Following the rules of rendaku in Japanese: Native speakers’ production of novel compound words

Sammanfattning: The goal of this thesis is to gain knowledge on the psychological reality of some known rules, conditions, and categorizations that have been observed in the phenomenon of Japanese “rendaku” (also known as sequential voicing). The results presented in this thesis are based on a quantitative study that was aimed at native speakers of Japanese. The conducted experiment for this study was designed as a forced-choice test, where the respondents had to select their preferred reading of novel compound words. Eight groups of words, each consisting of five compound words, were tested. Four groups tested the psychological reality of different lexical rendaku personalities proposed by Rosen (2001), and one of these groups consisted of words, found by Irwin (2009), which seemingly break Rosen’s Rule, a rule based on a prosodic size requirement. The four remaining groups of words tested four different known rules or conditions that tend to systematically either trigger or block rendaku. The results show clear differences in Rosen’s categorization of words, which suggests that native speakers of Japanese possess an internal categorization as well. However, it is also evident from the results that the internal and lexical categorizations differ considerably from each other. The results further show that lexically rule breaking words translate into irregularities in native speakers’ production of novel compound words as well. The results from the four groups of words that tested known rules and conditions only reinforce their positions as rules and conditions.