Tibetan chang

What it is

Chang (ཆང་) is the traditional alcoholic drink of the Tibetan Plateau. It is most often made from qingke (青稞), the hulless barley grown at high altitude. The grain is cooked, mixed with a powdered starter and left to ferment as moist grain with no free liquid. Water is then added to wash the sweet, sour, alcoholic material out of the grain, and the cloudy liquid is strained off. The drink is usually described as mildly sweet and sour, grain-scented and low in alcohol. Its strength depends on the fermentation and on how much water is drawn through the grain.

Qingke is a cultivated form of Hordeum vulgare, called Tibetan hulless or naked barley in English. Its hull separates easily from the grain, which suits it to tsampa as well as to chang. The Tibetan word for barley is nas (ནས་). Chang is the drink and qingke is the grain. Qingke jiu is the Chinese term for highland barley alcohol. It can mean fermented barley wine, but it is also used for distilled spirit, so how a product is made has to be checked.

The starter is part of the definition. Studies of Tibetan barley wine use the term Zangqu (藏曲), including qingke xiaoqu, for a mixed starter of moulds, yeasts and bacteria. Ethnographic accounts record powdered starters under names such as changdzi. In neighbouring Himalayan traditions the starter cakes are called phab or phabs. These are not interchangeable, because the raw materials, plants, microbes and local techniques differ. What they share is a culture that breaks barley starch into sugar and supports alcoholic fermentation.

Traditional chang is made in two stages: a solid fermentation, then a water extraction. The cooked, inoculated grain develops sugar, alcohol and acid while little free liquid is present. It is then steeped in water and strained, and more water can be added to the same grain for further extractions of falling strength. Conventional beer separates sweet wort from the grain before it ferments, which chang does not, and qingke baijiu is distilled to concentrate alcohol, which chang is not.

Qiangqin, also written Qiangji, in Qonggyai County, Shannan, is one well-documented place. Government accounts describe household barley wine, inherited techniques and a modern brewery that makes new products alongside the traditional ones. Chang is best treated as a regional family of fermentations, each with its own barley, starter and extraction method, and not as one recipe used by every Tibetan community.

History

The history of chang is separate from the much older history of barley on the plateau. Genetic and archaeological work shows qingke has been grown in Tibet for at least 3,500 years. A 2018 study in Nature Communications reconstructed the crop's ancestry and suggested that domesticated barley reached southern Tibet from the west between about 4,500 and 3,500 years ago. That dates the grain, not the drink. Growing barley, malting it and fermenting cooked barley with a mixed starter are different skills.

Early Tibetan texts mention drinking barley alcohol. A Chinese Tibetology article discusses the Old Tibetan historical documents kept at Dunhuang, including an account in which an elderly minister from the reign of Songtsen Gampo prepares barley alcohol for a royal gathering. It also names Tibetan-period wooden documents from Miran that list provisions of alcoholic drink. These are stronger evidence of early brewing than a modern origin story, but they give no full chang recipe. Their dates, terms and translations need checking against the primary texts before an earliest century can be named.

Qonggyai keeps an origin legend about Princess Wencheng, who married the Tibetan ruler Songtsen Gampo in the seventh century. Local accounts say she made barley wine from local grain and spring water, introduced a method of saccharifying and fermenting together, and taught village women to make it. County and Shannan descriptions of Qiangqin tell the story. No dated seventh-century document describes her making chang. The marriage is history and the invention is legend.

A second Qiangqin story tells of a visiting Buddhist teacher who praised the village's wine and so helped give it its reputation and name. A 2019 county publication presents both stories as local explanations. They show oral tradition and do not date the wine.

Technical documentation is much more recent. Nyima Tashi's account, in the 2005 ICARDA volume Food Barley: Importance, Uses and Local Knowledge, describes cleaning and cooking barley, adding powdered starter, several days of solid fermentation and extraction with water. The volume came out of a 2002 workshop supported by ICARDA and FAO.

Qonggyai's present-day records show the tradition alive. County publications describe brewing displays and vessels at Qiangqin, and a December 2024 Tibet Daily report describes household production, named practitioners and a new local brewery. The Qiangqin company was set up in 2019 and was running experimental modern production by 2024. This is recent development and not proof that the recipe has stayed the same for thirteen centuries. A 2024 ethnography of Tibetan-speaking communities in Nepal records the social role of chang there, which shows a wider Himalayan brewing culture and not a single origin.

