Chapter 37The 'No Adjective Class' Thesis: Property Concepts as Stative Verbs
Why pirka, poro, and other property words are analysed as stative intransitive verbs rather than adjectives, and whether any residual adjectival/quality-noun subclass survives.
37.1 Overview
Words expressing property concepts — pirka ‘be good, be beautiful’, hure ‘be red’, poro ‘be big, grow’, pon ‘be small’, wen ‘be bad, be poor’, tanne ‘be long’, pewre ‘be young’, méan ‘be cold’ — are intransitive verbs in Hokkaido Ainu. They take the nominative person affixes of the intransitive-subject set, predicate without a copula, derive manner adverbs with the suffix -no, and enter the same causative and applicative morphology as action verbs. No separate adjective class is warranted Nakagawa (2024: 118); Nakagawa (2022: 474); Satō (2008: 2); Shibatani (1990: 19); Refsing (1986: 76) ‹consensus›.
| form | gloss | 4.S predication | prenominal collocate | inchoative |
|---|---|---|---|---|
| pirka | 'be good, be beautiful' | pirka=an | pirka menoko 'beautiful woman' | same form (contextual) |
| hure | 'be red' | hure=an | hure cep 'red fish' | same form 'become red' |
| poro | 'be big, be large' | poro=an | poro cise 'big house' | same form 'grow, become big' |
| pon | 'be small' | pon=an | pon hekaci 'small child' | same form 'become small' |
| wen | 'be bad, be poor' | wen=an | wen kamuy 'bad deity' | same form |
| tanne | 'be long' | tanne=an | tanne ni 'long stick' | same form 'lengthen' |
| pewre | 'be young' | pewre=an | pewre menoko 'young woman' | same form |
| méan | 'be cold' | méan=an | — | same form 'become cold' |
The clause-level consequences — property verbs predicating without a copula, the person-index frame, the interaction with negation — are treated in Chapter 99 (Property-Concept (Adjectival) Predication without a Copula). The word-class system as a whole, including the diagnostic criteria that separate verbs from nouns, adverbs, and adnouns, is established in Chapter 33 (The Word-Class Inventory and the Noun/Verb Bipartition). The -no adverbializer is documented in Chapter 61 (The Adverbializer -no and Adverb Formation from Verbs); causative counterparts of property verbs (e.g. pirka-re ‘make good, cure’) are treated in Chapter 75 (The Transitivizer -ka and Lexical Valency Pairs).
37.2 The standard analysis: three lines of evidence
Three independently sufficient lines of evidence converge on the verbal classification of property words Nakagawa (2022: 474–476); Nakagawa (2024: 118).
37.2.1 Syntactic identity: the person-marking frame
Property verbs occupy the predicate slot and receive the intransitive-subject (S) person prefix — the same prefix set as action intransitives, and distinct from the transitive-subject (A) set that nouns cannot host in this position at all. pewre ‘be young’ and mina ‘laugh’ both take ku= as their first-singular subject:
‘That is truly a yukar I heard back when I was young.’
Nakagawa 2024: 118; Saru
Nakagawa ex. 1. The property verb pewre 'be young' takes the first-singular intransitive prefix ku= in the same slot as the action verb nu 'hear'; no morphological difference marks one as stative and the other as active.
Third-person predication is zero-marked — the same zero as on all third-person intransitive predicates, with no copula:
‘That mountain is high.’
Shibatani 1990: 19; Hokkaido (dialect not further specified)
Shibatani ex. 13b. The property verb ri 'be tall/high' predicates the third-person subject nupuri 'mountain' with zero person marking and no copula.
‘The wind is strong.’
Shibatani 1990: 19; Hokkaido (dialect not further specified)
Shibatani ex. 13c.
The same pattern holds prenominally: a property verb precedes the noun without any morphological change, identical to an action verb in attributive position. pon hekaci 'small child' (property verb pon ‘be small’ + noun hekaci ‘child’) deploys the same verb-before-noun construction as apkas utar arki ruwe ne 'the walking people came' (action verb apkas ‘walk’ prenominally) Nakagawa (2022: 475). In neither case does the verb carry an attributive morpheme or particle.
37.2.2 Morphological identity: causative and applicative morphology
The causative suffix -re (and its allomorphs -e, -ka) attaches to property verb stems when the semantics licenses a causative reading, producing transitive counterparts exactly as it does from action intransitives Nakagawa (2022: 476). pirka-re ‘make good, make beautiful, cure’ is derived from pirka ‘be good’ and wen-re ‘make bad’ from wen ‘be bad’. Even roots that are predominantly stative in corpus use feed the causative: kiyanne ‘be older, be senior’ yields kiyanne-re ‘treat as a senior, defer to as an elder’, showing that a latent inchoative channel is available even where spontaneous onset is infrequent Nakagawa (2024: 119). The full transitivizer system is treated in Chapter 75 (The Transitivizer -ka and Lexical Valency Pairs).
