Ihwaju

What it is

Ihwaju (이화주, 梨花酒) is a Korean rice wine so thick that it is eaten with a spoon. Cooked rice is mixed with a powdered rice starter called ihwagok (이화곡, also written ihwaguk) and very little brewing water. Over a couple of weeks the mass softens and turns sweet, slightly sour and alcoholic. It can be thinned with water for drinking. Korean reference works class it as a takju, a cloudy rice wine.

The name means pear-blossom wine, but the drink is not made from pears or blossoms. The name marks the season. The starter was made beforehand, in the first or second lunar month depending on the recipe, and the wine was brewed around the time the pear trees flowered. Nothing floral goes into the vessel. The fermented material is rice.

The starter is what sets ihwaju apart. Ihwagok is traditionally made from non-glutinous rice that is soaked, ground, shaped into small cakes and left to grow a working culture. Some old recipes wrap the cakes in mugwort and others lay them among pine needles. Those plants belong to starter making, not to the wine. The finished starter carries moulds and yeasts and the enzymes that turn rice starch into sugar and start alcoholic fermentation.

The rice itself varies by recipe. Historical texts use non-glutinous white rice, glutinous rice or a mixture. Many grind the rice to flour, shape it into ring-shaped cakes called gumeongtteok and mix them with starter. A high share of rice and almost no added water give the heavy, creamy body. Ordinary makgeolli thickened to a similar consistency is not ihwaju, because its starter and its fermentation differ.

Ihwaju has the reputation of a luxury wine of courtly and well-off households, partly because so much rice goes into so little drink. That does not mean ordinary households never made it. Modern Korean breweries make versions from spoonable paste to a softer, yoghurt-like wine. The name covers a family of methods, not one fixed texture, strength or culture.

History

Ihwaju belongs to medieval Korean drinking culture. The Encyclopedia of Korean Culture finds it in Goryeo-period writing, including Dongguk Isanggukjip, tied to the 13th-century writer Yi Gyu-bo, and the song Hallim Byeolgok, generally dated to the 13th century. These texts show that a drink called ihwaju was known among Korean liquors. They give no method, and they do not prove that the thick, spoonable wine of today was already standard.

The earliest detailed method is in Sangayorok, compiled around 1450 by the Joseon physician Jeon Sun-ui. It starts the starter in late winter and brews the wine as pear blossom nears. The starter rice is ground to flour, formed into pieces the size of duck eggs, wrapped in mugwort and kept in a warm place. After rounds of resting and turning, the pieces are broken, dried and ground to powder. The wine is made from cooked rice cakes and that powder, then matured in an earthenware vessel.

Sangayorok describes a connected process rather than just naming a drink. It gives quantities in the old measures mal and doe: five mal of rice for the starter, ten mal for the wine, and a further instruction involving five doe of starter powder. These are volume measures that vary by period, and they cannot be turned into metric weights without careful work. The text has the brewer check the vessel after three or four days for built-up warmth and open the wine around the fifteenth day of the fifth lunar month.

Later Joseon writers kept the tradition going. Suunjapbang (16th century), Eumsik Dimibang (17th century, by Jang Gye-hyang), Sallim Gyeongje and its enlarged successor, and 19th-century works such as Imwon Gyeongjeji each preserve a method, and Eumsik Dimibang gives more than one. Some use pine needles in place of mugwort. Others change how the rice is prepared, how much starter goes in, or what season to brew in. The same name does not mean the same recipe.

A 19th-century manuscript, Juchan, gives a clear second recipe. The starter is made when pear blossom opens, with rice-flour lumps laid among pine needles and dried after seven days. The wine uses powdered starter and cooked ring-shaped rice cakes: one mal four doe of starter for a normal batch, or one mal six doe for a sweeter one. It differs from the 15th-century account in both timing and starter preparation.

Modern interest centres on preservation, research and revival. A 1993 study reproduced a traditional method and examined its organisms and enzymes. A 2020 Kyung Hee University thesis compared traditional and improved methods and found large variation between batches. Breweries in Gyeonggi Province now sell ihwaju. Their products show that the tradition lives on, but they are modern work and should not be read back into the manuscripts.

