History of Soy Sauce, Shoyu, and Tamari - Page 2

A Special Report on The History of Traditional Fermented Soyfoods

A Chapter from the Unpublished Manuscript, History of Soybeans and Soyfoods: 1100 B.C. to the 1980s

by William Shurtleff and Akiko Aoyagi

Copyright 2007 Soyinfo Center, Lafayette, California

Page 1 | Page 2 | Page 3 | Page 4 | Page 5 | Page 6 | Page7 | Page 8


Philippines. Since the Philippines was a U.S. possession?? from 1898 to 1946, much of the early research on soy sauce was done by Americans and most of it was published in English. The main center of soy sauce research to date has been the Bureau of Science (renamed the National Institute of Science and Technology, or NIST, during the 1960s) in Manila. Soy sauce has been studied more than any other Philippine soyfood. The earliest known reference to or report on Philippine soy sauce ( toyo ) was in 1912 by Gibbs (Associate professor of Chemistry at the University of the Philippines) and Agcaoili. Both worked with the Bureau of Science. They noted that toyo was made principally by Chinese using soybeans imported from China, and that it was "a Chinese sauce of the Worcestershire type." The basic soy sauce was made with only soybeans, salt and water. After being aged and sunned for 2-4 months, the sauce was boiled for 12 hours, one or more times, to produce various grades or strengths of sauce. The Macao Chinese sometimes added wheat flour to the soybean koji. If molasses or sugar were added later the product was called si yao , a Chinese term similar to shih-yu . The authors gave analyses for two types of toyo (regular and with added molasses) and published a photograph showing toyo fermenting in a courtyard in nearly 100 earthenware jars, each covered by a conical woven bamboo lid. In 1919 Groff (working under Swingle at the USDA) published a lengthy article in the Philippine Journal of Science describing soy sauce production in nearby Kwantung province in China.

In 1925 Leopoldo Salazar, in his Bachelor of Agriculture thesis at University of the Philippines, Los Banos, developed improved methods for making toyo (and tofu), allowing a good sauce to be made in 8 weeks. His work was published in The Philippine Agriculturalist in 1926. Orosa (1932) of the Bureau of Science reviewed the foreign literature on soy sauce and gave some Philippine soy sauce recipes. In 1934 Prof. Filipe Adriano and co-workers at the Bureau of Plant Industry in Manila published a detailed, 15-page report on toyo. They noted that toyo was not as widely used in the Philippines as in other Asian countries because of the local popularity of patis , a fish sauce. Yet imports of sauces (mostly soy sauce) to the Philippines grew from 1.1 million kg in 1928 to 2.0 million kg in 1932. They described the method for making toyo, gave the names and address of 14 Philippine toyo manufacturers (most were located in or around Manila) and analyzed their samples, noting that many had molasses added. The best grade toyo was generally bottled without prior heating. The Philippine Department of Agriculture's Handbook of Philippine Agriculture (1939) described the soy sauce koji as being made with 1-2 parts wheat flour for every 2 parts soybeans, indicating a trend toward Japanese-style soy sauce. Fermentation time was 4 months to several years, followed by pasteurization but no addition of a sweetener.

In 1940 Yenko and Baens (the latter from the Bureau of Science) published the first of many studies by Philippine researchers on the use of indigenous crops to replace soybeans and wheat (which were not widely grown in the Philippines) in the manufacture of toyo. Using the Japanese method, they found that rice was a good substitute for wheat flour or barley. From this time on the majority of Philippine soy sauce research was done by women.

After almost 20 years of inactivity, research on soy sauce began to increase in the early 1960s. Masilungan^?? et al. (1960) noted that toyo, introduced to the Philippines by the Chinese, had found ready acceptance; from 1950-1960, retail value had increased eight-fold to 2.7 million pesos. New Japanese-style products and their manufacture were described. Mendoza (1961) discussed Coco Sauce, in which copra meal (defatted coconut meat) was used in place of soybeans to make a good-tasting sauce. Cruz (1963) wrote a thesis at Araneta University analyzing the protein content in soy sauces made from local raw materials. Baens-Arcega (1966a) patented a process for accelerating soy sauce production and in 1970, compared fish sauce and soy sauce. There were also studies on making soy sauce from copra and soybeans (Baens-Arcega 1966b), coconut paring meal and soybeans (Soriano et. 1967), corn and sorghum in place of wheat (Lebonifarel-Dumde 1967), and mecan peas in place of soybeans (Crisostomo et al. 1974). Although good results were obtained in ease case, these processes have apparently not been commercialized.

