somananda e-book

common cause of any divergence in the readings of that manuscript, on the one hand, from readings shared by G, P, and R, on the other. In other instances, the readings shared by P and R are superior to those shared by G and J.235 Put differently, there are occasions when the transmission of the text must have been one that involved the subsequent corruption of given readings witnessed in P and R but not in G and J (PR! GJ), or readings now found in G, P, and R but not in J (GPR! J). R, in turn, witnesses an (only slightly) earlier form of the text than what is found in P, this being discernible not simply on the basis of the relative antiquity, compared with P, of the manuscript (a fact that, as mentioned above, cannot serve on its own to prove the relative antiquity of the readings of the manuscripts in question), but also because one not infrequently finds R according with the superior readings of G and J against P, even if it is more common for P and R to witness identical variants. On the other hand, P also sometimes accords with the readings of G and J against those of R, as R sometimes may be shown to witness corruptions of readings found in P (and/or G and J).

There must be, then, three intermediate stages in the transmission of the ŚD and ŚDVṛ for which we currently have no witnesses. One must exist at a point of divergence of P and R, on the one hand, from G and J, on the other, the correct and better readings more often, but not exclusively, being transmitted to G and J, rather than P and R. A second stage of transmission not witnessed in our manuscripts must exist at the point of divergence of G from J, with the superior variants regularly appearing in J, but also sometimes, if rarely, in G when the readings of the two texts diverge, the variants of G sometimes according with P and R when they differ from J, and sometimes being unique among the witnesses consulted to G alone. Finally, a third divergence must exist in a hypothetical state of the text that would have existed prior to the divergence of R and P, this because R not infrequently shares variant readings with G and J (and sometimes with T, and even C) that are not witnessed in P; and, conversely, R sometimes witnesses unique variants that are not found in any of the remaining three manuscripts of the Northern Recension.

The relationships of these manuscripts, then, may be mapped in a diagram of the stemma, as seen in figure 1.

235Such true variants the preferable ones of which are witnessed in P and R, as opposed to G and J, are few in ŚD (and ŚDVṛ ad ŚD) chapters 1–3, but they do exist. For example: P, R, and Ked:read purobhāge in ŚDVṛ ad ŚD 1.19–20ab, while G and J read pare bhāge, an unlikely reading given the context, and also because it would place the power of eagerness, aunmukhya, within the highest condition of Śiva (which Somānanda never describes as a “part” [bhāga] of Śiva-nature, anyway) when it should be the first moment emerging from it. To offer a second example: P, R, and Ked:read taddṛśyānāṃ for tato dṛśyānāṃ attested by both G and J in ŚDVṛ ad ŚD 2.26cd–28ab.

but P diverges from R at a point where it may alternatively witness errors not found in R or, very occasionally, correct readings found in J and G but not R. There are also rare occasions where P and/or R record correct readings that are absent from both J and G. Finally, the line with the arrowhead that points towards C from the direction of the Northern Recension should be understood to represent the contamination of C with readings from the Northern Recension.

If the relationship of the various manuscripts consulted for the present edition is rather complex, one may take solace in the fact that this picture meets the expectations one would have for such a collection of sources, the provenances of which span a number of centuries and hail from all corners of the Indian subcontinent, from Kashmir to Kerala and from Bengal to Maharashtra and Rajasthan. Nevertheless, the state of the evidence is such that it demands that one weigh all of the readings of all the manuscripts wherever they vary, as any of the manuscripts theoretically could offer the best reading in any given instance, the one exception being C, which uniformly may be shown to be corrupted whenever it witnesses a unique reading of the ŚD. In practice, however, one must have good reasons not to follow the reading of J when it comes to editing both the commentary and the root text (mūla), as the readings there may regularly, if not exclusively, be judged to be the superior ones, with the caveat that T sometimes witnesses an earlier form of the text and therefore occasionally furnishes the preferable reading. It should be added, as well, however, that J itself, like T, sometimes shows itself to be witness to corruptions not found in the manuscripts that witness a subsequent form of the text of the Northern Recension, though these usually involve the expected sorts of trivial variants, such as the omission of the final visarga, small errors in the transcription of vowels (such as the recording of the short diphthong o for the short vowel a, the long vowel ā for its short counterpart, a, etc.), and the like. These readings, of course, have of necessity been filtered from the accepted reading of the present edition.

C O N V E N T I O N S O F T H E E D I T I O N

The apparatus of the edition is a positive one, as it explicitly notes both those manuscripts that record the text appearing above the line and those that do not. There are five registers of notes to the critical edition. The first and third registers record any comments, parallel passages, or the like that are related to the mūla and the commentary, respectively. The second and the fourth record the variant readings, with the former recording the variants of the mūla by verse number and by the quarter-verse (pāda), labeled a, b, c, and d, the latter recording the variants found in the commentary by the line number of the chapter in question. The fifth, bottom register of the notes records the folia on which the given passages of text are recorded, this by first identifying the manuscript in question, followed by folio number and an indication of whether the following text appears on the front (“r” for recto) or back (“v” for verso) of the given folio.

of the various Sanskrit proper names and/or key terms, with two exceptions: I regularly refer to paśyantī with “she” (and related pronouns) in the translation of the second chapter, as I do the same for śakti, “power,” in the third chapter.

