Chapter 73Valency and Voice Constructions

73.1 Arguments, indexes, and expressed phrases

Valency concerns the arguments a predicate takes. A clause need not express each argument with a separate noun phrase, and a verb's person indexes do not directly count its arguments. A three-place verb may have a subject and two objects while displaying only an A–O person combination Satō (2008: 147–148); Nakagawa (2024: 127, 135–137). These distinctions underlie the valency and voice constructions described in this part of the grammar.

Person marking can diagnose a change where the paradigm distinguishes S and A. Fourth-person =an versus a=, and first-plural =as versus ci=, supply such contrasts. First-singular ku= is shared by S and A Nakagawa (2024: 166–170). The latter form alone cannot identify a predicate's transitivity.

73.2 Common derivational effects

These are the principal effects described by Nakagawa and by Bugaeva and Kobayashi Nakagawa (2024: 137–140, 185–210); Bugaeva & Kobayashi (2022: §4). They do not assign one invariant valency value to every occurrence of a written prefix. An incorporated modifier, for example, need not remove a core argument Satō (2022: §§1–2).

73.3 Causation changes the participant structure

The participant playing changes from S to O; the mother is the added causer and transitive subject. Nakagawa uses this pair to show a distinction that would remain unexpressed if both participants were third person Nakagawa (2024: 431). See Chapter 74 (The Productive Causative -re, -te, and -e) and Chapter 76 (Causees and Multiple Causative Layers).

Indefinite -yar/-ar has a different effect. The causer replaces the base's subject, and the causee is unavailable as a core object, so the number of core arguments remains unchanged Nakagawa (2024: 200–201, 432–433); Bugaeva & Kobayashi (2022: §4.3.1). An oblique phrase can still identify a causee in documented constructions; see Chapter 77 (Indefinite Causative -yar and -ar).

73.4 Lexical pairs and morphological analysis

Lexical pairs such as mak-ke “open” intransitively and mak-a “open something” differ in their argument structure. Nakagawa treats their endings as forming suffixes on a root and assigns no independent arity to those suffixes Nakagawa (2024: 199). Bugaeva places such formations within a broader comparison of direct causation, while Kobayashi distinguishes forming suffixes from causative suffixes Bugaeva (2014: 130–132); Kobayashi (2015: 33–35) ‹contested›.

The analyses differ over how the related words are derived. A table of argument counts should therefore distinguish an operation on an existing verb from alternative formations of verbs from a shared root. The lexical patterns are treated in Chapter 75 (Lexical Verb Pairs and Causative -ka).

73.5 Voice, interpretation, and complex verbs

Changes in viewpoint do not always change the number of indexed arguments. Indefinite-agent constructions can receive passive translations, with competing accounts of the affected participant's syntactic status Satō (2008: 203–206); Shibatani (1990: 58–60) ‹contested›. See Chapter 87 (Middle ci- and the Passive Analysis) for the evidence behind those analyses.

Complex verbs can combine several operations, but each attested combination needs its own argument account. An inventory of double applicatives, causatives and incorporation patterns does not establish that every operation can be repeated together without restriction Nakagawa (2024: 207–210); Bugaeva & Kobayashi (2022: §6). The distinction between person markers and derivational elements is described in Chapter 56 (The Verb Word and Its Boundaries).