Colocasia esculenta (L.) Schott, a perennial herb grown for its starchy underground corm and for its leaves, and cultivated in two botanical varieties that behave quite differently in the ground and in the pot. Var. esculenta, the dasheen, throws a single large central corm and thrives standing in water. Var. antiquorum, the eddoe, makes a smaller central corm ringed with a cluster of cormels, and grows on dry land. Both are distinct from Xanthosoma sagittifolium, the New World tannia or cocoyam, which is a different genus altogether and is treated separately.
Taro is old enough that the question of where it was domesticated has more than one honest answer. The wild plant runs in a broad band from the swampy forest margins of north-eastern India through Upper Myanmar and southern China to island Southeast Asia and New Guinea, and it is a plant that suits people before anyone decides to farm it: it grows where water stands, it can be dug at any time of year rather than at a harvest, and it propagates from its own offsets, so a gardener who wants more simply replants the top of what has been eaten.
The oldest secure archaeological evidence for cultivating it is not in Asia at all. At Kuk Swamp in the Wahgi Valley of the New Guinea highlands, drainage ditches cut into wetland soil have been dated to roughly ten thousand years ago, and starch grains and use-wear recovered from stone tools at the site have been read as taro processing. UNESCO inscribed Kuk in 2008 as one of the few places on earth where agriculture can be shown to have been invented independently. The interpretation of the earliest phase is not unanimous, and the debate is a live one amongst archaeologists, but the later phases are not in serious dispute, and taro is central to them.
Genetic work has since supported what the archaeology implies. Studies of chloroplast and nuclear markers find deep divisions in cultivated taro that do not collapse into a single origin, and the majority reading is that the plant was brought into cultivation more than once, in Asia and again in New Guinea, from wild populations that were already distinct. The atlas therefore gives taro two origin points rather than one.
The two cultivated varieties diverged early and are still farmed as two different crops. Dasheen, var. esculenta, is the wetland plant: a single heavy corm grown in flooded terraces, which is what a Hawaiian loʻi, a Chinese paddy, and an Egyptian Delta field all produce. Eddoe, var. antiquorum, is the dryland plant, whose small central corm is ringed with a fistful of cormels, and it is those cormels that are eaten across much of India, the Caribbean, and Japan. The difference is not merely botanical. A dasheen corm slices and holds its shape or pounds to a smooth paste; eddoe cormels are firmer, drier, and more slippery, and they are cooked whole. A recipe written for one is frequently wrong for the other.
Everything about taro's diffusion follows from one fact: it is propagated vegetatively rather than from seed. Taro flowers rarely and sets viable seed rarely still, so what travels is a living cutting, and every plant at the far end of a voyage is a clone of the one that left. That is why taro moved with people and only with people, why its varieties carry names rather than numbers in the places that grow them, and why the loss of a named cultivar is final.
Taro travelled west and east from its Asian homeland along two entirely separate systems, and it is a rare plant that can be said to have completed both journeys before the classical world had finished writing itself down.
The eastern journey is the great Austronesian expansion, and taro was in the canoes for the whole of it. Carried out of Taiwan from around 3000 BCE, through the Philippines and island Southeast Asia, it met the older New Guinea taro somewhere in the Bismarck Archipelago, and the Lapita voyagers who reached Fiji, Tonga, and Samoa around three thousand years ago took a mixed inheritance with them. From western Polynesia it went on to the Society Islands and finally, around the turn of the second millennium, to Hawaiʻi, which built the most elaborate taro civilisation of all. It is one of the canoe plants, the small suite of species Polynesian voyagers carried deliberately because a landfall without them was not a home.
The western journey went by sea and by caravan into Egypt, where taro was established well before the Roman period and is farmed in the Nile Delta still. The classical record here is a trap worth walking into carefully. Greek kolokasion, which Dioscorides and Pliny both use, did not originally mean this plant at all: it named the root of the Egyptian lotus, Nelumbo nucifera, and the word was only later transferred to the aroid growing in the same fields. Linnaeus inherited the transferred sense, which is why the genus is Colocasia today. Anyone reading the ancient authors for taro's arrival date is therefore reading a name that has moved under them.
From Egypt the plant went three ways. North to Cyprus, where kolokasi became and remains a genuine European staple, the only one taro has anywhere in Europe, grown around Sotira and Avgorou and protected under its own designation of origin. East into the Levant, where Beirut and Damascus cook it with lemon and lentils. And west across the Sahel into the forest belt of West Africa, where it became the older of the two crops that English calls cocoyam.