37.2.3 Semantic identity: the stative–inchoative duality
Ainu intransitive verbs encode both the state and the transition into the state in a single form Nakagawa (2024: 118–119); Nakagawa (2022: 477); Shibatani (1990: 19). poro ‘be big; become big, grow’ reads as stative in one context and as inchoative in another; the same holds for hure ‘be red; become red’ and méan ‘be cold; become cold’. This is not a property peculiar to property words: the copula ne ‘be; become’, the existential an ‘exist; come into being’, and siknu ‘be alive; revive’ show the same duality Nakagawa (2024: 118–119). The inchoative imperative of a property verb therefore takes the bare stem directly:
‘Get well quickly!’
Shibatani 1990: 19; Hokkaido (dialect not further specified)
Shibatani (1990: 19). The bare property verb pirka 'be good / be healthy' receives an inchoative reading under the imperative frame. No dedicated inchoative form exists; the reading 'become good/healthy' is forced by context.
The periphrastic inchoative with the copula is blocked. *hure ne for 'become red' and *poro ne for 'become big' are ungrammatical: the inchoative reading is already carried by the bare property verb, and the copula ne in that position constructs a nominal predicate ('be/become a —'), a structurally distinct construction treated in Chapter 96 (The Copula ne and Predicate-Nominal Clauses) Nakagawa (2024: 119); Nakagawa (2022: 477).
37.3 Diagnostic evidence
The three arguments above translate into a battery of formal diagnostics that separate property verbs from the other word classes in Nakagawa's inventory Nakagawa (2024: 115–116); Nakagawa (2022: 473):
| diagnostic | property verb (pirka) | noun (menoko 'woman') | adnoun (tan 'this') |
|---|---|---|---|
| Takes intransitive-S person prefix | ✓ ku=pirka | ✗ (predication requires ne) | ✗ |
| Predicates without copula | ✓ pirka=an | ✗ requires kamuy ne | ✗ |
| Negated by somo…ki | ✓ pirka somo ki | ✗ (nominal negation somo ne) | N/A |
| Derives adverb with -no | ✓ pirka-no 'well' | ✗ | ✗ |
| Prenominal modification without morphological change | ✓ pirka menoko | ✗ (requires relative clause) | ✓ tan menoko |
| Takes causative -re/-e/-ka | ✓ pirka-re | ✗ | ✗ |
| Nominalised by -p/-pe | ✓ pirka p 'a good one' | ✓ (head noun directly) | ✓ tan-pe 'this thing' |
The first four diagnostics align property verbs with verbs and against both nouns and adnouns. Diagnostic five — prenominal modification without morphological change — is shared between property verbs and adnouns, which accounts for the apparent adjective-like quality of items like pirka and pon ‘be small’ in noun-modifying position. Diagnostic seven shows that property verbs also enter the -p/-pe nominalization slot exactly as action verbs do.
Ijäs (2023) presents property verbs as intransitive stative verbs throughout the pedagogical guide, with no separate adjective paradigm and with stative versus inchoative readings resolved by context. Satō (2008: 2) groups property words in the intransitive verb class without an adjective subclass. Refsing (1986: 76) reaches the same conclusion for the Shizunai dialect.
37.4 The candidate counter-class: adnouns, poro and pon, and the historical analysis
Three items have been taken to suggest a residual adjective class: (a) the adnoun (連体詞) class proper, (b) the specific items poro ‘be big, grow’ and pon ‘be small’, and (c) Kindaichi's pre-1942 grammar.
37.4.1 The adnoun class
Nakagawa establishes a genuine adnoun class (連体詞) defined by morphosyntactic criteria: members modify a noun, cannot serve as a predicate, and take no person affix Nakagawa (2024: 116, 283); Nakagawa (2022: 473). Its inventory comprises demonstratives (tan 'this', toan 'that', ne 'the aforementioned'), numeric prenominals (sine 'one-', tu 'two-'), interrogatives (hemanta 'which kind'), indefinites, and a residual set of general adnouns — oar 'one-of-a-pair', kes 'every', pinne 'male', matne 'female', oya 'other' Nakagawa (2024: 283). None of these carries a property meaning in the sense relevant to adjective typology; class membership rests entirely on the morphosyntactic diagnostics above. The demonstrative prenominals and their discourse-anaphoric subsets are treated in Chapter 50 (Demonstratives, Anaphora, and Definiteness).
37.4.2 poro and pon: verbs with prominent prenominal distribution
poro ‘be big, grow’ and pon ‘be small’ appear as predicates — a=kor cise poro=an 'my house is big' (constructed) — and as prenominal modifiers — poro nupuri 'a big mountain'. In prenominal position they are surface-identical to adnouns, because neither position displays person-affix morphology. The predicate data, however, show that they take S-affixes like other property verbs; Nakagawa classifies them as verbs Nakagawa (2024: 118); Nakagawa (2022: 474).
