HISTORY OF SÄMKHYA

The term "sämkhya" means "relating to number, enumeration, or calculation." As an adjective, the term refers to any enumerated set or grouping and can presumably be used in any inquiry in which enumeration or calculation is a prominent feature (for example, mathematics, grammar, prosody, psychology, medicine, and so forth). As a masculine noun, the term refers to someone who calculates, enumerates, or discriminates properly or correctly. As a neuter noun, the term comes to refer to a specific system of dualist philosophizing that proceeds by a method of enumerating the contents of experience and the world for the purpose of attaining radical liberation (moksa, kaivalya) from frustration and rebirth. These three dimensions of meaning in the word "samhhya" are not simply synchronie distinctions but indicate as well the diachronic or historical development of the word in the ancient period. That is to say, in the ancient history of South Asian culture there appear to be three identifiable phases «of development of the term "sämkhya" that roughly correspond to these three basic meaning dimensions.1 These can be briefly characterized as follows: ( 1 ) Intellectual inquiry in the oldest learned traditions of ancient India (from the Vedic period, ca. 1500 before the Common Era [B.G.E. }, through the Mauryan period in the fourth and third centuries B.G.E. ) was frequently cast in the format of elaborate enumerations of the contents of a particular subject matter — for example, the principles of statecraft as preserved in Kautilya's Arthasästra, the principles of medicine as preserved in the Caraka,samhitä and Susndasamhitä, and so forth. The Vedic corpus itself exhibits this tendency as do traditions of law {nitisästra ) and politics (räjadharma), and it is in such environments that one finds some of the early references to sämkhya. Kautilya, for example, refers to sämkhya as one of three traditions of änviksiki? The notion of änviksiki in these ancient contexts means something like the enumeration of the contents of a particular subject matter by means of systematic reasoning.3 The practice of änviksiki is not really "philosophy" in our usual senses of the term; it is, rather, a kind of general "scientific" inquiry by means of the systematic enumeration of basic principles.4 Such enumerations appeared in a variety of intellectual subject areas, including phonology, grammar, statecraft, medicine, law, cosmology, and iconography, and the compilations of these subject-area enumerations sometimes came to be called "tantras" (meaning a scientific work, and synonymous with such terms as "sästra" "vidyä", and so forth). Moreover, certain stylistic rules or "methodological devices" (yuktis) came to be accepted in composing scientific works — for example, a brief statement of a position (uddesa), a lengthy exposition of a position (nirdesa), an etymological explanation (nirvacana), the proper order or sequence in enumerating a subject (vidhäna), and so forth:5 Kautilya's Ärthasästra provides a list of such methodological devices, and the author illustrates how his work uses the various methodological devices, thereby establishing that his treatise is a scientific work. The medical texts [Caraka and Susruta) are also scientific works in this sense and likewise provide lists of methodological devices. This may well explain why the later technical Sämkhya philosophy is frequently referred to as a tantra, and it helps in understanding the reasons why the long introduction to the Tuktidipikä (the most important commentary on the Sämkhyakärikä), contains a detailed discussion of the methodological devices essential for any systematic inquiry. In this oldest period, however, it is undoubtedly an anachronism to interpret references to sämkhya, änviksiki, or tantra as themselves completed or distinct systems of thought, as some older scholars have suggested (Garbe, for example).6 It is more plausible to interpret these references in a much more general sense as the first and groping attempts at systematic thinking, which proceeded by determining and enumerating the components of anything (whether it be the components of the human body, the components of the sacrificial ritual, the components of the heavens, or the components of grammar). (2) A second phase in the development of the term "sämkhya" begins from the period of the oldest, pre-Buddhistic Upanisads, ca. eighth or seventh centuries B.C.E., and can be traced through traditions of the early ascetic spirituality in South Asia, namely, the various monastic [sramana and yati) groups, the early Jain and Buddhist movements, and so forth, reaching a culmination in the sorts of speculative thinking one finds in the Moksadharma portion of the Mahäbhärata, in the Bhagavadgitä, and in the cosmological descriptions of the oldest Purânas (or, in other words, reaching into the first centuries of the Common Era). If in the oldest period the term "sämkhya" could refer generally to any enumerated set of principles (in an environment of änviksiki for the sake of constructing a scientific work), in this second period the notion becomes linked to a methodology of reasoning that results in spiritual knowledge (vidyä, jfiäna, viveka) that leads to liberation from the cycle of frustration and rebirth. It is possible, of course, perhaps even likely, that in the oldest period the term "sämkhya" in its general sense of intellectual enumeration was applied on occasion in contexts of meditation and religious cosmology — the enumerations in Rg Veda 1.164, X.90, or X.I29, or the enumerations of the parts of the body or the breaths in the Atharva Veda or in the Brâhmana literature would suggest as much — but there is little doubt that it is primarily in this second period that "sämkhya" becomes a prominent notion in those environments in which meditation, spirititual exercises, and religious cosmology represent the crucial subject matters. The archaic ontology of Chändogya Upanisad VI.2-5, for example, with its emphasis on primordial Being [sat) in its tripartite manifestations as fire (red), water (white), and food (black), correlated with speech, breath, and mind, probably foreshadows the later Sâmkhya ontological notions of prakrti, the three gunas, and the préexistence of the effect. On one level, of course, this kind of reflection echoes older Vedic notions (for example, some of the number sequences and symbolism of RV.X.164), but, on another level, it represents a transition to later formulations such as those in Svetätvatara Upanisad— for example, "The One unborn, red, white, and black... ." (Suet. Up. IV.5), and "Two birds, companions (who are) always united, cling to the selfsame tree..." (Svet. Up. IV.6-7) — a text in which the older Vedic symbolism is clearly present and yet a text in which the terms "sâmkhya" and "yoga" are actually used. Cosmological speculations such as these are combined with elaborate descriptions of yogic experience in such texts as Kafha Upanisad, Moksadharma, Bhagavadgitä, and Buddhacarita. The same sorts of speculation are used in the medical literature [Carakasamhitä and Susrutasamhitä), and the hierarchical ordering , of basic principles {tattva) is given a cosmological turn with respect to the periodic creation and dissolution of the manifest world in Manusmrti and in most of the oldest Puränas. Certain characteristic notions become associated with Sämkhya, but throughout the period Sämkhya is primarily a methodology for attaining liberation and appears to allow for a great variety of philosophical formulations. Edgerton has expressed the matter well: "Any formula of metaphysical truth, provided that knowledge thereof was conceived to tend towards salvation, might be called Sâmkhya.7 ... It appears, then, that Sâmkhya means in the Upanisads and the Epic simply the way of salvation by knowledge, and does not imply any system of metaphysical truth whatever."8 On one level, Sâmkhya as a methodology for attaining salvation by knowing carries further many of the older cosmological notions of the oldest Upanisads as set forth in Chändogya Upanisad VI, and so forth. On another level, Sämkhya as a methodology for attaining salvation by knowing carries further the various psychological analyses of experience that first appear in the oldest Upanisads and then become dominant motifs in Jain and Buddhist meditation contexts and in such later Upanisads as Katha and Svetäsvatara. The enumeration of basic principles in a hierarchical order is a fundamental aspect of the methodology, but the precise number of enumerated items varies widely. In some passages seventeen basic principles are enumerated;9 in other passages twenty;10 or twenty-four;11 or the later, standard listing of twenty-five12 are enumerated. On occasion the highest principle is the old Upanisadic brahman or ätman, or, again, the highest principle is God (isvara). In some contexts the Sämkhya methodology implies a monistic perspective, in others a theistic or dualist perspective. Throughout the period, however, a characteristic terminology and a recurrent set of intellectual issues begin to develop around the methodology: reflections about a primordial materiality (pradhäna); enumerations of psychic states or conditions (bhävas, gunas) that can be construed psychologically and/or cosmologically ; analyses of the various aspects of intellectual experience in terms of intellect/will (hereafter translated simply as "intellect") (buddhi), egoity (ahamkära), and mind Çmanas); speculations about the nature of the inner self (purusa) in terms of a cosmic Self (ätman) or the self in the body or in the manifest world (jiva, bh ütätman ) ; elaborations