The Idli being a steam cooked dish made of ground and fermented paste of rice and black gram can be considered as one of the healthiest foods discovered and evolved in southern India. A number of steamed cousin dishes of idli are known in coastal India (especially Tulunadu) since times immemorial. The idli has been mentioned in some of the ancient Kannada texts.
Incidentally, now you can find numerous notes in the internet fueling controversies and disputes that the technology of steaming as well as fermenting were borrowed from abroad. The common myths floating in the internet are:
Myth 1: The process of fermentation was not known in India.
Myth 2: There were no vessels for making steam baking in India till 7-8 C CE.
Myth 3: Idli technology was derived after an Indonesian dish called kedli.
Myth 4: Idli was made from the pulse Urudu (black gram) only in 10th Century.
We need to accept that the exchange of ideas and technologies is a common phenomenon in the global village since antiquity. However, let us review the available data for obtaining a better perspective understanding on the antiquity of idli.
Antiquity of agriculture of Urudu and Rice
It is interesting to note that the pulse urudu (uddu, urad, or ‘black gram’) is an indigenous crop in origin. The archeo-botanical studies on ancient ash mounds of suggest that urudu (Botanical name: Vigna mungo Linn) was grown in parts of northern Karnataka as early as 2200 BC. It has been considered to be native crop of South India evolved from the wild varieties like Phaseolus sublobatus, a wild plant prevalent in the Western Ghats (Fuller, ).
It has been considered that the rice (Oriza sativa) was said to have been introduced in India during Harappa civilization ca 2500 BC. Archeo-botanical studies in semi-arid zone of Karnataka suggest that rice was introduced in south India ca 1800 during iron ages. The archeo-botanical data might not be representative, since the rice is a preferred crop in regions of intensive rainfall, where the ancient seeds either may not have been adequately preserved or available for studies.
The European words for rice (Old French ris; Italian riso; German/Welsh: reis etc) suggest derivation from Dravidian words like ‘arici’ (Tamil) and ‘ari’ (Tulu). The Arab and European countries were engaged in trade with south Indian ports during the early centuries and the along with the exports some of the Indian words also were absorbed in the Europe.
Antiquity of the process of Fermentation
The Myth No 1 is not true.
The knowledge on the process of fermentation is not new to Indians since remote antiquity. The intoxicating drinks were well known in India since Vedic times at least as early as ca.2000BC. Similarly brewing toddy from palms was a profession and tribes specializing in the art were known as idiga or devar since antiquity in India. The ancient art of making of curd (yoghurt) from milk also involved the understanding the intricacies of fermentation.
Similarly the art of marinating fruits and vegetables (pickling) in saline water was also akin to the process of fermentation. Coastal Indians were adept in marinating vegetables and fruits in saline water and preserving for lean seasons, since saline water was available in the proximity. Traditional pickles of jack fruit or mango in saline water, without chillies, were once popular in rural Tulunadu and are known as “neer uppad “(=watery pickle) in Tulu.
The grains of urudu and rice were both common home grown agricultural products in southern India since antiquity. It is possible that ancestors naturally learnt the art of fermenting the mixtures of ground urudu and rice pastes as a consequence of keeping the mixture overnight.
Urudu and urugu
Note that the Tulu word for fermentation (“urugu”) is related to the word for the black gram (“urudu”). The modern Kannada word for fermentation “hudugu” has subsequently been modified from the original Old Kannada/Tulu equivalent of “urugu”. It needs to be pointed out here that Tulu and Old Kannada cultures shared many common words till about 10th Century CE. With evolution of language, the modern Kannada (Hosa-Kannada) has modified the original old Kannada words, whereas the Tulu has preserved its antique and basic oral characters, probably because of lack of pervasive and well- spread script for writing the language. In any case, the word clearly suggests that ancient people of Tulu/Kannada regions were versed with the art of fermenting the ground urudu +rice pastes.
Cousins of idli: mooDe, gunDa, kottige and tondru
Myth No 2 that there were no steaming vessels in ancient India, attributed to Chinese Buddhist traveler monk Xuanzang (or Hsuan Tsung , ca 602-662 CE) may not be correct, in the sense that the art of steam cooking was known in India, even though it is difficult to date the chronological event in the absence of precisely dated texts.
Ancient Greek traveler reports dating back to the beginning of the Common Era (ca.100 CE) mention that Indians were using boiled grains (Post 11; 31 March 2007. It reminds me of the boiled rice was being used as a nourishing food in Tulunadu since ages.
Steam cooking does not any need elaborate designer vessels. In coastal Tulunadu since ages steam cooking is done using simple earthen pots.