How it works

Chang needs starch conversion and alcoholic fermentation to happen together. Barley stores most of its carbohydrate as starch, and brewer's yeast cannot ferment starch directly. Cooking hydrates and gelatinises the starch so that enzymes from the starter can reach it. Fungal amylases cut the starch into dextrins and fermentable sugars. Yeasts turn some of those sugars into ethanol and carbon dioxide, while bacteria add organic acids and other compounds. In traditional production all of this happens inside the cooked grain, before water is added to extract the drink.

A study of sixteen Zangqu samples from seven places in Tibet found Saccharomycopsis, Rhizopus and Aspergillus among the main fungi, with Lactobacillus and Leuconostoc prominent among the bacteria. Of 57 isolated strains, Rhizopus oryzae M12 had particularly strong saccharifying activity and Bacillus amyloliquefaciens Q4 had high protease activity. These are strains from particular starters, not required ingredients of every starter.

The enzymes do different jobs. Alpha-amylase cuts the inner bonds of starch chains into shorter carbohydrates, glucoamylase releases glucose from starch and dextrins, and other enzymes deal with branched starch. Proteases break barley protein into peptides and amino acids, which feed the microbes and add to taste and aroma. How active these enzymes are decides how fast the grain softens and how much sugar becomes available.

Yeast does the alcoholic work. Saccharomyces cerevisiae turns glucose into pyruvate by glycolysis. Pyruvate becomes acetaldehyde, releasing carbon dioxide, and alcohol dehydrogenase then reduces the acetaldehyde to ethanol and regenerates the NAD+ that glycolysis needs. Yeasts also convert amino acids into higher alcohols by the Ehrlich pathway, and these alcohols react with organic acids to form esters. That is why the drink can smell of fruit and flowers as well as cooked barley.

The bacteria matter too. A 2020 study followed the microbes through a highland barley wine fermentation. Rhizopus and Saccharomyces were the main fungi throughout. The bacteria changed a great deal: after 48 hours Acetobacter made up 74.60% and Leuconostoc 13.53% of the bacterial sequencing reads in the sampled fermentation. Those are shares of the reads in one experiment, not shares of living cells and not a standard starter makeup. Acetobacter can make acetic acid where oxygen is present, and Leuconostoc and related lactic acid bacteria add organic acids.

Chang is more than a solution of ethanol. A 2020 sensory and chemical study of Tibetan qingke jiu, using four commercial highland barley wines, identified 66 aroma compounds, 17 of them above their calculated odour thresholds. They included phenylacetaldehyde, beta-phenylethanol, ethyl acetate, ethyl butanoate, methional, methionol, sotolon and furaneol. The wines tasted of fruit, cooked potato, honey, sweetness, sourness and caramel.

Barley also brings beta-glucans, proteins and phenolic compounds. Beta-glucans affect the thickness of the mash and how well it extracts, and they are not fermentable glucose until enzymes break them down. Sequencing shows which organisms are present, but culture and biochemical tests are needed to show what each one does.

Recorded sequence and what remains unknown

The 2005 account names barley, water and powdered changdzi. Barley provides the grain, the starter supports starch conversion and fermentation, and water is used in cooking and later extraction. Starter composition differs between producers and cannot be rebuilt from a list of organisms found in other samples.

Nyima Tashi's account in the 2005 ICARDA volume begins by cleaning and washing the barley. Some brewers prefer purple barley for its flavour and appearance, and many mix malting barley and hulless barley at a ratio of 1:2. The prepared brewing grain is called changdru. It is boiled for 2 to 5 hours, which softens the kernels and opens the starch to the starter's enzymes. The boiled barley, poub, is cooled before the powdered starter, changdzi, is mixed in.

The inoculated barley ferments for 3 to 5 days as grain, with no free liquid left. The fermented grain is called lenmar, and small amounts can be eaten. To make the drink, the lenmar goes into an earthenware pot, water is added and the mixture steeps for 6 to 10 hours. The liquid is then strained. More water added to the same grain gives three or four extractions.

A description of production in Qonggyai County follows a close sequence. Cooked barley is spread on woven bamboo while still warm so that it loses some moisture. A fermentation cake is ground to powder and mixed into the grain, which goes into a covered, insulated ceramic jar for about two to three days in summer and three to four days in winter. Water is then drawn through the grain repeatedly, giving a first, second and third wine of falling strength. Quantities, vessels and conditions differ from Nyima Tashi's account, so the two are kept apart.

