Semantics of Nothingness Negation 2
Semantics of Nothingness: Bhartrhari’s Philosophy of Negation – II

Bhartrhari’s treatment of negation as ultimately grounded on being, and his assertion that there is no absolute negation of speech, removes it from the paradox that underlies negating something.

Compounds with Negative Particles and the Metaphysics of Negation

An oft-cited verse identifies six different meanings of the negative particle:

The negative particle has six different meanings: similarity with and contradiction to X, otherness and diminution, reproach and absence.

Accordingly, the negative particle expresses the following meanings, depending upon the context:

  1. Similarity: e.g. the term anaśva (non-horse) refers to an animal that is similar to but exclusive of a horse, such as a mule, donkey, or zebra.
  2. Contradiction: e.g. the term apunya (non-virtue) refers to vice.
  3. Otherness: e.g. abrāhmana (non-Brahmin) refers to Ksatriyas etc.
  4. Diminution: e.g. anudarā refers to a girl with a slim waist.
  5. Impropriety or reproach: e.g. apaśavah (non-animals) used to refer to the animals that are not to be sacrificed.
  6. Absence: e.g. abrāhmano grāmah (a village without Brahmins) demonstrates the absence of Brahmins in a village.

Opinions vary about which meaning is primary. Kaunda Bhatta maintains that only the final one, absence, is the primary meaning of the negative particle, the rest being secondary (Coward and Raja 1990, 288). According to Naiyāyikas, both contradiction and absence are the primary meanings of the negative particle. Patañjali’s commentary on the Astādhyāyī (AA) 2.2.6 also raises issues relevant to understanding the meaning of negation in a compound term. A section in Vrttisamuddheśa (VS 248–315) reads as a commentary upon Patañjali’s MB (AA 2.2.6). 14

Patañjali argues that the term abrāhmana can be explained in three different ways, being dependent upon the primacy of the second term (or brāhmana), an external term (someone else being referred to by this term), or the first term (the negative particle). 15 When reading Patañjali, it becomes clear that he considers the second term to be primary and he also defends the position that, in different contexts, even the negative particle can be primary. The position that an external term is primary has here been abandoned altogether. Bhartrhari’s treatment of negation in a compound (VS 248–315) primarily rests on these assumptions. We should not, however, conclude that all grammarians maintain the same position. 16

The issue of which among the parts of a compound is primary in comprehending its meaning is broadly philosophical, and what has been addressed by classical philosophers about negation rests on their underlying metaphysics and epistemology. However we interpret negation, all the interpretations provided lead to the question of whether negation negates something that exists or whether it negates something nonexistent. In his attempt to resolve the apparent paradox, Bhartrhari goes back to the absolute vāc, the reality constituting speech, and argues that even when negation is affirmed, the affirmation represents the foundation of all epistemic and ontological considerations. When we say, “There is no book on the table,” the negation relates to the book that has no existence while the application of the term ‘book’ affirms its categorical existence. Negation can negate only things that exist, and at the time of negation, there is no existence: a paradoxical situation. It would, however, be too hasty to draw further parallels. Sanskrit grammarians escape the paradox by adding one level to existence: the superimposed existence (upacārasattā). 17 The existence of what has been negated, accordingly, is superimposed upon what is being negated. Rather than nañ, the negative particle that has its own reference, the negative particle becomes a coreference (dyotaka), 18 indicating that the term ‘book’ refers both to its existence and nonexistence, and the particle indicates negation that is linked with the term ‘book.’ This position relates to the earlier argument regarding the primacy of the terms if compounded with nañ, with an application that the negative particle does not have an independent meaning.

The tradition of the Sanskrit grammarians weighs the primacy of the second term in a compound with the negative particle. 19 The issue, however, is how can something that does not exist (e.g., a brāhmana in the case of abrāhmana) be primary? When even the primacy of the first or second term in the compound is called into question, the primacy of an external term is impossible. Unlike sentential negation, this negation cannot give meaning by corresponding to the verb either. When, for instance, we say, “There is an asymmetrical pattern on the wall,” we cannot derive the meaning that there is the absence of a symmetrical pattern, but rather that there is existence of a pattern that is asymmetrical.

