Klaus Karttunen

In this article my intention is to discuss in a general way the somewhat obscure period of Indian literary and religious history falling between the last phase of the creation of commonly recognised Vedic literature (the Smrti part of it) and the beginnings of the modern scholarship dealing with it. In short - the mediaeval period. However, there is an inherent danger in using schematic terms of periodisation, especially when originally brought from a different geographic context. Therefore I must first try to define somewhat more exactly what I here understand under the notion of mediaevalism.

A concept like the Middle Ages is not very precise, even in European circumstances, and when applied to India it has a different meaning. In history, the concept of Indian Middle Ages is mostly used as roughly corresponding the early (pre-Mughal) Islamic period, sometimes also including the period immediately preceding the Islamic conquest.

In literature and religion, such a distinction as mediaeval, if used at all, must be stated differently. A text written in Sanskrit is often styled as ancient or classical, even when it is probably composed well into the second millennium A.D. In addition, there are fairly modern Sanskrit texts, and it is difficult to say where the borderline should be drawn.1 At the same time, a book in some NIA language is clearly mediaeval, even if it actually precedes many well known Sanskrit texts.2

We may further note that a tradition, like the Veda, does not necessarily follow the changes of political history. The orthodox Hindu society still went on in its usual way, at least in South India, although the rulers were Muslims.3 In this respect, the change was not so significant.

It seems to me that the best way to periodise the long history of the Vedas lies in the way of transmission. At present, we are not so much concerned with the creative period with its well known subdivisions (in Samhita, Brahmana, Upanisad and Sutra), although this, too, continued much longer than is often thought. While the canons of Samhitas and Brahmanas were closed at an early date, there are late Sutras (like Agnivexa and Vaikhanasa) dating into the early centuries A.D. Although to us the difference between the classical Upanisads and later, often sectarian works going under the same name, is significant and useful indeed, to the Hindus this was not always so clear at all. But what is important now is what followed after the creative period, when the actual canon was considered as a complete and closed whole. Here we have to look at the method of transmission. The ancient period is characterised by pure oral transmission, after it followed the period - mediaeval, if you like - using, in addition to oral transmission, written characters and manuscripts. The main texts, however, were still primarily transmitted orally. Thus the modern period is that of printed publications and rapidly declining oral tr which is just a recent development. We must also keep in mind that all these periods are overlapping.

We do not know for certain when exactly the most holy books of India were first written down. The major and the only really authoritative way of transmission was oral,4 and written text was important only in exegesis and ritual science, not in religion itself. In other genres of literature, for instance in the Dharmaxastra and Ayurveda, it is rather well established that the extant text tradition was only fixed in the earliest (written) commentaries. Before that, in oral transmission, the texts were open to additions and modifications. In the Veda the texts were fixed much earlier, and the transmission in the first place remained oral. But here, too, the texts were probably first written down in connection with commentaries.

For the textual integrity of the Veda this seems to have been rather fortunate, as the purely literary tradition is somewhat open to corruption. Due to the semi-moist climate of India, manuscripts are rather short-lived: in poor conditions a palm-leaf as well as a paper manuscript is liable to rapid decay, and even in best conditions - kept carefully and tightly bound in a dry place and not read too often - three or four centuries seem to be the average maximum of age. There are some older manuscripts extant, true, but generally they are very brittle and fragile, worm-eaten, and often darkened more or less into illegibility. I have had an opportunity to examine such palmleaf manuscripts myself. Birch-bark lasts somewhat longer, but it was available only in fartest northwest regions.5

Of course the extant manuscripts were often copied from older ones, but frequent copying seems to have introduced textual corruption more easily than the traditional method of oral transmission founded on careful training applying special mnemotechniques. In any case it seems that the written tradition of the Samhitas and Brahmanas hardly extends beyond the mediaeval period of Indian history. Manuscript books, their keeping, preservation and copying have never had such an important role in traditional Hindu religion as in Buddhism and especially in Jainism. For the Veda, oral transmission has always been the main way of transmission while in traditional schools xastric works of grammar and aesthetics and even some Kavyas were learnt by heart, and even physicians had to know their fundamental texts by heart.6

Thus, in the Veda, manuscripts are only a kind of subsidiary aid. It is an amazing tour-de-force of oral transmission that there still are Brahmans, who have learned their Veda in the traditional way, living gurukule at the teacher's house, practising the ancient method of adhyayana, and this way learning long texts by heart without any supporting written material. In the case of the most ancient text the line of such oral transmission goes back more than 3000 years, and the text has still remained virtually unchanged. There are even texts for which there are no manuscripts at all, with possible exception of modern notes written from an oral source. This is sometimes the case in rare Sakhas, e.g. with the more elaborate Ganas or song-books of the Jaiminiya Samaveda.

