It is important for the history of Sa@mkhya philosophy that Caraka's treatment of it, which so far as I know has never been dealt with in any of the modern studies of Sa@mkhya, should be brought before the notice of the students of this philosophy. According to Caraka there are six elements (_dhatus_), viz. the five elements such as akas'a, vayu etc. and cetana, called also puru@sa. From other points of view, the categories may be said to be twenty-four only, viz. the ten senses (five cognitive and five conative), manas, the five objects of senses and the eightfold prak@rti (prak@rti, mahat, aha@mkara and the five elements)[Footnote ref 324]. The manas works through the senses. It is atomic and its existence is proved by the fact that in spite of the existence of the senses there cannot be any knowledge unless manas is in touch with them. There are two movements of manas as indeterminate sensing (_uha_) and conceiving (_vicara_) before definite understanding (_buddhi_) arises. Each of the five senses is the product of the combination of five elements but the auditory sense is made with a preponderance of akasa, the sense of touch with a preponderanceof air, the visual sense with a preponderance of light, the taste with a preponderance of water and the sense of smell with a preponderance of earth. Caraka does not mention the tanmatras at all [Footnote ref 325]. The conglomeration of the sense-objects (_indriyartha_) or gross matter, the ten senses, manas, the five subtle bhutas and prak@rti, mahat and aha@mkara taking place through rajas make up what we call man. When the sattva is at its height this conglomeration ceases. All karma, the fruit of karma, cognition, pleasure, pain, ignorance, life and death belongs to this conglomeration. But there is also the puru@sa, for had it not been so there would be no birth, death, bondage, or salvation. If the atman were not regarded as cause, all illuminations of cognition would be without any reason. If a permanent self were not recognized, then for the work of one others would be responsible. This puru@sa, called also _paramatman_, is beginningless and it has no cause beyond itself. The self is in itself without consciousness. Consciousness can only come to it through its connection with the sense organs and manas. By ignorance, will, antipathy, and work, this conglomeration of puru@sa and the other elements takes place. Knowledge, feeling, or action, cannot be produced without this combination. All positive effects are due to conglomerations of causes and not by a single cause, but all destruction comes naturally and without cause. That which is eternal is never the product of anything. Caraka identifies the avyakta part of prak@rti with puru@sa as forming one category. The vikara or evolutionary products of prak@rti are called k@setra, whereas the avyakta part of prak@rti is regarded as the k@setrajna (_avyaktamasya k@setrasya k@setrajnam@r@sayo viduh_). This avyakta and cetana are one and the same entity. From this unmanifested prak@rti or cetana is derived the buddhi, and from the buddhi is derived the ego (_aha@mkara_) and from the aha@mkara the five elements and the senses are produced, and when this production is complete, we say that creation has taken place. At the time of pralaya (periodical cosmic dissolution) all the evolutes return back to prak@rti, and thus become unmanifest with it, whereas at the time of a new creation from the puru@sa the unmanifest (_avyakta_), all the manifested forms--the evolutes of buddhi, aha@mkara,etc.--appear [Footnote ref 326]. This cycle of births or rebirths or of dissolution and new creation acts through the influence of rajas and tamas, and so those who can get rid of these two will never again suffer this revolution in a cycle. The manas can only become active in association with the self, which is the real agent. This self of itself takes rebirth in all kinds of lives according to its own wish, undetermined by anyone else. It works according to its own free will and reaps the fruits of its karma. Though all the souls are pervasive, yet they can only perceive in particular bodies where they are associated with their own specific senses. All pleasures and pains are felt by the conglomeration (_ras'i_), and not by the atman presiding over it. From the enjoyment and suffering of pleasure and pain comes desire (_t@r@s@na_) consisting of wish and antipathy, and from desire again comes pleasure and pain. Mok@sa means complete cessation of pleasure and pain, arising through the association of the self with the manas, the sense, and sense-objects. If the manas is settled steadily in the self, it is the state of yoga when there is neither pleasure nor pain. When true knowledge dawns that "all are produced by causes, are transitory, rise of themselves, but are not produced by the self and are sorrow, and do not belong to me the self," the self transcends all. This is the last renunciation when all affections and knowledge become finally extinct. There remains no indication of any positive existence of the self at this time, and the self can no longer be perceived [Footnote ref 327]. It is the state of Brahman. Those who know Brahman call this state the Brahman, which is eternal and absolutely devoid of any characteristic. This state is spoken of by the Sa@mkhyas as their goal, and also that of the Yogins. When rajas and tamas are rooted out and the karma of the past whose fruits have to be enjoyed are exhausted, and there is no new karma and new birth,the state of mok@sa comes about. Various kinds of moral endeavours in the shape of association with good people, abandoning of desires, determined attempts at discovering the truth with fixed attention, are spoken of as indispensable means. Truth (tattva) thus discovered should be recalled again and again [Footnote ref 328] and this will ultimately effect the disunion of the body with the self. As the self is avyakta (unmanifested) and has no specific nature or character, this state can only be described as absolute cessation (_mok@se niv@rttirni@hs'e@sa_).