Although both Somānanda and Utpaladeva use a great deal of technical language, I have refrained wherever possible from the common practice of quoting such terms (in parentheses) in the body of a translation. I find that the frequent inclusion of numerous Sanskrit technical terms in the body of the translation often serves only to distract the nonspecialist reader, while the Sanskritist is regularly aware of how I have interpreted the particular terms and expressions. (Instead, I have identified in the notes the particular technical or other terms that were rendered in the translation when I feel it will aid the specialist or even nonspecialist reader.) I have similarly limited the use of untranslated Sanskrit words in the translation, though I do quote the Sanskrit when the terms in question are relatively well known technical ones. I don’t feel it is necessary, in other words, to develop new translations of relatively well-known Sanskrit terms just for the sake of avoiding the use of Sanskrit terms in the translation. Such a practice could easily create confusion when a well-known Sanskrit term is rendered with a new and therefore unfamiliar English term, thereby clouding the authors’ intended meaning. There is a second reason for quoting Sanskrit terms in the translation: I sometimes do so in places where Utpaladeva analyzes a given term, or simply defines the term in question with a pure synonym, as occurs for example in his commentary on ŚD 1.2. (It is, moreover, for this reason that I quote the Sanskrit verse of the ŚD prior to its translation: the presence of the verse within the translation allows even those who do not read Sanskrit to understand from where Utpaladeva has taken a given term he subsequently glosses.)

Any other of the relatively few Sanskrit terms that appear in the translation have almost without exception been explained in the present Introduction, and those few that have not are explained on the page, in the various notes to the translation, or in Utpaladeva’s very explanations of the terms in question. I should add that I have also generally endeavored, though not without exception, to translate a single Sanskrit term with a single English equivalent, not because I believe that doing so is a practice indispensable to any sound theory of trans- lation—I do not—but because I think it will simplify matters for the reader. (This translation practice is of course more difficult to sustain when rendering longer portions of a given work, as one inevitably encounters a greater variety of contexts and therefore a greater semantic range for certain lexical items, which must be presented to the reader with the appropriate emphasis and nuance in English.)

Whatever text I supply to the translation is included in parentheses: ( ). I also place in square brackets [ ] any language that is used to explain that which immediately precedes the bracketed material. For example, if the text reads, as it does on ŚDVṛ ad ŚD 1.25–25 “that, being a new sequential form,” I include in brackets the referent of the pronoun: “that [appearance], being a new

sequential form.” As these materials are added to, or offered in explanation of, what is found in the Sanskrit text, the translation has been composed in a manner that allows it to be read legibly even when excluding anything found in parentheses or brackets. This is so with only two types of exceptions: I also sometimes identify in square brackets the referent of a pronoun or the agent of an action who is signified obliquely by the conjugation of a given verb. (Note that I employ this same convention even when the action in question is conveyed through one of the many nominal constructions of Sanskrit.) The most common example of this is found repeatedly in the introductory passages Utpaladeva furnishes prior to a great number of the quoted portions of the ŚD: “[Somānanda] says.” For although Utpaladeva nowhere refers to his teacher by name in what survives of the ŚDVṛ, Somānanda is clearly the agent implied by the verb in these introductory passages. In such instances, one will have to include the term found in brackets for the translation to read legibly. I try whenever possible, however, to avoid the use of any parentheses or square brackets, as I feel, again, that such conventions often serve to distract the reader.

I frequently make reference to the ĪPK and ĪPVṛ of Utpaladeva in the translation and notes thereto, and in doing so I regularly refer to the numbering of the verses found in Torella’s edition of the text. References to passages of the Śivadṛṣṭi and Utpaladeva’s commentary correspond with the numbering of verses found in the edition presented here. When referring to passages of these texts that lie beyond the scope of the present edition, I refer the reader to the numbering of the verses of the published KSTS edition (Ked:). When referring to the VP, I follow Rau’s numbering of the verses.

The many notes accompanying the translation are meant to serve two distinct audiences. Some are purely technical notes, which explain philological problems, reveal textual problems or variants, explain my interpretations of the many compound words in the Sanskrit text, and the like. Others are rather meant for the general reader, as they explain the flow of logic found in the text or elaborate on a given term, concept, or concern found therein. When an important term is translated, I often record in the notes the Sanskrit term that was translated, as mentioned already. Wherever possible, I indicate the presence of parallel passages and identify any passages of text that are quoted elsewhere in the primary literature. Some of these parallel passages are recorded in the notes, but, whenever possible, I simply note the location of the parallel passage and quote it in the notes to the edition. Finally, I refer not infrequently to the various types of Sanskrit compounds—coordinative, determinative, descriptive, exocentric/possessive, and so forth—in my translation and notes (this in no small part because Utpaladeva, like all good Sanskrit commentators, regularly discusses these grammatical forms in his commentary). Those who, being unfamiliar with such Sanskrit grammatical constructions, wish to know more about them, are advised to consult the detailed and accessible, if compact, explanations thereof

This page intentionally left blank

{ PART II }

The Translation