Chapter 149The Oral-Literature Genre System and Its Grammatical Signatures

149.1 The genre system: parameters and macro-categories

No single Ainu term covers oral-literary activity as a whole; the domain is delimited descriptively. The most widely adopted modern taxonomy, developed by Nakagawa and followed by Endō (2022), organizes oral activity into four macro-categories by communicative purpose: songs (seated upopo, dance songs rimse / horippa, lyric yaysama, lullabies), word play (kikinasi bird-cry mimicry, tongue-twisters, riddles), chants directed to gods or humans (prayer kamuy nomi itak / inonno itak, incantation, dispute-speech caranke, funeral words, greeting words), and narrative (prose tales, divine epics, heroic epics, legend). Within the narrative category the formal parameters that differentiate genres are: presence or absence of melody; presence or absence of a sakehe ‘refrain (of a kamuy yukar)’ refrain; rhythmic tapping (rep ‘beat time’) with a stick (repni ‘rhythm-stick’); typical length; and narrator gender Endō (2022); Satō (2008: 259).

The table below gives the Chitose-based three-way split of the narrative genres and its primary formal diagnostics, which constitute the empirical baseline from which dialect variants and contested cases depart Satō (2008: 259).

The table identifies the sakehe as the primary formal divide between yukar and kamuy yukar: both are sung genres with melody, but yukar carries no sakehe while kamuy yukar does. A fourth narrative form, irupaye, presents yukar content in a prose frame with no melody or refrain; it was available to women narrators for whom verse yukar was conventionally restricted Satō (2008: 259). Two further genres stand outside the main three: upaskuma / ucaskuma (oral tradition, legend, historical lore), which makes no fixed verse or performance demands, and the coastal-Hidaka/Iburi oyna (culture-hero epic with a distinct protagonist class), treated separately below. The grammatical consequences of the sung/prose distinction — person-marking conventions, evidential defaults, and metrical packaging — are surveyed in Chapter 155 (Narrative TAM and Evidential Patterning by Genre).

149.2 Dialect-area variation in genre vocabulary

The Ainu genre vocabulary is not uniform across Hokkaido. Terms shift across dialect areas in ways that reflect both sound change and the independent naming of genre categories by distinct communities. The principal dialect correspondences, as documented by Endō (2022) and Chiri (1955), are:

Two naming collisions make this geography analytically difficult ‹contested›. First, the term oyna covers different genres in different areas: in Tokachi and Sakhalin it functions as a synonym of kamuy yukar (a divine epic with sakehe), while in coastal Hidaka and Iburi it designates a distinct genre whose protagonist is a culture-hero (Aynurakkur / Okikurmi / Oynakamuy) and which uses no sakehe Endō (2022); Nakagawa (1995: s.v. oyna). Second, the term tuytak means 'prose tale' across most of eastern Hokkaido but 'divine epic' in Samani — the same form names metrically and functionally opposite genres depending on locality Endō (2022). These are not irregular synonymies; they reflect genre systems that were named independently in different dialect communities, a distribution that Nakagawa treated as a problem in dialect linguistic geography Endō (2022).

The classification also shifts within a single dialect tradition. In Tamura's Saru materials, informant 平賀サダモ (Sadamo) does not use oyna as a genre label at all: the divine and culture-hero epics that Kindaichi and Kubodera classed as oyna are subsumed under yúkar in her usage Tamura (1996); Endō (2022). This shows that even the yukar/kamuy-yukar boundary, presented as basic in the Chitose table above, is not imposed uniformly on all bearers of the Saru tradition.

Kayano (1996) provides a metalinguistic definition of yukar as a practice of imitation and transmission: "誰かが聞かせてくれた長い長い物語，それを真似て次に誰か に聞かせる，それをユカㇻと言う" ('the long story that someone told you, which you imitate and tell the next person — that is called yukar'). The same entry records the dream-transmission convention: a yukar heard in a dream (wentarap or ta a=nu yukar) must be retold at once or it is lost.

