Page 145 - Encyclopedia of Philosophy of Language

alinguistic elements like the suffix -/ attached to verbal roots are called anubandha ('marker'). They do not belong to the language that is the object of the descrip- tion, i.e., Sanskrit. For example, the Sanskrit root referred to is yuj, not yuji or yuja. In the finally derived surface forms, the anubandha marker has to be removed. Patanjali explains how this is similar to ordi- nary usage—a crow that sits on the roof of a house, for example, may be utilized to identify the house: '"Which of these two is Devadatta's house?" "That where the crow sits." When the crowfliesawayand the house is no longer marked, one knows in consequence which house was indicated' (Mahabhasya, ed. Kiel- horn 84.21-85.3). Similarly, English grammar has to account formally for the fact that the past tense of 'drive' is 'drove,' not 'drive + ed' or 'drive +past suffix -ed.'
The sutras or rules of grammatical description belong to the metalanguage of grammar. But there is a higher, meta-metalevel to which the metarules (paribhasa) of grammatical description belong. The most famous of these is a rule that safeguards the consistency of at least part of the grammar by elim- inating contradictions: vipratisedhe param karyam 'in case of contradiction, the latter (rule) prevails (over the former)'(1.4.2).
In order for this rule to apply, the rules werelisted in a specific order. The discovery of rule order is there- fore closely related to that of metalanguage. The use made of metarule 1.4.2 proves, incidentally, that the Sanskrit grammarians recognized and utilized the principle of noncontradiction. This principle was for- mulated and adhered to at the same time or earlier in the ritual manuals, and subsequently in Indian logic and in most of the philosophical systems.
3. Philosophical Theories of Meaning
The following three sections include only a sample of Indian theories of meaning, excluding much that is of later date and all the Jaina and Buddhist contri- butions.
3.1 Mimamsa Theories
The Mimamsa (short for Pflrva- or Karma-Mlmamsa) is a system of ritual philosophy that provides a par- ticular interpretation of the Vedas. It is similar to the Brahmanas in this respect, but it has incorporated the methodology of the ritual and grammatical sutras and is therefore more principled and systematic. Accord- ing to the Mimamsa, the core of the Veda consists of 'injunctions'(vidhf). Thelogicalandsemanticalanaly- sis of these is of special interest from a comparative point of view because it complements the prevailing paradigm of sentence interpretation in Western logic and philosophy, which have long looked upon sen- tences as if they were primarily statements. The excess- ive preoccupation of the Mlmamsa with the optative
corresponds to the Western obsession with the indica- tive.
The stock example of vidhi is the injunction ex- pressed by yajeta 'he shall sacrifice.' According to the grammatical analysis which the Mlmamsa accepts, thisexpressionconsistsoftheverbalroot,yaji(where- / is the indicatory element) and the optative ending - eta. According to the Mlmamsa philosophers, the principal semantic feature of this composite expression is not the root but the ending, because it is through the ending that a word is brought in relation to other words. The ending in turn expresses two elements: 'general verbality' (akhyatatvd) and 'opta- tiveness' (lintva). Of these two, the latter is again the marked feature, for every verb denotes an action but only the optative force is prompted by the Vedic injunctions. This optative force is 'the ultimate of ulti- mates, the peg on which the whole system of Vedic duty hangs' (Edgerton 1928: 176).
The grammarians had already characterized the sentence in terms of its final verb. The Prabhakara Guru school of Mlmamsa extended the analysis of Vedic injunctions in a similar spirit to the more general semantic theory ofanvitabhidhana, according to which the meaning (abhidhand) of a sentence is a single entity that depends on the combined meanings of its con- stituent elements(anvita). The other Mlmamsa school, that of Kumarila Bhatta, was satisfied with the appar- ently simpler and more commonsensical theory, that the meaning of a sentence arises from abhihitanvaya, astringingtogetherorcollection(anvayd) ofthemean- ings of its constituent elements (abhihitd). The former theory is closely related not only to the analysis of Vedic injunctions but also to a logical or syntactic analysis in which sentences (not necessarily state- ments) are the main units of discourse. The latter theory is more easily related to a dictionary-oriented semantics.
The theory that sentence meaning depends on word meaning is especially unsatisfactory when trying to account for logical connectives such as negation and other syncategoremata. This problem is not solved by introducing dictionary entries such as 'Neg,' for 'not- A is not a function of not and A, but a recursively defined expression.' Thus, He did not do it is analyzed as 'it is not the case that he did it.' This goes back, in historical terms, to the Aristotelian insight that 'the negation of the sentence is the negation of the predi- cate,' which is reflected in turn by the logical sym- bolism F(a), formed so that the negation of F(a) is
~F(a), or, more explicitly (as in the Principia Mathematicd):
That the meaning of not-A is not a function of the meanings of not and A was obvious to the followers of the anvitabhidhana theory. But they went further and constructed a theory of negation which is richer
Indian Theories of Meaning
123