Phonology VIII: サ行とハ行の揺れ
There are various instances in Japanese in which fluctuation between the phonemes /s/ and /h/ can be observed. This is attributed to an ongoing merger of /s/ into /h/ which can be observed in many Modern Japanese dialects.
The History of サ行音
Thanks to comparative linguistics and other reconstructive methods that are available to researchers, even the smallest amount of data collected from discoveries of older inscriptions can tell us a lot about what Japanese might have sounded like in the Nara Period and earlier.
One notable feature that has been widely accepted is how the Modern Japanese phoneme /s/ was /*ts/ in Early Old Japanese and into Late Old Japanese. This was first pointed out by the linguist 有坂秀世 in his work 『上代に於けるサ行の頭音』, and the discovery was made thanks to research into the reconstruction of Chinese pronunciations1 at the time and how approximations for Japanese at the time seemed to indicate the lack of an /s/ phoneme.
By the start of the early Heian Period, however, it is also widely accepted that the /*ts/ phoneme of Proto-Japanese had transitioned to /s/, with the allophonic variant [*ɕ] appearing before /i/ and /e/.
While the palatalization of /s/ before /i/ is maintained in Modern Japanese, it may come off as strange that セ was once pronounced likewise as [ɕe]. However, thanks to transliterations found in Kirishitan Japanese-Portuguese (日葡辞書) dictionary inscriptions of the pronunciations found in Middle Japanese that it was indeed pronounced as so, and still is in many Western Japanese dialects.
/s/ → /h/
There is a universal tendency for /s/ to transition to /h/ over the course of natural language change. As such, its appearance in Japanese is not be surprising. Yet, its appearance in Japanese remains dialectal with varying social implications.
One such commonly held belief is that the pronunciation of /s/ as [h] is reflective of ‘lazy’ speech. While sound changes are often thought of as being motivated by easing articulation, sounds naturally drift in their qualities over time, and pinpointing exactly why a specific sound change occurs is rather difficult. As such, the assertion that pronouncing /s/ as [h] is a reflection of a natural sound change of /s/ to /h/ and NOT a reflection of lazy speech.
Examples By Dialect
A general tendency across Japanese dialects is that /s/ → /h/ is most prominent in non-Eastern Japanese dialects. In Eastern Japanese dialects, the opposite phenomenon in which /h/ manifests as [s] in most contexts may be observed. However, when /s/ is not palatalized, retention of [ɕ] with /e/ may appear in both regions in older populations.
| 標準語 | 非標準語の発音 | Meaning | 使われる方言 |
| ～ません | ～まへん | Polite negation | 西日本方言 |
| それなら | ほんなら | If so | 西日本方言 |
| ～さん | ～はん | Honorific suffix | 西日本方言 |
| 捨（す）てる | 捨（ふ）てる | To throw away | 西日本方言 |
| 布団を敷（し）く | 布団を敷（ひ）く | To lay out a futon | 西日本方言 |
| 質屋（しちや） | 質屋（ひちや） | Pawn shop | 西日本方言 江戸弁 九州方言 |
| お七夜（しちや） | お七夜（ひちや） | Name-giving ceremony | 西日本方言 名古屋弁 |
| 七宗町 | 七宗町 | Hichisō Town | 地名 |
| 冷やっこい | しゃっこい | Cold | 東日本方言 |
| 東（ひがし） | 東（しがし） | East | 東日本方言 |
| 先生 | 先生（へんへ） | Teacher | 津軽弁 |
| 先生 | 先生（しぇんしぇい） | Teacher | 博多弁 |
敷く vs 引く
Of the examples mentioned above, the one that is perhaps the most fascinating is the pronunciation of 敷く. The verb itself has various meanings and spellings, of which the greater meaning of “to spread/lay out” can be felt. While pronouncing it as ひく is indicative of the sound change /s/ to /h/, the verb has subsequently been convoluted with 引く, which happens to have the similar meaning of “to lay on.” Examples of the two verbs being used seemingly as one have been observed for since Classical Japanese2. Common examples of these verbs being confused include:
| Meaning | Phrase | 標準語 |
| To spread oil (in a pan) | 油を｛ひく・しく｝ | 油を引（ひ）く |
| To lay a cable | ケーブルを｛ひく・しく｝ | ケーブルを引（ひ）く |
| To carry out an order | 指令を｛ひく・しく｝ | 指令を布（し）く |
| To take up a position | 陣を｛ひく・しく｝ | 陣を敷（し）く |
| To be run over by a car | 車に｛ひかれる・しかれる｝ | 車に（ひ）かれる |
/s/ → /h/ in 近畿方言
Kansai Dialects (関西弁) are notorious for proliferent examples of the sound change /s/ → /h/, but because this is prevalent to a greater extent in the region, treating it as a phenomenon of 近畿方言, which encompasses all dialects in the region or even as exemplary of Western Japanese Dialects (西日本方言) in general would be most accurate.
While drawing the line between which dialects do or do not exhibit this sound change is ambiguous, as a learner of Japanese, common expressions attributed to 関西弁 which contain it are worth noting due to how often they are encountered.
- Other evidence includes Sanskrit-based transliterations (梵字) via Chinese characters through Buddhist influence. The 五十音図 falls in line with how sounds are ordered in Sanskrit with the exception of the /s/ and /h/ columns in Modern Japanese, indicating that their values must have been different at the time of borrowing. It is also worth noting that /s/ manifesting as /*ts/ does not overlap in time with the appearance of [ts] as an allophone of /t/ in Modern Japanese, as the two phenomena are separated by several centuries. ↩︎
- 前栽うゑさせたまひて砂子ひかせけるに・・・
On top of having laid out sand in planting the garden……
In this poem written by 伊勢, a poet (歌人) from the 10th century, ひく can be seen used in the sense of 敷き詰める. ↩︎ - かまへん is a shortening of 構わへん, which in turn features ～へん which derives from a shortening of ～まへん (= ～ません). ↩︎
- ほ is derived from そう. ↩︎