A 2020 study of Tibetan qingke jiu describes another method: the barley is washed, soaked for about 24 hours, steamed and cooled to 30 °C before Zangqu is added, and then ferments in solid state for three days. This is a variant and not the household method above.

Each step has a job. Washing removes debris, soaking and cooking hydrate the barley, and cooling protects the culture from heat. Mixing spreads the starter through the grain, and the solid fermentation lets the enzymes work. Water added later carries sugars, acids, alcohol and aroma out of the grain, and each extraction takes more of the soluble material, which is why extractions differ in strength and flavour. Times, temperatures and quantities have to stay with the method they were recorded for.

Traditional descriptions give no standard starter dose, water quantity, grain temperature or fermentation end point. Modern producers in the area use controlled cooking, mechanical processing, clarification, cold storage and microbial stabilisation. These make the product more consistent and are not part of a historical household recipe.

Regional and historical variations

Chang differs by place, barley, starter and extraction. Qiangqin (Qiangji) village in Qonggyai County is a well-documented centre of household production, where each family's wine tastes different. A December 2024 report names Lobsang Dorje as a recognised practitioner and describes efforts to change sweetness, acidity and colour while keeping parts of the traditional method. These are modern changes and not a single historical formula.

In the Qonggyai account the first extraction is aromatic, sweet and full-bodied, the second is lighter and refreshing, and the third is weaker. All of them come from one fermented grain and not from separate fermentations. Nyima Tashi likewise records three or four additions of water and reports about 7% alcohol in the first extraction and 5% in the second. Those are observations, not specifications for every region.

Barley colour matters. Nyima Tashi records a preference for purple barley, and Qonggyai producers describe the wine's colour changing with the proportions of dark and pale qingke. One practitioner makes a wine the colour of red wine without brown sugar or artificial colouring, which is a modern product and not evidence that red chang was ever standard. Rice, wheat and mixed-cereal versions exist, but barley defines this entry.

Nyima Tashi describes sanchang as a western Tibetan relative made from tsampa, the roasted flour of highland barley. Tsampa is mixed with warm water and powdered starter, fermented for about three days, then cut and dried. Some people eat the semi-fermented pieces and others steep them in water to make a drink. It uses roasted flour and not the cooked whole grain, and it deserves its own record.

Across the Himalaya the names chang, chhang and chyang can mean a drink of barley, rice, finger millet or mixed cereals. Ladakh has barley brewing with local starters, including methods that use malted grain, and communities in India, Nepal and Bhutan also brew from rice and millet. A 2016 study of chyang used finger millet and mixed cereals with marcha starter, and its results cannot be applied to Tibetan qingke chang. Phab from Bhutan or nearby cannot be assumed to match a given Tibetan Zangqu either. A study of plateau starters reports a barley-and-pea starter at 7:3, and in Sapi, Ladakh, the phap starter is made from wheat flour and local plants, moulded and dried.

Tibetan cultural writing names zla-chang, month-aged wine, and lo-chang, year-aged wine, which involve keeping the fermented grain for longer before extraction. They show that ageing and storage varied, but no tested ageing schedule exists. Tibetan communities also distil spirit from fermented grain. A distilled barley spirit, including products sold as qingke baijiu, is not ordinary chang just because it uses the same grain.

Qiangqin's brewery, set up in 2019, uses a more mechanised line and has tried changes to processing, colour, clarity and refrigerated distribution. Laboratory studies elsewhere test other starter blends and controlled fermentation. These show the tradition developing and do not define one authentic modern recipe.

How it's eaten and used

Chang is mostly drunk as an alcoholic drink, often fresh and cloudy and not clarified and bottled like commercial beer. Depending on the fermentation and extraction it is sweet, slightly sour, grain-scented and light. Nyima Tashi reports about 7% alcohol in the first extraction and 5% in the second, and says people commonly drink chang of about 5% or less. Four commercial highland barley wines in a 2020 study measured 3.0 to 3.5%. These are examples and not a standard.

Chang is a customary offering to guests in Qonggyai and appears at gatherings, weddings and festivals. Government accounts describe etiquette in which cups are offered, the drink is taken in stages and a little wine is used in a symbolic gesture before drinking; Qiangqin Village is known for presenting three cups. These customs vary by community and occasion and are not required of every Tibetan or at every event.

