抄録
Simplex truncation is a highly productive word formation process in Japanese, where words with five or more moras are shortened to forms with four moras or fewer. Numerous studies have been conducted to predict a single output for a given input, but each faces its own set of challenges. To empirically evaluate the mechanisms proposed thus far, we conducted a dictionary-based survey and examined three key hypotheses: the accent-based hypothesis, the pseudo-compound hypothesis, and the tripartite structure hypothesis. The results of our survey indicated that simplex truncation is best explained by the tripartite structure hypothesis, which posits that the prosodic structure of a word comprises three components. Additionally, we identified the possibility that Japanese contains true pseudo-compounds̶words whose underlying structure is bipartite.
