KOJI, Tanaka, (Linyanxin translation), Kyoto University Agricultural Dept., Japan

(Agricultural Archaeology 1996(3):122-132. Japanese>Chinese transl. by Lin Guangxin & Peng Shijiang, History Laboratory Researcher, S China Agricultural University, Guangzhou, PR CHINA. Trans. by Jiwu Wang, Jianming Liang & B. Gordon)

(3) Digging sticks (dibble) and plows

I continue with plow and dibble, the latter used on dry and wet land. The E Indonesian S Sulawesi plant mainly corn but use a 5 ft. iron-tipped dibble (lingis) to turn rice dryland soil or break clumps at the end of the dry season, followed by seeding and rake levelling. The dibble is also used in slash & burned fields and fallow for dry rice, corn, millet, other grain and potato. Metzner (1977:116-127) said dry rice was recently introduced to E Timor because there is less rain and it tasted better than paddy rice. Paddy rice is planted for 2-3 years, with fields fallow for a few years. Before seeding, burnt dryland is either unplowed (lere rai) or plowed (fila rai). After weed burning on lere rai, dibbles (ai suak) are used for seeding. After weed burning on fila rai, sharp 2.5-3 m long dibbles (ai suak boot) are used to turn soil, several people working together because one can only do 100 sq.m per day.

Both E & W Timorese use dibbles, the former for transplanting rice seedlings on unplowed slash & burned land. On dryland they use 1.2-1.5 m long dibbles (lingis, aesuak or asuwak) to plow by turning soil clumps and then transplant. They are iron or wood; but some wood ones have iron points and are used for dry rice, corn, millet, beans and potato. Some areas don't have dry rice (Poniman 1988:142-155). Foluolesi Islanders use the iron-pointed dibble (su'a) to cut grass between major dry rice crops. Year 1 dry rice is planted with corn and beans; year 2 rice is mixed with potato; year 3 field is just potato. Burnt wood is used to build dikes. Dibbles or spades (tambi) are used before planting to cut grass, and after planting to extract its roots (Shandao 1990:650, 753, 758).

E Indonesians also use dibbles to prepare dryland. Crop structure suggests they were used before rice was introduced, probably for grain and tubers in unplowed slash & burned and dry fields. Da Lin's (1990:549-552, 555, 557, 560, 561) Oceanian and SE Asian data say they were introduced from SW China, but were they for turning soil or planting in slash & burned fields? He did not discuss the issue, but they were certainly used from the E side of the SE Islands to Meilanixiya, where Barrau (1958:9, 19-24) and Brass (1941:557) say they were widely distributed for planting well-developed crops. They are ancient but persist (Golson 1977:601, 638, Goercki 1978:185-190, Nmag 1981:31-39). I believe E Indonesians used them after grain and rice were introduced.

Dibbles are used in dry field and paddy. The E Timorese use them (ai suak) to fix dikes after hoof plowing (Metzner 1977:129, 342), the same way they are used in turning soil. Sasaki (1987a:581-609) said the Philippines and other island "superior dry rice phenomenon" changed slash & burned land "root-planting to dry rice-root planting". Original dibble use on slash & burned land for paddy probably relates to cultivated dry rice evolution, or grew from dry rice land transformation to paddy.

Island dry rice and paddy planting similarities standardize paddy spades and dibbles; e.g., historic spades resemble paddles. The popular Yfujiaduo gaud is a 1.5-2.1 m long dibble with 0.5 m long pointed shovel. Its spade is often wood, the dibble iron-pointed, while others are all iron. According to soil condition, a spade has a narrow straight or angled point for digging, diking, turning soil (esp. dryland) and preparing fields (Conklin 1980:17-21).

The S Sulawesi Mountain Tuolacha use the same spade, the Boermasi a different one (peleko) like the Yifujiaduo, another 1.5 m long with iron point for turning paddy soil, and a longer one to prepare paddies and build bridges. All are iron-pointed wood, but some dibbles are covered with thin iron sheet.

We don't know if the island spade is from the E Indonesian dibble or if it were for paddy rice. As the Yifujiaduo plant rice in slash & burnt valleys using dibbles to turn soil (Conklin 1980:24) and the Tuolacha do not turn slash & burned soil, using dibbles to dig a hole for rice or corn seed, it is difficult to prove the spade evolved from the dibble, considering only slash & burn plowing. We conclude the dibble is used for root and paddy rice planting, stressing the spade is a paddy alternative rather than another dibble shape. Considering spade use in building terraces and preparing fields by hoof plowing, the spade is likely ancient. Indeed, Barrau says root crop dibbles and spades are for preparing fields and planting seed, like paddies. As both spades and dibbles are used for tubers (Barrau 1958:21), it is predictable some think the paddy spade evolved from root planting, but those who say the spade and hoof plow were used simultaneously from ancient times, are wrong. As spade origin is not my topic, I will discuss it later. First, I want to describe another SE Island trait, brushhooks to prepare fields instead of plowing.

