Chapter 76Causee Marking, Ditransitive and Double Causatives

76.1 Overview

The productive causative suffix -re/-te/-e raises arity by one. Applied to an intransitive base it creates a two-place transitive; applied to a transitive base it creates a three-place ditransitive; applied a second time to a derived causative stem it produces a four-place verb. The chapter follows the causee — the participant who is made or permitted to act — through each of these steps: its expression as the O of a derived transitive, the priority it takes over the theme when both objects compete for the single O-index slot, and the behaviour of double causatives under the two-slot ceiling. The allomorphy and the productive versus non-productive distribution of causative morphology are developed in Chapter 74 (The Morphological Causative -re / -e / -te). The indefinite causative -yar/-ar, which adds a causer while leaving the causee both unexpressed and unindexed, is treated in Chapter 77 (The Permissive/Sociative Causative -yar). Bugaeva (2014: §7)

76.2 The causee as object

When -re/-te/-e derives a transitive verb from an intransitive base, the former intransitive subject becomes the O of the causative verb. Person marking shows this shift directly. The intransitive sinot ‘play (intr)’ takes the first-person A prefix ku=; once the causative -te is added, that same participant reappears as the first-person object en=, and the new causer — hapo 'mother' — occupies the A position, unmarked because third-person A is zero Nakagawa (2024: 431).

The causee bears no case particle. Ainu objects are invariably unmarked, so the causee surfaces as a bare NP alongside any O-set affix that cross-references it, in exactly the same way the patient of any transitive does Bugaeva (2014: 142); Nakagawa (2024: 201). The same principle carries into the three-place frame that arises when a transitive base is causativized: the marking type does not change, only the number of competitors for the single O-index slot increases.

76.3 Causative of a transitive: the three-place frame

When -re/-te/-e applies to a transitive verb (arity +2), the result has arity +3. The new causer fills the A position, the original transitive subject (now the causee) fills one O position, and the original patient or theme fills a second O position. Bugaeva (2014: 142) characterizes the causee in this frame as an unmarked direct object in a double-object construction, parallel in its coding to a recipient in an applicative-derived ditransitive Bugaeva (2014: 142); Bugaeva & Kobayashi (2022: §4.1.2).

A corpus example from the Saru texts in the Nakagawa archive illustrates all three participants at once. The base verb is tuy-e ‘cut (transitive)’, the transitivized member of the tuy/tuy-e pair; adding the causative -re raises arity to three. The causer is indexed by a= (fourth person, functioning as narrative first person), the causee 'you' by e= (2SG.O), and the theme appears as a bare NP without any person affix on the verb:

Ijäs (Basic Lesson 3) uses the label 'ditransitive' for three-place verbs such as epakasnu ‘teach’ and kore ‘give’, and correctly identifies a three-argument structure. Her accompanying claim that the indirect object falls between subject and direct object in a fixed position is a pedagogical simplification: Nakagawa's data show the two objects of a ditransitive in variable order Ijäs (2023); Nakagawa (2024: 363–365) ‹consensus›.

The table below tracks arity through the causative chain. The third row covers double causatives, taken up in §5. Nakagawa (2024: §13.3); Bugaeva (2014: §7)

76.4 Person indexing: causee priority and secundative alignment

An Ainu verb carries at most two person affixes — one in the A-set and one in the O-set — the two-slot ceiling described in Chapter 73 (Valency, Voice, and Noun Incorporation: Overview) Nakagawa (2024: 210–212); Bugaeva & Kobayashi (2022: §2); Satō (2023). In the three-place causative frame two objects compete for the single O-slot. The attested indexing pattern, exemplified by a=e=tuy-e-re, gives the O-slot to the causee and leaves the theme as a bare NP.

This outcome places Ainu causative ditransitives in the secundative column of ditransitive indexing typology (see Chapter 66 (Object Indexing: Monotransitive and Ditransitive Clauses)): the causee, as derived O, patterns with the O of a monotransitive verb for person-marking purposes, rather than the original patient claiming that slot Satō (2023); Bugaeva & Kobayashi (2022: §4.1.2). The pattern mirrors what applies in applicative-derived ditransitives, where the applied object — not the original patient — is indexed in the O-position ‹corpus-suggested›.

Whether the reverse indexing (theme indexed, causee as bare NP) ever occurs is not settled in the sources consulted ‹speculative›.

