Chapter 140Topic and Focus Marking: anak(ne), patek, and ka

140.2.1 Form and attachment

anak and anakne mark the 提題 ('topic-raising') function: a phrase is singled out as the current topic of talk and implicitly set in contrast to other possible topics Nakagawa (2024: 320, §9.4.2①). Nakagawa and Satō agree that the two forms carry no difference in meaning: "whether anak and anakne differ at all is an open question" Satō (2008: 64); Nakagawa (2024: 320). The particle follows its host immediately. Three attachment sites are documented: after a bare noun phrase, after a postpositional phrase (noun + relational noun), and after a verb phrase that functions as the topic of the sentence Nakagawa (2024: 320–321). Ijäs's online pedagogical grammar gives a learner-oriented account of the same functions Ijäs (2023).

The particle does not mark the grammatical role of its host. A topicalized phrase may be the agent, the patient, or an oblique of the comment predicate, and the verb's person agreement is determined entirely by its valency, not by topic status. This is consistent with the absence of morphological case on core arguments and with the head-marking profile of the language described in Chapter 62 (Architecture of the Personal-Affix System: The Four Persons and the S/A/O Paradigms): it is the verb that tracks participants through its prefixal paradigm, and a separate topic particle does not alter that tracking.

140.2.2 Topic versus contrastive topic

Satō distinguishes two uses of a single form: plain topic (主題) and contrastive topic (対照) Satō (2008: 62, §7.2). In contrastive use, two phrases each receive anak or anakne and the comment clauses are set in explicit opposition, typically with korka ‘but’:

A contrastive pair does not require both members to be overtly topic-marked. In the following example the first member (wakka ‘water’) bears anak, but the second (sake ‘sake’) carries no particle; the contrast is established by the korka ‘but’ connective and the parallel structure Satō (2008: 62–63):

140.2.3 The given-information restriction

anak requires a referent that the hearer can already be assumed to know Satō (2008: 63). Satō illustrates with a pointing context: a demonstrative like toan pe 'that thing' is situationally identifiable but is being raised as a topic for the first time, so anak is blocked even though the referent is visible. The constraint is sharper than Japanese 「は」, which permits 'あれは何だ?' in the same situation; Ainu does not allow the equivalent construction Satō (2008: 63); Nakagawa (2024: 320). A participant already mentioned or recoverable from context can freely be topic-marked; given that Hokkaido Ainu regularly omits core arguments when they are contextually recoverable (see Chapter 139 (Reference Tracking and Argument Continuity)), anak signals that the referent is being made the active topic in contrast to something else, not merely that it is recoverable.

140.2.4 Scope and weight compared with Japanese wa

Nakagawa notes that anak is used considerably less than Japanese 「は」: a word-for-word substitution of 「は」 with anak produces unnatural Ainu sentences Nakagawa (2024: 320). The particle carries something of the weight of '〜というものに関しては' ('as for the thing that is X'), marking a topic that is actively contrasted with alternatives. Once a topic has been established in discourse, it need not be re-marked on subsequent clauses within the same topic chain. This assessment is corroborated by community linguistic discussion: the particle is reserved for contexts where singling-out or contrast is genuinely intended, and omitting it when topic continuity is clear from context is the natural choice (nukopoli, aynu-corpora Discord 2024-11-09; aynu-corpora Discord (2023–2026)) ‹corpus-suggested›.

140.2.5 Person affixes and independent pronouns

A delimiting particle cannot attach to a person-agreement prefix. Those prefixes are bound indexes that form a morphological unit with the stem; they are not free words and cannot host a postposition or particle. Topicalizing a first- or second-person participant therefore requires the free independent pronoun (kani ‘I’, eani ‘you (sg.)’, etc.) as the syntactic host Nakagawa (2024: 253, ex.323). The construct *ku= anakne or *en= anakne, where the particle would attach to a bare prefix, is ungrammatical Nakagawa (2024: 253):

The same constraint applies to all persons. For the contrastive topicalization of a fourth-person or logophoric referent through the free pronoun asinuma ‘self (logophoric)’, see Chapter 67 (The Indefinite/Fourth Person: Forms, Reference, and Agent-Defocusing), and for the full independent-pronoun paradigm, Chapter 49 (Personal, Reflexive, and Emphatic Pronouns).