Bhartrhari argues on this ground that this negation can only correspond to the subject and not to the verb (VS 251). This suggestion of the grammarians that the negative particle has only coreferentiality further distances them from the way Parmenides understood negation. Rather than accepting negation as a case of paradox, the approaches of the grammarians appear to resolve it by advancing different epistemic and ontic structures.

This position, however, does not satisfy Bhartrhari, who instead invites further scrutiny. If nañ is just coreferential, the issue of it having three different positions with regard to the primacy of meaning in a compound would not even arise. 20 If the argument is that the action that is subordinate to the agent in a compound, such as pācaka, is what the negative particle corresponds to when in a compound, since there is not even a subordinate action expressed by the term brāhman a, the negative particle in the compound abrāhmana would not corefer to anything (VS and VST 252). Helārāja (980 CE) concludes on this ground that in the case of compounds (and not in sentences), the particle has to be understood as referring directly to something. 21 What is directly negated then, when in a compound, is the existence of its referent (given that this existence is superimposed).

It has been postulated that being (sat) is given with every single term and what a term ‘book’ means, for instance, is ‘a book exists.’ Bhartrhari presents a position that rejects this understanding and instead maintains that the term nañ does not express nonexistence (asattā) but only prohibits (nisedha) what has been postulated by its referent. This position of the negative particle referring to prohibition, rather than negation, stems from the following reasoning:

  1. Existence is embedded with all terms (as all terms refer to something that exists).
  2. The term existence (sattā), if the above position is correct, then refers to the existence of existence (as existence is given to all terms including the term ‘existence’ itself).
  3. The negation of existence (asattā), then, would mean the negation of the existence of existence, a tautology (VS 253).

In Sanskrit, one can make compounds where the suffix expresses repetition of action, and so the meaning of the suffix is similar to that conveyed by the verbal root. Verbs, it must be noted, convey process. However, negation can be linked only with what can be negated, and if the negative particle is to be understood as nonexistence, there would be no relation between it and the referent term. The conclusion, then, is that the negative particle, instead of meaning nonexistence, only means negation. 22 By relying on the above-mentioned argument that words directly refer to reality, Bhartrhari escapes the tautology that the term ‘existence’ refers to the existence of existence.

The above argument, however, is not applicable in the case of abrāhmana, where existence is part of the base term (brāhmana) and what exists cannot be negated. This raises a bigger issue. One can make the same argument regarding the terms asan or akrtvā, as they both refer to a positive nature (bhāva). This leads to the argument that the negative particle is merely coreferential rather than having its own primary meaning. Helārāja points out that there are two options, even when the particle is understood as coreferential:

(1) nañ is understood as revealing an object which is the substratum of a particular action.

(2) It is understood as revealing an object which is the substratum of objects in general (Iyer 1974, 240).

The problem is, if the first position is maintained, a compound with the negative particle would not be possible, as there would be no semantic connection. Even if some connection is maintained, the three possibilities of interpreting negation will not arise. In the second alternative, if the action in general coreferred to by nañ is linked with the second term expressing only the objects in general, this will lead to an infinite regression. It cannot be linked to the second term that is expressive only of a particular. If the particle is considered to be coreferring, it will be linked with the meaning that is coreferred (dyotya) to, and there will only be primacy of the second term. Either way, the possibility of three alternative meanings of negation will not arise in this case. Only when negation is considered as having its own independent meaning among the terms that have come in the compound word would it make sense to argue whether the negative particle or the second term is primary.

Relating to the previous conversation, whether nañ refers to nonexistence or negation, Bhartrhari maintains that even when the meaning of the particle is maintained as nonexistence, there will be no consequence in examples such as asan or akrtvā. Bhartrhari argues:

The negative particle relates to the generic substrate of the agency of action. Due to this, [the particle] is [also] linked with particular substrates such as brāhmana [in the example abrāhmana]. (VS 255)

When the negative particle is explained as referring to nonexistence, absence is explained in terms of its substrate. Since the particle is explained as referring to the substrate, existence in general, there will be no consequence in the example asan. Since the verb √ as is intransitive, the instrument (sādhana) that accomplishes this action is the very agent itself. Accordingly, absence is explained in terms of its substrate (i.e., absence is the absence of something). The particle, then, is negation in general. The words that accompany the particle particularize negation. As Iyer summarizes, “In a compound, the negative particle is expressive of the substratum in general, colored by non-existence, of the action of existence. So the meaning of nañ amounts to nāsti = ‘it does not exist’” (Iyer 1974, 241). Accordingly, in terms such as abrāhmana, the negative particle expresses the nonexistence of the substrate in general and the term brāhmana refers to a particular substrate.

