Skip to content

Chapter 58Verbal Number and Stem Alternation

Lexical number pairs, participant and event number, dialect forms, additional pa and ci, and the effects of respect and narrative person.

58.1 Number is a property of particular verb stems

Some Ainu verbs have distinct singular and plural stems. The distinction is especially visible in existence, posture, movement, and certain transitive actions; most verbs have no such pair Nakagawa (2024: 141–148); Satō (2008: 101–104). Stem number is therefore a lexical property to learn alongside a verb's argument frame.

Intransitive alternations generally follow the subject's number. Transitive alternations often concern the object or repeated action, but there are exceptions. The transitive pair eramuan / eramuoka “understand,” for example, follows the subject in Nakagawa's Saru examples Nakagawa (2024: 142–143). A causative may retain the base verb's number contrast: the participant that was its subject becomes an object of the causative Nakagawa (2024: 146).

58.2 Pairs with different stems

The following selected pairs are given in Nakagawa's paradigms. The table separates ordinary intransitives from transitive and locative frames Nakagawa (2024: 146–148, tables 16–17).

Frame and meaningSingularPlural
Existanoka, okay
Comeekarki
Goarpa, omanpaye
Standasroski
Sitarok
Stand something upasiroski
Take somethingukuyna
Be in a placeomao

The alternatives need regional qualification. Arpa is used in Saru, Chitose, and Mukawa among the varieties listed; oman occurs elsewhere. Chitose has both oka and okay, while Saru normally has oka with conditioned occurrences of okay Nakagawa (2024: 147, notes 1–2).

For “put,” Nakagawa establishes Saru and Chitose anu / ari and Yakumo and Obihiro ama / are; the occurrence of a singular form in another region does not establish its plural partner there. For “kill,” he records several regional forms and specifically notes that Tokachi ronno lacks the corresponding singular rayke in that dialect's system Nakagawa (2024: 147–148, notes 4, 7–8).

58.3 Consonant alternation and stem-forming pa

A group of directional intransitives alternates final -n and -p: rikin / rikip “ascend,” ran / rap “descend,” san / sap “come down from the mountains,” and ahun / ahup “enter” Nakagawa (2024: 148, table 18). Other apparent pairs have narrower documentation: the plural repup is reported only from Shiranuka in that survey, and makap from Samani, Kushiro, and Sakhalin Nakagawa (2024: 148, notes 10–11).

A further pattern replaces a final vowel with -pa, as in tuye / tuypa “cut,” kaye / kaypa “break,” and komo / kompa “bend something.” Many such transitive pairs concern a change in the object, but ani / anpa “hold” illustrates a different meaning Nakagawa (2024: 149–150, table 20).

Derived intransitives can retain the alternation. Nakagawa relates hopuni / hopunpa “get up” to puni / punpa “raise something.” In the intransitive pair, the contrast follows the subject Nakagawa (2024: 150–151, table 21).

(1)
kunneywa morning
tunas, early
a=yupi 4.A=elder.brother
hopuni get.up.SG
hine CONJ
suke cook

‘Early in the morning my elder brother got up and cooked.’

Nakagawa 2024: 151, example 57; Saru

(2)
kunneywa morning
an exist
hine, CONJ
a=yuputari 4.A=elder.brothers
ka ADD
hopunpa get.up.PL

‘Morning came, and my elder brothers also got up.’

Nakagawa 2024: 151, example 58; Saru

58.4 Plural reference does not always require a plural stem

Nakagawa gives transitive singular stems with explicitly plural objects. In a Saru narrative, a=eci=rayke combines second-person plural object marking with singular “kill.” Other examples describe repeated hunting with either singular rayke or plural ronnu Nakagawa (2024: 144–145, examples 48–52). He consequently treats singular as a default in such transitive pairs, with plural making the multiplicity salient.

The intransitive pattern is different in his analysis. A single husband going back and forth repeatedly retains singular “go” and “come”; repetition alone does not require the plural stems Nakagawa (2024: 145, example 53, from Tamura 1984: 30). Shibatani discusses lexical selectional restrictions as an alternative to ordinary agreement for much of the number system Shibatani (1990: 52–54) ‹contested›.

Numerals and quantifiers often occur with singular forms, but plural forms are also documented. Their distribution supplies no fixed numerical threshold for switching stems Nakagawa (2024: 160–162); Refsing (1986: 117–118). See Chapter 55 (Numeral Syntax, Ordinals, and Quantity).

58.5 Additional pa and ci

Additional pa must be distinguished from the stem-forming pattern above. It can follow an already plural stem or causative morphology and can occur after a person suffix. Nakagawa analyzes it as an auxiliary; the Handbook discussion also considers clitic terminology Nakagawa (2024: 156–160); Nakagawa (2022: §5.1). Satō's Chitose description retains suffixal terminology and reports limited evidence for its exceptional position Satō (2008: 103–104) ‹contested›.

Additional pa can express participant plurality or repeated action. Nakagawa explains its occurrence after an already plural stem by assigning the two forms different contributions: the stem may count participants and the following pa repeated events, or the stem may express object/action plurality while pa identifies several actors Nakagawa (2024: 157–158).

Satō gives accepted yaynu-an-pa and rejected *yaynu-pa-an in 2008. His later discussion reports a different elicitation in which proposed yaynu-as pa was returned without pa; the further form he considers is a possibility, not an established correction to the earlier example Satō (2008: 104); Satō (2025: 26).

The distribution of ci also differs by dialect. Nakagawa contrasts Horobetsu iki-ci=as, with the number element inside the person suffix, and Shizunai hotke=an ci, with it after the inflected verb. Additional pa occurs in Saru and Chitose, but also Tokachi and Ishikari in his account Nakagawa (2024: 155–159). These distributions are more specific than a western–eastern division.

Nakagawa compares Sakhalin (a)hci with these forms, drawing on Murasaki's descriptions. Its person and number distribution, and its status as an affix or a following element, receive different analyses Nakagawa (2024: 153–156); Nakagawa (2022: §5.3) ‹contested›.

58.6 Respect and narrative person

Plural forms can address or refer respectfully to a single person. The imperative particle yan and honorific person choices can therefore accompany plural stems without several actual participants Refsing (1986: 150–151); Nakagawa (2024: 160). Narrative fourth-person forms likewise interact with stem number differently across dialects Nakagawa (2024: 160–162). Person paradigms and headcounts must be read separately; the dialect comparisons continue in Chapter 62 (Person Marking and Argument Roles).

References cited in this chapter

Nakagawa (2022) ·Nakagawa (2024) ·Refsing (1986) ·Satō (2008) ·Satō (2025) ·Shibatani (1990)