Chapter 143The Sentence-Final Particle System and Illocutionary Force
The inventory of clause-final particles (na, wa, ya, nankor, …) encoding assertion, emphasis, confirmation, and softening, and their grammaticalization from formal nouns.
143.1 The 終助詞 class: definition and position
Hokkaido Ainu has a closed class of sentence-final particles (終助詞, hereafter SFPs) that occupy the outermost position in the finite clause. The distributional test for the class, first systematised by Fukuda (1961: §20), is that an SFP follows a predicate and closes the utterance: nothing may follow within the same clause Nakagawa (2024: 331); Satō (2008: 71). This positional exclusivity distinguishes SFPs from the adverbial particles (副助詞) treated in Chapter 140 (Topic and Focus Marking: anak(ne), patek, and ka) and Chapter 141 (Cleft, Pseudocleft, and Nominalization-Based Focus), which are category-preserving and do not close a clause; a 副助詞 attached to a verb phrase must be re-verbalised by the light verb ki before a clause-linker may attach, while a SFP makes that attachment impossible altogether Nakagawa (2024: 317, 320); Fukuda (1961: §2).
Two frequency generalizations apply across the class. SFPs are less frequent in Hokkaido Ainu than in Japanese: a bare finite predicate routinely ends a sentence, and SFP absence signals nothing marked Satō (2008: 72); Nakagawa (2024: 333); Fukuda (1961: §20). Stacking is also restricted: only hani ‘admonitive final particle’ freely follows another SFP (na hani, yan hani); no other combination is freely productive Satō (2008: 72); Fukuda (1961: §28); Nakagawa (2024: 333).
The class encodes the speaker's illocutionary intent — command, invitation, question, assertion, admonition — and a stratum of discourse tuning that softens, reinforces, or appeals for a response from the addressee Nakagawa (2024: 331); Satō (2008: 71).
143.2 The illocutionary inventory
The Saru–Chitose baseline inventory is given below; dialect variants are noted where documented Nakagawa (2024: 331–337); Satō (2008: 71–75); Fukuda (1961: §20–33).
| SFP | illocutionary value | host / notes | dialect |
|---|---|---|---|
| yan | plural/polite imperative | plural verb stem required; C-final → an | pan-HA |
| ya (imper.) | sg-addressee polite command | imperative; homophonous with Q ya | ‹ISH› only |
| ya / a | polar question | a post-consonant; ya also COMP 'whether' | pan-HA |
| ro / rok / no | hortative (seeks addressee's assent) | any person, freq. 4th-pl; see debate below | rok ‹TOK› (Shiranuka 白糠); no (Urakawa 浦河, Samani 様似) |
| so | deliberative / self-directed volition | declarative VP; not addressed to interlocutor | pan-HA |
| wa | soft assertive (談話調整) | declarative only; never Q or imperative | pan-HA; ≠ sequential connective wa |
| na | assertive-appeal (demands attention/response) | declarative or command; also connective before commands | pan-HA |
| un | soft address / affirmation | VP, noun, adverb; an un = polite dubitative | pan-HA |
| hani | admonitive ('won't you; please do') | stacks after na, yan | pan-HA; yan hani → yanani |
| oka | counterfactual wish | concord with emphatic ta; optional | pan-HA |
| nek | emphatic assertion | concord with tas(i) / esta(p); mainly 雅語 | pan-HA; ≠ copula ne |
| yo | emotive reminiscence (strong affect) | after nominalizer p/pe or hi | pan-HA |
| tapan | formal register (ございます調) | replaces ne/an/un after ruwe/hawe/siri | pan-HA |
143.2.1 Imperative and hortative: yan, ya, ro, so
The polite imperative yan ‘come up from water; polite imperative particle’ requires the plural verb stem even when the addressee is one person: arki yan 'please come' uses the suppletive plural stem arki, not singular ek Nakagawa (2024: 332); Fukuda (1961: §26). Imperative predicates take no subject person-index; *a=arki yan is ungrammatical Fukuda (1961: §26) — the full morphological constraint is set out in Chapter 114 (Imperative and Prohibitive (Directive Mood)). When the stem ends in a consonant, yan reduces to an. Ijäs's pedagogical grammar title a-eyaypakasnu ro 'let us learn [it]' Ijäs (2023) illustrates the hortative in natural use (see below).
