A year on a hillside above a thousand metres is not a season and a crop. It is a schedule with almost no slack in it. The same household grew upland rice, maize and opium poppy in the same twelve months, and two of the three wanted the same weeks. The festivals were not laid over that schedule as ornament. They were cut into it — four days here, twelve days there, in the gaps the crops left — and while one ran, nobody went to the fields.
First, the names
ชาวเขา chao khao /chaaw kʰǎw/ — ชาว chao “the folk of” + เขา khao “hill, mountain” — is a Thai administrative category, not a people. It was the frame the state worked in: the 1971 study quoted throughout this page opens by placing the Akha among “the six major tribal groups included within the research program of the Tribal Research Centre located in Chiang Mai”. Six groups is an office's number, not a hillside's.
So the peoples get their own names here: Hmong (ม้ง), Akha (อาข่า), Lisu (ลีซู), Lahu (ลาหู่), Iu Mien (เมี่ยน), Karen (ปกาเกอะญอ), Lua (ลัวะ). Their calendars are not the same calendar.
ฝิ่น fin /fìn/ is the poppy and its gum both. Tradition holds — the word belongs to the Old-World chain that also gave English opium: Greek ὄπιον ópion, a diminutive of ὀπός opós “plant sap”, through Arabic and Persian أفيون afyūn. Which leg of the trade carried it into Thai, I have not pinned down and do not assert. กาแฟ kafae /kaa-fɛɛ/ came the short way, from French café, itself from Arabic قهوة qahwa.
The year, drawn
| Month | Upland rice | Maize | Opium poppy | Coffee | Ceremony |
|---|---|---|---|---|---|
| January | — | — | lancing and collecting | picking | Lisu New Year (lunar) · Lahu New Year, twelve days, no fixed date |
| February | — | — | lancing ends; field cleared | picking ends · flowering | Lisu & Lahu New Year |
| March | cut and dry the swidden | — | — | prune, mulch | — |
| April | burn, then sow | plant | — | prune · cherries set | — |
| May | sow, first weeding | weeding | — | cherries filling | — |
| June | weeding | weeding | — | cherries filling | — |
| July | weeding | first cobs | — | cherries filling | — |
| August | rice maturing | cobs in; stalks hoed under | field prepared in the maize plot | cherries filling | Akha Swinging Ceremony — four days, nobody works the fields |
| September | HARVEST | — | SOWING | cherries filling | — |
| October | harvest ends | — | sowing must be finished | cherries ripening | ~Akha rite to drive out sickness |
| November | thresh and store | — | weeding seedlings | PICKING BEGINS | Hmong New Year — noj peb caug, after the harvest |
| December | — | — | weeding; flowering begins | picking | Hmong New Year |
Why the poppy fitted the hill
Not romance, and not appetite. Fit.
It grew where the state was not, and where little else paid. Poppy went in above roughly 1,000 m; a British consul watching plantations in 1905–06 put them at five to six thousand feet. That is the band above the wet-rice valleys, and until roads arrived it was a band with no way to move a bulky crop out.
It followed the maize out of the same field. The maize cobs come in through July and August; in August the dry stalks and the weeds are hoed under, and the poppy is sown into that ground in September and October. Maize and poppy are one succession, not two crops. It also lasted: a rice swidden was good for two or three years, a maize field for something like eight.
It wanted the months nothing else wanted. Flowering in December, lancing through January and February — the cold, dry weeks when the rice is already threshed and stored and the next swidden has not been cut.
It was money that walked. Light, non-perishable, and bought at the village by a trader who came to you. Nothing else grown up there in 1960 had those three properties at once.
It was also the medicine chest. In the highlands opium was used against aches, pain, fever, diarrhoea and coughs — which is to say against exactly the things that kill people a long walk from a clinic. The manuscript tradition catalogued on this site has its own medical genre, written down in the valleys; this was the highland equivalent, and it was a crop rather than a text.
And it cost the household its address. Geddes argued that the commitment to poppy was the main reason the Blue Hmong of 1960s Thailand moved so often between settlements — not restlessness, but the search for new poppy ground. Inference — the crop that paid best was also the crop that exhausted soil you could not replace, so the calendar was stable and the map was not.
The labour, counted
Rice and opium have been measured at about the same intensity, roughly 220 person-days per hectare each; maize at about 80. Inference — that is the crunch in one line. Two crops of equal appetite, whose peak weeks are eight weeks apart at best and overlapping at worst, grown by the same hands.
The lancing
The harvest is the reason the whole calendar bends. A poppy capsule is scored shallowly — a millimetre or two, with a small multi-bladed knife — and the latex that beads out is left to thicken overnight and scraped off at first light. One capsule is worked several times, a couple of days apart, so a field is not harvested once but walked repeatedly for weeks, at dawn, by everyone who can walk it. Cut too shallow and nothing comes; too deep and the pod is spoiled. That is a skilled job with a short window and no way to hire your way out of it in a village of forty houses.
