Chapter 166Macro-Comparison and Deep-Contact Controversies

166.1 The isolate baseline

Ainu has no confirmed genetic relative. Every proposal to link it to another language family has been examined and set aside; Janhunen's verdict — that Ainu "does not show unambiguous indications of being associated with any other language family" — expresses the current consensus ‹consensus› Janhunen (2022: §2); Alonso de la Fuente (2022: §1); (Vovin 1993). Within its own lineage Ainu constitutes a small family — sometimes called Ainuic or Kurilic — of three closely related varieties: Hokkaido, Sakhalin, and the Northern Kuriles (Chapter 162 (Sakhalin and Kuril Ainu: The External Comparison)). The question addressed here is not the isolate status itself, which is settled, but two connected and genuinely open questions: the nature and direction of contact between Ainu and neighbouring languages, especially Japonic; and whether any proposed deep-time comparison or areal alignment survives methodological scrutiny. The isolate status and its evidential basis are set out in Chapter 5 (The Genetic Position of Ainu: Isolate Status and Macro-Comparison Controversies); the ethnolinguistic context in Chapter 3 (The Ainu People, Their Homeland, and History).

For an isolate the comparative method is unavailable: there is no sister language against which to project correspondences back into a proto-system. Three analytic tools remain — internal reconstruction, contact-induced-change analysis, and typology — with the important qualification that typological reasoning carries no probatory weight for genetic questions and must be used "with utmost caution" Alonso de la Fuente (2022: §1). The methodologically recoverable depth is also shallow: internal reconstruction reaches approximately 1,000 years without unsupported extrapolation, terminating at a Common Ainu stage no earlier than the thirteenth-century CE branching that separated the Sakhalin varieties from Hokkaido Alonso de la Fuente (2022: §1, §4). The reconstruction of Proto-Ainu and Common Ainu is treated in Chapter 156 (Proto-Ainu Reconstruction: Segments and Accent Classes) and Chapter 158 (Internal Reconstruction and Grammaticalization Pathways).

166.2 The macro-comparison catalogue

The table below surveys the external comparisons that have been proposed for Ainu and records their current status. Judgements follow Janhunen Janhunen (2022: §2–3c), Alonso de la Fuente Alonso de la Fuente (2022: §1), and the three published reviews of Vovin (1993): Sidwell (1996), Abondolo (1997), and Witczak (1999), all cited second-hand via Alonso Alonso de la Fuente (2022: §1).

A structural point underlies the table. Proposing a genetic relationship for an isolate places the entire burden of proof on the proposer, because the null hypothesis — independent origin — accounts for all of the data without additional apparatus. Lexical resemblances between any two languages occur by chance at a predictable rate; only systematic sound-to-sound correspondences across a significant portion of core vocabulary, independent of areal diffusion, constitutes evidence for common descent Alonso de la Fuente (2022: §1). No proposal for Ainu has satisfied this standard.

166.3 Old Japanese loanwords into Ainu

Contact between Ainu and Japonic is the best-documented diachronic layer in Ainu's external history, and the directionality is not in dispute for the early stratum. A set of Ainu items preserve phonological features of Old Japanese or Pre-Old-Japanese that had disappeared from Japanese before the Nara period, placing the borrowing events "around the middle of the first millennium CE" Janhunen (2022: §3c). The diagnostics are the retention of a labial stop where later Japanese shows a fricative (Pre-OJ *p- → later Japanese /h-/), and the preservation of sequences that subsequently monophthongized. The regularly cited items are:

• sippo ‘salt’ — preserves Pre-OJ *siopo before the labial lenition that yields Modern Japanese 塩 shio Janhunen (2022: §3c).
• pasuy ‘prayer stick, ritual chopstick’ — compared with OJ hashi 'chopstick'; Ainu retains the original stop p- lost in Japanese Janhunen (2022: §3c).
• kamuy ‘deity, spirit’ — an early cultural borrowing whose deep distribution across all Ainu lects confirms a pre-branching age Janhunen (2022: §3c).

