There might be some trends that brought new problematic issues into the paralokasiddhi of Dharmak]rtian Tradition. They might had derived from the Dharmak]rtian realm itself, and came to be noticed as the theme of the discussion among the commentators. They also might be, as can be testified clearly by the texts of the opponents, influenced from the external trend of discourse with other schools. As the late professor Frauwallner suggested, the theory of Lokqyata philosophy had some stages of development in the historical discourse with Buddhists. He pointed out the developed stage of the new theory that is derived from the discourse on the origination of cittasantqna of a new born baby. The objection not only arises from the theoretical position of Lokqyata but also anticipates the Buddhist theory of cittasantqna that had been presented as the reply to the criticism from the Lokqyata side. That is as follows: if the first cognition of the new born baby originates only from cognition and not from the four material elements, the only supposed material stuff for the first cognition of the baby should be the cognition of its parent's, more precisely, the mother's cognition. This is the developed stage of the new theory mentioned by Profs. Frauwallner and Steinkellner. ⁽¹⁾

This problem was developed in another way also. That is another counter attack on the cittasantqna theory of a new born baby. However, it does not presuppose the parent's santqna as material stuff, but proclaims the possibility of origination from other kinds of material stuff (vijqt]yasantqna) not only from the same series of sentient stuff(sajqt]yasantqna). This tenet is, in fact, already stated in the old theory of Lokqyata epistemology. However, this theory differs from the old one in dealing with the origination of monomatisantqna presupposing the svajqt]yacittasantqna theory of the Dharmak]rtian Buddhists. Furthermore, it differs from the undeveloped stage of the new theory, since the main point of the new phase of discourse, concerning the theory of two models of causality stated by Dharmak]rti, centers around the causality between fire and the sequence of smoke.

The first type of discourse is carried on by !qntaraksita ⁽²⁾and etc. However, !qntarak2ita never concerns himself with the latter. That may be another type of the developed phase of the new theory. On this topic, after !qntaraksita, commentators come to be interested in, and join in the discourse.

To clarify the problem, now we should investigate some passages from the text of !r]dhara.

Dharmak]rti, in his Pramqzavqrttika Chap.II, expounds cittasantqna theory especially concerning the doctrine of universal momentariness (k2azabhaxga). !r]dhara, commenting on Pra1astapqda's proof of qtman by means of pari1e2a,⁽³⁾ criticizes the Buddhist doctrine of k2azabhaxga precisely. At the end of the criticism, he develops an argument against the proof of previous / successive lives (paralokasiddhi) in the Dharmak]rtian Tradition.⁽⁴⁾

　 Buddhist j`qnasantqna theory concerning the theory of transmigration is not proper, because, there cannot be any reason for the statement that the first cognition [of the present life] in one's mother's womb could be derived from the immediately previous cognition that had existed [at the moment of death in the previous life].

In fact, the origination of matter is not restricted only to the same kind of causal matter in the series, but also could be originated from other kinds of causal matters. For example, a [new] sequence of smoke could be originated from fire that is not in the same series [of smoke].

　The problem which I would like to point out is the latter portion of the statement. The ground presented by !r]dhara obviously supposed the theory of svacittasantqna theory of Dharmak]rti, who demonstrated the possibility of transmigration of sentient beings without presupposing any eternal personal entity(qtman). These arguments can be traced back to the Pramqzavqrttika (PV), Chap.II. However, we should notice that: the theory mentioned here is not the one of Dharmak]rti himself, but the developed one. !r]dhara's way of argumentation is deeply influenced by the developed epistemological discourse between later Buddhist philosophers - such as Praj`qkaragupta, Kamala1]la and Dharmottara - and the Lokqyata circle, in which Jaryarq1i the author of Tattvopaplavasi/ha (TUS) 　may have been included.

　 The main point of the issue for them is the causality between fire and the sequence of smoke. That means the problem of the origination of conceptional mind from different kinds of causal matter(/body). This point mainly concerns the causal theory of Dharmak]rti in general, not only ontologically but also epistemologically and even logically.

