Chapter 75The Transitivizer -ka and Lexical Valency Pairs

75.1 Overview

Three morphological mechanisms build transitive verb forms from intransitive bases in Hokkaido Ainu, all operating within the arity calculus set out in Chapter 73 (Valency, Voice, and Noun Incorporation: Overview). The oldest layer consists of stem-specific forming suffixes — a vowel suffix -V or the suffix -ke — that produce lexically listed transitivity pairs; these are non-productive and the intransitive and transitive members are recorded as separate dictionary entries Nakagawa (2024: 198–199). A second suffix, -ka, overlaps in function with the forming suffixes but extends to stative and property bases and to derived stems, placing it on the boundary between the forming-suffix category and the productive direct causative; the literature has not settled its classification Bugaeva (2014: 129–130, 143); Kobayashi (2015: 33) ‹contested›. The third mechanism, the transitivizer -kar, productively adds an object to any intransitive base (arity +1) and must be distinguished from two surface-identical strings: a verse-filler -kar appended to transitives with no arity effect, and the verb kar ‘make, do’ that appears as the head of incorporated-object constructions such as cise-kar 'build a house' Nakagawa (2024: 209–210).

The productive indirect causative -re/-te/-e, which supplies the default transitivizing strategy and gradually displaces -ka in the more southerly dialects, is treated separately in Chapter 74 (The Morphological Causative -re / -e / -te). The valency classes and the transitivity test that diagnoses derived verb type are discussed in Chapter 57 (Transitivity, Valence Classes, and Ambitransitivity).

75.2 Forming suffixes and lexical transitivity pairs

Nakagawa's paradigm tables 表38 and 表39 set out two structural patterns Nakagawa (2024: 198–199). In the first, the intransitive is the bare root and the transitive adds a final vowel (a suffix -V that assimilates to the stem shape). In the second, the intransitive already carries -ke and the transitive replaces it with the vowel suffix. For the motion-verb class the pattern reverses: the bare root is intransitive and -ke marks the transitive. Table 1 samples all three patterns, ending with the -ka pair discussed in the following section.

Bugaeva's 387-verb Saru sample counts 111 verbs with -V and 9 with -ke, classifying all as direct causatives on semantic grounds Bugaeva (2014: 130, 143). The -V set is closed and non-productive; no new pairs of this type are formed, and the members are listed separately in the dictionaries Tamura (1996). The three motion verbs that take -ke as their transitive suffix (ran, ahun, yan) have plural counterparts formed with the productive suffix -te instead Nakagawa (2024: 202–203).

75.3 The suffix -ka: hosts, dialect distribution, and semantics

Among the motion-verb set, rikin ‘go up (vi)’ takes -ka as its transitive suffix in most Hokkaido dialects. Nakagawa's 表43 also lists a further set of verbs to which -ka attaches without belonging to the motion-verb class: retar ‘be white’, us ‘go out’, iunin ‘settle (vi)’, and several dialect-restricted forms (teyne in Ishikari; een in Yakumo, Saru, and Bihoro). The distribution across this second group is a lexical property of the individual verbs, not a regional feature Nakagawa (2024: 203).

In Saru, the pair nam-te ~ nam-ka 'make cold' shows intra-dialectal competition between the productive suffix -te and -ka on the same base Kobayashi (2015: 33); the analytical implications are taken up in the debate section below. Cross-dialectally, -ka covers a broader verb class in northeastern Hokkaido; the replacement of -ka by -re/-te/-e is most advanced in Saru, as Bugaeva's sample of 38 -ka verbs against 190 -re/-e/-te verbs shows Bugaeva (2014: 135–136, 143–144).

The semantic boundary between the forming-suffix causatives and the productive -te/-re is visible on verbs that admit both: for ahun ‘enter’, ahun-ke encodes direct causation with a non-agentive, inanimate causee (putting luggage inside), while ahun-te or ahun-re is used when the causee is an agentive human brought inside Nakagawa (2024: 202–203). The parallel holds for -ka: it signals that the causee is treated as a highly affected patient (direct causation), contrasting with the indirect or permissive readings accessible to -re/-te/-e Bugaeva (2014: 130, 139).

The suffix -ka also attaches outside the indirect reflexive prefix si- to produce a 'pretend to' reading. Satō (2007) analyses this construction within his account of si- as marking indirect self-involvement: a pretense is only effective through others' perception, which is precisely the situation the indirect reflexive encodes Satō (2007: 36).

75.4 The transitivizer -kar

The suffix -kar productively transitivizes intransitive verbs (arity +1): peray ‘fish (vi)’ → peray-kar ‘catch (a fish)’; cis ‘cry (vi)’ → cis-kar ‘cry over sth’; hotuye ‘shout (vi)’ → hotuye-kar ‘call sb’ Nakagawa (2024: 209). Its productivity range is wider than that of -ka: no restriction to a semantic class or phonological stem shape applies, and derived forms can be created from any intransitive base.