The corpus supplies further examples with possessed NPs as topic. In the following passage the narrator's dog is established as the topic of a multi-clause stretch by anakne on the possessed NP:

140.2.6 anak versus anakne and diachrony

Both forms function identically in the texts, but the frequency picture is contradictory across sources. Satō, working from Chitose texts, reports that anakne occurs roughly twice as often as anak Satō (2008: 64), while the whole aligned Hokkaido corpus (over 1.5 million tokens) gives the reverse — approximately 9,355 tokens of anak against 4,789 of anakne. Satō's counts cover a single dialect and a small text set; the corpus aggregates all Hokkaido varieties without disambiguating functions. Whether there is any conditioned distribution between the two forms remains open ‹contested›.

Etymologically, anak derives from an-yak — the existential verb an plus the conditional/contrastive particle yak — as recorded in the Saru dictionary Tamura (1996: s.v. anak) (corroborated by nukopoli, aynu-corpora Discord 2023-12-08; aynu-corpora Discord (2023–2026)) ‹speculative›. anakne adds ne (the copula or a reinforcing particle) to the same base. The corresponding Sakhalin forms have been proposed as neyke or neanpe (nukopoli, aynu-corpora Discord 2024-12-27; aynu-corpora Discord (2023–2026)) ‹speculative›; the correspondence remains to be established through systematic comparative work.

140.4.1 ka: additive, scalar, and NPI-forming

ka (southeastern Hokkaido variant kay) adds the host to a contextually established set and carries a scalar or contrary-to-expectation force: 'also, even, and — perhaps surprisingly' Nakagawa (2024: 321–322, §9.4.2②); Fukuda (1961: §5); Satō (2008: 64–65). Fukuda formulates its Saru semantics as 'beyond or contrary to expectation' Fukuda (1961: §5), and Satō illustrates with examples carrying emotional weight: 'even a dog won't eat it'; 'even two or three young men came' Satō (2008: 64). This distinguishes ka from the plain additive hem (§ below).

In negative clauses, ka reinforces the negation with a scalar 'not even' reading. When two or more noun phrases each carry ka, the construction lists members of a set all of which fall within the scope of the negation:

The combination of an indefinite pro-form with ka produces an indefinite or negative-polarity series: nep ka 'something / anything', nen ka 'someone / anyone', ney ta ka 'somewhere / anywhere'; under negation these shift to 'nothing / no-one / nowhere' Nakagawa (2024: 322–323); Satō (2008: 137). The interface between ka and the interrogative pro-form system — including the question-word vs indefinite-word distinction — is treated in Chapter 51 (Interrogative, Indefinite, and Negative-Polarity Pro-forms). For the evidential use of the same pro-forms in content questions, see Chapter 120 (The Nominalization-plus-Copula Evidential Schema).

140.4.2 hem, hemem, and hene: plain additive and alternative focus

hem (Saru / Chitose; Horobetsu variant hemem) is the plain additive particle: it adds one more item to a list with no scalar or contrary-to-expectation implication. In a paired construction X hem Y hem 'both X and Y', each member carries hem and the reading is symmetrical addition Nakagawa (2024: 321–322); Fukuda (1961: §4). The particle is attested only in Saru and Chitose; elsewhere plain juxtaposition or, in Tokachi and Kushiro, the coordinator usa serves the same function.

hene (Tokachi variant ene) introduces alternatives or illustrative examples: 'for example; or … or …'. A repeated hene … hene presents two or more members as mutually interchangeable options, as in nisatta hene oyasim hene 'tomorrow or the day after' Nakagawa (2024: 323–324); Fukuda (1961: §6); Satō (2008: 65).