The Cause of Seamless Integration: Revisiting the Dual Authorship of the Kavyaprakasa.

• Author: Bronner, Yigal

Who composed the Light on Literature (Kavyaprakasa), the dominant treatise on Sanskrit poetics in the second millennium CE and one of the most commented upon texts in the entire history of South Asia? There is a strong and longstanding tradition, harking back to the earliest written responses to the Light and its oldest extant manuscripts, that Mammata, the text's main author, left it incomplete, and that the final portion was composed by Alaka (also known as Alata or Allata). Later sources even identify the precise spot of the changing of the guard, the ornament parikara in the tenth and last chapter. But if we trust this tradition, what are its implications for our understanding of the Light, the process of its composition, and its overall congruity?

To the best of my knowledge, the main engagements with this question belong in the early decades of the twentieth century. In 1912 V. S. Sukthankar convincingly demonstrated that the author of the latter part of the last chapter of the Light suddenly relies almost verbatim on Rudrata's discussion, whereas the author of the earlier part of the chapter, while familiar with Rudrata and occasionally resorting to his text, depends on a far more varied pool of sources. (1) In the next decade H. R. Divekar noticed stylistic divergences that seemed to distinguish between the two parts of the book. (2) Unfortunately, Divekar muddied the water unnecessarily. The Light, like many texts of its genre, consists of two distinct textual layers: brief versified statements that provide the gist of the teaching in a compressed manner (karika) and prose discussions (with a plethora of examples, most of them in verse) that unpack them and take the discussion further (vrtti). Divekar decided, based on entirely flimsy evidence, that not only did Mammata leave the karika portion of his text incomplete, but that he also never got to write the vrtti, and, hence, that the vrtti in its entirety was composed by Alaka, the author of the remaining karikas. (3) This intervention derailed the debate, and for a while researchers were seeking to prove (or disprove) that the author of the karikas was familiar with the vrtti. (4) With this wild goose chase, the original exploration basically exhausted itself, (5) and from this point on, the already staked positions sometimes get rehashed but are never advanced. (6)

What no scholar, modern or premodern, seems willing to explore are specific discrepancies of opinion between the two parts of the text. The commentators who report a change of hands seem blind, perhaps by choice, to any concomitant change of hearts, and the modern scholars seem attuned primarily to questions of compositional methods and personal styles. This paper looks at one possible point of substantial difference and its implications to the way we understand the Light, the process of its constitution, and the history of its reception. More broadly, the paper probes writerly and readerly approaches to textual unity, especially, but not solely, in a possible coproduction: what kind of credit do authors claim and receive, and what premium is put on the text's overall integration.

1. THE LIGHT'S REJECTION OF "CAUSE" (HETU)

   The key passage in question is from the part attributed to Alaka. It belongs in the prose (vrtti) part of the text and offers a follow-up comment on the discussion of "garland of causes" (karanamala). The added comment explains the reason for not listing "cause" (hetu) among the Light's ornaments:

   "The ornament 'cause' is stating the effect along with its cause as inseparable." We do not provide here this definition of "cause." After all, when it has a form such as "clarified butter is long-life" it can never be an embellishment, as there is nothing striking in it. Now as for:

   "The lotuses are in bloom high and low, all bees abuzz, the cuckoos frenzied-- the ravishing season that excites the world has begun." The traditional view is that this takes the form of literature only on the power of "tender alliteration" and not because of fancying the ornament "cause." Hence "cause" is nothing but the "inferential sign of poetry" with which we have already dealt. (7) A bit of background is necessary here. The status of "cause" (hetu) has always been a topic of debate among Sanskrit theorists. Bhamaha rejected it as nonpoetic, to which Dandin responded by embracing it as a "number-one ornament" and turning its exposition into a tour de force. (8) Udbhata silently ignored it and instead introduced "the cause of poetry" (kavyahelu), which he also dubbed "the inferential sign of poetry" (kavyalinga), without explaining whether this is a new name for the old device or a new one altogether. (9) Rudrata, for his part, ignored Udbhata's nomenclature and reintroduced "cause" (hetu), defining it--and this is the definition that we find rejected in the above quote--as inseparable (or identical, or perhaps coreferential; abhedatah) statements of causes and effects, although even he seems to acknowledge that this ornament is set aside from the others (anyebhyah prthagbhutah). (10) It would appear that the Light, which has earlier defined kavyalinga and now rejects hetu, sides with Udbhata on this point. Moreover, the above passage provides a principled refutation of Rudrata's position. There is nothing striking (vaicitryabhavat) in statements that have the cause and the effect identified by being coreferential (e.g., "clarified butter is long-life")--which is how the passage seems to interpret abhedatah in the definition--whereas statements of causal relations that do have a striking quality come under the purview of kavyalinga, the real ornament "cause" when it comes to poetry.

