Volume 4 : 4 April 2004

Anirban Dash, Ph.D.



Semantikos is a Greek word derived from 'sema' (sign) going back to the Indo-European 'dhiei' (to see), which is paralleled by 'OIA dhyaanam' (introspection) and the reduplicated form from Persian 'deedan' (to see). Sign has come to mean a word, which is the symbol of expression, the symbol denoting an object. In this connection we can also compare the word 'varNa'- which originally meant 'colour' a sign, and then a sound or a letter.


India has occupied the highest position in the field of semantics. The following statement of Prof. M. B. Emeneau may prove a source of encouragement and inspiration to many:

"Certainly in one other slowly awakening department of Linguistics, that is concerned with meaning, the west still has to learn from India. Their grammarians, literary theoreticians and philosophers were all concerned with problems of meaning, and much was thought and written on the subject. Of this, the west is for all practical linguistic purposes innocent. The Hindu treatises are in a difficult style, and only a few in the west will be qualified to deal with them, as Sanskritists, philosophers and linguistic scholars. Yet, the results are likely to be worth the efforts: It is the subject that can be recommended to aspirants." [vi]


The Sanskrit term vyaakaraNa, which dissolves words into elements (stem and suffix) and thereby bringing out their exact meanings, is, in itself, an indication that grammar, etymology and semantics are intimately connected.

The ancient Indian scholars have recognized grammar as one of the eight[vii] methods of learning the meaning of words:

The other seven methods are:

  1. Lokavyavahaara (Popular usage)
  2. Aptavaakya (direct statement of trustworthy authority)
  3. Upamaana (analogy)
  4. Kosa (lexicon)
  5. Vaakyasesa (the rest of the passage in the context)
  6. Vivruti (explanation)
  7. Siddhapada-saannidhya (syntactic connection with words already known)


Grammar explains what a correct sentence is. But semantics tells us the full significance of a sentence with all its implications. There are many idioms and paraphrases in a language, which may be grammatically incorrect but semantically most expressive.

The term used in ancient days to denote a grammarian seems to have been vaagyogavid [viii], that is, one who knows the connection and usage of words. The word Shaabda-shaastra for grammar clearly suggests how closely semantics and grammar are related. AAchaarya ShaNkara explains that grammar, by division of words, enables us to comprehend their meaning, and in the modern times, Jesperson, in his book Philosophy of Grammar, argued that 'for a clear understanding of grammar, a psychological study of language is most essential'.


Among the stalwarts of Indian linguistics and grammar, we may recount the names of ShaakaTaayana, Yaaska, PaaNini, Kaatyaayana, and PataNjali, along with Bhartrhari, who summed up all the precious gains and gave a distinct individuality to the science of language as well as to the philosophy of grammar. A long path was already trodden in these fields, before Bhartrhari arrived on the screen.

It will be worthwhile if we take into account some of those achievements, in order to get a clear picture of Bhartrhari's contribution in this field.


The study of meaning in India commenced with the attempt to give the etymology of Vedic words. During the period of the BraahmaNas, etymological explanation of Vedic words had made sufficient progress[ix] .

Thereafter, scholars like Yaaska (8th century) and PaaNini (6th century) expounded the scientific treatise on science of etymology and grammar. Both these seers have adopted a scientific methodology to explain the words as well as their meanings, while laying down the principles of etymology. Yaaska declared: 'artho nityam pareekseta,' that is, 'while giving etymology, the meaning must be examined regularly'. Following this principle, he gives the etymology of the words, keeping constant touch with their meaning. PaaNini also, while giving the derivations of the words, concentrates upon their meanings.

Although both (Yaaska and PaaNini) describe the semantic manifestation based on phonetic changes in the words, PaaNini does not attempt a derivation of the non-derivatives (avyutpanna).


Yaaska declares in unambiguous terms that any query regarding words should start on the presumption that every word has some set of meaning. ShaakaTaayana held the view that all nouns originate from a verbal root, that is, naamaanyaakhyaatajaaeeti (SaakaTaayana, Nirukta. 1.1). This position may appear to be far-fetched for a few, but it has its deeper implication as well. It means, in other words, that every word originates from original sense.

