Sutra , a Comprehensive Study on meaning of the term and its usage
The word sūtra is used to designate two distinct categories of Sanskrit and Middle Indic literature. In one aspect, the word is used to designate collections of short aphoristic rules, each of which is called a sūtra: a Sūtra is in this way a collection of sūtras. Such Sūtras belong primarily, though not exclusively, to the various Śāstras (disciplines, sciences) of the Brahmanical tradition. The other kind of Sūtras are primarily, perhaps exclusively, found in the canonical literature of the Buddhists and the Jainas. These Sūtras are not short and aphoristic, and they can as a matter of fact be long and elaborate.
The difference between these two kinds of Sūtras is striking and has led some scholars to propose two different etymological explanations for the word. The first, mainly Brahmanical, Sūtra would be so called because the primary meaning of the Sanskrit word sūtra is “thread,” “string,” and a Sūtra text is “any work or manual consisting of strings of short sentences or aphoristic rules hanging together like threads.” Alternatively, a Sūtra is like a thread spun from different fibers, because the earliest Sūtras (the Śrautasūtras; see below) consisted of individual statements systematically collected from different sources and joined together (Klaus, 2000; 2004). The Buddhist and Jaina Sūtra, in contrast, would owe their name to the faulty Sanskritization of Middle Indic sutta. The correct Sanskritization of this word would be sūkta, that is, su + ukta (well spoken).
Some (e.g. L. Renou) maintain – mainly on the basis of the Baudhāyanaśrautasūtra, which is of a hybrid character – that the aphoristic Sūtra arose historically as a condensation of more elaborate prose. The existence of the probably older Vādhūlaśrautasūtra (see below), which displays a straightforward Sūtra style, casts doubt on the validity of this observation.
The second explanation of the word sūtra – as a faulty Sanskritization of sutta – has not gone unchallenged, and it may not be correct (Hinüber, 1994, 132n28). No better explanation of the word as used by the Buddhists and the Jainas has, to my knowledge, been proposed. This means that the custom of using the same word to designate two different genres of literature remains, for the time being, unexplained. Some authors do not consider that there is a fundamental difference between the two (Caillat, 1994, 81).
In what follows I will concentrate on the first kind of Sūtras, collections of short aphoristic sentences. In their vast majority they are associated with the Brahmanical tradition. Among the exceptions we may have to count the Prātimokṣasūtra, a collection of rules for Buddhist monks and nuns, which is commented upon in the Sūtravibhaṅga (Nolot, 1994).
An important number of Sūtras is part of vedic ancillary literature. They belong to the Vedāṅgas ("Limbs of the Veda"). Of these there are, traditionally, six:
A considerable number of Sūtras belong to the kalpa Vedāṅga. There are Sūtras that deal with the solemn ritual (the Śrautasūtras), others that deal with domestic ritual (the Gṛhyasūtras), others again that deal with ritual measurements (the Śulbasūtras), and those that concern correct behavior in general (the Dharmasūtras ). Many vedic schools had Sūtras of several or all of these four types. The Vaiṣṇava ritual tradition of the Vaikhānasa has its own Śrautasūtras, thereby emphazising their close association with the vedic tradition (Colas, 1996).
Apart from the Vedāṅgas, there are Sūtras that belong to the main Brahmanical philosophical schools: the Vaiśeṣikasūtra gives expression to the central tenets of Vaiśeṣika, the Nyāyasūtra to those of Nyāya, and the Mīmāṃsāsūtra to the reflections of the school called Mīmāṃsā, while the Brahmasūt
ra is accepted as a foundational text by the various subschools of the philosophy known as Vedānta. (The Tattvasamāsasūtra and the Sāṃkhyasūtra, expressive of Sāṃkhya thought, are both very late compositions, perhaps from the 14th cent. CE; Larson & Bhattacharya, 1987, 315f., 327f.) The existence of a Tattvārthasūtra of the Jainas may no doubt be taken as an indication that this movement wanted to be seen as a philosophical movement on a par with the Brahmanical schools enumerated.
Other Śāstras (disciplines/sciences) that were more or less closely connected with the Brahmanical tradition composed Sūtra texts to lay down their essential teachings. An example is the Kāmasūtra, which concerns itself with “erotic science” ( kāma ). Some theological schools base their doctrines on “revealed” or “discovered” Sūtras. The Pāśupatasūtra (Bisschop, 2006) belongs to a Śaiva sect of that name (Pāśupata) and describes its rituals; its commentator Kauṇḍinya steered the school in a philosophical direction, which is not yet present in the Sūtra (Hara, 2002). Foundational for the traditions of Kashmir Śaivism is the Śivasūtra by Vasugupta (9th cent.), which became the basis of a rich commentarial tradition, such as the Śivasūtravimarśiṇī by Kṣemarāja (Silburn, 1980).
The Yogasūtra is a special case (Maas, 2006, xiif.). It is part of a larger text, called Yogaśāstra, in which later interpreters distinguish between Yogasūtra and Yogabhāṣya (also known as Vyāsabhāṣya). Also the attribution of these “two” texts to two different authors (usually called “Patañjali” and “Vyāsa,” respectively) is late. The expressions Yogasūtra and Yogabhāṣya are not found in the Yogaśāstra, and they appear to have been imposed later. The Yogaśāstra presents itself as a manual of yoga practice and Sāṃkhya philosophy; in due course it came to be looked upon as a treatise on “Yoga philosophy”.
A similar situation may prevail in the Arthaśāstra, the classical Brahmanical text on statecraft ( artha ). It consists of prose and verse, but concludes with the statement that Viṣṇugupta composed both Sūtra and Bhāṣya. Unlike the Yogaśāstra, it is not obvious how the two should be separated (Bronkhorst, 1991, 214f.; Renou, 1961, 187f. [543f.]; Olivelle, 2012, 3-38).
The foundational treatise of a number of Śāstras is not a sūtra but a verse text. Interestingly, these verse texts are in a number of cases treated by their early commentators as if they consisted of sūtras. An example is the Sāṃkhya philosophy, which did not have its own Sūtra text until the 2nd millennium. The verse text called Sāṃkhyakārikā had to serve for a number of centuries as its foundational treatise. (An earlier Sāṃkhya text in prose, the Ṣaṣṭitantra, existed and was known to the Sāṃkhyakārikā; it did not, however, receive as much attention as the latter and was lost.) However, parts of the verses of the Sāṃkhyakārikā are referred to as sūtras in the Yuktidīpikā, its most important commentary dating from around the 7th century CE (Bronkhorst, 1994, 666f.; 2003).
Another example is the Āryabhaṭīya, a verse text on astronomy and mathematics composed in 499 CE. Its earliest surviving commentary, completed by Bhāskara in 629 CE, regularly uses the term sūtra to refer to (parts of) the verses it comments upon (Bronkhorst, 2001, 53). The same applies to the mathematical text contained in the so-called Bakhshali Manuscript (Hayashi, 1995, 84), and to other mathematical and astronomical texts (Bronkhorst, 2001, 70n38).
An example from outside the Brahmanical tradition is the Buddhist verse text called Abhidharmakośa, composed in the 4th or 5th century CE. Its commentary, the Abhidharmakośabhāṣya, was composed, according to tradition, by the same author, Vasubandhu. This did not prevent him from referring to parts of the verses as sūtras. Since this is a Buddhist text, it employs the word sūtra in two ways, both to refer to parts of the verses it comments upon, and to refer to canonical Buddhist Sūtras. Interestingly, the Buddhist author Saṅghabhadra appears to call Vasubandhu an “author of sūtras” (sūtrakāra; Collett, 1995, 56; Kritzer, 2005, xxxi–xxxii).
The Madhyāntavibhāgaśāstra is a Buddhist text that consists of verse and prose. Its colophons also have the name Madhyāntavibhāgakārikābhāṣya, which calls attention to this division (kārikā means verse). The commentator Sthiramati calls the same text Madhyāntavibhāgasūtrabhāṣya, which means no doubt that he, too, looked upon the verses (or part thereof) as sūtras.