149.3 The heroic epic: yukar and sakorpe

The yukar is the longest and formally most demanding of the sung genres: an epic sung to a melody that is specific to the narrator (not fixed to the story and subject to change mid-performance), stretching from a single evening to several nights Endō (2022). Its defining formal features in the Saru/Chitose baseline are the absence of a sakehe refrain and the presence of rhythmic tapping with a repni ‘rhythm-stick’ on the hearth-rim throughout the performance Satō (2008: 259). Listeners may add a chorus call (hetce). The narrator is traditionally a man; women conventionally told the prose form irupaye when they wished to narrate yukar content Satō (2008: 259), though modern documentation records women reciting verse yukar knowingly against that convention Endō (2022).

The protagonist is a parentless superhuman youth — in Saru texts canonically Poy-ya-un-pe or Pon Sinutapka-un-kur, in eastern Hokkaido Pon Otasam-un-kur — who flies, dives, travels underground, and battles without moral framing (heroes may annihilate whole villages; the worldview is entertainment, not instruction) Endō (2022). The sub-genre menoko yukar ('women's heroic epic') keeps the format but replaces battle with marriage as its thematic core; female heroes wield shamanistic power (tusu) and foresight Endō (2022). A speaker of the Saru dialect, 川上まつ子, recorded an explicit account of the menoko yukar genre: the genre is told with a sakehe refrain (sakekoye), but the narrator herself preferred to deliver it without the sakehe, as an uwepeker ILCAA Ainu materials (1976).

In eastern and central Hokkaido — Tokachi and adjacent areas — the heroic epic is named sakorpe (the eastern term for this genre; its internal morphology is debated) Endō (2022). For the lexical domain of 'hero', the noun rametok ‘bravery; brave person (rametok < ram-etok 'spirit-tip')’ can stand alone as 'brave person / hero' in epic contexts, attested with modifiers such as kamuy rametok and sonno rametok; a further term cipapa appears in Chiri Yukie (1923) but is absent from other dictionaries and texts (gengojiro, aynu-corpora Discord 2024-02-13; aynu-corpora Discord (2023–2026)) ‹speculative›.

A Saru narrative records the genre terms in disjunctive alternation, with the narrator uncertain whether the performance was yukar or oyna:

The Abuta dialect (Iburi, 虻田) yukar narrated by 遠島タンケ attests the same aspectual and comparative apparatus as Saru/Chitose verse narrative, including durative kor okay, resultative wa okay / wa an, and the comparative particle kane ‘as if; to the extent that’. The prospective anki ‘about to’ is also documented. A striking lexical item in this text is the genre-specific verb sakehawki 'to sing / recite a yukar' — a denominal verb built on the eastern heroic-epic root Satō (2009: §2.2).

A residual opening formula, the eyororpe, may repeat at the start of a yukar performance. Satō (2008: 259) reads this as a possible vestige of a refrain that yukar once possessed — an inference rather than an established fact ‹speculative›.

149.4 The divine epic: kamuy yukar, the sakehe refrain, and oyna

The kamuy yukar is shorter than the yukar (from a few minutes to around an hour), told mainly by women, and sung to a melody fixed per story — in contrast to the narrator-specific melody of the heroic epic Endō (2022). Its protagonist is a kamuy ‘god; divine being’ (animal, plant, fire, water, weather, household object) narrating its own experiences in the first person. The genre's primary formal diagnostic is the sakehe (折り返し, burden, refrain): a fixed string, typically inserted after every line or every few lines, fixed per story Endō (2022); Satō (2008: 259).

sakehe ‘refrain (of a kamuy yukar); the sake of —’ is the possessed form of sake; Tamura (1996: s.v. sákehe) marks the refrain sense with a distinct pitch accent (sákehe) to distinguish it from the homophonous 'for the sake of —'. Sakehe range from transparent onomatopoeia (the cuckoo-god's sakehe han kakkok kakkok imitates the bird's cry; the thunder-god's hum pakpak is the sound of rain or thunder, glossed by its narrator as humihiruy 'it's a loud sound' Biratori Ainu oral literature (1969)) to completely opaque compounds such as the long sakehe recorded by Kayano (1996): tunoyake renoyake kutunke kamuyke kamuy cikappo hum hum hum.