   Note that several aspects of this passage attributed to Alaka appear to be in strong harmony with earlier portions of the Light. The grounds for the rejection of hetu as defined by Rudrata, as the commentators point out, hark back to the discussion of semantics in the Light's second chapter. (11) Here the karikas enumerate six types of secondary capacities of language (laksana) and highlight two of the six as involving attributive transference (gaunt) based on similarity, something that does not happen in the other four. (12) The vrtti further explains that superimposition (saropa, as in "the VahTka is a cow") and determination (sadhyavasana, as in "this one is a cow") based on similarity stand apart from those based on some other relationship, such as that of cause and effect: "In statements such as 'clarified butter is long-life', or 'this one is long-life', there is some other relationship such as that of a cause and effect that is not based on similarity." (13) Later in the book, but still in the karika portion attributed to Mammata, the semantic processes of "superimposition" (aropa) and "determination" (adhyavasana) that involve attributive transference are identified as the bases of the ornaments "identification" (rupaka) and "intensification" (atisayokti), respectively. (14) Now, in the vrtti attributed to Alaka, the "pure" category of superimposition and even the same example from chapter 2 ("clarified butter is long-life") are reintroduced to reject the status of "cause" (hetu) as Rudrata is understood to have defined it. The passage also refers the reader back to the karika on "the inferential sign of poetry" (kavyalinga), in the portion attributed to Mammata, as the accepted form of "cause." It would thus seem that all the levels of the text--the karika and vrtti in chapter 2, the karikas attributed to Mammata in chapter 10, and the vrtti comment on hetu attributed to Alaka later in this chapter--are in harmony and anticipate or rely on each other.

   Indeed, note that the rejection of an ornamental variety, while rare in the Light, has a precedent in the vrtti portion attributed to Mammata, where a subtype of "identification" (rupaka) called "chain identification" (rasanarupaka), also coming from Rudrata, is explicitly set aside. In both rejections of Rudrata's categories, the example verses from his text are cited, and in both, the language used for rejection is very similar (na vaicitryavad iti na laksitam vs. na laksitah... vaicitryabhavat). (15) Thus, in both passages the author has Rudrata's text in mind, he highlights his differences from him, and then similarly justifies his views. In short, our passage seems to be in harmony with the earlier layers of the text.

2. "CAUSE" (HETU) AS A PROBLEM IN THE LIGHT

   The picture, however, is more complicated, for two main reasons. First, the Light does appear to accept the ornament "cause" in chapter 4. This chapter, considered one of the highlights of the work, offers an extensive exposition of suggestion and its numerous types and subtypes, including cases where ornaments either suggest further meanings or are, in turn, suggested. Only a smaller subset of ornaments is mentioned in this chapter (fourteen out of the sixty-three ornaments discussed in chapter 10), but "cause," conspicuously referred to as "the ornament 'cause'" (hetvalahkara), is one of them. Indeed, it is invoked apropos of three separate illustrations, which makes it one of the most cited ornaments in chapter 4. (16) And what is more, the chapter also identifies "the inferential sign of poetry" (kavyalinga) in three additional instances, and the two ornaments are invoked in close proximity. (17)

   Why would the author repeatedly identify "cause" as an ornament, if he does not believe it to be one? One possible explanation is that ornaments are not the focus of attention in chapter 4, and the author is being careless in naming them. Note that the author of this chapter is occasionally not fully consistent in his terminology: kavyalinga itself is twice called thus and once kavyahetu, a variation already known from Udbhata's treatise. (18) Likewise, there are references to "apparent contradiction" (virodhabhasa) and also to "contradiction" (virodha), but only the former is defined...