An example from Jainism is the Jambūdvīpasaṃgrahaṇī of Haribhadra Sūri (12th cent. CE). It consists of 30 verses. Its commentator Prabhānanda Sūri refers to Haribhadra as sūtrakāra, and to the verse text as Gāthā, occasionally also as Sūtra (e.g. Van Den Bossche, 2007 passim [sūtrakāra]; 192, 200, 251 [sūtra]; 251 [sūtrādikathitayā gāthayā]).
It is difficult to determine which of the surving Sūtras are most ancient. It is often assumed that some of the Sūtras on the solemn ritual (Śrautasūtras) are among the oldest. The Baudhāyanaśrautasūtra, in particular, has been claimed to be early, and to show the signs of a transition from the elaborate style of texts called Brāhmaṇas to the condensed style of Sūtras (Renou, 1947, 173; 1963, 180 [584]; Caland, 1903, 5). The same applies to the Vādhūlaśrautasūtra, whose surviving form may even be older (Voegeli, 2007). Also Pāṇini’s Aṣṭādhyāyī (Eight Chapters), a Sūtra of the vyākaraṇa Vedāṅga, is early. Unlike most other early Sūtras, it can be dated with some precision: it was composed around 350 BCE, probably somewhat later than that, perhaps even under the early Mauryas (Hinüber, 1990, 34; Falk, 1993, 304; 1994, 327n45). The development of the Sūtra style may be explained by the desire to facilitate memorization. The systematizing and didactic character of this literature has been emphasized by modern research (Renou, 1963, 165f. [569f.]).
Whatever the initial reasons to compose Sūtras, the evidence suggests that in later times it became desirable for self-respecting disciplines to have a foundational text in this style. This would explain various features, such as the very late composition of the philosophical Tattvasamāsasūtra and Sāṃkhyasūtra, the fact that a number of commentators treat their foundational texts as consisting of sūtras even though they do not, and the existence of a number of other, late, Sūtras. Some authors, among them the Jaina author Hemacandra, did not hesitate to adopt the Sūtra style for some of their works (Renou, 1963, 171f. [575f.]).
The Sūtra that has no doubt exerted the greatest influence on classical Indian culture in general is Pāṇini’s grammar, the Aṣṭādhyāyī. This text has been studied, for most of its history, along with and in the light of the Mahābhāṣya (Great Commentary) of Patañjali, who lived in the 2nd century BCE. The Mahābhāṣya is a composite text. It consists of two parts: on the one hand, there are the aphoristic comments called Vārttikas, most of them composed by someone called Kātyāyana, who may have lived towards the end of the 3rd century BCE and may, according to some, also have been the author of a Prātiśākhya (the Vājasaneyiprātiśākhya) and perhaps of a Śrautasūtra (the Kātyāyanaśrautasūtra; Thieme, 1935, 96; 1938; 1958, 41f.; Renou, 1947, 184; Scharfe, 1977, 140; Parpola, 1994, 299f.); on the other hand, there are Patañjali’s comments, both on the Vārttikas and on Pāṇini’s sūtras.
Read along with its Vārttikas, the Mahābhāṣya presents itself as a text in prose that is regularly interrupted by short nominal phrases (vārttikas). For a number of centuries, readers of this text did not realize that this peculiar style of the Mahābhāṣya resulted from the fact that in reality, (at least) two texts are involved, composed by (at least) two different authors: the Vārttikas and the Bhāṣya. This fact could escape them because such indicators in the text are few and subject to interpretation. The result of this confusion has been that a number of later commentators imitated what they considered to be the style of the Mahābhāṣya; some of them called their commentaries written in this style Vārttika (Bronkhorst, 1990). The Vārttika style that is adopted in these commentaries is characterized by short nominal sentences in the midst of ordinary prose. When used in a commentary that explains a Sūtra, it can easily give rise to confusion, because it will contain two types of short nominal phrases: the sūtras and the nominal phrases (called vākyas) that characterize the Vārttika style (see e.g. Lang, 1988). This situation may in some cases have led to contamination of the Sūtra text (Bronkhorst, 1991, 221f.).
The vārttikas of the Mahābhāṣya are important for another reason as well. Whereas the old Sūtras have virtually no place for the discussion of various points of view, the grammatical vārttikas are largely concerned with such discussions. Slightly more recent Sūtras, notably the Sūtras belonging to philosophical schools, argue for philosophical positions against real or imaginary critics. The suggestion has therefore been made that these Sūtras may continue a tradition that finds its earliest expression in the vārttikas included in the Mahābhāṣya (Renou, 1963, 169 [573], 181 [585], 191f. [595f.]; compare Parpola, 1994, 299n36).
Many Sūtras have only reached us along with a commentary. The authors of these commentaries are, with rare exceptions, different from the authors of the Sūtras. The sūtras are embedded in that commentary, from which they can usually be extracted without great difficulty. However, the extracted Sūtra text is inevitably the text accepted, or favored, by the author of the commentary. In cases where there is only one early commentary, we can only hope, while reconstituting the Sūtra, that the commentator has not modified the text he commented upon. This hope is not always justified. Some Sūtras have several independent commentaries; in these cases the Sūtra texts in the different commentaries are not always identical. In the case of certain other Sūtras, their inner logic makes it possible to identify insertions or modifications of the original order. The modifications that have been brought to light in these ways are, unfortunately, so numerous that we must fear the worst for Sūtras whose texts allow of no verification. Sūtras are, almost by definition, texts whose historical reliability is uncertain.
These doubts are strengthened by our uncertainty as to the way Sūtras were composed. Many are ascribed to legendary authors and serve the purpose of presenting an authoritative résumé of the teachings of a specific school. Thinkers of the school concerned, including commentators, would not normally express their disagreement with the content of “their” Sūtra. Indeed, commentaries usually display “textual deference” (Smith, 1991) toward the Sūtra text they comment upon by not criticizing it, and occasionally by keeping plain mistakes unchanged (Bronkhorst, 2006). The commentators were in this way more or less obliged to reinterpret sūtras, sometimes in very artificial ways, or to “correct” sūtras or adjust their order so as to obtain the “correct” sequence. Many of the Sūtras that have been preserved have undergone a multitude of such treatments, but only rarely is it possible to find out details.
The Mīmāṃsāsūtra has been handed down along with its earliest surviving commentary of Śabara, the Śābarabhāṣya, which may have to be dated to the middle of the 1st millennium, perhaps the 5th century CE (see Slaje, 2007, 131–132n61; along with Bronkhorst, 2007, 12n14; further Franco, 2002, 282f.). There were earlier commentaries. The Śābarabhāṣya contains a portion of one, known by the name of Vṛttikāragrantha (Frauwallner, 1968, 107–113). The names of other commentators are known, but apart from citations, none of the pre-Śabara commentaries has survived. Subsequent commentators comment upon the Śābarabhāṣya and take Śabara’s readings of the sūtras for granted. For most of the sūtras, we therefore depend exclusively on Śabara.
However, even the limited portion of the Vṛttikāragrantha that is cited in Śabara’s commentary shows that there was no unanimity with regard to the exact reading of all of the sūtras. The Vṛttikāragrantha comments on two sūtras (1.1.4, 5), and it is clear that both its reading and its interpretation of these two sūtras differ in important ways from Śabara’s reading and interpretation (see also Frauwallner, 1968, 106). Śabara’s reading in this particular case is not the only one, and not necessarily the original one. Similar doubts may be justified with regard to many of the remaining sūtras, with the difference that for those remaining sūtras we have no independent testimony concerning what variant readings existed.
It is impossible to know whether the Mīmāṃsāsūtra as we have it is a unified work, composed at one time by one single author (or a small number of them). The sūtras are difficult to understand without the help of a commentary. Few attempts have been made to do so. The most serious attempt to date was made by F.X. Clooney (1990; see also Franco, 2005).