A lexicalized sub-pattern involves the fourth-person object affix i= serving as an indefinite causee. Forms such as iyairaykere and irankarapte have been proposed by nukopoli (aynu-corpora Discord, November 2024) as i= + stem + -re/-te, glossable as 'cause people/someone to be grateful' and 'cause people/someone to greet' respectively. On this reading i= occupies the O-slot as an indefinite causee, and some lexicalized tokens drift toward an adjectival-modal reading analogous to English excit-ing alongside excit-ed. The analysis is not yet verified against the corpus and Nakagawa's (2024) treatment of these forms has been judged insufficiently detailed by the same discussant aynu-corpora Discord (2023–2026) ‹contested›.

If the i=-causee reading is correct, causee expression in the productive causative system presents a three-way gradient: a specific causee (named NP plus O-set affix on the verb) → an indefinite causee (i=-affix, cross-referenced but unspecified) → an unindexed causee (-yar, not cross-referenced at all). The antipassive prefix i- can similarly fill the causee slot with a generic object when applied to a causative-derived stem — Bugaeva's examples i-newsar-ka 'amuse people' and i-oripak-ka 'make people feel humble', reported via the Handbook — and forms a related but distinct indefinite-causee strategy Bugaeva & Kobayashi (2022: §4.2.3) ‹corpus-suggested›.

76.5 Double causatives

The productive suffix applies iteratively. A derived causative stem may itself serve as the input to a second causativization, and the allomorphy rules apply at each step. After the vowel-final first causative -e (the allomorph after r), the second causative surfaces as -re (the allomorph after a vowel), producing the sequence -e-re Nakagawa (2024: 200–201); Bugaeva (2014: 134).

kor ‘have, hold (transitive)’ is a two-place verb. The first causative suffix -e (after r) yields kor-e ‘give (causative of kor: cause to have)’ (+3): the giver is the new A, and the original possessor becomes a causee-beneficiary in the O position. A second -re yields kor-e-re (have-caus-caus, +4): an outer causer now makes the giver give. Bugaeva (2014: 142–143) records kor-e-re as attested once in her 387-verb Saru sample Bugaeva (2014: 142–143).

omarere ‘cause someone to put X in Y’ provides a second documented instance. The base oma ‘be inside (a place)’ is a locative verb; omare ‘put X in Y (causative of oma)’ is its three-place causative; omare-re is the double causative with four participants Bugaeva & Kobayashi (2022: §4.1.2).

The two-slot ceiling limits what can be cross-referenced in any four-place frame. Of four participants, only the outer causer (A-slot) and the immediate causee (O-slot) can be indexed by person affixes on the verb; the remaining objects must be expressed as overt bare NPs. The asymmetry between a morphologically productive stacking operation and its restricted expressibility in person marking explains the marginal corpus status of double causatives. Bugaeva and Kobayashi observe that the licensing of double-object frames and double causatives in Ainu is cross-linguistically unusual, most languages blocking iterative morphological causativization Bugaeva & Kobayashi (2022: §4.1.2) ‹corpus-suggested›.

76.6 Combinatorial limits and related morphology

The causative sits within a broader combinability system in which two applicatives, two causatives, two incorporated objects, reflexive, and agreement indexes can co-occur on a single verb — surveyed in Chapter 73 (Valency, Voice, and Noun Incorporation: Overview) Bugaeva (2014); Bugaeva & Kobayashi (2022: §4.1.1). The two-slot ceiling is the shared operational bound across all combinations Satō (2023); Bugaeva & Kobayashi (2022: §2).

Kobayashi's argument-structure analysis is relevant background here. She separates causative derivation — which always adds an agent-like participant, displacing the original subject to O — from applicative derivation, which adds a non-agent participant to the object set. The distinction predicts which slot a new participant targets and clarifies when applicative and causative are alternatives for expanding the argument frame Kobayashi (2015: 33); Kobayashi (2020). The non-causative side is taken up in Chapter 78 (The Applicative System: Overview and the Preverb-vs-Applicative Debate).

Takahashi (2017) reports that in Horobetsu the causative suffix acquires an additional switch-reference function in clause linkage: in coordinate chains it marks a different subject in the following clause; in subordinate chains it marks the same subject. This function is not attested in descriptions of Saru or Chitose and does not bear directly on causee marking, but it shows the causative suffix to be morphologically multifunctional in at least one dialect Takahashi (2017: 102–106) ‹corpus-suggested›.

For the non-productive lexical causatives (-V, -ke, -ka) and their relationship to transitivity pairs, see Chapter 75 (The Transitivizer -ka and Lexical Valency Pairs); for the arity classes themselves, Chapter 57 (Transitivity, Valence Classes, and Ambitransitivity).