Chapter 50Demonstratives, Anaphora, and Definiteness

50.1 Overview: a bipartite deixis system

Hokkaido Ainu demonstratives divide into two formally distinct sets. One set — the deictic (on-site) forms — points to referents in the physical scene; the other — the anaphoric forms — tracks referents established in discourse. Japanese uses a single three-way この/その/あの series for both functions; Hokkaido Ainu keeps the two apart with different lexical roots Nakagawa (2024: 284). Kindaichi and Chiri's classical grammar already identified the deictic prenominals as fusions of three locative roots (te 'here', ta 'there', to 'yonder') with the copula an, while the anaphoric ne-series points in the discourse stream, not in space Kindaichi & Chiri (1936: 57–58). Refsing's Shizunai description formalises the whole system as a two-dimensional root paradigm Refsing (1986: 109); the typological framing of Diessel's demonstrative taxonomy maps the anaphoric forms onto a recognitional–anaphoric cluster distinct from both exophoric pointing and grammaticalized definite articles (Diessel 1999).

Minor deictic prenominals in Saru include verse-register tanan (pl. tanoka(y)) and emphatic tapan (pl. tapoka(y)), the latter most frequent in Chitose address contexts Nakagawa (2024: 284). The independent personal pronouns are treated in Chapter 49 (Personal, Reflexive, and Emphatic Pronouns); the interrogative and indefinite pro-forms in Chapter 51 (Interrogative, Indefinite, and Negative-Polarity Pro-forms).

50.2 Deictic prenominals

The etymology of the deictic prenominals is transparent. Kindaichi and Chiri derive tan from ta + copula an, and toan from to + an; the medial taan is the long (information-richer) variant of tan, and the plurals tanokay / tonokay are formed by fusing the root with the plural existential okay Kindaichi & Chiri (1936: 58); Nakagawa (2024: 288). The functional difference between tan and taan parallels Japanese この vs. その: tan is used when the referent is immediately at hand, taan when it lies within the speaker's perceptual sphere but is not immediately present, and toan when it is away from both speaker and addressee Nakagawa (2024: 283–284). Community discussion has proposed an iconic account of the vowel gradient: longer form encodes more spatial distance-information, front raised vowel signals proximity, back lowered vowel signals remoteness ‹speculative› aynu-corpora Discord (2023–2026) (nukopoli, 2025-12-27). The functional distribution is uncontroversial Nakagawa (2024: 283–284); the iconicity explanation is unverified.

The distal toan is also the principal vehicle for nominal third-person reference: Hokkaido Ainu uses toan kur 'that person' or nea kur 'that person (aforementioned)' where languages with grammaticalized third-person pronouns use them, since the third-person pronouns of the paradigm treated in Chapter 49 (Personal, Reflexive, and Emphatic Pronouns) are rare in actual discourse Nakagawa (2024: 256).

50.3 Anaphoric and recognitional demonstratives

The anaphoric prenominals — ne, nea, newaan — are formally separate from the deictic set and cannot be used to point at a scene-present referent. nea picks up a referent already established in the discourse; in a narrative, nea menoko means 'the woman (who was just spoken of)' and does not admit a deictic (scene-pointing) reading Nakagawa (2024: 284). The weaker ne marks a referent as immediately prior in the discourse; newaan is more emphatic.

Corpus counts confirm the predominance of the anaphoric function in narrative text: the surface string nea occurs approximately 1,984 times in the ainu-corpora aligned set — substantially more than toan (c. 735 tokens) and tanpe (c. 391 tokens) combined — consistent with its role as the primary anaphora-tracking device of narrative ‹corpus-confirmed› Nakagawa (2024: 287). (Surface counts conflate any homophonous forms and are upper bounds.)

Third-person reference. Where Japanese uses 彼 / 彼女, Hokkaido Ainu characteristically uses a demonstrative NP — nea menoko 'that woman (referred to)', toan kur 'that person (over there)'. The third-person and fourth-person pronouns of the independent paradigm occur, but they are rare; the demonstrative NP is the default anaphora-tracking strategy in oral literature and in narrative prose Nakagawa (2024: 256). The interaction of this strategy with pronominal-argument interpretations of the person affixes is considered in Chapter 94 (Pro-Drop, (Non)configurationality, and the Pronominal-Argument Debate).

Dialect variation. nea is pan-Hokkaido; most dialects outside Saru and Chitose use nean (pronoun neanpe) — attested in Horobetsu, Shizunai, and Ishikari, with both forms occurring in some Horobetsu and Ishikari speakers Nakagawa (2024: 285–286). Shizunai and the dialects to the east additionally have ikia (pl. ikirok; pronoun ikiap), a second anaphoric prenominal functionally close to nea but more restricted in scope; it is rare in Saru Nakagawa (2024: 285). The Ijäs pedagogical grammar covers the core nea forms alongside the deictic series Ijäs (2023).

50.4 Demonstrative pronouns

Each prenominal may fuse with the formal noun p ‘thing (nominalizer)’ (~pe) to produce a demonstrative pronoun referring to a thing rather than modifying a noun head: tan-pe 'this (thing)', toan-pe 'that (thing)', newaan-pe / nea-p 'the aforementioned thing', and the eastern ikia-p (Shizunai-and-east) Nakagawa (2024: 287). The suffix alternation -pe / -p follows the phonological environment of the nominalizer, treated in Chapter 101 (The Nominalizers -p/-pe and -i/-hi (Participant, Event, Place, Fact)).