Ishikari has a distinct singular-addressee command ya — formally identical to the Q-particle but functionally opposed to it Nakagawa (2024: 332); Fukuda (1961: §26). The two yas are disambiguated by the predicate type that precedes them (imperative vs. declarative) and by dialect; the Q-particle treatment is in the section on boundary cases below.
‘Do come again, won't you.’
Satō 2008: 74; Chitose
The plural suppletive stem arki is required by yan even for a single addressee. hani stacks after yan to add the gentle admonitive 'won't you / I hope you will'. The cluster yan hani contracts to yanani in the Chitose speech-stream (Nakagawa 2024: 333).
The hortative ro ‘hortative particle’ (‹TOK› Shiranuka rok; Urakawa/Samani no) seeks the addressee's assent for a proposed joint action Nakagawa (2024: 333); Satō (2008: 73); Fukuda (1961: §23). An earlier analysis held that ro is a "first-plural-inclusive command particle", a claim advanced in Kindaichi & Chiri (1936: §215) and repeated in Chiri (1942: §118). Fukuda (1961: §23) rejects this: the particle is attested after first-singular, second-person, and third-person predicates, not only after the fourth-person (inclusive) plural; the frequent co-occurrence with the fourth person reflects the hortative meaning 'let us …', not a morphosyntactic person restriction Fukuda (1961: §23); Nakagawa (2024: 333) ‹contested›. Further discussion is in Chapter 115 (Hortative, Optative, and Cohortative).
The deliberative so ‘deliberative sentence-final particle’ converts a predicate into a self-addressed musing — consideration of an action for one's own benefit, not aimed at an interlocutor Nakagawa (2024: 334); Satō (2008: 74); Fukuda (1961: §22).
143.2.2 Assertive and softening: na, wa, un
The assertive-appeal particle na ‘assertive sentence-final particle’ rouses the hearer's attention and implies a response — verbal or behavioural — is expected Nakagawa (2024: 332); Satō (2008: 73); Fukuda (1961: §27). It attaches to declarative and command predicates alike and is among the most frequent SFPs in the Saru and Chitose corpus.
‘There are plenty of fine cushions here, you know.’
Nakagawa 2024: 332; Chitose
Nakagawa ex. 583. na follows the existential predicate and signals that the speaker expects the hearer to act on the information (sit down, use the cushions). The existential an is a 3rd-person (zero-marked) predication.
The softening particle wa ‘soft assertive sentence-final particle’ adds a mild conversational tone to a declarative, serving the function Nakagawa calls 談話調整 (discourse tuning) Nakagawa (2024: 334); Satō (2008: 74–75); Fukuda (1961: §21). It is restricted to declarative predicates: neither questions nor imperatives take it. In answers and in-dialogue assertions, wa is particularly frequent, pulling blunt propositions toward a gentler register:
‘She is no pauper of an old woman.’
Nakagawa 2024: 334; Chitose
Nakagawa ex. 582, first clause of a pair (the second is pirka huci ne wa 'she is a fine old woman'). wa softens the emphatic negation ka somo ne; the exchange is in dialogue. Fukuda 1961: §21 notes that the particle carries a feel comparable to the Tokyo-female / Hokkaido-unisex Japanese「わ」, without claiming it is female-restricted in Ainu — see §register-gender.
The soft-address particle un ‘soft address / affirmation particle’ conveys gentle appeal or mild affirmation and is unusual in the inventory for attaching to nouns and adverbs as well as to VPs Nakagawa (2024: 332–333); Satō (2008: 74); Fukuda (1961: §30). The sequence an un (existential + un) serves as a polite dubitative. In the affirmative-answer frame of the evidential tail (§evidential-frame below), un follows the formal noun directly: ku=nu ruwe un 'yes, I heard it' Nakagawa (2024: 396).