For the Sruti, therefore, the manuscripts have always been only an additional aid beside the mainstream of oral transmission. And even the Smrti part is traditionally learnt by heart. A Brahman performing, say, a samskara, may still be muttering not only the mantras but the entire passage of his Grhyasutra during the ceremony. However, the contents of his muttering may also consist of a later text of the same ritual (in any case containing the necessary mantras. The ritual manual does not have the same sacrosanct quality as the Sruti, and this has made room for written transmission and even modification.

Learning an orally transmitted text begins at tender years, which simplifies the learning of the great number of texts by heart. This is possible as the actual content of the Sruti is not considered so important. After the creative period, the archaic Vedic language of the Samhitas and the very content of the texts had soon become somewhat obsolete. Traditional sciences - the Vedges - like phonetics, grammar, metrics, and etymology tried to make amends for it, but the result was not always too fortunate. On the other hand, it is in the very nature of the Sruti that its content is quite unimportant. When certain verses are recited in the original language with correct pronunciation and accents (and, in case of the Samaveda, with the right musical tune) and on the appropriate occasion accompanying the prescribed rites, everything is in order. Also, in this respect it is entirely irrelevant whether the text is understood or not.

Fortunately for us, however, the Vedic Sakhas did not lack intellectual curiosity. Much was forgotten, but at least some tradition of interpretation was kept alive. In the beginning it consisted of oral explanations given by the teacher, later they were also written down. This resulted in written commentaries, which are often presented under the name of the famous Sayana (lived in the 14th century), and many others. These commentaries give word-for-word explanations, partly fantastic, partly justified, of the given text.7

It is important to note that these commentaries are often older or at least of the same age than the extant manuscripts of the text itself. Moreover, in an elaborate word-for-word commentary the text itself cannot suffer such corruption as is possible in the simple manuscript tradition, a fact well known to classical Indian scholars. It is often the case that in these commentaries a relatively old written version of the text has been preserved. To an extent, such versions confirm the testimony of linguistics, penetrating into a much more hoary antiquity, demonstrating that the long oral transmission has been more or less flawless.

Now it is time to consider more closely the Vedic ritual and ritual literature. It is a well known fact that the ritual rules have been laid down in Sutras, the Grhyasutras for the domestic and the Srautasutras for the major so-to-say official ritual, each exegetic school or Sakha having its own Sutra. These works are written in the so-called sutra style, in terse and condensed rules and statements, which were meant to be learnt by heart. Many details are excluded from the text and left to the training given by the teacher. With these details it is important to note that while the main lines of the ritual have been more or less static, the society has enormously changed during the many centuries which have elapsed since the codification of the ritual Sutras, and this is reflected as modifications and additions in ritual. In fact, this development had begun even earlier. There is already a discrepancy between the Sutras and the ritual accounts found in the Brahmanas. We may also note that the contextual relevance of the Samhita texts used in various rituals is often only apparent, more or less arbitrary, and with no real relation to the ritual context. Thus, it is hard to imagine that these texts would originally have been composed for their present use.

In the beginning it sufficed that the changes that had taken place in the ritual informally be taught orally by the guru to his pupils, without laying them down in a formal (even if oral) text. But as time went on the number of changes increased. Of course, in an extremely conservative tradition, major or really significant changes were rare or even impossible, the rules having been laid down in the Sutras. The main tradition was also closed, it was no longer possible to revise the Sutra. But the rites tended to become ever more sophisticated. Irrelevant details could extend into small rites,8 while some other parts were suppressed or modified.9 Even new rituals - never mentioned in the Sutras - were accepted.10 Theoretically, the difference between Vaidika and Laukika rites was (and still is) kept, but in practice both are in many ways intertwined in the actual ritual.

In a changed society the terse prescriptions, written centuries earlier, were not always very practical. Sometimes they were just too short, sometimes they seemed to miss the point, sometimes they were too difficult and no longer properly understood. There are two solutions to this problem: commentaries and new practical manuals. There are commentaries to nearly every Sutra, explaining their language and the ritual, given alternatives and discussing differing opinions. Especially in the Srauta tradition the ancient tendency to consider ritual as a kind of science developed now to its fullest extent. Generations of ritual scholars spent their lives studying and discussing the details of the major rituals and expounding their ideas in learned works.