The main features of the Sa@mkhya doctrine as given by Caraka are thus: 1. Puru@sa is the state of avyakta. 2. By a conglomera of this avyakta with its later products a conglomeration is formed which generates the so-called living being. 3. The tanmatras are not mentioned. 4. Rajas and tamas represent the bad states of the mind and sattva the good ones. 5. The ultimate state of emancipation is either absolute annihilation or characterless absolute existence and it is spoken of as the Brahman state; there is no consciousness in this state, for consciousness is due to the conglomeration of the self with its evolutes, buddhi, aha@mkara etc. 6. The senses are formed of matter (_bhautika_).
This account of Sa@mkhya agrees with the system of Sa@mkhya propounded by Pancas'ikha (who is said to be the direct pupil of Asuri the pupil of Kapila, the founder of the system) in the Mahabharata XII. 219. Pancas'ikha of course does not describe the system as elaborately as Caraka does. But even from what little he says it may be supposed that the system of Sa@mkhya he sketches is the same as that of Caraka [Footnote ref 329]. Pancas'ikha speaks of the ultimate truth as being avyakta (a term applied in all Sa@mkhya literature to prak@rti) in the state of puru@sa (_purusavasthamavyaktam_). If man is the product of a mere combination of the different elements, then one may assume that all ceases with death. Caraka in answer to such an objection introduces a discussion, in which he tries to establish the existence of a self as the postulate of all our duties and sense of moral responsibility. The same discussion occurs in Pancas'ikha also, and the proofsfor the existence of the self are also the same. Like Caraka again Pancas'ikha also says that all consciousness is due to the conditions of the conglomeration of our physical body mind,--and the element of "cetas." They are mutually independent, and by such independence carry on the process of life and work. None of the phenomena produced by such a conglomeration are self. All our suffering comes in because we think these to be the self. Mok@sa is realized when we can practise absolute renunciation of these phenomena. The gu@nas described by Pancas'ikha are the different kinds of good and bad qualities of the mind as Caraka has it. The state of the conglomeration is spoken of as the k@setra, as Caraka says, and there is no annihilation or eternality; and the last state is described as being like that when all rivers lose themselves in the ocean and it is called ali@nga (without any characteristic)--a term reserved for prak@rti in later Sa@mkhya. This state is attainable by the doctrine of ultimate renunciation which is also called the doctrine of complete destruction (_samyagbadha_).
Gu@naratna (fourteenth century A.D.), a commentator of _@Sa@ddars'anasamuccaya_, mentions two schools of Sa@mkhya, the Maulikya (original) and the Uttara or (later) [Footnote ref 330]. Of these the doctrine of the Maulikya Sa@mkhya is said to be that which believed that there was a separate pradhana for each atman (_maulikyasa@mkhya hyatmanamatmanam prati p@rthak pradhanam vadanti_). This seems to be a reference to the Sa@mkhya doctrine I have just sketched. I am therefore disposed to think that this represents the earliest systematic doctrine of Sa@mkhya.
