Chapter 163Japanese Loanwords and Their Phonological Adaptation

163.1 The Japanese contact layer

Vovin characterizes the lexical impact of Japanese on Hokkaido Ainu as making modern Ainu "inundated by loanwords from Japanese" Vovin (2022: §1). The flow has been almost entirely one-directional for the modern period: domestic animals, trade goods, administrative vocabulary, and — from the Meiji era onward — the full terminology of modern institutions all entered Ainu from Japanese. The reverse direction, Ainu elements entering Japanese through place names and regional vocabulary, is treated in Chapter 164 (Ainu Loanwords and Toponymy in Japanese); contact with Nivkh and Uilta, which shaped the Sakhalin lexicon but has no Hokkaido counterpart, is discussed in Chapter 165 (Northern Contact: Nivkh, Tungusic, and Manchu).

Vovin himself acknowledges that "there is no comprehensive study" of Japanese loanwords in Ainu Vovin (2022: §1). The account below draws on the etymology tags in Chiri's classified dictionary Chiri (1953), the word-formation remarks in Nakagawa Nakagawa (2024: §1.3), and attested corpus examples. Ainu is a language isolate with no genetic relationship to Japanese (see Chapter 5 (The Genetic Position of Ainu: Isolate Status and Macro-Comparison Controversies)), so the loans are a contact phenomenon and bring Japanese phonological structures that require systematic restructuring to fit the Ainu syllable canon.

163.2 Phonological adaptation

Nakagawa states the diagnostic plainly: free lexical words of three or more syllables are often loanwords (3音節以上の自立語は外来語であることが多い), citing tampaku ‘tobacco’ as a typical case Nakagawa (2024: §1.3). The underlying phonological mismatch is clear: Japanese contrasts voiced and voiceless obstruents; Ainu has no phonemic voicing (see Chapter 21 (Syllable Structure, Phonotactics, and Word-Edge Constraints)). Japanese also permits nasal-mora and long-vowel mora types that fall outside the Ainu (C)V(C) canon. Every borrowed form is therefore reshaped by one or more of the following repair operations, documented in Chiri's etymology tags Chiri (1953) and in Nakagawa Nakagawa (2024: §1.3).

The obstruent repairs reflect Ainu's lack of phonemic voicing: Japanese voiced stops surface voiceless or prenasalized (b → p / mp, d → t / nt, g → k). Prenasalization is the preferred repair for medial voiced stops in a nasal-compatible environment, as both tampaku ‘tobacco’ and kampi ‘paper’ confirm ‹corpus-confirmed›. Consonant clusters outside the (C)V(C) template are broken by epenthesis; the default epenthetic vowel is o or a copy vowel — a pattern attested in community comparisons of loanword shapes from multiple source languages ‹contested› aynu-corpora Discord (2023–2026) (nukopoli, aynu-corpora Discord, 2024-12-11 [227]). A further substitution affects the Japanese rhotic: Japanese /r/ has an [l]-allophone that maps to Ainu n rather than to Ainu r (a flap with no lateral allophone) ‹contested› aynu-corpora Discord (2023–2026) (antitwilight, aynu-corpora Discord, 2025-01-02 [400]).

The nasal spelling of prenasalized loans is an active orthographic debate. Nakagawa's convention writes the etymological labial nasal (hence tampaku, kampi, umma), because /n/ and /m/ neutralize to [m] before /p/ and /m/. Kana orthography uses ン throughout since kana encodes a single nasal mora. The choice between kampi and kanpi reflects an etymological or neutralization-based orthographic stance, not a phonological difference ‹contested› aynu-corpora Discord (2023–2026) (nukopoli, aynu-corpora Discord, 2024-11-16 [216, 217]).

One community-proposed case suggests that voicing may occasionally survive in a borrowed form: the garment term cinciri has been proposed as a loan of Japanese 綴れ (tsuzure) with the voiced affricate retained ‹speculative› aynu-corpora Discord (2023–2026) (scheuewsuiski, aynu-corpora Discord, 2024-11-05 [215]). The etymology is unverified against a primary dictionary source.

163.3 Chronological strata

163.4 Semantic distribution

Loans cluster in the semantic fields where post-contact material and administrative culture arrived: domestic animals, trade goods, clothing, tools, money and administration, and in the modern stratum technology and institutions Nakagawa (2024: §1.3). Core vocabulary — body parts, kinship terms, the natural-world lexicon, basic verbs, the native numerals one through ten, and pronouns — is native. The vigesimal numeral system treated in Chapter 54 (The Vigesimal Numeral System) is entirely native in its derivational base; classifiers and ordinals in Chapter 55 (Numeral Classifiers, Ordinals, and Quantifier Syntax) likewise draw on native morphology. Property-concept predicates in Chapter 99 (Property-Concept (Adjectival) Predication without a Copula) are native verbal predicates. The pattern is the classic "borrow the culture, keep the core" profile of contact lexicon.

