Chapter 41Nominal Number: Transnumerality, -utar, and the Locus of Plurality

41.1 Transnumerality: nouns unspecified for number

Hokkaido Ainu nouns carry no grammatical number. A bare noun names a referent without specifying whether one or several are intended: suma ‘stone’ is the form for 'a stone', '(the) stone', and 'stones' alike; the same holds for cep ‘fish’, cise ‘house’, and kamuy ‘spirit, deity’ Nakagawa (2024: 259); Kindaichi & Chiri (1936: 30). In Corbett's typology this is the transnumeral pattern: the noun is underspecified for number rather than syncretically encoding a singular or plural value through a single form (Corbett 2000: 9–11). Obligatory singular/plural agreement of the kind found in English or French is absent.

Plurality may be made explicit by the formal noun utar ‘plural; relatives, people’, but this addition is strictly optional. In the 209,878-sentence Hokkaido corpus utar records 8,051 tokens (rank 30 among distinct types) — high frequency reflecting pragmatic salience, not grammatical compulsion ‹corpus-confirmed›.

41.2 The formal noun utar — distributional status

utar belongs to the closed class of formal nouns — weakly autonomous items that form an NP only when preceded by another element Nakagawa (2024: 257–258); Nakagawa (2022: 481–482). Its syntactic distribution is uniquely wide within that class. Whereas hi ('thing/matter') takes only a VP host, and p/pe ('thing/one') takes a VP or prenominal, utar follows a VP, a bare common noun, or a prenominal — the only formal noun that can immediately follow a noun without an intervening predication Nakagawa (2024: 258); Nakagawa (2022: 482).

Sources: Nakagawa (2024: 257–258); Nakagawa (2022: 482).

When following a bare noun, utar adds a collective or associative plural reading without predicational content:

The broader distribution of formal nouns is treated in Chapter 35 (The Formal / Defective Noun Set (ruwe, hawe, siri, humi, hi, pe, kur)). For the p/pe nominalizer in clausal contexts see Chapter 101 (The Nominalizers -p/-pe and -i/-hi (Participant, Event, Place, Fact)).

41.3 Semantics: associative reading and the animacy debate

The meaning of utar in pluralizing use is associative rather than strictly additive. Kindaichi and Chiri's example acha utar arki translates as 'uncle and his group came', not 'several uncles came': the form implies a cluster of associates gathered around the named referent, potentially including members of different categories Kindaichi & Chiri (1936: 30). Ijäs's pedagogical grammar presents the same associative reading Ijäs (2023).

Whether utar is restricted to animate referents is a genuine disagreement ‹contested›. Kindaichi and Chiri state that utar applies to persons and animals (kamuy utar, ainu utar, seta utar 'dogs', cikap utar 'birds'), and that suma utar 'stones' and ni utar 'trees' are disallowed unless the referent is personified Kindaichi & Chiri (1936: 30–31). Nakagawa, working from Saru material, takes the opposite position: utar freely pluralizes inanimates (suma utar, ni utar, rus utar 'furs', coypep utar 'lacquerware'), the extension to all entities motivated by the animist conception that every being carries a spirit Nakagawa (2024: 259). The Saru corpus attests inanimate utar directly:

The discrepancy most likely reflects dialect variation rather than a uniform pan-Hokkaido norm: the animacy restriction appears stronger in eastern and classical-norm speech, while Saru extends utar to inanimates within the animist conceptual frame. The two textbook illustrations do not fare alike in the corpus: suma utar 'stones' does not occur, but ni utar 'trees' does, in a Saru/Chitose prose report — nítay or ta ni utar ka poronno a=momka 'in the forest many trees too were swept away'. The inanimate extension therefore has attested support, thin though it is: two occurrences against none for the stone example ‹corpus-confirmed›.

41.4 The lexicalized plural hekattar

One nominal has grammaticalized a dedicated plural form through phonological fusion with utar: hekaci ‘child’ + utar contracts to hekattar ‘children’, which serves as the fixed plural of 'child' Kindaichi & Chiri (1936: 31); Nakagawa (2024: 259). The corpus records 297 tokens of hekattar (rank 509), confirming lexicalization ‹corpus-confirmed›. The fused form can itself be followed by another utar — hekattar utar — for additional collective emphasis Kindaichi & Chiri (1936: 31).

A separate nominal strategy for inherent plurality is reduplication: chaichai 'twigs', kaukau 'hailstones', kankan 'intestines', and ramram 'scales' carry plural semantics through their reduplicated structure without requiring utar Kindaichi & Chiri (1936: 31). These are lexicalized; reduplication is not a productive synchronic plural strategy. The productive resource for ad hoc pluralization remains utar or the verbal number morphology described in Chapter 58 (Verbal Number: Suppletive Stems and the Pluractional -pa).