In spite of the numerous difficulties that accompany all attempts at interpretation, it is possible to make some educated guesses about the reasons why this text was composed in the first place. In spite of its proximity to the Śrautasūtras mentioned above, the Mīmāṃsāsūtra appears to address general questions that had not been dealt with in those ritual Sūtras. There is, to begin with, the ambition to write not just for one ritual school, but for all of them. In other words, the Mīmāṃsāsūtra implicitly proclaims the unity of ritual practice and the fundamental identity of the ritual acts prescribed in the different schools. Besides, it is possible that already the original Mīmāṃsāsūtra introduced some notions that characterize Mīmāṃsā thought henceforth, namely, the beginninglessness (anāditva) and authorlessness (apauruṣeyatva) of the Veda. These notions allowed subsequent Mīmāṃsakas to ignore all that pretends to be historical in the Veda: a text without temporal beginning cannot refer to historical events. Followers of the school could now concentrate on what remained – primarily injunctions (vidhi) – and were under no obligation to defend the sometimes improbable stories recounted in the Veda against the criticism of skeptical outsiders (Bronkhorst, 2001a). With these theoretical preliminaries in place, the text can then turn to the interpretation of injunctions, whereby the general principle is that the interpretation must be as close to the text as possible (Bronkhorst, 1997).
Sureśvara, a commentator and perhaps pupil of the famous Vedāntin Śaṅkara, claimed that Jaimini, the author of the Mīmāṃsāsūtra, also composed the Brahmasūtra (Bronkhorst, 2007). This claim is false and is contradicted by the testimony of other Vedāntins, both before and contemporary with Sureśvara. The claim cannot be dissociated from the attempt of certain Vedāntins to present their thought as a (better) form of Mīmāṃsā. Later on – probably not before the 2nd millennium CE, and perhaps for the first time in Yāmunācārya’s Ātmasiddhi – the name Uttaramīmāṃsā ("Later Mīmāṃsā") came into use, to distinguish it from Pūrvamīmāṃsā ("Earlier Mīmāṃsā"), which is the name these thinkers give to the Mīmāṃsā of the Mīmāṃsāsūtra and it commentaries.
The Brahmasūtra belongs to the current of Vedānta thought that considered itself a form of Mīmāṃsā. This is clear from the fact that it refers, with one exception, to exactly the same teachers as those who figure in the Mīmāṃsāsūtra. Late vedic literature provides ample evidence for the interest in ritual matters of these teachers. For their vedantic inclinations, however, there is no other evidence than their mention in the Brahmasūtra. The safest conclusion to be drawn from this strange situation is that these teachers were not associated with vedantic thought in any of its forms. They are mentioned in the Brahmasūtra because the author of this text (perhaps there were several) wanted to present his (or their) form of Vedānta as a school that recognized the same authorities as did Mīmāṃsā. They recognized the same authorities, because at bottom Mīmāṃsā and Vedānta are presented as one and the same school of thought. Not all Vedāntins shared this point of view, but those who commented upon the Brahmasūtra and their followers did. A number of commentators even commented on the combination in 20 chapters of Mīmāṃsāsūtra (12 chapters) plus Brahmasūtra (four chapters) plus a miscellaneous text in between the two, the Saṅkarṣakāṇḍa (four chapters; Bronkhorst, 2007; on the Saṅkarṣakāṇḍa, see Lariviere, 1981; Kanazawa, 1989). Their works have not been preserved.
The date of the Brahmasūtra is hard to determine. It has to be older than the earliest surviving commentary on it, by Śaṅkara (end of 7th cent. CE; Slaje, 2007, 116n1), and younger than the Mīmāṃsāsūtra from which it draws inspiration. A number of indications in the text suggest that it has to be a great deal younger than the Mīmāṃsāsūtra, for it is acquainted with Buddhist systems of thought that were in all probability created much later (Jacobi, 1911, 13f. [571f.]). The fact that Vedānta is not mentioned in lists of philosophical schools until the Madhyamakahṛdaya by the Buddhist Bhavya in the 6th century CE, and that even here Vedānta does not appear as a form of Mīmāṃsā, allows us to consider the possibility that the Brahmasūtra is a late text, much closer in time to its first surviving commentator (Śaṅkara) than is often supposed.
Many commentaries have been written on the Brahmasūtra, some of them already in the 1st millennium CE. A comparison of the texts used in the early commentaries leaves no doubt that there was no unanimity with regard to its exact form. This is not particularly disturbing in this specific case, for the Brahmasūtra is difficult to interpret in any of its forms. It does, however, illustrate the statement made earlier to the extent that the historical reliability of most Sūtras is uncertain.
As in the case of the Mīmāṃsāsūtra, it is possible to make educated guesses as to the overall aim of the Brahmasūtra. Since it presents Vedānta as a form of Mīmāṃsā, its goal is to show that the central tenets of Vedānta thought can be derived from certain upanishadic statements by applying essentially the Mīmāṃsā method of interpretation. Indeed, these tenets can only be known in this way.
A much earlier commentary on the Vaiśeṣikasūtra appears to have existed, a work called Kaṭandī composed by a certain Rāvaṇa (Bronkhorst, 1993). This work is now lost, and it cannot therefore help us in constituting an earlier version of the Sūtra it comments upon. The Kaṭandī itself was commented upon by someone called Praśasta (6th cent. CE; Hattori, 1994, 706): this subcommentary, too, is now lost. However, Praśasta also composed an independent work on Vaiśeṣika in which he frequently refers to sūtras. He called this work Padārthadharmasaṅgraha (Compendium of the Characteristics of the Categories), but it became better known under the name Praśastapādabhāṣya (Commentary of the Venerable Praśasta). This latter name is unfortunate, for the Padārthadharmasaṅgraha is no commentary (pace Tubb & Boose, 2007, 2).
About the date of the original Vaiśeṣikasūtra little can be said. It must, of course, predate Praśasta’s Padārthadharmasaṅgraha and Rāvaṇa’s Kaṭandī. More interesting in the present context is that there is reason to believe that some form of Vaiśeṣika was already known to Aśvaghoṣa, who appears to betray acquaintance with this system in his Buddhacarita (Bronkhorst, 2005). Aśvaghoṣa belongs to the 1st or 2nd century CE. If we assume that the Vaiśeṣikasūtra is the earliest text in which the Vaiśeṣika philosophy found expression, we may then conclude that this text, in one form or other, existed already in the early centuries of the Common Era. The relationship between the Vaiśeṣikasūtra and the medical Carakasaṃhitā ( āyurveda ) remains a debated issue. Both texts may date roughly from the same period (Meulenbeld, 1999–2002, vol. IA, 10f., 113f.; vol. IB, 200).
A detailed reconstruction of the original form of the Vaiśeṣikasūtra will probably never be possible. A comparison of the different versions sometimes allows plausible guesses about an earlier form. The internal analysis of the resulting text may occasionally help to get back even further (for examples, see Bronkhorst, 1993; 1994; Nozawa, 1994; Preisendanz, 1994; Meuthrath, 1999a). Discussions of elements of the Vaiśeṣika philosophy by critical outsiders teach us in some cases that the Vaiśeṣika they knew was different from the classical exposition in Praśasta’s Padārthadharmasaṅgraha, but the information thus obtained normally concerns the Vaiśeṣika of the Kaṭandī, not that of the original Sūtra (see e.g. Bronkhorst, 1996, 2004).
In spite of these numerous difficulties, it seems likely that the Vaiśeṣikasūtra was, from its beginning, concerned with the presentation of an ontological scheme. It is furthermore probable that it had developed this ontological scheme under the influence of the Sarvāstivāda school of Buddhism (Bronkhorst, 1992).
The Nyāyabhāṣya was probably composed in the second half of the 5th century CE (Oberhammer, 1964, 302n1; Franco & Preisendanz, 1995). Most scholars believe that the Nyāyasūtra had at that time already been in existence for a while. Some have made propositions as to what its earlier history may have looked like. The Nyāyasūtra as we know it from the Nyāyabhāṣya is, like the latter, divided into five chapters (adhyāyas), each of which has two “daily portions” (āhnikas). Some scholars have argued that chapters 1 and 5 together (or 1 and 5.2 together) once constituted an independent text dealing with rules of debate (Tucci, 1929, xxiiif.; Oberhammer, 1963, 70f.; 1992; Meuthrath, 1996). The remaining chapters 2–4 must then have been added to this original kernel. The question of whether these added chapters ever existed as an independent treatise is variously answered by different scholars.