The noun kampi ‘paper, document’ illustrates how a fully nativized loan can become a productive morphological base. Borrowed early as 'paper' (< kami 紙), it has generated a set of neologisms for educational and museum contexts in the revitalization period: kampi-nukat-cise 'school' (paper–look-at–house), kanpi pu 'library' (paper–storehouse), imeru-kampi 'email' (lightning–paper) Ijäs (2023: §5). A loanword functioning as a stem in noun incorporation and in productive compounding is indistinguishable in its morphological behavior from a native root.

163.5 Degree of integration

Early loans are fully integrated. They bear Ainu accentuation according to the prosodic rules described in Chapter 24 (The Prosodic Unit and the Accent-versus-Tone Analysis): the geminate in umma ‘horse’ creates a heavy first syllable that receives the accent nucleus by the default rule for CVC.CV words, so the form sounds native, not Japanese. Early loans also participate freely in compounding and incorporation (see examples in §3.1 above). Native obligatorily-possessed nouns carry affiliative suffixes in possessed contexts; borrowed nouns such as umma ‘horse’ and kampi ‘paper’ are not obligatorily possessed and do not display this morphology — a diagnostic that separates many loans from native inalienable nouns.

Modern loans are less integrated. They frequently retain Japanese syllable structure, and in the new-speaker texts produced in language-revitalization contexts, the boundary between a nativized loanword and a code-switch from Japanese is structurally difficult to establish ‹speculative›. The compound umma menoko illustrates how native and loan morphology interact productively: the bare juxtaposition is ambiguous between 'horse and girl' (appositive) and 'horse-girl' (compound); inserting the copula explicitly as umma-ne-menoko removes the ambiguity ‹speculative› aynu-corpora Discord (2023–2026) (nukopoli, aynu-corpora Discord, 2023-12-12 [356]).

163.6 Diagnostics for identifying Japanese loans

Four properties together identify a form as a Japanese loan in Hokkaido Ainu:

• Syllable count. A free lexical root of three or more syllables is a strong prior for borrowing Nakagawa (2024: §1.3). Native Ainu roots are predominantly mono- or disyllabic; polysyllabic roots arise through borrowing or through productive — and morphologically transparent — affixal derivation.
• Recognizable phonological repair. Prenasalization of a medial voiced stop (-mp-, -nt-), gemination of a stem-initial nasal consonant, or an unexpected initial m- where a Japanese source has n- all point to a Japanese etymon.
• Absence of affiliative morphology. Obligatorily-possessed native nouns display a concept/affiliative alternation (所属形; see Chapter 45 (Morphophonology of the Affiliative Suffix and Echo/Copy Vowels)). Most Japanese loans are not obligatorily possessed and lack this alternation.
• Chiri dictionary etymology tags. The tags [<日本語…] and [<Jap.…] in Chiri (1953) mark the identified Japanese-loan inventory and remain the most systematic primary record. Vovin notes the lack of a comprehensive successor survey Vovin (2022: §1).

The name for the source language itself illustrates one of the finer integration distinctions. The two-word phrase sisam itak 'the speech of the Wajin' (a descriptive NP) and the single compound sisamitak 'Japanese [language]' (a language name) are argued to differ semantically and phonologically: the former has two independent accentual words, while the latter functions as a single compound with one accent domain. Writing them together obscures the accentuation and conflates the two meanings ‹speculative› aynu-corpora Discord (2023–2026) (nukopoli, aynu-corpora Discord, 2024-11-05 [299]; 2024-12-20 [549]).

Ochiai's study of sikerpe ‘Amur cork tree fruit (Phellodendron amurense)’ demonstrates a methodological corollary from the reverse direction: a Japanese dialect tree-name preserving a reconstructible intermediate Ainu form — *si kor pe before the change to sikerpe — can be used both to date the borrowing and to anchor the proto-Ainu reconstruction Ochiai (2026: 199). The evidence for Ainu loans in Old Japanese, including Vovin's Man'yōshū material, is taken up in Chapter 164 (Ainu Loanwords and Toponymy in Japanese) and Chapter 166 (Macro-Comparison and Deep-Contact Controversies).