Tungrymbai

What it is

Tungrymbai is a traditional fermented soybean food of the Khasi and Jaintia communities of Meghalaya, northeastern India. Whole soybeans are cooked until soft, placed in leaf-lined baskets and allowed to ferment naturally. The beans become sticky, soften further and develop a strong, pungent smell. After fermentation they may be pounded into a coarse mass and cooked with seasonings, particularly ginger, garlic, chilli and black sesame. Tungrymbai can therefore refer both to the fermented soybean preparation and to the finished cooked dish made from it.

The fermentation belongs to the alkaline bacterial family of soybean foods. It is related in broad microbiological terms to Nepalese kinema, Japanese natto, Korean cheonggukjang and other fermented soybeans of northeastern India. Its main organisms are Bacillus bacteria rather than the moulds used for tempeh or certain Chinese soybean ferments. Fermentation develops flavour primarily through protein breakdown, while some strains produce extracellular substances responsible for the sticky texture.

A 2015 study by Chettri and Tamang examined thirty-nine tungrymbai samples collected from villages and markets in Meghalaya. Of the 211 Bacillus isolates identified from those samples, approximately 55% were Bacillus subtilis, 25.5% Bacillus licheniformis and 19.5% Bacillus pumilus. These percentages describe cultured isolates, not the exact abundance of each species in every batch. No evidence establishes that traditional producers deliberately inoculate beans with a fixed mixture of these three organisms.

The principal raw material is soybean, Glycine max. Traditional production uses locally available soybean varieties, bamboo baskets, leaves and a cloth or jute covering. Researchers have identified leaves called lamet, described botanically as Clinogyne dichotoma or Phrynium pubinerve in different publications. Leaves and equipment may contribute microorganisms, but their influence varies and does not amount to a controlled starter-culture system.

A characteristic feature is the treatment after fermentation. Unlike natto, which is commonly eaten with its beans substantially intact, tungrymbai is often mashed and subsequently cooked. The prepared food may be a thick savoury mixture rather than a loose mass of beans. Black sesame is especially important in familiar Khasi and Jaintia dishes, although a contemporary account from Pasyih distinguishes plain fermented tungrymbai from a more expensive cooked version containing black sesame.

History

Tungrymbai belongs to the indigenous food traditions of Meghalaya, particularly the Khasi and Jaintia communities. Its preparation has historically been associated with household work, local markets and the transmission of practical knowledge between generations. Researchers describe the role of women in preparing and selling fermented soybean foods throughout northeastern India, including tungrymbai. The first community to make tungrymbai and the century of origin are not known.

A Meghalaya Biodiversity Board account records interviews with residents of Pasyih village, West Jaintia Hills. The report says that the village's inhabitants have practised tungrymbai fermentation for centuries. This is valuable evidence of local understanding and oral tradition, but it is not a securely dated historical attestation. The document does not present a dated manuscript, archaeological specimen or continuous sequence of production records establishing a particular starting century.

One early substantial scientific record is the 2006 study by Agrahar-Murugkar and Subbulakshmi, which investigated the preparation techniques and nutritional composition of fermented foods eaten by Khasi communities. The researchers examined tungrymbai alongside fermented bamboo and fish products. Their work documented tungrymbai as a locally important protein-containing food and provided laboratory measurements of its composition. This establishes a published account of the tradition without suggesting that the preparation was newly developed.

Sohliya and colleagues investigated tungrymbai in research submitted in 2007 and published in October 2009. They examined traditional processing materials, including raw and soaked soybeans, leaves and wooden grinding equipment, to identify microorganisms associated with spontaneous fermentation. A separate 2009 paper by Tamang and colleagues documented indigenous women's knowledge of fermented soybean production and described the stages of tungrymbai preparation.

Research expanded during the following decade. A 2012 study examined microbial and chemical changes between fermented and subsequently cooked tungrymbai. In 2014 and 2015, Chettri and Tamang investigated functional properties and Bacillus diversity in samples of tungrymbai and bekang. Researchers at North-Eastern Hill University subsequently experimented with selected lactic acid bacteria to standardise aspects of fermentation and investigate storage behaviour. Those trials are modern research adaptations rather than evidence of the traditional organisms being deliberately cultivated.