143.2.3 Admonitive and stacking: hani
The admonitive hani ‘admonitive final particle’ adds gentle urging to commands and admonitions — a 'won't you; I hope you will' coloring Nakagawa (2024: 333); Satō (2008: 72); Fukuda (1961: §28). Its structural property is free stacking after na or yan, yielding bipartite clusters na hani and yan hani; the latter contracts to yanani in Chitose. hani is not restricted to imperatives, as the following shows:
‘I am going to report this to my father, you know.’
Satō 2008: 72; Chitose
Satō ex. = Nakagawa 2024: 333. na hani on a purposive declarative; neither clause is an imperative, demonstrating that hani attaches regardless of illocutionary mode. or-o is the possessed form (or-POSS) of the relational noun or 'place, location of'.
Members of the Discord-based aynu-corpora community have proposed that hani derives from the motion verb hanke 'approach, come near' by grammaticalization into a command particle, parallel to the proposed path of yan from a lexical verb meaning 'come up from the water' (antitwilight, aynu-corpora Discord 2024-11-18; aynu-corpora Discord (2023–2026)) ‹speculative›. Both etymologies are plausible given the cross-linguistically common motion-verb-to-imperative cline, but neither has been established in the primary literature; the proposals are leads for diachronic work.
143.2.4 Kakari-musubi concord: oka, nek
Three pairs in the Ainu particle inventory exhibit a concord relationship structurally parallel to Japanese 係り結び: an emphatic adverbial particle (副助詞) inside the clause licenses or requires a specific SFP at sentence end. The 係助詞 label for the emphatic members of these pairs was introduced by Chiri (1942: §118) (via Fukuda (1961: §11)) and is endorsed by Nakagawa (2024: 330); Satō (2008: 73).
| emphatic 副助詞 | concordant SFP | illocutionary reading | sources |
|---|---|---|---|
| ta 'こそ (wish)' | oka (optional in Saru) | counterfactual wish '〜だったらなあ' | Nk 330; Fk §12/§25; St 73 |
| tas(i) 'まさに〜こそ' | nek (optional in Saru) | assured assertion '〜のだよ (indeed it is)' | Nk 330; Fk §14/§24; St 73 |
| esta‹SHI› / estap‹TOK› | ne | assertion 'こそ〜だ' | Nk 330 |
In Saru the concordant SFP is optional — the closing particle "is not necessarily present" Fukuda (1961: §11) — which raises the question of whether the concord is a productive grammatical constraint or a set of conventionalized collocations. The sources consulted do not settle this ‹contested›.
‘If only I were a bird!’
Nakagawa 2024: 330; Saru
Nakagawa ex. 560 (= Tamura 1988b: 33; the Fukuda 1961: §12.2 citation gives the form with oka explicit). The emphatic ta triggers the counterfactual wish reading; the concordant oka is here present but may be omitted in Saru. The couplet continues rera ta ku=ne ('if only I were the wind!').
‘About that alone am I truly troubled.’
Nakagawa 2024: 334; Saru
Nakagawa ex. 594 (source: Chiba University Ainu materials 2015: 160). The emphatic tasi triggers the concordant nek; patek 'only' restricts scope. The 4th-person A prefix a= tracks the narrator/speaker in the oral-literature register.
Satō (2008: 65–66) identifies a structurally more opaque intensification construction in oral-literary register: a bare concept-form noun (hum 'sound', ru 'trace') followed by the particle konna/kan and a verb ending in the concord suffix -natara, -atki, or -rototke:
‘The sound of his coming thundered on.’
Satō 2008: 66; Hokkaido (dialect not further specified)
Satō ex. (oral-literature register). The bare concept form hum (not possessed humi) appears before konna. Satō classifies konna/kan provisionally as a 強調の助詞 (emphatic 副助詞) but flags its grammatical status as 'unclear'; the appearance of the concept form rather than the possessed form before konna is unexplained — the same nouns take possessed forms when functioning as formal-noun evidential heads. See formal-defective-nouns for the possessed/concept-form alternation.