Questions that concern the composition of the Nyāyasūtra cannot be separated from those that concern its date or dates. This can be illustrated with the help of the scholarly discussion about the relationship between the Nyāyasūtra and the Vigrahavyāvartanī of Nāgārjuna. Similarities to part of the latter text occur in chapters 2 and 5 of the Nyāyasūtra. If these similarities are close enough to warrant the conclusion of mutual influence (which is not certain), the direction of that influence still remains to be decided. One possibility is that the relevant sūtras in chapter 5 of the Nyāyasūtra influenced (and therefore preceded) Nāgārjuna, a Buddhist philosopher from the 2nd century CE, whereas the sūtras concerned in chapter 2 reacted to (and are therefore more recent than) Nāgārjuna. Other configurations are possible, too (see Bronkhorst, 1985; Oetke, 1991; Meuthrath, 1999).
To the extent that the Nyāyasūtra deals with rules of debate, it belongs to a category of texts of which several exemplars have survived, the oldest perhaps as part of the medical Carakasaṃhitā (Prets, 2000). Other parts of the Nyāyasūtra contain discussions of an ontological and epistemological nature. Some sūtras, perhaps inserted, show that their authors thought of Nyāya as a science concerning the self and did not hesitate to include self-oriented yogic practices (Preisendanz, 2000).
"the author of the Yogabhāṣya brought the Yogasūtras together, perhaps from different sources, and wrote a commentary which in some cases demonstrably deviated from the original intention of the sūtras. It seems probable that deviations from the original meanings were made primarily to suit the theoretical tastes of the author of the Yoga Bhāṣya" (Bronkhorst, 1984, 203).
They also suggest that the author of the Yogabhāṣya may not have had any direct experience with yogic states (Bronkhorst, 1984, 203).
We further saw that the Yogaśāstra presents itself as a manual of yoga practice and Sāṃkhya philosophy (Maas, 2006, xx–xxi). The theoretical side does not find much expression in the sūtras. The strong influence of Buddhism on the yoga practice of the sūtras has long been recognized (Senart, 1900; La Vallée Poussin, 1936–1937; Bronkhorst, 1986, 65–70 [68–75]; Bronkhorst, 2009). The yoga of the Yogasūtra is for this reason in many respects quite different from the yoga we find in earlier texts such as the Mahābhārata .
However, fragments of the Sūtra text it once possessed, and of the commentaries on it that once existed (at least two), have been preserved in the works of its critics. These fragments have been collected (most recently in Bhattacharya, 2003) and allow us to correct at least some of the incorrect notions with which later tradition has burdened the memory of the Cārvākas.
The collection of Cārvāka fragments is confronted with serious problems. It appears that Cārvāka texts were known until about the 12th century CE. More recent texts – most notably the Sarvadarśanasaṃgraha, a 14th-century text that has been the point of departure of modern Cārvāka studies – were no longer acquainted with any Cārvāka texts. They still attributed opinions to the Cārvākas, but it seems clear that some of these were based on prejudice rather than reliable information.
This Cārvākasūtra is sometimes referred to as Bārhaspatyasūtra and ascribed to the mythical seer Bṛhaspati. It claims that the four elements earth, water, fire, and wind are the ultimate building blocks of our universe. Even consciousness is stated to be derived from these. The soul is nothing but the body endowed with consciousness. Claims like these are responsible for the modern custom of referring to the school as materialistic. This custom, though justified, should not make us forget that the primary concern of the school does not appear to have been to develop a materialistic worldview, but rather to reject certain beliefs – rebirth and karmic retribution, life after death – held by others. This negative purpose led to other denials, such as, “There is no means of knowledge for determining (the existence of) the other world,” and “There is no other world because of the absence of any other-worldly being (i.e. a transmigrating self)” (Bhattacharya, 2003, 605, 612).
The exact date of composition of the Tattvārthasūtra remains uncertain, but may have been in or near the 4th century CE (Balcerowicz, 2008, 35n23; for bibliographical information, see Wiles, 1998; for a translation, Tatia, 1994). The two oldest commentaries on the Tattvārthasūtra are the Tattvārthādhigamabhāṣya (whose authorship and date will be discussed below) and the Sarvārthasiddhi of Devanandin (perhaps second half of 5th century CE; Bronkhorst, 1985a, 161). These two commentaries belong to the two main divisions of Jainism: Śvetāmbara and Digambara, respectively. Both divisions claim that the original Tattvārthasūtra belonged to them. The Śvetāmbaras go further and maintain that the Tattvārthasūtra and the Tattvārthādhigamabhāṣya were written by one and the same person, Umāsvāti. Neither of these claims resists a detailed consideration of the evidence. The original Tattvārthasūtra may have belonged neither to the Śvetāmbaras nor to the Digambaras, but to the Yāpanīyas, a third division of Jainism that no longer survives. Further, the author of the Tattvārthādhigamabhāṣya did not also compose the Tattvārthasūtra. What he did do was incorporate the Sūtra text into his work in a way that is reminiscent of the Yogaśāstra of “Patañjali” and of the Arthaśāstra (discussed above). Umāsvāti may have composed this work in the first half of the 5th century CE.
It goes almost without saying that the Tattvārthādhigamabhāṣya and the Sarvārthasiddhi do not comment on exactly the same Sūtra text. There are some indications to the effect that the Sarvārthasiddhi sometimes comments on a version closer to the original. A full reconstruction of the original texts seems impossible (Bronkhorst, 1985a).
It seems clear that the aphoristic Sūtra style is primarily a feature of Brahmanical technical literature. Where Buddhist or Jaina works are composed in this style, Brahmanical influence is most often evident.
The difference between these two kinds of Sūtras is striking and has led some scholars to propose two different etymological explanations for the word. The first, mainly Brahmanical, Sūtra would be so called because the primary meaning of the Sanskrit word sūtra is “thread,” “string,” and a Sūtra text is “any work or manual consisting of strings of short sentences or aphoristic rules hanging together like threads.” Alternatively, a Sūtra is like a thread spun from different fibers, because the earliest Sūtras (the Śrautasūtras; see below) consisted of individual statements systematically collected from different sources and joined together (Klaus, 2000; 2004). The Buddhist and Jaina Sūtra, in contrast, would owe their name to the faulty Sanskritization of Middle Indic sutta. The correct Sanskritization of this word would be sūkta, that is, su + ukta (well spoken).
Some (e.g. L. Renou) maintain – mainly on the basis of the Baudhāyanaśrautasūtra, which is of a hybrid character – that the aphoristic Sūtra arose historically as a condensation of more elaborate prose. The existence of the probably older Vādhūlaśrautasūtra (see below), which displays a straightforward Sūtra style, casts doubt on the validity of this observation.
The second explanation of the word sūtra – as a faulty Sanskritization of sutta – has not gone unchallenged, and it may not be correct (Hinüber, 1994, 132n28). No better explanation of the word as used by the Buddhists and the Jainas has, to my knowledge, been proposed. This means that the custom of using the same word to designate two different genres of literature remains, for the time being, unexplained. Some authors do not consider that there is a fundamental difference between the two (Caillat, 1994, 81).
In what follows I will concentrate on the first kind of Sūtras, collections of short aphoristic sentences. In their vast majority they are associated with the Brahmanical tradition. Among the exceptions we may have to count the Prātimokṣasūtra, a collection of rules for Buddhist monks and nuns, which is commented upon in the Sūtravibhaṅga (Nolot, 1994).
An important number of Sūtras is part of vedic ancillary literature. They belong to the Vedāṅgas ("Limbs of the Veda"). Of these there are, traditionally, six:
- 1.
- the science of proper articulation and pronunciation (śikṣā);
- 2.
- meter (chandas);
- 3.
- grammar (vyākaraṇa);
- 4.
- etymological explanation (nirukta);
- 5.
- astronomy and calendar (jyotiṣa); and
- 6.