Place names attest their prenominal frequency particularly clearly: Poro-pet 'big river' and Pon-bets 'small river' are compounds of the type intransitive-verb-root + noun → nominal Nakagawa (2024: 267–268), the same formation available to any intransitive verb root. The broader verb-centredness of Ainu nominal derivation — in which nearly every noun has a simultaneous intransitive-verb reading and noun-formation operates principally through verbal roots — supports classifying poro and pon as verbs whose prenominal distribution happens to be prominent (antitwilight, aynu-corpora Discord, 2026-06-02) aynu-corpora Discord (2023–2026) ‹contested›.
37.4.3 Kindaichi's adjective class and its rejection
Kindaichi & Chiri (1936) §189–190 posits a 形容詞 (adjective) class with predicative person-inflection (形容詞の人稱變化), attributive use (連体形), and a participial form (分詞形容詞). The predicative paradigm uses the same intransitive-subject prefixes as verbs — the observation that ultimately enabled the merger. Chiri (1942) argued that since adjective-meaning words and action verbs share their full inflectional frame, their person-marking morphology, and their clause-final behaviour, positing two classes serves no formal purpose. Chiri coined 用詞 (yōshi) for the merged class, and the no-adjective analysis has been standard since then ‹consensus›. The term 用詞 is no longer in use; the merged class is simply 動詞 (verb) in current work Nakagawa (2022: 474); Nakagawa (2024: 118).
A residual nuance concerns the clause-linking particles wa and no. wa links clauses whose predicate describes a state (pirka wa 'since it is good'), while no links clauses describing a dynamic event (somo ipe no 'without eating'). Because property verbs cluster at the static pole, the contrast was sometimes read as a syntactic trace of an adjective/verb boundary. Nakagawa (2022: 479) demonstrates that the split tracks the static/dynamic semantic axis of the whole VP — it does not mark a part-of-speech boundary within the verbal system ‹consensus›. The linking contrast is examined in Chapter 99 (Property-Concept (Adjectival) Predication without a Copula).
37.5 Comparison and degree
Degree and comparison are expressed through verbal and adverbial means, not through dedicated adjectival morphology. The postpositional adverb akkari ‘more than (X), exceeding’ takes an object-marked antecedent and frames the comparison with a bare property verb predicate unchanged in form: en=pirka akkari pirka=an 'you are more beautiful than I am' (constructed). Degree adverbs such as earkinne ‘very, extremely’ and poronno ‘greatly, very much’ modify a property verb directly, and the equative uses neno ‘similar to, like’. In every pattern, the property verb itself remains morphologically unchanged — which follows from its verbal status, since Ainu verbs in general carry no degree inflections. These strategies are documented with corpus examples in Chapter 146 (Adverbs, Degree Words, and Comparison Strategies). Kanazawa (1898); Biratori Ainu oral literature (1969)
37.6 Typological framing
In the typological survey of intransitive predication, languages range from those encoding property concepts in a dedicated open-class adjective (English, Russian) to those using a verbal strategy, where property words are simply intransitive stative verbs, to those using a nominal strategy (Stassen 1997); Nakagawa (2022: 474). Hokkaido Ainu falls squarely in the verbal-encoding type by every formal diagnostic: property words pattern with verbs in their morphology and predication, and no morphological or syntactic slot is reserved for adjectives alone ‹consensus›.
This places Ainu with Japanese and Korean, whose 形容詞 class is itself a closed verbal category. Nakagawa draws the parallel explicitly: Ainu "adjectives" are verb-type — modifying a noun directly without an attributive morpheme — in contrast to the noun-type strategy where a linking particle or agreement marker intervenes Nakagawa (2022: 475). Shibatani (1990: 19) states the outcome concisely: "adjective-meaning words become predicates without any copula, in exactly the same way as verbs."
A community hypothesis in the aynu-corpora discussion archive proposes that the existential an ‘exist, be present’ in forms like pirka an 'be good; become good' might represent a distinct aspect marker for onset of state — parallel to ne an 'become X' alongside copular ne 'be X' (nukopoli, aynu-corpora Discord, 2024-11-15) aynu-corpora Discord (2023–2026) ‹speculative›. The bare corpus data show pirka an collocating with inchoative-oriented translations in certain narrative frames, but no primary source in the descriptive literature distinguishes a grammaticalized onset-of-state use of an here from its ordinary existential value. The hypothesis is noted as a lead for further corpus investigation.
References cited in this chapter
aynu-corpora Discord (2023–2026) ·Biratori Ainu oral literature (1969) ·Chiri (1942) ·Ijäs (2023) ·Kanazawa (1898) ·Kindaichi & Chiri (1936) ·Nakagawa (2022) ·Nakagawa (2024) ·Refsing (1986) ·Satō (2008) ·Shibatani (1990) ·Stassen (1997)