Nondualism in Indian philosophy of language

·期刊原文
The word is the world: Nondualism in Indian philosophy of language

By Ashok Aklujkar

Philosophy East and West

V.51 n.4 (Oct 2001)

pp.452-473

Copyright 2001 by University of Hawai'i Press

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

P.452

As someone writing on texts and thinkers in the Indian philosophical tradition, I should begin by criticizing the departments of philosophy at Western universities for the narrow and colored view they have of Eastern philosophies. However, since I am at present speaking at a department of philosophy (at the University of Hawai'i) that has probably done more for Eastern philosophies than any other department of philosophy in the Western world, I may not come across as credible if I voice the by-now-familiar but justified complaint of scholars interested in Eastern philosophies: I may be seen as an ungracious guest if not as someone whose statement is contradicted immediately by the evidence around him. If, after all the hospitality I have received from Professors Arindam Chakrabarti, Vrinda Dalmiya, Ramanath Sharma, and Eliot Deutsch and several students, I act as if I do not remember it, that karma will make me, in my next life, a politician!

Although Hawai'i's enviable philosophy department has thus deprived me of a provocative opening, I retain the freedom to express apprehension about certain tendencies I detect in contemporary philosophers. Since I am primarily a Sanskritist and Indologist and do not manage to keep up with writings in philosophy to any significant extent, my knowledge of these tendencies could be flawed. In fact, I would be happy if I turn out to be wrong. But, let me, while I have the floor, give the philosophers among you the benefit of knowing how a dabbling outsider sees their field.

The first tendency I would like to talk about has a direct bearing on the title of my talk. In fact, I have made it the first tendency because it can be related straightaway to my title. A reduction is implicit in the phrase "the word is the world." It clearly speaks of reducing the diversity we call "world" to a singularity named "word." I sense a general apprehension about such reductive endeavors among philosophers of the present and the recent past. Anyone coming to a conclusion like "X is Y" in the case of fundamental entities is suspected to be naive, unaware of the complexities of philosophical investigations. What this person may see as a remarkable summation following a great deal of hard thinking can be seen by many others as something too simple to be valid. This may, to some extent, be due to the psychology of specialists: they become accustomed to complexity to such an extent that in their view a simple, two-item pronouncement at a basic level amounts to being ultimately useless. Another possibility is that the reduction is sometimes so unexpected or so breathtaking that its very unexpectedness or breathtaking nature makes academics jittery. However, we must remember that simplicity is not identical with validity. Nor does a dizzying sweep necessarily mean skipping the logical steps

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

P.453

in between;1 otherwise, a phenomenon like the mathematician Srinivasa Aiyangar Ramanujan (1887-1920) would not have occurred. Supercomputers would not be working on figuring out the steps to the valid deductions Ramanujan made without specifying the stages through which his thinking passed. Thus, it does not behoove a philosopher to reject a reductive pronouncement out of hand. If he is not convinced by it, neither its simplicity nor its air of reckless abandon absolves him of his responsibility to counter it with good empirical and logical arguments.

The second tendency that generates resistence against formulations of the type "the word is the world" is reluctance to see truth as relative--as susceptible to levels.2 It is quite understandable why certain philosophers, including Madhva and Vallabha in the Indian tradition, feel that truth would not be truth if it were to change its nature according to the station or stance of the person doing the determination. One must concede to them, at least as far as the ultimate truth is concerned, the right to cry "foul" if they sense that a definition agreed upon in the beginning is being changed midstream. There is a sense of methodological security in the supposition that truth is something out there, waiting to be discovered and articulated. If it is thought to be unstable or changeable and open to the influence of angle, approach, or intellectual equipment of the searcher, one can understand why the reaction of some philosophers is not different from their reaction to the sudden removal of a stool from under their feet. Now, while the method that insists on sticking to initial definitions may lead to defensible philosophies and hence cannot be declared inferior, one cannot insist that it alone be the method. Doing so would commit the philosopher to certain positions which, themselves, ultimately, stand in need of justification. Regardless of what he feels, a philosopher's journey is not unidimensional or linear. It is not the case that there is a method lying around like a tool, and all he has to do is to pick it up and employ. As far as philosophizing goes, the problem of method and results (or consequences) can be like the chicken-and-egg problem. The choice of the method belongs to the metaphilosophical level, but that choice is not immune to the implications of individual object-level philosophies. That truth is invariable and must remain so is itself a determination of a philosophical kind. Without making it, a philosopher will not be able to rule out the possibility of using a different method-let us say a context-sensitive or "pay-as-you-go" kind of method.

Another way of phrasing my critique of the second tendency would be to say that, at the higher level-that of philosophizing about a philosophy or philosophies--the possibility that ontology and epistemology could be interdepedent needs to be kept open. To make the notion of invariable truth a part of one's ontological thinking is to create bounds for one's epistemological thinking. And to start the inquiry of how one knows with the supposition that how one knows has no bearing on what one comes to know--to assume that the truth out there is impervious to one's knowing capabilities and so forth-is to limit the potential of one's ontological thinking.

The third tendency that I find disquieting in contemporary or near-contemporary philosophy is the attachment--in fact, over-attachment-to discipline or department labels, especially in relation to religious studies. The presupposition behind the effort

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

P.454

the philosophers make to distance themselves from what they view as the territory of departments of religious studies is that the foundations of philosophy and of the study of religion are so different that a philosopher, while he is philosophizing, must not get into even the theoretical (including the psychological) component of religion. This presupposition is relatively recent in the history of philosophy.3 However, it has become a defining character of most philosophy departments, especially in the West and particularly at those universities in which separate departments of philosophy and religious studies exist. It has led to theory of religion being narrowed to a few fields such as theology and to stocking of Eastern philosophy books under "religion" or "occult" in bookstores and public libraries. It arises out of the view that concern with religious or spiritual life or with metaphysical reality is proper only for those departments which, unlike the departments of philosophy, do not distinguish between faith, belief, and so forth on the one hand and propositions derived solely from empirical evidence and reason on the other.

The sand layers on which this prevailing stance of philosophers is based are many. I have already suggested that the definition of philosophy presupposed in the stance has a very short history. Most pre-twentieth-century philosophers have not followed it. Nor does the narrow definition satisfactorily cover all activities such as the study of aesthetics that go on in philosophy departments.4 It has rightly been suggested that, in practice, the stance amounts to using ignorance or prejudice to keep the study of Indian philosophies out of the departments of philosophy ignorance if the department members have never taken the trouble of studying Indian philosophy to any significant extent and prejudice if they are aware of how much Indian philosophy covers but are unwilling to concede the possibility that some of it may have nothing to do with the ordinarily assumed characteristics of religion. More importantly, the assumption that a philosophical inquiry can proceed without any recourse to faith is itself of questionable validity. The faith involved in such an inquiry may not be faith in a particular book (dogma, tradition of thinking) or person, but is the difference of direction or object sufficient to establish that there is no element of faith in philosophical investigations?

