Chapter 31Person Markers: Contraction, Accent, and Suffix Boundaries

31.1 Person markers and their phonological neighbors

Person markers can affect the vowels, syllables, and accent of the form to which they attach. Southern Hokkaido descriptions distinguish prefix-vowel loss, the conversion of stem-initial /i/ to /y/, and the preservation or relocation of stem accent Nakagawa (2024: 52–53); Satō (2008: 110–111). Suffixes introduce their own patterns at the following boundary.

The effects differ by marker and dialect. This chapter describes the phonological combinations; the grammatical functions and combined-person paradigms are treated in Chapter 62 (Person Marking and Argument Roles).

31.2 Vowel loss in ku= and ci=

Before a stem beginning with /a, e, o, u/, ku= can appear as k=, and ci= as c=. Nakagawa identifies this contraction in Saru, Chitose, and Mukawa, while most other dialects retain the prefix vowel Nakagawa (2024: 52). Satō describes the corresponding Chitose forms Satō (2008: 110–111).

The examples mean “I put something into something” and “our nose,” respectively Nakagawa (2024: 52). The prefix retains its person-marking function when its vowel disappears.

31.3 The separate treatment of initial /i/

The contraction rule excludes /i/. With ku= or e=, an initial /i/ in an open stem syllable can instead become the coda /y/ of the preceding syllable. Nakagawa gives ku=ipe as kúype “I eat” and e=itak as éytak “you speak” Nakagawa (2024: 52). The prefix vowel remains in these forms.

Weakening is blocked when it would create an inadmissible consonant sequence. With inkar “look,” Nakagawa gives kuínkar; with ikka “steal,” kuíkka Nakagawa (2024: 53). Satō gives the same restriction for ku- and e- before inkar Satō (2008: 110–111).

Satō treats ci- before initial /i/ separately. He reports no examples in his description and tentatively predicts a sequence cii Satō (2008: 111, point 4) ‹speculative›. His documented ku- and e- patterns therefore leave this combination unresolved.

31.4 Accent and the order of changes

Nakagawa derives contracted forms by applying accent relocation before vowel loss. The unprefixed verb omáre receives stem-initial accent with ku=, producing ku=ómare. Contraction then gives k=ómare Nakagawa (2024: 52).

The surface initial syllable kó is open. Its accent reflects the order of operations in this account, so it cannot be predicted by applying the ordinary open-first-syllable rule only to the final contracted shape.

A= and eci= preserve the stem's accent: a=omáre and eci=etú Nakagawa (2024: 53). Shiraishi describes the same contrast between person-prefix classes Shiraishi (2022: §5.1). See Chapter 26 (Accent in Compounds and under Affixation and Cliticization) for complex forms more generally.

31.5 The boundary before person suffixes

Satō reports that -an and -as can be preceded by a glottal stop after a consonant other than /n/. His example is sap-an, pronounced [sapʔan]. After /n/, he describes joined pronunciation as usual: san-an is [sanan] Satō (2008: 111, point 9).

The suffix boundary can thus remain audible in some environments while allowing close phonetic joining in others. Shiraishi's account of resyllabification likewise recognizes morphological and accentual conditions on its application Shiraishi (2022: §4.4).

Secondary accent supplies another distinction. Shiraishi gives iˈruska-ˌan “we, including you, are angry,” but no secondary accent in ˈsap-an or ˈcis-an, whose stems are closed monosyllables Shiraishi (2022: §5.1, examples 20–21). The accent and glottal patterns concern different aspects of the suffix boundary.

31.6 Connected speech, dialect forms, and spelling

Lexical person forms also vary by dialect. Nakagawa and Fukazawa contrast Saru and Chitose kani “I” with kuani elsewhere Nakagawa & Fukazawa (2022: §4.1). The full and shortened pronouns have this regional distribution.

Transcriptions also vary. Nakagawa explicitly compares spellings such as komap, k omap, k-omap, and k=omap Nakagawa (2024: 59–61). Contracted forms can therefore appear in Roman-letter texts, with different representations of the boundary. The conventions are discussed in Chapter 15 (Orthographic Standardization and the Word-Division Problem).

31.7 Phonology and the status of person markers

Bugaeva combines accent behavior with separability in classifying person markers. She treats ku= and the other accent-affecting prefixes as affixes, a= and eci= as clitics, and discusses word-like properties of the person suffixes Bugaeva (2012: 472–473). This is a structural analysis of several properties, beyond the vowel alternations alone.

Nakagawa's equals sign has a more limited purpose: it identifies person markers without deciding their word or clitic status, and does not specify a pronunciation boundary Nakagawa (2024: 52, 59–61). The notation can consequently remain the same across markers whose phonological behavior differs.