of the five sense capacities (indriya) correlated with the five gross elements (bhüta), the five action capacities (karmendriya), and the ûve contents or "objects" (visaya) of the senses; and a general polarity between subjectivity and objectivity in terms of "the knower of the field" (ksetrajna) and "the field" (ksetra). Clearly there is a system (or systems) in the process of developing, but the focus in this second period is rather on the process or methodology itself and not on the contents that result from the process. v In contrast to methods of spiritual discipline {yoga ) that emphasize posture, breathing, recitation, and ascetic practices (tapas), sämkhya is the intellectual or reasoning method. The follower of sämkhya is one who reasons or discriminates properly, one whose spiritual discipline is meditative reasoning. This is probably the sense of the term "sämkhya" in the compound sämkhya-yoga-adhigamya ("to be understood by proper reasoning and spiritual discipline") in Svetäsvatara Upanisad VI. 13. It is probably also the sense meant in the twelfth chapter of Asvaghosa's Buddhacarita, in which reference is made to older spiritual methodologies studied by Gotama the Buddha prior to the discovery of his own unique method of meditation. Regarding the specific contents of this reasoning methodology, J. A.B. van Buitenen has offered the following comment

There must have existed scores of more or less isolated little centres where parallel doctrines were being evolved out of a common source. Occasional meetings at pilgrimages and festivals, reports from other and remote äsramas brought by wandering ascetics, polemic encounters with other preachers must have resulted in a laborious process of partial renovation and conservation, more precise definitions of doctrines and eclecticisms, readjustments of terminology, etc. At this stage to credit these little centres with the name "schools" is to do them too much or too little honor. . . . Most of the process must elude us necessarily, but we stand a better chance of recovering the little that is left by allowing for the greatest diversity, rather than the greatest uniformity of doctrine.13 In the Moksadharma portion of the Mahäbhärata various names of ancient teachers are associated with these developing traditions, including Kapila, Äsuri, Bhrgu, Yâjnavalkya, Sanatkumâra, Vasistha, âuka, Asita Dévala (or Asita and Dévala), Vyàsa, Janaka, and Pancasikha. Some of these names can be traced back to the older Upanisads, and many of them also appear in the later Purânic literature. Three of them are frequently referred to in the later technical philosophical literature as important precursors of Sâmkhya philosophy, namely, Kapila, Äsuri, and Pancasikha. The Sämkhyakärikä and its commentaries refer to Kapila and Äsuri as the founders of the philosophical system and to Pancasikha as a teacher who greatly expanded or revised the original teachings. Unfortunately, all three teachers are lost to antiquity. References to Kapila and Äsuri are brief and largely eulogistic, and the situation is not much better with Pancasikha. Fragments here and there are attributed to a certain "Pancasikha/' and Pancasikha on occasion is referred to as the author of a massive treatise in verse on Sâmkhya philosophy called Sastitantra. The views attributed to Pancasikha in the Moksadharma^ however, appear to be clearly different from the views that can be pieced together from the fragments, suggesting that there was more than one Pancasikha or that the name Pancasikha was a revered name in the tradition to which a variety of views were ascribed.14 Moreover, the claim that Pancasikha is the author of the Sastitantra is contradicted by other references that attribute authorship of Sastitantra to Kapila or to a certain Varsaganya. It is reasonable to suppose, however, that Pancasikha was a revered teacher of sâmkhya in the sense that has been indicated in this second period, that is, sâmkhya not yet as a fixed philosophical system, but as a general methodology of salvation by knowing or reasoning. It is also reasonable to suppose that practitioners of sâmkhya in this sense represent various kinds of ancient lines of teachers (guruparamparä) that traced their lineages to archaic figures such as Kapila and Äsuri