That a philosopher is willing to examine or give up a particular faith as soon as its existence is discovered--to scrutinize or get rid of a mole in his and his companions' thinking as soon as it is exposed--is reassuring and may give the philosopher the right to adopt a holier-than-thou attitude toward a religionist, but it still does not prove absence of faith in his undertaking. As Arindam Chakrabarti (1992) has nicely argued in an article of his, an assumption of general unreliability of all heard information should not be made. Further, testimony or śabda-pramāṇa can be said to infiltrate even those judgments that are thought to be products of perception (pratyakṣa) and/or inference (anumāna) alone. Faith or reliance on something that is not independently proved, which is accepted because the rest of the philosophical model needs it-because not all circularity in reasoning can be avoided--is found everywhere. Goedel's proof to the effect that no system is complete in it self can be said to be another way of arriving at the same realization.

One can and, to a significant extent, one must distinguish the "logical" faith in

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

P.455

philosophical system-building--the faith in a given or an a priori--from the faith in religious discourse, but one can distinguish the former from the latter in categorial or disciplinary terms only if one decides to essentialize all religions narrowly or arbitrarily--only if one allows ethnocentrism to play a role. In our contemporary world, we should fare better, for we know now that there are religions in which no faith in a person like God is required or for which God is simply a convenient metaphor. Similarly, there are religions in which infallibility of scriptures finds scope only on a lower level; the guiding texts are declared to be like boats one does not need after crossing to the other shore, or they are compared to torches one discards after passing through a dark tunnel.5 Also, if in a religious tradition scriptural infallibility is maintained through a process in which linguistic competence and hermeneutic logic are employed to effect a reinterpretation in such a way that they amount to secular logic, then faith and the notion of infallibility become simply a mask to wear. The innocence, in the currently widespread stance of philosophers, of the possibility that logical reasoning, grounded in empirical experience, can exist camouflaged or can make a familiar bedfellow to faith now needs to be removed.

A further possibility of which philosophers need to spread awareness is that at least some of the ancient systems of thought can be models rather than direct enumerations of certain entities. One can, for example, look upon S ā ṁkhya and Vai śeṣika as attempts to create models made of what must be minimally assumed if the world is to be explained-a kind of creation of ideal languages. A view such as "there is no intellect or mind, there is only brain" or such as "there is nothing in S ā ṁkhya buddhi that corresponds to something in the real world" will not necessarily incapacitate the S ā ṁkhya model. Nor would the discovery of subatomic particles necessarily render useless the Vaisesika model incorporating anu, "atom," for a modern person trying to make sense of the world and man's reaction to it.

At this point some of you have probably begun to wonder if I am ever going to go beyond a critique of the heavyweights of the philosophical establishment and come to the specific topic of my presentation. What I have actually done in arguing that a time has not arrived to set aside philosophies culminating in reduction, taking a level-relative view of truth, treading fearlessly in all areas of inquiry, and giving lists of world constituents is to cut the ground from under some objections that could be made against the specific philosophy I am going to discuss. Usually, a speaker's practice is to explain a philosophy first and then consider its strengths and weaknesses. I have indeed put the cart before the horse, but I hope that once the horse is brought to the other side from the very brink of philosophizing and yoked where it should be, we can go galloping.

The other clarification I should offer is that I am not going to follow the usual, or at least the more common, method of Sanskritists or Indologists in which the activity of determining that a certain philosophical view can indeed be attributed to a philosopher is carried out first and then, if and where necessary, an explanation or justification of why the philosopher held the view attributed to him is attempted. Commonly, textual or linguistic philology precedes the philology deriving itself from an awareness of philosophy in general and of the larger historical and cultural con-

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

P.456

text of a particular tradition. However, I will pretend that I have understood my proponent of "the-word-is-the-world view" correctly and will mostly refer you to my 1970 Ph.D. dissertation for textual evidence. Further, using the idiom of our day, I will try to recreate my proponent's argument in a sequence I find convenient. While I am reasonably sure that if the proponent were present here today and addressing this gathering in English, after having acquired a working knowledge of Western philosophy, he would not say something very different from what I am going to say, I cannot assert that each and every detail of my recreation of his background thinking was known to him or would have occurred to him. For each proposition I attribute to him, I have tried my best to ensure that he was aware of it or could reasonably be said to have been aware of it and that he accepted it or would have accepted it in the relevant context at the proper level. However, I cannot deny that there is an element of "creative" philosophy on my part.

Bhartṛhari, the Grammarian-Philosopher most eligible for attributing the view "the word is the world," lived sometime around A.D. 425-450, if not earlier. In the current state of our resources, he can be viewed as the originator of the view, although it is clear from his major surviving work, the Trikāṇḍī or Vākyapadīya, that the view must have preceded him by quite some time, probably many centuries, in several important details if not in its entirety. The argumentation in support of the view, as distinct from the statement of the view, is at present found for the first time only in Bhartṛhari's incompletely preserved magnum opus, albeit it is not Bhartṛhari's principal intention, except maybe for a part of the first book, to argue systematically in favor of the view. My aim here is to identify myself with Bhartṛhari, unearth the many facets of his argumentation, and give him the best possible hearing that I can.

The word, even when understood as a linguistic unit of any length--from one phoneme, through a free morpheme, to a phrase, clause, sentence, or entire discourse--is clearly not the same as the world. It is rather something which exists, is made, or is realized in the world. Logically, therefore, it is possible to establish the identity of word and world only if we go below the common understanding of "word" or above it. The former way, way (a), provides two options:

(a-i) Understand "word" as standing for what it physically is, that is, as sound.

(a-ii) Understand "word" as standing for the form in which it survives or exists, that is, as the impressions or images in the mind which correspond to the sound realizations we call linguistic units.6

The former captures the physical common element of words. The latter turns words into a set of mental entities.

The second way, (b), that of expanding the meaning of "word" beyond phoneme, morpheme, phrase, clause, and so forth, can also bifurcate:

(b-i) By "word" what we mean here is "a whole language, an entire system of linguistic symbols" or "a set of languages, a collectivity of linguistic systems" that a person knows.

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

P.457

(b-ii) The word is ultimately that principle, power, or force which is at the bottom of all language knowing. It is the entity that makes all individual languages possible.

Both the major ways, (a) and (b), of making the word include the world would obviously involve collapsing the subject-object duality. However, they achieve this collapsing in manners that have important repercussions for philosophy and religion. Following Yogi Berra's weirdly brilliant advice, "When you come to a fork in the road, take it," I will now consider them, along with the two possibilities under each of them.

Bhartṛhari must have been aware of the view that there is some intrinsic relationship between a sound, that is, a sound sequence,7 and the object it denotes. Statements such as sa bhūr iti vyāharat, bhūmim asṛjat, "he uttered the word bhū and he created bhūmi" (the plane or basis for things to exist, the earth) (Taittirīya Brāhmaṇa 2.2.4)--reminiscent of "And the Lord God said, Let there be light: and there was light" (Genesis 1 : 3)--are available in texts that definitely antedate his work and must have been known to him.8 One well-known kārikā or verse of his reads chandobhya eva prathamam etad viśvaṁ vyavartata, "All this first appeared as diversity with the Chandas body (i.e., the revealed Veda words) as its source." However, we do not come to know from him the reasoning behind the assertion that there is some kind of intrinsic physical relation between word sounds and objects. Nor can we be certain that the assertion was universalized, extended to all words, by him and adopted as a view. Sounds producing things and things being ultimately made of sounds is evidently not asserted beyond the "in-the-beginning-there-was" type of statements. Also, even when the creation context is transcended, it is only the sounds uttered by certain individuals, commonly called ṛṣis or seers, that are said to have a genetic relationship with things. Bhartṛhari's general theory of language may use the relationship as one argument but is unlikely to make a foundation of it. It can be an adequate foundation only for mantra philosophy. I should, however, observe in passing that the postulation of a relationship of the described kind is not as outlandish as it may seem at first blush. If physical objects ultimately consist of constantly moving subatomic particles, that is, of vibrations, it makes sense that sounds, which also consist of vibrations, should have a family connection with them.