The Atlantic crossing was made in the holds of slaving ships, as provisioning rather than as cargo, and taro took root across the Caribbean, where it acquired the name it still carries there. Dasheen is generally derived from the French de Chine, of China, a plausible reading that has never been proved, and the leaf rather than the corm became the point: callaloo, the dish and eventually the word for the green itself, is dasheen leaf, and it is one of the few cases anywhere in which the transported plant mattered more for its foliage than for its root.
Taro feeds a great many people and is counted badly. The production figures put Nigeria, China, Cameroon, and Ghana at the head of the table, but the statistics for West and Central Africa fold Colocasia together with Xanthosoma, the New World tannia that arrived in the nineteenth century and has substantially displaced it, so the African numbers describe two crops under one heading. What is not in doubt is the direction of travel: taro remains a staple for tens of millions of people across the Pacific, South and Southeast Asia, West Africa, and the Caribbean, and it is nowhere near as marginal as its absence from the northern supermarket suggests.
In Hawaiʻi the plant is not a vegetable but a relative. The Kumulipo and the Hāloa tradition make kalo the elder sibling of the Hawaiian people, born before them and buried, and the taro that grew from that burial the ancestor of both. Sugar and cattle took the water that the terraced loʻi needed, and the acreage collapsed through the twentieth century. Its restoration since the 1970s, valley by valley and stream by stream, has been as much a matter of water rights and Hawaiian sovereignty as of agriculture, and the fight over stream flows is still being conducted in the courts.
At the other extreme, taro has become a flavour before it is a food. The purple that a Western consumer now associates with the word usually comes from ube, a Philippine yam, or from a flavouring, rather than from taro, whose flesh is more often cream or grey flecked with violet; but taro milk tea, taro ice cream, and taro crisps have carried the name into markets that have never seen the corm.
One caution belongs at the end of any account of this plant, because it applies to every part of it. Raw taro is inedible and unpleasantly so: the corm, the leaf, and the stem all carry needle-shaped calcium oxalate raphides that lodge in the soft tissue of the mouth and throat and burn for hours. Thorough cooking destroys them, which is why every cuisine that has ever taken taro seriously boils, steams, bakes, or ferments it at length, and why the instruction to cook it far longer than seems necessary is the one instruction in a taro recipe that must never be shortened.
A note on the entry's boundaries. Xanthosoma sagittifolium, called tannia, malanga, yautía, macabo, and in West Africa the new cocoyam, is a different genus with a Neotropical domestication of its own, and it is not covered here despite sharing kitchens, names, and frequently the same pot. Alocasia macrorrhizos, the giant taro of the Pacific, and Cyrtosperma merkusii, the giant swamp taro that is the great famine insurance of Micronesia and Kiribati, are likewise separate plants. Where a dish could be made with either taro or tannia, the entry says so rather than pretending the distinction is always observed at the market.
Historical Journey of Taro
Kuk, the Wahgi Valley & the New Guinea Highlands — c. 8000 BCE
The wetland floor of the Wahgi Valley holds the oldest evidence anywhere that somebody was growing taro rather than gathering it. Drainage ditches cut into the swamp at Kuk have been dated to roughly ten thousand years ago, and the starch grains and wear traces on stone tools from the site have been read as taro processing. UNESCO inscribed Kuk in 2008 as one of a handful of places on earth where farming was invented rather than imported. The earliest phase is argued over and should be reported as argued over; the later phases are not, and taro is at the centre of them.
What the ditches represent is a specific solution to a specific plant. Taro wants standing water but not stagnant water, and a swamp cut through with drains is a swamp converted into a taro garden. That is a considerable amount of engineering to expend on a crop, and it was expended nine thousand years before the first pot was fired in the region.
Melanesia has cooked it the same way for most of that time. Laplap, grated taro and banana bound with coconut cream, wrapped in leaves and baked on hot stones in an earth oven, is what the pit at Kuk was ultimately for, and it remains the dish of the Melanesian feast from the New Guinea highlands through Vanuatu.
Assam, Bengal & Upper Myanmar — c. 5000 BCE
The wild taro of the eastern Himalayan foothills and the Bay of Bengal is the population from which most of the world's cultivated taro descends, and it is here that the two farmed varieties parted company. Var. esculenta, the dasheen, went into standing water and made one heavy corm. Var. antiquorum, the eddoe, stayed on dry land and made a small corm ringed with cormels. India cooks the eddoe, and calls it arbi.