Tibetan cultural writing documents chang in offerings, celebrations and communal exchange. A 2024 ethnography of Tibetan-speaking communities in the Nubri Valley of Nepal examines fermented and distilled drinks in household ritual, labour exchange, social relations and obligation. The authors also note that drinking alcohol is contested within Tibetan religious life, so chang has ceremonial roles in some communities without being accepted or required by all.

Barley wine is shared at Losar, the Tibetan New Year, at weddings and at household celebrations, and offered to honoured guests. Qiangqin Village has built tourism and cultural events around its brewing reputation. These practices show chang's importance now, not that each custom is unchanged since the Tibetan empire.

The fermented grain is food too. Nyima Tashi says lenmar is eaten in small amounts and can be fried in oil and served with sugar. He also describes changuel, made by boiling chang and tsampa with cooked rice, dried cheese and sugar for about ten minutes. Tsampa may also be mixed with chang as a soft grain dish. These foods keep more of the cereal solids than strained chang does.

Modern barley wine is sold in restaurants, to visitors and in packages. A stabilised bottle should not be assumed to have the microbial activity or serving character of freshly extracted household chang.

Claims that chang aids digestion, prevents altitude illness or has medicinal value need their own evidence. Qingke barley contains beta-glucans, phenolic compounds and other nutrients, but that does not show how much remains in strained alcohol or that drinking it helps anyone. Chang is an alcoholic drink whatever its grain.

Reading results and problems

A good fermentation brings several changes together: the grain softens, starch breaks into sugar, alcohol develops, and microbes make acids and aroma. A sweet taste or an alcoholic smell does not show that fermentation is finished or safe. Because chang is drawn out of solid grain, the amount of water added and the number of extractions change the strength and flavour of what is served.

The starter varies most. The Zangqu study of sixteen samples from seven Tibetan places found wide fungal and bacterial diversity, and isolated strains differed in saccharifying and protease activity. Dose alone therefore cannot guarantee the same result: a culture with weak amylase activity may release little sugar even if it holds living yeast.

Temperature affects microbes and flavour. The December 2024 Qonggyai report names too high a fermentation temperature as a cause of unwanted sourness, and says uneven cooking gives grain of uneven texture and overcooking can give cloudy or scorched results. Other producers say small differences in processing change bitterness and sourness. These are producers' observations and not tested limits, and no safe upper fermentation temperature is given for the local inherited starter.

Acid needs careful reading. Lactic acid bacteria can add mild acidity, and acetic acid bacteria can give a sharper, vinegar-like note where oxygen is present. A high share of Acetobacter after 48 hours in one 2020 fermentation shows an active bacterial population but does not show that the finished drink is spoiled. Measurements can describe a batch, but no safe threshold is established for this method.

No single smell proves safety. A harsh solvent smell, a strong smell of rot or unexpected growth on the surface calls for investigation, not for a longer fermentation or more sugar. Barley variety, starter and heating all change the normal aroma, so a defect has to be judged against the actual process.

Mixed starters can carry unwanted microbes next to the useful ones. Tibet's health authorities issued food-safety standards for brewing Zangqu in 2024 and for traditional highland barley wine xiaoqu in 2025, covering production, inspection and storage. These show that starter quality is regulated and do not make uncontrolled household mould safe. The hazards include unwanted bacteria, mycotoxin-producing fungi and contaminated grain, and finding Rhizopus or Aspergillus in a culture does not make an unidentified culture safe.

Renewed fermentation is a documented problem in Qonggyai. A 2024 account from the local brewery says one additive-free refrigerated product kept only 7 to 10 days and that fermentation starting again was a major technical problem. That is one producer's experience and not a shelf life for chang in general. Leftover carbohydrate and living yeast can start fermenting again, changing the sweetness and alcohol and making carbon dioxide, and a sealed container can build pressure.

Appearance does not show alcohol content. Within one traditional account, extractions from the same grain differ in strength, and industrial processing, longer fermentation and different water additions change it further. Chang should not be called alcohol-free or safe for people who avoid alcohol. Claims that it protects against cold, speeds recovery from drunkenness or has medicinal effects are cultural reports and not clinical evidence.

No universal finishing pH, minimum alcohol, pathogen limit or household shelf life exists for Tibetan chang. A fully measured laboratory fermentation with a commercial yeast gives technical guidance, but it cannot stand in for an inherited Qonggyai starter without new testing.