Possibility a-ii would lead us to a Yogācāra or Vijñāna-vāda kind of view, "the external world truly or ultimately exists only in our mind" -- a kind of idealism, if we permit the application of the term "idealism" even to a plane of reality that transcends the ordinary worldly one.9 The Vijñāna-vāda talks in terms of vijñāna or citta; the option we are considering would talk in terms of linguistic units. Apart from this difference of idiom, it may not differ from the former in any essential way. By way of clarification, however, we should note that it will not have to hold that for each object of the external world there is a corresponding designation or linguistic unit in the mind-that there is a one-to-one mapping. It needs to assume only that the extent of the physical world cannot exceed what language can bring to the perceiver as existing. If any objects exist that are not reflected in language even through a pro-

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

P.458

noun or an expression like "this is something," these objects are of no use to a philosopher trying to determine what really exists. A philosopher's ontological investigation must begin with what language brings to him as existing and cannot outpace language. The moment one begins to speak of an object not reflected in language, it either ceases to be an object that is not reflected or becomes useless for the philosopher, very much like a lost space vehicle to NASA scientists.

Now, we see a clear articulation of this position in Bhartṛhari's work, as the Vṛtti 1.129 line sad api vāg-vyavahāreṇānupagṛhītam artha-rūpam asatā tulyam informs us.10 Thus, the thesis that the collectivity of physical objects that the world is is subsumed by the linguistic units that constitute a language or set of languages known to a person can also be attributed to Bhartṛhari. It is one way in which he can defend the thesis of "word : world" identity, but it is a somewhat "pedestrian" way. The epitaph on Mr. Yeast's tomb in Ripley's Believe It or Not Collection reads "Pardon me for not rising." A philosopher establishing the "word : world" identity only on the ground just now specified will have to extend a similar apology.

Coming to way (b), we first need to ask if there indeed could be any justification for extending the meaning of the word "word" to language totalities residing (in some sense) in an individual. In Sanskrit, such an extension would mean equating śabda and vāc in specific contexts.11 Bhartṛhari clearly equates them thus, and is preceded and followed by authors who show no hesitation in assuming the same equation.12 There is thus philological support for extending the meaning of śabda. In English and other languages, too, a precedent can be seen, depending on the theology and interpretation one prefers, in statements such as "In the beginning was the Word, and the Word was with the Lord God" and "And the Word was made flesh, and dwelt among us, and we beheld His glory, the glory as of the only begotten of the Father, full of grace and truth" (John 1 :1, 14). However, what we need in the present context is not a philological justification but a logical one. The word "word" can definitely cover a particular sound (i.e., a particular sound sequence) and the anchor of that sound in the mind. Words are not lost when the sounds vanish. Most of them are preserved somewhere in us for significant lengths of time. Also, it makes sense to assume, at least in a preliminary model of how articulation or communication takes place, that, first, a thought or string of linguistic symbols occurs to a speaker, and then the speaker gives a garb of sound to that string. Since thought incarnates as sound or sound is given a mental counterpart, there should be no objection to thinking of "word" as standing for both.13

Now, the thought infused with linguistic signs that is postulated as an entity preceding an utterance cannot come into existence unless the linguistic signs are stored, metaphorically speaking, in a deeper layer of our mind. We have to imagine a storing place or stage, somewhat like the freezers of affluent American families, in which all the linguistic signs, or the impressions of significant sounds known to the speaker, are preserved and from which they can be taken out for use as they are needed. Some may be stored for longer times, some for short durations; some in sharp shapes and some as diffused images; some as small as phonemes and some as long as whole discourses; and some grammatical and hence recommended to all

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

P.459

like fresh vegetables and some ungrammatical and thought to be fit, like meat, only for some persons.

Assuming a storehouse of the described kind, now recall that the transformation of the latent elements of thought or language into an identifiable thought or linguistic string-into something with sequenced expressions, into mental speech (and later into audible speech) -- is a process that must be accepted by any theoretician trying to explain linguistic communication.14 A theoretician may differ in choosing terminology but must admit that a transition takes place. Something that was not at the surface comes to the surface. The unmanifest becomes manifest and amenable to reproduction in sound that can be heard by others. The unmarked provides markers that we can use to give it an identity. The transformation involved is no less intractable than that of one energy form into another.15 For it to occur (and it must be included in the explanatory model), we must assume that there is no barrier between the stage of latency and the stage of manifestation--that one flows into the other, that the mind pancake is simply flipped. We cannot justifiably assume a physical separation between thought and the reservoir which lies under or behind it. In fact, thought would only be a ripple on the boundless reservoir, only a phase among the countless phases through which the reservoir constantly passes. Other than associating what seems to be one thought with one moment or phase, there is no real separation that we can defend with valid arguments.

Further, the contents of the seamless reservoir, even while remaining largely the same, can change and do change, especially as a person learns more of a language or a new language. In theory, therefore, we need to assume a basic or pure form of the reservoir that takes on various colorings in the life of an individual. The basic or pure form may never be realized, unless we entertain the possibility that the mind can be trained not to "mind" its usual business, to empty itself, to cease to exist in its worldly garb. But such a form must be postulated in theory, for, without it, there will be no accounting of the continuity and change in an individual's language system or set of language systems. Given its priority in postulation, we may call it the original form, adopting a pseudo-chronological idiom, and think of it as the source of all linguistic emanation.16 However, we cannot think of it as a physically separable entity unless we admit the possibility of transcending the mind or mental make-up of the worldly kind, that is, unless we admit the possibility of a state like mokṣa or nirvāṇa.

The moment the mind of an individual reaches a state in which it is devoid of its peculiarities or colorings, it ceases to be an individual's mind. It becomes indistinguishable from any other mind, as well as from whatever it is that makes the individual a possessor of mind and makes his or her body tick with life. In the theoretical model presupposed at this point, mind, then, of necessity, turns out to be the same as consciousness (buddhi), life force (prāṇa, jīva), sentience (cit or caitanya), soul or Self (ātman-this last in its non-worldly aspect) -- everything that the physical body is not, even if mind and body are in reality mutually dependent and mutually influential and even if distinctions are made between mind and other manifestations of life such as breath (prāṇa) and speech (vāc).17

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

P.460

There should thus be no objection to extending the meaning of "word" and to equating it with one or more languages and with what underlies all languages. Bhartṛhari accepts such extensions through the notions of paśyantī and para paśyantī -rūpa, respectively. Under the alternative "word as language" or "word as paśyantī ," then, the word is equal to the world in a manner similar to that of sub-way or option a-ii. The known or knowable world cannot exceed what language contains. However, unlike sub-way a-ii, there is no necessary claim here that worldly objects do not actually exist or can exist only in dependence on mind. A genetic relationship between bāhyārtha or vastvartha, "external objects," on the one hand and vāc or language on the other is not postulated or proposed.18

Finally, in sub-way b-ii, in which śabda or word is taken to mean the entity that underlies all individual manifestations of language, that is, the language principle, we have the possibility of equating word with the world in a causal sense. Like the brahman of Vedāntins, the śabda-tattva brahman of Bhartṛhari can be viewed as the single ultimate source of all the diversity we perceive and infer.19 Under this alternative, the world will not be distinct from word in the sense that it will not come to exist without the latter, and if it ceases to exist it will revert to the latter.

