The Mythico-Ritual Syntax of Omnipotence

·期刊原文
The Mythico-Ritual Syntax of Omnipotence
By Lawrence, David
Philosophy East & West
V. 48:4 (1998.10)
pp. 592-622
Copyright 1998 by University of Hawaii Press
Hawaii, USA

--------------------------------------------------------------------------------

p. 592

Introduction
This essay will examine the interpretation, by the tenth- and eleventh-century Kashmiri philosophers Utpaladeva and Abhinavagupta, of monistic, tantric `Saiva myth and ritual in terms of Sanskrit action syntax. I will focus primarily on their works on II`svarapratyabhij~naa but will also refer in Abhinavagupta's Tantraaloka and Tantrasaara. [1] Aside from the importance or these particular theories, I believe that the area of Sanskritic philosophical speculation they exemplify is of interest for general comparative research.

Propaedeutic Overview of Monistic `Saivism and the Pratyabhij~naa Apologetics
Monistic Kashmiri Saivism contains a plethora of complex, interweaving symbolic and ritual traditions, of which Utpaladeva and Abhinavagupta adhere to a lineage called Trika. For our purposes it will be most useful to describe the broader patterns. For these traditions, the only reality is the omnipotent deity `Siva. `Siva, out of a kind of play, divides Himself from His consort and power `Sakti, and in sexual union emanates the universe through Her. Liberation is the realization of one's true self as `Siva. This is achieved through diverse ritual practices that reenact the cosmogonic myth, ranging from sexual rites, in which the partners become `Siva and `Sakti, to theosophical contemplations.

The Pratyabhij~naa system developed by Utpaladeva and Abhinavagupta is both a philosophical apologetics and a gnoseological internalization of tantric praxis. Its primary mode of explanation is epistemological. The thinkers explain the myth of `Siva's emanation of the world through `Sakti as His self-recognition (ahampratyavamar`sa); they also equate this self-recognition with Bhart.rhari's principle of the World Absolute (`sabdabrahman/paraavaak) in order to interpret creation as linguistic in nature. In debates with the Buddhist logicians, the `Saivas attempt to show that this recognition is the reality underlying and constituting all states of affairs. As a ritual enactment, by thus disclosing the necessity and ubiquity of `Siva's self-recognition, the system leads the student to complete participation in it. [2]

Indian Semantics of Mythico-Ritual Narrative
Now Utpaladeva and Abhinavagupta also explain the myth and ritual of `Siva-`Sakti, or `Siva's self-recognition, as His kriyaa. [3] This word, derived from the root k.r, is cognate to the English word "create," and means both creation and action. The universe is `Siva's "cre-ation."

p. 593

This interpretation is also based on semantic theory, just as the `Saivas follow Bhart.rhari in viewing the universe synthesized through `Siva's self-recognition as inherently linguistic, they follow the dominant Vyaakara.na, 'Grammarian', view that kriyaa, action, is the chief meaning of all language. Language is said to express being (sattaa) as an action to be accomplished. [4] This view extrapolates to all discourse the priestly interpretation of the Vedas as expressing injunctions for the ritual kriyaas of sacrifice. [5]

I believe that the `Saiva and related semantic concepts of action have some analogy to contemporary Western theories of the intrinsically narrative character of experience. [6] However, the `Saiva and other Indian theories neither relativize narrative like Lyotard nor "revisionistically" metaphorize it like Ricoeur. [7] The mythical story that is paradigmatic for human ritual action is not understood to be a mere quality of experience generated by imagination, but is posited of existence itself. We are talking here of a narrative or mythico-ritual narrative ontology.

Diverging Mythico-Ritual Syntaxes
Now, the `Saivas along with other Indian philosophies further interpret the action of myth and ritual with theories of Sanskrit syntax. This is provided for by the system of the Sanskrit kaarakas. While the six kaarakas are historically related to the Indo-European cases, their conception is more subtle, and bridges in a particular way the areas of semantics and syntax. [8] Kaarakas describe various logical relationships of the referents of declined nouns to the main action expressed by a verb: the action's agent (kart.r), its chief instrument (kara.na), the object it effects or affects (karman), the location (adhikara.na), where it comes from spatially, causally, and so forth (apaadaana), and for whom or what one performs it (sampradaana). The same kaaraka may actually be expressed in more than one case declension. The genitive is not even considered a kaaraka, as it usually articulates relationships between nouns, rather than between nouns and a verb.

The word kaaraka is a derivative from the same root as the word "action" (k.r), having the causal significance of 'actor', 'maker', 'factor.' The kaarakas are understood as functioning to accomplish the action expressed by the verb. The action is said to be 'that which is to be established' (saadhya) and the kaarakas are 'establishers' (saadhanas). Take the typical example: "He cooks rice in the pot with fire." The pot contains the rice and water, and the fire heats them. The rice, that is, the direct object expressed in the accusative, is explained to be the locus of the result (phala) of the action of cooking. This result is a transformation of the nature of the rice.

In accordance with its general emphases on language and action, Sanskritic speculation about many philosophical topics often has recourse

p. 594

to argument about the interrelations between the various kaarakas and the overarching kriyaa. Even traditions such as Nyaaya-Vai`se.sika, Advaita Vedaanta, and Buddhism, which do not accept the notion that semantics is primarily the expression of action, are influenced by the view and also engage in these discussions.

To help thematize one of the trajectories within these discussions, I point out an analogous area of Western theory, propounded by Kenneth Burke in his Grammar of Motives. [9] Developing his categories from a model of drama, Burke identifies five intrinsic factors or "motives" necessary to all accounts of action: the act itself, the agent, the scene, the agency (i.e., instrument), [10] and the purpose. [11] The last four of these roughly overlap with the Sanskrit kart.r, adhikara.na, kara.na, and sampradaana, and an interesting parallel may also be observed between the causal significance of motive and that of kaaraka. [12] However, I am not concerned here with a detailed comparison of the schemes.

Rather I am interested in Burke's insight that different philosophical and literary expressions tend, explicitly or implicitly, to emphasize reductionistically or "feature" particular motives at the expense of the others. For example, Burke observes that Hegelian idealism features the agent. Materialism, from the Greek atomists through modern science and, I would add, historicism and social scientific contextualism, features the scene. [13] Burke interprets scientific and technological thought, pragmatism, and utilitarianism as featuring the instrument/agency, in an analysis which may be assimilated to other studies of contemporary "instrumental" rationality. Outside the classification of philosophies, Burke's grammar has notably been used in sociological theories to typologize different sorts of explanatory accounts and justifications of action. [14]

It is not necessary to attempt to use Burke to classify different South Asian philosophies. I believe that an important dimension of the Sanskritic debates about action syntax is precisely an effort of the different schools to articulate their own, literally grammatical, grammars of motives. This is exemplified in the area of inquiry that is the chief focus of this essay. I will here sketch how the `Saivas and other Indian theories either feature or downplay the role of the agent in relation to the syntactic nexus of action and its objects and results. [15]

The Orthodox Hindu and Buddhist Denigration of the Agent in the Syntactic Nexus of Action and Result
Now the mainstream of orthodox Indian philosophical systems, Hindu as well as Buddhist, has in various ways denigrated the role of the agent within the syntactic nexus. First, the agent is devalued in the same manner as all of the other kaarakas. As I have said, action is the chief meaning and referent of language. All of the kaarakas are accessories

p. 595

(saadhanas) helping to accomplish that action which is to be accomplished (saadhya).

In his study "What is Karma," Edwin Gerow places this point in a more complicated pattern of considerations that syntactically subordinate the agent. On the one hand, the word karman means action. On the other, it designates the kaaraka indicating the direct object, which is (usually) expressed in the accusative case. As stated previously, the direct object is understood to receive the result (phala) of the verbal process (vyaapaara), for example the rice which is cooked.

Gerow points out a strong tendency in Indian linguistic and philosophical speculation toward the identification of the verbal process with the result. The reasoning for this identification is complex. However, it may be stated briefly that the case of intransitive verbs, which to us would seem exceptional, became paradigmatic. In "He sits," the sitting is a unity of both process and result. The meaning of the word karman as both action and direct object articulates the identity.

Furthermore, explains Gerow, the unitary nexus of process and result came to be understood as the dominant feature of syntax. This is observed in a proclivity in Sanskrit toward passive constructions -- a proclivity that in itself emphasizes the agent's accessory status. Now, in passive syntax, the verb agrees with the karman, for example "The rice is cooked by him." In Gerow's view, the culmination of these patterns of thinking is the late grammarian Naage`sa's treatment of what would seem to us the oddest case as most paradigmatic, that is, passive intransitive syntax. Thus "It is sat (by him)." The process-result is featured strongly.

According to Gerow, this grammatical understanding is the same as the religious conception of karman as a chain of process and result extending across lives. The agent is seen as a kind of adjunct bound within the nexus. Gerow is careful to distinguish his views from earlier arguments that Sanskrit causes an alleged "passive character" of the Indian mind. [16] Nevertheless he still contends that it is primarily a fact about the Sanskrit language that was reflected in the various forms of the doctrine of karman, and was increasingly realized in the linguistic and philosophical speculation.

I believe that the range of ideas about karman in Indian history is far too vast to be explained this way. I also view Sanskrit and other languages in themselves as semantically more indeterminate. [17] My interest is rather in how the second-order, scholastic speculation reflects divergent religious-cultural understandings within the South Asian cultural area. In any event, Gerow has made the important observation that a number of philosophical and theological systems do think in precisely the way described, using grammar to support their views. I believe that this pattern of thinking articulates not only the agent's bondage to karman in rebirth for Hindus and Buddhists but also its subordination

p. 596

to the order of objective ritual behavior -- pertaining to caste, sacrifice, life cycle, and so forth -- in orthodox Hindu society.