Defending the theory of Kautsa, an argument was advanced, which was in itself the greatest rebuttal to the divine theory regarding the origin of language. It was in this context that the parity in the Vedic and folk languages, on the basis of their having the same words, was established along with the declaration that former amongst them is as much meaningful as the latter.[x] Yaaska summed up the whole argument and declared: 'Where the meaning is not accompanying and the separation of 'root' and 'suffix' is not possible in normal ways, even in those cases the query regarding the 'root' should be persisted with because every word is used only because of its sense.'[xi] Thus, the theory of ShaakaTaayana was taken to its logical end.

PaaNini followed ShaakaTaayana and AApishali in his search for finding out a 'root' of even almost every doubtful word, and read the UNaadi section as separated from the general category.[xii]

In the meantime, Meemaamsaa and Nyaaya Schools of Indian philosophy, along with Bauddha and the likes, also developed their distinct theories regarding the nature of meaning and its resting in the word. It was here that the 'direct' and 'indirect' forms of meaning were discussed first, which later gave birth to the theory of the three kinds of word - powers, namely, abhidhaa (direct), laksaNaa (related), vyaNjanaa (indirect).[xiii]

PataNjali is the first grammarian of the PaaNinian School who had deeper concern for the linguistic problems. For him the simplest definition of meaning was 'the cognition, which is received simultaneously with the utterance of the 'word' [xiv]. A word can convey four types of meaning, namely, jaati (universal); drvaya (object); guNa (quality); and kriyaa (action). According to PataNjali, the meaning should be distinguished from the word, which in fact is the utter sound, capable of conveying the object, for which it stands.


Bhartrhari, the Linguist-philosopher, is well known for his Vaakyapadeeya (VP), based on Indian grammatical philosophy and several traditional schools of thought. According to Jan Gonda, "The extremely difficult VP… of this outstanding philosopher … proves to be an intellectual achievement of considerable importance, and though its author, of course, follows traditional lines of thought, of remarkable originality." "The Vaakyapadeeya, is considered to be the major Indian work of its time on grammar, semantics and philosophy of language". (George Cardona, 1980)


Bhartrhari has recorded as many as fourteen contextual factors in determining the sense in doubtful cases. They are:

  1. Samsarga: Close connection of one individual object with another .
  2. Vipragoga: Separation as contrasted with samsarga
  3. Saahacarya: Association; company
  4. Virodhitaa : Contrary nature; dissimilarity; enmity as one.
  5. Artha: Purpose
  6. PrakaraNa: Context
  7. LiNga: Significant expression supplying the factor needed by another expression to complete its sense.
  8. Sannidhi: proximity with other words.
  9. Saamarthya: Capability
  10. Aucitee: Propriety
  11. Desha: A place where a particular word is uttered
  12. Kaala: The time factor notion of an individual object.
  13. Vyakti: Person.
  14. Svara: Accent

It is noteworthy that, six out of these fourteen contextual factors also figure in Bruhaddevataa kaarikaa. These are artha (purpose), prakaraNa (the subject matter under discussion), liNga (gender), aucitya (propriety), desha ( place; an indication from another place) and kaala (time). These are the contextual factors for determining the meaning of the Vedic mantras or the expressions in the classical language. [xvi]


Taking hint from a PaaNinian rule svam rUpam shabdasyaashabdasamjNaa (Panini 1.1.68)[xvii], Bhartrhari developed the idea of two-fold meaning. When a word is uttered, it reveals two types of meanings:

  1. Its phonetic form.
  2. An object for which it stands.

Thus, it is said that, this twofold nature of word is compared with knowledge and light" "As the knowledge itself and its object are two inseparable aspects of one and the same thing, so are the two aspects of the word, i.e., word and meaning, inseparably united and belonging to one and the same thing." "Light has two aspects: as a receptacle thing by itself, and as an instrument in the reception of other things. In the same way, all the words have two aspects: they are receptacles in their own forms, and they are instrumental in reception of the sense, lying behind themselves." [xviii]


Bhartrhari's conception of shabda is analogous to a certain extent to the modern concept of linguistic sign [xix]. A linguistic sign is considered to be a two-sided entity. Even for Bhartrhari, shabda is related to the phonetic structure on the one hand and to the semantic fact on the other hand.[xx] He says that the grammarians recognize two kinds of words: sound word and semantic word. The latter is not a material word but a psychological entity or mental equivalent of an articulate sound (buddhistha).