The formal noun / possessed-form puzzle that the konna construction poses is discussed in Chapter 35 (The Formal / Defective Noun Set (ruwe, hawe, siri, humi, hi, pe, kur)); the construction is also adjacent to the kakari-musubi debate in Satō (2008: 65–66) ‹contested›. Whether any of these three concord pairs constitute a productive focus-concord (like Japanese 係り結び) or merely a set of frozen collocations remains an open descriptive question Nakagawa (2024: 330); Fukuda (1961: §11) ‹contested›.
143.2.5 Register-marked forms: yo and tapan
The emotive particle yo ‘emotive sentence-final particle’ attaches after a clausal nominalizer (p/pe or hi) and signals strong affective involvement — reminiscence, wonder, or pathos Satō (2008: 75). Its position after the nominalizer, rather than directly after the verb, makes it one of the few SFPs whose attachment site is lexically conditioned. See Chapter 101 (The Nominalizers -p/-pe and -i/-hi (Participant, Event, Place, Fact)) for the nominalizer inventory.
The formal-register marker tapan replaces ne, an, or un after the evidential formal nouns ruwe, hawe, and siri, yielding the ございます-register frame ruwe tapan Fukuda (1961: §33). It is the one SFP that enters the evidential tail proper (the predicator slot of the formal noun) rather than stacking outside it.
143.3 Ordering and stacking at the clause edge
The general clause-final ordering runs: verb stem → TAM auxiliaries → evidential tail (ruwe/hawe/siri/humi + ne/an) → SFP. The ordering is confirmed by the grammaticality of … ruwe ne na and the ungrammaticality of *… na ruwe ne Nakagawa (2024: 393); Fukuda (1961: §20). The two exceptions within this slot are un and tapan, which intrude into the evidential predicator position itself (ruwe un, ruwe tapan) rather than stacking outside it — see Chapter 120 (The Nominalization-plus-Copula Evidential Schema) and Chapter 121 (ruwe ne — the Inferential / Visual-Trace Evidential).
Within the outer SFP slot, only hani freely follows another SFP Fukuda (1961: §28). No other combination of two SFPs is freely attested in the sources consulted; longer stacks have not been systematically studied in the Hokkaido corpus ‹corpus-suggested›. tapan does not add to the outer SFP slot: it occupies the predicator position of the formal noun and is not a stacking particle in the same sense.
143.4 Boundaries: SFP, connective, and the evidential tail
143.4.1 na as clause-boundary connective
When na immediately precedes a command clause, it functions as a clause-boundary signal rather than a sentence-closing SFP: … na [command] reads 'since … / because … [do X]', where na provides the motivating ground for the instruction Nakagawa (2024: 332); Fukuda (1961: §27). The following corpus instance shows na closing a declarative clause inside a direct quote, where the SFP value is foregrounded by the sekor hawean framing:
‘I plan to go [there] travelling the whole night.’
ILCAA Ainu materials 1976; Saru, Kawakami Matsuko, 'isoytak dokowa 2', ILCAA Ainu materials (Saru)
na closes the quoted declarative clause inside sekor hawean '(she) said'. The sequence kusu ne na (purpose + copula + na) is common in discourse: the assertive-appeal force of na urges the hearer to acknowledge the speaker's stated intention.
The ambiguity between the SFP and connective readings is resolved by whether a further clause follows: sentence-final na ends the utterance, while the connective na always precedes something else. The connective function is treated fully in Chapter 131 (Clause Linkage in Ainu: Typology and Inventory) and Chapter 147 (Conjunctions and Discourse Connectives).