I would like to present the philosophical context of the Buddhist paralokasiddhi behind the statement of !r]dhara, tracing back the changing phases of the arguments up to Dharmak]rti.

I. Buddhist problems discussed after !qntarak2ita

　 Dharmak]rti in his PV II 40 states:⁽⁵⁾

In every moment, a newly originated moment has every previous moment [as its cause].

Therefore, at any time, it could be seen that the cause itself has a previous cause. 　

This statement is the main theme of the discourse concerning the origination of mind. Dharmak]rti here advocates the theory of svajqticittasantqna depending upon this premise. Dharmak]rti, afterward in his PV II 1 113, in discussion with an opponent who objects to the theory of svajqticittasantqna, presents a prasaxga.⁽⁶⁾

If one states that [a series of] cognition, once originated directly from a body at the first stage, will be restricted in its further origination by the cognition itself, he should explain why the body which still has potency [to generate the next moment of cognition] ceases its productive function.

The opinion of the opponent introduced here is noteworthy for its significance for the post- Dharmak]rtian Buddhist philosophers. Concerning this point, the causal relation between fire and smoke comes to be discussed. We have two texts which deal with the problem. The first is the additional comments of Kamala1]la to the Dharmak]rtis statement. The other is the more precise epistemological investigation of Dharmottara.

Kamala1]la develops an additional discussion on the body-mind relation in commenting on Tattvasa/graha (TS) XXII 1885, independent from the context of !qntarak2ita. Handling the matter, he introduces the argument on the origin of consciousness discussed in PV II. 113. The problem is presented in the following manner. ⁽⁷⁾

Buddhist assertion: Desire[, hate and] etc. become awakened from the seeds accumulated 　 previously in the same series of mind (svab]jabodhaprav3tti) .

Therefore, consciousness can be proved as originated from the consciousness of a previous moment.

Objection from Lokqyata: Even if consciousness is originated from previous consciousness, the consciousness that arises at the initial stage is originated from the body as a material cause (upqdhqnahetu).

Therefore, it cannot be proved by the reason that consciousness does not have a beginn- ing. [Thus the series of consciousness of the previous lives could not have existed. ] The objection can be criticized and avoided on the following three grounds.

The first criticism is that it does not match the second type of causal model which Dharmak]rti himself had shown by the theory of two models of causality in his Hetubindu.⁽⁸⁾

The 2nd criticism is an application of the prasaxga that is already mentioned by Dharmak]rti himself in PV II 113ff.⁽⁹⁾ The 3rd criticism also only develops the argument in PV II 113. It is carried out in assuming these three cases. ⁽¹⁰⁾ Quoting PV II 113, the arguments come to a finish.

　　 In the Paralokasiddhi (PLS) of Dharmottara, there are advocated two Lokqyata tenets: that cognition, being an attribute/function, is the same as a body itself, and that cognition, being produced, is independent from the body itself. The latter is categorized as the new theory of Lokqyata by Profs. Frauwallner/ Steinkellner.⁽¹¹⁾ The theory is stated as followings:⁽¹²⁾

Smoke etc., being originated from fire at the initial stage, can only be generated afterward from the previous moments of the same sequence of a causal series (svajqt]yasantqna) . That is the same case as the origination of cognition. The sequence of cognition, once being produced from four material elements(mahqbh[ta), at a later stage can only be generated from the previous moments of the same causal sequence of cognition itself.

This theory is precisely analyzed and counter objected to from the Buddhist epistemological ground of the Dharmak]rtian tradition, as Prof. Steinkellner's investigation shows. The criticism of Dharmottara in his PLS (Discussion 2.22. according to the synopsis by Prof. Steinkellner) may be a parallel discourse to this. In this case, most of the criticism by Dharmottara is directed at this theory.

Dharmottara shows two models of causality which represent the generation of result.⁽¹³⁾

i. The generation of the result as a that depends on a .

