On Opacity
Continuing from...
B. Language and Interpretative Techniques
Here I assume that truth-conditions alone do not automatically trigger any process of verifying sentences or reviewing beliefs or other propositional attitudes. Rather it is the users of natural languages that play a proactive role in the comprehension and evaluation of sentences. The proactive role of speakers is evident by the fact that the normal usage of such languages does not support the famous deflationist claim that adding the predicate is true to a sentence ? adds no content to it .
The fact that a string like it is true that contributes to the meaning of a natural language sentence is made evident in the cases the same sentence is evaluated differently by language users. Or in other words, a sentence may be deemed true by the person that says it and false by those who hear it, regardless of the conditions obtained in a certain world or situation. This means that the claim that a sentence like (2a) means (2b) can only be made from the perspective of the person who utters it, if and only if he is sincere:(2) a. Tom Jobim composed a new song.
b. It is true that Tom Jobim composed a new song.
It cannot be made from the perspective of the hearer, who may doubt (2a). And, if the utter is a conscious liar, it cannot be made from his perspective either. Notice that this discrepancy of opinions may occur independently of whether Tom Jobim has or has not composed a new song in a certain world or situation. In part this observation shows a reactive role played by hearers, when they accept or doubt a sentence. But the possibility that even the utter of a sentence may deem it false evidences the proactive role he plays. And if one takes into account that interlocutors in a conversation also have their intentions and communicate them, then their evaluation of any sentence may also reflect a proactive role, more than a reactive one.
Thus, a sentence ? is not inherently construed as true (T?), false (??), undecidable (U?), possible (??) or necessary (o?). Those are judgements made by the language users when they handle sentences.
But the language users' proactive role is not limited to their capacity of merely ascribing truth-values to sentences. It is often the case that language users are able to construe gibberish utterances in a manner that they make sense, without incurring into some paradoxes or into some explosive inconsistency a la Pseudo-Scotus (99). And they usually do so, unless they choose not to.
Let us illustrate this idea with examples, like the sentences like (3). While the equivalents to in any artificial logic language are absurdities or paradoxes that require a sophisticated philosophical engineering to solve them, users of natural languages somehow manage to extract non-paradoxical coherent meanings from them:(3) a. This sentence means the converse of whatever it means.
b. There's no such thing as legacies. At least, there is a legacy, but I'll never see it.
c. The ambitious are more likely to succeed with success, which is the opposite of failure.
There is an evident self-referential paradox in (3a), a clear contradiction in (3b) and a redundant or circular thought in (3c). And they can be interpreted in this manner, if the language users that read them proactively choose to see the possible paradoxes, contradictions and redundancies. But, for the reasons that I shall go into further on, that is not what they frequently do in the everyday common usage of a natural language. Accordingly, (3a) is usually interpreted either as referring to another sentence or as a potential metaphor for something, while (3b) may be taken as a review of statements and (3c) as an attempt to stress something. This evidences that the hearers/readers are able to recover the intended messages behind clumsily constructed sentences. Methinks that, in the exercise of this capacity, language users proactively employ certain techniques. But these techniques are not just any techniques: they are not merely ad hoc inventions, neither are they hazardously chosen.C. Paraconsistency
Sch?ter (1994), among others, claims that the semantics of natural languages exhibit four basic properties that are already acknowledged and have been investigated in non-classic logic theories: paraconsistency (76), defeasibility, which contrast with monotonicity defined in (98), partiality, which contrasts to totality defined in (102), and relevance (101). Though paraconsistent approaches in Logic have been developed since the seminal works of Jas?kowsky (1948) and da Costa (1963, 1997) , and though they have important consequences for Linguistics, similar approaches in natural language semantics are only beginning.
As Priest (2002) explains, most of Paraconsistent Logic consists of proposing and applying strategies against non-consistency (Cf 93). There are several possible techniques to avoid or contain explosion within a logic system or semantics for artificial or natural languages, such as propositional filtration, non-adjunction, non-truth functional approach of negation, de Morgan algebra, etc. But those are all techniques invented by logicians for artificial languages. Should we accept the idea that in construing sentences the users of natural languages also use techniques to control explosions, then such techniques must be available as inherent interpretative devices of the human linguistic systems. In other words, as Logicians have their paraconsistent techniques for artificial languages, so the users of natural languages have techniques of their own, which are made possible by the fundamental properties of such languages.
Accordingly, the capacity of humans to ascribe values to sentences independently of the actual conditions obtained in a certain world or situation, which underlies the phenomenon called opacity in human languages, has to do with at least one of these techniques humans naturally posses: the contextualisation of sentences. I shall explore and unfold this matter in the following.
Continue
Posted by Tony Marmo
at 14:16 BST
Updated: Wednesday, 21 July 2004 14:20 BST