It is only under this final alternative that the thesis of "word : world" identity rises to a truly philosophical height, like real-life yeast and unlike Mr. Yeast of Ripley's Believe It or Not. The identity through sound or vibrations can only be limited to mantra theory, as we noted. Coming to realize that the extent of the world cannot outreach the extent of language is a significant step in philosophy -- a step which tells us that the island of ontology has a coextensive ocean of language under it and which suggests that a philosopher should pay attention to language. However, it may be viewed as a mere methodological insight or as an explication of something that a philosopher must abide by, whether he likes it or not.

So far I have explained only the possible ways in which the world could be reduced to word and have given evidence of Bhartṛhari's awareness of these ways. To say the same thing differently, I have given a demonstration of how multifaceted his treatment of the issue turns out to be, if we systematically collect all his relevant statements. The demonstration, in turn, indicates that, at the metaphysical level, Bhartṛhari is a perspectivist. He is willing to consider the possibility that a given view or position may be correct, provided it is placed at the right level in the midst of other logically related ideas. This does not mean that he would defend each and every idea or that there is no scope for gradation of views in his philosophy, but only that he is in favor of maintaining an attitude of intellectual generosity and willing to replace a plank at a time while remaining afloat in his boat. At the metaphilosophical level, philosophization seems to have been to him like the piece of leather which the Indian ascetic Kalanos is said to have used to demonstrate to Alexander the Great the instability of the empire that Alexander was building.20 Kalanos, according to this account, stepped on one side of a rather large piece of dried leather. The part on which he stepped went down and the part opposite to it went up. Then he stepped on the part that went up and showed how the other part would rise. The territory won by Alexander, he said, was going to be like the leather.

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

P.461

At the most the region in which Alexander was present would remain quiet. At least one of the other regions would rise against him when he was away from it.

The analogy Kalanos used for giving Alexander the message of futility of imperialist ambitions can also be used to point out the interrelatedness of philosophical ideas suggested by Bhartṛhari's procedure. The conclusions reached in one area invariably affect the conclusions in another. Hence, attention to the setting and level of a view is necessary. Its metaphilosophical implications need to be noted. If it pertains to the epistemological half, what it would do to the ontological half of the Siva of philosophy must be taken into account. Just as in the history of Western philosophy the nature of philosophizing has become clearer as a result of the linguistic approach, so also Bhartṛhari's approach to what we would call philosophy has been shaped by his awareness of the role language plays in the constitution of darśanas.

To anyone who knows the historiography of Indian philosophy, it should not come as a surprise that Bhartṛhari's statement that śabda-tattva brahman ultimately contains the world has been viewed by many modern scholars as only a linguistic version of Vedānta philosophy. "Only the name differs; it is essentially the same kind of thinking," some of them have assumed if they have not so bluntly stated.21 Furthermore, it is not surprising that the two alternatives entertained in the case of the Vedā fsntic brahman's creation of the world, namely pariṇāma, "real transformation," and vivarta, "only apparent, illusory transformation," have been entertained also in the case of Bhartṛhari's brahman. Consequently, his verse (1.124) śabdasya pariṇāmo 'yam ity āmnāyavido viduḥ / chandobhya eva prathamam etad viśvaṁ vyavarrtata // has been subjected to much discussion. The absence of contrast between pariṇāmaḥ and vyavartata (a finite verb from the same root vi+ vṛt, which figures in the noun vivarta) in this verse has naturally puzzled scholars, given the distinction of pariṇāma and vivarta stated sharply and frequently by post-Śaṁkara authors.22 Dependent, to some extent, on whether we declare Bhartṛhari to be an upholder of pariṇāma or an upholder of vivarta is the issue of whether he is a monist or a dualist.

Strictly speaking, Bhart ṛhari is not a Vedāntin, although broad similarities exist between his thought and that of the Vedāntins like Maṇḍana-mi śra and Śa ṁkara and although the traditions of pariṇāma proponents and vivarta proponents were probably older than his own time (cf. V ṛtti 1.124-128; 1.141 and its Vrtti). It would be more accurate to speak of Bhart ṛhari as a Trayyanta-vādin,23 a term having a close connection with Vedāntin but coming from a period in which Vedanta was not a system or school. More importantly, as I shall outline below, the argument that is associated with śabda-tattva-brahman is very different in detail, as well as thrust, from the argument surrounding the Vedāntin's brahman.24

As far as the creation of the world from the language principle is concerned, what principally sets Bhartṛhari apart from the Vedāntins is that he has two versions of this creation. We may call them a weak version and a strong version. According to the latter, even the physical things of the world could be said to come from brahman. If they so come without changing the nature of brahman, they will have only an illusory or relative existence. Bhartṛhari would be a language monist very

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

P.462

much in the fashion of Śaṁkara's Advaita. If the physical things come as a real transformation of brahman--a transformation in which brahman generates physical things, continuing into them without losing its own nature--Bhartṛhari would be close to Vallabha's Śuddha Advaita. Bhartṛhari seems to be aware of these two possibilities under the strong version (see Vṛtti 1.141). He does not reject the strong version and may even be said to speak of it as having support in some ancient texts, perhaps the Veda (cf. Vṛtti 1.124). However, it is not the version he primarily needs for his philosophy, and it is not the only version he considers absolutely proved. In keeping with and as part of the perspectivistic thinking displayed elsewhere in his work, he settles for the minimum he needs and keeps open the possibility that, at the ultimate level, more than one philosophical conception may be valid. In other words, he seems to refuse to force the issue at a plane where no determination, one way or the other, can be exclusively defended and where such a defense would be more a matter of personal preference and scriptural interpretation than of having a logical proof. He opts for a semantic ascent that is at the same time an insightful identification of the limitation besetting all attempts at conceptualizing the Absolute or the final reality.

To describe the same situation differently, Bhartṛhari is perhaps unique among Indian philosophers to suggest the possibility that a philosopher's determination of fundamental reality is relative to whether or not he is willing to accept the possibility of mokṣa, spiritual liberation. Bhartṛhari's own preference almost certainly was to accept the possibility, but, if we were to acquaint him with our distinction between philosophy and religion, he would probably have said that his preference, while not being illogical, was religious or cultural rather than philosophical.

