Abstract
In this paper, we consider an effect on inference results for truth-qualified natural language propositions involving fuzzy quantigiers by truth-qualifiers. We proposed an inference method previously. For examples, for the proposition "Many tall men are heavy is τ", we can infer a modified proposition "Most tall men are more or less heavy is τ" by the method, where the fuzzy quantifier "Most" in the inferred proposition can be resolved analytically, according to the modifier "more or less". Here τ is a monotone and injective truth-qualifier.In this paper, we propose a method which effects on the inference results following to adding a weight attribute to truth-qualified fuzzy sets by given truth-qualifiers. By the method, when we infer the proposition "Most tall men are more or less heavy is true" from the proposition "many tal men are heavy is true", it is possible that we infer the proposition "Almost all of tall men are more or less heavy is very true" from the proposition "Many tall men are heavy is very true".By the method, it is possible that we get different inference results depending on the effect by truth-qualifiers.