Let us take an example of the Nambudiri Brahmans of Kerala. They were rich landowners, the estates going solely to the eldest sons. Other sons had not much to do and the family estate ensured their living. While some of them became active in literature or science, many devoted their lives to ritual scholarship, keeping the tradition alive until the present century. Though the last Agnicayana, the ritual of constructing the fire-altar in 1975 was supported and financed from outside, the ritual expertise still existed and other great rituals of the same kind performed earlier and without any outside participation or attendance were still well remembered by old people.11 The Tamil Jaiminiyas, too, have been very active, though their tradition is now nearly extinct. Among them even such an exceptional literary achievement could be found as a Mantravrtti, a special commentary of the mantras used in rituals.

It should perhaps be emphasised that for a long time the majority of Brahmans, and all other classes of society, were no longer really interested in the Vedas. To them, the new sectarian religions of Visnu and Siva were more important than old gods and the sacrificial mysticism. They preferred the temple puja to elaborate Vedic ritual and, instead of the Vedas, they read the holy books of the second generation such as the epics Dharmaxastra and Puranas. Even to them, Veda had always been a principle to be kept but only relatively few saw it as a still living tradition and practised it beyond a few simple mantras and the necessary samskaras.

We come now to practical manuals. They are often quoted under two names, Prayoga and Paddhati. The difference seems to lie mainly in the name, translated as 'practical manual' and 'guide-book', respectively. What has been said of one, is mostly applicable to the other, and there are cases where both names are used for the same work. Occasionally such terms as Vidhi and Prakarana have also been used and a metrical work is known as a Karika.

What I have to say about these ritual manuals applies as well to the solemn ritual as to the domestic. That I have taken the Grhya viewpoint is only due to the fact that my own studies in this field have always been concentrated on the domestic side. To be honest, I have no taste for the elitist intricacy of the Srauta tradition, while the connection of the Grhya with the real everyday life has always seemed fascinating to me.

The earliest (or at least the earliest extant) phase of this later Vedic ritual literature contains of the so-called Parixistas. More or less still retaining the old sutra style, they supplement the Sutras and are sometimes included in their manuscripts as kinds of appendices. Nevertheless, their late origin as compared to the corresponding Sutras can be easily noticed. The Baudhayana-Grhyaparixista or Xesasutra, for instance, contains such late elements as references to post-Vedic gods such as Siva, Durga, Skanda, and even Krsna, to temples with images of gods, a Grahaxanti with the Greek planetary sequence and so on. Thus it might well belong to the latter half of the first millennium A.D.12

We cannot say when the first Prayogas or Paddhatis were compiled. Probably they have not been preserved. The extant ones are often rather late, but - as it often happens in Indian literary history - it is extremely difficult to determind the dates certainly or even approximately. In Indian manuscript libraries and private collections there is a great number of these texts but very few have been edited, and these are often modern compilations published and sold only locally. Therefore they have been more or less ignored in Western Indology, although they can, despite the included late modifications, greatly help us to elucidate many complicated problems of the Vedic ritual.13 In addition, they can also be studied as such, as documents of the later history and development of Vedic religion. This could help us understand how a millennia old tradition could, at least to some extent, retain its position despite the many changes taking place around it.

The situation with the commentaries is somewhat better. They have been indispensable for understanding the often difficult text of the Sutra that quite many have been printed. Printed, yes, but not edited. Even when several manuscripts have been carefully collated for the Sutra text,14 the commentary is just printed, without much care. Very few scholars have ever paid the same attention to a commentary as to an independent text, or studied its relation not only to the explained text, but also to other works of the same school, its position in the ritual and textual tradition and its possible influence on them (see Gonda 1977a:659ff). And still the number of unpublished commentaries greatly exceeds those published.

As I mentioned above, there is a great number of Prayogas and Paddhatis (and also metrical Karikas) in the Vedic ritual and most of them are unpublished. Quite many of them follow closely the Sutra (Grhya or Srauta) of their school, discussing the same rituals in the same order, but adding new rites and even complete rituals, which are missing in the Sutra.15 Some are restricted to one particular ritual - e.g. for the domestic ritual there is a great number of Vivahapaddhatis, explaining the most complicated of all Grhya rituals, the wedding - while some Srautaprayogas tend to concentrate on one particular ritual or also on the offices and functions of one particular priest (or even the yajamana or sacrificer). Some concentrate on additional rites, some are hardly more than a mere paraphrasis of the Sutra. Sometimes the bulk of short Prayoga's text consists of the Mantras quoted from the Sutra and - for auxiliary rites - from other sources, and these Mantras are bound together by short ritual directions, which can be even more condensed than in the Sutra. In a recent work these directions are no longer in Sanskrit but in a modern language.16