There is a number of traditional steam boiled dishes made out of fermented ground mixtures of rice + urudu. These ancient cousins of modern idli are known by Tulu names such as mooDe, gunDa, kottige and tondru adde.
Traditional Steam cooking
The fermented batter is poured into containers fashioned nout of specific leaves and kept systematically in upright vertical position in a abutting series inside a large earthen pot. The pot is filled with water to the half full level. Then the pot is closed with a lid and kept over the fire. The water in the pot boils generating steam that boils the contents of the containers kept inside the pot. The traditional rural cooks can identify the appropriate degree of baking based on the emanation of specific fragrance of the steamed dishes couched within the leafy containers.
mooDe: leafy tubular container
Of these mooDe is steam -baked in tubular structures fashioned out of ribbon like leaves of the shrub known as “munDevu “ (Indian Screw pine or Pandanus utilis) which grows in coastal areas and preferentially along the river banks. The leaves are shorn of their spines and warmed over fire to make them supple. Thus fire-stoked tender leaves were fashioned into tubes of desirable size, making use of vegetal pins derived from fronds of coconut palm. The tubes are closed at one end so as to contain the fermented mixture of rice+ urudu batter.
In olden days, in rural Tulunadu, mooDe tubes of about 10 to 14 inches long with 3 to 4 inch diameter were being knitted using the tendered munDev leaves, especially on the eve of specific festivals. (Nowadays, the size of the mooDe tubes available commercially in the market or in local hotels have shrunken and it is common to find shorter than six inch long mooDe tubes of less than two inch diameter).
Incidentally the word mudepini or mudeyuni in Tulu is usually applied to the art of knitting ribbon like leaves into containers or knitting palm frond into mats. Thus moode is a tubular leafy container fashioned out of fire-tendered ribbon like leaves of mundev plant.
The term munDevu (ಮುಂಡೇವು) is quite interesting. The common screw pine plant (Pandanus utilis) apparently was named after Munda tribes, who are Austro-Asiatic tribes considered to have been migrated to India from South East Asian countries early in the history.
The containers for gunDa or kottige are fashioned out of leaves of jack fruit tree. To make a single gunDa container four jack leaves are selected and placed like a star and using pins made from coconut fronds fashioned into small tumbler shaped containers. Rest of the steam baking procedure is common as described above.
Another related steamed rice+ urudu dish is known as “tondru adde” which we covered in Post : 227 ( dated 2 February 2010).
These are the ancient pattern of preparing steamed dishes from the fermented paste mixtures of rice and urudu (normally in the ration of 4:1).
Secrets of Fermentation
The batter (banda in Tulu) for preparing steamed dishes from the fermented paste mixtures of rice and urudu were traditionally ground separately using manual grinders and mixed subsequently. Our grandmothers preferred the rice and urudu combination in a ratio of 4:1. In fact the process of uruguni (=fermenting, Tulu) of the urudu paste overnight holds the key for production of tasty moodes, gundas or idlis using steam cooking techniques. The fermentation is a secret natural process abetted by microbes. Intelligent observers (specially the biologists with a flair for cooking) have noted that fermentation in some houses (where specific set of fermenting microbes survive in abundance) result in tasteful moodes / idlis, whereas in utterly ‘clean’ (microbe free) the optimum degree of taste and savor can be deficient! Some cooks prefer the ratio of 1:2 for the urudu: rice proportions to enhance the strength of fermentation. And besides, the rice used is, preferentially composed of equal proportions of boiled and raw rice.
In rural Tulunadu, toddy was added to the batter to increase the degree of fermentation. Increased level of fermentation leads to pronounced taste of the steam cooked dishes. Some persons add (chilled) coconut water, and /or green chillies, ginger and/or other spices etc to soften the batter.
The exact version of the coastal dish moode might not have been in usage in Karnataka mainland possibly as the shrub Mundev (Screw pine) is not found growing in interior regions. However, it seems in northern Karnataka some other leaves were used to make containers for iddali-ge.
Tulu - Old Kannada: lingual affinity
The Tulu language though geographically restricted to coastal Karnataka shared equivalent words with Old Kannada obviously as a consequence of consanguineous evolution and development. Thus we find many ancient words common to Tulu and Kannada - as we can verify from the text of inscription at Halmidi or works of poet Pampa or even texts like Vaddāradhāne. Consequently, some of the old Kannada words now non-existent in modern Kannada can be traced in Tulu language with regard to clarification of their original meanings.
Tigale: One interesting example that immediately comes to my mind is a term “tigale” ( ತಿಗಲೆ) which can be found in the text of Vaddāradhāne. Prof. D.L. Narasimhachar, in the notes of Vaddāradhāne has given (probably guessed) the meaning of the word “tigale” as neck. However, it may be noted that the “tigale “ is a very commonplace word in Tulu and it means the chest.