The Biratori examples are from Biratori Ainu oral literature (1969); the three marlin-tuna variants and the cuckoo sakehe are cited via Endō (2022).

The function of the sakehe is disputed ‹contested›. Nakagawa (2001) proposed that sakehe originally served as an index of the protagonist. Against this, the marlin-tuna data above show that the same story-type can carry different sakehe in different performance areas, and that the protagonist may shift between variants (Chitose changes the fish species and replaces the sakehe). The correspondence between story-type and sakehe is therefore not fixed, and many sakehe cannot be related to the protagonist through any recoverable semantic link Endō (2022).

The Ainu Shin'yōshū (Chiri Yukie (1923)) — Iburi (Horobetsu) kamuy yukar compiled and translated into Japanese by 知里幸惠 (Chiri Yukie, 1903–1922) — is the canonical first published Ainu-language oral-literature collection and remains a primary grammatical resource for Iburi kamuy yukar. Its narrator consistently uses the exclusive first-person plural prefix ci= / suffix =as as a singular hero-narrator marker, throughout every text Satō (2008: 261); Satō (2004). Table-of-contents titles in the collection such as AEKIRUSHI have been morphologically reanalysed as a=e-kir-usi 'the place one has memory of' (0penke, aynu-corpora Discord 2024-03-19, after 片山龍峯 2003; aynu-corpora Discord (2023–2026); ‹contested›) rather than the earlier parsing from ekiru 'turn back'.

The coastal-Hidaka/Iburi oyna is a distinct genre with a culture-hero protagonist (Aynurakkur, Okikurmi, Oynakamuy) and no sakehe, making it formally closer to the yukar than to the kamuy yukar Endō (2022); Nakagawa (1995: s.v. oyna). It should not be conflated with the eastern use of oyna as a synonym of kamuy yukar. The grammatical analysis of the logophoric narrator's voice (the protagonist's first-person perspective) and the evidence for classifying the narrator function as logophoric rather than indirect is in Chapter 150 (First-Person Narration, the Logophoric Fourth Person, and Reported Discourse).

149.5 Prose narration: uwepeker, tuytak, and ucaskuma

The prose tale — uwepeker in Saru/Chitose, tuytak in most of eastern Hokkaido — uses no verse, no melody, and no sakehe. Length is intermediate. The narrative range runs from near-conversational to lightly elevated prose; corpus token counts give uwepeker 104 tokens and tuytak 24 tokens in the aligned Hokkaido corpus (the counts conflate metalinguistic and performance contexts).

Endō (2022) identifies four sub-genres within the prose tradition:

1. Human prose tales (the default): an unnamed human protagonist narrates in the first person within a 'real event' frame, typically with a reward-virtue / punish-vice arc.
2. Kamuy prose tales (kamuy uwepeker): a deity narrates; the register leans toward elegant language (atomte itak), paralleling kamuy yukar in prose form.
3. Japanese-borrowed prose tales: Ainu-ized versions of Japanese story types, told in the third person.
4. Pananpe–Penanpe tales: the dyadic pair 'downstream-dweller' / 'upstream-dweller', opening with the formula pananpe an penanpe an hine … 'there were Panampe and Penampe, and …', told in the third person and recognized as a children's genre.

The standard human prose tale is bracketed by formulaic structure: an opening that establishes the protagonist's family and lifestyle ("I had father, mother, elder brother; father hunted bear and deer; we lacked nothing…") and a closing identification formula in which the first-person narrator discloses its own identity via the reportative quotative frame sekor … isoytak / ye ('so X said / told'), revealing retrospectively who the "I" was throughout Endō (2022). This closure connects to the broader system of narrative reference tracking treated in Chapter 139 (Reference Tracking and Argument Continuity).

The opening move of a Saru uwepeker shows the narrator naming the genre in the first person before the tale proper begins — the performative self-announcement that frames the prose narration:

Ucaskuma / upaskuma designates a looser genre of oral tradition, legend, and historical account; it imposes no fixed verse or performance constraints and covers knowledge about places, events, and customs passed down across generations Endō (2022).

