Chapter 87Middle ci- and the Passive Analysis
Middle and resultative descriptions, person restrictions, agent phrases, competing analyses of passive patients, and anticausative meanings.
87.1 Affected participants and different constructions
Ainu has several constructions that present an affected participant without an ordinary personal actor as subject. They include middle ci-, passive-like uses of fourth-person A marking, and verbs in si- with spontaneous or anticausative meanings. Their morphology, participant structure and lexical distribution differ Nakagawa (2024: 189–190, 428–430); Bugaeva & Kobayashi (2022: §§4.2.2, 4.2.4).
87.2 Middle ci- and resultative descriptions
Nakagawa derives one-place ci-cari “be scattered” from two-place cari “scatter.” In his middle analysis, the affected participant becomes the subject Nakagawa (2024: 189). The Handbook uses the descriptive gloss RES, “object-resultative,” while noting that the construction has received limited study Bugaeva & Kobayashi (2022: §4.2.4).
‘Now human flesh lay scattered on the snow, and the monster's flesh also lay scattered, and …’
Nakagawa 2024: 189; Chitose, Example 167
The first and third ka are additive particles; the ka after upas is the spatial noun ‘top’. The snow is the location of the scattered flesh.
The resultative description is also illustrated by a phrase from Tamura's Saru dictionary.
‘A place where blood clots are scattered.’
Bugaeva & Kobayashi 2022: §4.2.4; Saru, Example 28, quoting Tamura 1996:48
The final i in ci-car-i is the verbal TR.SG suffix in the source's segmentation. Possessive morphology occurs on kem-ihi and uske-he.
Nakagawa reports that these middle verbs occur mainly in the third person, with other persons uncommon. Shirasawa Nabe rejects *ku=cimontapire for “I was caught off guard and flustered” and prefers a=en=montapire. This is a particular Chitose judgment, alongside his more qualified statement about the class Nakagawa (2024: 189).
In Saru and Chitose, ci- often loses its vowel before a, u, e, o, giving ca, cu, ce, co. Nakagawa relates it historically to personal ci=, without equating their present functions ‹speculative› Nakagawa (2024: 189–190). Refsing reports no middle a-/ci- examples in her Shizunai material and describes spontaneous events with si- Refsing (1986: 31). These descriptions have different dialect and material coverage.
87.3 Indefinite-person forms in passive constructions
Fourth-person A forms can background the actor. An agent can be expressed with or(o) wa, a locative-noun and ablative expression. Nakagawa gives the following comparison to illustrate the construction Nakagawa (2024: 428–429).
The affected participant has O-series person marking when overtly indexed. The agent phrase can still identify a particular actor, so the traditional term “indefinite” does not require an unknowable agent in every passive use Satō (2008: 205–206); Shibatani (1990: 58–60). Dialect forms and the other uses of the series are described in Chapter 67 (Fourth-Person Forms and Their Uses).
Nakagawa distinguishes this construction from middle ci-, which he describes as unable to take the same agent phrase. He also discusses different uses of ci- in poetic expansions Nakagawa (2024: 430). A similar English passive translation can therefore correspond to grammatically different Ainu expressions.
87.4 The syntactic status of the passive patient
Satō and Nakagawa retain object status for the affected participant in the a= construction. Satō supports this analysis with the desire construction: a speaker who wants to be loved uses a nominalized desired event plus ku=ki rusuy, as when the desirer differs from the subject of the desired event Satō (2008: 205–206); Nakagawa (2024: 428–429).
‘I would like my grandmother to love me, but …’
Satō 2008: 206; Opening of the grandmother example
The sentence continues by saying that grandmother does not love the speaker. En= indexes the speaker as object in the nominalized event; ku= marks the desirer on ki.
Shibatani argues for syntactic subjecthood despite the object morphology. His evidence includes the patient's position before an ablative agent phrase and honorific plural verbs referring to the patient in Classical Ainu. He interprets the mismatch as syntactic reanalysis of an indefinite-person construction, with older morphology surviving ‹contested› Shibatani (1990: 58–60). His account and Satō's differ in the syntactic diagnostics and material they emphasize. Their disagreement concerns grammatical relations as well as the choice of the term “passive.”
87.5 Spontaneous change and anticausatives
Some verbs in si- present a change without an expressed causer. Bugaeva distinguishes these anticausatives from self-induced motion, which she calls autocausative, and proposes a historical development through reflexive, autocausative and anticausative functions ‹speculative› Bugaeva (2025: 178–181). The lexical constructions and competing accounts of si- are described in Chapter 85 (Reflexive si- in Verbal and Spatial Constructions). This family of meanings does not make every reflexive verb spontaneous or every passive construction intransitive under all analyses.
References cited in this chapter
Bugaeva (2025) ·Bugaeva & Kobayashi (2022) ·Nakagawa (2024) ·Refsing (1986) ·Satō (2008) ·Shibatani (1990)