(4) Non-plowing/reverse tilling using brushhooklike tool

Brushhooklike tools are absent in Malayan and Sarawakan unplowed rice fields, but common in coastal S Borneo and E Sumatran unplowed fields; e.g., Malaccan Channel and S China wetland especially suited to rice. The E Sumatran Liaonei use short brushhooklike tools (parang) and long dibbles (tajak) to cut and uproot grass before collecting it after rotting, repeating the process to hasten new grass and transplant seedlings. Like a tajak, a parang or short-handled 50-100 cm knife is used to cut grass, mostly individually but also by two people, one cutting, the other upending, and often repeated, after which field preparation is finished.

The plow is absent in wetland due to water, long grass and thick peat layers; i.e., no inorganic soil because it developed from rotting grass. Rice cultivation checks quick-growing grass more effectively than upending soil. As it is impossible to upend all grass roots by spading, the parang and tajak become major field preparators.

Wetlands along rivers resemble the coast. As rivers transport soil, rain hastens grass and land is submerged in rainy season, rice is planted in dry season. As excess water and grass inhibit plowing and transplanting, grass is cut with parang and tajak in unplowed rice fields, but some people use the hoof plow, as aforementioned. Land partly inundated by seawater is unplowed and needs multiple transplanting with dibbles.

Except Liaonei State, tajak and parang use without plowing occurs elsewhere in Sumatra. Kormering River people heavily use them (Poniman & Takatani 97-127), like people in W Sumatran Aqi and Gulu States. Elsewhere, grass is burnt and upended with a spade (ibid.:2-6, 80-88); e.g., N Sumatrans to prepare rice fields (ibid.:15-16, 33-41). It is used in mountain valleys and on salt and freshwater banks.

NE Malayans do not plow rich organic soil, using a W Malaccan, Kerian, Tanjong Karang and Sungei Manik method brought by early 20th century S Borneo and Java immigrants using tajak and parang to prepare unplowed rice fields (Jackson 1972:90-94).

S Borneo people do not plow coastal wetland, adopting the Liaonei tajak to cut and leave grass for 10-15 days; then upend it every few days while new grass is cut (Beusechem 1939:4; Noors-yamsi & Kidayat 1974:11). Rotted grass is used as fertilizer. S Sumatrans also use this method. Cultivated coastal wetland does not have a long history. E Sumatran and Malaccan Channel people began rice cultivation in early 20th century, peaking in 1930-40's, but plowing is not popular because the technology is new. Suggesting pre-plowing, wetland people retain lebak and paya, the simplest methods on the coast and lower rivers, the last region developed for island rice cultivation and the start of the next part.

I now summarize paddy preparation by describing some island findings related to the change from hoof to non-plowing, while ignoring Javanese and Balinese foot-plowing12, replaced by plow and spade. As elsewhere, plowing began with hoof-plowing which disappeared. As my examples are personal observation; it is impossible to re-observe them, but SE island field preparation includes the following traits: 1. Borneo and Philippine valleys have mainly hoof & foot-plowing and spading; 2. Sumatran and Malayan valleys and wetland have mainly hoof & foot-plowing and non-plowing; 3. Xiaozhenta region has hoof-plowing and dibbles for paddies and fallow dry land; and 4. new coastal rice fields were unplowed, but plow is now common, especially in Java and Bali.

2. Rice planting methods

Comparing Chinese rice transplanting and Indian dryfield broadcasting, SE island people mix methods; i.e., under Chinese-type transplanting, they also used dibbles, multi-transplanting, dry field broadcasting and bunch planting, the last quite unique and a trait of Malaysian rice cultivation.

    1. Dibbling
    2. Bunch planting is widespread in SE islands and mainland, starting with slash & burn, then dibbling to insert rice, a typical Asian method13. SE islands use slash & burn in dry and wetland; e.g., Sumatrans slash & burn dry and river wetland for rice, and after early 20th century, for rubber plantations. They burn brush and plant rice between widely-spaced rubber trees. As highland and valley paddies often have too little or too much water, respectively, bunch planting using dibbles to insert rice occurs in those seasons when both are dry but the air is wet. Dry fields are prepared with plow and spade, wetland by hoof-plow, but better water distribution means all is spaded.