From this emerges the crux of the problem: when two meanings are combined, which meaning is primary, the particle referring to the substratum in general, or the second term referring to the particular? The issue is this: if the particle is interpreted as prohibition, three possibilities (of the primacy of the second term in a compound, the primacy of the external term, and the primacy of negation or the particle itself) would not arise. When negation in general is linked with a particular entity, with negation referring to its substrate, or when the substratum is negated, this raises the issue of the primacy of the terms in a compound. That is, shall this be understood as ~(P), or P (~), or ~ P (Q), where Q stands for an external term?

It needs to be explained how the particle corefers in a sentence while directly referring to something in a word. Bhartrhari adds that particles do not just negate; they also involve action. For instance, when one uses the term ‘nis kauśāmbih ‘(one who has exited out of Kauśāmbi), the prefix nis does not simply stand for negation but also for the act of exiting (VS 256). The particle here does not just denote nonexistence, but existence. If the particle here were to simply negate existence, it could not combine with the second term that refers to being. It is therefore reasonable to say that the compound term denotes the agent of the action of exiting from Kauśāmbi. This leads to the issue of which meaning is primary. It is commonly seen that when there are multiple qualifiers, it is upon the speaker to decide which among them is primary. 23 This makes the discussion of three alternative ways of describing negation relevant. We shall now turn to these three possibilities.

(1) The Primacy of the Final Term in a Compound

A significant section in VS (259–315) focuses on the primacy of meaning among terms when a term is combined with a negative particle. 24 In this discourse, Bhartrhari first examines the case where there is a primacy of the final term. In the example “non-Brahmana” (abrāhmana), while the negative particle refers to nonexistence, the second term describes a positive entity. When primacy is given to the second term, this means ‘someone in whom the quality of being a brāhmana does not exist.’ While the particle nañ refers to nonexistence in general, brāhmana particularizes this nonbeing. Bhartrhari compares this negation with any other qualifier. If we say, “a fair lady,” the term ‘fair’ makes explicit what is implicit in the lady, and the term ‘fair’ does not invent something what is not in the lady.

One can argue that if the very existence of brāhmana has been denied by the particle, why would someone utilize the positive term that also contains the meaning of existence? Bhartrhari addresses this question by maintaining a distinction between language and reality. In language, we do use san brāhmana, where existence that has already been affirmed by the term brāhmana, is twice affirmed by the use of the qualifier sat (VS 261). Along these lines, the negative particle as a qualifier and that which is being negated are thus inherently related and inseparable. When we say, “a blue lily,” there is no such thing as ‘blue’ that is distinct from the lily. In language or in conceptualization, however, this distinction is obvious. When a compound with a negative particle is used, Bhartrhari argues, a mistaken identity is corrected (VS 262). When we say, for instance, “This dairy is non-organic,” we are correcting a mistaken identity (or a presumed mistaken identity) that the dairy was organic.

A problem, however, still remains. In terms such as abhāva, nonexistence, what is comprehended by negation is not “some other entity similar to existence” (bhāva). Rather, a simple negation of existence is comprehended. In such instances, Bhartrhari maintains that the meaning is comprehended by assuming a cognitive entity even in the absence of its external reference (VS 263). This explanation of negative compounds rests on “superimposed existence,” rather than the real entity. In other words, every word, when pronounced, gives a mental presence of the entity, and negation denies the entity having a real reference. Here, Bhartrhari deviates from Patañjali, who maintains that a word, brāhmana for instance, refers to a constellation of properties, and by means of referring to the constellation of properties, it also stands for the particular properties. 25 Bhartrhari does present this position (VS 264), albeit as an alternative to what he has presented before. The negative particle nañ, along these lines, reveals the absence of those properties, and accordingly, the term abrāhmana describes the lack of the properties of a brāhmana in someone other than the brāhmana (VS 264). In other words, in the term abrāhmana, the term brāhmana stands for the partial qualities of a brāhmana. A question emerges: if the negative particle refers to the absence of properties in its substrate, how can it be connected with the term in making a compound? The question is, the part that has been negated is what has not even been described by the second term. When we say abrāhmana, the nañ is negating the properties that do not exist in its referent, ksatriya (a warrior) for instance. Bhartrhari responds to this by saying that when something is half-accomplished, both terms are used to denote it, as accomplished and not accomplished (VS 265). What he means is, when we use the term ‘abrāhmana,’ it refers to someone who is neither a brāhmana nor someone that entirely lacks the properties of a brāhmana.