(ex. flame ==>smoke)

ii. The one that depends on the of immediately previous moment. ( ex. objective stuff /eye /light/ and etc. ==> visual perception)

These two models of causalities correspond to the one that Dharmak]rti had shown in his Hetubindu, as Prof.Steinkellner has commented on. ⁽¹⁴⁾

Opponents, acknowledging this Buddhist theory, and depending on the ground of the first model of causality that represents the origination of a new series [ex. smoke] from the other kind of material cause [fire] in the initial stage, counter attacks by stating the possibility of the generation of cognizing series from a particular cause, namely four material elements. However, the model becomes useless to explain the . In this case, cognition originates from material elements at the initial stage and from the sequence of cognition which subsequently follows at every later moment. Nevertheless, in fact, sense perceptions are generated as instantaneous results of the existence of (ekasqmqgr]) which precedes the series of cognition, at any moment.

After　the discourse on the epistemological problems centering on the difference between sense perception and conceptional consciousness, Dharmottara continues to present counter objections, he proves the adequateness of the Dharmak]rtian theory of causality. ⁽¹⁵⁾

　　These two post-!qntarak2ita Dharmak]rtian philosopher, equally have intention to prove the svajqt]yacittasantqna theory by interpreting two models of origination mentioned by Dharmak]rti. However, Dharmak]rti himself did not acknowledge the necessity to fix the argument to that theory. Additionally, !qntarak2ita himself also did not mention the issue. We can suppose some opponent who attacks Dharmak]rti's theory of svajqt]yasantqna theory concerning the two models of causality.

　 To clarify the situation of these arguments, we should consider the statement of Bhqvaviveka presented against a Lokqyata who might have been in the circle that had been advocating the new theory. Bhqvaviveka had once reply to an attack by the new theory of Lokqyata philosopher(/s) who admitted a pramqt3 derived from body in secondary meaning. However, he did not resolve the problem, since he neglected the difference between conceptional cognition and perception. The question concerns a pramqt3 who cognizes logical concomitance between fire and smoke. He ha only skirted the issue from the view point of the aspect of sense perception, criticizing by proving that plural different momentary conditions are originated corresponding with the variety of momentary external objects.⁽¹⁶⁾ The problem could have been solved by the causal theory of Dharmak]rti himself advocated in the Hetubindu. However, Dharmak]rti recorded only a simplified prasaxga in PV II 113 and left the problem for the philosophers of a later time. The followers of Dharmak]rti came to acknowledge the deep meaning of the matter, and tried to confirm the ontological / epistemological coordination of the theory of svajqticittasamtqna reconsidering the theory of two models of causalities already mentioned by Dharmak]rti. In the Dharmak]rtian tradition of paralokasiddhi, a new stage appeared in which the core of the arguments is the nature of consciousness (manomati) that is independent from sense perception, keeping a certain identical factor, and appears momentarily. In this way, Dharmottora confronts the necessity to demonstrate the inner stream of cognition (pramqt3) which appears momentary in every moment from the awakening of the previous consciousness.

For the post- Dharmak]rtian Buddhist philosophers, the well-known example of the bh[tacaitanyavqda theory (i.e. intoxicating function that is dependent on wine), comes to be replaced by the example of the new tenet (i.e. a series of smoke originated from a fire). This is also known to Buddhist logician as the typical example of the inference through causal concomitance(kqryqnumqza).

II. Buddhist problems discussed concerning Logical concomitance.

As the evidence to clarify the situation, in which post-Dharmak]rtian philosophers confront these difficulties pointed out by Lokqyata philosophers even in their logical aspect, we have the following texts.

Praj`qkaragupta commenting on the verses of Dharmak]rti (PV II 39f.) develops an argument with which Dharmak]rti was not concerned.⁽¹⁷⁾ The matter concerns the cognizing subject of causal concomitance between fire and smoke, that is between a cause and its result which are of a different kind of causal sequence(vijqt]yasantqna), such as a lotus flower and caw dung as its cause. In this context of argument, we confront a logical issue, that is not so clearly mentioned within the Dharmak]rtian tradition, but in the Lokqyata tradition. The issue, pointed out as the impossibility of kqryqnumqna (inference of fire from smoke), not only as an ontological but also as a logical problem against the paralokasiddhi of the Dharmak]rtian tradition.