Support is accorded to this mode of explanation by Alexis Sanderson's study, "Purity and Power among the Brahmans of Kashmir." [18] Sanderson examines conceptions of agency with a social-historical rather than a grammatical focus, and he is especially concerned with the associated symbolism of purity and impurity. Nevertheless, his analysis of the extreme but divergent approaches to denigrating agency of the orthodox traditions of Puurva Miimaa.msaa and Advaita Vedaanta is based on the very texts discussing the syntax of action.

The case of the orthodox ritualistic system of Puurva Miimaa.msaa is particularly interesting. The Miimaa.msakas actually stress the importance of the agent in carrying out ritual injunctions. However, the agent is "depersonalized" in various ways. As with most Indian traditions, he is subject to karman as action and result in rebirth. More specifically, he must follow the injunctions to perform rituals. Salvation comes from performing rituals that completely transcend all worldly purpose. Sanderson explains:

This contradiction, that of the "solipsistic conformist" was his self-representation as ritual agent. The notion of autonomous agency individualized the person, but his determination by a world of revealed duties, his wish to conform to the Brahmanical ideal, depersonalized this individual, purging him of all independent motivations. [19]

As Sanderson observes, Advaita Vedaanta accepts the necessity of social-ritual action described by Miimaa.msaa, but contends that it is ultimately illusory and escapes from it in a liberation equated with non-agency. [20] I will take `Sa^nkara as the representative of Advaita because his name is so well known in the West. [21] `Sa^nkara is thoroughly grounded in the ritualistic school of Puurva Miimaa.msaa. He rigidly defends the orthodox Hindu patterns of behavior. However, I would say that he is equally terrified of them. For him an agent is inexorably subordinated to the syntax of acting and enjoying results. `Sa^nkara thus systematically rejects the Puurva Miimaa.msaa process of exegesis and ritual action as the way to ultimate salvation. He divorces the saving knowledge of the real Self/Brahman from agency and action. The following passages from the Brahma Suutra Bhaa.sya are typical:

We maintain [as opposed to the Puurva Miimaa.msaa] that the knowledge of the Self does not pertain to something which is to be done. [This knowledge] is not for the purpose of avoiding or pursuing [anything]. Our excellence is [maintaining that], when there is the realization of the Self as Brahman, there is the abandonment or everything which has to be done, and the completion what has to be done. [22]

p. 597

This is what one who knows Brahmin realizes: "I am Brahman, which is completely different from that [limited self] known previously as agent and enjoyer; and is neither agent nor enjoyer in any of the three times [i.e., past, present, and future]. Thus previously I was neither agent nor enjoyer. Nor am I now. Nor will I be at a future time." Only thus is liberation possible. For otherwise, if there were no destruction of the karmas which have been proceeding for beginningless time, there would be no liberation. [23]

This gnosis of nonagency is the individuality of the renunciant as interpreted by Louis Dumont. [24]

The valuation of subordination to action in Puurva Miimaa.msaa and the attempt to escape from the same in Advaita Vedaanta indicate the diversity of ways of proceeding from the same or similar assumptions. Sanderson also observes that there are moderate devaluations of agency in the Hindu "middle ground," for example by the Vai.s.navas. [25] Of course, the mainstream Buddhist positions completely deny the existence of a self/ agent. I have said enough to provide a foil to the `Saiva conceptions.

The `Saiva Grammar of Omnipotence
As Sanderson explains, the monistic `Saivism of Utpaladeva and Abhinavagupta pursues the value of "power" as opposed to the orthodox Hindu "purity." The omnipotence of identity with `Siva-`Sakti or participation in `Siva's self-recognition stands in continuity with other limited magical powers (siddhis) that are pursued. And, as Sanderson points out, the tantric praxes transgress the ritual proscriptions concerning purity, for example regarding caste, diet, sexuality, and death. Likewise, for the monistic `Saiva traditions karman as bonding process and result extending across lives is not an inexorable consequence of action. It is described as one of three illusory taints (malas), the taint of karman (kaarma mala). [26] This taint is understood as only an incomplete realization of one's omnipotent, cosmogonic Action `Sakti. Freedom from karman is the realization of that `Sakti. [27]

In developing a grammar of power or omnipotence, Utpaladeva and Abhinavagupta take up several earlier understandings of the (albeit often delimited) role of the kart.r or agent, particularly from the Vyaakara.na and Nyaaya-Vai`se.sika traditions, and radicalize them. The alternatives provided by these considerations were not given sufficient attention by Gerow.

I will summarize the most important of these to the `Saivas. As I have explained, all of the kaarakas are understood to function in accomplishing the overall action or process (vyaapaara) expressed by the verb. They do this through their own subordinate processes. The pan holds the rice, the fire heats it, and so forth. Where are all the subordinate processes synthesized into the larger one? This is understood to be accomplished by the agent, who is the locus of the overall process (vyaapaaraa`sraya).

p. 598

Furthermore, according to grammar, it is the agent who is the instigator (prayojaka) of all of the subordinate processes comprising the larger one. He arranges the equipment, lights and controls the fire, and so forth. [28] Sometimes further subjective factors in this instigation are identified. The Naiyaayikas in particular stress that the agent has the intention or desire (icchaa) and makes the effort (yatna) that brings about the action. The followers of Paa.nini state that the direct object (karman), as receiving the result of the process, is that which is most desired (iipsitatama) by the agent.

I will make one more point. While the agent controls the processes of the other kaarakas, no other kaaraka has a similar influence on him. The other kaarakas are 'determined by another' (paratantra), but he is 'self-determined' or 'free' (svatantra) with regard to their operations.

The `Saivas explain the Lord's/Self's cosmogonic omnipotence as His creation of the universe from His mere intention (icchaa) and self- determining freedom (svaatantrya). Most important to their syntactic theory is their interpretation of `Siva's character as locus of all subordinate processes of all things in the universe. This is explained as His synthesis (anusa.mdhaana) of them by recognizing Himself through the recognitions of each of them.

This philosophical syntax is exemplified in the following explanation by Abhinavagupta, which uses the analogy of ordinary agency to elucidate that of the Lord:

Here [according to this system], action is really nothing but the Supreme Lord's intention [icchaa]. [This intention] consists of uninterrupted self- recognition [svaatmaaparamar`sa], which has the nature of unobstructed freedom [svaatantrya], and is not dependent on another... For [limited individuals such as] Caitra or Maitra, etc., the inner intention [icchaa, such as that expressed] "I cook" is the action. Thus, even though there is the relation of [one who is cooking] with numerous movements [spanda] such as putting something on the fire, etc., the [intention] "I cook" is uninterrupted. It is nothing but the intention [icchaa] "I cook" which appears as such movements... Thus is that recognitive judgment [vimar`sa] of the Lord, which has the nature of intention [icchaa, which may be expressed] "I lord," "I appear," "I manifest in cosmogonic vibration [sphuraami]," "I create through agitation [ghuur.ne]" [29] and "I recognize [pratyavam.r`saami]." The essential nature [of such recognitive judgment] is nothing but "I"... [30]

Thus there is a reductionism in a direction opposite to that which Gerow observed. In a sentence such as "Devadatta cooks rice in the pan with wood," the factors such as the pan, wood, and rice appear (prakaa`sante) as merged in the action. The action in turn rests (aa`sritam) in the agent. [31] According to the `Saivas, even action that seems to be situated primarily in the object is actually located in the agent through His unifying recognitive synthesis. [32]

p. 599

The `Saiva Syntax of Causation. The `Saivas elaborate this syntax in a number of different spheres of philosophical explanation. We may first look at their theorization about the important Indian philosophical topic of the cause-effect relation. [33] To begin with, the term most frequently used for cause in discussions of causality is kaara.na. This word is situated in syntax in the position of the kaaraka called kara.na, which corresponds to our instrumental case. [34] Whereas all the kaarakas have a kind of causal role in the accomplishment of an action, this case indicates that the cause proper is conceived as the most efficient means (saadhakatama). Sometimes the cause is also declined in the apaadaana or ablative, indicating that from which the action comes. The effect is the result (phala) that is understood to occur in the direct object (karman).

The `Saivas subsume the cause-effect relation within their idealistic agential syntax. Through His recognitive synthesis, the omnipotent agent `Siva/the Self is the substratum of the overarching verbal process that contains the subordinate operations of all the other kaarakas, and it is He who instigates these operations. He moves and, as underlying substance, connects through time what is ordinarily considered the cause with the effect in the object As such it is He who is the real or essential cause. Utpaladeva puts this densely, referring to `Siva's action as His Action `Sakti:

That [Action] `Sakti is the existence [sattaa] of both [cause and effect, which are at different moments each] existent [sat] and nonexistent [asat]. That [Action `Sakti] does not belong to what is insentient. Therefore the essential nature of the cause-effect [kaaryakaara.na] relation is the agent-direct object [kart.rkarma] relation. [35]

In the following passage of his commentary, Abhinavagupta more fully explains the Lord's instigation and synthesis of the different moments through His Action `Sakti, using the example of germination:

[At the time of the seed, regardless of whether or not] the sprout [is conceived to be either] existent [Implicitly within the seed] or nonexistent, [36] the insentient seed does not have the capacity to make it have such existence that it is experienced [i.e., as developed into the sprout]. In the sprout being born from the seed, there is no power of the sprout because it does not yet exist. In the sprout being born, how can there be the power of the seed, since it is other than the sprout? Since this is so, it is concluded that the effect [kaarya] is nothing but the direct object [karma] being manifested through the Action `Sakti. This [fact] is evinced in the gerundive suffix [used to derive the word for effect, kaarya, from the verb for action, k.r]. [37] The effect is caused to be produced by Him. [38] [He is the] agent by reason of exercising the capacity for this [production of the effect]. Therefore the cause [kaara.na] rests in the agent [kart.r], who is consciousness. [39]

Within the perspective of this idealism, furthermore, the apparently unconscious process of a sprout arising from a seed is really not different

p. 600

from the manifestly conscious process of a porter making a pot. The Lord manifest from Himself the seed, along with the additional necessities such as earth and water -- and then the sprout. [40] Also, it is ultimately the Lord Himself, and not the potter qua limited individual, who makes the pot. The Lord manifests the potter, the tools and materials, and their activities. Abhinava places the process in the monistic idealistic perspective with the analogy of the reflection in a mirror of a potter making a pot. [41]

The `Saivas elaborate some of the same basic syntactic considerations to produce an interesting refutation of the Buddhist understanding of causation as "dependent origination." According to this conception, causality is a mere regularity of succession between evanescent entities, without any continuous or substantial "connection" between them. Such causality may be described in the manner "When there is this, then there is this" (asmin satiidam asti). Now, this expression in Sanskrit uses the grammatical construction that in English is called the "locative absolute."