This response, however, invites further problems. Even the term brāhmana would then refer to abrāhmana, for there may be some expected properties lacking in any individual brāhmana. Furthermore, there will be no distinction between the tasks that have been fully accomplished and those half accomplished, if the term ‘accomplished’ is used in such a loose sense (VS 266). Bhartrhari responds to this objection by showing how a term can be used in both a primary and secondary sense. Brāhman a, for instance, is used to denote a person with a particular blood lineage, and to describe the person’s tendency to ascetic practices and studies. There is a secondary use of the term to describe someone with a tendency of austerities and studies. Bhartrhari says that the term abrāhmana is used to negate the primary meaning in a subject who has some qualities of a brāhmana (VS 267). This exposition is required particularly when the compound is interpreted with respect to the primacy of the second term. If the negative particle were to negate the meaning of the second term, the particle could not be considered an attribute, because it does not reject its substrate but only qualifies it. Along these lines, a negative particle only negates a part of the attributes among those expressed by the second term in a generic sense and thus the remaining attributes describe the referent (VS 268). When we use a term ‘non-chemical’ to refer to something, the entity being referred to does not have a total absence of chemical properties. Instead, it only lacks certain properties, such as toxicity.

Another problem arises. There may be a lack of one or another property in every brāhmana. The above interpretation of the term abrāhmana with a negative particle presumes that the term brāhmana refers to one with all the properties assigned to a brāhmana, and its negation will relate to someone with only some of such properties. In the absence of such brāhmanas that have all the assigned properties, this negative particle will be meaningless (VS 269–70). Bhartrhari responds, when in a compound, the term loses its primary meaning and is used only in the secondary sense. As has been described earlier, Bhartrhari maintains that both existence and nonexistence are embedded in the terminal meaning and the negative particle only underscores nonexistence that is intrinsic to the word meaning. He further explains that the same is found in other compounds, such as a ‘king’s officer,’ where the qualifier stands for something that already exists in what is qualified (VS 271–2).

However, a problem still persists. How can a term be used to refer to something that has only partial properties? 26 Bhartrhari responds to the objection by saying that a word in a compound cannot exclude both its primary meaning and the reference by the negative particle that is identified by the term (VS 277–8). Independent of external referents, words convey the meaning that is linked with their concepts, and negation, along these lines, limits the scope of meaning of the word only to the negative aspect (VS 279).

Bhartrhari claims, following a similar argument, that the term aneka can also be explained where the second term means ‘one,’ and the compound stands for ‘many.’ If the second term were not primary, he argues, the term would not be declined similar to the term eka, and the formations such as anekasmai would not be possible (VS 281–3). Bhartrhari explains that, just as abrāhmana describes some qualities ascribed to a brāhmana, oneness is superimposed in what is more than one (VS 284). This passage is an explicit reference that describes a superimposed concept on the primary meaning. What the negative particle represents in such instances, according to Bhartrhari, is the removal of the error that has superimposed oneness onto something that is more than one. This negation, therefore, does not negate the word ‘one’ but rejects the erroneous cognition of oneness (VS 285). In his understanding, the negative particle in a compound aneka is similar to its use in a sentence: such as, “This is not one,” where the term ‘one’ refers to other than one due to its association with ‘not’ (VS 286). Although the meaning of the term ‘one’ is not what is derived from ‘not-one,’ the second term is used only to reject what has been superimposed on a substrate that is not one. Just as the term ‘non-white’ refers to something that is of any color but white, so does the term ‘aneka ‘refer to something that is not singular (VS 287–8).

