Chapter 0Introduction
What this grammar sets out to do, the data and conventions it relies on, how to read its glossed examples, and the dialect labels and scholarship behind them.
This book is a reference grammar of Sakhalin Ainu, the branch of the Ainu language family spoken on the southern half of Sakhalin until the late twentieth century and documented in a record that runs from the 1860s to the 1990s. It is organised to be consulted rather than read straight through: six parts lead from the speakers and their history through phonology and morphology to the syntax of the complex clause, and every numbered example is glossed, sourced, and tagged for dialect. This chapter explains how the book works — what it sets out to do (§0.1), what data it admits and under what conventions (§0.2), how a numbered example is put together (§0.3), what the dialect labels mean and which variety is the default (§0.4), and whose scholarship the description rests on (§0.5).
0.1 Aims and scope
Sakhalin Ainu has been described in parts, but never whole. Chiri planned a complete grammar and finished only the phonology and a portion of the morphology (Chiri 1942); Murasaki described the sound system and morphology of a single west-coast variety, Rayciska (Murasaki 1979); the recent literature treats individual subsystems — evidentiality, negation, person marking, number — one article or dissertation at a time. The aim of this book is the obvious missing item: a single, systematically organised description of the language from phoneme to discourse, written in English, in which each statement can be traced to the evidence for it.
Two consequences of that aim shape the text. First, the grammar is a synthesis where a synthesis is possible: points established by earlier scholars are credited to them by name and page, and the reader should be able to move from any paragraph here to the passage of Chiri, Murasaki, Piłsudski, or Dal Corso that stands behind it. Second, where the literature is silent — and for much of the syntax it is — the grammar does not merely record the silence. It analyses the texts directly and states its own conclusion, marked as such, so that a gap in previous research becomes a description that later work can confirm or correct. The data underneath such analyses remain attested; only the analysis is new.
The grammar treats Sakhalin Ainu as a system in its own right, not as a divergent dialect of the better-described Hokkaidō varieties. No knowledge of Hokkaidō Ainu is assumed, and Hokkaidō forms appear only where an explicit comparison sharpens the Sakhalin picture — most often in phonology (Chapter 4), where the correspondences are systematic, and in the verb, where the two branches have grammaticalised differently (Chapter 13, Chapter 16).
0.2 Data and conventions
Every numbered example in this grammar is attested unless it carries the explicit label constructed example. Constructed material is used sparingly — to complete a paradigm table or to spell out the prediction of an analysis first made here — and is never the sole support for a descriptive claim. The attested examples come from the published record of the language: Piłsudski’s Materials (Piłsudski 1912), the texts embedded in Dobrotvorsky’s dictionary (Dobrotvorsky 1875) in Sakaguchi’s edition (Sakaguchi 2021), the letters of Sentoku Tarōji (Tangiku & Ogihara 2001), the folktales of Asai Take (Murasaki 2001), the post-war recordings and elicited sentences of Hattori and Murasaki, and the examples of the grammatical literature itself. The full inventory, with an assessment of what each source can and cannot show, is given in Chapter 2.
Examples are glossed morpheme by morpheme after the Leipzig Glossing Rules (Comrie, Haspelmath & Bickel 2015). Each grammatical gloss is set in small capitals and expands on hover; the complete list, with the handful of additions this grammar makes for Sakhalin-specific categories, is kept on the abbreviations page, and the build of this site checks every gloss in every example against that list, so the two cannot drift apart.
Citation runs in an unbroken chain. A statement in the prose is supported by a numbered example; the example cites its source down to the page, text, or letter; author–year citations such as Chiri (1942: §76) link to the references, where works that could only be consulted second-hand are flagged as reported. In the other direction, every morpheme in an example, and every italicised Ainu form in the prose such as cise ‘house’, links to its entry in the dictionary of this site, which in turn indexes the corpus attestations of the word. The reader who distrusts a claim is meant to be able to reach the primary record without leaving the page’s links.
0.3 How to read an example
The following, from a Tarayka tale in Piłsudski’s collection, shows all the parts.