Bengal is where the plant is used most exhaustively. Kochu covers the whole of it: the corm mashed with mustard oil and green chilli as kochu bata, the leaf as kochu shaak, and the loti, the pale runner the plant throws out sideways to start a new corm, cooked with prawns or a fish head until it gives. Bengal eats the stolon and Kerala eats the stem, and between them very little of the plant is left unused anywhere on the subcontinent.
The hill regions of Assam and Upper Myanmar kept it closest to its wild habit, gathering and planting it along the margins of paddy where nothing else would grow, which is more or less how the plant behaves without human help. From this homeland taro went in three directions across the subcontinent: south down the peninsula to Malabar, west to Gujarat and the Konkan, and east into the Chinese river systems.
Guangzhou & the Pearl River Delta — c. 3500 BCE
Taro reached the Yangtze and the southern Chinese river systems early, and Chinese agricultural texts treat it as an established famine reserve long before the sweet potato arrived to take that role. It suited the flooded field, it stored in the ground, and it did not compete with rice for the same calendar. The character 芋 yù is ancient, and 南方草木狀 and the later agricultural manuals list named cultivars.
What Guangdong built with it is not a famine food at all. Yùtou kòuròu layers thick slices of taro with pork belly, both fried first, then steams the assembled bowl for hours so that the fat runs down through the taro and the taro gives its dry starchiness back to the pork; it is inverted onto the plate at the table, which is what kòu means. The Hakka claim it and cook it at the New Year.
The other Cantonese contribution is a dim sum. Wu tau gou, taro cake, binds diced taro with rice flour, dried shrimp, and Chinese sausage into a steamed loaf that is then sliced and fried, and it sits on the trolley beside its more famous radish-cake sibling. Between the two dishes, taro in southern China spans the celebratory table and the mid-morning one.
Taipei, Dajia & Taiwan — c. 3000 BCE
Taiwan is the staging post for half the world's taro, because Taiwan is where the Austronesian expansion began. The peoples who left the island from around 3000 BCE carried a small deliberate suite of plants with them, taro amongst the first of them, and every taro in the Philippines, Micronesia, Melanesia, and Polynesia that is not descended from the New Guinea line traces back through here. It is a transit point in the largest sense and also a place that never stopped growing the crop.
Modern Taiwan has made taro into something the ancestors would not recognise. The betel-nut purple of Dajia taro, grown on the Taichung coast and the most prized cultivar on the island, turns up in cakes, in ice cream, and above all in yùyuán, taro balls: cooked taro mashed with sweet potato starch, rolled, cut into stubby cylinders, boiled, and served hot in sweet soup or cold over shaved ice with red beans. The mountain village of Jiufen, which emptied when its gold ran out and refilled with tourists after a film was shot there in 1989, sells them by the tonne on a staircase of stone steps, and taro balls are now as firmly a Taiwanese emblem as the tea that goes with them.
Kerala, Tamil Nadu & the Dravidian South — c. 2500 BCE
Taro moved down the peninsula from the north-eastern homeland in prehistory and became, in the Dravidian south, an entirely ordinary vegetable: not a famine reserve, not a ceremonial crop, but something in the weekly rotation of almost every household kitchen. Malayalam calls the corm chembu, Tamil seppankizhangu, Telugu chamagadda, and Kannada kesuvina gadde.
The south keeps its tubers carefully separated in a way English does not. Chembu is Colocasia esculenta; chena is elephant foot yam, Amorphophallus paeoniifolius, a different genus entirely; kaachil is a true yam, a Dioscorea. English calls all three yam and loses every distinction that matters at the stove, which is a reasonable illustration of why the vernacular names are worth keeping.
Kerala goes further with the plant than anywhere else in the subcontinent, eating three parts of it. The corm is boiled and then mezhukkupuratti, smeared with coconut oil and tossed with shallots and curry leaves until the edges catch. The leaf is used as a wrapper and a green. And the thal, the thick fleshy stem between them, is de-stringed, cooked hard with tamarind or kudampuli, and made into thoran with grated coconut, which is a use of the plant found in very few other cuisines anywhere.
Tamil Nadu concentrates on the corm and on making it crisp. Seppankizhangu varuval boils the corm, flattens each piece with the heel of the hand to open up the surface, dusts it with rice flour and sambar powder, and shallow fries it until the crust shatters.