This conversation reinforces Bhartrhari’s underlying philosophy that terms directly refer to concepts rather than external entities. When one term is uttered, all possible meanings have the potential to manifest, and the negative particle restricts meaning, limiting it only to what has been meant. When one says “Seat them all,” the first term, ‘seat,’ does not express the scope of the imperative. In the same way, anekam āsaya (seat no-one) restricts the scope of the injunction (VS 289). A mere negation of number cannot be the meaning of aneka, as the term is understood in the meaning of the number that is more than one. Accordingly, this cannot be interpreted as the nonimplicative (prasajya) type of negation (VS 290). Whether in a sentence or in a compound, this category of negation needs to be explained along these lines (VS 294). When one says, “I do not take sugar,” he may still take a sugar substitute. If what is meant is a total rejection of all that is sweet, that is, a nonimplicative negation, the negation would be different from when a substitute is comprehended through negation, a case of implicative (paryudāsa) negation.

(2) The Primacy of the External Term

Having extensively addressed the argument that a compound with a negative particle places primacy in the second term, Bhartrhari addresses the position that there can also be primacy in an external term in a compound with a negative particle (VS 296–315). Although it does have a certain logical appeal to it, this alternative seems to have been abandoned by Patañjali as well as by Bhartrhari. The argument in favor of the primacy of the external term goes along these lines. The term abrāhmana has been analyzed in earlier conversation by accepting that the second term in the compound denotes the superimposed meaning, rather than the meaning of the term itself. This is to say that the term ksatriya (or one like it) is understood to be superimposed on the term brāhmana, and when the terminal meaning is derived, there is still the primacy of the second term in a negative compound. However, if the negative particle is interpreted as a qualifier, with the second term being qualified by it, the meaning derived is something external. Accordingly, the term abrāhmana would be interpreted as “someone who lacks the properties that qualify a brāhmana.” In this unique situation, although the negative particle in general denotes nonexistence, it denotes the same external object as does the second term that is positive in essence, and thus the negative particle and the positive term both have the same substrate, affirming the primacy of an external term. 27 In this regard, this compound functions similarly to a relative or adjectival compound.

If the view that words refer to a class and not particular entities is adopted, its negation relates to a particular entity external to both terms in the compound when the second term refers to the class (VS 302). When this position is followed, a clear distinction can be made between a compound with a negative particle and a relative compound. In Sanskrit, there are two ways a compound with the negative particle can be made:

(1) If the term is referring to someone who lacks cows, the compound will be agu (a case of the bahuvrīhi).

(2) If the term is referring to an animal that is not a cow, the compound will be agau (a case of the tatpurusa) (VS 303).

This is to show that the primacy of the external entity cannot be confused with that of the second term in compounds.

(3) The Primacy of the Negative Particle

Bhartrhari eventually addresses the position that, in a compound, primacy belongs to the negative particle which is the first term. If this position is followed, the positive term, for instance brāhmana, would qualify the negative particle, since the meaning would be derived with the primacy of negation (VS 305). An interlocutor raises the question: if negation is considered the terminal meaning, how can it be linked with action, the verbal meaning, as in the case of ‘feed the nonbrāhmana ’? Bhartrhari replies, that what is considered as nonexistent also exists in another way. In other words, while negation is primary, the verb still relates to something existent, a case of implicative negation (VS 306). Bhartrhari points out that there should be no confusion over the issue of the primacy of negation in a compound, as the Sanskrit language offers two different types of compounds that explicitly highlight this difference. 28 Words have innate power, but unlike in a sentence where a negative particle can simply negate something, they denote what has been negated when applied in compounds, and thus what has been negated has primacy over the particle or negation as its meaning. Because of this, the compound terms follow the gender and number of the final term (VS 308). Furthermore, if they are not related to verbs, there would be no grammatical applicability of such terms, and sentences such as “bring a nonbrāhmin” would have no meaning (VS 309).