The number in parentheses counts examples within the chapter; cross-references in the prose use it. The first line, in italics, is the sentence as the source prints it — here Piłsudski’s own transcription, with his accent marks and his x for the velar realisation of the coda we write h. When source spelling and our romanisation coincide, this line is omitted. The second line is the morphemic segmentation in the romanisation of this grammar (Chapter 3): a hyphen separates affixes, as in ray-ahci ‘die’ plus the plural suffix, and an equals sign marks clitics. Each piece links to its dictionary entry.
Beneath it, the gloss line gives one gloss per word: lexical glosses in lower case (‘all’, ‘die’), grammatical categories in small capitals (PL), each of the latter expandable by hovering. Then comes the free translation in single quotes, sometimes followed by a literal rendering marked lit. where the idiom and the morphology pull apart. The last line is the source: the citation (Piłsudski 1912, text 2, line 4), the dialect tag (§0.4), and, where known, the locality and the speaker — here Sisratoka of Tarayka. Some examples close with a small-print note on points the gloss cannot carry, such as the note on -ahci spelt -axći in Chapter 26, where this tale is glossed in full.
0.4 Dialect labels and the West-coast baseline
Sakhalin Ainu was not uniform: the record splits, broadly, into a west-coast and an east-coast dialect group, surveyed in Chapter 2. Every example in this grammar carries one of three tags.
| Tag | Variety | Chief sources |
|---|---|---|
| WS | West Sakhalin: Rayciska, Odasu, Maoka, Tarantomari, and neighbours | Murasaki’s and Hattori’s recordings; the Asai Take folktales |
| ES | East Sakhalin: Tarayka, Otasan, Ay, Tunayci, and neighbours | Piłsudski’s Materials; the Sentoku letters |
| SA | Sakhalin, not further localisable | early wordlists; generalising statements in the literature |
Unmarked descriptive statements in the prose describe the west-coast varieties. This is a policy, not a judgement of importance: the west coast is where the twentieth-century record is deepest, with sound recordings, elicitation, and a described phonology, so it is the only variety for which negative evidence — the claim that something does not occur — carries weight. East-coast data are flagged ES wherever they diverge from the baseline or supply what the west-coast record lacks; since no descriptive grammar of any east-coast dialect exists, Piłsudski’s texts and Sentoku’s letters must serve there in place of recordings. A WS example for comparison with the ES example above:
Hokkaidō Ainu is not a dialect of this grammar’s object language and receives no tag; where a Hokkaidō form is cited for contrast, the prose says so explicitly, and the references page marks Hokkaidō-centred works as such. Nothing from Hokkaidō is ever silently folded into the description of Sakhalin.
0.5 The scholarship behind this grammar
The documentation of Sakhalin Ainu is the work of a short and traceable line of scholars, and this grammar is best read as the latest link in it. The first substantial record is Dobrotvorsky’s dictionary of 1875, compiled on the west coast while Sakhalin was a Russian penal colony (Dobrotvorsky 1875). The largest body of connected text is owed to Bronisław Piłsudski, whose Materials of 1912 transcribe and translate east-coast narratives collected in 1902–1905 (Piłsudski 1912). Chiri Mashiho then gave the language its first linguistic grammar, the 1942 アイヌ語法研究, written with the Sakhalin dialect at its centre though never carried as far as syntax (Chiri 1942).
After the war, Hattori Shirō brought the methods of structural phonology and dialect geography to the surviving speakers, producing the comparative dialect dictionary and the first systematic account of Sakhalin person marking (Hattori 1964), (Hattori 1961). From the 1960s Murasaki Kyōko recorded the last fluent generation on the west coast — above all Asai Take, the final native speaker, who died in 1994 — and published the materials, the grammar, and the folktale corpus that anchor the modern study of the language (Murasaki 1976), (Murasaki 1979), (Murasaki 2001).
The present generation has re-edited the old records and pushed the analysis into territory the documenters never treated: Anna Bugaeva on nominalisation and Ainu typology at large (Bugaeva 2016), (Bugaeva 2022), Elia Dal Corso on evidentiality, alignment, and the Piłsudski and Murasaki corpora (Dal Corso 2018), (Dal Corso 2024), Sakaguchi Ryō on number, numerals, and the philology of the early sources (Sakaguchi 2021), (Sakaguchi 2024), and Tangiku Itsuji on the Sentoku letters and the Sakhalin–Hokkaidō divide (Tangiku & Ogihara 2001), (Tangiku 2022). The full history of this research, source by source, is the subject of Chapter 2.