Chapter 14Historical Orthographies and the Batchelor Tradition

14.1 The Hepburn-derived tradition

Before the phonemic romanization documented in Chapter 11 (The Latin Phonemic Transcription) took hold in the 1950s and 1960s, Hokkaido Ainu was written in Latin script mainly in orthographies derived from the Hepburn romanization of Japanese. The system that fixed this tradition in print is John Batchelor's. His An Ainu–English–Japanese Dictionary appeared in 1889 and went through a second (1905), third (1926), and fourth edition (1938) Batchelor (1905); Batchelor (1938); Nakagawa (2024: 90), and his Bible translations — above all the 1897 New Testament チコロ・ウタラパ (Chikoro utarapa), transmitted together with the Psalms and Jonah — constitute the largest body of connected Ainu text in this orthography Batchelor (1897); Refsing (1986: 18). Batchelor wrote ch for the affricate and sh for the palatalized sibilant, on the model of the Hepburn values for Japanese, and until phonemic transcription displaced it this style was the ordinary way of romanizing Ainu Nakagawa (2024: 53); Nakagawa (2006: 3); Kirikae (1997).

The tradition was transmitted to the Ainu writers whose texts anchor the modern corpus. Kannari Matsu, whose notebooks preserve a vast corpus of heroic epics, was a graduate of 愛隣学校, the mission school Batchelor founded in Hakodate, where she was taught English and the Latin alphabet; her correspondence with her sister was conducted entirely in romaji Nakagawa (2006: 24). Chiri Yukie's アイヌ神謡集 Chiri Yukie (1923), the collection of divine epics that became the model for later Ainu-authored writing, is likewise written in a Hepburn-style orthography, and Kirikae judges the romanization of both women to be a direct continuation of Batchelor's practice, nurtured by his publications and by the aunt–niece connection between them Kirikae (1997) ‹consensus›. On the academic side the same conventions passed to Kindaichi Kyōsuke and his students; Kubodera Itsuhiko retained the Hepburn style throughout his career Nakagawa (2006: 3); Kirikae (1997).

Batchelor was a missionary who learned the language for evangelical work, and his transcription reflects the ear and habits of an English speaker rather than an analysis of the sound system Refsing (1986: 18). Contemporary and later assessments diverge sharply. Chiri Mashiho's アイヌ語入門 devotes some fifteen pages to mistranslations, phantom words, and wrong forms in the dictionary and concludes that "it would be closer to the truth to say that it consists solely of flaws" Chiri (1956) (translation from Refsing (1986: 18–19)). Laufer had already written that Batchelor "almost neglected" the sound system and that his transcription was "no more than an attempt at adopting the English alphabet to the writing of Ainu" Laufer (1917: 198), quoted in Majewicz (2022: §5). Refsing's assessment is more measured: the dictionary and the grammar bound with it are too inaccurate to serve as a primary source, yet given the scarcity of early materials they retain value when used critically, the English glosses often being closer to the mark than the Japanese ones Refsing (1986: 18–19). Nakagawa notes that after Chiri's critique researchers stopped citing the dictionary openly, while judging it still highly usable for its period and coverage Nakagawa (2024: 90). In this grammar Batchelor's works are treated as philological witnesses: their forms are cited for what they attest about nineteenth-century usage, and every form is read through the correspondence rules of the next section. The place of these works in the source base is set out in Chapter 7 (Written Sources: Grammars, Dictionaries, and the Philological Record).

14.2 Systematic features of Batchelor's orthography

Batchelor's spellings deviate from the modern standard in recurrent, largely rule-governed ways. The table gathers the correspondences; the subsections below document each group from the published comparisons — Kirikae Kirikae (1997), Nakagawa's orthographic surveys Nakagawa (2006: 3–5); Nakagawa (2024: 53–54), and the Sakaguchi–Watanabe re-edition of Batchelor's 1924 Ainu Fireside Stories Sakaguchi & Watanabe (2019: 243–245) — and from the 1897 New Testament Batchelor (1897).

14.3 From Batchelor to the modern standard

Nakagawa tracks the passage from the Hepburn tradition to the modern standard through one example set — aynu ‘person’, horkew ‘wolf’, cis ‘cry’, cis=an 'we cry' — respelled across the successive systems Nakagawa (2006: 3–5):

Each row resolves one layer of the Hepburn inheritance. Chiri Mashiho, who had used the Hepburn style in the 1936 grammar co-authored with Kindaichi Kindaichi & Chiri (1936), moved in 1942 to one-letter-per-phoneme spellings for the glides and the affricate (j ŭ č), while still splitting /s/ between s and š by position; in his 1950s work he abandoned the diacritic letters — Nakagawa attributes the retreat to printing costs — leaving ch as the one remaining digraph Nakagawa (2006: 3–4); Nakagawa (2024: 55–56). The phonemic system established in the 1950s by Hattori and Tamura, and disseminated through the 1964 dialect dictionary Hattori (1964), completed the one-phoneme-one-letter principle with c, wrote the glottal onset ʼ as a phoneme, and wrote transitional glides (ʼuwepeker); the style codified in Akor Itak Hokkaidō Utari Kyōkai (1994) drops both and segments the person markers with = Nakagawa (2006: 4–5); Nakagawa (2024: 56–57). The resulting standard, its remaining points of variation, and the boundary-marking conventions are documented in Chapter 11 (The Latin Phonemic Transcription) and Chapter 15 (Orthographic Standardization and the Word-Division Problem); the parallel kana tradition, which the same writers used alongside romanization, is treated in Chapter 12 (Katakana Orthography and the Extended Small-Kana Codas).

14.4 Reading Batchelor materials today

The working method for making a Batchelor-orthography text usable is the paired re-edition: the original spelling is reproduced intact and a modern respelling (表記変更) is printed beside it. Sakaguchi and Watanabe's edition of the Fireside Stories states the policy explicitly: capitalization of the original is retained; where the text's rV spelling leaves it undecidable whether a vowel is present, the vowel is bracketed rather than silently resolved (or(o)wano); and the respelled text marks morphophonemically fused sequences with an underscore (or_ ta, itak h_i), keeping the underlying words recoverable while recording the sandhi Sakaguchi & Watanabe (2019: 245–247). The same principles transfer to the New Testament and the dictionary citations.

Most of the distance between the two orthographies is mechanical. The segmental correspondences of the features table are exceptionless rewrite rules, and the echo vowels resolve against a lexicon of coda-final stems. A rule-plus-lexicon conversion of the 1897 New Testament, checked against respelling practice of the Sakaguchi–Watanabe type, resolves roughly seven-eighths of tokens; the residue is concentrated where conversion stops being orthographic — the segmentation and placement of person markers, which depend on morphological analysis of the clause rather than on spelling ‹corpus-confirmed›. Reading Batchelor therefore means re-doing the person-marking analysis that his orthography never encoded; the analysis itself is given in Chapter 69 (The Personal-Affix Template: Position Classes and Affix Ordering), and the citation conventions for examples drawn from pre-phonemic sources in Chapter 10 (Interlinear Glossing, Abbreviations, and Citation Conventions).