The establishment of the ASEAN Sub-Committee on Protein in 1974 stimulated extensive research and publication on soy sauce, starting in the mid-1970s. Important papers were presented by Soriano (1975) and Milono (1975) (both at the Sub-Committee's third meeting in Manila), Abdon (1978), Soriano and Pardo (1978), Pardo et al. (1979), and Soriano et al. (1980). Using a 2:1 ratio by weight of soybeans to roasted wheat flour, Soriano and Pardo (1978) found that soy sauce tasted best after about 8 weeks.

Soy sauce production and exports grew steadily. Between 1969 and 1976 imports grew from 3,494 liters to 7,127 liters, but exports jumped from 3,000 liters to 70,337 liters; 80% of these exports went to the USA. Yet total exports were less than 1% of domestic production. Most producers made a fermented sauce containing primarily soybeans (Soriano 1978). Dietary surveys conducted by the Food and Nutrition Research Institute showed that per capita consumption of soy sauce during the 1970s was 1.0 gm per day, being highest in Mindanao (1.05 gm) and Luzon (0.997 gm). The population of 50 million had a total annual consumption of about 18,250 metric tons. Of the 23 soy sauce plants registered with the Philippine FDA, 18 are operating in Luzon (the main Philippine island; most of these are in the Metro Manila area), 4 in the Visayan Islands, and 1 in Davao, Mindanao. The majority of companies were still run by Chinese. In 1974 large Philippine manufacturers sold 13,786 metric tons of soy sauce. There were three grades or classes according to the number of the drawing, class of raw materials, method of manufacture, and time and manner of fermentation. Artificial color (caramel) and flavor (MSG) were added only to the second and third drawings or extracts. The residue from the first drawing was processed and sold as yellow salted soybeans.

Thailand. Little is known of the early history of soy sauce in Thailand, in part because the Thai script is illegible only to Thai speakers. Called see-ewe?? (from the Chinese shi-you), its production is thought to date back to ancient times, when it was introduced by the Chinese (Sundhagul et al. 1978). The earliest known reference to soy sauce dates from only 1974, when Areekul and co-workers reported that Thai soy sauce contained 0.14 micrograms of vitamin B-12 per 100 grams, compared with 1.91 micrograms for the more popular Thai fermented fish sauce, nam pla . From 1974 on, with the establishment of the ASEAN Sub-Committee on Protein, soy sauce research flourished. The main centers of research were the Applied Scientific Research Corporation of Thailand (ASRCT) in Bangkok, Kasetsart University (Biology Dept.), and Mahidol University (Dept. of Microbiology). In 1977 the Symposium on Indigenous Fermented Foods was held in Bangkok. Papers on Thai soy sauce were presented by Yeoh??

Sundhagul et al. (1978), in a study of the Thai soy sauce industry, reported that there were approximately 50 soy sauce plants in the country; 41 of these were located in and around the capital city of Bangkok, far from the center of soybean production in northern Thailand. The majority of plants were small, family-run concerns, making less than 100 kl a year. Total production of fermented soy sauce was estimated at 6,000 kl per year (about 7,200 metric tons, or 148 gm per capita), but it was not increasing. Four types of fermented soy sauce were identified; light-colored, salty and dark, dark by aging, and sweet. Typical fermentation time was 2 months in open air. HVP soy sauce, however, was growing in popularity. Sundhagul also developed a bibliography of soy sauce for the years 1973-77. Lotong (1978), in a study of soy sauce production in Thailand, surveyed 35 soy sauce plants and found that the largest used 12,600 kg of soybeans a month. Some varieties used rice flour as an ingredient, with or in place of wheat flour. Many soy sauce plants also make miso ( tao chieo ) and fermented tofu.

In 1980 Bhumiratana and co-workers described in detail the traditional Thai process for making soy sauce. From a soy sauce plant in Bangkok, they isolated a strain of koji mold (identified as Aspergillus flavus var. columnaris ) that was a superior protease producer and did not produce aflatoxin. They used it as a pure culture in a traditional plant to make soy sauce superior in quality to that usually produced. Flegel and Bhumiratana (1980) subsequently developed a method of whereby soy sauce producers could make high-quality pure-culture inoculum in plastic bags. These may be the first instances of producing Thai soy sauce using a pure culture starter.

