Chapter 18/s/ and the s ~ š Palatalization Alternation

18.1 One sibilant phoneme, several realizations

Hokkaido Ainu has a sibilant phoneme /s/ whose realizations include [s] and palatalized [ʃ] or [ɕ]. These symbols describe variation within /s/; the consonant inventory also contains the separate fricative /h/ Shibatani (1990: §2.2); Bugaeva (2012: 465). The degree and distribution of palatalization vary among descriptions and speakers.

The conditioning by /i/ is well documented. Broader claims need their dialect scope kept visible: Bugaeva's Chitose description limits the usual palatalization to positions next to /i/, whereas other accounts describe palatalized final /s/ more generally Bugaeva (2012: 465); Nakagawa (2024: 29).

18.2 At the beginning of a syllable

Before /i/, /s/ has the palatal realization heard in the syllable si. Satō explicitly distinguishes this from [si] with a plain [s] Satō (2008: 11). Shiraishi illustrates palatalization with sini [ʃini] “rest,” citing earlier descriptions Shiraishi (2022: §3).

Before other vowels, the evidence is less uniform. Nakagawa notes older kana spellings corresponding to シャ、シュ、シェ、ショ, but leaves open whether they record an earlier pronunciation or a different transcription practice Nakagawa (2024: 28). Bugaeva describes most Chitose instances away from /i/ as comparatively unpalatalized Bugaeva (2012: 465). The older spellings alone cannot establish a historical sound change.

18.3 At the end of a syllable

Nakagawa distinguishes final /s/ after /i/, realized as [ɕ], from final /s/ after the other vowels, described more approximately as [sʲ]. His examples contrast cis “cry” with as “stand” Nakagawa (2024: 29). Bugaeva's Chitose description reports less palatalization away from /i/, including in final position Bugaeva (2012: 465).

Final /s/ must still be distinguished from an open syllable ending in /si/. Nakagawa uses the following pair to illustrate the difference Nakagawa (2024: 29–30):

The second form contains an additional vowel and syllable. Palatalized pronunciation of the consonant does not eliminate that difference. The possessive formation is treated in Chapter 45 (Affiliative Suffix Patterns).

18.4 Connected-speech environments

A final consonant should also be examined in its phrase, where the following segment may affect its realization. Refsing describes final /s/ as frequently palatalized but gives an exception before a following word beginning with /s/ Refsing (1986: §10.3). A statement about a word pronounced in isolation therefore need not exhaust its behavior in connected speech.

18.5 Comparison with Japanese

Japanese kana and Japanese sound labels are useful approximations in the descriptive literature, but authors differ in how closely they identify the Ainu sounds with them. Bugaeva explicitly describes Chitose /s/ as less palatalized in many environments than the corresponding Japanese comparison suggests Bugaeva (2012: 465). A kana spelling is consequently best read together with the author's phonetic explanation.

18.6 Speaker and dialect variation

Bugaeva reports Tamura's observation of individual variation in Saru palatalization, while presenting her own Chitose observations separately Bugaeva (2012: 465). These reports favor an account that identifies the speaker and environment. Combining all published variants into one obligatory rule would overstate their agreement.

18.7 Related consonant processes

The broader consonant system is described in Chapter 16 (The Consonant Inventory and Its Phonetic Realization); related changes at morpheme and word boundaries are discussed in Chapter 28 (Assimilation and Cluster Simplification: Coda /r/, Nasals, and Clusters).