What is missing in all of these environments, however, is a critical appreciation for the need to argue for or establish an intellectual basis for these speculative intuitions. Reasoning, to be sure, is being used, but it is a reasoning not yet distinguished from the immediacy of personal experience and the accumulated heritage of ritual performance and priestly wisdom. There is, of course, some groping for independence and a growing recognition that thinking itself may be a unique human activity that can exert its own identity against the established and received ordering of things. The very fact that much Upanisadic speculation appears to have been developed in princely {räjanya) or warrior (ksatriya) circles (as opposed to priestly groups) and that the early independent ascetic movements (Jains, Buddhists, and so forth) were especially successful among the newly emerging commercial classes in towns where commerce and a monied economy were developing, certainly suggest that thoughtful persons were in need of new and independent ways of thinking and behaving. Moreover, that, the political consolidation achieved under the Mauryans appears to have been legitimized by a notion of dharma and a theory of the state that owed more tö Jain and Buddhist paradigms than to older Vedic models is also symptomatic of changes that were occurring in other areas of intellectual life. Similarly, the rise of devotional and theistic movements (the Krsna cult, and so forth) in the last centuries before the beginning of the Common Era is an additional symptom of a broadly based cultural need to develop new and different patterns of intellectual formulation. Many of these tensions and changes come together intellectually in the Bhagavadgitä, and it is surely no accident that the so-called "philosophy" of the Gitä is little more than a potpourri of Upanisadic speculation, cosmological and psychological sänikhya reasoning, Jain and Buddhist ascetic motifs, varnäsramadharma as karmayoga, tied together with an apologia for early Vaisnava bhaktiyoga — a potpourri that confuses a modern reader almost as much as it confused Arjuna. In older German scholarship there was an interesting debate as to whether the kind of "philosophy" one finds in the epics (including the Gitä) and the Puränas is pure syncretism {Mischphilosophie, as in Garbe) or transitional philosophy {Übergangsphilosophie, as in Oldenberg).15 The resolution of the debate is surely the correctness of both, or possibly neither, for the crucial point is that there is no evidence of serious independent philosophizing of any kind in these texts. Whether one wishes to call these traditions syncretistic religion (or what we usually mean when we use the terms "Hinduism" and "Buddhism") or prephilosophical speculation on the way to becoming philosophy (or what we usually mean when we use the expressions "the philosopy"

of the Vedas and Upanisads" in regard to the Vedic corpus or "early Buddhist philosophy" in regard to the Buddhist canonical texts of the Tripitaka) makes little difference. They all have in common a predi-, lection for speculative intuition in an environment of received authority. Returning, however, to Sämkhya, the point to be stressed is that in this ancient period there is only a Proto-Sämkhya. There was, of course, an incipient philosophical Sämkhya gradually distilling itself out of this diffuse and varied intellectual heritage, but the evidence suggests that it was not at first taken very seriously. Whenever it is referred to (in the Moksadharma or the Gitä, for example), it is simply discounted and characterized as not really being different from Yoga.16 Taken overall, then, it is heuristically permissible to refer to this second period of development of Sâmkhya as Kapila-Pancasikha-Sämkhya, or to carry through the association of the term "sämkhya" with the term "tantra" from the oldest period, to refer to this second period as KapilaPaficasikha-Tantra, or simply as Kapila-Tantra. (3) The third phase in the development of the term "sämkhya" marks the beginning of the technical philosophical tradition and coincides with the end of the second period, namely, from about the last century B.G.E. through the first several centuries G.E. Until recently this third phase was as shrouded in obscurity as the second phase, and Edgerton, for example, in 1924 claimed that Sâmkhya as a technical philosophical system was not really in existence prior to Isvarakrsna's Sämkhyakärikä.17 Since then, however, three sources have become available that clearly indicate that Sämkhya as a technical system existed prior to Isvarakrsna, and that Isvarakrsna's own formulation comes at the end of the normative period of formulation rather than at the beginning. These three sources are (A) the publication of a previously unknown commentary on the Sämkhyakärikä called Tuktidipikä (edited by P. Chakravarti in 1938, and edited a second time by R. C. Pandeya in 1967) ;18 (B) the reconstruction of a pre-Kärikä interpretation of Sämkhya epistemology based on quotations from older Sämkhya texts cited in Dignâga, Jinendrabuddhi, Mallavädin, and Simhasüri by E. Frauwallner;19 and (C) the reconstruction of a Sämkhya "emanation text" or a "short instructional tract" from the earliest Puränas and the Moksadharma, which Puränic editors then brought into conformity with the normative view of an established Sâmkhya philosophical system, by P. Hacker.20 (A) From the Tuktidipikä it becomes clear that there was a tradition of philosophical Sämkhya in the early centuries of* the Common Era that was more than a methodology of liberation by knowing (that is to say, more than the rather diffuse Sämkhya-Yoga traditions characteristic of the second period described above), and, specifically, that this tradition (1 ) attempted to establish certain instruments of knowledge (pramänas) and to offer careful definitions of these instruments

