Preface

Three names are credited with the production of this book. The idea that developed into this project was first mooted by Alex WATSON. In his doctoral thesis he had re-edited much of the Buddhist section of the Paramoksanitasakarikavrtti on the basis of two Devanagari manuscripts and parallel passages in others of Ramakantha's texts, and in 2002 he suggested to Dominic GOODALL that they work together on a critical edition and first translation of the whole Buddhist section. Dominic GOODALL was immediately enthusiastic about covering not just this one section, but the whole text. From our first reading sessions together in the Pondicherry Centre of the Ecole francaise d'Extreme-Orient, S. L. P. ANJANEYA SARMA not only took part, but took the role of the expounder of the text as it was then constituted, drawing attention to anomalies of style, grammatical difficulties and resolutions, and to parallel discussions in other branches of literature.

Dominic GOODALL took up the task of collating the readings of the various sources and furnishing a preliminary edition to serve as the basis for discussion.

It was to Alex WATSON that the lion's share of the work fell, for he logged our vigorous discussions and, for most passages of the work, drafted out the first translation. The majority of the annotation too is his, as is the introduction (except for the section entitled 'Sources').

Over the years, various people too numerous to remember, let alone list, participated in our reading sessions and offered suggestions. To these should be added the participants of the various sessions of the "International Intensive Sanskrit Summer Retreats" (co-organised by the EFEO and the Indo European Studies department of Eotvos- Lorand University, Budapest) at which parts of the text were studied. Their emendations and comments are recorded from time to time in our apparatus and notes. Four individuals made an especially important contribution and their names will be found rather frequently. The first is Professor MANI DRAVIDA, who kindly came to Pondicherry on several occasions to expound particularly rich sections of the text that baffled us. He did so with his characteristic magisterial ease and clarity, invariably leading us to reconsider our interpretations of several points. The second, Professor Alexis SANDERSON, attended no reading session, but generously furnished us with his annotated photocopy of the Devakottai edition. His jottings led us to important parallels and pointed up a number of corruptions in the transmission of the text. The third, Professor Harunaga ISAACSON, discussed many tricky passages with us through e-mails, Skype and on his visits to Pondicherry. The fourth is Professor Kei KATAOKA, who not only furnished us with crucial photographs at the beginning of the project, but also sent us lists of corrections in its final hours. To each of them we owe a considerable debt of gratitude.

We are grateful too to the two anonymous readers for their suggestions and to the following libraries for allowing us to consult and to photograph or copy manuscripts used for our work:· the French Institute of Pondicherry; the Oriental Research Institute, Mysore; the Government Oriental Manuscripts Library, Madras; the Adyar Library; the Dharmapuram Adheenam; and the Bhandarkar Oriental Research Institute, Pune. We benefitted too, of course (as does almost every manuscript-related in do logical project in this day and age), from the use of microfilms made by the Nepal-German Manuscripts Preservation Project, for even though no manuscript transmitting our text survives in Nepal, our annotation draws upon other works that are transmitted there.

A word about the style of the translation is perhaps expected. It will at once be obvious that it does not read as an independent text in smooth English, but that it is intended as a tool to enable the reader to follow the Sanskrit text. Accompanying the translation is a very considerable body of annotation, some of it devoted to laying bare doubts about the constitution of the text, about how to interpret the syntax, about the author's idioms, word-choice, and particle-use, and about details of Saiva theology that are alluded to, but most of it intended to help the reader follow each step of argumentation. This style of translation-more like the Pompidou Centre, with all its pipes and ducts and, cables picked out in bright colours and exposed to view, than the Musee du Quai Branly, clothed in a "living wall" of leafy green naturalness-will not appeal to all. But it is perhaps most suited to this sort of text, a fascinatingly rich, but rather tough piece of exegetical writing, the study of which casts light not only on the history of Saiva thought, but on a number of theological and philosophical doctrines for which little other testimony survives.

Introduction

1 Preliminary Remarks

The Paremoksanirasokarika: of Sadyojyotih (675-'725 CE)2 is a text of 59 verses that lists and then refutes twenty positions regarding the nature of liberation (moksa). Its commentary by Ramakantha (950-1000 CE) expounds the twenty positions, not necessarily in the way Sadyojyotih understood them, and then refutes them, occasionally just by elaborating Sadyojyotih's refutation, but frequently by adding long digressions and new arguments.

The twenty positions are listed in the left-hand column of Figure 1. They are given there in the order in which they are listed by Sadyojyotih, and expounded by Ramakantha (which happens to be different from the order in which they are refuted).

The proponents of these positions are never named by Sadyojyotih, and only very occasionally by Ramakantha. But enough evidence can be garnered to propose identifications of the proponents in almost all cases. These are listed in the right hand column.

