Skip to content

Chapter 88Object (Patient) Noun Incorporation

Productive object incorporation (N+V → intransitive), its diagnostics, and its interaction with the antipassive and valency prefixes.

88.1 Object incorporation in the valency system

Object noun incorporation takes a nominal stem and absorbs it into a transitive verb as its Patient argument, producing a single verb word that expresses subject, incorporated object, and predicate together Nakagawa (2024: 480–483); Satō (2022: §1–2). In Nakagawa's arity calculus the transitive host carries arity +2 — one slot for a subject argument, one for an object argument. The incorporated noun fills the object slot at arity −1, and the resulting compound has arity +1: an intransitive verb Nakagawa (2024: 137, 480). cep ‘fish’ and koyki ‘catch, strike’ yield cep-koyki 'fish-catch': the nominal is inside the verb word and the object argument position is saturated.

The criterion for wordhood is the outer position of the person index Nakagawa (2024: 482–483); Satō (2022: §2). With a phrasal object the index attaches to the verb alone: cep ku=koyki has the 1SG prefix on koyki, with cep as a separate noun phrase. Under incorporation the index precedes the entire noun-verb complex: ku=cep-koyki 'I fish-catch'. With 4th-person marking the diagnostic is decisive: cep-koyki=an can only be one intransitive word, because the transitive root koyki cannot host the intransitive S suffix =an on its own Nakagawa (2024: 482–483). The overview of the valency system in Chapter 73 (Valency, Voice, and Noun Incorporation: Overview) introduces this diagnostic; the present chapter develops object-type incorporation specifically: the detransitivizing morphology, the semantic vocabulary of incorporated nouns, and the typological position of Hokkaido Ainu Object NI.

88.2 Detransitivization and person-marking consequences

Incorporating the object reduces the verb's argument count by one and shifts its person marking from the transitive to the intransitive paradigm. Satō's minimal pair makes the shift visible Satō (2022: §2): the transitive pukusa a=ta carries the 4th-person A prefix a=, which marks an indefinite transitive agent. Once pukusa ‘lily bulb’ is incorporated, the predicate becomes intransitive pukusa-ta=an, bearing instead the 4th-person S suffix =an.

(1)
pukusa lily.bulb
a=ta 4.A=dig

‘Someone digs lily bulbs.’

Satō 2022: §2

Transitive: a= marks an indefinite 4th-person agent; pukusa 'lily bulb' is a free NP object external to the verb.

(2)
pukusa-ta=an lily.bulb-dig=4.S

‘Someone digs lily bulbs.’

Satō 2022: §2

After object incorporation the intransitive S suffix =an replaces the transitive A prefix a=. The noun pukusa is now inside the verb word and can no longer appear as an external NP.

The suffix =an is impossible on a transitive verb: a transitive verb with an indefinite 4th-person agent must carry the A prefix a= Satō (2008: 96); Nakagawa (2024: 137, 140). It therefore serves as the definitive morphological test for intransitivity after Object NI. The same alternation appears with kinakar ‘herb-harvest’: the incorporated intransitive is kinakar=an, while the phrasal construction gives the transitive kina a=kar with the A prefix; this pair is consistent with the analysis in Satō (2022: §2) and has been independently noted in community discussion (thegodofneet, aynu-corpora Discord, 2024 aynu-corpora Discord (2023–2026)) ‹corpus-confirmed›.

The same intransitivization is attested across the Saru oral corpus. Kawakami Matsuko uses cep-koyki=an repeatedly in narrative; the 4th-person S suffix =an appears on what is otherwise the transitive root koyki ILCAA Ainu materials (1976):

(3)
pet river
or place
ta LOC
ran=an descend.SG=4.S
wa CONJ
suy again
cep-koyki=an fish-catch=4.S
rok PL.PRF
ayne then

‘I went down to the river and caught fish again, and at last …’

ILCAA Ainu materials 1976; Saru, Kawakami Matsuko, uwepeker 13; ILCAA aa-irc/016#14

The 4.S suffix =an on cep-koyki is the diagnostic: after O-NI the compound is intransitive. The compound cep-koyki 'fish-catch' occurs 144 times in the Hokkaido corpus (rank 848).

