkadian or the and It should be that the or tree was already known in the Old period This would be before or the as the oldest archaeobotanical evidence of citron in South Asia, which is Muthukumaran that Greco-Roman and “apple,” could refer to any and fleshy fruit. He suggests that is in fact the citron fruit, which is reminiscent of the Greek for citron, apple,” The reviewer notes that is already in Old Babylonian Nippur 11 Thus, Muthukumaran’s identification of with citron, it should have been known in since at least the early second millennium BCE. This would be to is known the of the dispersal of the lemon The next is to the eastern of the plant The earliest archaeobotanical finds of from BCE, may be a domesticated form of the earliest archaeobotanical finds from in Mesopotamia third quarter of the third may be Asian from South Asia, presumably one of its such as the melon Frantz-Szabo an identification of with 2005). However, identification with the can be out since it is a Powell in RlA 10 the identification of and Akkadian with to be (Powell Muthukumaran identifies and Akkadian instead with melon as well as Hebrew or Greek refers to the melon Akkadian refers to the melon and to other such as the while and of the should be identified with the Even such as are an BCE the already between and of and In the opinion of the reviewer, be identified with the domesticated form of the identifies Greek with the and as well as with of the however, these terms could also refer to other known melon such as the as the author himself notes Muthukumaran that Akkadian or both terms which have no in be a cognate of Sanskrit or as there are other Sanskrit words that South Asian cucurbits, the reviewer no chance to identify or with a especially as the author himself that are and there is a very taxonomy for these plants even the some of the Greek or terms is a for the of which according to archaeological finds in the Mediterranean in the first or second century CE This is also supported by the philological evidence to the or the of the plant and its fruit Muthukumaran identification of seeds in the century of (see see Nesbitt earliest records are from and The seeds from sixth-century BCE or the fourth-century BCE of on the Sea be melon seeds However, all the Muthukumaran he does not to out the that is already some of the millennium of cucurbits, which are to be with In the next chapter learn the lotus an plant with seeds and that in water or The first archaeological evidence from dates to the early fourth century BCE Since Muthukumaran the evidence from and in the late second millennium and since there is no textual or archaeological evidence from ancient Mesopotamia and Iran it is that lotus has not been found in the Middle East as the reviewer it was directly across the Red Sea to and from there to in the first millennium lotus as a of the to Muthukumaran, this shows that the in the fifth century BCE were of the lotus Since the fourth century BCE, Greek authors lotus as or to Muthukumaran, this is a of its importance as a food in millennium BCE where cultivated lotus in of as by Theophrastus and In the lotus plant seems to have been in the Greco-Roman less as a food than as a and example, its were used to The next with another and very to South and It is known by from the third century BCE and possibly as Greek Latin after the fourth century as Latin and as Greek and in the of the early third century CE as Muthukumaran that the of used for in the Mediterranean suggests the of routes of or the of more than one via the Red Sea to and then to The short chapter the of a already in together with et al. the is not in the and is also from the tree Muthukumaran then another South Asian plant that has migrated to the known as Indian or identification with Akkadian a is already in the (see This is by an Achaemenid with Old Persian and which is a of Sanskrit The of is from to may have a Dravidian as other goods from Magan and Meluḫḫa in texts South Asian finds from the anthropogenic use of the sissoo tree at least by the millennium BCE A with sissoo from in dating to the third millennium BCE the of sissoo rather than its cultivation This well with the fact that and there is no cuneiform evidence from Mesopotamia that came directly from Meluḫḫa to The of that it was from South Asia to and from there to This is supported by the fact that the earliest cuneiform evidence from the late third millennium BCE identifies Meluḫḫa and Magan as the of the The could be for of which be and which is in India, that it was imported from India to and from there to Muthukumaran that there is no evidence that sissoo was cultivated in the late third or second millennium BCE and that cultivation only in the century BCE However, the reviewer notes that an in an Old Babylonian from lists from (see This could be a to the fact that sissoo was already cultivated in in the early second millennium the early sissoo was cultivated in tree in southern and perhaps also in It was not until the century BCE that sissoo was cultivated on a large in Mesopotamia, both on in the and in the There is cuneiform evidence for sissoo in Babylonia, and India as late as the Achaemenid period to In the last chapter, Muthukumaran the to the to He from the previous in which plants migrated from South Asia to Southwest Asia and the Mediterranean. Since were he the role of such as the palace or in the and use of that had the to with crops, this is very However, must be to at least as as the written evidence is since and other are usually the only to written at least in ancient a must first with plants, the first is Muthukumaran the importance of ancient long-distance However, he also that plants did not come only with for or as a also exchange and especially in first-millennium BCE had with a plant, they could on to the next and large-scale cultivation it was to and and crops such as cotton also on the Thus, the lower Mesopotamian with its and irrigation such as a and were also As with rice, and and were in a was or not Muthukumaran a role to the as the for cultivation a had been by farmers and consumers and had found its and the third was according to the author, the fourth was the of the plant, which could often of this work is that it together the of a of and in a comprehensive and The is The of on some years of cultivated plant history an area some from west to is a some of the literature of the will The reviewer has therefore some of these in the Even the monograph offers completely for in the it is a valuable that the field of literature from and it for the In the to literature and with ancient ancient and ancient Persian plant to the reviewer, were a great the is A more comprehensive one would be to the In the of ancient terms and nomenclature of cultivated plants should have been in the or in a with to of the as a in this field is of importance and, the reviewer’s the author has a in the for It was a to