How it works

Ihwaju is made by saccharification and alcoholic fermentation running together. Rice holds starch, not sugar. When rice flour is mixed with hot water and cooked, the starch granules swell and open up, so enzymes can reach them. The organisms in ihwagok make those enzymes and cut the starch into smaller carbohydrates and then into fermentable sugars. Yeasts turn part of that sugar into ethanol and carbon dioxide. Both jobs go on in the same thick mass, so sugar is being made and eaten at the same time.

A 1993 study of a reproduced ihwaju starter found Aspergillus oryzae and a yeast recorded as Hansenula sp. as the main cultured organisms, at about 1.2 million colony-forming units per gram each. That describes one starter, not every ihwagok. Hansenula is an older genus name whose species have since been reclassified, so the study does not identify a modern species. A 2020 thesis made ihwagok that held wild yeasts and lactic acid bacteria, which shows that different starter methods can grow different microbial communities.

Aspergillus oryzae is the enzyme-maker. Its alpha-amylase breaks the long starch chains into shorter dextrins, and other amylases, such as glucoamylase, release glucose that yeast can ferment. The 1993 study measured alpha- and beta-amylase and still found activity in wine matured for a year. Moulds in Korean rice starters also produce proteases, which free peptides and amino acids from the rice protein and feed the yeast.

The alcohol comes from yeast. Glucose passes through glycolysis to pyruvate, pyruvate decarboxylase turns pyruvate into acetaldehyde and carbon dioxide, and alcohol dehydrogenase reduces the acetaldehyde to ethanol. That last step recycles the NAD+ glycolysis needs to continue. Yeasts also make small amounts of higher alcohols and esters that add aroma. Lactic acid bacteria, where present, add organic acids and compete with other microbes. Which acid-formers live in historical ihwagok is not established, so organisms found in ordinary wheat nuruk should not be assumed here.

The small amount of added water shapes everything. Dense rice solids make a paste, not a free-flowing mash. As the enzymes work, the rice softens and soluble sugar builds up, and the texture reflects the balance of remaining starch, dextrins, water, ethanol and suspended rice. Little free liquid affects how nutrients and oxygen move and how heat leaves the mass. Old instructions to move or cool the mass when it grows warm fit that picture, since active microbes give off heat, but they are not a validated temperature-control method.

A 2020 Kyung Hee University thesis measured 11.86 to 13.53% alcohol by volume and 156 to 182 mg/mL reducing sugar in traditional-style test batches. A two-stage method gave 12.58 to 13.30% alcohol and 164 to 183 mg/mL reducing sugar. So ihwaju can be both strong and quite sweet. The spread between batches shows how much starter activity, water handling and method affect the result.

How it's made and why each ingredient and step matters

Traditional ihwaju starts with the starter. In the Sangayorok method, non-glutinous white rice is soaked overnight, ground to a fine flour and kneaded into firm pieces the size of duck eggs. The pieces are wrapped in mugwort, laid in a straw container in a warm room and turned at intervals. Then they are broken up, dried, ground and sieved. Making the starter ahead of time gives the enzyme-producing fungi and other organisms time to grow on the rice.

Mugwort and pine needles appear in different starter methods. Sangayorok uses mugwort. Juchan lays the rice-flour lumps among pine needles in a steamer kept somewhere cool. The plants help to separate the pieces and are part of the microbial surroundings. Their exact effect on which microbes grow, or on flavour, has not been tested for every method. They are not interchangeable preservatives, and their presence does not make a starter safe.

The rice for the wine is washed, ground and cooked separately. The classic shape is gumeongtteok, a ring-shaped cake with a hole in the middle. Cooking gelatinises the starch so that the starter's enzymes can reach it. Depending on the text, the cooked cakes are cooled or worked while warm, then kneaded with the powdered starter until it is evenly spread. Historical recipes add little or no free water to the fermenting mass. Water is still needed to wash, soak and cook the rice. The difference is between water used to prepare the rice and water added to the mash.