- ceremonial (kalpa).
A considerable number of Sūtras belong to the kalpa Vedāṅga. There are Sūtras that deal with the solemn ritual (the Śrautasūtras), others that deal with domestic ritual (the Gṛhyasūtras), others again that deal with ritual measurements (the Śulbasūtras), and those that concern correct behavior in general (the Dharmasūtras ). Many vedic schools had Sūtras of several or all of these four types. The Vaiṣṇava ritual tradition of the Vaikhānasa has its own Śrautasūtras, thereby emphazising their close association with the vedic tradition (Colas, 1996).
Apart from the Vedāṅgas, there are Sūtras that belong to the main Brahmanical philosophical schools: the Vaiśeṣikasūtra gives expression to the central tenets of Vaiśeṣika, the Nyāyasūtra to those of Nyāya, and the Mīmāṃsāsūtra to the reflections of the school called Mīmāṃsā, while the Brahmasūt
ra is accepted as a foundational text by the various subschools of the philosophy known as Vedānta. (The Tattvasamāsasūtra and the Sāṃkhyasūtra, expressive of Sāṃkhya thought, are both very late compositions, perhaps from the 14th cent. CE; Larson & Bhattacharya, 1987, 315f., 327f.) The existence of a Tattvārthasūtra of the Jainas may no doubt be taken as an indication that this movement wanted to be seen as a philosophical movement on a par with the Brahmanical schools enumerated.
Other Śāstras (disciplines/sciences) that were more or less closely connected with the Brahmanical tradition composed Sūtra texts to lay down their essential teachings. An example is the Kāmasūtra, which concerns itself with “erotic science” ( kāma ). Some theological schools base their doctrines on “revealed” or “discovered” Sūtras. The Pāśupatasūtra (Bisschop, 2006) belongs to a Śaiva sect of that name (Pāśupata) and describes its rituals; its commentator Kauṇḍinya steered the school in a philosophical direction, which is not yet present in the Sūtra (Hara, 2002). Foundational for the traditions of Kashmir Śaivism is the Śivasūtra by Vasugupta (9th cent.), which became the basis of a rich commentarial tradition, such as the Śivasūtravimarśiṇī by Kṣemarāja (Silburn, 1980).
The Yogasūtra is a special case (Maas, 2006, xiif.). It is part of a larger text, called Yogaśāstra, in which later interpreters distinguish between Yogasūtra and Yogabhāṣya (also known as Vyāsabhāṣya). Also the attribution of these “two” texts to two different authors (usually called “Patañjali” and “Vyāsa,” respectively) is late. The expressions Yogasūtra and Yogabhāṣya are not found in the Yogaśāstra, and they appear to have been imposed later. The Yogaśāstra presents itself as a manual of yoga practice and Sāṃkhya philosophy; in due course it came to be looked upon as a treatise on “Yoga philosophy”.
A similar situation may prevail in the Arthaśāstra, the classical Brahmanical text on statecraft ( artha ). It consists of prose and verse, but concludes with the statement that Viṣṇugupta composed both Sūtra and Bhāṣya. Unlike the Yogaśāstra, it is not obvious how the two should be separated (Bronkhorst, 1991, 214f.; Renou, 1961, 187f. [543f.]; Olivelle, 2012, 3-38).
The foundational treatise of a number of Śāstras is not a sūtra but a verse text. Interestingly, these verse texts are in a number of cases treated by their early commentators as if they consisted of sūtras. An example is the Sāṃkhya philosophy, which did not have its own Sūtra text until the 2nd millennium. The verse text called Sāṃkhyakārikā had to serve for a number of centuries as its foundational treatise. (An earlier Sāṃkhya text in prose, the Ṣaṣṭitantra, existed and was known to the Sāṃkhyakārikā; it did not, however, receive as much attention as the latter and was lost.) However, parts of the verses of the Sāṃkhyakārikā are referred to as sūtras in the Yuktidīpikā, its most important commentary dating from around the 7th century CE (Bronkhorst, 1994, 666f.; 2003).
Another example is the Āryabhaṭīya, a verse text on astronomy and mathematics composed in 499 CE. Its earliest surviving commentary, completed by Bhāskara in 629 CE, regularly uses the term sūtra to refer to (parts of) the verses it comments upon (Bronkhorst, 2001, 53). The same applies to the mathematical text contained in the so-called Bakhshali Manuscript (Hayashi, 1995, 84), and to other mathematical and astronomical texts (Bronkhorst, 2001, 70n38).
An example from outside the Brahmanical tradition is the Buddhist verse text called Abhidharmakośa, composed in the 4th or 5th century CE. Its commentary, the Abhidharmakośabhāṣya, was composed, according to tradition, by the same author, Vasubandhu. This did not prevent him from referring to parts of the verses as sūtras. Since this is a Buddhist text, it employs the word sūtra in two ways, both to refer to parts of the verses it comments upon, and to refer to canonical Buddhist Sūtras. Interestingly, the Buddhist author Saṅghabhadra appears to call Vasubandhu an “author of sūtras” (sūtrakāra; Collett, 1995, 56; Kritzer, 2005, xxxi–xxxii).
The Madhyāntavibhāgaśāstra is a Buddhist text that consists of verse and prose. Its colophons also have the name Madhyāntavibhāgakārikābhāṣya, which calls attention to this division (kārikā means verse). The commentator Sthiramati calls the same text Madhyāntavibhāgasūtrabhāṣya, which means no doubt that he, too, looked upon the verses (or part thereof) as sūtras.
An example from Jainism is the Jambūdvīpasaṃgrahaṇī of Haribhadra Sūri (12th cent. CE). It consists of 30 verses. Its commentator Prabhānanda Sūri refers to Haribhadra as sūtrakāra, and to the verse text as Gāthā, occasionally also as Sūtra (e.g. Van Den Bossche, 2007 passim [sūtrakāra]; 192, 200, 251 [sūtra]; 251 [sūtrādikathitayā gāthayā]).
It is difficult to determine which of the surving Sūtras are most ancient. It is often assumed that some of the Sūtras on the solemn ritual (Śrautasūtras) are among the oldest. The Baudhāyanaśrautasūtra, in particular, has been claimed to be early, and to show the signs of a transition from the elaborate style of texts called Brāhmaṇas to the condensed style of Sūtras (Renou, 1947, 173; 1963, 180 [584]; Caland, 1903, 5). The same applies to the Vādhūlaśrautasūtra, whose surviving form may even be older (Voegeli, 2007). Also Pāṇini’s Aṣṭādhyāyī (Eight Chapters), a Sūtra of the vyākaraṇa Vedāṅga, is early. Unlike most other early Sūtras, it can be dated with some precision: it was composed around 350 BCE, probably somewhat later than that, perhaps even under the early Mauryas (Hinüber, 1990, 34; Falk, 1993, 304; 1994, 327n45). The development of the Sūtra style may be explained by the desire to facilitate memorization. The systematizing and didactic character of this literature has been emphasized by modern research (Renou, 1963, 165f. [569f.]).
Whatever the initial reasons to compose Sūtras, the evidence suggests that in later times it became desirable for self-respecting disciplines to have a foundational text in this style. This would explain various features, such as the very late composition of the philosophical Tattvasamāsasūtra and Sāṃkhyasūtra, the fact that a number of commentators treat their foundational texts as consisting of sūtras even though they do not, and the existence of a number of other, late, Sūtras. Some authors, among them the Jaina author Hemacandra, did not hesitate to adopt the Sūtra style for some of their works (Renou, 1963, 171f. [575f.]).
The Sūtra that has no doubt exerted the greatest influence on classical Indian culture in general is Pāṇini’s grammar, the Aṣṭādhyāyī. This text has been studied, for most of its history, along with and in the light of the Mahābhāṣya (Great Commentary) of Patañjali, who lived in the 2nd century BCE. The Mahābhāṣya is a composite text. It consists of two parts: on the one hand, there are the aphoristic comments called Vārttikas, most of them composed by someone called Kātyāyana, who may have lived towards the end of the 3rd century BCE and may, according to some, also have been the author of a Prātiśākhya (the Vājasaneyiprātiśākhya) and perhaps of a Śrautasūtra (the Kātyāyanaśrautasūtra; Thieme, 1935, 96; 1938; 1958, 41f.; Renou, 1947, 184; Scharfe, 1977, 140; Parpola, 1994, 299f.); on the other hand, there are Patañjali’s comments, both on the Vārttikas and on Pāṇini’s sūtras.