The Kannada/Tulu term “iDe” means small leaf cup with space inside to hold the batter (There is also another word for leafy cups known as “donne” in Karnataka).
It can be noted that there are many words “–De” that signify space in Tulu also such as iDe (=gap; here); aDe (= thereto); ode (=whereto); hooDe (=sandy place) etc. There are rice dishes having suffixes of –DDe. : aDDe =rice dish (usually steam baked); oDDe (or oDe)= fried donut of urudu (black gram) paste. Therefore, iDDe+lige is the word applied to a related steamed dish of rice + urudu mixture. The suffix –lige is also used as in the case of polige or holige.
It is clear that the ancient term “IDDalige” has been abbreviated to “iDli” with passage of time. However in the instance of sweet dish ‘polige’ (which has become holige in modern Kannada) an abbreviated term ‘poli’ exists in Maharashtra.
Idada: It is interesting to note that “Idada” is an idli equivalent made of rice and urudu and is popular in Madhya Pradesh.
Iddalige in Vaddārādhane
In one of the earliest known (Old) Kannada texts discovered, in a descriptive list of dishes in a story, the name of iddalige can also found. The ancient Kannada text is a compilation of Jain religious stories, apparently translated from its Prakrit original is popularly known as Vaddaradhane, even though there are disputes that it is not its original name.
It is necessary to emphasize that only the name iddalige is mentioned within a list of dishes prepared in a feast in a story and that grain composition or method of making has not been explained in the text as assumed by some of the writers in the net.
Age of Vaddārādhane
The earliest known text in early Old-Kannada ‘Vaddārādhane’ attributed to author Shivakotacharya was considered to have been composed around Kogilu area in western part of present Bellary district, Karnataka. The precise date of composition of the text has been difficult to decide. Prof. D.L. Narasimhachar editor of the Vaddārādhane considered the work to have been compiled ca 920 CE. However, another veteran Manjeshwar Govinda Pai based on the style of language used in the text is said to have estimated the date of compilation at not later than ca. 600 CE. The Old Kannada version has been understood to be a translation of 19 stories of Jain tradition originally compiled by Shivakoti-Acharya in Prakrit language known as “Bhagavati Aradhana” of ca. 2nd Century CE. It is not clear whether the iddalige has also been mentioned in the original Prakrit work. Anyway considering various possibilities and uncertainties of dating the old texts, we may conclude the textual reference to the iddalige to be anytime between ca. 2nd and ca.10th Century CE.
The Prakrit was the administrative language in Karnataka in the early centuries of the Common Era (Shettar, 2007 ). Even in Coastal Karnataka the influence of Prakrit period is also well entrenched in Tulu language also as we find numerous Prakrit words in Tulu. Subsequently probably 4th Century CE onward Kannada came into prominence in Karnataka. With accent on Kannada, it appears the rulers of the land encouraged translating and rewriting older Prakrit texts in to Kannada.
Lokopakara & Manasollasa
Iddalige and Iddarika are further also found in Kannada and Sanskrit works of ca 11 Century known as "Lokopakara" and "Manasollasa" respectively.
The Lokopakara (Kannada) was complied by Chavundaraya II (ca.1025 CE) . Chavundaraya II was Sarvadhikari under Kalyana Chalukya king Jayasimha II. He belonged to Mudgal ( now in Raichur district, Karnataka.) Earlier he was a minister at Banavasi. The 7th Chapter of Lokopakara deals with cooking. Narasimha Murthy ( 2008 ) mentions only one line about the idli in Lokopakara: “While referring to the popular Indian dish iddali, he mentions mung dal, cumin seeds, coriander and pepper”
Similarly, in “Manasollasa” (ca. 1130th Century CE) a Sanskrit work compiled by Someshwara, the ruler of Kalyana (Bidar district, Karnataka,) the dish has been referred to as “iddarika”.
Dhokla (Dhokra) popular in Gujarat is a related dish. It was known as Dukkia earlier as mentioned in jain text, 1066 CE. Flours of rice and chickpea (bengal gram Cicer arietinum or basan) are mixed and fermented and steam boiled to make the dish. A similar steamed dish made exclusively of Bengal gram is known as Khaman. There is another Gujarati dish known as Khandvi made of gram flour and yogurt. The latter is known as “Suralichi Vada” or “Patuli” in Maharashtra.
Achaya and Kedli ?
Achaya (2000, 2003) wrote that there is a suggestion that idli originated from an Indonesian dish called Kedli and brought to India by Hindu kings of Indonesia. Note that Achaya mentioned that it is a suggestion only.