Sangayorok sets the wine out in detail. The starter begins with five mal of rice in the early second lunar month. The wine uses ten mal of white rice, made into fine flour and cooked cakes, and a later step combines rice with five doe of starter powder. The pieces go into an earthenware vessel with a space left in the middle. After three or four days the brewer checks for built-up heat and, if needed, takes out the mass, cools it and puts it back. The wine then rests somewhere cool until its opening date.

Juchan differs. It uses one mal of rice for the starter and two mal for the wine, with one mal four doe of starter powder, or one mal six doe for a sweeter result. The rice is made into cooked ring-shaped cakes and mixed with the starter while warm. The starter spends seven days among pine needles instead of Sangayorok's longer mugwort wrap. These are separate recipes, and their figures should not be merged.

The modern recipe from Korea's Rural Development Administration, which is the recipe given here, simplifies the process and starts from a ready-made starter. It uses 1 kg of non-glutinous white rice, 300 g of ihwagok and 1 g of yeast. The rice is ground fine, because a coarse flour would spoil the smooth texture a spoonable wine needs. A hot-water dough is shaped into ring-shaped cakes, crushed and cooled, then mixed with the powdered starter and held in a cloth-covered earthenware vessel at 15 to 20 °C for 15 to 20 days. The cloth keeps dust out and lets carbon dioxide escape. The recipe leaves out the water quantities, the point at which the yeast goes in, and whether the cakes are boiled or steamed. Historical recipes do cook them, so the gap is worth settling before you brew.

A 2020 study changed the process for consistency. It first grew the microbes in a wetter mixture of ihwagok and water, then added more rice for the alcoholic stage. This gave more even results than a copy of the traditional method. It is a modern adaptation, not what Sangayorok describes.

Regional and historical variations

Ihwaju recipes differ more by manuscript, household and period than by modern province. Medieval writing gives the name. Joseon cookbooks give steadily more detail. Sangayorok, Suunjapbang, Eumsik Dimibang, Sallim Gyeongje, Jeungbo Sallim Gyeongje, Haedong Nongseo and Imwon Gyeongjeji vary in starter preparation, rice treatment, starter quantity and maturation. Each is a separate witness. None is the original Korean method.

Sangayorok, compiled around 1450, makes a relatively elaborate starter from non-glutinous rice flour shaped into small pieces and wrapped in mugwort. The starter develops over successive seven-day periods and is then dried. The wine uses cooked rice cakes, cooling and maturation. The early 19th-century Juchan recipe grows its starter for seven days among pine needles before drying and grinding it. The starter stage became shorter in the later text. That may be linked to changes in season and production method, but no text proves the reason.

Eumsik Dimibang, a 17th-century household cookbook associated with Jang Gye-hyang, holds several ihwaju methods. They differ in the ratio of rice to nuruk, how the rice dough is made and how the fermented mass is handled. One calls for 3 or 5 doe of powdered nuruk per mal of rice. One adds a good deal of glutinous rice for sweetness, and one allows brewing outside pear-blossom season. The Korea Traditional Knowledge Portal also keeps a version in old spelling, 니화듀. So neither the rice mixture nor the brewing season was fixed.

Sallim Gyeongje describes finely ground rice formed into ring-shaped cakes, cooked and cooled, mixed with powdered starter and packed into a jar. The Encyclopedia of Korean Culture reports that it separates a sweeter wine, made with more starter, from a clearer, stronger one, made with less. These are historical descriptions, not tested targets. The figures in different manuscripts cannot be safely combined by picking the most convenient ones.

Modern producers add more variation. Contemporary breweries make a dense, yoghurt-like ihwaju sold in small jars or bottles. The Sulseam brewery in Yongin makes one from Korean rice and a traditional starter, labelled 8% alcohol by volume. Other makers may use selected cultures, controlled temperatures, different rice proportions or a changed maturing schedule. These are modern interpretations, not evidence about how any old ihwaju was fermented.

Ihwaju should also be told apart from other thick Korean rice wines. Home-brewed creamy wines may use wheat nuruk, other rice starters or other methods. What defines ihwaju is its link to the pear-blossom rice starter and the historical brewing methods, not just a spoonable texture. How each variant was spread across Korea is not well recorded.