Read along with its Vārttikas, the Mahābhāṣya presents itself as a text in prose that is regularly interrupted by short nominal phrases (vārttikas). For a number of centuries, readers of this text did not realize that this peculiar style of the Mahābhāṣya resulted from the fact that in reality, (at least) two texts are involved, composed by (at least) two different authors: the Vārttikas and the Bhāṣya. This fact could escape them because such indicators in the text are few and subject to interpretation. The result of this confusion has been that a number of later commentators imitated what they considered to be the style of the Mahābhāṣya; some of them called their commentaries written in this style Vārttika (Bronkhorst, 1990). The Vārttika style that is adopted in these commentaries is characterized by short nominal sentences in the midst of ordinary prose. When used in a commentary that explains a Sūtra, it can easily give rise to confusion, because it will contain two types of short nominal phrases: the sūtras and the nominal phrases (called vākyas) that characterize the Vārttika style (see e.g. Lang, 1988). This situation may in some cases have led to contamination of the Sūtra text (Bronkhorst, 1991, 221f.).
The vārttikas of the Mahābhāṣya are important for another reason as well. Whereas the old Sūtras have virtually no place for the discussion of various points of view, the grammatical vārttikas are largely concerned with such discussions. Slightly more recent Sūtras, notably the Sūtras belonging to philosophical schools, argue for philosophical positions against real or imaginary critics. The suggestion has therefore been made that these Sūtras may continue a tradition that finds its earliest expression in the vārttikas included in the Mahābhāṣya (Renou, 1963, 169 [573], 181 [585], 191f. [595f.]; compare Parpola, 1994, 299n36).
Many Sūtras have only reached us along with a commentary. The authors of these commentaries are, with rare exceptions, different from the authors of the Sūtras. The sūtras are embedded in that commentary, from which they can usually be extracted without great difficulty. However, the extracted Sūtra text is inevitably the text accepted, or favored, by the author of the commentary. In cases where there is only one early commentary, we can only hope, while reconstituting the Sūtra, that the commentator has not modified the text he commented upon. This hope is not always justified. Some Sūtras have several independent commentaries; in these cases the Sūtra texts in the different commentaries are not always identical. In the case of certain other Sūtras, their inner logic makes it possible to identify insertions or modifications of the original order. The modifications that have been brought to light in these ways are, unfortunately, so numerous that we must fear the worst for Sūtras whose texts allow of no verification. Sūtras are, almost by definition, texts whose historical reliability is uncertain.
These doubts are strengthened by our uncertainty as to the way Sūtras were composed. Many are ascribed to legendary authors and serve the purpose of presenting an authoritative résumé of the teachings of a specific school. Thinkers of the school concerned, including commentators, would not normally express their disagreement with the content of “their” Sūtra. Indeed, commentaries usually display “textual deference” (Smith, 1991) toward the Sūtra text they comment upon by not criticizing it, and occasionally by keeping plain mistakes unchanged (Bronkhorst, 2006). The commentators were in this way more or less obliged to reinterpret sūtras, sometimes in very artificial ways, or to “correct” sūtras or adjust their order so as to obtain the “correct” sequence. Many of the Sūtras that have been preserved have undergone a multitude of such treatments, but only rarely is it possible to find out details.
The Philosophical Sūtras
Mīmāṃsāsūtra
The Mīmāṃsāsūtra, traditionally attributed to someone called Jaimini, is by far the longest of the surviving philosophical Sūtras: all by itself it is, with some 2,700 sūtras in 12 chapters (2,745 according to Renou, 1961, 207 [563]), longer than the Sūtras of Nyāya, Vaiśeṣika, Sāṃkhya, Yoga, and Vedānta combined. The original Mīmāṃsāsūtra is close – in terms of subject matter, presentation, and date – to the Śrautasūtras (Parpola, 1981, 164f.; 1994). Like the latter, it deals with the solemn ritual.The Mīmāṃsāsūtra has been handed down along with its earliest surviving commentary of Śabara, the Śābarabhāṣya, which may have to be dated to the middle of the 1st millennium, perhaps the 5th century CE (see Slaje, 2007, 131–132n61; along with Bronkhorst, 2007, 12n14; further Franco, 2002, 282f.). There were earlier commentaries. The Śābarabhāṣya contains a portion of one, known by the name of Vṛttikāragrantha (Frauwallner, 1968, 107–113). The names of other commentators are known, but apart from citations, none of the pre-Śabara commentaries has survived. Subsequent commentators comment upon the Śābarabhāṣya and take Śabara’s readings of the sūtras for granted. For most of the sūtras, we therefore depend exclusively on Śabara.
However, even the limited portion of the Vṛttikāragrantha that is cited in Śabara’s commentary shows that there was no unanimity with regard to the exact reading of all of the sūtras. The Vṛttikāragrantha comments on two sūtras (1.1.4, 5), and it is clear that both its reading and its interpretation of these two sūtras differ in important ways from Śabara’s reading and interpretation (see also Frauwallner, 1968, 106). Śabara’s reading in this particular case is not the only one, and not necessarily the original one. Similar doubts may be justified with regard to many of the remaining sūtras, with the difference that for those remaining sūtras we have no independent testimony concerning what variant readings existed.
It is impossible to know whether the Mīmāṃsāsūtra as we have it is a unified work, composed at one time by one single author (or a small number of them). The sūtras are difficult to understand without the help of a commentary. Few attempts have been made to do so. The most serious attempt to date was made by F.X. Clooney (1990; see also Franco, 2005).
In spite of the numerous difficulties that accompany all attempts at interpretation, it is possible to make some educated guesses about the reasons why this text was composed in the first place. In spite of its proximity to the Śrautasūtras mentioned above, the Mīmāṃsāsūtra appears to address general questions that had not been dealt with in those ritual Sūtras. There is, to begin with, the ambition to write not just for one ritual school, but for all of them. In other words, the Mīmāṃsāsūtra implicitly proclaims the unity of ritual practice and the fundamental identity of the ritual acts prescribed in the different schools. Besides, it is possible that already the original Mīmāṃsāsūtra introduced some notions that characterize Mīmāṃsā thought henceforth, namely, the beginninglessness (anāditva) and authorlessness (apauruṣeyatva) of the Veda. These notions allowed subsequent Mīmāṃsakas to ignore all that pretends to be historical in the Veda: a text without temporal beginning cannot refer to historical events. Followers of the school could now concentrate on what remained – primarily injunctions (vidhi) – and were under no obligation to defend the sometimes improbable stories recounted in the Veda against the criticism of skeptical outsiders (Bronkhorst, 2001a). With these theoretical preliminaries in place, the text can then turn to the interpretation of injunctions, whereby the general principle is that the interpretation must be as close to the text as possible (Bronkhorst, 1997).
Brahmasūtra
Of all surviving Sūtra works, the four chapters that constitute the Brahmasūtra – attributed by tradition to a Bādarāyaṇa – may carry the dubious distinction of being the least intelligible. Numerous modern researchers have commented upon this and have even made the suggestion that the author(s) of the text cultivated this kind of unintelligibility on purpose (Deussen, 1923, 28; Renou, 1942, 122 [444, 328]; 1961, 206 [562]; Rüping, 1977, 2; further Thibaut, 1890–1896, vol. I, xiii–xiv; Renou, 1961, 197 [553]; 1962: 202 [628]). It is as a result extremely difficult to extract useful information directly from this text. Attempts are yet sometimes made (see e.g. Ghate, 1918; Modi, n.d.; Adams, 1993).Sureśvara, a commentator and perhaps pupil of the famous Vedāntin Śaṅkara, claimed that Jaimini, the author of the Mīmāṃsāsūtra, also composed the Brahmasūtra (Bronkhorst, 2007). This claim is false and is contradicted by the testimony of other Vedāntins, both before and contemporary with Sureśvara. The claim cannot be dissociated from the attempt of certain Vedāntins to present their thought as a (better) form of Mīmāṃsā. Later on – probably not before the 2nd millennium CE, and perhaps for the first time in Yāmunācārya’s Ātmasiddhi – the name Uttaramīmāṃsā ("Later Mīmāṃsā") came into use, to distinguish it from Pūrvamīmāṃsā ("Earlier Mīmāṃsā"), which is the name these thinkers give to the Mīmāṃsā of the Mīmāṃsāsūtra and it commentaries.