The person-marking shift is the sole morphological trace of Object NI. There is no overt incorporation morpheme, no case suffix on the noun stem, and no change to TAM morphology Nakagawa (2024: 480–483); Satō (2022: §2). The incorporated noun leaves no external argument slot: the Patient NP can no longer appear in the clause as a co-referential noun phrase. The two-affix ceiling — at most one subject index and one object index on any verb — bounds the system: Object NI uses up the object slot, leaving only the subject slot available Satō (2023: 39); Bugaeva & Kobayashi (2022: §2).

88.3 The four NI types and Chitose frequencies

Satō classifies Hokkaido Ainu noun incorporation into four types by the grammatical relation of the incorporated noun to the base verb Satō (2016: 83–84); Satō (2022: §3); Nakagawa (2024: 483–491). The distribution below comes from a count of 517 NI tokens in the Chitose-dialect speech of the narrator 白沢ナベ Satō (2016: 83–84); Satō (2022: §3). No pan-Hokkaido NI-type frequency study has been published; the percentages are Chitose-specific and derive from a single narrator's corpus ‹corpus-confirmed›.

type% (CHI)nMithun (1984) paralleltreated in
Object NI85.9444Type I — lexical compoundingthis chapter
Intransitive natural-force subject NI6.835Type IChapter 89 (Subject Incorporation and Possessor-Stranding)
Intransitive possessor-requiring subject NI5.629Type II — manipulates external case
Transitive natural-force subject NI1.79(Type I)

Object NI accounts for nearly seven eighths of all incorporation tokens. Its alignment with Mithun's Type I — lexical compounding in which the incorporated noun names a general activity type rather than a specific referent — is consistent with the Chitose data, though the theoretical status of Ainu NI more broadly is a live debate (Mithun 1984); Satō (2022: §5); Satō (2016: 83–84). The three subject-NI subtypes, including the possessor-stranding construction that is typologically the most marked, are treated in Chapter 89 (Subject Incorporation and Possessor-Stranding).

Ijäs's Lesson 3 introduces the two-argument transitive frame with the constructed example numan hekaci cep koyki 'the boy caught some fish' Ijäs (2023), the phrasal form of the root that appears incorporated as cep-koyki above. Her lesson does not treat incorporation; the contrast between the phrasal and the incorporated frame is developed here ‹consensus›.

88.4 Semantic fields of incorporated objects

Satō's Chitose count yields a breakdown of the semantic domains from which Object NI compounds draw their incorporated nouns Satō (2022: §3.1):

semantic fieldshare of O-NI tokens (CHI)representative examples
Space / location26%soy-osma 'plunge outside'
Body parts18%sik-erayke 'stare at'
Plants11%ni-tuye 'cut trees'; pukusa-ta 'dig lily bulbs'
Tools / house7%cise-kar 'house-build'; wakka-ta 'water-scoop'
Fire5%ape-ari 'make a fire'
Other / miscellaneous33%kem-eyki 'sew' (lit. 'thread-pierce')

The space category reflects the frequency of spatial compounds in which a location noun functions as the direct Patient of the incorporating verb: these are Object NI, with the locative noun already in the object grammatical relation. Oblique incorporations — instruments and locatives that were first promoted to object position by an applicative before incorporating — are a distinct construction treated in Chapter 90 (Oblique Incorporation and the Syntactic-vs-Lexicalist Debate) Shibatani (1990: 62, 66–67); Nakagawa (2024: 484). Body-part compounds cluster around perception and contact verbs; plant, tool, and fire nouns reflect the subsistence activities that are most elaborated in Ainu oral narrative. The following Chitose example illustrates the tool / house type, with cise ‘house’ absorbed into kar ‘make, build’:

(4)
toy land
ku=hok 1SG.A=buy
wa CONJ
or-o place-POSS
ta LOC
ku=cise-kar 1SG.S=house-build
wa CONJ
k=an 1SG.S=exist
ruwe track-POSS
ne COP
wa CONJ

‘I bought land, built a house there, and lived on.’