How it's eaten and used

Ihwaju is eaten with a spoon. A well-made thick batch looks like a smooth rice paste or a dense, creamy porridge, not a pourable wine. It is often compared with yoghurt, though it contains no milk and is not a dairy ferment. The flavour combines rice sweetness, alcohol, fermentation acidity and the aroma of a rice starter. Depending on the recipe, it can taste sweeter and stronger than ordinary makgeolli.

Korean reference works and the modern recipe also describe thinning ihwaju with cold water before drinking, especially in summer. That makes a lighter drink that keeps the flavour of the fermented rice but changes the sweetness, acidity, strength and body. No standard dilution is recorded, so any ratio is a matter of taste. Some modern products are already soft and creamy and can be eaten as they are.

Ihwaju was tied to seasonal brewing around pear blossom. Korean historical writing treats it as an expensive rice wine, because a lot of rice went in and little water. It appears in Goryeo literary works and Joseon household manuscripts, which marks its place in Korean drinking culture. It is often linked with aristocratic households, but that is no evidence that only the court drank it.

Today ihwaju is sold as a traditional alcoholic speciality and presented as a spoonable, dessert-like wine. Sulseam packs it in small jars and suggests serving it with fruit and biscuits. Strength, starter and consistency differ from one product to the next, so check the label before assuming a serving method or a strength.

No standard list of dishes cooked with ihwaju is recorded, and no pairings are documented. Pairing it with fruit or desserts is a modern habit, not a feature of the old recipes. Its sweetness does not make it alcohol-free, and its creamy look is not a reason to treat it as yoghurt, rice pudding or a plain dessert.

Reading results and problems

You can follow an ihwaju fermentation by how the rice texture, sweetness, acidity and alcohol change, and by how active the starter is. As amylases break down the cooked starch, the rice mass softens and soluble sugar builds up. Yeast then consumes some of that sugar and makes ethanol and carbon dioxide. A sweet taste does not mean fermentation is finished, and a thick body says nothing about alcohol. The result depends on the rice, the starter's activity, the water content, the microbes present and how the wine is matured.

Starter quality is the biggest source of variation. In the 1993 study, Aspergillus oryzae and Hansenula sp. were isolated from the starter, and amylase activity was still measurable in wine aged for a year. The authors reported that this experimental product kept well without pasteurisation or preservatives. That is an observation about one product. It is not a shelf-life rule, and it does not show pathogen control, safe packaging or the safety of home-grown mould cultures.

The 2020 Kyung Hee thesis shows how uneven the fermentation can be. Traditional-style batches differed in saccharifying activity, alcohol, total acidity and reducing sugar, and alcohol ran from 11.86 to 13.53% by volume. A two-stage method, with extra water in a first growth stage and more rice added afterwards, gave steadier results. A historical recipe cannot be expected to behave the same way with a starter of unknown strength.

Watch for heat and for runaway starter growth. Old recipes have the brewer examine the mass when it warms and move or cool it. Temperature and mould growth were clearly practical worries for old brewers, but nothing in the historical recipes tells safe moulds from unsafe ones. Studies of Korean nuruk have found Aspergillus flavus alongside the useful fungi, and some Aspergillus strains make aflatoxins. A 2020 survey found raised total aflatoxin in 14 of 61 nuruk samples. Do not treat arbitrary mould as a substitute for a verified food-grade starter.

Cooked rice brings a different risk. Bacillus cereus spores occur in rice and survive cooking. If contaminated cooked rice is held somewhere that allows growth, some strains make toxins, including the heat-stable cereulide. The usual advice to chill cooked rice quickly cannot be applied directly to a long fermentation, which needs its own proven control. The historical recipes give no modern controls for pathogens, starter purity, safe cooling or toxin prevention.

Keep active fermentation and finished storage apart in your mind. Leftover sugar and live microbes can keep making alcohol and carbon dioxide after packaging, and cold slows this without necessarily stopping it. The studies set no universal finishing pH, safe alcohol level, pressure limit or storage time. If you see unexpected surface growth, a rotten smell or visible contamination, do not scrape it off or add more starter. Discard the batch.