Chapter 128Negative Predicates of Ability and Cognition

128.1 A class of lexically negative predicates

Most clauses in Hokkaido Ainu are negated analytically, by the preverbal adverb somo ‘not’ placed before the predicate (treated in Chapter 126 (Standard Clausal Negation with somo)). A small set of verbs negates differently: the negation is part of the verb's lexical meaning, so the negative clause is built on a verb distinct from its affirmative counterpart rather than on somo plus the positive verb. The members that express ability and cognition are the most clearly auxiliary-like part of this set — eaykap ‘cannot’, koyaykus ‘cannot (the situation prevents it)’, niwkes ‘cannot finish’, eramiskari ‘not know; have never’, and erampewtek ‘not understand’. Nakagawa lists them among the 他動詞起源の助動詞 ‘auxiliaries of transitive origin’ (his Table 75); Bugaeva describes the same forms as verbs that contain negation as part of their meaning Nakagawa (2024: 314–315); Bugaeva (2012: 496). The existential and privative members of the wider lexical-negative class, isam ‘not exist’ and sak ‘lack’, are treated in Chapter 127 (Negative Existential and Possessive: isam).

The negatives are well represented in the aligned Hokkaido corpus: eaykap occurs 1,082 times, eramiskari 464, erampewtek 454, niwkes 177, and koyaykus 154, against affirmative easkay 1,324 and amkir 143. These are lemma-level token totals, normalized across dialects with personal affixes stripped, so they measure the salience of each item rather than the rate of any one construction; on that measure the class belongs to the core grammar of negation and not to an incidental lexical field ‹corpus-confirmed›.

128.2 Morphology and derivation

The ability pair is morphologically transparent. Easkay is e-askay, the applicative e- on askay ‘be skilled’, and eaykap is e-aykap, the same prefix on aykap ‘be unskilful’ Tamura (1996: s.v. eaykap); Kayano (1996: s.v. easkay); Ijäs (2023). The negation therefore resides in the root: askay and aykap are themselves a lexical antonym pair ‘skilled’ versus ‘unskilful’, differentiated by s ~ y and not by a productive negating morpheme. The applicative e- supplies the slot for the thing or event one is or is not able to do, which is what makes both verbs transitive and lets them take a clause as their object.

The cognition negatives are lexicalized as wholes. Eramiskari stands as the negative of amkir, and erampewtek as the negative of eraman / eramuan; Tamura classes all four as transitive verbs and tags eramiskari and erampewtek explicitly as 否定動詞 ‘negative verbs’ Tamura (1996: s.v. eramiskari, erampewtek). The two negatives share the element eram(u)- ‘mind, understanding’ with their affirmative counterparts, but neither segments into a recognizable negator, and the accessible grammars do not analyse them further. Koyaykus and niwkes are likewise treated as unanalyzed units; koyaykus attaches as a whole after a verb, as in the dictionary forms mokor koyaykus ‘be unable to sleep’ and hunara koyaykus ‘be unable to find’ Kayano (1996: s.v. koyaykus).

The cognition negatives carry minor dialect variants. Saru has eramiskari, with eramuskari recorded for a single Nibutani speaker and eramuskare in Ishikari; erampewtek appears as erampetek elsewhere Tamura (1996: s.v. eramiskari); Nakagawa (2024: 314–315).

128.3 Argument structure and the auxiliary frame

Because they are transitive in origin, these predicates can stand as the main verb of their clause, indexing an experiencer or agent as subject and taking the event as object, or they can follow another verb as an auxiliary. In the auxiliary use the lexical verb keeps its personal marking and the negative auxiliary adds none, exactly as the formal verb ki ‘do’ does in the somo ki periphrasis: the negation sits in the auxiliary, while person stays on the predicate it scopes over Nakagawa (2024: 314, 360, 372).

The transitive origin remains active: the auxiliary may instead carry its own personal prefix, with the preceding clause as its object. In the next example the indefinite person a= attaches to eaykap, not to the verb paye ‘go’, so that eaykap predicates ‘one cannot [go further upstream]’.

The auxiliary may be preceded by the particle ka ‘even, at all’, in the frame V ka eaykap (also V ka eramiskari), with no clear difference in meaning from the plain frame. Nakagawa notes that this ka slot is available to the transitive-origin auxiliaries but is lacking for the native auxiliaries, which makes it a diagnostic of the class Nakagawa (2024: 314). The strengthening that Ijäs assigns to the slot pedagogically is the ordinary scalar contribution of ka, not a separate negative morphology Ijäs (2023).

The shared post-verbal slot identifies these predicates with the auxiliary zone rather than with immediately preverbal somo. Satō treats eaykap as an auxiliary occupying the same position as somo ki; the difference is that somo ki negates through the formal verb ki, whereas eaykap and its peers are themselves the negative predicate, so somo is not added to them Satō (2008: 124); Bugaeva (2012: 496).

128.4 Ability: eaykap, koyaykus, niwkes

Eaykap is the broad negative of easkay. Nakagawa defines its range as inability in a wide sense, covering both a lack of capacity and a situation that does not permit the event, and Tamura adds the further reading ‘be poor at’ Nakagawa (2024: 315); Tamura (1996: s.v. eaykap). The modal semantics of the easkay / eaykap pair as a category of ability and possibility are developed in Chapter 117 (Abilitative easkay / eaykap (Ability and Possibility)).

