The central issue of any account of humorous metaphors is to answer the question of why some metaphors are humorous and some are not (the same question, mutatis mutandis applies to blends, or any other cognitive phenomenon involved in humor). Most accounts of humorous metaphors rely on variants of the "distance theory" (Fónagy 1982, Morrissey 1989, Pollio 1996) which essentially consists in postulating a threshold of "semantic distance" beyond which the linkage between the two domains in the metaphorical construal becomes "stretched" and is therefore perceived as humorous. The most obvious problem for the distance theory is that no precise (or even approximate) quantification of the threshold exists. In this paper I will present an alternative (but not necessarily antagonistic) approach grounded in the most recent developments of humor research within the incongruity/resolution framework. The proposed explanation for humorous metaphors (stricto sensu) is that they are metaphors in which the incongruity of the mapping of different domains is not fully resolved by the interpretation (making sense) of the metaphor (Oring 2003). This is a primarily semantic explanation. The explanation for failed metaphors is primarily pragmatic and relies on an extension of the principle of cooperation (CP), which has been justified on independent grounds elsewhere. Essentially, this entails the postulation of a "principle of non-cooperation" (NCP) which governs the violation of the principle of cooperation. (Salvatore Attardo)