      N Sumatrans also bunch plant, first plowing, then spading a month later, followed by rice planting 2 weeks later with two dibbles, a woman with a bamboo tube inserting the rice and covering the hole. Holes are also covered with coconut leaves trodden by buffalo. Since the 1950's, bunch planting also occurs with machine-dug holes, 12 at a time, but the prevading dryland trait remains bunch planting (Poniman 1988:17-27).

      Malayan transitional paddy/slash & burn dryland bunch planting (padi tugalan) begins after early rainy season plowing (Hill 1951:61), but detailed records are absent (Jackson 1972:84). Since plow introduction, Indian methods were adopted, but bunch planting persists14.

      Mature Javanese rice cultivation includes bunch planting. Slash & burn was in central and E Java until early 20th century (Damu 1987:2-21). Late 19th century Holland official De Bie said it preceded dibbling (pandja gedjig, djemblong or aseuk), seed insertion and leaf covering, like Sumatra. Elsewhere, people converted rice field from burnt forest (hoema or ladang) or open land (tegal), while dryland was used after fallow (tipar). Fallow land was plowed, burnt land not, but both were bunch planted (Bie 1901:15, 86).

      Paddy bunch planting after plowing is mainly in Sawah tadahan or Sawah tadudan land (sawah geledoek). E Java fields (rancah) with watered maturing seedlings gogo (dry slash & burn) or sawah (paddy), but Damu (1989:32) said their difference is "not as great as imagined, with cultivation inseparable" in paddy, dryland and slash & burn. It replaced transplanting in rainy Sawa tadahan and persists in Java and some Indonesian dry areas.

    3. Broadcasting in irrigated field
    4. The E Timorese mostly broadcast in wetlands with some transplanting. After hoof plowing, dike fixing and drainage, they broadcast 1 hectare/day/person, wait 5 days and water. When seedlings have 4-5 leaves, fields are dried but may be rewatered until harvest (Metzner 1977:128-131). W Timorese transplant after hoof plowing, but non-irrigated fields are broadcast (ngkakik). Central Timorese wetland broadcast is from Sawu Island (Poniman 1988:133-159). Transplanting is replacing broadcasting overall, but Timor wetland broadcasting was traditional.

      Direct planting (30%) supplemented broadcasting (70%) in Puliwali County in the early 1980's, with Holland-imported transplanting common 20 years ago. Direct planting includes wetland broadcasting (pengambok) and bunch planting (pan-bubu). Broadcasting is often on high cost hoof-plowed and spaded valley fields levelled with a luisan. After overnight watering, broadcast seed is rewatered, ignored for 2 weeks, drained, rewatered and weeded until harvest. Bunch planting involves one hour seed soaking, stirring into slurry, planting, watering after a little growth and weeding twice, a technology imported in 1950. Broadcasting embodies the whole region.

      In Sumatra, both methods predate transplanting (Poniman 1988:1-2, 25-7), while Java broadcast persisted in late 19th century. Damu (1990:64) said W Javanese bunch plant in dry and wetland, inserting 3-4 seeds per hole. Different regions call it sawur, tandur acret, sebar tu ur, tandur kunclongan, etc. Wetland bunch planting is in Waru Jajeng, Grise, Pekalongan, Krawan, Sidayu and "common in Java wetland before irrigation", according to De Bai, who said it was in 20% of wetland in an early 20th century Encyclopedia of Dutch Indonesia. He said bunch planting is very common throughout the region."

      I don't know if rice bunch planting or broadcasting is Malayan because the latter is in Philippine wetland. As bunch planting is in dryland, broadcasting in wetland and transplanting in paddy, I can't say if broadcast is from bunch planting in dryland or transplanting. Furukawa (1982:64) thought its paddy use resulted from dryland bunch planting, but dryland bunch planting tools are Chinese. As paddy broadcasting complexity suggests importation, it likely originated elsewhere. The Malayan example implies broadcasting from dry to wetland because it and bunch planting co-occurred, with later import of the Indian technique. It is unclear how broadcasting was imported, but it is ancient in the SE islands. Thus, it is not unreasonable to say wetland broadcasting is ancient.