These items are morphologically opaque in Ainu — no segment maps onto a recoverable Ainu morpheme — whereas inherited vocabulary often remains internally analyzable. Compare cep ‘fish’, reconstructed as Pre-PA *ti=(ʔ)e pe 'the thing eaten by everyone' (< fourth-person prefix ci= + e ‘eat’ + nominalizer -p), whose internal structure remains transparent after a millennium of phonological change Alonso de la Fuente (2022: §4, Table 18). (Chapter 158 (Internal Reconstruction and Grammaticalization Pathways).) Opacity versus transparency of this kind is a standard heuristic for separating loans from inherited vocabulary.

Later loanwords from Japanese, received after the post-Nara sound changes, carry the expected modern phonological shapes: cikenki (Saru; < Standard Japanese 付け木 tsukegi 'matches'), antuki (Saru; < 小豆 azuki 'azuki bean', with prenasalization from Tōhoku Japanese) Alonso de la Fuente (2022: §1). The two strata — pre-Nara and post-Nara — are therefore separable through phonological shape alone.

The Eurasian Wanderwort kotan ‘settlement, village’ — attested throughout Hokkaido and appearing at 4,210 corpus tokens — sits outside this Japanese-specific stratum. Janhunen (2022 §5) classifies it as a continental Wanderwort; whether it entered Ainu via Japanese, via Tungusic, or through an older continental network is unresolved ‹speculative› Janhunen (2022: §5); the community has raised the same question without resolving it aynu-corpora Discord (2023–2026) (antitwilight, aynu-corpora Discord, 2024-11-15, speculative). The word sisam ‘Japanese person’ is, by contrast, an Ainu coinage: reconstructible as Pre-PA *sii+sam/o with native Ainu morphology, surviving in the Yakumo dialect as samó Janhunen (2022: §4).

166.4 The reverse direction: Ainu elements in early Japonic

The claim that Ainu was once spoken more broadly across the Japanese archipelago and left loanwords in Old Japanese is the most contested diachronic proposal currently appearing in peer-reviewed Ainu scholarship. Vovin (2022) argues that Ainu extended through Kantō and Tōhoku before the Yayoi expansion, and that the Man'yōshū — particularly East Old Japanese (EOJ) in books 14 and 20 and West Old Japanese (WOJ) in books 17 and 19 — contains loans from this substrate ‹contested› Vovin (2022: §§1–6).

Among the specific etymologies Vovin proposes:

• EOJ kuta 'house' ← Ainu kotan ‘settlement’ (village→house shift parallel to Latin vicus/villa) Vovin (2022: §3).
• WOJ porö 'big' ← Ainu poro ‘big, large’ Vovin (2022: §4).
• OJ iruka 'dolphin' ← Ainu rika 'whale-fat' Vovin (2022: §2).
• OJ kanimba 'cherry' ← Ainu karinpa 'cherry' Vovin (2022: §5).
• Place names in Man'yōshū: Kanipa ← Ainu ka-ne-pa 'upper bank'; Ikuri ← Ainu ikure Vovin (2022: §3).

The loan-phonology proposed is systematic: Ainu geminate clusters simplify across the borrowing boundary (Ainu -kk- → OJ -k-); Ainu word-final consonants are lost in EOJ surface forms; and the Proto-Japonic vowel shifts *o > u and *e > i in EOJ environments are said to reflect the Ainu source shapes Vovin (2022: §2).

The argument has not been independently corroborated in either Old Japanese philology or Ainu linguistics. The structural problem is that short monosyllabic resemblances — like WOJ porö ~ Ainu poro — fall within the range expected by chance between unrelated languages, and the required semantic shifts must be stipulated rather than derived from cross-linguistically attested patterns. Janhunen (2022) characterises the shared features under the more general rubric of "Jōmonization of Japanese" — Ainu-type languages influencing Japonic typologically — without endorsing individual lexical etymologies of this kind Janhunen (2022: §3c). The substrate hypothesis and the lexical-etymology program address different phenomena and should be kept analytically separate: the former has structural support; the latter remains unconfirmed ‹contested›.

166.5 The Jōmon-substrate thesis and Ainu ethnogenesis

Independent of the lexical-etymology debate, Janhunen (2022) develops a substrate account of Japonic that is better evidenced and more broadly accepted. The Japonic family arrived on the archipelago with the Yayoi agricultural expansion from the Korean Peninsula, roughly between 900 BCE and 500 CE, and displaced the earlier "Jōmonic" linguistic communities whose long prior presence shaped Japonic phonology, morphology, and vocabulary in ways that remain incompletely disentangled ‹contested› Janhunen (2022: §3a–3c). On this account, Ainu is the sole surviving Jōmon language — its survival "an accident" in the northeast while the Japonic expansion replaced the other Jōmon languages of the archipelago Janhunen (2022: §3a).

