Published: April 25, 2022

Author: Madelyn Weber
Nominator: Rebecca Scarborough
Course: LING 3100 Language Sound Structures, Fall 2021
LURA 2022

Japanese is a language only spoken widely in Japan, however there are approximately 47 dialects spoken within the language, which makes for some interesting variations in vocabulary, grammar, and pronunciation. 

I examined pronunciation differences between the Kanto and Kansai dialects. The Kanto dialect is spoken in and around Tokyo, the capital, located in the Kanto region of Japan. The Kansai dialect is spoken around the cities of Osaka and Kyoto, located in the Kansai region of Japan. 

Japanese makes use of a syllabary writing system — a written system in which each ‘letter’ represents a consonant-vowel pair, pronounced as a syllable. Japanese spoken pronunciation, however, is based off of a system of moras — a phonological unit of sound, similar to a syllable, that consists of a consonant-vowel (CV) construction and denotes timing in pronunciation, where each mora is pronounced for the same duration of time. Due to this, Japanese can only accept a CV sound structure, with the exception of the sound /n/, which is composed of a single consonant. As an example, we can look at the Japanese word for “Japanese language,” 日本語 /ni.ho.n.go/, which consists of four moras (syllables). Each mora has been split up by periods within the transcription above. So, when pronouncing the word /ni.ho.n.go/, each mora should be pronounced for the same approximate measure of time. 

I am specifically analyzing how mora (syllable) length changes between the Tokyo and Osaka dialects, especially in monosyllabic words. A well-known trait in the Tokyo dialect of Japanese is that monosyllabic words — or words that consist of only a single mora/syllable — are pronounced for the same measure of time as any other mora within a given sentence, but in the Osaka dialect, such single-mora words tend to be lengthened so that they will take up the measure of two moras. For example, we can look at the Japanese word for “eye”, 目 [me], which consists of one mora. For an Osaka dialect speaker, it is expected that they would lengthen the final vowel [e], to produce the two-mora pronunciation /me.e/ or [meː]. 

Using the software Praat, I was able to measure the length of different monosyllabic words spoken by native Japanese speakers; two native Tokyo dialect speakers and two native Osaka dialect speakers. 

Data 1: Example sentences, along with IPA transcription and English translation.

In order to examine whether or not Osaka dialect speakers truly lengthen the vowel sound in monosyllabic words, and to measure the extent of lengthening, we can look at monosyllabic word pronunciation lengths and compare them against non-monosyllabic words. This is done by taking words which contain the same sounds as the monosyllabic words, but only measuring the individual sound itself, rather than the entire word. We can then compare these against the length measures of a lengthed sound. 

Measuring Process Example: 

The monosyllabic word for “letter; character” in Japanese is 字 [ʑi]. We compare this against the Japanese word for “Chinese character”, 漢字 [kanʑi], as well as the word for “grandfather”, お 祖父さん [oʑiːsan]. Here, a measure for the length of pronunciation is gathered for the monosyllabic word [ʑi], and is compared to the measure for the lengthened sound [ʑiː] in [oʑiːsan], as well as the sound [ʑi] embedded within the word [kanʑi]. 

Data 2: Words examined along with approximate average pronunciation lengths per dialect.

After collecting the data of different pronunciation length measures, we can compare them in the tables below. 

Data 3: Difference in length measures between monosyllabic and word-embedded sounds.

Data 4: Difference in length measurements between monosyllabic and lengthened sounds:

From Data 3 and 4, it appears as if the Osaka dialect, on average, has much larger differences

in length of pronunciation between moras for monosyllabic words and for that of the same sound embedded within another word. Seeing such large gaps in difference in pronunciation time supports the expectation that monosyllabic words are lengthened from the length of one mora to two. This is further backed up by looking at the difference in pronunciation length of monosyllabic words and those of the words containing the same or similar lengthened sounds, in which the Osaka dialect is showing smaller differences in the length of pronunciation. This backs up the idea that the two-mora pronunciations are similar in length to the monosyllabic words of the Osaka dialect. We can go one step further by examining half the time it takes for both dialects to pronounce monosyllabic words. So, if the Osaka dialect truly does pronounce their monosyllabic words for the length of two moras, then the halved length of such words should more closely reflect the length of pronunciation for that of similar sounds embedded in other words. 

Data 5: Division on pronunciation length by dialect. 

As can be seen from the table above, on average, the divided measures for monosyllabic words from the Osaka dialect displays pronunciation lengths closer to that of the single mora sounds embedded within words. So we can conclude that while the exact length of monosyllabic words spoken in the Osaka dialect are indeed lengthened when compared to their Tokyo dialect counterparts, they are not always pronounced for the approximate measurement of two mora lengths.a