The southern kitchen's handling of the plant's oxalate is the most systematic anywhere in this entry, and it is chemical rather than superstitious. Acid goes into the cooking water rather than onto the finished dish: tamarind, or a piece of kudampuli, the Malabar gambooge, which is the same souring agent that defines the region's fish curries. The plant itself asks little in return, growing on paddy bunds, backwater margins, and the wet shaded corners of a homestead where nothing else will crop.
Gujarat, the Konkan & the Western Deccan — c. 2200 BCE
Taro reached the western seaboard in prehistory and Gujarat and the Konkan did something with it that almost nobody else has attempted at scale: they took the plant seriously as a leaf. Colocasia foliage is large, free, and grows back within weeks, and it is also inedible raw, being the most oxalate-heavy foliage in ordinary cultivation. The western Indian answer was to spread the leaves with a thick paste of gram flour soured with tamarind and sweetened with jaggery, stack them, roll the stack into a tight log, steam it, and cut it across into pinwheels.
The dish that results is called patra in Gujarat, alu vadi in Maharashtra, from alu, the local name for the leaf, and patrode along the Konkan and in coastal Karnataka. They are one dish under three names rather than three dishes, and the differences are of degree: Gujarat sweetens it more heavily with jaggery, Maharashtra makes it hotter and drier, and the Konkan frequently steams it in a different shape.
What makes it worth a point on the map is that the technique is a solution to the plant rather than merely a use of it. The acid in the paste attacks the calcium oxalate that makes raw colocasia burn the mouth, so the tamarind is doing chemical work before it does any culinary work at all, and the sourness that now defines the dish began as the thing that made it safe. It is one of the clearer cases in this atlas of a technique inventing a taste.
The corm is eaten here too, as arbi, boiled and fried or dropped into a sour curry, and it belongs to the same western sweet-and-sour axis, the khatta-mitha balance of tamarind held against jaggery, that governs so much of the Gujarati kitchen. Patra is monsoon food, made when the leaves are largest and most tender, and it turns up in the tiffin box, on the festival thali, and at the roadside stall in roughly equal measure.
Bicol & the Philippine Archipelago — c. 2000 BCE
Taro came south out of Taiwan with the first Austronesian settlers of Luzon and spread through the archipelago ahead of rice, which is why gabi is one of the oldest food words in the Philippine languages and why it turns up in the sourest and most everyday of Filipino dishes: the corm thickens sinigang, dissolving into the broth and giving it body without flour.
Bicol, on the south-eastern peninsula of Luzon, took the leaf instead and built its signature dish on it. Laing simmers dried taro leaves in coconut milk with chilli, ginger, and a little shrimp paste or pork until the leaves collapse into a dark, oily, fiercely hot mass and the coconut fat separates out on top. The rule that governs it is one every Bicolano cook states first: do not stir it. Agitating the pot before the leaves have surrendered releases the oxalate and the dish will scratch the throat of everyone who eats it, so the pan is left alone for the better part of an hour. It is a superstition with a chemical basis, and it is correct.
Viti Levu & the Fijian Islands — c. 1100 BCE
The Lapita voyagers who crossed from the Bismarck Archipelago into Fiji around three thousand years ago were carrying a mixed taro: the Asian plant brought down the Austronesian chain and the older New Guinea plant picked up on the way. Fiji is where the two lines met and went on together, and it is the western gate of the Polynesian world.
Dalo is still the ceremonial starch of Fiji, presented in bundles at a sevusevu and lifted from the lovo at any occasion worth digging a pit for. The everyday dish, though, is the leaf: rourou, young dalo leaves simmered slowly in coconut cream with onion until they are black-green and silken, eaten with fish or with more taro. It is the Fijian member of a family that runs the length of the Pacific under a dozen cognate names, and the technique is identical in every one of them.
Apia, Samoa & Western Polynesia — c. 900 BCE
Taro is one of the canoe plants: the deliberate cargo of a voyaging society, packed alongside breadfruit, banana, coconut, sugarcane, and paper mulberry because a landfall without them was a camp rather than a home. Reaching Samoa and Tonga around three thousand years ago, talo became the prestige starch of western Polynesia, the thing planted first and presented at every formal exchange.
Both the corm and the leaf are cooked, and the leaf carries the region's most characteristic dishes. Samoan palusami wraps coconut cream in young taro leaves and bakes the parcels on the stones of an umu until the leaf braises in the fat and turns to silk. Tongan lū pulu layers the same leaves with onion, coconut cream, and corned beef in the ʻumu, and the word lū in its name is simply the taro leaf itself. The cooking time in both is long past the point of apparent doneness, and for the usual reason.
