|
|
|
|
|
|
on the one hand, the felicity conditions governing the use of expressions (and, therefore, the pragmatic appropriateness of sentences); on the other hand, the mutual knowledge of participants in the communicative process. (Let us call this ideal couple of cooperators Sender and Addressee, hereafter S and A.) The pragmatic approach sounds closer to the nature of presuppositional activity in natural language communication. However, the notion of felicity conditions is not completely adequate to express the full relationship between lexical item and textual insertion. Moreover, when describing presupposition as depending on the knowledge or beliefs of S, on the beliefs that S attributes to A, and on the agreement of S and A on a common set of background beliefs or assumptions, the crude pragmatic approach states what happens, but not why it happens. |
|
|
|
|
|
|
|
|
A limitation of most pragmatic theories is the lack of a textual perspective: very often presuppositions are tested in ad hoc-constructed sentences, removed from any context of utterance. Such sentences do not belong to natural discourse, and it seems unconvincing to base a grammatical theory on artificial examples. We assume that properties of sentences in textual contexts more reliably reflect grammar than do sentences in abstraction. We have said that presupposition is a fuzzy category. The term does not seem to define a series of homogeneous grammatical phenomena; it is more an open category which we assume can be explained only inside a theory of discourse.
2 In fact, a textual approach analyzing presuppositions from the point of view of discourse functions allows a homogeneous explanation, since this homogeneity is no longer on the level of formal structure but, rather, on the level of discourse functions, that is, stated in terms of the textual effects that they produce for the Addressee. |
|
|
|
|
|
|
|
|
Thus we will hypothesize a general functioning of information in discourse, which can accountin general termsfor all different presuppositional constructions. Such a textual account is weaker than a grammatical theory for presuppositions. The grammatical constructions which are traditionally called presuppositions interact with general textual principles in such a way that presuppositions could be defined as the result of both semantic rules and discourse. In other words, we assume that presuppositional constructions are recorded in the lexicon or otherwise encoded in the language system, but they are activatedor downplayedby means of general discourse principles. Particularly, we will argue that, for the presuppositional phenomena we are considering here (P-terms), those textual effects depend on meaning organization. Thus text theory has to be linked to meaning representa- |
|
|
|
|
|