Inference — carry that shape into the coffee section below. Repeated passes, ripeness judged by eye, a window measured in weeks, every hand in the household. The crop changed. The shape of the work did not.
The festival months
August — the Akha Swinging Ceremony
It is always in August, the tenth lunar month of the Akha calendar, and it runs four days. But it does not fall on a date. It falls on the village headman's own auspicious animal-day, and each year the old men meet at his house to work out which day that is. In Saen Chai village in Mae Chan district in 1967 the headman's day was the day of the buffalo, so the four days ran ant, buffalo, tiger, horse — the ceremony starting one day before his day so that his day is the day the swings go up.
The ceremony is held for the maturing of the planted rice, and it belongs to a female god: Umsahyeh, whom the Akha regard as their first female ancestor and who is credited with starting it. (The scan of the 1971 article renders her name three ways — Umsahyeh, Umsahyee, Umsahyeb — and I am not going to silently pick one.) Because the god is female the ceremony belongs to the women, who wear everything they have made that year. It has since been called the women's new year.
A girl's passage into womanhood is measured in these ceremonies, one piece of dress at a time: at fifteen her hat is decorated with red and white beads; then the brassiere; then the belt; then, at the fourth ceremony, the high headdress. Four steps, four ceremonies. You cannot grow up faster than the calendar.
The village's great swing stands all year and nobody may cut or chip its posts: the fine is one pig, paid to the headman. The small swings built for the children come down when the four days end.
And twice, plainly, in the ethnographer's own words: no one goes to the fields during the ceremony. Nobody works the fields.
November–December — Hmong New Year
noj peb caug — noj “eat” + peb “three” + caug “ten” — “eat thirty”, for the thirtieth and last day of the twelfth lunar month, the end of the harvest year. It is held once the crop is in and stored, and it is explicitly the point at which people rest.
January–February — Lisu and Lahu New Year
Lisu New Year (Nyi Ma) follows the lunar calendar, so late January or early February, and opens with a sombre ancestor rite before anything else. Lahu New Year (Kho Jouw We) runs twelve days and has no fixed date at all — it happens when the harvest is finished.
How a date gets set: the twelve
The Akha count in twelves. A year has twelve months of thirty days; a month has five weeks of five, six or seven days; the days run in a cycle of twelve animals, and the same twelve name the years. Printed exactly as the 1971 article prints them:
- 1 the day of the ant
- 2 the day of the buffalo
- 3 the day of the tiger
- 4 the day of the horse
- 5 the day of the rabbit
- 6 the day of the termite
- 7 the day of the mule
- 8 the day of the giraffe
- 9 the day of the monkey
- 10 the day of the chicken
- 11 the day of the dog
- 12 the day of the pig
Read that list against the twelve this site already knows — the cycle behind the phrommachat and the horā almanac, in which the horse is the seventh year and the elephant the twelfth, and in which the horse carries a whole literature of its own.
Inference — this looks like the Sino-Tibetan twelve reworked by people who kept the count and swapped the animals for the ones on their own hillside: an ant where the rat goes, a termite where the dragon or the serpent goes, a mule where the goat goes. The giraffe at position eight is odd enough that I would want it checked against an Akha-language source before anyone builds on it — it is far more likely to be a 1971 gloss of a local word than a giraffe.
So an Akha ceremony has a month from the farming year and a day from the animal cycle, two registers running independently — which is how the four days of the Swinging Ceremony can be named “ant, buffalo, tiger, horse” and still be in August. And there is no rat in the list.
The rat-catching season
Here is what the literature does support, and it is enough to say the season is a mechanism rather than a quirk. Upland rice in this region suffers rodent outbreaks driven by bamboo masting: a bamboo stand flowers gregariously, seeds enormously, the rats eat the seed and multiply, the seed runs out, and the swollen population moves onto the rice. Trap-barrier systems have been studied against exactly that. Separately, and routinely, field rats are taken in Thai fields with bamboo bow-and-deadfall traps, the catching pegged to the harvest, and eaten — a lean, high-protein meat, and in the north a seasonal one.
Inference — put those together and a rat-catching season is not a hunting hobby squeezed between crops. It is crop defence that happens to feed you, at the one moment it works: when the rice is cut, the cover is gone, the stubble is open, and the animals that have been eating the harvest are concentrated, fat and findable. A calendar drawn by farmers would mark that. A calendar drawn by an agronomist would call it pest control and put it in a different chapter.
To close it: photograph the panel, front on, with any caption; ask the museum what it was drawn from. A chart of that kind most likely descends from the Tribal Research Centre in Chiang Mai, whose survey work is named in the 1971 article above. Until then this section stays marked as open, and no later pass should quietly promote it.
Then coffee
What replaced the poppy replaced its schedule almost exactly, and then broke something else.