Satō 2022: §3.1; Chitose

ku=cise-kar: the 1SG prefix is the sole argument index on the compound, showing that cise 'house' has been absorbed and the resulting verb is intransitive. k=an is contracted ku=an 'I exist / live'.

Satō's corpus attests several NI↔phrase near-pairs — apa asi / apa-asi 'close the door', cep koyki / cep-koyki, wakka ta / wakka-ta — but these alternations are few and not systematically productive; most Object NI compounds have no attested phrasal counterpart Satō (2022: §3.1) ‹corpus-confirmed›.

88.5 The genericity restriction and the lexical character of Object NI

The incorporated noun must be semantically unbounded: a generic, non-referring stem Satō (2022: §2); Satō (2008: 225–226). A common noun such as cep 'fish' incorporates freely into cep-koyki 'fish-catch'. A specific proper noun cannot: *Sikotpet-tomotuye 'cross the Shikotsu river' is ungrammatical because a river name is an inherently referring expression Satō (2022: §2); Satō (2008: 225–226).

Ainu Object NI groups with Mithun's Type I — lexical compounding in which the incorporated noun names the type of object typically affected rather than a referential argument (Mithun 1984); Satō (2022: §3.1, §5); Satō (2016: 83–84). Nakagawa places the same character in a distributional frame: NI compounds are commoner in verse than in ordinary prose or conversation, and they represent a lexicalized stratum of the verb vocabulary rather than a productive derivational slot Nakagawa (2024: 480). Whether Ainu NI is best analysed as purely lexical, as Baker's syntactic head-movement, or as a hybrid that must reference both grammatical relations and semantic roles as Shibatani proposes is a dispute the sources consulted do not resolve; it is taken up in Chapter 90 (Oblique Incorporation and the Syntactic-vs-Lexicalist Debate) (Shibatani 1988); Shibatani (1990: 60–66); (Baker 1988) ‹contested›.

A small set of verbs permits simultaneous incorporation of two nominal stems, producing rare three-element compounds. Nakagawa cites cep-ya-o-kuta 'dump fish onto the shore' — incorporands cep 'fish' and ya 'shore', mediated by the applicative o- — and itak-mak-kus-te 'use hidden words' as attested cases Nakagawa (2024: 488). The general limit is one incorporated noun per verb Shibatani (1990: 72), and these forms are marginal in the corpus ‹corpus-suggested›.

NI compounds that have lost any productive phrasal alternant are synchronically lexicalized denominal verbs; the derivational pathway that produces them connects to the broader denominal processes treated in Chapter 60 (Denominal and Deverbal Verb Derivation). A diachronic line also links Object NI to the antipassive prefix i- ‘antipassive’: Bugaeva proposes that i- arose from grammaticalization of O-incorporation of a generic noun, with the 4th-person pronoun i= derived secondarily from the antipassive Bugaeva (2021: 235–239), while Nam reconstructs the source as a distinct generic-noun stem Nam (2021: 63–74); the dispute belongs to Chapter 83 (The Antipassive / Detransitive i-) ‹contested›.

Where Object NI is blocked — most critically at the possessor-requiring type (impossible because an inflectional possessor makes the incorporated noun specific, and no reflexive rescue is available) — the function passes to quasi-incorporation (QI), a tight phrasal unit that is idiomatic but not yet a single verb word Satō (2022: §4, §6); Satō (2016: 88–92). The QI construction and the markedness calculus governing the NI–QI division are developed in Chapter 90 (Oblique Incorporation and the Syntactic-vs-Lexicalist Debate).

References cited in this chapter

aynu-corpora Discord (2023–2026) ·Baker (1988) ·Bugaeva (2021) ·Bugaeva & Kobayashi (2022) ·Ijäs (2023) ·ILCAA Ainu materials (1976) ·Mithun (1984) ·Nakagawa (2024) ·Nam (2021) ·Satō (2008) ·Satō (2016) ·Satō (2022) ·Satō (2023) ·Shibatani (1988) ·Shibatani (1990)