The `Saivas interpret the regular priority and posteriority expressed with the locative construction as a sort of expectation (apek.saa) between the moments. They contend that such an expectation could not exist between discrete entities that in themselves lack recognitive synthesis (anusa.mdhaana). [42] Here is another terse statement by Utpaladeva:

The cause-effect relation [is formulated by the Buddhists as dependent origination expressed] "When there is this, then there is this." This [relation] cannot belong to insentient things, which are devoid of expectation [apek.saa]. For the referents of the seventh case cannot be supposed to be those [insentient things] which are situated only in themselves and are devoid of recognitive synthesis [anusa.mdhaana], whether they are existent [sat] or nonexistent [asat]. [43]

In his commentary, Abhinava again invokes the `Saiva interpretations of the syntactic principles regarding the agent as free and as substratum uniting the constituent processes of the kaarakas in the overarching process. It is He who unites what is expressed in the locative with what is expressed in the main clause:

The meaning of the seventh [locative] or other declension is classified as the relation between the kriyaa and karaka. It is this [relation between the kriyaa and kaaraka which is the only] regular succession [44] between things. There is no other dry [45] [form of regular succession]. That [regular succession] is possible if the pair of things [which are related] rest on [Him] who is free [svatantra] and has the nature of consciousness, and not otherwise... [The Buddhists have advocated dependent origination as a regular succession of cause and effect having such typical expressions using the locative as] "When there is the seed, there is the sprout" and "When there is fire, there is smoke." All this

p. 601

regular succession is possible only when there is rest [of the things occurring in succession] on the subject [pramaat.r] who is free [svatantra] and has the nature of consciousness. And not otherwise. [46]

For the `Saivas there is no unrelated component of the sentence. All syntax is related through the agent.

We see that in the discussions of causality the agent's ability, as substratum and instigator, to unite the processes of cause and effect is crucial. It will be useful for us briefly to consider the Pratyabhij~naa thinkers' evaluations of the Saa.mkhya and Advaita Vedaanta expressions of satkaaryavaada, the view that the effect preexists in the cause. According to the Saa.mkhya, the effect is a transformation (pari.naama) of a continuous underlying material cause (pradhaana, prak.rti). The `Saivas applaud the Saa.mkhya for asserting the continuity between cause and effect. [47] However they disagree with the Saa.mkhya understanding of the underlying material cause as insentient matter. Here n an excerpt from Abhinavagupta's argument that the different moments of the action of transformation can only be synthesized and instigated by a conscious agent:

Action is explained to be a differentiation between forms belonging to a unitary essential nature, which is unitary due to the force of recognition [pratyabhij~naa]. Since [action] is accompanied by temporal sequence, these mutually differentiated forms do not appear simultaneously... So, because the material cause [pradhaana] is characterized by a particular action, it has agency. It does not have a mere dry causality... [48] The material cause [in its ultimate nature as source of the universe, viewed by the Pratyabhij~naa the agent `Siva] abandons one form, establishes another distinct from it, and becomes intent [prahvataa] upon a third... [49]

The Pratyabhij~naa syntax articulates a distinctive satkaaryavaada of emanating Consciousness. [50]

Advaita Vedanta also advocates the satkaaryavaada, and maintains that the underlying material cause is the Self/Consciousness rather than insentient matter. However, it claims that the effect that is the universe is an illusion (maayaa), a false supposition or "projection" (vivarta) on this Self. The `Saivas adduce many of the well-known (nonsyntactic,) arguments against the Advaita Vedaantin understanding of maayaa -- the impossibility of denying a world that all experience, the difficulty of specifying the subject of cosmic illusion, and the illogicality of the Advaita view of the ontological inexplicability (anirvacacaniiyatva) of illusion. [51] On the basis of such considerations, the `Saivas claim that the unity of the multiplistic universe car only be explained with their narrative syntax of "cre-ation" by the conscious agent `Siva. [52] To quote again an excerpt from Abhinavagupta:

Therefore, even if it is admitted [with Advaita Vedaanta] that the real Consciousness is a unity, there will not be possible [by their view] creation/action

p. 602

[kriyaa], which involves entrance into diverse forms. [This creation] has agency [kart.rtva] as its definitive characteristic. However, all this is possible if there is [in accordance with our view, agential] freedom [svaatantrya], which has the essential nature of recognitive judgment (paraamar`sa). For recognitive judgment is intention [icchaa], which has the nature of the desire to create/act [cikiir`saa]. Everything to be created exists within that [intention] in non-differentiation... Thus the Great Lord causes the Universe, which has the nature of awareness, to appear in diversity. This universe is real [but at the same time] is essentially nothing but His own Self, has the ultimate nature of awareness, and is a unity of unbroken awareness. Lordship [ai`svarya] is explained to be freedom [svaatantrya] having the character of doing this which is extremely difficult... Because the subject of discussion here is the cause-effect [kaaryakaara.na] relation, both the kart.r and the karman are mentioned. However, the other kaarakas also really follow within [anuprave`siini] the unitary agency [kart.rtva]. Otherwise, how could that [creation/action] be undifferentiated, when there is a differentiated multitude of kaarakas, such as the kara.na, etc.? [53]

The divergence from `Sa^nkara's grammar of the agent's bondage to action within the realm of illusion could not be greater.

The `Saiva Syntax as Ontological. The explanation of all things as the results of `Siva's/the Self's cre-ation underlines the fact that the `Saivas are articulating a theory of existence, a narrative ontology that reductionistically features the agent. Thus Abhinava explains in the conclusion to the discussion of causality:

[The expression] "The pot exists" has this [true] meaning: The Great Lord, who is awareness [prakaa`sa], desires to exist as the pot and assumes that existence through His freedom [svaatantrya]... [54]

To relate ontology and syntax even more explicitly, I quote two more statements by Utpaladeva and Abhinavagupta:
Being is the condition of one who becomes, that is, the agency of the act of becoming... [55]

Being is the agency of the act of becoming, that is, freedom [as is characteristic of an agent] regarding all actions. [56]

The Epistemic Syntax of Omnipotence. Now the `Saivas' ontological syntax as explained so far is parasitic on their idealistic epistemology of `Siva's emanating self-recognition. However, the category of ontology may be understood as in a certain way comprehending epistemology, inasmuch as knowledge of various sorts must be assigned some sort of existence. The `Saivas thus also situate epistemology within their narrative ontology. They explain cognition as action. As a statement of Somaananda cited by Abhinavagupta puts it, "At the lime of the cognition of a pot, there is the action 'He knows the pot.'" [57]

p. 603

Like other Indian philosophical schools, the Pratyabhij~naa makes frequent use of syntax to explain the structure of processes of human knowing. To begin with the epistemological triad (tripu.tii) of subject, means, and object of cognition: The subject of cognition (pramaat.r) is, of course, the agent (kart.r). The word for means of cognition, pramaa.na, is derived in the same way, and has the same syntactic function as the kaaraka usually used to express the cause, that is, kara.na. The pramaa.na is the most efficient means (saadhakatama) for the accomplishment of the action of cognizing. The object of cognition is the direct object (karman) of the action of cognizing. And another relevant category outside the triad, cognition itself (pramaa, pramiti), is the result (phala) of the process (vyaapaara) of cognizing the object of cognition.