    5. Multi-planting and Dibble transplanting
    6. I discussed brushhooklike tool preparation and rice seedling transplant in the last section. Multi-planting transplants rice seedlings first to nursery, then to second and third fields, involving dibbling and wood seedling beds. After E Sumatrans cut weeds with a tajak or parang, they seed the first field (samir) in highland or riverbanks nearby, putting soil in a 100x500x2-3 cm palm leaf box or nursery covered with plantain leaf. When rice is 10cm high 10 days later, they coil the box and carry it to the second field (lecak), dividing in 20-30 pieces for transplant. After another 40-50 days and 60 cm growth, seedlings are transplanted to a prepare field (ladang). Leaves are cut and 4-5 seedlings inserted in holes 30 cm apart; i.e., 10 kg rice/5 sq.m samir, from which they are retransplanted to a 280 sq.m lecak. Lecak seedlings can be retransplanted to a 0.33-0.5-hectare ladang.

      Some highland people substitute a 1 m high soil-covered wood platform for the first field; others float a soil & palm leaf-covered platform on water (samir or palian), transplanting seedlings with a dibble (tanjan or penyucuk), rather than their hands because the brushhooks leave many abrasive roots.

      S Carimandan rice seedling dry fields are taradak, wood platforms palaian and second and third fields ampak and lacak. Dryland seeds are bunch-planted, a hole dug every 15 cm with 50 seeds, with 5 kg seed/hectare (Noorsyamsi & Kidayat 1974:9-11). W Malayan wetland imported this method from Carimandan and is common (Jackson 1972:81, 91) and on coastal paddies, but unconfined to wetland. It has been in low Sumatran and Malayan paddies since ancient times, where N coast Sumatrans transplant with a kind of dibble (kuku kambing). The dibble is common in Malaya (Ho 1967:54). Double transplanting in some Sumatran wetland involves 7-8 cm holes, with 20-day seedlings transplanted to paddies. Elsewhere, first field seedlings are transplanted to a higher second field, and a month later to a prepared field, depending on water conditions. Common in Sumatra, the Mingulu transplant seedlings in mountain valley paddies. After seeding the first field, seedlings are transplanted to the second field 2-4 weeks later (2-30/hole) and retransplanted in the prepared field after 1-2 weeks (Poniman 1988:86-87). In sum, plowless rice cultivation by multi-transplanting is unconfined to coastal wetland, but occurs in lower wetland; a very ancient technique.

      Until the early 20th century (Da Mu 1990:65), lowland Sumatrans, Carimandans and Javan paddy farmers (De Bie 1901:40) also used the wood platform. Javans put Ipomoea aquatica weed on wet earth, cover with rice straw and then soil. A few make a seedling bed on the platform. Philipinos still use a technique dapog to transplant seedlings, using palm leaves to make seedling beds in a 1.5 m wide box, its length relating to prepared field needs. Seedlings are transplanted after 10-14 days, a method identical to the Sumatran.

    7. Casting seedlings in nursery

Although SE Asian islanders bunch plant, they also broadcast rice seed, a method distinct to paddy or dry field. I now introduce a method seen in 1979 in W Java using rice ears placed on the field to grow into seedlings. I saw seedlings a few cm high in flooded paddy, ears immersed, later seen in W and central Javan mountain paddies.

Wushi Lan (1984:34) said sticky rice is differentiated from japonica in W Java by broadcasting japonica (seba), while sticky rice ears are immersed, with seedlings put in beds (tebar). Villagers say they plant ears (ngoe rit) because it is simple, while gummy soil makes individual seeds difficult to remove. People usually immerse the ears, then plant them in the seedling bed (Yu Jing 1944:5). De Bie (1901:40-41) said locals broadcast (sebar) and plant ears (oeritan, ranggeujan, tebar). Purposes for planting opened ears with root in the soil to prevent removal are to protect seed from rain, prevent rice mixing and honor rice (place root in soil), as locals think it disrespectful to use hands or feet to pull rice from ear.

The N Luzon Yifujia (Conklin 1980:20) and Bangdu (Jenks 1905:97) also used this method, the former preparing their fields carefully by irrigating with warmer water from small ponds, using much fertilizer and plowing. After draining, rice ears are placed in the field like in Java, then rewatered 4-5 cm. The Bangdu repeated the Yifujia method, but left several inches between ears.

Ancient ear planting was in Java, Bali and Luzon, but other regions are hazy. It remains common in SE Asian islands, but is ubiquitous elsewhere. It is also very common where harvested ears are stored; i.e., ears are central from planting to harvest, a distinct trait in Malayan rice cultivation. (to be cont'd.)