Contemporary village production remains economically important. A July 2026 NESFAS report describes Pasyih as a recognised centre of tungrymbai production and records sales to markets in Jowai, Shangpung and Khliehriat. The report describes producers boiling beans, fermenting them in baskets, pounding them and preparing a black-sesame version. This provides direct modern documentation of an ongoing practice. Claims of prehistoric invention, an exact ancient recipe or a single origin point remain unsupported.

How it works: bacteria, enzymes and chemistry

Tungrymbai is predominantly an alkaline bacterial fermentation. Thorough cooking hydrates and softens the soybean material, denatures proteins and reduces many of the vegetative microorganisms originally present. Some bacterial spores may survive cooking, while additional organisms can enter through handling and processing materials. During the subsequent warm fermentation, suitable Bacillus bacteria multiply and produce extracellular enzymes that act on the soybean substrate.

Proteolysis is central to the transformation. Soybeans contain storage proteins, especially glycinin and beta-conglycinin. Bacillus proteases cut these large proteins into smaller peptides, while peptidases release free amino acids. This increases the pool of soluble nitrogenous compounds and generates substances associated with savoury taste. Further amino-acid metabolism can release ammonia and contribute to alkalinity and pungent aromas. These reactions distinguish tungrymbai from a primarily lactic-acid fermentation, where accumulating organic acids drive the pH downward.

The 2015 Chettri and Tamang study found an average Bacillus population of approximately 8.2 log colony-forming units per gram in the investigated tungrymbai and bekang products. Among tungrymbai isolates, Bacillus subtilis was the most frequent identified species, followed by B. licheniformis and B. pumilus. The researchers used molecular identification techniques alongside phenotypic tests. Isolation confirms that organisms were present and culturable under the experimental conditions; it does not prove that each isolate contributed equally to fermentation.

Some Bacillus strains produce poly-gamma-glutamic acid, a polymer of glutamate residues associated with the sticky threads found in natto-like fermented soybeans. A 2014 functional study measured poly-gamma-glutamic acid production in isolates from tungrymbai and bekang. The tungrymbai-derived Bacillus subtilis strain TS1:B25 produced 2.8 mg per millilitre under the reported laboratory conditions. This supports a plausible mechanism for stickiness, but that laboratory value is not a measured concentration in every traditional batch.

Other microorganisms may occur alongside Bacillus. A 2009 study recovered Enterococcus faecium, yeasts and other organisms from production materials or fermented samples. Its results showed that the raw beans, soaked beans, leaves and wooden equipment carried different microbial populations. These observations help explain why spontaneous fermentation can vary between producers. They do not establish that all recovered yeasts and bacteria are necessary, harmless or useful as starters.

Further biochemical effects include changes in available sugars, lipids, phytate and soybean isoflavones. Experimental investigations of isolates have demonstrated various enzyme activities and breakdown of some antinutritional compounds. Laboratory studies using added Lactobacillus cultures also reported changes in isoflavone aglycones and antioxidant assays. These results come from inoculated experiments; they do not carry over to naturally fermented tungrymbai or show a clinical health benefit.

How it is made and why each step matters

A documented Khasi preparation begins with cleaning and washing the soybeans, followed by soaking for approximately four to six hours. The softened seed coats may be removed by rubbing the beans between the palms. The beans are then cooked until they can be crushed easily. One published method describes cooking for one to two hours after soaking. Cooking softens the soybean structure and makes nutrients more accessible to microbial enzymes. Removing some seed coats also changes texture and the exposed surface of the beans.

After cooking, excess water is removed and the beans are cooled. They are placed inside a bamboo basket lined with leaves, traditionally called lamet, and covered with a cloth or jute bag. The basket is kept near or above a fireplace. The 2009 ethnographic account describes fermentation at approximately 25–40°C for three to five days. A separate microbiological paper published the same year describes three to four days. These temperatures and durations are field observations, not independently validated safety limits.

The leaves provide a traditional wrapping surface and help contain the beans. Their microbial contribution is not merely theoretical: Sohliya and colleagues isolated microorganisms from the wrapping leaves and other equipment. However, finding Bacillus on leaves does not prove that a particular leaf species reliably inoculates every batch with the desired strains. Botanical identity, leaf condition and handling matter. The plants are named inconsistently, so no substitute leaves can be recommended.