The `Saivas' basic strategy is to reduce alt the other categories in essential nature to the process (vyaapara) of self-recognition internal to the subject/agent. The `Saivas claim that the other factors in the syntax of cognition only appear to function independently through a sort of contraction (sa.mkoca) of this self-recognition. [58]

Now, the relation between means of cognition (pramaa.na) and cognition (pramaa, pramiti) is a particular form of the relation between cause and effect, which we have already considered. The underlying process is perhaps even clearer here where the focus is explicitly on knowledge. In the following passage, Abhinavagupta places both means and cognition within the synthetic nexus of the agent's self-recognition:

Recognitive judgment [vimar`sa] has been established previously to be, in its essential nature, inwardly directed [as self-recognition]. Only when that [recognitive judgment] is contracted in the condition of [having apparently separate] objects, it is cognition as a result [phala]. [We may take the following as an illustration:] There is the awareness "I who am brave am victorious." One may analyze into a cause-effect relation the two conditions of bravery and victory, which are actually situated in a single [agent]. [This would be expressed] "Since I am brave, therefore I am victorious." Similarly [there is the analysis] that, since there is the perceptual awareness of blue, therefore there is the judgment 'This is blue.' Even though they really have a unitary nature, [through this artificial analysis, they are understood in] a cause-effect [hetu-phala] relation. [59]

What are normally regarded as cognitive means and effect are equally reduced to subjective conditions. [60]

In this vein, there are a large number of examples where the `Saivas reinterpret expressions of cognitive states in forms that uncover the syntactic role of the agent. Thus Abhinava glosses the sentence "The blue appears to me [mama niilam bhaati]" as "The blue is cognized by me [mayaa niilam j~naayate]." [61] Likewise Utpaladeva explains regarding the memory of an earlier experience:

p. 604

[An earlier experience] may be remembered as different [from the subject, as expressed in the form "That cognition of mine was thus." [The memory having this expression] is nothing but a grammatical analysis [vyaakara.na] [62] of the memory [expressed] "It was seen by me." [63]

In another interesting example, Abhinava refutes the apparent perception of a wife as an independent "other." He controverts the idea that the direct object (karman) is the locus of cognition as result (phala) of cognizing:

One may claim that cognition rests on the lotus face of one such as a wife [i.e., as an object of cognition], and is not produced within the Self. For there is the past passive participle affix as referring to the direct object [karman]. [We respond] that the one [claiming this] does not understand his own speech. For the direct object [karman] is what is to be obtained by means of the action [kriyaa] of the agent [kart.r]. Thus the action [kriyaa] of cognizing is established in the agent [kart.r]. [64]

This is the opposite of the direction of thinking observed by Gerow.

Sometimes the thinkers parse the underlying syntax of experiences with expressions that do not even mention a direct object. [65] Thus, Abhinava states:

Thus even [apparently separate objects such as] blue, etc., existing in a ... recognitive judgment [paraamar`sa] such as "This is blue," are established to be constituted by the Self, because they rest upon the root recognitive judgment [paraamar`sa] "I." Even "I cognize this to be blue" really amounts to "I am aware [prakaa`se]." [66]

Abhinava states that some believe that the expression that does not mention the object is the most proper one. [67]

Some Practical Implications
I must emphasize that this narrative syntax that we have considered is ritually axiomatic as well as doctrinal/mythical. In his Tantraaloka and Tantrasaara, Abhinavagupta uses syntactic theory in explaining the inner significance of preliminary ceremonies of the tantric ritual: nyaasa, the divinizing projection of mantric syllables onto the body and other parts of the ritual; and purifications [68] with the sacrificial vase. Abhinava identifies various components of the ritual, such as the location, ritual implements, object of sacrifice (e.g., flowers), and oblations, with the grammatical cases of adhikara.na, apaadaana, kara.na, karman, and so forth. He explains the overarching ritual process as the aspirant's identification with `Siva as agent of all the cases. It is this identification that is facilitated by nyaasa and purifications with the vase. The experience of this identification is cultivated in the rites of worship (arcaa, puujaa) with the intent of carrying it over into the rest of one's life. To quote Abhinava:

p. 605

All things are without separation from `Siva, who has supernatural power [siddhi], or is perfectly complete [puur.na]. Here, through [the kriyaa which is] worship, the kaarakaas [are realized to be without separation from `Siva]. In the case of worship, the entire collection of kaarakaas is revealed to be undifferentiated from `Siva. This also obtains in [ordinary worldly actions such as] walking, etc. A horse which has been freed from carrying loads and roaming in an enclosed area, does not betray its training [`siiksaa] even when it has entered in battle. [69] Similarly, one who has identified the kaarakaas with `Siva -- through training [abhyaasa] in the kriyaa of worship -- eliminates the duality of kaarakaas [and the aspirant's true Self as `Siva] even [when engaged in ordinary worldly actions such as] walking of standing. So, to the one occupied with this training [abhyaasa] of unity, this universe appears immediately [akramaat, 'without temporal succession'] and powerfully [ha.thaat], like a woman dancing in a frenzy [k.sobha] of the perfectly complete state of `Siva... [70]

Thus, since the kriyaa of worship bestows this identification [of everything with `Siva], its ultimate nature is the unity of all the kaarakaas with Him. By the ceremonies of purification of the place, there is [realized] the identity [with `Siva] of the seat of the sacrificer. By means of the divinizing-meditative-projection [nyaasa], the sacrificer, object of sacrifice [i.e. the direct object or karman], place of that [sacrifice -- the locative], kara.na [instrument], aadaana [ablative] and sampradaa [dative] attain [adhi`serate] identity with the state of `Siva. [71]

The Pratyabhij~naa system itself participates fully in this ritual syntax. As I have explained, the epistemological disclosure of `Siva's self-recognition as the inner reality of all experiences is supposed to lead us to participate in the same. The whole philosophical inquiry is equally explained as a meditation on agency. As Utpaladeva states:

That one succeeds [siddhyati], who places his feet on this [the path of the Pratyabhij~naa `saastra], and, contemplating that the status of the Agent [kart.r] of the world belongs to himself, submerges himself incessantly in the state of `Siva. [72]

Conclusion
While this essay can hardly have been exhaustive, I hope, by thematizing aspects of philosophical kriyaa-kaaraka theory as articulating grammars of motives, to have pointed to a fruitful direction of inquiry for both Indology and general comparative studies. A great deal more research can be done on the particular question taken up here of Sanskritic theorization on the relation of the syntactic agent to action, its objects, and results. The ramifications of the `Saiva view in myth and ritual can be further explored. More can be done on the intricate theories of Puurva Miimaa.msaa and Advaita Vedaanta, as well of the other schools -- Nyaaya, Jainism, Buddhism, and so forth. It is also possible that some schools (the Vai.s.navas?) may develop a more "moderate" syntax than either of the polarized orthodox and tantric theories I have presented. Again, more work can be done on how the various other kaarakas -- for

p. 606

example, location (adhikara.na), instrument (kara.na), purpose (sampradaana), and so forth -- have been said to relate to the action.

Given contemporary academic preoccupations with action, language, and narrative, I believe that Kenneth Burke's pioneering study of competing grammars of motives deserves still greater attention than it has already received. Used flexibly, Burke's own pentad of categories still has heuristic value -- particularly in attempting to characterize implicit narrative grammars. However, I believe it would often be of greater interest to examine various cultures' indigenous, explicitly grammatical, reflections on action syntax and its relation to religious myth and ritual -- as I have done with Sanskritic discussions. Among classic intellectual traditions, the Sanskritic may have been exceptionally concerned with linguistic speculation. Nevertheless, I believe that "grammars of motives" could be developed into a useful comparative problematic to order some of the vast and growing body of scholarship on theories of language and action in other cultures -- Greece, Islam, China, Judaism, Christianity, and so forth.

The dialogical and critical confrontation with foreign worldviews/myths and praxes/rituals in the comparative philosophy of religion should also be facilitated by an understanding of how their grammatical reflections rationalize understandings of motivation. The syntax of omnipotence elaborated by Utpaladeva and Abhinavagupta further elucidates the distinctiveness of their religious philosophy. It clearly diverges from the subordination of the human to God in the mainstream Judeo-Christian narratives, although it may have analogues in more radical appropriations of Neoplatonism [73] and in nineteenth- and twentieth-century romanticism and Neo-Hegelianism. [74]

The `Saiva syntax likewise diverges from the modem "metanarratives" of individual agents who scientifically understand a world of independent objects [75] in terms of causal relationships and exploit them through technology, and who organize their societies with the same rationality. Scientific and technological thought have been widely analyzed in proto-syntactical terms as "objectivist" and "instrumentalist."

"Postmodern" theorists, in turning the critical resources of modernity upon itself, have also attempted to deconstruct its correlative understandings of the individual rational agent. One could speak of a "fragmented" agent of the paralogical postmodern narratives. However, the "fissures" posited in the agent by thinkers such as Derrida, Foucault, and Lyotard are in a sense "hyper-modern": (crypto-)objective cultural-historical context and linguistic structure or "post-structure," and instrumentalities of power and legitimation, sexual drives, and so forth.

Carrying further such speculation about implicit and explicit Western grammars of motives is beyond the scope of this essay, and it is not necessary for the reader to agree with all of it. I only wish to underline

p. 607

my suggestion that understanding the Pratyabhij~naa thinkers' syntactic theories makes the intellectual and spiritual alternative they pose to contemporary thought stand in sharper relief.

NOTES
This essay develops one of the themes in my "Argument and the Recognition of `Siva: The Philosophical Theology of Utpaladeva and Abhinavagupta" (Ph.D. diss., University of Chicago, 1992), being prepared for publication. An earlier version of this essay was presented as a paper at the 24th Annual Conference on South Asia, Madison, Wisconsin, 1995. At various points since the early 1980s, I have benefited from studying and discussing the `Saiva and related grammatical theories of action with several other scholars, including Navjivan Rastogi, Hemendra Nath Chakravarty, Saudmini Deshmukh, Srinarayan Mishra, and Edwin Gerow.

The following abbreviations are used in this article:

BIPV Bhaaskarii by Bhaaskaraka.n.tha, commentary on IPV.

IPK II`svarapratyabhij~naakaarikaa by Utpaladeva.

IPKV II`svarapratyabhij~naakaarikaav.rtti by Utpaladeva, commentary on IPK.

IPV II`svarapratyabhij~naavimar`sinii by Abhinavagupta, commentary on IPK.

IPVV II`svarapratyabhij~naaviv.rttivimar`sinii by Abhinavagupta, commentary on Utpaladeva's II`svarapratyabhij~naaviv.rti.

TA Tantraaloka by Abhinavagupta.

TAV Tantraalokaviveka by Jayaratha, commentary on TA.

TS Tantrasaara by Abhinavagupta.