The ethnolinguistic implication is that the Ainu homeland was not Hokkaido but Honshu. The present Hokkaido distribution results from northward movement: the Satsumon culture (c. 700–1200 CE) carried the Ainu linguistic ancestors northward, reaching northern Hokkaido around 1000 CE, southern Sakhalin around 1300 CE, and the Kuriles around 1500–1600 CE Janhunen (2022: §3a–3b). The Emishi (蝦夷) of northern Honshu are most plausibly treated as a proto-historical complex partly ancestral to the Ainu Janhunen (2022: §3a). The Okhotsk culture (c. 600–1600 CE), whose language Janhunen associates with early Ghilyak/Amuric, was absorbed by the Satsumon population in northern Hokkaido and Sakhalin, contributing a borrowing residue treated in Chapter 165 (Northern Contact: Nivkh, Tungusic, and Manchu).

Community discussion connects the OJ ethnonym emi₁si with the Sakhalin Ainu ethnonym enciw. Antitwilight (aynu-corpora Discord, 2024-12-23) proposes that OJ emi₁si borrows from a pre-Proto-Ainu form *æmi-ti, with the segmentation *ti = first-person plural, regularly giving > enciw aynu-corpora Discord (2023–2026) ‹speculative›. This partially converges with Janhunen's (2022 §4) note that enciw ~ Emishi/Ebisu/Ezo *emisV~*emiciV (citing Kindaichi 1960), though the specific morpheme segmentation proposed by the community goes beyond the published source and cannot be verified against it.

The word tunakay ‘reindeer’ — which entered Japanese as tonakai — is a probable loan from a pre-Proto-Ghilyak source into Ainu, illustrating contact at the Okhotsk cultural interface ‹contested› Janhunen (2022: §3c).

166.6 Toponymic substrate on Honshu

A Honshu Ainu homeland should have left a toponymic residue; place-name substrates are well attested where a speech community has been displaced but not entirely erased. Two bodies of candidate evidence circulate in the literature: analysis of Tōhoku place names with elements cognate with productive Ainu geographical vocabulary, and Vovin's (2022 §3) identifications of Ainu-origin proper names in the Man'yōshū (§4 above).

Ainu geographical vocabulary is morphologically productive and well-documented in Hokkaido: nay ‘stream, small river’, pet ‘river’, nupuri ‘mountain’, wakka ‘fresh water’, and or ‘place, location’ appear in hundreds of Hokkaido place names. The community has noted specific Tōhoku cases: nukopoli (aynu-corpora Discord, 2026-01-22, proposed) analyses the Iwate surname and toponym 乙茂内 (Otomonai) as Ainu o-tum-nay 'head-middle-stream', citing the concentration of 内 (-nai) toponyms in the surrounding area as convergent evidence aynu-corpora Discord (2023–2026) ‹speculative›. The morphological segmentation is plausible against known Ainu geographical morphology, but confirmation requires independent phonological correspondence evidence that is not yet available.

A systematic survey of Tōhoku toponymy against the Ainu geographical lexicon remains a research gap. The existing evidence — from the matagi hunter-gatherer vocabulary of Tōhoku, from the Emishi cultural zone extending considerably south of modern Hokkaido, and from the individual proposals above — is suggestive but not yet probative ‹speculative›.

166.7 Separating contact, inheritance, and chance

The central methodological problem for all the proposals above is the triage between three origins for any given resemblance: inheritance from a common ancestor; borrowing through contact; and chance. For an isolate, the inheritance route is ruled out by definition for resemblances with established families. The issue is therefore always whether a similarity is contact-induced or accidental.