Malaysia. Malaysian soy sauce is the Chinese type, and it was probably introduced in ancient times. There the two basic varieties: the thin variety called pak yau or san chau and a thick sweet variety called sai-yau , lak-yau , or hak yau ). Each comes in several grades. The thin sauce requires 1.5 to 6 months of fermentation compared with 6 to 24 months for the thick sauce. The thick sauce, sweetened with caramel, sells for slightly more than the thin sauce, and is used for frying rice, noodles, or meat. All these terms derive from Cantonese. The Malaysian soy sauce industry is operated mainly by Chinese or by Malaysians of Chinese descent, who inherited the traditional techniques from the forefathers from China. The few Malay-run plants produce mainly a blend of the thicker sauce, while the Chinese plants produce more of the thin sauce.

In 1975 research on soy sauce began at the Malaysian Agricultural Research and Development Institute (MARDI) in conjunction with the ASEAN Sub-Committee on Protein. Much research was conducted thereafter.

The 1971 Survey of Manufacturing Industries, Peninsular Malaysia showed that in 1970 there were 156 soy sauce plants operating in West (Peninsular) Malaysia. Soy sauce exports from 1972 to 1975 averaged 726 metric tons, while imports averaged 800 metric tons, increasing to 1,000 metric tons in 1978. A 1977 survey listed 153 plants. Most were located in the middle and northern part of peninsular Malaysia, in the states of Penang and Province Wellesley (44), Perak (35), and Selangor (28). About 95% of these were owned by Chinese. More than half of these plants were cottage industries. Only 16 employed more than 10 full-time employees, and only 2 had more than 50. In 1977 in Western Malaysia (population 7 million) soy sauce consumption is estimated at 5.7 liters (6.7 kg) per capita per year. Malays and Indians living on the East Coast of Peninsular Malaysia consumed primarily fish sauce, and little or no soy sauce. Thus annual consumption was about 39,900 kiloliters (46,900 metric tons). All plants but one made fermented soy sauce. The typical soybeans-to-wheat flour ratio was 1.44 to 1. The koji mold fermentation averaged 7 days and the brine fermentation 3 months. Most plants also made other soyfoods (Ong 1977; Merican 1977; Yeoh et al. 1978).

Ismail (1977) described a fish-soy sauce, but it is not clear whether or not this is a traditional product and what its indigenous name is.

Merican (1978) reported that Malaysian soy sauce plants used from 3 to 120 metric tons of dry soybeans a year. Chinese consumed the most soy sauce, roughly 18 liters per family per year, compared with 9 liters for Malay families on the West Coast of Peninsular Malaysia. A typical plant, making 10,000 liters of high grade thin soy sauce, would use 2,000 kg soybeans, 1,000 kg wheat flour, 3,600 kg salt, and 1,000 kg sugar. Noranizam (1979) studied Rhizopus molds in Malaysian soy sauce koji, and Ng (1979) described how Malaysian soy sauce ( tau eu ) was made.

Wood (1982) estimated that in 1980 140 soy sauce plants in Malaysia produced 21,000 tons a year. It is not clear why this figure, obtained from the proprietor of a leading brewery in Kuala Lumpur, is less than half the 1977 estimate given above.

Singapore. Although soy sauce must have been introduced to Singapore (the majority of whose population is Chinese) from China long ago, very little is known of its history or present status. The two major types are light and dark.

Burkill (1935) outlined the process for making soy sauce in Singapore and noted that the koji could be made year-round. Some recent research has been done by the Singapore Institute of Standards and Industrial Research (SISIR), supported by a grant from the ASEAN-Australian Economic Cooperation Programme. At an ASEAN workshop on soy sauce manufacturing techniques, held in Singapore in 1978 Tang discussed worldwide developments in soy sauce fermentation, while noting briefly that in 1976 in Singapore there were 16 establishments that produced a total of S$6.5 million of soy sauce; the volume of production was not given. That year exports were valued at S$1.7, slightly larger than imports worth $1.5. Tang also compiled a literature survey on soy sauce microbiology and technological developments.