The texts thus give us a view outwards on to what traditions Saiddhantikas (= those belonging to Sadyojyotih and Ramakantha's tradition, the Saiva Siddhanta) in the final centuries of the first millennium saw surrounding their own, whom they regarded as their rivals, and which doctrines and arguments of these opponents they considered to require refutation. Of Ramakantha's nine surviving texts+ five have up to now been partially translated into a Western language" and one completely. 6 This publication adds a second complete translation.

This is one of the most interesting of Ramakantha's texts for Indologists who are not specifically concerned with Saiva Siddhanta, because of the snapshot it provides of the religio-philosophical landscape of tenth-century India. About half of the twenty positions are well known from other sources, but the other half have left little trace elsewhere in Sanskrit literature. The text thus offers a unique glimpse of certain forgotten conceptions that came to be swamped by those of the classical traditions. Some of them seem to be unknown even to Ramakantha, having presumably been pushed into obscurity in the centuries between Sadyojyotih's time and his own.? The value, for the historian of ideas, of this record of archaic views is not only that it provides a fuller picture of the variety of conceptions of liberation, but also that it helps to explain the genesis of some of the more well-known classical views."

The verses and Ramakantha's commentary contain sections only for the twenty positions that are refuted, having no separate section giving the authors' own Saiddhantika view. This is presented in. Sadyojyotih's Moksakarika and its commentary by Ramakantha, the Moksakarika being considered to some extent a separate text, and to some extent part of a larger text encompassing it, the Paramoksanirasokarika and others.? It should not be thought, though, that our texts are of no value to those seeking to understand the philosophy and theology of Saiva Siddhanta. We get insights into Saiddhantika thinking at every stage of the refutations, for in refuting rival traditions Sadyojyotih's and Ramakantha's own presuppositions are brought to bear, and we see what separates their own thinking from that of their opponents. The Saiddhantika view of liberation, furthermore, is expounded in passing at several points in Ramakantha's commentary (ad verses 6-'7, 27, 31-32, 37-42).

The Saiddhantika view is that liberation consists in the manifestation of the soul's innate qualities of omniscience and omnipotence. The soul is then the same as God (isvarasama), where 'same' means qualitatively identical but numerically distinct. This is very close to views 18, 19 and 20, which also hold that the liberated soul is the same as God, being omniscient and omnipotent. They differ from each other and from the Saiddhantika view in their explanations of how omniscience and omnipotence become associated with the liberated soul. In view 18 those two qualities arise from scratch (Utpattivada); in view 19 they are transferred from God (Sankrantivada); in view 20 the soul is possessed by them, as one can be possessed by a spirit (Avesavada); in the Saiddhantika view, omniscience and omnipotence already exist in souls prior to liberation in an unmanifest state, and at liberation they become manifest as a result of the removal of the soul's Impurity (Abhivyaktivada). Sadyojyotih and Ramakantha divide views 18, 19 and 20 off from the rest (see verses 6 and 7). Ramakantha describes the proponents of these three views as 'belonging to our own religion' (samanatantrikas), meaning not that they are quite Saiddhantikas but that, unlike the proponents of all the other seventeen positions, they belong to the same wider (Saiva) religion.

The twenty views are not classified by our authors in any way other than by this dividing off of the last three, and the aligning of them with the Saiva Siddhanta. But here are two ways in which they could be arranged thematically.

(1) The views can be differentiated through the following sequence of dichotomies (see Figure 2). First there are those that are theistic and those that are non-theistic, a 'theistic' view being one according to which the liberated soul exists alongside, below or, in one case, above God. Those which are non-theistic can then be subdivided into those according to which not only is there no God, but there is also no self, and those for whom what exists in the liberated state is a self. Into the former category fall the Buddhist and Carvaka views. Finally, those which accept a self can be subdivided into those for whom individuality is preserved in liberation, and those for whom liberation consists in the dissolution of the individual self. In the first camp the principal proponents are Sankhya and Nyaya; in the second they are Advaita Vedanta and Pancaratra.

(2) Liberation has been contrasted with another goal of Indian religion, the acquisition of supernatural powers (siddhis), by regarding the pursuit of the former as a search for 'freedom from', and the pursuit of the latter as a search for 'freedom to' There is no denying that the liberated states of the Buddhists, Naiyayikas, Vaisesikas, Sankhyas and Advaita Vedantins, involving as they do a complete lack of cognition and action, are strongly marked by a propensity for 'freedom from'. But this tendency was rejected, and indeed ridiculed," by other traditions, for example the theistic ones. In .many of these we find a pronounced predilection for 'freedom to' in the omniscience and omnipotence that they claim to be the culmination of the path they teach. Once the diversity of liberation doctrines is taken into account, the dichotomy of the two kinds of freedom becomes useful not because we can equate one kind with liberation, but because we see how the two kinds are differentially present within the various liberation doctrines. The twenty views can be laid out on a continuum, the two poles of which are the two types of freedom.