The Nile Delta & Egypt — c. 400 BCE
Taro reached Egypt from the Indian Ocean world, by way of the Red Sea and the caravan roads, and was established in the Delta well before the Roman period. Reading the classical authors for the date, however, requires care. Greek kolokasion in Dioscorides and in Pliny does not originally name this plant at all: it names the root of the Egyptian lotus, Nelumbo nucifera, and the word slid across to the aroid growing in the same wet fields only afterwards. Linnaeus inherited the shifted sense, which is why the genus is Colocasia today, and why an ancient reference to colocasia in Egypt cannot simply be taken as a reference to taro.
What is not in question is that Egypt has grown and eaten it continuously for a very long time. Qolqas is a winter vegetable there, sold in great muddy corms in the markets from November, peeled under running water because the sap irritates the hands, and stewed with chard or with coriander and lemon until it thickens the pot on its own account. It is the plant's only substantial foothold in North Africa and the point from which it reached both Cyprus and the Levant.
Yamagata & the Japanese Archipelago — c. 300 BCE
Taro arrived in Japan from the south before rice did, and the language still records the fact. Satoimo, 里芋, is the village tuber, named in explicit contrast with yamaimo, 山芋, the mountain tuber, which is a wild yam gathered rather than grown. The pair of words marks the moment a foraged plant became a farmed one, and the older tuber culture it belongs to was substantially displaced by rice without ever quite disappearing: satoimo remains the traditional offering at the harvest moon, and in much of the country the New Year ozōni is not right without it.
Yamagata prefecture turned it into a public occasion. The imonikai is an autumn riverbank gathering at which taro, beef, konnyaku, and spring onion are simmered together in an iron pot over a driftwood fire and eaten outdoors by everyone present. The largest of them, held on the banks of the Mamigasaki in Yamagata City each September, cooks for tens of thousands of people in a cauldron six metres across, lowered and stirred by a construction crane with a specially made shovel, and seasoned with sake by the cask. The scale is the joke and the point.
Seoul & the Korean Peninsula — c. 100 CE
Taro came into Korea from China and settled into a single, precisely placed role. Toran, 토란, is Sino-Korean 土卵, earth egg, and it is autumn food: harvested in the ninth lunar month and eaten at Chuseok, the harvest thanksgiving, in a clear brisket soup with dried radish greens and seaweed.
Torantang is the dish, and it is one of the few in the Korean repertoire that is genuinely seasonal rather than nominally so, because fresh toran is available for a matter of weeks. Korean cooks parboil it in rice-washing water or in salted water and discard that liquid before the soup is built, which draws out the slipperiness and the irritation together; the technique is old, and it is the same instruction every taro-eating culture arrives at independently.
Tahiti & the Society Islands — c. 200 CE
The Society Islands were settled from western Polynesia, and taro went with the canoes. In Tahiti it became the first crop of the valley floors, terraced and irrigated wherever a stream could be diverted, and the archaeology of those terraces is a good part of what is known about the islands before Europeans arrived.
Tahitian keeps the leaf and the corm firmly apart in the vocabulary. Taro is the corm; fafa is the young leaf, and it names a dish of its own. Poulet fafa braises chicken with a great heap of fafa until the leaves collapse, then finishes it with coconut milk, and it belongs to maʻa Tahiti, the Sunday meal cooked in the ahimaʻa and eaten at length after church. The dish's name carries the island's whole modern history in two words, one French and one Tahitian, and the Tahitian one is the older by a very long way.
Sotira, Famagusta & Cyprus — c. 400 CE
Cyprus is the only place in Europe where taro is a traditional staple rather than an import, and it has been grown on the island since antiquity, reaching it from Egypt. Kolokasi is farmed around Sotira and Avgorou in the Famagusta district in the east of the island on heavy red soil, lifted from November, and protected under its own designation of origin, which distinguishes the large parent corm from the poulles, the small offsets, and treats them as different products for different dishes.
The Cypriot technique is unlike any other in this entry and is the reason the crop survived here. The corm is never washed before cutting, because wet kolokasi turns to slime; it is wiped, and then, instead of being sliced through, each piece is scored and snapped off with the point of a knife, so that it breaks along its own grain and the broken face seals in the pan. It is then browned in oil, and only afterwards stewed with pork or chicken, celery, tomato, and lemon. Cooks on the island will tell you that a kolokasi cut with a knife falls apart, and they are describing something real.