A common factor among these works is that they are practical. They are intended to be manuals for the actual rituals, not scholarly works like some commentaries and various Smrti and Dharma studies on ritual. As such they were also used, even so much that in some schools or minor sections they have wholly superseded the Sutra itself.17 Frequently the Sutra is preserved in one or a few manuscripts only, as a kind of scholarly curiosity, while actual rituals are learned and performed according to the Prayoga.18

My own studies have been concentrated on the Jaiminiya tradition, but a comparison to other Sakhas suggests that it can well serve as an example. Samavedic Jaiminiyas form, at least currently, a minor Sakha with only a few subdivisions. There are some traces of Jaiminiyas formerly living in Karnataka and &ndhra Pradesh, but at present there are only three extant groups: the Tamil Ayyangars of Tamilnadu, the Tamil Ayyars of Kerala, and the Malayalam speaking Nambudiris of Kerala. Especially the Nambudiris and to some extent also the Ayyangars have formerly, as rich and tradition-bound groups, taken a considerable interest in the Vedic ritual, and the level of scholarship has been high. Among these two groups the Jaiminiyagrhyasutra has therefore been preserved, but in relatively few manuscripts. This scholarly interest is also seen in commentaries to the Sutra written by Bhavatrata and Srinivasa. In addition, there is such a rarity as a Mantravrtti. All three are known both in Kerala and in Tamilnadu, though the communication between the two groups was broken centuries ago.

In the actual ritual, however, the Jaiminiyagrhyasutra is nowhere in use. The Nambudiris are using a Malayalam Prayoga called the Jaiminiyacadaööu, known in many manuscripts, the oldest of which is some 300 years old. The Ayyars use a metric Prayoga, the Vainateyakarika by Vinatananda, a text which is also known in Tamilnadu. The author was a pupil of the commentator Srinivasa, and they both lived in the old Jaiminiya centre of Tittakkuti (South Arcot district of Tamilnadu). Among the Ayyangars, several text have been in use. There is a short Prayoga called Anukramanikai,19 and a more elaborate (but partly identical) Jaimunisamaprayoga, the latter preserved in one manuscript in Tittakkuti. There are also some metric texts like the Srinivasakarika (also called Srinivasa's Prayogaratnamala) and Tarunagnihotrakarika. Both seem to have been rather well known. More rare, at least presently, is the anonymous Grhyakarmakiriyakrama, of which one Malayalam manuscript (and quotations in the Jaiminiyacadaööu) is known, and the text is included in the Jaimunisamaprayoga.20 In addition, there seems to be some shorter Prayogas on individual rituals, but their manuscripts (said to be extant at least in Tittakkuti) have never been fully examined (see Parpola 1973). In addition to these manuscript texts there are now a few printed Prayogas of the daily ritual (in Tamil and in Malayalam) and an extensive, but unfinished synthesis, the Jaiminiyaprayogagavivarana, compiled in Tamil and Sanskrit and published by A. Rangasvami Ayyangar in Kumbakonam in 1923.

In larger Sakhas the amount of extant literature is much grater, of course, but I hope this helps to form an idea of the situation as it has been a fairly long time. It is an interesting case of interaction between oral traditon and literacy. While the most important texts canonised by tradition have always been learnt by heart, and still are, albeit by fewer and fewer Brahmans, in the course of centuries there arose an extensive scholarly and practical literature around them and this is often transmitted in written form. But in spite of this written tradition, we should not undervalue the ability of an Indian Vedic scholar of the traditional kind, trained from the age of seven, to learn by heart every text he considers important enough.

Used special fontface = "Book Antiqua Sanskrit"
[.pdf] For exact view, here is a pdf version of this article,
veda.pdf, size 170 kb.



An early version of this paper was read at the University of Kracow in October 1992 and the present one at the University of Tartu in March 1997. Thereby I would like to thank Professor Asko Parpola, one of my teachers, for the opportunity to observe and study living Vedic traditions in South India, and all my Jaiminiya friends in villages in Kerala and Tamil Nadu. Margot Stout Whiting has kindly checked my English.

1 Thus for instance the vast amalgamation of Pauranic tradition known as the Skandapurana, as far as we can speak of it as a single work at all, cannot be older than the 16th century, as has been shown in the Groningen Skandapurana project (see Adriaensen et al 1994). Many scientific manuals and commentaries were composed during the 17th and 18th centuries, and a 19th century compilation, the Sukraniti, passed for a long time as a genuine ancient work. And of course Indian scholars of traditional learning are all the time producing new Sanskrit literature.

2 For instance, the earliest classics of Hindi, Bengali and Telugu. Tamil, of course, has a different position as the second classical language of ancient India besides Sanskrit.