The Brahmasūtra belongs to the current of Vedānta thought that considered itself a form of Mīmāṃsā. This is clear from the fact that it refers, with one exception, to exactly the same teachers as those who figure in the Mīmāṃsāsūtra. Late vedic literature provides ample evidence for the interest in ritual matters of these teachers. For their vedantic inclinations, however, there is no other evidence than their mention in the Brahmasūtra. The safest conclusion to be drawn from this strange situation is that these teachers were not associated with vedantic thought in any of its forms. They are mentioned in the Brahmasūtra because the author of this text (perhaps there were several) wanted to present his (or their) form of Vedānta as a school that recognized the same authorities as did Mīmāṃsā. They recognized the same authorities, because at bottom Mīmāṃsā and Vedānta are presented as one and the same school of thought. Not all Vedāntins shared this point of view, but those who commented upon the Brahmasūtra and their followers did. A number of commentators even commented on the combination in 20 chapters of Mīmāṃsāsūtra (12 chapters) plus Brahmasūtra (four chapters) plus a miscellaneous text in between the two, the Saṅkarṣakāṇḍa (four chapters; Bronkhorst, 2007; on the Saṅkarṣakāṇḍa, see Lariviere, 1981; Kanazawa, 1989). Their works have not been preserved.
The date of the Brahmasūtra is hard to determine. It has to be older than the earliest surviving commentary on it, by Śaṅkara (end of 7th cent. CE; Slaje, 2007, 116n1), and younger than the Mīmāṃsāsūtra from which it draws inspiration. A number of indications in the text suggest that it has to be a great deal younger than the Mīmāṃsāsūtra, for it is acquainted with Buddhist systems of thought that were in all probability created much later (Jacobi, 1911, 13f. [571f.]). The fact that Vedānta is not mentioned in lists of philosophical schools until the Madhyamakahṛdaya by the Buddhist Bhavya in the 6th century CE, and that even here Vedānta does not appear as a form of Mīmāṃsā, allows us to consider the possibility that the Brahmasūtra is a late text, much closer in time to its first surviving commentator (Śaṅkara) than is often supposed.
Many commentaries have been written on the Brahmasūtra, some of them already in the 1st millennium CE. A comparison of the texts used in the early commentaries leaves no doubt that there was no unanimity with regard to its exact form. This is not particularly disturbing in this specific case, for the Brahmasūtra is difficult to interpret in any of its forms. It does, however, illustrate the statement made earlier to the extent that the historical reliability of most Sūtras is uncertain.
As in the case of the Mīmāṃsāsūtra, it is possible to make educated guesses as to the overall aim of the Brahmasūtra. Since it presents Vedānta as a form of Mīmāṃsā, its goal is to show that the central tenets of Vedānta thought can be derived from certain upanishadic statements by applying essentially the Mīmāṃsā method of interpretation. Indeed, these tenets can only be known in this way.
Vaiśeṣikasūtra
The Vaiśeṣikasūtra is the earliest surviving text of the philosophical school of the same name. It survives in a number of different versions. Three of these are commented upon by Candrānanda (probably between the 7th and 10th cents. CE), Bhaṭṭa Vādīndra (13th cent. CE), and Śaṅkara Miśra (15th cent. CE), respectively. Two further versions, these without commentaries, have recently been brought to light (Isaacson, 1994; 1995 includes a concordance). The differences between these versions are considerable.A much earlier commentary on the Vaiśeṣikasūtra appears to have existed, a work called Kaṭandī composed by a certain Rāvaṇa (Bronkhorst, 1993). This work is now lost, and it cannot therefore help us in constituting an earlier version of the Sūtra it comments upon. The Kaṭandī itself was commented upon by someone called Praśasta (6th cent. CE; Hattori, 1994, 706): this subcommentary, too, is now lost. However, Praśasta also composed an independent work on Vaiśeṣika in which he frequently refers to sūtras. He called this work Padārthadharmasaṅgraha (Compendium of the Characteristics of the Categories), but it became better known under the name Praśastapādabhāṣya (Commentary of the Venerable Praśasta). This latter name is unfortunate, for the Padārthadharmasaṅgraha is no commentary (pace Tubb & Boose, 2007, 2).
About the date of the original Vaiśeṣikasūtra little can be said. It must, of course, predate Praśasta’s Padārthadharmasaṅgraha and Rāvaṇa’s Kaṭandī. More interesting in the present context is that there is reason to believe that some form of Vaiśeṣika was already known to Aśvaghoṣa, who appears to betray acquaintance with this system in his Buddhacarita (Bronkhorst, 2005). Aśvaghoṣa belongs to the 1st or 2nd century CE. If we assume that the Vaiśeṣikasūtra is the earliest text in which the Vaiśeṣika philosophy found expression, we may then conclude that this text, in one form or other, existed already in the early centuries of the Common Era. The relationship between the Vaiśeṣikasūtra and the medical Carakasaṃhitā ( āyurveda ) remains a debated issue. Both texts may date roughly from the same period (Meulenbeld, 1999–2002, vol. IA, 10f., 113f.; vol. IB, 200).
A detailed reconstruction of the original form of the Vaiśeṣikasūtra will probably never be possible. A comparison of the different versions sometimes allows plausible guesses about an earlier form. The internal analysis of the resulting text may occasionally help to get back even further (for examples, see Bronkhorst, 1993; 1994; Nozawa, 1994; Preisendanz, 1994; Meuthrath, 1999a). Discussions of elements of the Vaiśeṣika philosophy by critical outsiders teach us in some cases that the Vaiśeṣika they knew was different from the classical exposition in Praśasta’s Padārthadharmasaṅgraha, but the information thus obtained normally concerns the Vaiśeṣika of the Kaṭandī, not that of the original Sūtra (see e.g. Bronkhorst, 1996, 2004).
In spite of these numerous difficulties, it seems likely that the Vaiśeṣikasūtra was, from its beginning, concerned with the presentation of an ontological scheme. It is furthermore probable that it had developed this ontological scheme under the influence of the Sarvāstivāda school of Buddhism (Bronkhorst, 1992).
Nyāyasūtra
The Nyāyasūtra, attributed variously to Gautama or Akṣapāda, has come down to us embedded in the commentary called Nyāyabhāṣya of Vātsyāyana or Pakṣilasvāmin. (There are manuscripts that contain only the Nyāyasūtra, but these turn out to stand in a complex relationship with surviving commentaries; see Muroya, 2007.) The Nyāyabhāṣya is written in the Vārttika style discussed above (Windisch, 1888; Bronkhorst, 1990). While it is true that some manuscripts contain indications to distinguish Sūtras from commentary (Muroya, 2006, 27f.), extraction of the Sūtra text is yet hindered by ambiguities, whose resolution has led to disagreements among both traditional and modern scholars (see e.g. Preisendanz, 1994a, 293f., 422f., 524,f., 642f., 667f.; Muroya, 2006, 37f.). If we add to this the variant readings found in the manuscripts of the Nyāyabhāṣya (e.g. Muroya, 2006, 28f.), it will be clear that our knowledge of the text of the Nyāyasūtra known to and used by the author of that text is not perfect. Still the best reconstruction of that Nyāyasūtra as a whole is the one presented in W. Ruben (1928).The Nyāyabhāṣya was probably composed in the second half of the 5th century CE (Oberhammer, 1964, 302n1; Franco & Preisendanz, 1995). Most scholars believe that the Nyāyasūtra had at that time already been in existence for a while. Some have made propositions as to what its earlier history may have looked like. The Nyāyasūtra as we know it from the Nyāyabhāṣya is, like the latter, divided into five chapters (adhyāyas), each of which has two “daily portions” (āhnikas). Some scholars have argued that chapters 1 and 5 together (or 1 and 5.2 together) once constituted an independent text dealing with rules of debate (Tucci, 1929, xxiiif.; Oberhammer, 1963, 70f.; 1992; Meuthrath, 1996). The remaining chapters 2–4 must then have been added to this original kernel. The question of whether these added chapters ever existed as an independent treatise is variously answered by different scholars.