1. The chief Pratyabhij~naa texts interpreted here are Utpaladeva, II`svarapratyabhij~naakaarikaa (IPK) and Abhinavagupta, II`svarapratyabhij~naavimar`sinii (IPV), for which I will use the edition II`svarapratyabhij~naavimar`sinii of Abhinavagupta, Doctrine of Divine Recognition: Sanskrit text with Bhaaskarii, 2 vols., ed. K. A. Subramania Iyer and K. C. Pandey (reprint, Delhi: Motilal Banarsidass, 1986). I will sometimes refer to the eighteenth-century commentary on the IPV, Bhaaskarii, by Bhaaskaraka.n.tha (BIPV). Other Pratyabhij~naa writings within the scope of this essay are: Utpaladeva, Siddhitrayii and the II`svarapratyabhij~naakaarikaav.rtti, ed. Madhusudan Kaul Shastri, Kashmir Series of Texts and Studies, no. 34 (Srinagar Kashmir Pratap

p. 608

Steam Press, 1921), and The II`svarapratyabhij~naaviv.rtivimar`sinii by Abhinavagupta, 3 vols., ed. Madhusudan Kaul Shastri, Kashmir Series of Texts and Studies (reprint, Delhi: Akay Book Corporation, 1987). The II`svarapratyabhij~naakaarikaav.rttti and II`svarapratyabhij~naaviv.rtivimar`sinii will henceforth be referred to as IPKV and IPVV, respectively. For Abhinavagupta's writings on symbolic and ritual theology I will use, first, The Tantraaloka or Abhinavagupta with the Commentary of Jayaratha, 8 vols., ed. Madhusudan Kaul Shastri and Mukunda Rama Shastri, Kashmir Series of Texts and Studies, republication, ed. R. C. Dwivedi and Navjivan Rastogi (Delhi: Motilal Banarsidass, 1987). Abhinava's book will henceforth be referred to as TA, and Jayaratha's commentary, Tantraalokaviveka, will be referred to as TAV. I will also refer to The Tantrasaara of Abhinavagupta, ed. Mukunda Ram Sastri, Kashmir Series of Texts and Studies, no. 17 (reprint, Delhi: Bani Prakashan, 1982). This text will be abbreviated TS.

2. I investigate the thinkers' understanding of the Pratyabhij~naa as both philosophical proof and spiritual exercise in "Tantric Argument: The Transfiguration of Philosophical Discourse in the Pratyabhij~naa System of Utpaladeva and Abhinavagupta," Philosophy East and West 46(2) (1996): 165-204. In "`Siva's Self-Recognition and the Problem of Interpretation," Philosophy East and West 48 (2) (1998): 197-231, I more fully examine and defend the `Saiva philosophy of recognition in its analogies with the Western philosophical theology of logos.

3. `Siva's epistemic-recognitive and active aspects are often described, respectively, as His Knowledge (j~naana) and Action (kriyaa) `Saktis.

4. See Vaakyapadiiya of Bhart.rhari, kaa.n.da 3, ed. K. A. Subramania Iyer (Pune: Deccan College, 1963-1973), pt. 2, 3.8, 1-40, particularly 3.8.1.1; 3.8.27, 22; 3.8.35, 26.

5. The ritualistic philosophical school of Puurva Miimaa.msaa elaborated at great length on the injunctive nature of language, describing the dispositional and concrete 'bringing about' (bhaavanaa) of actions conveyed by the inflected verbs in sentences.

6. An influential earlier study is Stephen Crites, "The Narrative Quality of Experience," Journal of the American Academy of Religion 39 (September 1971): 291-311.

7. See Jean-Francois Lyotard, The Postmodern Condition: A Report on Knowledge, trans. Geoff Bennington and Brian Massumi (Minneapolis: University of Minnesota Press, 1989), and Paul Ricoeur, Time and Narrative, 3 vols., trans. Kathleen McLaughlin/Blamey and David Pellauer (Chicago: University of Chicago Press,1984-1988).

p. 609

8. Most of the basic ideas on karakas explained here may be found in Pata~njali's Vyaakara.na-Mahaabhaa.sya: Kaarakaahnika (P.1.4.23-1.4.55), ed. with introd., trans., and notes by S. D. Joshi and I. A. F. Roodbergen (Pune: University of Poona, 1975). Also see the Kriyaasamudde`sa, in Vaakyapadiiya, kaa.n.da 3, pt. 2, 3.8, 1-40, and the Saadhanasamudde`sa, in Vaakyapadiiya, kaa.n.da 3, pt. 1, 3.7, 230-370. Some of the more important secondary studies, representing a variety of points of view, are: K. A. Subramania Iyer, Bhart.rhari: A Study of the Vaakyapadiiya in the Light of the Ancient Commentaries (Pune: Deccan College, 1969), pp. 283-344; George Cardona, "Paa.nini's Syntactic Categories," Journal of the Oriental Institute, Baroda 16 (1967): 202-215; idem, "Paa.nini's Kaarakas; Agency, Animation and Identity," Journal of Indian Philosophy 2 (1974); 231-306. Sergiu Al-George, "Le sujet gramatical chez Paa.nini," Studia et Acta Orientalia Bucaresti 1 (1957): 39-47; idem, "The Extra-Linguistic Origin of Paa.nini's Syntactic Categories and Their Linguistic Accuracy," Journal of the Oriental Institute, Baroda 18 (1968): 1-7; Rosane Rocher, "'Agent' et 'objet' chez Paa.nini," Journal of the American Oriental Society 84 (1964): 44-54; Bimal Krishna Matilal, "Indian Theorists on the Nature of the Sentence (Vaakya)," Foundations of Language 2 (1966): 377-393; idem, "The Doctrine of Kara.na in Grammar-Logic, "Journal of the Ganganatha Jha Research Institute 17 (1960): 63-69; idem, "The Kaaraka Theory," chap. in Bimal Krishna Matilal, The Word and the World: India's Contribution to the Study of Language (Oxford: Oxford University Press, 1990), pp. 40-48; Frits Staal, "Syntactic and Semantic Relations in Paa.nini," Foundations of Language 5 (1969): 83-117; Edwin Gerow, "What is Karma (Ki.m Karmeti): An Exercise in Philosophical Semantics," Indologica Taurinensia 10 (1982): 67-116.

9. Kenneth Burke, A Grammar of Motives (Berkeley: University of California Press. 1962).

10. I prefer the term "instrument" rather than Burke's "agency" in order to avoid confusion with the category of the "agent." The term "instrument" is also more commonly used in grammar and philosophy.

11. After I had developed my ideas on the analogy between Burke and aspects of kaaraka theory, I discovered the following: the use of terminology for dramatic roles is applied to understanding the sentence by D. Terence Langendoen. The roles are actually used to describe semantic relationships that are turned into syntactic relationships through transformational rules (D. Terence Langendoen, Essentials of English Grammar [New York: Holt, Rinehart and Winston, 1970], pp. 61 ff.). This approach in turn is suggested as a way

p. 610

of understanding kaarakas by S. D. Joshi and I. A. F. Roodbergen, in the introduction to Pata~njali's Vyaakara.na-Mahaabhaa.sya: Kaarakaahnika, pp. iii-iv.

12. Though he attempts to derive his categories from Aristotle, was Burke actually reflecting on the Indo-European case structures?

13. According to Burke, Marxism also features the scene. However, it attributes to the scene a degree of its own conscious agent-hood. While it is illuminating to view such approaches as emphasizing the "scene," it seems to me that Burke should have included a separate category of the "object." Scientific or empirical research methodologies in the "hard" sciences as well as culture studies have of course been preoccupied with "objective" knowledge. This does not contradict Burke's analysis, as the approaches he identifies emphasize the "scene" as what is known objectively.

14. For example, Burke has been used to analyze different sociological and legal approaches to the problem of drunk driving. Some explanations emphasize the category "drunk driving" as an act carrying with it liability, and others emphasize the "drunk driver," an agent acting with intention and at fault. See Joseph R. Gusfield, The Culture of Public Problems: Drinking-Driving and the Symbolic Order (Chicago: University of Chic ago Press, 1981).
A study that, although it only mentions Burke, laid the ground- work for later uses of the grammar of motives is C. Wright Mills, "Situated Actions and Vocabularies of Motive," American Sociological Review 5 (1940): 904-913. For assessment of Burke's grammar of motives along with other aspects of his theory of symbolic action, including reviews of applications of his ideas, see Joseph R. Gusfield, introduction to Kenneth Burke on Symbols and Society, ed. Joseph R. Gusfield (Chicago: University of Chicago Press, 1989), pp. 1-49: Sonja K. Foss, Karen A. Foss, and Robert Trapp, Contemporary Perspectives on Rhetoric (Prospect Heights, Illinois: Waveland Press, 1985), pp. 153-188, 291-304 (with a good bibliography); Herbert W. Simons and Trevor Melia, eds., The Legacy of Kenneth Burke (Madison: University of Wisconsin Press, 1989); and George K. Zollschan, "Reasons for Conduct and the Conduct of Reason: The Eightfold Route to Motivational Ascription," in Social Change: Explorations, Diagnoses and Conjectures, ed. George K. Zollschan and Walter Hirsch (New York: John Wiley, 1976), pp. 270-317.

