Japanese Homophone Frequency Analysis: Acoustic Study

Abstract

Japanese homophones—words that share the same phonetic form but differ in meaning—are a prominent feature of the language, largely due to its limited phonemic inventory and rich prosodic system. Distinctions between such homophones are often conveyed through pitch accent and syllable duration, which serve as crucial phonological cues in spoken discourse. This study investigated two homophonic word pairs: ame (“rain” vs. “candy”) and hashi (“bridge” vs. “chopsticks”). Recordings were collected from native Japanese speakers, who were asked to produce the target words in a range of natural sentence contexts. Acoustic analyses were conducted using Praat software to examine pitch contours and syllable durations associated with each meaning. Results revealed consistent prosodic patterns that differentiate meanings within each homophone pair. In the case of ame, the pronunciation of “rain” featured a longer vowel duration on the first syllable [a] and a generally higher pitch contour, while “candy” exhibited greater length on the second syllable [me]. Similarly, for hashi, “chopsticks” was characterized by a longer [ha] syllable and elevated pitch at both the beginning and end of the word, whereas “bridge” showed a longer [shi] and a higher overall maximum pitch. These findings suggest that prosodic elements—particularly pitch accent and vowel duration—play an essential role in disambiguating homophones in Japanese. The results further highlight the distinct phonological structure of the Japanese language, where subtle variations in prosody contribute significantly to lexical differentiation in spoken communication.