The intonational profile of prose narration differs observably from conversation. Tamura (1996) reports that uwepeker narration uses a one-directional, low-variation melodic line, contrasting with the rising contour of conversation that seeks a response; within the Saru basin, upper-Saru (mid-river) narration tends toward near-monotone, while lower-Saru shows more pitch undulation (nukopoli, aynu-corpora Discord 2024-11-16; aynu-corpora Discord (2023–2026)) ‹corpus-suggested›. The full TAM and evidential patterning by genre is in Chapter 155 (Narrative TAM and Evidential Patterning by Genre).

149.6 Genre-conditioned grammar: person marking and number in sung epic

All three principal narrative genres are told in the first person from the protagonist's viewpoint, but the person prefix used for that first person is not the everyday ku=. In Saru and Chitose both verse and prose narrative, the protagonist is indexed by the fourth-person prefix/suffix a= / =an — the form that in conversational Ainu marks an indefinite or inclusive-plural subject or, with a second-person referent, a deferential addressee Satō (2008: 260–261); Endō (2022); Bugaeva (2008). The full paradigm and typological analysis of this device are in Chapter 67 (The Indefinite/Fourth Person: Forms, Reference, and Agent-Defocusing), Chapter 68 (Honorific and Logophoric Uses of the Fourth Person), and Chapter 62 (Architecture of the Personal-Affix System: The Four Persons and the S/A/O Paradigms); the present section notes only the genre-conditioned contrasts.

Satō (2008: 259–261) names the use of non-everyday person marking in oral literature the "indirection principle" (間接性の原理): on his hypothesis, the narrative was originally framed as a god or spirit speaking to a human inside a dream, and the non-everyday first-person signals the taboo-laden indirectness of that divine address. The competing "shamaness-oracle-origin" account holds that the form arose from a ritual medium's stance Satō (2008: 259–260) ‹contested›. The relationship between this narrator-perspective device and typological logophoricity is analysed by Nikitina & Bugaeva (2021), who argue that logophoric speech constructions in Ainu are direct — not indirect — reported speech; the consequences for the genre-narration system are pursued in Chapter 150 (First-Person Narration, the Logophoric Fourth Person, and Reported Discourse).

Region and genre interact on the person marking. In Iburi (Horobetsu) kamuy yukar — the tradition documented in Chiri Yukie (1923) — the hero-narrator consistently uses the exclusive first-person plural forms ci= / =as as a singular narrator throughout each text Satō (2008: 261). In Chitose, ci= / =as may appear only at the narrative opening, with the body of the text reverting to a= / =an Satō (2008: 261). The hero-person choice is thus both genre-conditioned (verse vs prose may condition the choice in Chitose) and region-conditioned (Iburi vs Saru/Chitose use different forms throughout).

A second genre-conditioned morphological feature is the number-agreement override. In everyday Saru/Chitose, the suffix =an never co-occurs with a singular verb root: one says k=ek 'I come.SG' in conversation, and =an appears only with the suppletive plural (arki=an 'we come.PL'). In the elegant register of sung epic, the combination is grammatical:

The broader elegant-register morphology (applicative case-marking by prefixes e- / o- / ko- rather than particles, more polysynthetic words, genre-specific lexicon) is covered in Chapter 153 (The Elevated/Poetic Register: Archaic Morphology and Formulaic Diction).

149.7 Verse structure: the two metres

Verse in Hokkaido Ainu is not organized by rhyme. Kindaichi (1931) noted the absence of systematic rhyme; Okuda (2022), following Philippi (1979), confirms that repetition in Ainu poetry largely reflects the small vowel inventory and widespread reduplication rather than a prosodic constraint ‹contested›. Against this consensus, Tangiku (2020) argues that an 18th-century yukar text (ルウェサニウンクㇽ) uses deliberate head-rhyme, end-rhyme, and imperfect rhyme — a finding that, if sustained, would require revising the standard account (nukopoli, aynu-corpora Discord 2023-12-13; aynu-corpora Discord (2023–2026)) ‹contested›.