143.4.2 ya/a as Q-particle vs. complementizer
The Q-particle ya ‘question particle; complementizer 'whether'’ / a ‘post-consonant question particle’ closes a polar question as a SFP Nakagawa (2024: 332–333); Fukuda (1961: §29). The same ya (but not a) also nominalizes a clause as an embedded polar question ('whether …'), functioning as a complementizer; in that use it is not a SFP because the clause it heads is embedded Nakagawa (2024: 333); Satō (2008: 74). The category of ya/a is therefore function-dependent: SFP in main clauses, COMP in embedded indirect questions ‹contested› Fukuda (1961: §29/§31); Nakagawa (2024: 333).
The Q-副助詞 he ‘polar question focus particle’ and the Q-SFPs ya/a occupy distinct positions: he attaches to the constituent in doubt and can appear mid-clause, while ya/a are clause-final Nakagawa (2024: 323–324); Fukuda (1961: §8/§31). The full interrogative strategy — the polar vs. content split, wh-in-situ, and the ne/an/∅ alternation on the evidential tail — is in Chapter 144 (Interrogative Strategies, Question Particles, and Evidential Questions).
143.4.3 wa as SFP vs. sequential connective
The SFP wa (soft assertive, §assertive-softening) is formally identical to the sequential clause-linker wa (CONJ 'and then, having done X') treated in Chapter 132 (Sequential wa and Clause Chaining). They are distinguished by position: the linker always precedes another predicate, the SFP ends the utterance. In many Hokkaido dialects the wa morpheme surfaces phonologically as ma (nukopoli, aynu-corpora Discord 2024-07-09, asserted; aynu-corpora Discord (2023–2026)). Whether this phonological shift applies equally to both the connective and the SFP, or whether only one of the two undergoes it in a given dialect, is not documented in the sources consulted ‹speculative›.
The community discussion has also proposed a morphological analysis of newa 'and' as ne + wa (copula + sequential linker) and neya 'or' as ne + ya (copula + Q-SFP) (aomidori_cs_ and nukopoli, aynu-corpora Discord 2023-12-12; aynu-corpora Discord (2023–2026)) ‹contested›. If correct, the SFP ya would participate directly in the formation of a common NP-coordinator — but the analysis remains community-proposed and the morphological reanalysis of these particles has not been verified in the primary literature.
143.4.4 The evidential tail as sentence-type switch
The evidential formal nouns (ruwe, hawe, siri, humi) that anchor the nominalization-plus-copula evidential construction (see Chapter 120 (The Nominalization-plus-Copula Evidential Schema) and Chapter 121 (ruwe ne — the Inferential / Visual-Trace Evidential)) also host a sentence-type switch: the predicator following the formal noun — copula ne, existential an, or zero — signals declarative, content-question, or polar-question force respectively Satō (2008: 136–137); Nakagawa (2024: 394–395).
| sentence type | evidential-tail predicator | example |
|---|---|---|
| declarative | ruwe ne (COP) | sirpirka ruwe ne 'the weather is fine (recognized fact)' |
| content (wh) question | ruwe an (existential) | hemanta ne kusu ene siran hi an 'why is it like this?' |
| polar (yes/no) question | bare ruwe (verb omitted) | e=eraman ruwe 'did you understand?' |
| affirmative answer | ruwe un (SFP un) | ku=nu ruwe un 'yes, I heard it' (Nakagawa 2024: 396) |
| formal assertion | ruwe tapan (SFP tapan replaces COP) | ceremonial/ございます-register; Fukuda 1961: §33 |
Satō (2008: 137) proposes that in a declarative the copula predicates the subject's nature; in a content question the existential asserts that the subject exists while its nature is questioned; in a polar question even existence is at issue, so the verb is omitted. The un and tapan entries in the table are the two SFPs that enter this predicator slot. The outer SFPs (na, wa, hani, …) stack outside the evidential frame: ruwe ne na is grammatical; *na ruwe ne is not Nakagawa (2024: 393).
‘Is it all right now?’