Koyaykus narrows the same space to circumstantial inability: the subject cannot do the event because the situation or the moment is against it, rather than for want of skill. Tamura draws the contrast directly — koyaykus for what outside factors prevent, eaykap for what is beyond one's ability or generally impossible Tamura (1996: s.v. koyaykus); Nakagawa (2024: 314).

Niwkes is more specific again, ‘be unable to carry an action through to completion’ Nakagawa (2024: 314). The lexicon of ability pairs it with eaykap as a doublet, the two glossed together as ‘cannot do’ and ‘cannot manage’ Kanazawa (1898). The articulation of the field has a consequence for glossing: a negative ability clause is rarely a bare ‘not able’, since the predicate chosen also classifies where the impossibility comes from ‹consensus›.

128.5 Cognition: experience and understanding

The cognition negatives split a domain that English often keeps together. The pair amkir / eramiskari concerns knowledge by acquaintance or experience. As a main verb eramiskari is ‘not know, not recognize’; after a verb phrase it gives an experiential negative, ‘have never V-ed’ Tamura (1996: s.v. eramiskari); Nakagawa (2024: 314–315).

The pair eraman / eramuan versus erampewtek concerns understanding and knowing how. As a main verb erampewtek is ‘not understand, not know’; after a verb phrase it means that the subject does not know how to carry the event out, and so cannot Tamura (1996: s.v. erampewtek); Bugaeva (2012: 496). Tamura keeps the two cognition negatives apart by source of knowledge: failure to recognize a person or thing is eramiskari, while failure to understand or to have the knowledge is erampewtek Tamura (1996: s.v. eramiskari, erampewtek). Ijäs's pedagogical contrast between experiential knowledge and understanding follows this same source split Ijäs (2023) ‹consensus›.

128.6 Interaction with somo, and the question of blocking

Two facts show that the lexical negatives do not stand in complementary distribution with standard negation. First, although somo is not added to the lexical negative itself, it freely negates a different predicate in the same clause:

Second, the lexical negative may sit inside a clause whose outer frame is itself negative. In Nakagawa's example the experiential negative a=eramiskari falls within the proposition, while the rhetorical ru-we somo ne is the outer conviction frame of the evidential, not a second analytic negation of eramiskari Nakagawa (2024: 393). The layering of negation with the formal-noun evidentials is treated in Chapter 100 (TAM, Evidentiality, and Negation on Nonverbal Predicates).

Whether the lexical negative outright blocks standard negation of the corresponding positive verb is a genuine question. Ijäs states the rule categorically: where a lexical negative exists, somo plus the positive verb is not normally possible, so that one says k=erampewtek and not *somo k=eraman Ijäs (2023). For at least one pair this is borne out by a primary source: Tamura's entry for amkir states that negation with somo is not formed, and the negative verb eramiskari is used instead Tamura (1996: s.v. amkir). The generalization does not hold across the whole class, however. Dal Corso records southern Hokkaido somo kor ‘not have’ beside privative sak, with a distinct pragmatic value, and Bugaeva describes these verbs only as containing negation in their meaning, not as barring somo Dal Corso (2025: 41); Bugaeva (2012: 496). The blocking is best stated as graded: near-obligatory for the cognition pair amkir / eramiskari, a strong preference for the ability verbs, but not a categorical rule, so that Ijäs's blanket statement overstates it ‹contested›.

The lexical negatives license the same polarity-sensitive environment as somo and isam. Nakagawa groups the ka-indefinites such as nep ka ‘anything, and under negation nothing’ with eaykap alongside the standard and existential negators Nakagawa (2024: 323); the licensing and scope of these items are treated in Chapter 129 (Negative-Polarity Indefinites and the Scope of Negation). In clause linkage they pattern with negative predicates: Takahashi shows that a negated subordinate clause selects the stative linker no rather than successive wa, and the lexical-negative auxiliaries take no in the same way, even without somo Takahashi (2016: 75–77); Nakagawa (2024: 340) ‹consensus›.

128.7 Boundaries and dialect evidence

The class is narrow, and an ordinary antonym is not a member of it. Ijäs lists lexical opposites such as eese ‘agree’ / kopan ‘refuse’ and pa ‘find’ / turaynu ‘lose’, which remain ordinary verbs and take somo in the normal way Ijäs (2023). The source check treats this as plausible but not separately verified for every pair, so the grammatical generalization is confined to the well-described negatives in the table above ‹corpus-suggested›.

The secure baseline for the class is the Saru description of Nakagawa and Tamura, with Ishikari and Tokachi variants where the examples mark them. Takahashi documents the Tokachi auxiliary use of the experiential negative:

Refsing's Shizunai description confirms that eaykap and somo ki both occur in Shizunai, but the available text for the relevant pages is too damaged to support Shizunai-specific claims Refsing (1986: 214–215). Dal Corso's history of Sakhalin negation bears on the class only diachronically: his proposal that negative-existential material underlies wider Ainu negation helps explain why the lexical negatives and standard somo form one connected domain, but the ability and cognition predicates described here are analysed from Hokkaido sources, and that history remains a hypothesis rather than a settled Hokkaido-internal fact Dal Corso (2025: 41–42) ‹speculative›.