41.5 utar and utari(hi): the form–meaning distinction

The string utar covers two related but functionally distinct items. The first is the formal noun pluralizer described above. The second is a full common noun meaning 'relatives, kin, companions', with its own affiliative form utari (long form utarihi), which surfaces under possession Nakagawa (2024: 250–252):

The referential scope of k=utari is strictly third-person: in k=utari ku=tura 'I took my family', the speaker does not belong to the group named. Likewise, a=utari opitta totek wa oka 'is your family all well?' takes a third-person verb because the referent is the addressee's relatives, not an inclusive 'you all' Nakagawa (2024: 251). The -utari that enters Tokachi anutari 'we (inclusive)' combined with a fourth-person verb edges toward pronominal function — Nakagawa characterises the form as 'halfway between pronoun and common noun' Nakagawa (2024: 252) — but the full treatment of pronominal -utari belongs to Chapter 49 (Personal, Reflexive, and Emphatic Pronouns).

The corpus records 1,501 tokens of utari (rank 130), mainly as a possessed group or kinship noun. Community analysis has proposed that the affiliative form is nearly obligatory in this function because 'companions belonging to no one' is semantically anomalous — the noun almost always surfaces possessed aynu-corpora Discord (2023–2026) (thegodofneet, 2025) ‹contested›. This aligns with the general behaviour of the affiliative form across the Ainu noun paradigm: the possessed shape presupposes a contextually available referent to which the named entity belongs Nakagawa (2024: 251–252).

The semantics of utar as a common noun extends to kinship idioms. utar ni nare 'become utar with someone' carries the meaning 'to marry (into a family)', since marriage incorporates one into a kin group aynu-corpora Discord (2023–2026) (thegodofneet, 2025) ‹corpus-suggested›. Kinship compounds on -utari can also name a pair rather than a plurality of the base: onautari and unuutari (literally 'father-group' and 'mother-group') denote 'parents' collectively aynu-corpora Discord (2023–2026) (nukopoli, 2026) ‹corpus-suggested›. For kinship-term morphology in detail see Chapter 48 (Kinship and Honorific Possession).

41.6 Locus of plurality: utar vs verbal number

When plurality is information-structurally relevant, Hokkaido Ainu has two independent resources: nominal utar and the verb's own number morphology — suppletive stem alternations (ek / arki 'come', arpa / paye 'go') and the plural suffix -pa Nakagawa (2024: 141–159). The two are independent: a noun phrase can carry utar without triggering a plural verb, and a plural verb can appear without nominal utar. Certain mass-like nouns — supuya 'smoke', hura 'smell' — consistently govern plural verb forms (supuya at, hura at) even without utar, encoding the diffuse nature of the referent at the verb rather than the noun Kindaichi & Chiri (1936: 31).

The plural verb stem is also sensitive to quantitative context. Small, specified counts (two, three) regularly co-occur with the singular stem (tu hekaci ek 'two children came'), while large or unspecified quantities take the plural stem (inne hekattar arki 'many children came'); uncountable referents remain with the singular stem even alongside 'many' aynu-corpora Discord (2023–2026) (nukopoli, 2024; attributed to Satō) ‹corpus-suggested›. On transitive verbs the plural stem tracks plurality of the object, following the absolutive-aligned pattern: rayke (singular object) / ronnu (plural object) aynu-corpora Discord (2023–2026) (nukopoli, 2023) ‹corpus-confirmed›. These agreement thresholds and the S/O alignment of verbal number are developed in Chapter 70 (Person–Number Interactions and Plural Agreement) and Chapter 58 (Verbal Number: Suppletive Stems and the Pluractional -pa).

In verse, speakers sometimes apply utar to a semantically singular referent for metrical reasons, or omit it from a clearly plural one — evidence that the form adds meaning rather than enforcing obligatory number agreement aynu-corpora Discord (2023–2026) (nukopoli, 2024) ‹contested›.

41.7 Sakhalin contrast

The Sakhalin varieties develop the cognate of utar into a more morphologically explicit system. West Sakhalin has a suffix -(a)hcin (East -(a)hsin) attaching to the affiliative form of the noun — eci=setaha-hcin 'your dogs', an=niyehe-hcin 'our firewood' — alongside -utara as a separate suffix Murasaki (1979: 85); Nakagawa (2024: 259–261). Sakaguchi distinguishes the two Sakhalin suffixes: -(a)hcin marks a homogeneous plural (all members standing in the same possessive relation), while -utara may mark homogeneous or approximative plurality and extends to personal names Sakaguchi (2024: 12); Baek (2021: 73). Despite the richer morphology, Sakhalin number remains optional and non-obligatory: -(a)hcin can attach even to nominalized clauses, so Sakhalin too lacks a grammatical number category in the strict typological sense Nakagawa (2024: 261). An etymology linking Sakhalin utah to Hokkaido utar through historical final *r has been proposed but is disputed on comparative-phonological grounds aynu-corpora Discord (2023–2026) (antitwilight, 2024) ‹contested›.