3 An exception to this is the partial loss of the part of the cult connected with king and government.

4 We can here bypass theories, some even recently proposed, supposing that such an extensive corpus as the Veda could not originate without the help of written texts. Even in our fast-paced century, many traditionally schooled Brahmans know by heart the entire text corpus of their Sakha. What has been said about the oral tradition by non-Indologist scholars like Parry & Lord and Ong mainly applies to epic texts (including the Indian epics). A case completely different from that of canonised religious texts.

5 All this holds true only for India, of course. The arid climate of Central Asia has preserved much older manuscripts. But they are not Vedic.

6 The farce Bhagavadajjukiya makes fun of a charlatan, who cannot even remember his own Sastra.

7 It was a great moment in the history of Western Vedic scholarship when Pischel and Geldner showed that these commentaries rally contain valuable ancient traditions beside the empty speculation ascribed to them by the earlier school of Roth et al.

8 Like some preliminary and auxiliary rites, such as the Yajñopavitadharana, Punyahavacana, Pratisarabandha, Aökurarpana etc.

9 Increasing passivity of women, disappearance of other varnas than Brahmans, disappearance - at least partly - of animal sacrifice.

10 E.g. some astrological rites and the new samskaras of Karnabhusana and Vidyarambha, while the sectarian Brahmans naturally added their own special cults.

11 I have myself seen an earlier Agnicayana altar in Paññal and heard an account of its origin. More information can be found in the great Agnicayana book (Staal 1983) and other accounts of Staal and Parpola, founded on fieldwork in Kerala.

12 See Gonda 1977a:586ff. Other wellknown examples discussed by Gonda, op. cit., are e.g. the Axvalayana-Grhyaparixista (the earlier work with this name, the later one is just a Prayoga; ibid. 605), and the two supplements of the Gobhilagryasutra viz. Gobhilaputra's Grhyasamgrahaparixista and the Karmapradipa (ibid. 609ff).

13 Certainly they are much more than such clumsy and unreliable additions to the Sutras, as many early Vedic scholars used to think.

14 However, very few text editions mention oral informants in their critical apparatus.

15 Especially the so-called aögas, or auxiliary rites, which originally were very simple and quick, have tended to develop into complicated, independent rites. Some were already mentioned in Parixistas, and in manuals they are often very important.

16 I have myself studied printed examples, where there directions have been written in Hindi, Tamil, or Malayalam. Without doubt, further examples could be found in other linguistic areas of India. It is possible that these kinds of works are a reaction to the declining knowledge of Sanskrit. While there still are many great scholars of Sanskrit in India, a simple Vadhyar, or a lay Brahman mainly pursuing worldly aims, but still to some extent interested in the traditions of his Sakha, might welcome instructions in his own language, accompanied by the text of the original mantras (which, as I have seen, can be given even in the clumsy form of the Tamil script, its very nature completely unsuitable for Sanskrit, instead of the traditional Grantha). We are probably entitled to suppose that these works are fairly modern.

17 Thus for instance the Vadhulagrhyasutra has been only preserved in connection with its commentary - the Vyakahya.

18 It must be noted that all prayogas are not Vedic. There are, for instance, Prayogas and Paddhatis on the Pañcaratra Vaisnava ritual (Gonda 1977b:41ff) and on the Saivagama ritual (ibid. 213ff., e.g. the extensive Somaxambhupaddhati (Brybber-Lachaux 1963) of the 11th century).

19 It is written in Sanskrit, but the title is always given with the Tamil ending.

20 On the interrelationship of these texts see further Karttunen 1989-90:141-156.


Adriaensen, R., Bakker, H.T., & Isaacson, H. 1994. «Towards a critical deition of the Skandapurana» Indo-Iranian Journal 37, pp. 325-331.
Brybber-Lachaux, H. (ed.) 1963. Publications de l'Institut Français d'indologie. Pondichéry.
Gonda, J. 1977a. Ritual Sutras. History of Indian Literature 1:2. Wiesbaden.
Gonda, J. 1977b. Mediaeval Religious Literature. History of Indian Literature 2:1. pp. 41 ff. Wiesbaden.
Karttunen, K. 1989-90. «Medieval Texts on the Grhya Ritual of the Jaiminiya Sakha», Bulletin d'études indiennes 7-8, pp. 141-156.
Parpola, A. 1973. The Literature and Study of the Jaiminiya Samaveda in retrospect and prospect. Studia Orientalia 43:6. Helsinki.
Staal, F. (ed.). 1983. Agni, The Vedic Ritual of the Fire Altar. Vol. 1-2. Berkeley.