The objection regarding the gender of the term, that the term should be indeclinable if the negative particle is primary, is also not tenable, as the gender, following Sanskrit grammar, is considered inherent to words themselves (VS 310). The genders of name-words, along these lines, are not necessarily congruent with the objects they identify, since gender in Sanskrit is grammatical and not natural. The term dāra, referring to wife, for example, is in the masculine. The term ksātra has a neuter gender and possesses the same meaning as the term ksatriya (warrior) in the masculine. Accordingly, the nonexistence that is expressed by the negative particle is comprehended in the form of some positive substrate, along with the gender and the number which correspond to the substrate (VS 312). Bhartrhari concludes by maintaining that the terms that refer to two distinct entities refer to a single object in a compound, and since the negative prefix represents all the deviations from the second term, there is primacy of the first term (VS 313). The second term, brāhmana in the aforementioned example, is only coreferring to what is meant. 29 Consequently, the second term expresses the particular from among what have been generically referred to by the first term.

Bhartrhari assimilates the position that there is primacy of an external term within the perspective of the primacy of the first term, arguing that the primacy of negation also accommodates the arguments for the primacy of the external term (VS 314). Helārāja expounds on this by saying that the notion that the external term has primacy in negation arises due to error. 30 If the immediate term after a compound is a noun, the primacy of that noun is assumed due to association, and in cases where it is not, a pronoun is superimposed. In either case, the primacy of the external term does not stand a chance. The issue regarding gender and number is followed in this case with a meta-rule (atideśa) of negation according to which negation implies a superimposition of the properties of an associated term. Bhartrhari points out that the same problems regarding gender and number that forced one alternative to be dropped solidify the alternative that there is primacy in the first term. Following this argument, if the second term is not predominant, the number and gender of the compound cannot be regulated (VS 315). The argument is, in the absence of the compound term qualifying a second term, the term should be referring to something generic, and when there is no particular as a referent, the term should be used in neuter gender singular. 31 Helārāja explicitly concludes on this ground that the only supportable position to be maintained is that there is primacy in the second term. 32 He also rejects that gender and number are understood with superimposition. He gives an example in which both the cases of the external term being in neuter and masculine, the compound term remains in the masculine (VST 315). This argument, however, does not counter the argument that gender and number are inherent with the term. On this ground, Helārāja, returning to the primacy of the first term, maintains that even this position should not be abandoned. 33 He also points out that Patañjali has supported this argument by refuting the objections against it. 34 This, then, allows the declination of the compound term with a pronoun. 35 Helārāja’s final word on this matter is that the primacy of the first term is established. 36 Therefore, it would be wrong to countenance that there is only the primacy of the final term in the particular compound under consideration and not a primacy to negation.


This discussion of Sanskrit semantics has multiple philosophical implications. Bhartrhari’s treatment of negation as ultimately grounded on being, and his assertion that there is no absolute negation of speech, removes it from the paradox that underlies negating something. This speech, or vāc, of Bhartrhari is not just a means of communication though. When speech is identified with the absolute, the Brahman, the consequence is that no negation of the foundational being is possible, a rejection of the Nāgārjunian approach.

Bhartrhari’s treatment of sentence negation and word negation further illuminate other issues. It has been discussed above that negation in a sentence relates to the verb and negation in compounds relate to the second word. The consequence is that word negations do not simply negate something. Bhartrhari’s logic rests on three-tiered negation:

P = Q P = ~ = P’

(Where P’ stands for something that is neither identical to P nor is absolutely different from it, in the sense that P’ shares many of the constituents of P but not all.) While Bhartrhari rejects the position that ~ P = Q, this is only in the context of the compound terms. His analysis of three-tiered negation (which stems from Patañjali’s analysis) still has relevance in understanding negation in the issues outside of semantics.

The consequence of Bhartrhari’s conclusion is that ~ P = P’ leaves negation as affirming something existing. For the Advaitins, ‘ignorance’ (avidyā) is of the essential character of being (bhāvarūpa). This understanding of ignorance as something phenomenal (while not having its own intrinsic being) would be semantically impossible had not the philosophy of language allowed such interpretation. Along the same lines, the SvātantrikaPrāsangika discourse on negation also stems from the semantic issue of whether the negative terms simply negate being or affirm something else. The discourse on language is therefore pivotal to understanding a wide range of philosophical issues that originated in classical India. Although I have restricted myself to the philosophy of Bhartrhari, his answers to the issues regarding negation are relevant for a wider discourse not only on language but also epistemology.