A Meghalaya Biodiversity Board account documents a different method from Pasyih. It describes boiling 20 kg of soybeans for eight to ten hours, replenishing water as necessary, draining through muslin and transferring the beans to leaf-lined baskets. The baskets are covered with jute and kept above a fireplace for three days and three nights. The fermented beans are then pounded, with a little cold water added if the mass is too dry. The report gives a yield of approximately two kilograms of fermented product per kilogram of original beans, but supplies no controlled fermentation temperature.

A July 2026 NESFAS account from the same village reports an even longer boiling time of twelve to thirteen hours. It describes two undisturbed days of fermentation, gently turning the beans and leaving them for another night. The beans are then pounded and mixed with turmeric. The reason for the differences is not recorded.

After fermentation, the beans can be prepared as a cooked dish. Traditional accounts describe mashing them, adding water and boiling while stirring until the water evaporates. Fried onion, ginger, garlic, chilli, salt and ground black sesame may then be incorporated. The black sesame is a culinary ingredient added to the preparation, not an essential starter for the bacterial fermentation. This is traditional practice, not a method with all the controls a validated, safe home process needs.

Regional and historical variations

Tungrymbai is associated with both Khasi and Jaintia food traditions, but published accounts show variation in how the beans are cooked, wrapped, fermented and finished. The differences do not necessarily correspond to clearly defined regional categories. Practice differs between villages, and no survey covers every one. It is therefore more accurate to identify a particular method by its source and locality than to present one recipe as universal.

The Khasi preparation documented in 2009 involves soaking, removal of soybean skins, relatively short cooking and fermentation in a basket lined with lamet leaves. The basket is kept near a fireplace for several days. Published descriptions of this process commonly include temperatures around 25–40°C, but the exact conditions can vary with household heating and the quantity of beans. Once fermented, the beans may be mashed and cooked with seasonings.

Pasyih in West Jaintia Hills is an important documented production locality. The Meghalaya Biodiversity Board describes producers working with comparatively large soybean batches and using lengthy boiling, muslin filtering, leaf-lined baskets and fireplace fermentation. The report records sales of the fermented product and identifies winter as an important production season. It also reports that some local producers regard summer batches as more susceptible to spoilage. These are valuable observations about production practice, not validated shelf-life measurements.

The July 2026 NESFAS account distinguishes ordinary tungrymbai from a version cooked with black sesame. It reports that both are sold through regional markets, with the black-sesame preparation commanding a higher price. This demonstrates a distinction between the fermented base and a more extensively prepared food. Black sesame was not necessarily part of the primary fermentation, and the two commercial forms may be stored differently.

The terminology also varies. NESFAS records tungrymbai or tungtoh among Khasi and Jaintia communities and dodim as a Garo name for a related fermented soybean food. Such names are useful for geographical comparison, but should not automatically be treated as interchangeable names for an identical recipe. The organisms and methods of every locally named preparation require their own evidence.

Scientific experimentation has introduced further variations. Studies published in 2018, 2019 and 2024 examined preparations inoculated with selected Lactobacillus cultures, including L. fermentum, L. plantarum and other strains. These trials investigated acidity, storage, sensory qualities and biochemical properties. They are modern adaptations designed for research or standardisation. They do not replace the evidence that naturally fermented tungrymbai is predominantly Bacillus-led.

How it is eaten and used

Tungrymbai is eaten as a strongly flavoured accompaniment to rice and other foods in Meghalaya. The fermented beans can be mashed and cooked into a thick savoury preparation, with black sesame, ginger, garlic, chilli, onion and salt among the documented ingredients. Its taste is concentrated and pungent, and its texture differs from that of ordinary boiled soybeans. In many prepared dishes the beans are no longer served separately; they become part of a thick mixture of seasonings.

A traditional preparation described in a 2009 ethnographic study begins after fermentation. The beans are transferred from their leaf wrapping, mashed and placed in a cooking vessel with water. They are boiled while being stirred until the water evaporates. Fried onion, garlic, ginger, chilli, ground black sesame and salt are then added. The result is a thick curry or side dish served with steamed rice. No precise quantities are given.

Black sesame is particularly associated with the dish in Khasi and Jaintia cooking. Meghalaya Tourism describes fermented soybeans cooked with black sesame as an accompaniment to jadoh, a rice-and-meat preparation, and other foods. Ground sesame contributes a nutty taste and thickens the mixture, while the fermented soybeans provide savoury flavour. This is a culinary combination, not evidence that sesame participates in the original soybean fermentation.

