Chapter 3Orthography and transcription

3.1.1 Dobrotvorsky and the earliest records

Dobrotvorsky, a Russian military physician, lived on Sakhalin from 1867 to 1872 and compiled an Ainu–Russian dictionary of over ten thousand headwords, more than half of them collected by himself, mainly from speakers of the central West coast such as Sambaku of Nayoro Sakaguchi (2021: 43–44). At the end of the dictionary he printed two short connected pieces — Ciwokanke's speech and a reconciliation dialogue — which Piłsudski already singled out as the first usable Sakhalin texts (Piłsudski 1912: XXIV).

Dobrotvorsky wrote Ainu with the pre-reform Russian alphabet, marking stress but not vowel length (Sakaguchi 2021: 45–47). Several of his conventions are phonetically revealing. He alternates б/п and к/г for /p/ and /k/, recording the intervocalic voicing that is allophonic in Sakhalin Ainu (see phonology). Word-initial /r/ he consistently heard with an occlusive onset and wrote тр, as in трéкоро for reekoro ‘very’ — an observation discussed at length by Piłsudski (1912: 8) and matched by Chiri's report that initial r wavers with a t-like stop (reekoro ~ teekoro) (Chiri 1942: 第1章). A silent ъ closes every consonant-final word, as in contemporary Russian spelling, and /c/ is normally ч. Example (1) shows the opening of his first text with our retranscription.

3.1.2 Sentoku Tarōji's letters of 1906

Sentoku Tarōji (1872–1929), an East-coast Ainu of Otasan schooled under the Russian administration, wrote three letters in Ainu to Piłsudski, dated 4 June, 15 June and 11 August 1906. They are kept in the Sakhalin Regional Museum and were deciphered and published with a Japanese translation by Tangiku Itsuji (丹菊逸治) and Ogihara Shinko (荻原眞子) (Tangiku & Ogihara 2001). They are the earliest known Ainu-language texts written by an Ainu hand, antedating Chiri Yukie's (知里幸恵) アイヌ神謡集 of 1923 by seventeen years (Nakagawa 2006: 21).

Sentoku's orthography is an unstandardised but internally consistent adaptation of Russian cursive. In the letters we observe the following habits: the coda fricative is х (яхка yahka 'even if'); /c/ is written both ц and ч (цеп cep 'fish', нучя ~ нуця nuca 'Russian'); б appears for /p/ after nasals (камбі ~ кампі kampi 'letter'); both и and the pre-reform і serve for /i/; and — most striking — the devoiced /i/ of syllables between voiceless consonants is simply not written: цсе for cise ‘house’, цкін for cikin 'if', цнукара for ci=nukara 'we read it'. This matches the strong devoicing of unaccented /i/ that Chiri describes for the dialect (Chiri 1942: 第1章). Sentoku also mixes in Arabic numerals and undigested Russian words (помощникъ старосты 'assistant village head'), exactly as a bilingual letter-writer would.

3.2.1 Piłsudski's practical orthography

The twenty-seven East-coast texts of Piłsudski's Materials are printed in a practical Latin orthography which he explains in the prefatory "Remarks on Ainu phonetics, and on the transcription of the texts" (Piłsudski 1912: 1–10). Its values are Polish rather than English: j is the glide /y/, v alternates with w, and a palatalised sibilant "between s and sh" is written with the Polish acute, wavering with plain s in his own hearing — iśam ~ isam 'there is not', śumari 'fox' (Piłsudski 1912: 5–6). The letter x renders the velar fricative that he correctly identified as found "only in Saghalien", corresponding to Hokkaidō p, t, k: axto 'rain' beside Yeso apto, cux beside cup 'sun', cex beside cep 'fish' (Piłsudski 1912: 7, 9). Voiced letters b, d, g appear wherever he heard intervocalic voicing; stress is marked on all non-oxytone words; whispered word-final vowels are set as small superscript letters; and words are divided etymologically rather than by sandhi (Piłsudski 1912: 9–10). Dal Corso has recently mapped this notation systematically onto a phonemic analysis (Dal Corso 2024). Example (4) shows one of his sentences in the original orthography and in ours.

3.2.2 Linguistic romanisations

Chiri Mashiho (知里真志保) gave the Sakhalin dialect its first systematic linguistic transcription. In his 1942 grammar he replaced the Hepburn-style digraphs of earlier work with č, š, j and ü, and wrote the Sakhalin coda fricative as x (max 'woman', tex 'hand') (Chiri 1942: 凡例); vowel length he left unwritten, regarding it as non-distinctive (Chiri 1942: 第1章), a decision since reversed by all later workers (see Itabashi (2001) on the phonemic status of length). Nakagawa traces how this notation fed into the strict one-letter-one-phoneme romanisation established in the 1950s and codified in the dialect dictionary edited by Hattori Shirō (服部四郎) (Nakagawa 2006: 3–5): there every vowel-initial syllable is preceded by an apostrophe for the glottal stop, /c/ is written c, the coda fricative h, and long vowels are doubled (Hattori 1964: 34).

Murasaki Kyōko (村崎恭子) adopted the Hattori system, in simplified form, for her grammar and for all her text editions: glottal stop ' before vowel-initial syllables, doubled long vowels, particles written as separate words Murasaki (1976), Murasaki (2025: 凡例). Her transcription of the tales of Asai Take (浅井タケ) is the single largest body of Sakhalin text in this notation (Murasaki 2001). Recent philological work on the older corpora (Sakaguchi, Dal Corso) keeps the same letter values but omits the apostrophe and separates person clitics with an equals sign (Sakaguchi 2021: 45), which is also the practice of this grammar.