By 1982 many of the soy sauces sold in supermarkets were the HVP-based nonfermented types, with high levels of caramel and/or sugar. Two large soy sauce manufacturers were Tai Hwa Food Industry and Yeo Hiap Seng.

Indonesia. In Indonesia the two basic types of soy sauce are the thick sweet type ( kecap manis ) and a thinner saltier type ( kecap asin ). When most Indonesians refer to kecap, they mean the thick sweet type. Kecap manis is unique in at least three respects: it contains large amounts of palm sugar to make it quite sweet; it is simmered for a long time (4-5 hours) which, combined with the sugar, makes it very thick; and it is seasoned with a variety of spices and herbs, and may even contain small amounts of fish and poultry. It is not known when or why the spices first started to be added.

The Indonesian word for soy sauce (formerly spelled ketjap , now spelled kecap , but pronounced KET-jap) and the American word "ketchup" (spiced tomato paste) are believed to share a common ancestry, having both derived from the Malay word kechap , which refers to a kind of fish sauce. The Malay word, in turn, is thought to have come from the Chinese (Amoy dialect) ketsiap , the name for a Chinese-invented fermented fish sauce made by pickling fish in brine. Some dictionaries derive the English and Indonesian words directly from the Chinese with no Malay intermediary (T. de L. 1889-90). It is not clear how the Western tomato paste came to be identified with a fish sauce, except that both are seasonings with similar consistencies. The American spelling catsup was derived by folk etymology from ketchup. This soy sauce was probably brought to Indonesia by Chinese from south China, where a related sauce ( ketsiap , ket-jap ) is made.

The earliest known references to Indonesian soy sauce were by Dutch travelers and scientists. In 1787 Mr. Isaac Titsing, in an article on "Bereiding van de Soija," first described its preparation and noted that it was called ketjap in Batavia (Jakarta). In 1895 (in Dutch) and in 1896 (in German) the Dutch scientist H.C. Prinsen Geerligs described at length the preparation of soy sauce by Chinese in Java. He noted that black soybeans were boiled, drained, and sun-dried, covered with hibiscus leaves which he felt caused them to become inoculated with Aspergillus molds. The molded beans were dried, immersed in a cold solution of salt water, fermented in the sun for 8 days, then cooked. The soy sauce was poured off, the beans re-boiled several times with plain water, then the various drawings mixed together, filtered, and boiled. To this was then added palm sugar, star anise, and various "soy herbs" sold by Chinese herbal pharmacists, and the mixture was finally simmered for a long time. Various qualities were sold, Prinsen Geerligs praised the flavor and aroma of the best grade and gave the earliest known analysis of its composition. It contained 15% sugar and glucose, 17.1% sodium chloride, and 57.1% water.

Boorsma (1901) repeated briefly Prinsen Geerligs analysis of ketjap. According to the Indonesian publication Pemimpin Pengoesaha Tanah (15 Jan. 1915) ingredients used in ketjap besides soybean were ground fish ( ikan pikak ), djamoer koeoping ( Auricularia sp.), aren sugar, salam leaves ( Eugenia polyantha ), pandan wangi leaves ( Pandanus odorus ), laos ( Alpina galanga ), ginger, sereh wangi , brambang abang , and an infusion of chicken flesh. Each soy sauce maker was said to add special ingredients to give the product a better taste and aroma (Heyne 1913, 1927). Heyne (1927) summarized earlier work, including that by van den Burgh, reporting that various additional flavorings were used in "kechap" including the fungus Hirneola , citronella, and red onions. Ochse (1931) noted that ketjap was sometimes made with pangium fruits. Stahel (1946b, in Surinam) and Ko (1961) summarized earlier information.

In 1972 Slamet and Gandjar in the earliest known study of the kecap microbiology by Indonesians, investigated the influence of Rhizopus oryzae and Aspergillus oryzae on the quality of soy sauce. Winarno et al. (1976) made the earliest known study of Indonesia's soy sauce industry. He reported that Indonesia's 588 kecap manufacturers employed 2,901 workers and used about 501 metric tons of soybeans a year (0.4% of all soybeans used in Indonesia) to make 3,336 metric tons of soy sauce worth US$890,000. More than half of this soy sauce was made in West Java. The sweet soy sauce (which may contain up to 50% palm sugar) accounted for an estimated 90% of the nation's total production. Solomon (1976) gave 29 Indonesian recipes using soy sauce.