Questions that concern the composition of the Nyāyasūtra cannot be separated from those that concern its date or dates. This can be illustrated with the help of the scholarly discussion about the relationship between the Nyāyasūtra and the Vigrahavyāvartanī of Nāgārjuna. Similarities to part of the latter text occur in chapters 2 and 5 of the Nyāyasūtra. If these similarities are close enough to warrant the conclusion of mutual influence (which is not certain), the direction of that influence still remains to be decided. One possibility is that the relevant sūtras in chapter 5 of the Nyāyasūtra influenced (and therefore preceded) Nāgārjuna, a Buddhist philosopher from the 2nd century CE, whereas the sūtras concerned in chapter 2 reacted to (and are therefore more recent than) Nāgārjuna. Other configurations are possible, too (see Bronkhorst, 1985; Oetke, 1991; Meuthrath, 1999).
To the extent that the Nyāyasūtra deals with rules of debate, it belongs to a category of texts of which several exemplars have survived, the oldest perhaps as part of the medical Carakasaṃhitā (Prets, 2000). Other parts of the Nyāyasūtra contain discussions of an ontological and epistemological nature. Some sūtras, perhaps inserted, show that their authors thought of Nyāya as a science concerning the self and did not hesitate to include self-oriented yogic practices (Preisendanz, 2000).
Yogasūtra
It was pointed out above that the Yogasūtra was part of a larger text called Yogaśāstra attributed to one single author who appears to have lived around the year 400 CE (Bronkhorst, 1985a). This does not necessarily mean that this single author, presumably called “Patañjali,” did indeed compose both the Yogabhāṣya and all of the Yogasūtra. Some scholars think that there was a commentary on the Yogasūtra older than the Yogabhāṣya (Angot, 2008, 25), but there is no evidence to support this. There are, however, indications that suggest that"the author of the Yogabhāṣya brought the Yogasūtras together, perhaps from different sources, and wrote a commentary which in some cases demonstrably deviated from the original intention of the sūtras. It seems probable that deviations from the original meanings were made primarily to suit the theoretical tastes of the author of the Yoga Bhāṣya" (Bronkhorst, 1984, 203).
They also suggest that the author of the Yogabhāṣya may not have had any direct experience with yogic states (Bronkhorst, 1984, 203).
We further saw that the Yogaśāstra presents itself as a manual of yoga practice and Sāṃkhya philosophy (Maas, 2006, xx–xxi). The theoretical side does not find much expression in the sūtras. The strong influence of Buddhism on the yoga practice of the sūtras has long been recognized (Senart, 1900; La Vallée Poussin, 1936–1937; Bronkhorst, 1986, 65–70 [68–75]; Bronkhorst, 2009). The yoga of the Yogasūtra is for this reason in many respects quite different from the yoga we find in earlier texts such as the Mahābhārata .
Cārvākasūtra
The Cārvākas or Lokāyatas constituted a Brahmanical school of thought that resisted the belief in rebirth and karmic retribution, and more in general the existence of an “other world” (Bronkhorst, 2007a, 142–159, 309–328). No doubt as a consequence of the refusal to accept the belief in rebirth and karmic retribution – a belief held by all of their opponents, whether Brahmanical, Buddhist, or Jaina – this school of philosophy did not survive roughly beyond the end of the 1st millennium CE. Criticism of this school continued well after its disappearance, and positions came to be attributed to it that it never held.However, fragments of the Sūtra text it once possessed, and of the commentaries on it that once existed (at least two), have been preserved in the works of its critics. These fragments have been collected (most recently in Bhattacharya, 2003) and allow us to correct at least some of the incorrect notions with which later tradition has burdened the memory of the Cārvākas.
The collection of Cārvāka fragments is confronted with serious problems. It appears that Cārvāka texts were known until about the 12th century CE. More recent texts – most notably the Sarvadarśanasaṃgraha, a 14th-century text that has been the point of departure of modern Cārvāka studies – were no longer acquainted with any Cārvāka texts. They still attributed opinions to the Cārvākas, but it seems clear that some of these were based on prejudice rather than reliable information.
This Cārvākasūtra is sometimes referred to as Bārhaspatyasūtra and ascribed to the mythical seer Bṛhaspati. It claims that the four elements earth, water, fire, and wind are the ultimate building blocks of our universe. Even consciousness is stated to be derived from these. The soul is nothing but the body endowed with consciousness. Claims like these are responsible for the modern custom of referring to the school as materialistic. This custom, though justified, should not make us forget that the primary concern of the school does not appear to have been to develop a materialistic worldview, but rather to reject certain beliefs – rebirth and karmic retribution, life after death – held by others. This negative purpose led to other denials, such as, “There is no means of knowledge for determining (the existence of) the other world,” and “There is no other world because of the absence of any other-worldly being (i.e. a transmigrating self)” (Bhattacharya, 2003, 605, 612).
Tattvārthasūtra
It was pointed out above that the fact that the Jainas laid down a summary of their doctrine in a Sūtra text can be taken as an indication that they wanted to be seen as a philosophical movement on a par with those of the Brahmans. The use of Sanskrit for this work, to be contrasted with their ongoing use of Middle Indic languages, points in the same direction. Written in Sanskrit, the Tattvārthasūtra could be read by Brahmanical thinkers (and by Buddhist thinkers, who had turned to Sanskrit in the early centuries of the Common Era). What is more, by using Sanskrit, Jaina philosophers were well prepared to take part in public debates that opposed thinkers from different schools, and in which the Jainas came to participate with gusto (Bronkhorst, 2007b).The exact date of composition of the Tattvārthasūtra remains uncertain, but may have been in or near the 4th century CE (Balcerowicz, 2008, 35n23; for bibliographical information, see Wiles, 1998; for a translation, Tatia, 1994). The two oldest commentaries on the Tattvārthasūtra are the Tattvārthādhigamabhāṣya (whose authorship and date will be discussed below) and the Sarvārthasiddhi of Devanandin (perhaps second half of 5th century CE; Bronkhorst, 1985a, 161). These two commentaries belong to the two main divisions of Jainism: Śvetāmbara and Digambara, respectively. Both divisions claim that the original Tattvārthasūtra belonged to them. The Śvetāmbaras go further and maintain that the Tattvārthasūtra and the Tattvārthādhigamabhāṣya were written by one and the same person, Umāsvāti. Neither of these claims resists a detailed consideration of the evidence. The original Tattvārthasūtra may have belonged neither to the Śvetāmbaras nor to the Digambaras, but to the Yāpanīyas, a third division of Jainism that no longer survives. Further, the author of the Tattvārthādhigamabhāṣya did not also compose the Tattvārthasūtra. What he did do was incorporate the Sūtra text into his work in a way that is reminiscent of the Yogaśāstra of “Patañjali” and of the Arthaśāstra (discussed above). Umāsvāti may have composed this work in the first half of the 5th century CE.
It goes almost without saying that the Tattvārthādhigamabhāṣya and the Sarvārthasiddhi do not comment on exactly the same Sūtra text. There are some indications to the effect that the Sarvārthasiddhi sometimes comments on a version closer to the original. A full reconstruction of the original texts seems impossible (Bronkhorst, 1985a).
It seems clear that the aphoristic Sūtra style is primarily a feature of Brahmanical technical literature. Where Buddhist or Jaina works are composed in this style, Brahmanical influence is most often evident.
Comments
Post a Comment