Now, what is the weaker version I am talking about which enables Bhartṛhari to achieve a semantic ascent? What is the proof that we can declare it to be acceptable to him as a philosopher? Let me take up the latter question first. Given the limitation of time, I will present only a part of the evidence. The opening verse of Bhartṛhari's Trikāṇḍī or Vākyapadīya has been cited countless times. It runs thus: anādi-nidhanaṁ brahma śabda-tattvaṁ yad akṣaram / vivartate 'rtha-bhāvena, prakriyā jagato yataḥ //. Regardless of how one translates this verse, one must reckon with the fact that here Bhartṛhari speaks of the prakriyā of jagat as coming from brahman, not the jagat itself. The word prakriyā has a well-established sense of "derivation, bringing constituents together in such a way that a final product emerges." In the tradition of Pāṇinian grammar, of which Bhartṛhari was one of the greatest exponents, prakriyā is regularly used for the process which leads us from the elements made available by Pāṇini's rules to the actual forms used in the language he describes. So, what Bhartṛhari is causally attributing to his brahman, the language principle, is primarily the explication or systematization of the world.25 The world as we conceive it is the creation of śabda-tattva-brahman. Apart from this conceiving, we have no access to the world. In that sense, or at least in that sense, the world exists only because of śabda-tattva-brahman, and its existence is coextensive with that of śabda-tattva-brahman. In the Vṛtti commentary on the quoted verse, too, the expression jagadākhyā vikārāḥ prakriyante is found, not jagadākhyā vikārāḥ kriyante or jagadākhyā

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

P.463

vikārā utpadyante.26 Another fact that an interpreter must reckon with is that the verse speaks of vivarta "diversification" through artha-bhāva "being an artha" and that artha in Bhartṛhari's text stands for vastvartha or bāhyārtha, "thing, external object, referent," as well as śabdārtha "sense, meaning, concept."

In the weaker version, the existence of the physical universe is taken for granted. It is not considered necessary to think of it as a product of brahman. However, its analysis or our generation of a conceptual scheme (not only the one done by a philosopher) depends on the language principle. In other words, it is the world's generation for the observer that the word brings about. At the bottom or head of this process-depending on how one chooses to look at it--are time and space. Without them, the separation of objects or the creation of the discrete entities needed for systematization would be impossible. Bhartṛhari, then, appropriately, regards time and space as the basic and inseparable potencies or powers (śakti) of the language principle. Once they become operative, things attain individuality, and the talk of relating or distinguishing things can begin. Out of such relating and distinguishing arise the categories of logical and linguistic thinking. Among the categories that the Indian tradition was aware of were different phases (avasthā) of a thing; collections or combinations (sāhitya, sāmagrī, samūha, or saṁsarga); absence (abhāva); capabilities, potencies, or powers (śakti); relations (saṁbandha); qualities (guṇa); universals (jāti); numbers (sāṁkhyā); time (kāla); and space (diś). Since we have noted that the last two have been taken as part and parcel of the language principle as its powers, it is only the ontological status of the remaining that we need to note. Although the remaining categories may ordinarily be thought to exist over and above the physical things, it is evident that they do not exist in the same sense as physical things. They are, therefore, viewed as products of the interaction between the language principle and physical things, with the interaction taking place through the time and space powers of the former. They are denizens of individual language systems, which we noted earlier as the third defensible meaning of vāc or śabda, as Bhartṛhari's paśyantī.

To make essentially the same point differently, a physical object is not directly accessible or accessible in itself according to Bhartṛhari. Each such object is a bundle of potencies or capacities. The potencies are (a) universal or class membership; (b) property, quality, or attribute; and so forth. In each act of cognition of a physical object, one or more potencies are activated. To account for why a particular potency and not another is activated, we postulate time as the activator and prohibitor of potencies. To account for the discreteness or physical separation, we postulate dig or a mathematical type of space as a force emanating from the language principle.27 The acceptability of this explanatory model is to be judged on the basis of its comprehensiveness and economy. Its assertion as an ontological truth is not independent of its epistemological efficacy.

Does the weaker version make Bhartṛhari a dualist, since it contains the physical world and the language principle as the two ultimates of his philosophy? Jan Houben (1995, p. 17 n. 40) has suggested this as a consequence of attributing the version to Bhartṛhari. However, the "dualist: nondualist" distinction, strictly speaking, belongs

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

P.464

only to the plane of ultimate experience.28 If Bhartṛhari was willing to entertain the possibility of there being a mokṣa (nirvāṇa or spiritual liberation) state, as he evidently was, even the weaker version leaves open to him the possibility of saying that, in the mokṣa state, the world ceases to exist for the person concerned, and only the language principle remains. It is not necessary that the world cease to exist absolutely or unqualifiedly for a monist.

It is suggested by the preceding account that Bhartṛhari takes a linguistic approach to philosophy in that he dissolves several categories such as relations, qualities, universals, and numbers as linguistic constructs after an analysis of how they find expression in the usage of the thinkers of his tradition. This approach is different in details and perhaps too radical in its outcome for many of the modern taste. However, it should not be viewed as invalid on that count. Furthermore, the approach is accompanied by a philosophy of language and a philosophy of grammar and, as mentioned earlier, gives his consideration of the world and ultimate reality a context that is significantly different from the context for the same issue in Vedanta.

In order to reduce the world to the word in either of his two versions, Bhartṛhari needs to accept that the mind (manas, buddhi, ātman in its worldly aspect) or the subject is the same as one of the levels of language or vāc. Although he can retain the duality of mind and language in usage, he must conclude that, at some point, they are the same ontological entity. If mind, as it is ordinarily understood, exceeds language in its scope, and language, as an abstracted or internalized entity, occupies only a part of mind, there would be a part of the world to which mind can gain access but language cannot. As we have seen, Bhartṛhari denies this possibility. Second, there are indications in his statements to the effect that mind is the same as language 3 or paśyantī.29

A consequence of30 the philosophical move just now specified is that a state of mind, whether we call it cognition, thought, or emotion, must be held to be identical with a state of language, namely the sentence or proposition.31 At a lower level, Bhartṛhari can speak of cognitions as being always infused by language, as he in fact does, but at a higher level he must hold that cognition and the formation of a sentence are not really different as events or states. As a mental state does not have a measurable duration, it can only be assigned the (theoretical) duration of one moment.32 The same would then necessarily hold true of a language state, which means that a sentence or proposition must be thought of as unitary. This unitary nature of sentence and sentence meaning has been argued for at great length by Bhartṛhari.33

The many insightful arguments advanced in favor of the unitary character of the sentence also establish that in Bhartṛhari's view the sentence is the primary unit of language. Words and word meanings (pada and padārtha) have a secondary or derivative reality, as they depend on being abstracted from sentences.34 Word meanings are not the same thing as external objects or referents.35 They are not even the same thing as images of external objects. In fact, ultimately, ontologically speaking, they are the same thing as words. It is not the case that words and their meanings are stored separately in the mind (or language 3). Meanings, if they are to

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

P.465

be thought of at all as separate from their signifiers, exist as impressions of signifiers. They are a vivarta, "temporary manifold creation," from words, which are ultimately one with language 3 and the language principle. While they may not subsume all physical objects, they do subsume the reflections or repercussions of all known physical objects, as well as the categories created to make sense of the physical world. As such, the language principle, the ultimate source of signifiers and meanings, can be said to subsume the physical world.

