The document discusses challenges to the idea of semantic determinism in language. It argues that the full propositional content of an utterance is not fully determined by semantics alone, but requires pragmatic enrichment based on context. Explicatures, or contextually supplemented meanings, are needed to arrive at a basic propositional meaning. Indirect speech acts also rely on pragmatic reasoning rather than just conventional meanings.
The main point of the talk: each approach serves as a heuristics for the others no matter epistemological incompatibilities and variations in objectives and scopes posited and worldviews (ethos) attached to a theory / philosophical grounding. Though the notion of procedural pragmatics, an idea notably developed in Geneva, I want to observe a set of pragmatic phenomenons, leading in the end to a question on interfaces between RT, in particular that specification I call procedural pragmatics, with other approaches.
(5) Would normally be taken as an implicature standard. But to explain it, we must assume that the hearer extracted straight away an implicit premiss like « Flat is boring for bicycling ». But the hearer can well fail to find this very premiss and need first to take an implicature like « Bicycling in Holland is easy », taking it as premiss together with another implicit premiss such as « the speaker is a good biker » in order to find, thanks to a deduction without explicit premiss, implicature 5. We notice that the deduction is still non-demonstrative. Implicature (3) has the degree of certainty that implicit premiss (2) has, which is weaker of course than the one of the explicit premiss (1). In the case of the following deduction (5), we notice that (5) is entertained at the degree of the weakest of the premisses, be it the implicature (3) or the contextual premiss (4). Now, it may also be the case that (5) is NOT INTENDED BY THE SPEAKER. In such a case, (5) cannot be treated as an implicature standard: it’s not part of the informative intention. It’s a speculation that the hearer makes, assuming it’s of importance to him. Let’s take a clearer example of that.
The weakest premiss is an implicature of current utterance. It’s already an element on which truth-conditions the speaker doesn’t commit himself. The information is already weakly reliable. Adding a premiss about Mary, even weaker, leads to a weak and risky, but possibly very useful, inference about possibly something else than the informative intention of the speaker. There are other consequences to inferences like that. For example: if the speaker didn’t explicitely talked about that, he may fear a refusal. Anyhow, such an information may lead the hearer to adopt a specific attitude when in turn endorsing the role of speaker of a new informative intention. Any conclusion corresponding to an informative intention can enter as premiss for further inferences about the speaker’s cognitive environment and intentions, that may not concern the informative intention proper. These are not communicated, they are not part of communication, but they are part of the aspects that determine this particular dimension of discourse, which is interaction. We see that RT tools can apply also to this important aspect of discourse studies, at least to some extent. The hearer can, this way, speculate argumentative goals on the part of the speaker. For instance, if there are other contextual assumptions that lead to think that the bicycling holiday won’t be easy anyway, for example if it’s a rainy season, the hearer may suspect some covert intention from the speaker’s side.
The hearer can come to the conclusion (7), which is in no way allowed by the sole utterance Holland is flat together with an implicit premiss. However, conclusion (7) can arise unconscioulsy and automatically to the hearer’s mind. (7) Is in no way an implicature of the original utterance. It’s a speculation based on no explicit premiss. It may be the case that the hearer evaluates the necessity to get confirmation of this information, or to block any consequence of it. He can do so by uttering a new utterance. This happens if the information is relevant to the hearer, and if the degree of reliability of the inference is weak and need confirmation. What assumptions will need confirmation? This is a difficulty, because it’s function of a multiplicity of factors, among which the capacity of the hearer to create new informative intentions, to adapt his goals during the conversation, his capacity to attribute mental states to others and finally to the possibly intrinsically undertermined dimension of informative intentions generations. It’s also function of his psychological condition. For example, some effects of paranoïa are due to attributing higher-degree inferences degrees of reliability that do not comply with the deductive-non-demonstrative model. It’s now time for me to get to the conclusions.
AD: Two utterances make a discourse when they are produced by the general laws of discourse as a specific subtype of human activity. CP: Two utterances make a discourse when the first one is a contect that provides more relevance for the second one. This latter assumption allows for thinking that psychosocially, for example praxeologically, defective discourses are in fact defective at the level of the potential of contextualisation an utterance should offer for the next ones.