15. It should be understood that this study, in purposes and method, is completely different from earlier efforts to characterize civilizations on the basis of their languages, as for example in Hajime Naka-

p. 611

mura, Ways of Thinking of Eastern Peoples: India, China, Tibet, Japan (Honolulu: University of Hawai'i Press, 1971). I view such projects of generalization as highly problematic. I am concerned here with competing philosophical theories of Sanskrit syntax.
I must also distinguish my approach from Frits Staal's syntactical analyses of Vedic ritual. See his "Ritual Syntax," in Sanskrit and Indian Studies: Essays in Honour of Daniel H. H. Ingalls, ed. M. Nagatomi et al. (Dordrecht: D. Reidel. 1980), pp. 119-143. Staal does not examine Sanskritic understandings of the motivation of action. Moreover, whereas I am concerned with theories of mythico-ritual syntax, he abstracts the syntax of ritual from semantics on the basis of his well-known conception of the "meaninglessness of ritual." He explains the highly convoluted synchronic structure of ritual in terms of Chomskian categories such as "phrase structure," "transformational," and "self-embedding" rules. (It would be interesting to relate such analysis to the indigenous understandings considered here.) He also claims that the meaningless ritualistic syntax preceded, and still affects, the structures of actual languages. Steal further elaborates his ideas in his Rules without Meaning: Ritual, Mantras and the Human Sciences, Toronto Studies in Religion, vol. 4 (New York: Peter Lang, 1989).

16. Nakamura continues these discussions. See my remark in the previous note.

17. It should be understood that the critical remarks made here and below are not meant to disparage Gerow's innovative study. They endeavor only to advance further the inquiry into these subjects.

18. Alexis Sanderson, "Purity and Power among the Brahmans of Kashmir," in The Category of the Person: Anthropology, Philosophy, History, ed. Michael Carrithers, Steven Collins, and Steven Lukes (Cambridge: Cambridge University Press, 1985), pp. 190-216.

19. Sanderson, "Purity," p. 196.

20. Ibid., pp. 196-197.

21. The explanations and quotations of `Sa^nkara are my own. I note that while the `Saivas indicate familiarity with Advaita Vedanta, they never quote `Sa^nkara. It does not matter here whether they knew his works, or those of other representatives of Advaita, for I only wish to show how they subvert a common mode of thinking.

22. Baadaraaya.na, `Sa^nkara, Vaacaspatimi`sra, Amalaanandasarasvatii, and Appayyadiiksita, The Brahmasuutra `Saa^nkara Bhaa.sya: With the Commentaries Bhaamatii, Kalpataru and Parimala, 2 vols., ed. K. L. Joshi (Delhi: Parimal Publications, 1987), 1.1.4, 1:130.

p. 612

23. Ibid., 4.1.13, 2:954. Cf. 3.4.16, 2:876.

24. Dumont emphasizes that this individuality is achieved through the renunciation of action in the sphere of worldly caste and other social relationships. I agree with his sociological observations about the reciprocal and complementary relationships that nevertheless exist between renouncers and the worldly, and about the cultural creativity of renunciant traditions. See Louis Dumont, "World Renunciation in Indian Religions," appendix in Homo Hierarchicus: The Caste System and Its Implications, trans. Mark Sainsbury, Louis Dumont, and Basia Gulati (Chicago: University of Chicago Press, 1980), pp. 267-286.

25. Sanderson, "Purity," pp 197-198.

26. The other two malas are those of limited individuality, the aa.nava mala, and of maayaa, the maayiiya mala.

27. See The Spandakaarikaas of Vasugupta with the Nir.naya by Ksemaraaja, ed. and trans. Madhusudan Kaul Shastri, Kashmir Series of Texts and Studies, no. 42 (Srinagar: Kashmir Pratap Steam Press, 1925), 3.16, 71-72; Abhinavagupta, `Srii Maaliniivijaya Vaarttikam, ed. Madhusudan Kaul Shastri, Kashmir Series of Texts and Studies, no. 31 (Srinagar: Kashmir Pratap Steam Press, 1921), 1.313-315, 30-31; TA and TAV, 13.266-268a, 5:2363-2365; IPK and IPV 3.2.4-10, 2:248-256.

28. In this respect, there are analogies between the ordinary agent and the agent of the causative conjugation.

29. The `Saivas commonly describe the unitary, eternal Lord's emanation of temporal diversity as a kind of agitation or vibration. This is the understanding articulated in the doctrine of `Siva's spanda. Abhinava emphasizes the connection with this doctrine here in describing the movements of the one cooking as spandas.

30. IPV 2.1.8, 2:24-25.

31. Utpaladeva, Sambandhasiddhi, in Siddhitrayii and the II`svarapratyabhij~naakaarikaav.rtti, 9.

32. IPVV 2.4.5, 3:189-190.

33. The classic contemporary reading of Indian philosophical positions in terms of the issue of causality is Karl H. Potter, Presuppositions of India's Philosophies (Englewood Cliffs, New Jersey: Prentice-Hall, 1963). Unfortunately, I cannot here get into all of the technicalities of the `Saiva theorization on causality. I can only point out some of the key ways in which they structure the subject with their understanding of syntax. Causality is the focus of the whole of IPK and IPV 2.4, 2:150-209.

p. 613

34. For elucidation of the correspondence of kaara.na and kara.na, see Matilal, "The Doctrine of Kara.na in Grammar-Logic."

35. IPK 2.4.2, 2:153. Abhinavagupta is not sure exactly how to take Utpaladeva's reference to the existent and the nonexistent. In my translation, I have taken the assertion as indicating the fact that the successive moments of an action, and thus the constituent "processes" of cause and effect, become existent and then nonexistent. The action itself is the existence/Being (sattaa) that unifies these moments. This interpretation seems to be supported by the following statement of Bhart.rhari: "Therefore that which is nonexistent [asat] has disappeared. And that which is existent [sat] is experienced. [Action] is understood as the unitary essential nature [aatman] of the existent, [sat] and the nonexistent [asat]" (Vaakyapadiiya, kaa.n.da 3, pt. 2, 3.6.19. 18). Another possibility is that Utpaladeva is asserting the inadequacy of both the Saa.mkhya and Nyaaya theories of causality. According to the former, the effect preexists in the cause, and arises as a sort of transformation of it. (This view will be discussed further below.) According to the latter, the effect is non-existent in the cause, and "emerges" as something new. For Abhinava's discussion of the alternative interpretations, see IPV 2.4.2, 2:154, and IPVV 2.4.2, 3:186. Also see BIPV 2.4.2, 2:154. In any case the basic point expressed by Utpaladeva regarding the role of the agent remains the same.

36. Abhinava is referring to the interpretation of the verse in terms of the Saa.mkhya and Nyaaya theories, as explained in the previous footnote.

37. As a gerundive, kaarya, 'effect,' literally means "that which is to be done."

38. I have used this awkward phrase to indicate that the verb kaaryate is conjugated in the causative.

39. IPV 2.4.2, 2:153-154. Utpaladeva explains the Lord's creation of things through His agential intention (icchaa) at IPK 2.4.1, 2:152. Abhinavagupta epitomizes this verse as also asserting that the relation of cause and effect is nothing but the relation of agent and direct object, at IPV 2.4, introduction, 2:151. Cf. the explanation more focused on the idealistic character of causality in IPK 2.4.4, 2:153.

40. This is spelled out at IPVV 2.4.8, 3:192.

41. See IPV 1.8.9. 1:411; IPV 2.4.4, 2:157-159; IPV 2.4.9, 2:169-170.

p. 614

42. See IPV 2.4.14, 2:188; IPVV 2.4.14, 3:218.

43. IPK 2.4.14-15, 2:187-190. See the discussion of the qualifications "existent" and "nonexistent" above.

44. This term, samanvaya, also has the significance of syntactic coordination.

45. It is notable that Abhinava refers to other ostensible relations as "dry" (`su.ska). Bhaaskaraka.n.tha glosses this term as ni.hsaara, 'without essence', and lists as examples "conjunction" (sa.myoga), and so forth. It is clear that Abhinava's "moist" relationship is that which involves action by the agent.

46. IPV 2 4.16, 2:192-193.

47. Abhinavagupta describes the Saa.mkhyas as "having long-ranging insight" and "resorting to recognition (pratyabhij~naa)" (IPV 2.4.18, 2:194-195). As asserted above, Utpaladeva may have briefly expressed disagreement with the Saa.mkhya account in IPK 2.4.2, 2:153.

48. Cf. the mention of "dry" relations above.

49. IPV 2.4.18, 2:196-197. The translation of the last sentence is rough, but the idea is dear for our purposes. See the discussion of the Saa.mkhya view throughout IPK and IPV 2.4.17-19, 2:193-200. The thinkers further argue that only a conscious agent can logically arbitrate unity and multiplicity at IPK and IPV 2.4.19, 2:197-200.

50. The Nyaaya also maintains that an agent must be conscious. See Cardona, "Paa.nini's Kaarakas: Agency, Animation and Identity." However, they hold the causal theory of asatkaaryavaada, and would never admit that all creation materially emanated from God as sole agent.

51. The Advaita Vedaantins maintain that illusion is not strictly nonexistent, because it is experienced. However, it is not existent, because it disappears when one attains self-realization.

52. `Sa^nkara's theistic sagu.na brahman, as ultimately illusory, does not address the `Saivas' considerations on the requirement for a genuine creator.

53. IPV 2.4.20, 2:203-205. See the whole discussion through IPK and IPV 2.4.20, 2:201-206. There are scattered arguments against Advaita Vedaanta throughout the `Saiva literature.

54. IPV 2.4.21, 2:207.

p. 615

55. sattaa bhavattaa bhavanakart.rtaa (IPKV 1.5.14. 19).

56. sattaa ca bhavanakart.rtaa sarvakriyaasu svaatantryam (IPV 1.5.14, 1:258-259).