It should be evident from the preceding quick summary of Bhartṛhari's extensive discussions falling in the area of philosophy of language and, to a small extent, in the area of philosophy of grammar, that the content and general nature of his epistemology and ontology are very different. While his brahman may actually or factually be the same as the brahman of the well-known Vedāntins, the way its nature is determined and described is unlike that of the Vedāntins. He makes no claim of originality, and his thinking clearly has affinities with the philosophy of sound preserved by mystics and Tantrics, yet, at the same time, we must recognize that without him we would not have known the philosophical depth of the religious teachings and practices of the mystics and Tantrics or of the several pronouncements in Vedic literature. To be sure, the profoundity of his conclusions comes from an uncompromising investigation that is not cowed by the thought of divisions of expertise or by the thought of the possible reduction of cherished category differences. However, the profundity also comes from the scope he offers to ontology and epistemology to go hand in hand-to affect each other. If philosophy was not like Kalanos' dried leather to Bhartṛhari, it was like a slab of jello to him. The moment one tries to hold one side, the middle or the other end wiggles.

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

Notes

1 - The situation is comparable to distillation of principles that make a society civil or better. That one can provide complex formulations of interrelationships does not mean that the items to be interrelated cannot be the simple basic truths such as "avoid violence" and "speak the truth." The "motherhood" notions of a civil society can be connected to each other in very complex ways. [Back]

2 - In this context, "truth" should be understood as standing for "truth or reality." [Back]

3 - The presupposition can also be viewed as an implication or consequence in the present context. [Back]

4 - The existence of metaphysical entities such as mind is frequently assumed in statements of aesthetic theories. [Back]

5 - There may also be religions in which the infallibility of scriptures is viewed at least by some individuals or by a group merely as a fiction that is to be maintained for reasons of convenience and social good.[Back]

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

P.466

6 - Evidently, a word is not lost when it ceases to exist in the medium of sound. If it leaves an impression on the mind, as it commonly does, it can be regenerated in sound from that impression. There is, thus, at least in a large number of instances, a mental counterpart connected with the sound form. [Back]

7 - Since even what appears to be a single sound to human ears can be thought of as a divisible entity, we will not be wrong in using the expression "sound sequence" or, in the present context, in thinking of "sound" as an abbreviation for "sound sequence." [Back]

8 - Bhartṛhari was definitely familiar with Yāska's Nirukta. The so-called etymologies of the Nirukta may be based on the assumption that there is some kind of connection between the nature of a thing and the form of the Vedic word used to refer to it. The word might have been seen as creating or bringing to the forefront a particular aspect in a thing. [Back]

9 - Reality captured in a philosopher's reconstruction is also a part of ordinary worldly reality in the mode of thinking implicit in traditions of Indian philosophy such as Vijñāna-vāda. Transcendence of ordinary worldly reality does not take place only through a theoretical or intellectual grasping of it in its entirety. Even when such grasping is valid, it must be accompanied or followed by the right kind of change in the cognitive capabilities or mechanism (the "spirit") of the person concerned. [Back]

10 - There is much evidence in several old and mutually independent sources that supports the ascription of the Vṛtti commentary to Bhartṛhari. Even if the merely logical possibility that the Vṛtti was authored by a student of Bhartṛhari were to be admitted, it would be very difficult to demonstrate that there is any significant difference of views between Bhartṛhari and the Vṛtti author. Even Houben's (1995, 1997, 1998, 1999) painstaking attempt at such a demonstration has not gone beyond making the claim that there is a difference of emphasis in the main verse text of the TK/VP and the Vṛtti. Apart from the difficulty of proving a difference of emphasis in the case of badly preserved ancient works, we face in this respect the difficulty that the difference Houben perceives may simply be due to the fact that one work is a kārikā text and the other is a commentary, expected by definition to be clearer, complementary, and detailed. [Back]

11 -Other Sanskrit words capable of expressing the meaning "language" (e.g., bhāṣā, vāṇī, gir do not seem to have the meaning range of vāc. [Back]

12 - Note, for example, the Mahābhāṣya gloss mahān devaḥ śabdaḥ ... mahatā devena nah sāmyaṁ yathā syād ity adhyeyaṁ vyākaraṇam under atha śabdānuśāsanam (Kielhorn edition, p. 3) for an earlier use of śabda in an extended sense. That Patañjali was aware at least of a philosophy that understood vāc in an extended sense is confirmed by the immediately following Mahābhāṣya statements and by Mahābhāṣya at the end of the Śiva-sūtra discussion, that is, the end of āhnika 1.2 (Kielhorn edition, p. 36). For a later use of śabda in an

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

P.467

extended sense, cf. Daṇḍin, Kāvyādarśa/Kāvya-lakṣaṇa 1.4: idam andhaṁ tamaḥ-kṛtsnaṁ jāyeta bhuvana-trayam / yadi śabdāhvayaṁ jyotir āsaṁsāraṁ na dīpyate //. [Back]

13 - These sentences seek to capture the vaikharī and madhyamā levels or phases of vāc as viewed by Bhartṛhari. [Back]

14 - At this level, it is not possible to decide whether one is dealing with thought or with language. Use of either term could be defended. [Back]

15 - The stages of articulation or utterance may be summarized thus: seeds of thought / latent language, a sequenceless group - sentence formation in mind, the "light" of thought infused with sequence and subtle operation of the breath - thought / sentence given the form of physical sound. [Back]

16 - We do not need to go beyond the original form. A theoretical model should aim at economy and should not be exposed to infinite regress. [Back]

17 - Every thinker needs to presuppose a model like the following if it is agreed that occurrence of cognition or knowledge takes place and one should account for it:

(a) sentience, pure consciousness, or life principle (cit, citi, caitanya),

(b) a cognizor, determinate consciousness, or intellect (buddhi), serving as the agent of specific acts of cognizing, and

(c) a state, act, or event of cognizing, a particular cognition (jñāna).

A thinker may choose names other than the ones given above. In his higher-level or ultimate ontology, he may also dispense with some of the entities included in the model. Further, he may choose to characterize certain entities as processes or flows, rather than as entities in the usual sense of the term, that is, as objects with a stable or constant character (some Buddhist philosophers in fact do so when they speak in terms of saṁtāna, saṁtati, pravāha, etc.). Yet, while he arrives at or defends his explanation of our experiencing the world, he needs to assume that there are in us or there take place in us things of the above three kinds. [Back]

18 - We may think of this as a weaker version of the philosophy of linguistic monism, to which I will return later. [Back]

19 - Some writers may have used śabda-brahman in the sense of śabda-tattva brahman. However, there is no clear evidence suggesting that Bhartṛhari used the two terms as synonyms. His śabda-brahman(Vṛtti 1, 12, 38,139-140) occurs, except perhaps in Vṛtti 1.139-140, in contexts in which one could take it as a reference to the Veda collectivity or to the collectivity of acceptable linguistic expressions (sādhu śabdas, something analogous to paśyantī but confined to standard Vedic and Sanskrit usage). Since the essence of the Veda collectivity (praṇava, om) is the same as śabda-tattva brahman in his philosophy, there is a close connection between śabda-brahman and śabda-tattva-brahman, but, in

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

P.468

the present state of the evidence, one cannot be certain that the former designates an entity as fundamental as the latter. [Back]

20 - The Latin spelling of this Greek transcription (possibly of Sanskrit kalyāṇa) is Calanus. I have not been able to locate the incident I summarize below in the Greek and Latin sources on Alexander's India expedition. My authority is at present only the series Cāṇakya shown on Dūra-darśana (Durdarshan), India's national television channel. However, the references to Kalanos and his compatriots made by the Classical authors make it plausible that Alexander could have been given the blunt message mentioned below. [Back]

21 - (a) Cf. the expositions of Bhartṛhari's thought, for example, in Ramaswami Sastri 1938, Gaurinath Sastri 1969, and Bishnupada Bhattacharya 1985. The modern major Sanskrit commentator of Bhartṛhari's Trikāṇḍī, Pandit Raghunātha Śarmā, gives an essentially Advaita Vedāntic reading of Bhartṛhari's thought in his commentary Ambākartrī.