In _Mahabharata_ XII. 318 three schools of Sa@mkhya are mentioned, viz. those who admitted twenty-four categories (the school I have sketched above), those who admitted twenty-five (the well-known orthodox Sa@mkhya system) and those who admitted twenty-six categories. This last school admitted a supreme being in addition to puru@sa and this was the twenty-sixth principle. This agrees with the orthodox Yoga system and the form of Sa@mkhya advocated in the _Mahabharata_. The schools of Sa@mkhya of twenty-four and twenty-five categories are here denounced as unsatisfactory. Doctrines similar to the school of Sa@mkhya we have sketched above are referred to in some of theother chapters of the _Mahabharata_ (XII. 203, 204). The self apart from the body is described as the moon of the new moon day; it is said that as Rahu (the shadow on the sun during an eclipse) cannot be seen apart from the sun, so the self cannot be seen apart from the body. The selfs (_s'ariri@na@h_) are spoken of as manifesting from prak@rti.
We do not know anything about Asuri the direct disciple of Kapila [Footnote ref 331]. But it seems probable that the system of Sa@mkhya we have sketched here which appears in fundamentally the same form in the _Mahabharata_ and has been attributed there to Pancas'ikha is probably the earliest form of Sa@mkhya available to us in a systematic form. Not only does Gu@naratna's reference to the school of Maulikya Sa@mkhya justify it, but the fact that Caraka (78 A.U.) does not refer to the Sa@mkhya as described by Is'varak@r@s@na and referred to in other parts of _Mahabharata_ is a definite proof that Is'varak@r@s@na's Sa@mkhya is a later modification, which was either non-existent in Caraka's time or was not regarded as an authoritative old Sa@mkhya view.
Wassilief says quoting Tibetan sources that Vindhyavasin altered the Sa@mkhya according to his own views [Footnote ref 332]. Takakusu thinks that Vindhyavasin was a title of Is'varak@r@s@na [Footnote ref 333] and Garbe holds that the date of Is'varak@r@s@na was about 100 A.D. It seems to be a very plausible view that Is'varak@r@s@na was indebted for his karikas to another work, which was probably written in a style different from what he employs. The seventh verse of his _Karika_ seems to be in purport the same as a passage which is found quoted in the_Mahabhasya_ of Patanjali the grammarian (147 B.C.) [Footnote ref 334]. The subject of the two passages are the enumeration of reasons which frustrate visual perception. This however is not a doctrine concerned with the strictly technical part of Sa@mkhya, and it is just possible that the book from which Patanjali quoted the passage, and which was probably paraphrased in the Arya metre by Is'varak@r@s@na was not a Sa@mkhya book at all. But though the subject of the verse is not one of the strictly technical parts of Sa@mkhya, yet since such an enumeration is not seen in any other system of Indian philosophy, and as it has some special bearing as a safeguard against certain objections against the Sa@mkhya doctrine of prak@rti, the natural and plausible supposition is that it was the verse of a Sa@mkhya book which was paraphrased by Is'varak@r@s@na.
The earliest descriptions of a Sa@mkhya which agrees with Is'varak@r@s@na's Sa@mkhya (but with an addition of Is'vara) are to be found in Patanjali's _Yoga sutras_ and in the _Mahabharata;_ but we are pretty certain that the Sa@mkhya of Caraka we have sketched here was known to Patanjali, for in _Yoga sutra_ I. 19 a reference is made to a view of Sa@mkhya similar to this.
From the point of view of history of philosophy the Sa@mkhya of Caraka and Pancas'ikha is very important; for it shows a transitional stage of thought between the Upani@sad ideas and the orthodox Sa@mkhya doctrine as represented by Is'varak@r@s@na. On the one hand its doctrine that the senses are material, and that effects are produced only as a result of collocations, and that the puru@sa is unconscious, brings it in close relation with Nyaya, and on the other its connections with Buddhism seem to be nearer than the orthodox Sa@mkhya.
We hear of a _Sa@s@titantras'astra_ as being one of the oldest Sa@mkhya works. This is described in the _Ahirbudhnya Sa@mhita_ as containing two books of thirty-two and twenty-eight chapters [Footnote ref 335]. A quotation from _Rajavarttika_ (a work about which there is no definite information) in Vacaspati Mis'ra's commentary on the Sa@mkhya karika_(72) says that it was called the _@Sa@s@titantra because it dealt with the existence of prak@rti, its oneness, its difference from puru@sas, its purposefulness for puru@sas, the multiplicity of puru@sas, connection and separation from puru@sas, the evolution ofthe categories, the inactivity of the puru@sas and the five _viparyyayas_, nine tu@s@tis, the defects of organs of twenty-eight kinds, and the eight siddhis [Footnote ref 336].