57. The `Sivad.r.s.ti of `Srisomaanandanaatha with the Vritti by Utpaladeva, ed. Madhusudan Kaul Shastri, Kashmir Series of Texts and Studies, no. 54 (Pune: Aryabhushan Press, 1934). 1.24, 19. Abhinavagupta cites this statement in the course of explaining the ultimately internal nature of the Action `Sakti (IPV 1.1.4, 1:74). For more on the `Saivas' understanding of the reciprocal encompassment of knowledge and action see my "Argument and the Recognition of `Siva," pp. 194-196.

58. I note that the Pratyabhij~naa syntax of the tripu.tii also rationalizes features of Krama tantric contemplations of circles of `Saktis -- of great importance to Abhinavagupta. In these, the pramaat.r, pramaa.na, and prameya are identified, respectively, with fire, the sun, and the moon. The latter are in various ways contemplated as "absorbed" in the former. See TA 4.122 ff., 3:740 ff.; TA 5.19 ff 3:945 ff.

59. IPV 2.3.1-2, 2:74-75.

60. I note that in this explanation the `Saivas are reformulating and subverting within their own syntax the Buddhist understanding of the means or cognition. The Buddhists also identify the means (pramaa.na) with the result of cognition itself (pramaa), with the purpose of claiming artificial the notion of an enduring, agent-located process (vyaapaara) connecting them. At IPV 2.3.1-2, 2:75, Abhinavagupta cites Pramaa.navaarttika 2.308 on the means-result identity. This Buddhist conception of the identity of cognitive means and result may be understood as a non-agential formulation of esse est percipi. It also instantiates in a particular epistemological context the syntactic trajectory against agency exposited by Gerow. While accepting the Buddhist idealism, the `Saivas place it within the processual nexus of an enduring cognitive agent. See the discussion throughout IPV 2.3.1-2, 2:73-76. Also see the denial of the phenomenal objectivity of the Buddhist logicians' 'unique particulars' (svalak.sa.nas) at IPV 2.3, introduction, 2:67. For a discussion of Buddhist philosophy as denying the essential features of action-syntax, which as explained here is articulated with reference to this syntax, see Nandita Bandyopadhyay, "The Buddhist Theory of Relation between Pramaa and Pramaa.na," Journal of Indian Philosophy 7 (1979): 43-78.

61. IPV 1.1.3, 1:63.

62. The grammatical analysis of a unitary discourse into sentences, and sentences into parts of speech, roots, suffixes, declensional endings,

p. 616

and so forth, is frequently said to be artificial and heuristic. The idea here is that it is artificial to reify cognition as a separate entity rather than as a process integral to the subject.

63. IPK 1.4.6, 1:178. Also see on this IPV 1.4.6, 1:177-182.

64. IPV 1.4.6, 1:180. Bhaaskaraka.n.tha here cites the familiar principle of Paa.nini's A.s.taadhyaayii 1.4.49 (BIPV 1.4.6, 1:180).

65. In the sentence in Utpaladeva's verse, "It was seen by me," the direct object is not given a pronoun but is expressed, according to normal Sanskrit usage, as the referent of the participle.

66. IPV 1.5.17. 1:279. The thinkers engage in an involved grammatical discussion of epistemology throughout IPK and IPV 1.5.17,1:273-280, in the course of treating the apparent objectification inherent in any designations of one's self-recognition as "Lord," "Self," "`Siva," "subject," and so forth.

67. IPV 1.4.6, 1:182. This is explained in terms of an alternative gloss of Utpala's verse. Utpaladeva uses the objectless syntax to explain the awareness in memory of an original experience at IPK 1.4.4, 1:167. See Abhinavagupta's grammatical explanation at IPV 1.4.4, 1:169-170.

68. It must be understood that tantric "purity" (`suddhi) pertains to the realization of unity as described here, and is transgressive of the ostensibly objective (and hierarchically defined) purity of Hindu orthodoxy. Abhinava himself is very careful to make the distinction. See TS 4. 31; TA. 118-119, 3:737; and TA 4.218-220, 3:858-859.

69. My interpretation of the analogy of the horse is influenced by that of Luce delle Sacre Scritture (Tantraaloka), trans. Raniero Gnoli (Torino: Unione Tipografico-Editrice Torinese, 1980), p. 454.

70. TA 15.147-151, 6:2516-2518.

71. TA 15.157-158, 6:2520. Also see TS 13, 135 ff.

72. IPK 4.1.16, 2:309.

73. Thus Meister Eckhart talked of an experience of identity with the Godhead, in which the aspirant could create the universe with God.

74. As even mainstream monotheistic theology has been marginalized, absolute idealism has virtually disappeared, except for isolated revisionists such as J. N. Findlay.

75. In my broad syntactic observations here, I am including the category of "object." See note 13 above.

p. 617

BIBLIOGRAPHY
Abhinavagupta. The II`svarapratyabhij~naaviv.rtivimar`sinii by Abhinavagupta. 3 vols. Edited by Madhusudan Kaul Shastri. Kashmir Series of Texts and Studies. Reprint, Delhi: Akay Book Corporation. 1987.

___. Luce delle Sacre Scritture (Tantraaloka). Translated into Italian by Raniero Gnoli. Torino: Unione Tipografico-Editrice Torinese, 1980.

___. `Srii Maaliniivijaya Vaarttikam. Edited by Madhusudan Kaul Shastri. Kashmir Series of Texts and Studies, no. 31. Srinagar: Kashmir Pratap Steam Press. 1921.

___. The Tantrasaara of Abhinavagupta. Edited by Mukunda Ram Sastri. Kashmir Series of Texts and Studies, no. 17. Reprint, Delhi: Bani Prakashan, 1982.

Abhinavagupta and Jayaratha. The Tantraaloka of Abhinavagupta with the Commentary of Jayaratha. 8 vols. Edited by Madhusudan Kaul Shastri and Mukunda Rama Shastri. Kashmir Series of Texts and Studies. Republication, edited by R. C. Dwivedi and Navjivan Rastogi. Delhi: Motilal Banarsidass, 1987.

Abhinavagupta, Utpaladeva, and Bhaaskara. II`svarapratyabhij~naavimar`sinii of Abhinavagupta, Doctrine of Divine Recognition, vols. 1-2. Includes Utpaladeva's II`svarapratyabhij~naakaarikaa and the commentary Bhaaskarii on Abhinavagupta's commentary. Edited by K. A. Subramania Iyer and K. C. Pandey. Reprint, Delhi: Motilal Banarsidass, 1986.

Al-George, Sergiu. "The Extra-Linguistic Origin of Paa.nini's Syntactic Categories and Their Linguistic Accuracy." Journal of the Oriental Institute, Baroda 18 (1968): 1-7.

___. "Le sujet gramatical chez Paa.nini." Studia et Acta Orientalia Bucaresti 1 (1957): 39-47.

Baadaraaya.na, `Sa^nkara, Vaacaspatimi`sra, Amalaanandasarasvatii, and Appayyadiiksita. The Brahmasuutra `Saa^nkara Bhaa.sya: With the Commentaries Bhaamatii, Kalpataru and Parimala. 2 vols. Edited by K. L. Joshi. Delhi: Parimal Publications, 1987.

Bandyopadhyay, Nandita. "The Buddhist Theory of Relation between Pramaa and Pramaa.na." Journal of Indian Philosophy 7 (1979): 43-78.

Bhart.rhari. Vaakyapadiiya of Bhart.rhari (with commentaries). 4 vols. Edited by K. A. Subramania Iyer. Kaa.n.da 1, Pune: Deccan College, 1966. Kaa.n.da 2, Delhi: Motilal Banarsidass. 1983. Kaa.n.da 3, pt. 1, Pune: Deccan College, 1963. Kaa.n.da 3, pt. 2, Pune: Deccan College, 1973.

p. 618

Burke, Kenneth. A Grammar of Motives. Berkeley: University of California Press, 1962.

___. Kenneth Burke on Symbols and Society. Edited with introduction by Joseph R. Gusfield. Chicago: University of Chicago Press, 1989.

___. The Rhetoric of Religion: Studies in Logology. Berkeley: University of California Press, 1970.

Cardona, George, "Paa.nini's Kaarakas: Agency, Animation and Identity." Journal of Indian Philosophy 2 (1974): 231-306.

___. "Paa.nini's Syntactic Categories." Journal of the Oriental institute, Baroda 16 (1967): 202-215.

Carrithers, Michael, Steven Collins, and Steven Lukes, eds. The Category of the Person: Anthropology, Philosophy, History. Cambridge: Cambridge University Press, 1985.

Crites, Stephen. "The Narrative Quality of Experience." Journal of the American Academy of Religion 39 (September 1971): 291-311.

Dharmakiirti. "Dharmakiirtis Sambandhapariik.saa." edited and translated by Erich Frauwallner. Wiener Zeitschrift fur die Kunde des Morgenlandes 41 (1934); 261-300.

Dharmakiirti and Manorathanandin. Pramaa.navaarttika of Acharya Dharmakirtti with the Commentary 'Vritti' of Acharya Manorathanandin. Edited by Dwarikadas Shastri. Varanasi: Bauddha Bharati, 1984.

Dharmakiirti, Viniitadeva, and Dharmottara. Nyaayabindu of Acharya Dharmakiirtti: With the Commentaries by Arya Vinitadeva and Dharmottara, and Dharmottara-Tika-Tippani. Edited by Dwarika Das Shastri. Varanasi: Bauddha Bharati, 1985.

Dumont, Louis. "World Renunciation in Indian Religions." Appendix in Homo Hierarchicus: The Caste System and Its Implications. Translated by Mark Sainsbury, Louis Dumont, and Basia Gulati. Chicago: University of Chicago Press. 1980.

Foss, Sonja K., Karen A. Foss, and Robert Trapp. Contemporary Perspectives on Rhetoric. Prospect Heights, Illinois: Waveland Press, 1985.