Some preparations also contain pork or animal fat. A 2018 paper on traditional meat products in northeastern India describes tungrymbai as a pork dish made using fermented soybeans, sesame paste and chopped ginger. A government cultural description identifies a Jaintia preparation in which fermented soybeans are fried in lard with red ginger and black sesame. These references demonstrate that the finished dish can contain animal ingredients, although the fermented soybean base itself is made from beans.

Tungrymbai is also sold as a fermented preparation for later cooking. Research and contemporary reports describe producers selling it at local markets, sometimes wrapped in leaves. The Pasyih accounts distinguish the mashed fermented beans from a cooked, seasoned black-sesame version. The condition of the product at sale therefore matters when describing how it is eaten. Fresh fermented beans and a cooked dish do not necessarily require the same handling or storage.

Tungrymbai is the fermented soybean preparation; cooked black-sesame tungrymbai, pork preparations and other local variations are dishes made from it. This avoids treating garlic, sesame, pork or lard as indispensable ingredients of the microbial fermentation. It also preserves the distinction between the original fermentation, pounding and later cooking, which can substantially alter the microbial population.

Reading results, identifying problems and understanding safety

Traditional tungrymbai is described as soft, sticky, strongly aromatic and sometimes pungent. These changes are consistent with bacterial growth, protein breakdown and extracellular polymer formation. The finished beans may be mashed easily, while a cooked preparation may have a thick, cohesive texture. These are descriptions of recognised products, not validated acceptance criteria. An unfamiliar smell or characteristic stickiness cannot establish whether the desired microorganisms are present or whether harmful microorganisms are absent.

Spontaneous fermentation presents important microbiological uncertainties. The 2009 field study detected microorganisms on raw beans, soaked beans, leaf wrapping and wooden grinding equipment. It isolated Bacillus subtilis, Enterococcus faecium and several yeasts among the organisms associated with production. Different materials carried different microbial populations. These are potential sources of inoculation and contamination, and not every organism found is necessary or safe.

The 2015 investigation identified B. subtilis, B. licheniformis and B. pumilus among tungrymbai isolates. These species should not be combined into an improvised starter mixture. Food suitability is strain-dependent, and the ability to isolate a bacterium does not establish its safety. B. cereus and other potentially hazardous spore-forming bacteria are recognised concerns for cooked soybean fermentations generally. They can be difficult to distinguish from desirable Bacillus strains without appropriate microbiological methods, and certain toxins cannot reliably be removed by ordinary reheating.

The alkaline character of tungrymbai is another safety consideration. Protein breakdown can produce ammonia and contribute to an elevated pH. Unlike lactic-acid vegetable fermentation, this process does not rely on substantial acidification as its principal preservation mechanism. Reported fermentation temperatures and durations describe traditional practices; they are not validated microbiological critical limits. Moisture, microbial competition, sanitation and post-fermentation handling affect the result.

Storage evidence is limited and method-dependent. The Meghalaya Biodiversity Board records local expectations of shorter keeping times in summer and longer periods in winter. These are producer observations, not validated storage recommendations. A 2018 study examined pre-cooked and post-cooked preparations using added Lactobacillus cultures at different storage temperatures, but those results apply to the experimental formulations. They give no shelf life for naturally fermented market tungrymbai.

Cooking the fermented preparation changes its microbial community. A 2012 study found differences in lactic acid bacteria, yeasts and other organisms between fermented samples before and after cooking, although spore-forming bacteria were comparatively persistent. This means that a cooked dish should not automatically be described as containing the same live microorganisms as the original ferment. Conversely, cooking does not retrospectively validate the safety of an earlier uncontrolled fermentation.

Laboratory studies have reported antioxidant activity, enzyme inhibition, isoflavone changes and other potentially interesting properties. These include experiments using deliberately introduced Lactobacillus strains. Such measurements do not establish clinical benefits from eating ordinary tungrymbai. Soybeans remain an allergenic ingredient after fermentation. For a reproducible food-safe production method, the starter or spontaneous microbial process, hygiene, temperature, storage conditions and relevant microbiological hazards would need validation. Unidentified mould growth or suspected contaminated batches should not be eaten or salvaged.