In fact, the only preserved work of Lokqyata, Tattvopaplavasimha of Jararq1i, mainly concerns this issue when the author deals with the Dharmak]rtian theory of logical concomitance between fire ad smoke, to wit, kqryqnumqza. Now, we can also trace another problematic issue on cittasantqna concerning the demonstration of previous lives. Jayarq1i gives criticism to the Buddhist definition of the twofold means of right cognition (pramqza).

First, the cittasantqna theory, being connected with the theory of direct perception (pratyak2a ), is criticized from five points as Dr. Franco shows.⁽¹⁸⁾ Every criticism is an attack on the actual cittasantqna theory of the Dharmak]rtian Buddhists. Jasyarq1i carries out the discussion only by criticizing the theory of opponents. He himself never intended to assert any independent theory against the opponents, but just tries to point out sophistically the theoretical inappropriateness of them (paraparyanuyogapara). In the fifth discussion, Jayarq1i deals with the difficulty of the Buddhist proof of previous existence presented by the Buddhist circle. Thus, he shows that the first cognition in this present existence is not the one brought about from the last cognition of the previous existence. He points out the inappropriateness of the example (d32wqnta) or the major premise of the syllogism.⁽¹⁹⁾ As the reasons for his statement, he develops criticisms to deny the Buddhist theory of kqryqnumqza .⁽²⁰⁾ The third criticism is very serious as it concerns the Buddhist theory of kqryqnumqza.

Jayarq1i, when he deals with Buddhist kqryqnumqza theory, reexamines the causal concomitance between fire and smoke again here also. At first, he states the impossibility of the causality between fire and smoke.⁽²¹⁾ He introduces the svajqt]yasantqna theory here also.⁽²²⁾ In this logical discourse, Jayarq1i, sophistically introduces the argument of ontological issue.⁽²³⁾ Not differentiating between the logical concomitance and the ontological causality of fire and smoke, he insists on the incompatibility of the Buddhist syllogism that demonstrates the existence of previously perished cognition in the former life.

On this matter we cannot find any direct Buddhist response to Jayarq1i. However, by the later Buddhist logicians, the syllogism is mentioned as kqryahetu, without any reinforcing of the logical validity against Bqrhaspatyas.⁽²⁴⁾ Anyway, here also, the issue to be discussed is the origination of a sequence of smoke that derived from a different kind of material cause at its initial stage. Of course the logical concomitance that can be shown in saptam] and the ontological causality between a material result and its material stuff, which may be shown in pa`cam], should not be confused as Jayarq1i did sophistically. However, the problem shows deep significance in considering the causal sequence of svajqticittasamtqna.

　　Now, we have shown the background of the Buddhist demonstration of previous lives. In this historical context, the post-Dharmak]rti Buddhist philosophers acknowledged the necessity of explaining the ontological / epistemological / logical coordination of the theory of svajqticittasantqna， not only in the Buddhist / Lokqyata circles but also even for the later Nyqyavai1e2ikas.　

1.

¹ 　E.Steinkellner,　Dharmottara's Paralokasiddhi, 　WSTB 15, Wien 1986, S.9-12,37(Anm.4).　

2. !qntak2ita, Tattvasa/graha XXII 1892ab. Cf.E.Steinkellner, op.cit., S.11.

3. Pra1astapqd, Padqrthadharmasa/graha:1e2qd qtmakqryatvqt te qtmq samadhigamyate ed.D.Jhq, Pra1astabha2ya with Commentary Nyqyakandal], Gaxganqthajhq Granthamqlq, Vanarasi 1977, p.175.

4. !r]dhara, Nyqyakandal]: ed.D.Jhq, p.197f.,