Semantically speaking, the speech can not be divided into any 'parts'. According to Bhartrhari, the analysis of a sentence into words and of words into stem and suffix is made for the practical purpose and has no real value. The phonetic similarity and dissimilarity is the sole basis of such an analysis. Therefore, the individual words have no real meaning, sentence is the only meaningful unit of the language.[xxi]

The essence of these statements lies in the claim that the division of a sentence into parts is a result of grammatical analysis. It is a means for teaching the language for those who are ignorant.[xxii] Worldly behavior depends on the capability of words and their meanings.[xxiii] Similarly, phoneme is a device, employed in analyzing and understanding the correct pronunciation. It has its length measured only at the time of its pronunciation.[xxiv] Otherwise, even the phonemes make themselves explicit only in the form of a plosion, the only mode of reception of speech.


According to Bhartrhari, the sole purpose of speech is to help someone to express his own self. The word is the only medium of expression and its reception. Therefore, it is the desire for self-expression, which becomes the basic criterion for ascertaining the unit of speech or the semantic minimum. And, as this desire for 'self-expression' is indivisible into any parts, the 'statement' carrying this desire must also be indivisible. [xxv]The expressional form of this very 'statement' is called 'sentence' or vaakya, which thus proves to be indivisible into words or phonemes.[xxvi]

In other words, only a statement can convey any meaning, not the so-called parts of it. And 'phoneme' is nothing, but the smallest part of a sentence, having no distinct semantic value.[xxvii] It is interesting to note that as far as the Semantics is concerned, on the one hand, varNa or phoneme has no semantic value at all. But, on the other hand, even single varNa or phoneme might make a complete statement in itself, if a certain semantic value is ascribed to it.[xxviii]

Some declare the 'phoneme' as a basic and compact semantic unit, generating the bigger semantic unit.[xxix] The protagonists of the 'sentence theory' recognize sentence as the basic expressive unit. They are, however, divided amongst themselves, about the definition, length and form of the sentence.

The 'sphoTa' theory, as propounded by the post-Bhartrhari grammarians, leaves no room for recognition of either the 'word' or the phoneme as a semantic unit of speech. Though sphoTa is equally the basic mode of reception for the phoneme, word or sentence, yet, semantically, this sphoTa or explosion is dependent solely on the unit of the sentence.[xxx] Hence only the sentence may be called as the true semantic minimum, or the unit of speech.[xxxi]


The sentence sphoTa of Bhartrhari is unique in nature. According to Bhartrhari, a sentence as a meaningful linguistic unit can not be sub-divided further into smaller significant units. A sentence has parts, which constitute its external structure. But all parts of a single integrated sentence do not convey the meaning of the sentence severally. The meaning of the sentence is something over and above the meaning of the parts. Thus, the meaning of the sentence is understood as a flash. According to Bhartrhari, it is pratibhaa.

When a speaker utters the sounds (dhvani), they die away in the next moment revealing the sphoTa, which is an auditory impression of the sound. Immediately the pratibhaa, as flash, translates the impression of the sound into meaning. The gap between the flash and understanding of a sentence meaning is not noticeable. Due to this, people think that they are identical. Sometimes the meaning of a sentence is understood even before the sentence is auditorily perceived. Thus, the pratibhaa plays an important role in understanding the overall meaning of the sentence.


To sum up, the theory of indivisible sentence meaning can play pivotal role in Bhartrhari's discussions on semantic issues. Besides this, Bhartrhari has discussed other important issues such as words and meaning relationship, the nature and expression of the meaning, problem of homonyms, etc.

The impact of these theories is also visible on the works of subsequent thinkers such as KaiyaTa, Naagesha BhaTTa, KauNDa BhaTTa, and others. Both ancient as well as modern scholars also receive semantic issues with great appreciation.

Bhartrhari enjoys a unique position in the history of Indian linguistics. He is the first grammarian, who gave serious consideration to many linguistic issues, which were left untouched or unanswered by his predecessors. As it is quite well known, before Bhartrhari the main emphasis of the Sanskrit Grammarians was on the formal aspects of the language. The questions related to meaning were left at the mercy of Meemaamsakas and logicians. Bhartrhari changed the focus of his attention from the formal to the notional or philosophical aspect of language. He focused his attention on meaning and explored its multiple dimensions. He collected ideas scattered in the works of PataNjali and the other Shaastrakaaras and developed out of them a theory, that grammarians could call their own. His Vaakyapadeeya. marks a beginning of the tradition that was solely devoted to arthaprakriyaa (meaning analysis).