AA = Astādhyāyī MB = Mahābhāsya of Patañjali MS = Mīmāmsāsūtra T = Tractatus Logico-Philosophicus VP = Vākyapadīya VPT = Vākyapadīyatīkā of Helārāja VS = Vrttisamuddeśa (Vākyapadīya 3.14) VSM = Vaiyākaranasiddhāntamañjūsā VST = Vrttisamuddeśatīkā of Helārāja


  1. I am thankful to Drs. JeeLoo Liu, Douglas Berger, Madhav Sharan Upadhyaya, Diwakar Acharya, David Buchta, and Boris Marjanovich for their valuable insights. I am equally thankful to Mrs. Mary Hicks for reading an earlier draft of this chapter.
  2. See Timalsina (2009a and 2009b).
  3. I am directly referring to Rorty (1992, particularly the Introduction pp. 1–39).
  4. Patañjali’s MB on AA 8.4.47. According to Kaiyata, this reading is corrupt due to a copier’s error. Even in his suggested reading, both terms of negation are present. Cf. AA 8.4.47.
  5. Jaimini’s terminology for threefold negation is pratisedha, vikapa, and paryudāsa. For a detailed analysis of negation in Mīmāmsā, see MS 10.8.1–22. See Staal (1962, 52–71) for the Mīmāmsā and Vyākarana exegeses of negation. Stall (1962, 61) analyzes two types of paryudāsa: 1) “the door should be unlocked”; and 2) “another door should be locked,” basically expanding upon the concept of vikalpa and paryudāsa. While pratisedha and vikalpa are spelled in MS 10.8.1, itaraparyudāsa is mentioned in MS 10.8.15. That Pānini does consider vikalpa as a type of negation is affirmed by ‘na veti vibhāsā ‘AA 1.1.44.
  6. While the particles mā / mān refer to negation, this imperative negation requires a much wider discussion that involves the ritual theory of Mīmāmsā, and due to limitation of space, I am not addressing this aspect of negation in this paper. In English, besides ‘not’ and ‘no,’ the varied appearances of negation in compound words such as unimaginative, non-reliable, atheist, incomplete or dysfunctional can be included in Sanskrit nañ.
  7. Cited in Mīmāmsānyāyaprakāśa (262, Nisedha section). For prasajya and paryudāsa, see also VP (2.84). A slight variant of this verse is cited by Staal (1962).
  8. What underlies Bhartrhari’s argument is that all negation rests on something positive as its foundation. Nyāya philosophers argue that the absence of a jar, for instance, is located on the surface upon which the jar is negated. Advaita Vedantins argue that Brahman is the foundation upon which entities are affirmed or negated but which in itself cannot be negated. Bhartrhari here appears to counter the Buddhist doctrine of emptiness, particularly the one grounded on Nāgārjunian arguments that rest on negation. Bhartrhari’s discussion of sentential negation also has a great relevance for the historical analysis of the Svātantrika and Prāsangika arguments of negation.
  9. When Pānini gives a rule that prātipadika is any meaningful word that is exclusive of the verb and the suffix (AA 1.2.45), implicative negation is implicit. In another rule (AA 8.4.47), Pānini states that all the consonants except for / h/ that are subsequent to a vowel can be optionally duplicated, and so both formations (e.g., klpta and klppta) are correct. However, this rule does not apply to those consonants if there is a vowel subsequent to it. This negation is non-implicative, and in this case the sentence is divided into two when analyzing meaning.
  10. The Sanskrit sentence is: na kalañjam bhaksayet | It is likely that kalañja meant the red garlic. For discussion, see Edgerton (1986, 164n213).
  11. For a detailed treatment of this sentence in the context of negation, see Staal (1962, 59).
  12. This argument, evident in the following verse cited by Punyarāja, resonates the position of Dharmakīrti:

satām ca na nisedho ’sti so ’satsu ca na vidyate| jagaty anena nyāyena nañarthah pralayan ̇gatah || Pramān aviniścaya 226. Cited by Punyarāja in his commentary (VP 2.241).