The indefinite pronoun nep 'something / what' — formed from the same root ne- plus the nominalizer — is an interrogative/indefinite, not a demonstrative. For anaphoric reference to a thing, Hokkaido Ainu uses newaanpe or neap, not nep Nakagawa (2024: 287). This gap in the anaphoric pronoun paradigm is an incidental consequence of nep's grammaticalization as an indefinite/interrogative; the full indefinite and interrogative series are treated in Chapter 51 (Interrogative, Indefinite, and Negative-Polarity Pro-forms).

50.5 Locative roots and case-marked forms

The three roots te ‘here’, ta ‘there’, and to ‘yonder’ function as locative nouns taking case particles directly, not as prenominal modifiers. Only ta and to enter the prenominal by fusing with the copula (ta-an → tan, to-an → toan). The proximal root te does not produce a prenominal and cannot precede a head noun — *te menoko is ungrammatical; te only takes case particles (te ta 'here', te un 'this way', te wano 'from now') Nakagawa (2024: 289).

Case-marked forms of the deictic prenominals arise by ellipsis of the formal noun hi before the case particle: taan ta 'here' < taan hi ta 'the-place-of-this at'; toan un 'over there (toward)' < toan hi un Nakagawa (2024: 288). These locative uses intersect with the relational-noun system; see Chapter 52 (Relational and Spatial (Locative) Nouns and Their Possessed Forms).

50.6 The Shizunai demonstrative grid and dialect variation

Refsing's Shizunai description assembles the demonstrative, interrogative, and indefinite forms into a two-dimensional paradigm of roots and semantic dimensions Refsing (1986: 109). It is the most explicit systematic statement of the opposition's internal logic in any published Hokkaido description.

Two observations follow from the grid. The anaphoric root ne- is formally identical to the interrogative root of nep 'what', nen 'who', and neyta 'where', and to the copula ne. Refsing flags this convergence as possibly more than coincidence Refsing (1986: 108) ‹speculative›. The place-row entry for the anaphoric column is empty — a systematic gap: there is no form *neta or *neyta functioning as 'the aforementioned place'.

Intra-Hokkaido variation affects chiefly the anaphoric prenominal and the distal pair.

• nea vs. nean. Saru and Chitose use nea; most other Hokkaido dialects (Horobetsu, Shizunai, Ishikari) use nean (pronoun neanpe), with both forms occurring in some Horobetsu and Ishikari speakers Nakagawa (2024: 285–286).
• ikia / ikiap. Shizunai and the dialects east of it have a second anaphoric prenominal ikia (pl. ikirok; pronoun ikiap), functionally close to nea but appearing to be more restricted in discourse scope. It is rare in Saru Nakagawa (2024: 285).
• toan vs. toon. Saru favours toan; both forms occur in Chitose; Obihiro and Tokachi avoid toan; Ishikari lacks toon Nakagawa (2024: 285). Takahashi's Tokachi description documents the local demonstrative system and the range of deictic forms available there Takahashi (2011).

50.7 Definiteness in an article-less language

Hokkaido Ainu has no grammatical article. Neither the deictic nor the anaphoric demonstratives are obligatory in contexts where a definite-article language would require an article; bare nouns occur freely in argument position without referential specification Nakagawa (2024: 283) ‹consensus›. The language encodes identifiability through a cluster of constructions, none of which individually constitutes an article system:

• Anaphoric demonstrative nea / ne marks a referent as established and salient in discourse — the functional counterpart of the definite article in post-introduction contexts. Its high corpus frequency relative to the deictic forms (§3 above) is consistent with this article-like role ‹corpus-suggested›.
• Possessive morphology. The personal possessor prefix combined with the affiliative form of an inalienable noun singles out the referent as uniquely identifiable relative to its possessor. The alienable/inalienable split and the semantics of obligatory possession are treated in Chapter 43 (The Alienable/Inalienable Split and Possessive Classification).
• Proper nouns and naming are inherently referentially unique and carry no additional definiteness morphology.
• Discourse position. A sentence-initial topic NP is typically presupposed as identifiable; first-mention NPs in non-topic position are indefinite by discourse default, without any morphological signal.

Whether nea should be analysed as a grammaticalized definite article is contested ‹contested›. Against the article analysis: nea is absent from many clearly definite contexts where English or Japanese would require the article; it alternates with a bare noun in narrative without apparent semantic contrast; and its distribution is better described as anaphoric-recognitional — marking a referent as recoverable from discourse — than as an absolute definiteness category Nakagawa (2024: 284). Refsing glosses the Shizunai cognate as 'the said', which aligns with recognitional use rather than a grammaticalized article Refsing (1986: 109). In Diessel's typology, recognitional demonstratives track referents that are identifiable through shared knowledge or discourse memory but are not fully grammaticalized articles (Diessel 1999). The related claim — that the affiliative form of certain nouns carries definiteness-like semantics in post-possessive contexts — is evaluated in that chapter. For the clause-level question of how topics and foci interact with the demonstrative system, see Chapter 42 (The Structure of the Noun Phrase).

The interaction of the demonstrative system with clausal interrogation — in particular with sentence-final question particles and with the interrogative existential frame — is treated in Chapter 144 (Interrogative Strategies, Question Particles, and Evidential Questions).