The ASEAN Protein Project stimulated the first extensive research on soy sauce by Indonesians. Major centers of research were the National Institute for Chemistry in Bandung and the Department of Microbiology at Gadjah Mada University in Yogyakarta. Saono et al. (1976, 1977) studied the microbiology and enzymes of kecap. In 1976 a 2-year survey began in Java to investigate techniques and problems related to soy sauce. Molds were collected and favorable strains selected. In 1977 at the Symposium on Indigenous Fermented Foods Saono and co-workers, and Poesponegoro and Tanuwidjaja gave good summaries of this "essential condiment in the Indonesian diet." They estimated 1973 production at 5,712 liters, listed the spices currently in use, and discussed biochemical changes and nutritional composition.

In 1978 Indonesia's first large, modern soy sauce plant was established in Jakarta, using defatted soybean flakes and cracked wheat. Most traditional companies used only whole soybeans, without wheat. Researchers developed soy sauce inoculum. Spices commonly used in Indonesian soy sauce (with average concentrations in mg per 100 ml of sauce) were reported to be star anise (powder, 250), coriander (powder, 50), sereh ( Andropogon sativum , extract, 3), laos root ( Alpina galanga , extract, 2), and salam leaves ( Eugenia polyantha extract, 1) (ASEAN 1978; ASEAN 1980).

Shurtleff and Aoyagi (1979) in The Book of Tempeh (professional hardcover edition) gave a detailed review of soy sauce in Indonesia, including industry statistics and a bibliography of Indonesian publications.

Vietnam . Two types of seasoning sauces have long been used in Vietnam: fish sauce ( nuoc-mam , pronounced nuk-mam, literally "water of salted fish") and soy sauce ( tuong , pronounced tung, a term derived from the Chinese jiang ). Nuoc-mam is most widely produced along the coasts and in south Vietnam where fish are abundant, the climate is too rainy for sun-drying fish, fuel is too scarce for fire-drying fish, and soybeans are not widely grown. This famous salty fish sauce, produced largely by enzymes in the tissues of small sea fish (autolysis), has been widely investigated since the early 1900s (Westenberg 1941; van Veen 1965).

The earliest known description of tuong in a European language was by Bui in 1905. A political writer, he described in detail processes for making two types of tuong (from soybeans and either glutinous rice or roasted corn) in Tonkin (north Vietnam). In 1932 the French pharmacist colonel J.C. Peirier and the Indochinese pharmacist Nguyen Kim Kinh wrote a detailed 8-page article on the "Chemical Analysis of Tuong-dau." They described how the product was made and noted that, as used by the Annamites (in today's central Vietnam), it had the advantage over fish sauce ( nuoc-mam ) of being less expensive and of being able to be made in any homes with a minimum of materials. For this reason it was called "the nuoc mam of the poor." Crevost (1933) described a village named Cu-da in the province of Ha-Dong which specialized in production of tuong-dau, making it only from April to July each year. Giraud (1942), in his book on soybeans, summarized earlier writings. Because we consider tuong to be more closely related to jiang and miso than to soy sauce, we have given details of these descriptions in Chapter 35.

The only other known accounts of tuong are very recent; Nathan (1979), Hoang (1981), and Leviton (1982c). Of these, Hoang is the most detailed. Nuoc mam's counterpart in the colder north has been various soy sauces, of which there are four types. All have been made since ancient times but there are no filtered soy sauces resembling those in other parts of Asia: (1) smooth tuong ( tuong cu da ) has traditionally been made in the village of Cu Da in North Vietnam, and has a brick-red color and smooth, thick consistency. It accounts for about 60% of all Vietnamese soy sauce; (2) chunky tuong ( tuong ban ), which comes from the village of Ban Yen Nhan, is a typical relative of Chinese jiang, resembling a soft miso more than a soy sauce. It accounts for about 35% of Vietnamese soy sauce; (3) soy nugget sauce ( xi dau , pronounced si-zao, the Vietnamese pronunciation of si-yu ) is a dark soy sauce, darkened by many months of exposure to air. It was traditionally made by Chinese rather than Vietnamese, and served in restaurants (especially Chinese ones) in both North and South Vietnam; and (4) nuoc tuong (pronounced nuc tung) is the clear yellowish liquid filtered off after cooking soybeans then allowing them to stand (while some autolysis and fermentation takes place) for 9 to 30 days. It is salted then used at home for pickling eggplants, but not widely. Tuong is typically made in either cylindrical or bulbous earthenware jars, capped with a conical woven bamboo lid to keep out rain but to allow aeration (Hoang 1982, personal communication).

