In a message dated 10/10/2001 8:48:29 PM Central Daylight Time, jjllambias@hotmail.com writes:
la pycyn cusku di'e Ok. Let me unpack "bogus": "The claim that I have stopped beating my wife violates its own presuppositions." I see no presuppositions coming up from the included sentence to affect the meta-claim. If there are some, what are they? Similarly, I see no presuuposition from the included claim to "John knows whether I ahve stopped beating my wife" as a report that John knows that some presuppositions of the included question are not met. <The way I see it, the questioner, when asking a question, presents the listener with a set of answers, from which the listener is supposed to pick one. When the listener finds that no member of the set is adequate, the response is {na'i}. It says that the set is inadequate, it is not just another member of the set. The listener is not playing along with the questioner in this case. A similar case happens when the response is {ki'a}. This again is not yet another member of the presented set. It is rather an indication by the listener that they can't make out what set the speaker means to present.> On the other hand, this looks like a plausible aternate theory. The parallel with {ki'a} is particularly convincing. And getting rid of the {na'i} form would simplify the schema. Perhaps we need to distinguish between responses and answers. The problem may be that the standard theory, while talking a Gricean line, in fact takes the set of answers as being generated by the sentence (in context, to be sure) and not by the questioner. Of course, the questioner cannot have unlimited control -- loaded questions are the exception, and questions get asked to which the answer is beyond the questioners previous ken, so some middle position remains to be spelled out. |