At one extreme we have the Buddhist view according to which liberation consists not just in freedom from suffering but freedom from existence itself. Here there is clearly no 'freedom to' know or do anything. This was not the only Buddhist view but it is the one that is given in our text (view 16). With the Buddhists fall the Carvakas. They too maintain (view 17) that the individual completely ceases to exist, though for them this 'liberation' happens to everyone at death, and is not an achievement accruing only to the enlightened. To the right of these two come the Naiyayikas and Vaisesikas, They do maintain (view 15) that the individual continues to exist in liberation, but without any consciousness or agency. Here we have freedom from knowing and doing, but not freedom from existence. Next come the Sankhyas (view 1), Advaita Vedantins (view 3) and Pancaratrikas (view 4). Consciousness continues in liberation for these, but it is a consciousness that is completely devoid of objects of experience. The light of consciousness is switched on in the liberated souls of these traditions, unlike in liberated Naiyayikas and Vaisesikas, but it is a light that, as it were, shines out into empty space without illuminating anything.

So all of these conceptions of liberation deny the presence of any changing states of consciousness, something that attracted comment from Andre BAREAU. Having said of Buddhist nirvatta (1973: 94) that it either must be pure nothingness in which nothing of the person remains, or 'must have resembled a profound and dreamless sleep, a complete unconsciousness', he goes on to write: To people who, like all Indians. believed themselves to pass without ceasing, without rest, immediately, from one existence to another, that is to say from one series of states of consciousness to another, that eternal and complete peace of psychic nothingness must have seemed desirable, whereas it has always terrified people in the West.

This is an important reflection on the question of why a complete lack of experience was promoted by some as the highest aspiration and the upper limit of human achievement But as we continue along the continuum, we will see that such a contentless liberation was desirable neither to all Indians, nor even to all of those who believed themselves to pass ceaselessly from one incarnation to another.

Next comes view 12 in our text, according to which liberation consists just in freedom from impurity (mala). Here we reach the first view that postulates knowing and doing in the liberated state. Then in views 8, 11 and 13, which are examined in some detail in section 5.2 of this Introduction, the power of knowing becomes expanded in liberation into omniscience, though the power of action is not found. In the views dealt with before this paragraph, the goal of freedom from suffering is taken to necessitate freedom from all cognitive experience, including that which is either pleasurable or neutral. In the views so far mentioned in this paragraph, cognitive activity continues, but it is not clear whether pleasure is present. Pleasure may have been regarded as only possible if alternating with suffering, its nature and existence deriving from a contrast with the latter. But in view 2 we find an explicit rejection of the presupposition that freedom from suffering requires also freedom from pleasure; the upholders of this view maintain that in liberation souls experience pure, uninterrupted, unexcelled and unbounded (suddhanirantaraniratisayanavacchinna) pleasure.

With the postulation of this kind of pleasure and of omniscience, we have arrived at views according to which liberation entails not just the removal of life's possibilities (such as suffering), but the addition of things not possible in the life of the unliberated. In other words we have arrived at conceptions of liberation that involve an element of 'freedom to'. This becomes more pronounced as we continue through the remaining views. These last two views still deny any action on the part of the liberated; though cognitive powers may increase, agency decreases. That changes at this point of the continuum.

Liberation as conceived of by the proponents of view 7 involves becoming one of God's principle attendants (mahagana), with all the extra powers and privileges that this promotion involves. This and view 14 are the only two of the twenty that conceive of liberation as an embodied state. According to the latter, the liberated soul sheds its samsaric body and sense faculties, but takes on new, highly elevated (tarakatara) ones. This new body and sense faculties, which unlike the previous ones are not caused by karma and not characterised by pain, allow a liberated existence on another planet (tarakabhuvane, literally 'in a world in the stars').

We are nearing the end of the continuum, and the advocates of all of the remaining four views claim that the liberated soul is omnipotent. For these proponents, to leave behind one's body and sense faculties as one enters the liberated state is not to leave behind the possibility of action; rather it is to expand its potential range. It is not the body that acts, nor is the body a necessary instrument of agency. It is the agent, i.e. the soul, that acts; and its agency consists not in moving, but in causing movement, as a magnet causes movement in iron-filings without itself moving. Having thrown off the bonds that limit the full expression of its power of action, and without a spatially limited body to restrict its sphere of operation (its 'magnetic field'), this sphere becomes equal in extent to that of the soul itself, i.e. all-pervading.

The difference between the four views (9, 18, 19 and 20) is just that, though they all postulate omnipotence and omniscience, in view 9 the operation of this omnipotence is subject to God's instigation, so that such souls lack complete autonomy.'?

We now introduce more detail about those views whose treatment by Sadyojyotih and Ramakantha is of most philosophical or historical interest.