ILCAA Ainu materials 1976; Shizunai, Oda Suteno, 'reyep kamuy', ILCAA Ainu materials (Shizunai)
Nakagawa 2024: 304, ex. (corpus Shizunai). The bare formal noun ru-we (no copula) combines with the Q-副助詞 he and the existential an, realising the reduced polar-Q frame ruwe he an (Nakagawa 2024: 409e). This illustrates the sentence-type switch at the evidential tail: copula ne (declarative) is replaced by existential an and the Q-particle he.
143.4.5 sekor is not a sentence-final particle
The quotative sekor ‘quotative marker’ is sometimes encountered in clause-final position of a quoted clause, preceding a speech verb (sekor hawean 'said, …'). It is not a SFP: it functions as a quotative adverb/particle (QUOT) that heads the quoted clause as an adverbial, not as a particle that closes a main clause Nakagawa (2024: 470); Bugaeva (2008: 35–36). The etymological analysis sekor < se 'make a sound' + kor 'while' (nukopoli, aynu-corpora Discord 2024-11-09; aynu-corpora Discord (2023–2026), asserted, corroborated by Nakagawa 2024: 470) confirms its adverbial origin: it is 'saying while', not a clause-closing illocutionary marker. The full syntactic and typological treatment of sekor is in Chapter 105 (Quotative Complementation and Reported Speech (sekor)) and Chapter 150 (First-Person Narration, the Logophoric Fourth Person, and Reported Discourse).
143.5 Register, genre, and the gender question
SFPs in Hokkaido Ainu are not conditioned by the speaker's biological sex. Satō (2008: 72) states this explicitly: "Ainu final particles are in principle used by both sexes; there is no particle specific to men or women" — a direct contrast with Japanese, where particles are gender-indexed. Fukuda (1961: §21) noted that the SFP wa carries a feel similar to the Tokyo-female / Hokkaido-unisex Japanese「わ」, but did not claim the Ainu particle is female-restricted. The two descriptions converge: HA SFPs are gender-neutral Nakagawa (2024: 331) ‹consensus›.
A register observation does correlate with speaker sex, but at the level of the evidential tail, not the SFP class. Community discussion in the aynu-corpora Discord identifies ruwe ne as "used for objective fact, frequent in folktales/legends, favoured by men, like のだ, and doing scene-shift and summary work" (nukopoli, aynu-corpora Discord 2024-11-09; aynu-corpora Discord (2023–2026)) ‹corpus-suggested›. This skew belongs to the evidential-nominal layer, not to the SFP layer; it does not contradict Satō's gender-neutrality claim for SFPs, which operates at a different level of the clause. The sociolinguistic conditions behind the ruwe ne frequency pattern, and register differences in evidential and particle use across genres, are treated in Chapter 155 (Narrative TAM and Evidential Patterning by Genre) and Chapter 145 (Gendered Speech and Pragmatic Registers).
Mirative meaning in Hokkaido Ainu is carried by the perfective auxiliary aan/anan, not by a dedicated SFP Nakagawa (2024: 393); Dal Corso (2018: 336). When ruwe ne appears in surprising contexts the surprise comes from the discourse frame, not from the formal noun itself. The mirativity–evidentiality interface is in Chapter 125 (Mirativity, Evidential Scope, and Grammaticalization to Sentence-Final Particles).
The Shizunai dialect materials in Refsing (1986) contain an SFP inventory that likely includes forms not in the Saru–Chitose baseline described above; the Refsing treatment (pp.253, 265–267) was not re-read for this chapter and the Shizunai inventory has not been compiled at Fukuda's level of rigour. Whether a form glossable as awa belongs to the Shizunai SFP inventory is one unsettled gap; the present description is limited to the Saru–Chitose core supplemented by documented dialect variants ‹speculative›.
References cited in this chapter
aynu-corpora Discord (2023–2026) ·Bugaeva (2008) ·Chiri (1942) ·Dal Corso (2018) ·Fukuda (1961) ·Ijäs (2023) ·ILCAA Ainu materials (1976) ·Kindaichi & Chiri (1936) ·Nakagawa (2024) ·Refsing (1986) ·Satō (2008)