Thus, Bhartrhari can be called the father of Indian Semantics in the significant sense of the term.


[i] Semantic change in Sanskrit. J.L. Kamboj, Nirman Prakashan , p. 3

[ii] Semantic in the technical term used to refer to the study of meaning, Palmos, F. R. Semantic, P. 1

[iii] Jesperson, O., The Philosophy of Grammar, p.33

[iv] Ullmann, S. Semantics, P. 12.

[v] Wilbur Marshal Urban: Language and Reality , 1939 , p. 95

[vi] Lyons. J., "The Meaning of Meaning" The times literary supplement, 23rd July, 1970, p. 795.

[vii] ShaktigrahaM VyaakaraNopamaanakoshaptvavaakyaad vyavahaaratashca |
vaakyasya shesaad vivrter vadanti saannidhyataH siddhapadasya vrddha1/2 || Keshava Ramrao Joshi, NSM. P. 173, BORI, 1985

[viii] yastu prayuNkte kushalo visese shabdaanyathaavadyavyavahaarakaale || Mbh.1.1, p.2 ||

[ix] India and linguistics , JAOS 75, 1955, p.151.

[x] ArthavantaH shabdasaamaanyaat || Nir. 1.1||

[xi] atha ananvite arthe apraadeshike vikaare arthanityaH parIkSeta || Nir. 2.1 ||

[xii]For detail. See 'History of Sanskrit Grammar' written by Pt. YuddhiSTira MImaaMsaka, part I.

[xiii] Studies in Indology , Satyakama Verma, Bharatiya prakaashana, 1976, p194.

[xiv] yenoccaaritena saasnaalaaNUlakakudakhuraviSaaNinaaM saMpratyayo bhavati shabda1/2 || Mbh.1.1,p.1||

[xv] Connection , separation , association , opposition , meaning, context, indication, the presence of another word, suitability, propriety , place, time, gender and accent etc. these are the causes of determining the meaning of a word when there is no definiteness in it. Vk 315-316 The Vaakyapadeeya of Bhartrhari, Chapter-II, English translation, K.A. S. Iyer, first edition , p.136-137.

[xvi] Linguistic Thought in Ancient India , Pushpendra Kumar, Nag publisher, p. 103)

[xvii] " When a meta -linguistic item is mentioned in a rule for purpose of grammatical operation, then) the own (phonetic) form of the meta-linguistic item( is to be understood ), with the exception of a technical name for the meta-lingustic item. , The ASTaadhyaayee of PaaNini with Translation and explanatory Notes, by S.D. Joshi and J.A.F. Roodbergen, Vol.1, Saahitya Akademi, p. 121

[xviii] aatmarUpaM yathaa jnaane jneyarUpaM ca drushyate |
artharUpaM tathaa shabde svarUpaM ca prakaashate || Bk. 51||
graahyatvaM graahakatvaM ca dve shaktI tejaso yathaa |
tathaiva sarvashabdaanaamete pruthagavasthite || Bk. 56||

[xix] This has been already pointed out by K.K Raja in "Indian Theories of meaning", p. 121.

[xx] dvaavupaadanashabdeSu shabdau shabdavido viduH |
eko nimittaM shabdaanaamaparo'rthe prayujyate || Bk.44||

[xxi] Vk. 11-14

[xxii] shabdasya na vibhaago'sti kuto'rthasya bhaviSyati |
vibhaagaiH prakriyaabhedamavidvaanpratipadyate || Vk. 13||

[xxiii] vyavahaarashca lokasya padaarthaiH parikalpitaiH |
shaastre padarthaH karyaarthaM laukikaH pravibhajyate ||Pk. 3. 88||

[xxiv] Bk.77

[xxv] Vaakyapadeeya. 10-16

[xxvi] pade na varNaa vidyante varNeSvavayavaa na ca |
vaakyaatpadaanaamatyantaM pravibhaago na ka‹cana || Bk. 74||

[xxvii] yathaa saavayavaa varNaa vinaa vaacyena kenacit |
arthavantaH samuditaa vaakyamapyevamiÒyate || Vk. 54||

[xxviii] Vk. 40

[xxix] Vk.41-43

[xxx] naarthavattaa pade varNe vaakye caivaM vishiSyate |
abhyaasaatprakramo'nyastu viruddha iva drushyate || Vk. 402||