A linguistic look at riddles

Abstract

The riddle has been the subject of considerable interest among anthropologists and folklorists for some time. Recently, linguists, too, have begun to examine the riddle, perhaps as an offshoot of the burgeoning linguistic interest in humor in general. This paper offers a new look at one type of riddle, the so-called ‘conundrum’, or ‘punning riddle’. Adopting Arthur Koestler's notion of ‘bisociation’, linguistic triggers (the ‘punch lines’) in conundrums can be viewed as phonetic forms linking the semantics of two disparate worlds. Conundrums can then be classified according to the nature of these linguistic triggers. Based on a variable which is termed the ‘similarity factor’, a ‘similarity cline’ is constructed, along which the triggers, and hence the conundrums, can be classified. One end of this cline marks maximal similarity of form, namely total identity, while the other end marks absolute dissimilarity of form. In between these poles are five stages - polysemy, homonymy, homophony, paraphony, and hahaphony - each stage being progressively further removed from the identity pole. These stages mark increasingly weaker degrees of phonetic similarity between the bisociated elements in the linguistic trigger. This classification serves to illuminate the nature of the conundrum itself, while at the same time accounting for a variety of subtypes which fertile minds have invented.