(2) developed a special interest in inference (anumäna) and constructed a sequence for making inferences made up of ten members (avayavas); (3) attempted, after much debate, to fix the number of basic principles, together with the precise order of their enumeration, including the technical term "subtle element" (tanmätra); (4) fully developed the related notions of prakrti, the three gunas, the transformation of the gunas (gunaparinäma), and the effect's préexistence in the cause (satkärya); (5) finally accepted after much controversy one primordial prakrti but a plurality oîpurusas; (6) maintained a rich fabric of internal debate involving such teachers as Paurika, Pancädhikarana, Patanjali, Värsaganya, and various schools such as the "followers of Värsaganya," including Vindhyavâsin and Isvarakrsna,21 and (7) maintained as well a vigorous polemic of external debate with certain Buddhist philosophers and with the followers of early Vaisesika. (8 ) It also identified itself with a tradition known as sastitantra, which apparently referred to a scheme of sixty topics made up often principal topics (mülikärtha) and fifty subsidiary categories (padärtha) and which also apparently referred to a text (or possibly texts, that is to say, more than one version) by the same name (Sastitantra) \ and (9) it received its final normative formulation in Isvarakrsna's Sämkhyakärikä, which, though a brief text, nevertheless encompassed all of the important issues of the system in a concise and cogent fashion. (B) From Frauwallner's reconstruction it becomes clear that PreKärikä philosophical Sämkhya operated with a definition of perception ("the functioning of the ear, etc.", eroträdi-vrttih) and a definition of inference ("because of the perception of one aspect of an established relation, one is able to infer the other aspect of a relation," sambandhäd ekasmät pratyaksät sesasiddhir anumänam, based on a scheme of seven established relations, or saptasambandha) that Isvarakrsna clearly built upon and improved. Frauwallner speculates that this older Sâmkhya epistemology derives from a revised version of Sastitantra composed by Värsaganya at the beginning of the fourth century of the Common Era. Such may or may not be the case, but the reconstructed passages do point to a pre-kärikä philosophical Sâmkhya epistemology.22 (C) Finally, from Hacker's reconstruction it becomes clear that there was an older Sämkhya ontology-cosmology that, again, formed the bases for Isvarakrsna's normative conceptualization in the Samkhyakärikä.23 Apparently, this philosophical tradition of Sämkhya developed some time between the sorts of speculation one finds in the Moksadharma and the Bhagavadgitä, on the one hand, and the sort of normative conceptualization one finds in the Sämkhyakärikä, on the other. Moreover, it appears to coincide with the development of comparable conceptualizations within traditions of early Buddhist thought and early Vaisesika. It is tempting to suggest with Frauwallner that this Sämkhya philosophical tradition is the oldest of the technical schools of Indian philosophy (Hindu, Buddhist, or Jain) and that Buddhist ontology, Vaisesika atomism, and Nyâya epistemology may all have arisen out of an earlier Sämkhya philosophical environment, but this is perhaps to claim too much. To be sure, all of the later technical systems undoubtedly derive from the sorts of fluid speculation one finds in the "middle''-verse Upanisads (Katha, and so forth), the Moksadharma, and the Bhagavadgitä, in which Sämkhya is primarily a methodology for liberation by knowing. When the term "sämkhya" becomes linked with a technical philosophical system, however, one has the impression that there has been a definite turn away from the older diffuse speculations and that philosophical Sämkhya has become a parallel or sibling intellectual movement alongside Vaisesika and the early Buddhist schools, rather than a parental tradition to these schools. Unfortunately, although the Tuktidipikä refers to a number, of older Sämkhya philosophical teachers, it is difficult to ascertain even rough approximations of their dates. Paurika, who evidently accepted a plurality of prakrtis along with a plurality of purusas, was probably an older teacher whose views were finally rejected during the final stages of normative