Okuda (2022) argues that Hokkaido Ainu verse deploys two metrically distinct systems, genre-differentiated. The first, syllable-count orientation, targets five syllables per line as basic (four-to-seven is the attested range), as described by Kindaichi (1931: 165) and Tamura (1973: 56); it is characteristic of lyric songs (yaysama). The second, accent orientation, places the first pitch accent in a fixed rhythmic slot (the first disyllabic foot) and admits four-syllable lines when the accent is line-initial; it characterizes divine epics (kamuy yukar). Heroic epics (yukar) use both systems depending on the musical rhythm of the performance Okuda (2022) ‹contested›. Okuda reports no other language in which one reciter alternates between two distinct metrical systems — a typological claim that rests on an argument from the present state of the literature rather than exhaustive survey ‹speculative›.

Techniques for achieving the five-syllable target in syllable-count verse include: the insertion of the light verb ki ‘do’ between a verb and a following element, uncontracted person prefixes, a prefixed initial u-, and retention of vowels that elide in conversational speech Satō (2008: 264); Okuda (2022). These techniques are also poetic-register diagnostics; they are analysed as part of the couplet and formulaic system in Chapter 152 (Parallelism, Couplets, and Word-Pair Doublets).

149.8 Competing taxonomies and contested genre boundaries

Each major scholar of Ainu oral literature has organized the genre system around different primary criteria. Kindaichi (1933) and Kubodera classified the corpus primarily by protagonist type (human vs divine vs culture-hero). Chiri Mashiho (Chiri (1955)) supplied the dialect-distribution data that revealed the naming mismatches across regions. Tamura (1996) showed that the Saru category of oyna is absent from at least one primary narrator's repertoire. Nakagawa (2001) established the four-function classification adopted by Endō (2022) and followed here. These schemes are not mutually falsifying — they weight different parameters — but they produce different genre maps and different descriptions of what counts as a variant of what.

At the level of individual texts, genre boundaries can be indeterminate. A kamuy yukar from the Shizunai collection narrated by 織田ステノ, titled "怠け者の子供" ('The lazy child'), carries a long sakehe (manayta sanke tar panke tar kocupu) but is shorter than typical kamuy yukar and concludes as a yayeisoytak (first-person self-narration) by a single human boy rather than a divine protagonist. Having melody rules out classification as prose, yet the text fits neither standard kamuy yukar (wrong protagonist type) nor yukar (carries a sakehe) (gengojiro, aynu-corpora Discord 2025-12-12; aynu-corpora Discord (2023–2026)) ‹contested›. Such cases show that the sakehe-vs-no-sakehe and divine-vs-human-protagonist parameters do not always co-vary as the taxonomy predicts.

The elegant-vs-everyday register distinction (see Chapter 153 (The Elevated/Poetic Register: Archaic Morphology and Formulaic Diction)) raises a related question. Kindaichi (1935: 381–382) ranked the density of elegant-language features across genres from least to most: greeting < prayer < incantation < caranke < yukar, treating the gradation as qualitative differences. Nakagawa (2001a[1987]: 222, cited via Endō (2022)) argues instead that elegant and everyday language are points on a continuum keyed to verse/prose and fixedness/improvisation, and proposes further that a genre-phrase classification (yukar-phrase vs prayer-phrase) may be more analytically effective than the elegant/everyday binary ‹contested›. That proposal predicts that the register differences between yukar and prayer, though real, are not of the same kind as the difference between elegant and everyday language generally; the evidential and TAM consequences of this distinction are tracked in Chapter 155 (Narrative TAM and Evidential Patterning by Genre).

The Ijäs pedagogical grammar (Ijäs (2023)) does not present a systematic genre taxonomy and is focused on spoken, conversational Ainu rather than the verse tradition. Where vocabulary or constructions associated with the oral-literature register appear in that work, the present chapter routes the reader to the philological and descriptive literature treated here.