(b) The tendency I have attributed to modern scholars has some antecedents in the works of some later grammarians of the Pāṇinian tradition. Bhaṭṭoji-dīkṣita, for example, remarks in his Śabda-kaustubha (p. 12): tad evaṁ varāṭikānveṣaṇāya pravṛttaś cintā-maṇim labdhavān iti vāsiṣṭha-rāmāyaṇoktābhāṇaka-nyāyena śabda-vicārāya pravṛttaḥ san prasaṅgād advaita aupaniṣade brahmaṇy api vyutpadyatām ity abhiprāyeṇa bhagavān bhartṛharir vivarta-vādādikam api prasaṅgād vyudapādayat. tat tu tantrāntare sphuṭaṁ prakṛte nātīvopayuktaṁ ceti neha tanyate. Obviously, Bhaṭṭoji thinks of brahman as properly belonging to the Upaniṣads (and thus to Vedāntas such as Śaṁkara's) and regards Bhartṛhari's engagement with it as prompted more by a desire to bring incidental or additional benefit to readers than by any organic connection it may have with the recognized concerns of Vyākaraṇa. [Back]

22 - (a) The remark most commonly cited in this context is satattvato 'nyathā prathā vikāra ity udīritaḥ / atattvato 'nyathā prathā vivarta ity udīrītaḥ //.

(b) Madeleine Biardeau (1964) has used the absence even as an argument in support of her view that the author of the Vākyapadīya kārikā (main verse) text must be different from the author of the Vṛtti commentary. The Vṛtti is ascribed to Bhartṛhari by a long tradition. [Back]

23 - Trayyanta is used in Vṛtti 1.10, 2.22, 2.233. 1 have pointed out the implications of the term in an as yet unpublished article on the term "vedānta." [Back]

24 - There is absence of contrast between pari + nam and vi + vṛt in the verse śabdasya pariṇāmo 'yam ity āmnāyavido viduḥ / chandobhya eva prathamam etad viśvaṁ vyavartata // because pari + nam was used in Bhartṛhari's time to speak of an effect or effects viewed as a single entity, and vi + vṛt was used when mention was to be made of several effects coming into being from a single cause, without really changing the nature of the cause. There was a difference of connotations (singularity and diversity) between the employment of the two terms, not a theoretical or philosophical difference (Aklujkar, forthcoming). [Back]

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

P.469

25 - Vṛṣabha derives the same import through a rather arbitrary interpretation of jagat. His explanatory gloss is: prakriyā iti prathamata utpattiḥ. jagataḥ iti sakalāgamopalakṣaṇam, "The word prakriyā stands for the first coming into existence. The word jagat is employed to speak of all inherited knowledge or lore." [Back]

26 - The reading of the ancient commentator Vṛṣabha is pratīyante, "are cognized," which would support my interpretation even more directly. However, since the exchange of t and k is found elsewhere in the manuscripts of Vṛṣabha's commentary, one cannot rule out the possibility that pratīyante is a corruption of prakriyante. Another possibility is that Vṛṣabha's original wording was prakriyante pratīyante with pratīyante intended as an explanation of prakriyante and that a haplography has occurred between the two pras. [Back]

27 - The preceding statements of this paragraph contain views that have a complex history in the tradition of Indian philosophy. In some cases, Bhartṛhari may be said to have borrowed the views; in others, he seems to have imparted them to others. Secondly, the acceptance of a view enumerated here does not mean that the remaining views were also acceptable to the provider or borrower. The same view or device appears in highly different configurations. For example, the implications of the view that a thing is not directly or in-itself accessible are drawn differently in a Yogācāra Buddhist philosopher's epistemology, particularly in his discussion of the means of cognition (pramāṇas). The Mīmāṁsakas, while generally being champions of śakti, do not, in their ontology, go to the extent of using it as a device for explaining our varying perceptions of physical things.

For one important way of uniting the word and the world, Bhartṛhari seems to have been indebted to the mahān ātman notion, probably developed by the early Sāṁkhyas. Furthermore, his thinking has a close connection with the proponents of Sattādvaita or Bhāvādvaita (e.g., Maṇḍana-miśra). These varieties of monism, in turn, seem to have been, to some extent, made possible by the Nirukta discussion of ṣaḍ bhāvavikāras, "six modes of action or state." These six are: jāyate, "is born"; asti, "is, exists"; vipariṇamate, "changes"; vardhate, "grows, increases"; apakṣīyate, "decreases"; and vinaśyati, "disappears, dies." A reduction to the following three is mentioned as the next step of analysis in later literature: bhū, "become"; kṛ, "does, makes, make become"; and as, "be, exist." The same literature speaks of reduction to as, "exists," as the final step. [Back]

28 - The word "experience" is used here as an unavoidable loose or approximate expression. There is no "experience" in the usual sense of the word in the state of liberation according to any school of Indian philosophy. In fact, there cannot be. If it were there, liberation would be no liberation. [Back]

29 - (a) The theoretical economy that Bhartṛhari thus achieves by collapsing the model needed for explaining the phenomenon of articulation and linguistic communication with the model needed to account for our experience of

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

P.470

the world -- by uniting the linguistic and the cognitive-psychological -- is a remarkable achievement. It can be charted as follows:

vaikharī, "language 1, speech"

madhyamā, "language 2" =jñāna,"cognition"

paśyantī, "language 3" =buddhi, "mind, intellect"

paraṁ paśyantyāḥ rūpam, sūkṣmā =cit/citi/caitanya "sentience,consciousness"

nityā atīndriyā vāc,vācaḥ uttamaṁ

rūpam,śabda-tattva-brahman,

"language principle,language 4"

The equations above require an expansion of the notion of language, but, as we have seen, the expansion can be defended and, in any case, is a small price to pay for the gains to be made.

(b) One may wonder about how Bhartṛhari deals with meaning conveyed through gesticulation or with sign languages. As his commentators clarify, Bhartṛhari takes language to be the most fundamental system of communication and views the conventions of gestures and so forth as ultimately based on language.