Three surface-identical strings require disambiguation through the arity computation Nakagawa (2024: 209–210). First, the transitivizer -kar itself, raising arity by one on an intransitive base. Second, a homophonous verse-filler -kar that appends to transitive stems with no arity change and no semantic contribution — a poetic register device extending syllable count: kocari (a three-place verb) remains three-place as kocari-kar. Third, the verb kar 'make, do' that appears as the head of incorporated-object constructions: in cise-kar 'build a house', the noun cise 'house' fills the object slot of the two-place verb kar, and the arity is that of kar itself Satō (2022: §3.1).

Community observation identifies a productive 'treat-verb' pattern within the third type: forms such as nimaw-e-kar 'treat a wound with medicinal bark', osoro-kar 'wipe', and sikaye-kar 'hammer in a peg' each incorporate an instrument or body-part noun into kar 'do' aynu-corpora Discord (2023–2026) (nukopoli, aynu-corpora Discord 2024). These are incorporated-object constructions and belong to the account in Chapter 88 (Object (Patient) Noun Incorporation), not to the transitivizer suffix ‹corpus-suggested›.

75.5 Direct causative or forming suffix: the analytical debate

Nakagawa treats the forming suffixes (-V, -ke) as 形成接尾辞 listed in 表38–39, a distinct category from the causative suffixes of 表41–43, which include -re/-te/-e, -ke for motion verbs, and -ka Nakagawa (2024: 198–203). The category distinction rests on closed membership and the absence of productive alternation: the -V and -ke pairs are dictionary-listed lexical alternants.

Bugaeva (2014: 129–130, 143) unifies -V, -ke, and -ka as direct causatives, opposing them as a group to the indirect causative -re/-te/-e by the non-agentive, highly-affected causee they encode. The 387-verb Saru count — 111 verbs under -V, 9 under -ke, 38 under -ka — establishes the proportions but does not settle the morphological-status question independently Bugaeva (2014: 130, 143).

Kobayashi (2015: 33) adduces evidence for separating -ka from the most lexicalized forming suffixes. Two properties distinguish it: intra-dialectal alternation with -re/-e/-te (the nam-te ~ nam-ka Saru pair noted above), and attachment to derived stems — ar-us-tek-ka 'annihilate' and ipe-rusuy-ka 'make hungry' both show -ka appended to a multi-morpheme derived stem, a combinatorial freedom the closed -V and -ke forming suffixes do not share Kobayashi (2015: 33). The si-…-ka 'pretend' construction described above provides a further case: -ka attaches outside the reflexive prefix si-, which has already reconfigured the argument frame.

Satō (2008: 235–238) includes -ka in the causative allomorphy list alongside -re, -te, -ke, and -e, stating that it tends to follow stems in m or y — "大体 as a tendency, with exceptions." The qualifier matters: a pure phonological allomorph would show regular conditioning; the exceptions reflect a lexical residue that fits neither a clean allomorphy account nor a uniform forming-suffix categorisation. Satō also notes that some -ke and -e forms are bleached transitivizers with no causative semantics (san-ke 'put out a fine' from san 'go out'), which aligns with Nakagawa's forming-suffix category and places the productive/lexical boundary inside, not between, these suffixes Satō (2008: 235–238).

Ijäs's Lesson 9 states that -re 'turns intransitive verbs into transitive verbs', which captures the right intuition for the valency-increasing set as a whole but supplies no allomorphy and does not distinguish -ka, -ke, or the lexical forming-suffix pairs as separate sub-types Ijäs (2023) ‹consensus› (the source gives no allomorphy detail).

75.6 Attachment to the base stem: rikinka and the absence of rikipka

The verb rikin ‘go up (vi)’ belongs to a class of Hokkaido Ainu motion-verb intransitives showing a -n/-p stem alternation in reduplication and some plural contexts, paralleling san/sap, yan/yap, and ahun/ahup aynu-corpora Discord (2023–2026) (nukopoli, aynu-corpora Discord 2026) ‹speculative›. In reduplication contexts rikin alternates with rikip.

When -ka attaches, only rikin-ka is attested; the form *rikip-ka is absent from the corpus aynu-corpora Discord (2023–2026) (nukopoli, aynu-corpora Discord 2026) ‹corpus-suggested›. The distribution shows that -ka targets the citation form of the stem — the singular -n variant — not the alternate -p shape that arises in reduplicated environments. The attachment point is the lexical base, prior to any stem alternation, a property shared by other causative suffixes that affix to the verb stem before verbal-number morphology applies. This behavior is consistent with Kobayashi's evidence that -ka operates at the same morphological level as the productive causative rather than with the most lexicalized forming suffixes Kobayashi (2015: 33).

The diachronic account of the -V and -ke/-ka pairs as anticausatives in origin is explored in (Bugaeva 2025). The middle and anticausative patterns that overlap synchronically with the transitivity pairs are treated in Chapter 87 (Middle, Anticausative, and the (Non-)Passive Question), and deverbal derivation more broadly in Chapter 60 (Denominal and Deverbal Verb Derivation).