But the content of the _Sa@s@titantra_ as given in _Ahirbudhnya Sa@mhita_ is different from it, and it appears from it that the Sa@mkhya of the _Sa@s@titantra_ referred to in the _Ahirbudhnya Sa@mhita_ was of a theistic character resembling the doctrine of the Pancaratra Vai@snavas and the _Ahirbudhnya Sa@mhita_ says that Kapila's theory of Sa@mkhya was a Vai@s@nava one. Vijnana Bhiksu, the greatest expounder of Sa@mkhya, says in many places of his work _Vijnanam@rta Bha@sya_ that Sa@mkhya was originally theistic, and that the atheistic Sa@mkhya is only a _prau@dhivada_ (an exaggerated attempt to show that no supposition of Is'vara is necessary to explain the world process) though the _Mahabharata_ points out that the difference between Sa@mkhya and Yoga is this, that the former is atheistic, while the latter is theistic. The discrepancy between the two accounts of _@Sa@s@titantra_ suggests that the original _Sa@s@titantra_ as referred to in the _Ahirbudhnya Sa@mhita_ was subsequently revised and considerably changed. This supposition is corroborated by the fact that Gu@naratna does not mention among the important Sa@mkhya works _@Sa@s@titantra_ but _@Sa@s@titantroddhara_(revised edition of _@Sa@s@titantra_) [Footnote ref 337]. Probably the earlier @Sa@s@titantra was lost even before Vacaspati's time.
If we believe the @Sa@s@titantra referred to in the _Ahirbudhnya Sa@mhita_ to be in all essential parts the same work which was composed by Kapila and based faithfully on his teachings, then it has to be assumed that Kapila's Sa@mkhya was theistic [Footnote ref 338]. It seems probable that his disciple Asuri tried to popularise it. But it seems that a great change occurred when Pancas'ikha the disciple of Asuri came to deal with it. For we know that his doctrine differed from the traditional one in many important respects. It is said in _Sa@mkhya karika_ (70) that the literature was divided by him into many parts (_tena bahudhak@rtam tantram_). The exact meaning of this reference is difficult to guess. It might mean that the original _@Sa@s@titantra_ was rewritten by him in various treatises. It is a well-known fact that most of the schools of Vai@s@navas accepted the form of cosmology which is the same in most essential parts as the Sa@mkhya cosmology. This justifies the assumption that Kapila's doctrine was probably theistic. But there are a few other points of difference between the Kapila and the Patanjala Sa@mkhya (Yoga). The only supposition that may be ventured is that Pancas'ikha probably modified Kapila's work in an atheistic way and passed it as Kapila's work. If this supposition is held reasonable, then we have three strata of Sa@mkhya, first a theistic one, the details of which are lost, but which is kept in a modified form by the Patanjala school of Sa@mkhya, second an atheistic one as represented by Pancas'ikha, and a third atheistic modification as the orthodox Sa@mkhya system. An important change in the Sa@mkhya doctrine seems to have been introduced by Vijnana Bhik@su (sixteenth century A.D.) by his treatment of gu@nas as types of reals. I have myself accepted this interpretation of Sa@mkhya as the most rational and philosophical one, and have therefore followed it in giving a connected system of the accepted Kapila and the Patanjala school of Sa@mkhya. But it must be pointed out that originally the notion of gu@nas was applied to different types of good and bad mental states, and then they were supposed in some mysterious way by mutual increase and decrease to form the objective world on the one hand and thetotality of human psychosis on the other. A systematic explanation of the gunas was attempted in two different lines by Vijnana Bhik@su and the Vai@s@nava writer Ve@nka@ta [Footnote ref 339]. As the Yoga philosophy compiled by Patanjali and commented on by Vyasa, Vacaspati and Vijn@ana Bhik@su, agree with the Sa@mkhya doctrine as explained by Vacaspati and Vijnana Bhik@su in most points I have preferred to call them the Kapila and the Patanjala schools of Sa@mkhya and have treated them together--a principle which was followed by Haribhadra in his _@Sa@ddars'anasamuaccaya_.
The other important Sa@mkhya teachers mentioned by Gaudapada are Sanaka, Sananda, Sanatana and Vo@dhu. Nothing is known about their historicity or doctrines.