Smooth tuong differs from typical soy sauces in three basic ways: (1) the koji (called moc ) is made from glutinous rice or corn; (2) the soybeans, not the grain, are roasted, then they are ground and fermented alone in water before the koji is added; and (3) there is no final separation of the liquid sauce from the solids residue; rather, the fermented mixture is ground until it has a smooth, somewhat thick consistency. Thus smooth tuong is the world's only suspended or "whole bean" soy sauce. Vietnamese soy sauce (smooth tuong) was first made in the West by Mr. Hoang of the Vietnam Food and Drink Co. in Maryland, starting in June 1979.

Tuong has always been made in only a few villages in North Vietnam, and by only a few families in those villages. Tuong has largely disappeared from North Vietnam since 1955 and from South Vietnam after 1975, since the communist regime took control (Hoang 1981). Because Vietnam is not part of ASEAN and because of the Vietnamese war that lasted until 1975 and the demands of reconstruction that followed, Vietnam was not as active in soy sauce research during the 1970s and 1980s as many of its Southeast Asian neighbors.

Korea . The earliest known reference to soy sauce and soybean paste ( kanjang and jang ) dates from AD 683 (Wang and Lee 1978). In 1949 A.K. Smith of the USDA visited Korea and described the process for making soy sauce in a Korean home, where most of it was and is still made, starting in January or February. He described the preparation of soybean koji (called meju in Korea and miso-dama in Japan) in the form of naturally-inoculated balls, which serve as the basis for both Korean soy sauce and miso. After a mash fermentation of 2-3 months in an earthen crock, the liquid was poured off, boiled for about 2 hours, and filtered to become the sauce. The sediment remaining in the crock was jang (soybean paste).

Starting in the 1950s, after Korea gained independence from Japanese rule, a number of indigenous studies on soy sauce began, most of them published in Korean, but generally with English summaries or summaries in Chemical Abstracts . Studies on the biochemistry of Korean soy sauce and changes during fermentation were published by S.H. Choi (1957, 1958), Park (1959, 1969), C.H. Chang (1965, 1968), and T.S. Lee (1970). Microbiological research was done by C.H. Cho (1970, 1971), W.J. Lee (1971), and Z.U. Kim (1975). T.W. Kwon (1972) gave an overview of soy sauce as one of Korea's fermented foods. TIME magazine (1969) reported research by Dr. David Seel in Korea showing that aflatoxins had been found in homemade Korean meju and soy sauce, and that this might be a cause of the high incidence of stomach cancer among Koreans. Several reports from 1977 showed aflatoxins in both homemade meju and soy sauce (Marjorie Baldwin 1983; get full citations). J.M. Lee et al. substituted corn and barley (1972) and potatoes and sweet potatoes (1972b) for imported wheat, with good results. Solomon (1976) gave 18 Korean recipes using soy sauce. J.R. Wang et al. (1978), in one of the best articles on Korean soy sauce to date, reported that it was the fourth largest food use of soybeans in Korea (after oil, tofu, and soybean paste), using 47,000 metric tons of soybeans a year or 10.6% of the total. Kanjang contained 31.1% solids, including 23.5% salt (both much higher than Japanese soy sauce). Per capita consumption was about 20 ml a day or 7.3 liters a year; with a population of 38.9 million in South Korea, this translates into 284 million liters or 284,000 kiloliters a year domestic consumption in 1978. Compare with Japan?? Taiwan?? Most of this soy sauce ( kanjang ) was still made in private homes for home consumption.

Wood (1982) gave details of Korean soy sauce exports, which nearly quadrupled (p.78) ... where is document?? Total exports??

Given the major importance of soy sauce in Korea, surprisingly little has been published on any aspect of the subject.


Page 1 | Page 2 | Page 3 | Page 4 | Page 5 | Page 6 | Page7 | Page 8