Gautama, Vaatsyaayana, Uddyotakara, Vaacaspati Mi`sra, and Vi`svanaatha. Nyaayadar`sanam: With Vaatsyaayana's Bhaa.sya, Uddyotakara's Vaarttika, Vaacaspati Mi`sra's Taatparya.tiikaa and Vi`svanaatha's V.rtti. Edited by Taranatha Nyaya-Tarkatirtha and Amarendramohan Tarkatirtha, with introduction by Narendra Chandra Vedantatirtha. Reprint, Delhi: Munshiram Manoharlal, 1985.

p. 619

Gerow, Edwin. "What is Karma (Ki.m Karmeti): An Exercise in Philosophical Semantics." Indologica Taurinensia 10 (1982): 87-116.

Gonda, Jan. Remarks on the Sanskrit Passive. Orientalia Rheno-Traiectina, edited by Jan Gonda and H. W. Obblink, vol. 4. Leiden: E. J. Brill, 1951.

Goudriaan, Teun, ed. Ritual and Speculation in Early Tantrism: Studies in Honor of Andre Padoux. SUNY Series in Tantric Studies, edited by Paul Eduardo Muller-Ortega. Albany: State University of New York Press, 1992.

Gusfield, Joseph R. The Culture of Public Problems: Drinking-Driving and the Symbolic Order. Chicago: University of Chicago Press, 1981.

Iyer, K. A. Subramania. Bhart.rhari: A Study of the Vaakyapadiiya in the Light of the Ancient Commentaries. Pune: Deccan College, 1969.

Jaimini, `Sabara, Kumaarila Bha.t.ta, Some`svara Bha.t.ta, Govindaam.rtamuni, and Mahaaprabhulaala Gosvaami. The Miimaa.msaa Dar`sana of Mahar.si Jaimini: With `Saabarabhaa.sya of `Sabaramuni, Tantravaartika by Kumaarila-Bha.t.ta with its Commentary Nyaayasudhaa of Some`svara-Bha.t.ta, Bhaa.syavivara.na of Govindaam.rtamuni and Bhaavaprakaa`sikaa by Dr. Mahaaprabhulaala Gosvaami. 2 vols. Edited by Mahaaprabhulaala Gosvaami. Varanasi: Tara Printing Works, 1984-1986.

Kau.n.da Bha.t.ta. Vaiyaakara.nabhuu.sa.nasaara. Edited with Hindi commentary by `Srii Brahma Datta Dvivedii. Varanasi: Chaukhambha Orientalia, 1985.

Kumaarila Bha.t.ta and Paarthasaarathi Mi`sra. `Slokavaarttika of `Srii Kumaarila Bha.t.ta: With the Commentary Nyaayaratnaakara of `Srii Paarthasaarthi Mi`sra. Edited by Dwarikadas Shastri. Varanasi: Ratna Publications, 1978.

Kunjunni Raja, K. Indian Theories of Meaning. Madras: Adyar Library and Research Centre, 1977.

Langendoen, D. Terence. Essentials of English Grammar. New York: Holt, Rinehart and Winston, 1970.

Lawrence, David Peter. "Argument and the Recognition of `Siva: The Philosophical Theology of Utpaladeva and Abhinavagupta." Ph.D. diss., University of Chicago, 1992. Revised manuscript being prepared for publication.

___. "`Siva's Self-Recognition and the Problem of Interpretation." Philosophy East and West 48 (2) (1998): 197-231.

p. 620

___. "Tantric Argument: The Transfiguration of Philosophical Discourse in the Pratyabhij~naa System of Utpaladeva and Abhinavagupta." Philosophy East and West 46 (2) (1996): 165-204.

Lyotard, Jean-Francois. The Postmodern Condition: A Report, on Knowledge. Translated by Geoff Bennington and Brian Massumi. Minneapolis: University of Minnesota Press, 1989.

Madan, T. N. Non-Renunciation: Themes and interpretations of Hindu Culture. Delhi: Oxford University Press, 1987.

Matilal, Bimal Krishna. "The Doctrine of Kara.na in Grammar-Logic." Journal of the Ganganatha Jha Research Institute 17(1960): 63-69.

___. "Indian Theorists on the Nature of the Sentence (Vaakya)." Foundations of Language 2 (1966): 377-393.

___. The Word and the World; India's Contribution to the Study of Language. Oxford: Oxford University Press, 1990.

Mills, C. Wright. "Situated Actions and Vocabularies of Motive." American Sociological Review 5 (1940): 904-913.

Naage`sa Bha.t.ta. Paramalaghuuma~njuu.sa. Edited with commentary by Kaalikaaprasaad Shukla. Baroda: University of Baroda, 1961.

Nakamura, Hajime. Ways of Thinking of Eastern Peoples: India, China, Tibet, Japan. Honolulu: University of Hawai'i Press, 1971.

O'Flaherty, Wendy Doniger, ed. Karma and Rebirth in Classic Indian Traditions. Berkeley: University of California Press. 1980: Delhi: Motilal Banarsidass, 1983.

Padoux. Andre, Vaac: The Concept of the Word in Selected Hindu Tantras. Translated by Jacques Gontier. The SUNY Series in the Shaiva Traditions of Kashmir, edited by Harvey Alper. Albany: State University of New York Press, 1990.

Paa.nini, Vaamana, and Jayaaditya. Kaa`sikaa: A Commentary on Paa.nini's Grammar. 2 vols. Edited by Aryendra Sharma, Khanderao Deshpande, and D. G. Padhye. Hyderabad: Sanskrit Academy, Osmania University, 1969-1970.

Pata~njali. Pata~njali's Vyaakara.na-Mahaabhaa.sya: Kaarakaahnika (P.1.4.23-1.4.55). Edited with introduction, translation, and notes by S. D. Joshi and J. A. F. Roodbergen. Pune: University of Poona, 1975.

Potter, Karl H. Presuppositions of India's Philosophies. Englewood Cliffs, New Jersey: Prentice-Hall. 1963.

Renou, Louis. "Les connexions entre le rituel et la grammaire en sanskrit." In A Reader on the Sanskrit Grammarians, edited by J. F. Staal,

p. 621

pp. 435-469. Cambridge: MIT Press, 1972; reprint, Delhi: Motilal Banarsidass, 1985.

Renou, Louis. "Grammaire et Vedaanta." In A Reader on the Sanskrit Grammarians, edited by J. F. Staal, 470-478. Cambridge: MIT Press. 1972; reprint, Delhi: Motilal Banarsidass. 1985.

Ricoeur. Paul. Oneself as Another. Translated by Kathleen Blarney. Chicago: University of Chicago Press, 1992.

___. Time and Narrative. 3 vols. Translated by Kathleen McLaughlin/Blamey and David Pellauer. Chicago: University of Chicago Press, 1984-1988.

Rocher, Rosane. "'Agent' et 'objet' chez Paa.nini." Journal of the American Oriental Society 84 (1964): 44-54.

Sanderson, Alexis. "Purity and Power Among the Brahmans of Kashmir." In The Category of the Person: Anthropology, Philosophy, History, edited by Michael Carrithers, Steven Collins, and Steven Lukes, pp. 190-216. Cambridge: Cambridge University Press. 1985.

Sanderson. Alexis. "`Saivism and the Tantric Traditions." In The World's Religions, edited by Stewart Sutherland et al., pp. 660-704. London: Routledge, 1988.

`Saantarak.sita and Kamala`siila. Tattvasa.mgraha of AAcarya Shaantarak.sita with the Commentary "Pa~njikaa" of Shri Kamalashiila. 2 vols. Edited by Dwarikadas Shastri. Varanasi: Bauddha Bharati, 1981-82.

Scott, Marvin B., and Stanford M. Lyman. "Accounts." American Sociological Review 33 (1968): 46-62.

Simons, Herbert W., and Trevor Melia, eds. The Legacy of Kenneth Burke. Madison: University of Wisconsin Press, 1989.

Somaananda and Utpaladeva. The `Sivad.r.s.ti of Srisomaanandanaatha with the Vritti by Utpaladeva. Edited by Madhusudan Kaul Shastri. Kashmir Series of Texts and Studies, no. 54. Pune: Aryabhushan Press.

Staal, Frits. "Ritual Syntax." In Sanskrit and Indian Studies: Essays in Honour of Daniel H. H. Ingalls, edited by M. Nagatomi el al., pp. 119-143. Dordrecht: D. Reidel, 1980.

___. Rules Without Meaning: Ritual, Mantras and the Human Sciences. Toronto Studies in Religion, vol. 4. New York: Peter Lang, 1989.

___. "Syntactic and Semantic Relations in Paa.nini." Foundations of Language 5 (1969): 83-117.

p. 622

___. ed. A Reader on the Sanskrit Grammarians. Cambridge: MIT Press, 1972; reprint, Delhi: Motilal Banarsidass, 1985.

Utpaladeva. The Siddhitrayii and the II`svarapratyabhij~naakaarikaav.rtti. Edited by Madhusudan Kaul Shastri. Kashmir Series of Texts and Studies, no. 34. Srinagar: Kashmir Pratap Steam Press, 1921.

Vasugupta and K.semaraaja. The Spandakaarikaas of Vasugupta with the Nir.naya by K.semaraaja. Edited and translated by Madhusudan Kaul Shastri. Kashmir Series of Texts and Studies, no. 42. Srinagar: Kashmir Pratap Steam Press, 1925.

Zollschan, George K. "Reasons for Conduct and the Conduct of Reason: The Eightfold Route to Motivational Ascription." In Social Change: Explorations, Diagnoses and Conjectures, edited by George K. Zollschan and Walter Hirsch. New York: John Wiley, 1976.