Artha: Semantics versus Pragmatics

Let’s say Devadatta is going to a party and brings a bottle of wine. When he gets there, he finds that Yajñadatta has already arrived—and that Yajñadatta has brought a bottle of the exact same wine that Devadatta picked up.

If their host were a grammarian, he would probably say, “Great! Now we have two bottles.” If he were a Mīmāṃsaka, he would probably say, “We already have one of those.”

I was reading Śabara this morning and thinking, once again, about the seemingly idiosyncratic way that Mīmāṃsakas use the word artha. It’s well known that the dictionary definitions of artha include “meaning,” “purpose,” “a thing,” and “a good thing,” and we can see how these definitions might fit with each other. Mīmāṃsā is concerned with the meaning of sentences in the Vedas, but in a different way than we might expect, especially if we are used to thinking of meaning in exclusively semantic terms.

In semantic terms, there might be a many-to-one relationship between signs and meanings. There is generally no constraint on different signs, including different tokens of the same sign, conveying the same meaning. But in this case, the verb we are interested in is “conveying.” The word “duck” might convey the meaning of a duck. But consider the sentence: “That is a red duck duck.” Both instances of the word convey their meaning—this is a debateable proposition but let’s assume it for the argument’s sake—but only one of them contributes its meaning to the sentence. In this case, the “extra” word is also poorly integrated into the syntax of the sentence. But we can think of a similar cases (“That is a red duck. That is a red duck.”) that don’t have clear syntactic problems but nevertheless illustrate the contrast between conveying and contributing a meaning. We can characterize this contrast in a couple of ways, but context is a place to start: a sign conveys a meaning independently of any context apart from the rules that characterize the semantic system itself, while it contributes a meaning only in a particular context that is external to this rule-system. While semantics looks at the conveyance of meaning, pragmatics is more concerned with the contribution of meaning.

When Mīmāṃsakas are discussing whether a given word or sentence is “meaningful” (arthavat) or “meaningless” (anarthaka), or whether its meaning is what is intended by it (vivakṣitārtha), they are usually talking about the pragmatic dimensions of meaning, that is, whether the word or sentence in question makes a contribution. One place is the mantra section (1.2.4). The objector says that mantras play a role in the ritual only by being pronounced. What they mean is inconsequential: they are “that of which the meaning is not intended” (avivakṣitārtha). This is not the same as saying that mantras are meaningless. In fact, it is crucial to the objector’s argument that mantras do indeed convey a meaning. The problem is that there are passages wherein a mantra is followed by a directive (vidhi) and both elements convey similar meanings (the example, for those who are interested, is “he takes the halter of the horse with the words ‘they grasped the halter of truth’”). The problem is essentially that only one of the elements can properly contribute its meaning. If the mantra does, then the directive can’t, and that is a big problem for Mīmāṃsakas. This problem is ultimately resolved by considering the directive to exclude (parisaṃkhyā) possibilities that would obtain if only the mantra were available—at least until the very idea of exclusionary directives is attacked in the sixteenth century. But the point I want to make here is simply that it often helps to think of meaning in more pragmatic than semantic terms, according to which “what is meant” includes not only the meaning that is conveyed but the meaning that is contributed (or not) to an existing context.