consolidation. Similarly, Pancädhikarana, who accepted only ten organs instead of the normative thirteen, was also probably an older teacher. Moreover, Pancädhikarana appears to have had a somewhat eccentric view concerning the subtle body, which later teachers rejected. Also, Patanjali (not to be confused with the compiler of the Togasütra and/or the grammarian ) is apparently an older figure, for his views that there was a new subtle body for each rebirth and that egoity has no separate existence as a basic principle apart from the intellect were discounted in the final formulation of the Sämkhya system. Värsaganya, however, and the followers of Värsaganya, including Vindhyaväsin, appear to have been closer to the time of Isvarakrsna. Indeed, it could well be the case that Isvarakrsna was himself in the lineage of Värsaganya. Frauwallner has suggested, basing his opinion primarily on citations of Varsaganya's views in the works of Vacaspati Misra, that Värsaganya was the author of a revised version of the $astitantra, older versions of which had been attributed to Kapila or Pancasikha. Vindhyaväsin is said to have been a pupil of Värsaganya, to have revised the developing system further, and, according to Paramärtha's "Life of Vasubandhu," to have defeated Vasubandhu's teacher (Buddhamitra, according to Paramärtha, or Manoratha, according to Hsüan Tsang's pupil, Kuei-chi) in a debate during the reign of Candragupta II (ca. fourth century).24 Vasubandhu, according to Chinese sources, then composed a rejoinder to Vindhyaväsin. Also, Hsüan-tsang (seventh century) refers to a later debate between Gunamati and a certain Sämkhya teacher, Mädhava, by name.25 It is interesting to observe, however, that the views of Vindhyavâsin (as set forth in the Tuktidipikä) and Mâdhava (as set forth in Dignäga) diverge considerably from the views of Isvarakrsna. Vindhyavâsin clearly preceded Isvarakrsna, for the author of the Tuktidipikä indicates that Isvarakrsna refrained from discussing the tenfold inference, since it had already been discussed by Vindhyaväsin. Moreover, the author of the Tuktidipikä claims that Vindhyaväsin rejected the notion of a subtle body (because the sense capacities are ubiquitous and do not, therefore, require a subtle vehicle for transmigration ) ; and that he accepted neither the contention that the subtle elements emerge out of egoity (since they emerge, rather, along with egoity from the intellect) nor the notion of a thirteenfold instrument (trayodasakarana) (since he argued instead that experience occurs in the mind, thus reducing intellect, egoity, and mind to one organ of internal experience, which, along with the ten sense capacities make a total of eleven organs instead of thirteen ). These variant views of Vindhyaväsin are suspiciously similar to the views of Vyâsa in his Togasütrabhäsya, a similarity that has inclined both Chakravarti and Frauwallner to suggest that the Värsaganya-Vindhyaväsin line of Sämkhya is preserved in the Pätanjala-Sämkhya of classical Yoga philosophy.26 Mâdhava, on the other hand, appears to have been later than Isvarakrsna, for the reported debate with Gunamati occurred around the time of Dignäga (ca. 480-540 ) a period in which the normative view of Sämkhya was already established. Moreoever, Dignäga refers to Mâdhava as a Sämkhya heretic or "destroyer of Sämkhya" (sämkhyauainäsika, sämkhya-näsaka) because he interprets the notion of prakrti and the three gunas as a plurality of primordial materialities (thus taking prakrti in the direction of Vaisesika atomism). Then, too, Mâdhava appears to have believed that action (karman) resides in this plurality of kinds of stuff and that the cycle of rebirth (samsara) is beginningless (thereby implicitly denying the Sämkhya notion of emanation). In all of this, it is quite clear that Sämkhya was a vigorous and polemical philosophical system, and one is tempted to believe the old Chinese claim that there were as many as eighteen schools of philosophical Sâmkhya (though the parallel with the eighteen Buddhist schools is probably no accident). This must have been intellectually a remarkable stage in the development of Sämkhya, and of Indian philosophy generally, for it was evidently in this creative and formative period in the first several centuries of the Common Era that the main issues of Indian philosophy were first formulated and polemically discussed: the number and definition of the instruments of knowledge, theories of ontology and causation, the role and function of knowing and ignorance, the theory of error, the problem of selfhood, the problem of

action and rebirth, and the problem of freedom and bondage. All of these issues had been discussed earlier, but the crucial task in this first philosophical period was that of systematic formulation, overall intellectual coherence, and persuasive presentation