(c) If one were to concede that sentience or brahman can be "experienced" in itself-that there can be an experience which constitutes or resembles the mokṣa or nirvāṇa stage, one should note which such an experience is not linguistic in the sense of being accessible to or pervaded by specific linguistic expressions; it is incommunicable; language, in the usual senses of the term, ceases to exist in it. However, one should also note at the same time that the implication or feature identified does not make it unjustifiable to view the ultimate as the pure form of language. Extending to the fourth form a term applicable to the first three forms does not depend on the criterion "Do the characteristics of the first three forms hold in the fourth?" when the fourth is, by definition, pure-when it, by definition, transcends the first three. just as the applicability of the term to the fourth form cannot be proved through evidence of the usual kind, it cannot be precluded, either, by evidence of the usual kind. Whether to extend the term or not then becomes a question of neatness of theory. As we have seen above, something beyond language 3 needs to be assumed in order to account fully for the characteristics of language 3, particularly its capability for change of constituents. That this assumed thing should not change its own nature while enabling language 3 to be what it is follows logically. If the assumed thing also were to change, we will have to postulate another wellspring for it (and so on) and settle for a situation of infinite regress. It is better, then, to conclude our search for a "base" with just one base. We need to think of it as different in nature from language 3 yet related to the latter. This makes it sensible that the term "language" be extended to it even if it is free from the particulars and operations of a specific language. If the earlier

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

P.471

constituents and equations in the model are acceptable, there is no gain in postulating two mutually exclusive enities at the final or most fundamental juncture. [Back]

30 - Depending on one's starting point, one could also use the expression "a basis for" in this context. [Back]

31 - I use "proposition" simply to cover the content part. The problems seen in the notion of proposition in the Western tradition need not be thought of in the present context. [Back]

32 - Even if it were to be assumed that technology would enable us to measure the temporal lengths of cognitions, there would remain the theoretical problem of deciding how cognitions should be individuated-what the definition of a single cognition should be. [Back]

33 - Practically the whole second book, the real Vākyapadīya of his Trikāṇḍī, can be said to have been devoted to a consideration of the issue. [Back]

34 - There is, for the same reason, considerable room for individual preference in how they are set up. One is free to imagine them in any form as long as consistency and, preferably, economy, are maintained in the method of analysis. Words as a class must be thought of as associated with meaning from a beginningless time. If that is not held to be the case, the fact that individual words are invested with new meanings would be inexplicable. The investing would require language, and language would not serve the intended function unless it already has meaning-bearing words. Individual words have only relatively stable or settled meanings. These meanings always get adjusted or become specific in the context of a sentence. The key to solving paradoxes (which can also include the problems felt with statements of cause-and-effect relationship, etc.) lies in realizing the gap between the conventional or lexical meanings and the contextually adjusted meanings of words. [Back]

35 - However, external objects may be viewed as word meanings in those cases in which they are found to exist and to correspond to word meanings. [Back]


References

Aklujkar, Ashok. 1970. "The Philosophy of Bhartṛhari's Trikāṇḍī." Ph.D. dissertation. Cambridge, MA: Harvard University. Unpublished.

Bhartṛhari. Trikāṇḍī or Vākyapadīya. I have reproduced the text of the kārik ās and the Vṛtti from my critical edition under preparation. Those wishing to verify my references to the Vṛtti prior to the publication of my edition should consult Subramania lyer 1966 and Subramania lyer 1983. I have followed the enumeration of k ārik ās in Rau 1977; hence, the numbers in my edition and those in

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

P.472

the editions by Subramania lyer do not always match. However, they are not far removed from each other. For the third Kāṇḍa of the TK, which has a commentary by Helā-rāja, see Subramania lyer 1963 and 1973.

Bhattacharya, Bishnupada. 1985. Bhartṛhari's Vākyapadīya and Linguistic Monism. Post-graduate and Research Department Series No. 25 / Prof. K. V. Abhyankar Memorial Lectures (Third Series). Poona: Bhandarkar Oriental Research Institute.

Bhaṭṭoji-dīk ṣita. 1898. Śabda-kaustubha. Edited by Rama Krishna Shastri, alias Tatya Shastri Patwardhana. Benares: Chowkhamba Sanskrit Book-depot.

Biardeau, Madeleine, ed., trans. 1964. Vākyapadīya Brahmakāṇḍa avec la Vṛtti de Harivṛṣabha. Publications d l'Institute Civilisation Indienne. Série IN-8. Fasc. 24. Paris.

Chakrabarti, Arindam. 1992. "On Knowing by Being Told." Philosophy East and West 42 (3): 421-439.

Houben, Jan E. M. 1995. "Bhartṛhari's Perspectivism (2): Bhartṛhari on the Primary Unit of Language." In History and Rationality. The Skōvde Papers in the Historiography of Linguistics, pp. 29-62, edited by Klaus D. Dutz and Kjell-Ake Forsgren. Münster: Nodus.

___________. 1997 [1999]. "Bhartṛhari's Vākyapadīya and the ancient Vṛtti (2): The Vedic Background of the Author of the Vākyapadīya-Vṛtti." Studien zur Indologie und Iranistik 21 : 71 -77.

___________. 1997. "The Vṛtti and Bhartṛhari's Perspectivism in the First Kāṇḍā of the Vākyapadīya." In Beyond Orientalism: The Impact of the Work of W. Halbfass on Indian and Cross-cultural Studies, pp. 317-358, edited by Karin Preisendanz and Eli Franco. Amsterdam: Rodopi.

___________. 1998. "Bhartṛhari's Vākyapadīya and the Ancient Vṛtti (1): The Vṛtti and Vṛṣabha-deva's Paddhati on Vākyapadīya 1.46a ātma-bhedaṁ / ātma-bhedas...." Annals of the Bhandarkar Oriental Research Institute 78:177-198.

___________. 1999. "On Syntactic and Stylistic Evidence Regarding the Authorship of the Vakyapadīya-Vṛtti (Bhartṛhari's Vākyapadīya and the ancient Vṛtti 3)." Wiener Zeitschrift für die Kunde Südasiens 43:167-197.

Ramaswami Sastri, V. A. 1938. "Bhartṛhari -- A Pre-Śaṅkara Advaitin." Annamalai University Journal 8:42-53. Also in the Proceedings of the All-India Oriental Conference (Trivandrum, 1937) 9: 548-562. Trivandrum, 1940.

Rau, Wilhelm, ed. 1977. Bhartṛharis Vākyapadīya: Die Mūla-kārikās nach den Handschriften herausgegeben und mit einem Pāda-Index versehen. Abhandlungen für die Kunde des Morgenlandes XLII, 4. Wiesbaden: Kommissionsverlag Franz Steiner GMBH.

Sastri, Gaurinath. 1968 [1969]. "Monism of Bhartṛhari." In Beiträge zur Geistesgechichte Indiens: Festschrift für Erich Frauwallner, hrsg. von G. Oberhammer,

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

P.473

pp. 319-322. Wiener Zeitschrift für die Kunde Süd- and Ostasiens and Archiv für indische Philosophie, Bd. 12-13. Leiden: E. J. Brill.

Subramania lyer, K. A., ed. 1963. Vākyāpadīya of Bhartṛhari with the Commentary of Helā-rāja. Kāṇḍa III, Part 1 [samudeśas 1-71. Deccan College Monograph Series no. 21. Poona: Deccan College Postgraduate and Research Institute.

___________.1966. Vākyāpadīya of Bhartṛhari with the Vṛtti and the Paddhati of Vṛṣabhadeva. Deccan College Monograph Series no. 32. Poona: Deccan College Postgraduate and Research Institute.

___________.1973.Vākyāpadīya of Bhartṛhari with the Prakīrṇaka-prakāśā of Helārāja. Kāṇḍa III, Part II [samudeśas 8-14]. [Continuation of Deccan College Monograph Series, no. 21?] Poona: Deccan College.

___________.1983. The Vākyāpadīya of Bhartṛhari, Kāṇḍa II with the Commentary of Puṇya-rāja and the Ancient Vṛtti. Delhi: Motilal Banarsidass.