The standard diagnostics applied in the Ainu contact literature Shiraishi & Tangiku (2022: §3) are:

• Direction of spread: does the item appear only in lects geographically adjacent to the supposed source, or is it pan-Ainu? A pan-Ainu distribution implies a pre-branching (Common Ainu or Proto-Ainu) age, excluding recent contact.
• Morphological analyzability: inherited vocabulary tends to be internally structured; loans tend to be opaque. sippo ‘salt’ carries no recoverable Ainu morpheme; cep ‘fish’ (< ci=e-p) is fully transparent — see Chapter 158 (Internal Reconstruction and Grammaticalization Pathways).
• Loan phonology: for a directed borrowing, the donor-language form and the recipient-language form must be connected by phonologically plausible, preferably recurrent, processes. Ad hoc adjustments stipulated for a single item weaken the case considerably.
• Systematic correspondence: a genetic or deep-contact relationship requires recurrent sound-to-sound correspondences across a significant portion of core vocabulary, not sporadic resemblances.

By these criteria the early Japanese→Ainu stratum (§3) meets most diagnostics and is well established ‹consensus›. Vovin's reverse-direction program (§4) proposes systematic loan-phonology but produces thin correspondence sets and semantically variable links, and has not been replicated ‹contested›. The Jōmon-substrate thesis (§5) operates at a depth that forecloses direct lexical testing; it is evaluated primarily through population history and typological alignment and is best treated as a working hypothesis rather than a demonstrated result ‹contested›.

Two community discussions illustrate the methodological divide. The proposal that Ainu si= (reflexive prefix, treated in Chapter 62 (Architecture of the Personal-Affix System: The Four Persons and the S/A/O Paradigms)) is cognate with Korean and Old Japanese reflexive pronouns — aomidori_cs_, aynu-corpora Discord, 2023-12-08 aynu-corpora Discord (2023–2026) — invokes forms short enough to match by chance and a semantic field (self-reference) common enough to arise independently; without a systematic correspondence framework the proposal cannot advance beyond conjecture ‹speculative›. The set of proposed Korean-origin vocabulary items (casi, notak, mose, nay, toy — antitwilight, aynu-corpora Discord, 2024-11-15, proposed aynu-corpora Discord (2023–2026)) faces the same difficulty: casi 'fortification' in particular has a plausible Ainu-internal structure and is not compelled to be a contact loan ‹contested›. Both sets are the kind of etymologies that Alonso (2022 §1) and Janhunen (2022 §2) explicitly bracket as undemonstrable under the current methodological standard for Ainu.

166.8 The Amur / North Pacific areal layer

At the typological level Ainu shares features with North Pacific Rim languages — polysynthetic morphology, head-marking, verb-final word order, extensive nominalization-based clause linkage — that place it loosely within a broad Northeast Asian typological zone, though Janhunen (2022 §3c) notes that Ainu "stands apart" even within that grouping. The typological characterisation is descriptive, not a genetic or even a firm areal claim.

The contact-driven Amur areal layer has a real lexical residue, but its Hokkaido signature is thin. The primary northern-contact zone was Sakhalin, where Ainu, Nivkh, Uilta, and Manchu (via the Santan trade) met; the lexical traffic is documented in Chapter 165 (Northern Contact: Nivkh, Tungusic, and Manchu). The items that reached Hokkaido Ainu directly — kaya ‘sail’ (< Nivkh, with the 'fish-skin garment' sense retained in Sakhalin but reduced to 'sail' in Hokkaido), sintoko ‘cask’ (morphologically opaque; possible ultimate Eurasian Wanderwort), san ‘offering shelf’ (< Nivkh cʰaŋ, Uilta saan) — are documented borrowings, not indicators of genetic relationship Shiraishi & Tangiku (2022: §3). Whether the Sakhalin contact situation constituted a true Sprachbund is left open by the most recent survey Shiraishi & Tangiku (2022: §1).

Alonso de la Fuente (2022 §2.4) observes that the reconstructed Proto-Ainu consonant inventory looks typologically like a mixed Southwest/Northeast Eurasian language that may have acquired some Southeast Asian-type features through contact: the merger of *l and *ɾ, the single fricative /s/, and the bare stop series /p t k/ are consistent with this areal profile. The observation is typological inference, not a contact-language argument, and it post-dates the genetic-affiliation question addressed in this chapter. The inventory is documented in Chapter 16 (The Consonant Inventory and Its Phonetic Realization) and reconstructed in Chapter 156 (Proto-Ainu Reconstruction: Segments and Accent Classes) ‹speculative› Alonso de la Fuente (2022: §2.4).