The Bight of Biafra & the Gulf of Guinea — c. 600 CE
Taro crossed the Sahel from the Nile into the West African forest belt long before the Atlantic trade, and became the older of the two crops that English lumps together as cocoyam. In Igbo it is ede, a word that later travelled the other way across the Atlantic and gave the Caribbean its eddoe.
Its role in the region is structural rather than starring. Boiled and pounded to a smooth paste, cocoyam is the classical thickener of the Igbo soup pot, and ofe onugbu, the bitterleaf soup that is the flagship of the Igbo kitchen, is built on it: the pounded corm is dropped into the simmering palm-oil broth in lumps that dissolve slowly and give the soup its characteristic heavy, drawing body. Where a cook uses a stock cube of thickener instead, the difference is the first thing anyone comments on.
One honest qualification belongs here. Since the nineteenth century the American Xanthosoma, the new cocoyam, has substantially displaced taro across West Africa, and the two are sold, cooked, and counted together. A West African cocoyam dish today may be made with either, and frequently with whichever is at the market. The atlas notes the ambiguity rather than resolving it by assertion.
Beirut, Damascus & the Levant — c. 900 CE
Taro reached the Levantine coast from Egypt and stayed close to it. Kilkass is a winter vegetable of the Lebanese and Syrian littoral rather than of the interior, grown in the coastal plain around Tripoli and Sidon where the water table is high, and it appears in the markets for a few months and then vanishes.
The Levantine treatment is the most acid of any in this entry, and deliberately so. Kilkass bi adas boils the peeled corm with brown lentils and then floods it with lemon juice and crushed garlic, in quantities that would be excessive with any other vegetable; here the acid does two jobs, cutting the flat starchiness and finishing what the long boil began on the oxalate. It is a Lenten dish and a fasting dish, eaten hot with bread and eaten cold the next day, and the version with tahini instead of lentils is the same argument made in a different accent.
Waipiʻo Valley & the Hawaiian Islands — c. 1100 CE
Nowhere took taro further than Hawaiʻi. Settled from the Society Islands around the turn of the second millennium, the archipelago built terraced pondfields, loʻi, fed by engineered irrigation channels, ʻauwai, drawn off the streams of the windward valleys; Waipiʻo on Hawaiʻi Island held some of the largest. Hundreds of named cultivars were selected, each for a purpose, and Hawaiian keeps distinct words for the plant, the corm, the leaf, and the paste at several stages of dilution.
Kalo is not a vegetable in Hawaiian thought but a relative. The Hāloa tradition holds that the first child of Wākea and Hoʻohōkūkalani was stillborn and buried, that taro grew from the grave, and that the second child, the ancestor of the Hawaiian people, was named for it. Kalo is therefore the elder sibling, and the obligation runs from the younger to the older.
Poi is the corm steamed, peeled, and pounded on a wooden board with a stone pounder until it is smooth, then thinned with water and left to sour for a day or two; it is graded by how many fingers are needed to lift it. Laulau wraps pork and fish in lūʻau leaves and steams the bundles for hours, and heʻe lūʻau, squid stewed with taro leaf in coconut milk, completes the set. The lūʻau feast is named for the leaf.
The twentieth century took the water. Sugar and ranching diverted the streams that the pondfields depended on and the acreage collapsed. The restoration of the loʻi since the 1970s has been carried out valley by valley and has been as much a matter of water rights and of Hawaiian sovereignty as of farming, and the litigation over stream flows continues.
Port of Spain, Trinidad & the Eastern Caribbean — c. 1650 CE
Taro crossed the Atlantic in the holds of slaving ships as provisioning rather than as cargo, and took hold across the Caribbean under two names that between them summarise the ocean that carried it. Eddoe came with the people, out of the Igbo ede and the Efik and Ibibio edé. Dasheen is generally derived from the French de Chine, of China, a colonial guess about the plant's origin fossilised into its name; the derivation is repeated everywhere and demonstrated nowhere.
What is unusual about the Caribbean is that the leaf mattered more than the corm. Dasheen bush became the green of the region, and callaloo came to name first the dish and then the leaf itself, so that a Trinidadian asking for callaloo at the market is asking for taro leaves. The dish is the West African leaf-and-palm-oil soup rebuilt with what the islands had: dasheen leaf and okra simmered with coconut milk, thyme, and a whole scotch bonnet lifted out before it bursts, and crab in it on a Sunday.
The word did not stay put either. Callaloo now means the national mixture in Trinidadian usage, the metaphor a Trinidadian reaches for to describe the island's own population, which is a considerable career for the leaf of a swamp plant from the Bay of Bengal.