  1. VP 2.242.
  2. Most of the philosophical issues raised by Bhartrhari while addressing negation are seminally present in MB in AA 2.2.6. There is a stylistic difference in their presentation though. Patañjali first addresses the primacy of the second term, and with a brief objection, goes on to address the sides of the primacy of the external and the first terms. Bhartrhari, on the other hand, addresses the central position, the primacy of the second term, and then addresses the positions of the primacy of external and first terms.

For a summary of this section, see Iyer (1992, 390–401). For this discussion, I have primarily relied on Helārāja’s commentary for understanding the VP passage. I have also utilized Iyer (1974, 236–271) in summarizing the concepts.

  1. See MB (AA 2.2.6) for discussion.
  2. This is visible in Nāgeśa’s statement. See Uddyota on MB (AA 2.2.6).
  3. Helārāja uses the term in this particular context: (upacārasattā VP 3.14.250).
  4. See VP 3.14. 250cd.
  5. Pānini reads nañ (AA 2.2.6) as an independent sūtra in the sequence of addressing tatpurusa, a type of compound where the second term is primary. Patañjali’s elaboration upon the passage maintains the same position, demonstrating the primacy of the second term.
  6. VP 3.14.250.
  7. VP 3.14.252.
  8. In the case of terms such as akrtvā, the particle simply negates what is denoted by the verbal root, and so the relation between the negative particle and the referent term is established. Based on VS 254 and Helārāja’s Tīkā thereon.
  9. For discussion, see VS 257–258.
  10. Bhartrhari raises a wide range of issues, primarily inspired by the MB, with regard to the meaning of the negative particle in a compound. For the primacy of the final term in a compound with a negative particle, see VS 259–295; for the primacy of the external term, see Sam bandhasamuddeśa 296–304; and for the primacy of the first term or the negative particle, see VS 305–315.
  11. Mahābhāsya on AA 2.2.6. This passage is read with ‘hi’ instead of ‘ca’ in VST 264.
  12. The situation is even worse with terms that are used in two opposite senses. For instance, the term ārāt is used to refer to both ‘near’ and ‘far’ (VS 273–5). Bhartrhari clarifies his position by adding that, while the base words have a wide range of meanings, particles exclude certain aspects. In this sense, a negative particle only highlights what is already there in the term itself. He gives an example of the root √ sthā, which has both the meanings of staying and going away, but the second meaning is manifest only when the term is combined with the prefix pra (as in prasthāna) (VS 276).
  13. See Helārāja’s exposition for illustration. VP 3.14.296.
  14. Bhartrhari addresses the problem with an acceptance of the above position as follows. When the negative particle is primary in a compound, the formation would be ‘abrāhmanya,’ which is an indeclinable (avyaya), while a formations such as abrāhmanāh etc. would not be possible.
  15. VST 313.
  16. VST 314.
  17. Although verse VS 315 comes in a sequence of arguments rejecting the position that there is primacy in the external term, in the discussion initiated in VS 314, Helāraja expands on this and also defends the position that there is a primacy of the first term. Helārāja provides several examples in the list where the gender of the qualifier does not match that of the qualified. See VST 315.
  18. VST 315.
  19. VST 315.
  20. VST 315.
  21. In consequence, formation of the terms such as asarvasmai remains possible.
  22. tasmād upapannam pūrvapadārthaprādhānyadarśanam | VST 315. Noteworthy in this context is that Iyer (1974, 270–1; 1992, 399–40) considers that grammarians maintained only the position of the primacy of the final term. This understanding not only contradicts Helārāja and in contemporary times, the commentator Raghunātha Śarmā, this goes explicitly against Patañjali’s defense of this position by maintaining that the gender and number of the terms are innate and not controlled by any external terms.

The paper appears as a chapter in the book “Nothingness in Asian Philosophy” edited by JeeLoo Liu and Douglas Berger and has been republished with author’s permission.

Featured Image:

Disclaimer: The facts and opinions expressed within this article are the personal opinions of the author. IndiaFacts does not assume any responsibility or liability for the accuracy, completeness,suitability,or validity of any information in this article.
  • Sakameedha

    Beautiful explanations. As a person who is studying mimamsa and vyakarana i can tell that not many articles dig this deep. Keep it up