The Royal Project began in 1969. Eradication did not begin until 1985 — the sequence is the point: the projects had to be earning the growers a living before the poppy was taken away, and removal was mostly negotiated rather than imposed. More than 150 replacement crops were introduced over the years, arabica coffee among them, along with tea, cabbage, apple and cut flowers. At Doi Tung, from the late 1980s, Princess Srinagarindra's project gave residents land-use titles and backed coffee and macadamia. Poppy went from 12,112 hectares in 1961 to 281 in 2015, a 97 per cent fall between 1985 and 2015, and it has not come back. Coffee, for its part, took more than three decades to become worth growing in Chiang Rai.
Arabica in the north now runs across Chiang Rai, Chiang Mai, Lampang, Mae Hong Son and Tak, between about 800 and 1,600 m. It flowers in February, fills its cherries from April to October, and is picked from November to February.
What that did to the year
Inference — four things, and only the first is the one usually told.
- It took the same slot. Poppy was lanced January and February; coffee is picked November to February. Both are hand work, judged by eye, walked in repeated passes over weeks, at the cold dry end of the year. The peak of the highland year did not move.
- It vacated September and October. The crunch — sow the poppy while the rice must come in — simply stops existing, because coffee wants nothing at all in those two months.
- It put weight into November and December, where the poppy asked only for weeding, and where Hmong New Year sits. Whether picking now runs across the festival, or the festival is worked around, or neither — I do not know, and it is not answerable from a desk. It is answerable in a village in December.
- It changed the tenure, not just the crop. Poppy was an annual on a shifting field; the field moved, and the household moved with it. Coffee is a perennial that yields nothing for three or four years and then yields for decades. You cannot walk away from it, and you cannot plant it unless you are certain of still being there in year four. That is why the answer at Doi Tung was land-use titles and not merely seedlings — and it is the deepest change of the two. The calendar survived the substitution. The mobility did not.
What is not settled
- The rat block. Above. Needs the panel photographed.
- The giraffe at position eight of the Akha twelve, and the three spellings of Umsahyeh — both artefacts of a scanned 1971 page, both wanting an Akha-language check.
- One wheel, several years. The crop rings drawn here are the Hmong-type upland succession. The Akha, Lisu, Lahu, Iu Mien, Karen and Lua farmed years that overlap it without matching it, and one wheel is drawn where there were several.
- The October Akha rite against sickness is marked ~ because I have its approximate month from a Thai provincial source and no more.
- Whether coffee picking collides with Hmong New Year in practice, now, in a particular village.
- None of this is in the catalogue. The manuscripts this site is built on are Tai Yuan, Tai Lue and Tai Khün monastic material from the valley wats — law, liturgy, medicine, astrology, chronicle. Highland swidden agriculture is not in them, and the absence is structural, not an accident of collecting: a calendar kept on a hillside above the wat was never going to be in the wat's library. Search the corpus for the poppy and you will get nothing, and that nothing is the correct answer.
Where this comes from
Plain prose is from a source in this list. Inference — is my reasoning across them. Tradition holds — is hedged background. A month written ~ is one I could not verify.
- The Akha ceremony, the twelve-day cycle, the Tribal Research Centre, and both statements that nobody works the fields — Chob Kacha-ananda, “The Akha Swinging Ceremony”, Journal of the Siam Society 59.1 (1971), pp. 119–128, read from the Siam Society's own scan. Fieldwork at Saen Chai village, Mae Chan district, 1967.
- The rat-catching season on a hand-drawn year-wheel — my own visit to the House of Opium, Sop Ruak. Marked open above.
- The Hmong crop calendar (maize in April, cobs June–July, poppy ground prepared in August, sown September–October against the rice harvest, weeded November and December) and the maize–poppy succession, field lifetimes and labour intensities — summaries of W. R. Geddes, Migrants of the Mountains: The Cultural Ecology of the Blue Miao (Hmong Njua) of Thailand (Oxford, 1976), and of Gary Yia Lee on agriculture and Hmong society. Read at one remove, and worth reading at first hand before the numbers are used for anything.
- Altitude, opium as household medicine, and the 1985 cut-off — a 2026 study of rice and food security in the northern Thai highlands, working across Mae Hong Son, Chiang Mai, Chiang Rai, Lampang and Tak with Karen, Lua, Khmu, Hmong, Lahu, Akha, Mien and Lisu communities.
- Lancing method — agricultural extension literature on Papaver somniferum, described here historically and in outline only.
- Hmong New Year — noj peb caug and its position at the close of the harvest year, from Hmong-language vocabulary sources and community accounts.
- Lisu and Lahu New Year — Nyi Ma on the lunar calendar; Kho Jouw We, twelve days, timed to the end of the harvest.
- The substitution figures — Thai Ministry of Foreign Affairs, “From Poppy to Coffee”; the Doi Tung project record; UNODC-derived hectare series for 1961 and 2015.
- The arabica calendar — northern Thai producer and trade accounts: flowering February, cherries filling April–October, picking November–February at 800–1,600 m.
- Bamboo masting and rodent outbreaks in upland rice — trap-barrier research on rodent damage during masting events; field-rat trapping and consumption from Thai reporting.