However, information on an Indonesian steamed dish called kedli could not be traced. Where went the mysterious kedli ? A little wandering in the jungle of internet revealed an interesting twist to the story of the elusive kedli !
Kadele to Kadalai
Instead we could find an Indonesian dish popularly known as Tempeh which was mysteriously based on “kedalai “or “kadele.”
There were profuse cultural exchanges between South India and South East Asian countries during the first millennia of the Common Era. During the period tenets of Buddhism as well as Hinduism formed lasting impressions in the cultural settings of the South East Asian countries like Java, Indonesia, Cambodia and Thailand.
Under these events we find evidences for the export of south Indian words to those cultures. The “kaDale” (Botanical name: Cicer arietinum) is a common (Tamil- Kannada - Tulu) Dravidian word for the pulse alternately known as Chick-pea or Garbanzo bean or the Bengal gram or Chana.
The term kaDale travelled to Java. However, it was applied to Soy bean there. The word kadele means Soybean in Javanese.
In Indonesian it was modified to kedalai. The kedalai is Soybean in Indonesia. Whole grains of soy bean are fermented and used in the making of the dish called tempeh.
It is reported that the term “kadele” has been mentioned in the Javanese manuscript called “Serat Sri Tanjung “ considered to have been composed in 12th or 13th Century CE. The manuscript also contains a reference to a species of sugarcane known by the Indian name of “chintamani “ variety of sugarcane. Thus, the case of Indian influence in adoption of the word “kadele” is supported by another Indian word Chintamani.
The main object of this review is to discuss the available data and to refute some of the confusing myths floating in the internet. The process of fermentation was known in India, as suggested by the existence of fermented ancient products like somras, toddy and watery pickle (neeruppad). Ancient Indians probably made use of simple earthen pots for steam cooking, samples of which were existed in rural kitchens of Tulunadu till modern times, till some five to six decades ago. There is an overwhelming evidence to suggest that idli was invented in parts of ancient Karnataka, probably as early as ca.2nd Century CE, as referred to in ancient Kannada/Prakrit texts and further complemented by data on steam cooked rice-urudu dishes akin to idli such as moode and gunda from the rural coastal Karnataka.
The term idli or iddalige is essentially an Early Kannada word and as such there is no proof to suggest that the technique of idli making was copied from an Indonesian dish supposed to have been called kedli. The origin of the mysterious word kedli happens to be the Dravidian (Kannada/Tamil/Tulu) term kaDale. The Javanese term kadele or the Indonesian term kedalai though derived from the Dravidian kadale refer to Soybean and not urudu.
Similarly, there is no staunch evidence to conclude that the ancient idli (Iddalige) was made exclusively of urudu and without rice.
Similarly, there is no staunch evidence to conclude that the ancient idli (Iddalige) was made exclusively of urudu and without rice.
(Note: Copies of these two works -Lokopakara and Manasollasa - were not available to me so far for perusal hence would not be able to verify the details discussed by other fellow contributors in the net)
Achaya, K.T . (2000) The story of our food. Universities Press. p.140.
Achaya, K.T. (1993) Indian Food: A historical companion.
Achaya, K.T. (2003) Indian food;
Narasimha Murthy, A.V. (2008) Some aspects of agriculture as described in Lokopakara, Chapter 14. in Lalanji Gopal & Vinod Chandra Srivastava (ed): History of agriculture in India, upto c.1200 AD.
William Shurtleff, H.T. Huang, Akiko Aoyagi. (2014).History of Soybeans and Soy foods in China and Taiwan, and in Chinese cookbooks, restaurants, and Chinese work with soy foods outside China (1024 BCE to 2014):.
ನರಸಿಂಹಾಚಾರ್, ಡಿ. ಎಲ್. ,ಡಾ: (೧೯೪೯; ೨೦೦೩) (ಸಂ) ಶಿವಕೋಟ್ಯಾಚಾರ್ಯ ವಿರಚಿತ ವಡ್ಡಾರಾಧನೆ. ಡಿವಿಕೆ ಮೂರ್ತಿ,ಮೈಸೂರು-೪, ೩೪೪ ಪುಟಗಳು.
ಷೆಟ್ಟರ್,.ಷ.(೨೦೦೭; ೨೦೧೦) ಶಂಗಂ ತಮಿಳಗಂ ಮತ್ತು ಕನ್ನಡ ನಾಡು –ನುಡಿ: ಆರಂಭ ಕಾಲದ ದ್ರಾವಿಡ ಸಂಬಂಧದ ಚಿಂತನೆ. ಅಭಿನವ,ಬೆಂಗಳೂರು.೩೨೦,ಪುಟಗಳು.