Chapter 30Citation Forms and Morphophonological Alternations

30.1 Forms in isolation and in combination

A word's dictionary form does not specify every pronunciation it takes in a phrase. Final consonants can change before the next sound, vowels can contract at a prefix boundary, and a consonant can become the onset of a following syllable. Possessive morphology introduces another kind of difference between forms Nakagawa (2024: 30, 36–40, 226–227); Shiraishi (2022: §§3.2, 4.3–4.5).

These patterns have separate conditions. A phonetic release after /r/, a contracted person prefix, and a possessed noun are not interchangeable instances of one stem alternation. The following comparisons distinguish the segmental, morphological, and lexical factors involved.

30.2 Final /r/ in different environments

Final /r/ often has a short vowel-like release. Its quality can resemble a neighboring vowel and can vary within one speaker's speech. Nakagawa describes this variation in kor “have,” kar “make,” pirka “good,” and arki “come, plural” Nakagawa (2024: 30). The release does not by itself establish a separate final vowel phoneme.

Before a consonant, final /r/ can undergo sandhi. For example, ku=kor tar “my carrying strap” can be pronounced kukot tar, and ku=kor nima “my wooden bowl” as kukon nima Nakagawa (2024: 36). These are conditioned changes within a combination. Their wider distribution is described in Chapter 28 (Assimilation and Cluster Simplification: Coda /r/, Nasals, and Clusters).

30.3 Final /t/ before /i/

Combining a /t/-final form with a following /i/ can produce /ci/. Nakagawa gives mat “woman” plus ikor “treasure” as macikor, and nit “handle” plus -ihi as nicihi Nakagawa (2024: 36). He distinguishes this regular change from the variable consonant alternations discussed on the following pages Nakagawa (2024: 38).

Shiraishi treats the pattern together with resyllabification and also gives rekut-i becoming rekuci “his or her throat” Shiraishi (2022: §4.4, example 9). The original final consonant precedes the vowel of the next syllable in the combined pronunciation.

The transcript provides a contextual occurrence of the combined form. The phonological analysis follows the descriptions above. Related possessive formations are treated in Chapter 45 (Affiliative Suffix Patterns).

30.4 Shortened person prefixes

In the southern varieties described by Nakagawa, ku= and ci= lose their vowel before stems beginning with /a, e, o, u/. His examples include ku=ómare becoming k=ómare “I put something into something,” and ci=étu becoming c=étu “our nose” Nakagawa (2024: 52). Satō gives the corresponding pattern in Chitose Satō (2008: 110–111).

Stem-initial /i/ behaves differently. With ku= or e=, it may become /y/, as in kúype “I eat” from ku=ipe. A closed initial stem syllable blocks that weakening in examples such as kuínkar “I look” Nakagawa (2024: 52–53). Prefix-vowel loss and stem-vowel weakening are described separately in Chapter 31 (Person Markers: Contraction, Accent, and Suffix Boundaries).

30.5 Lexical forms and proposed etymologies

A lexical word can preserve the outcome of an older change without participating in a productive modern alternation. For the robe name attus, Tamura's dictionary gives a tentative segmentation at-tus, explicitly marked uncertain Tamura (1996: s.v. attus) ‹speculative›.

Sakhalin comparison. Hattori's comparative entry gives Saru áttus beside Sakhalin ahrus Hattori (1964: item 566). The two regional forms differ in the consonant sequence following the initial vowel.

30.6 Koré: accent and lexical history

Nakagawa relates koré “give” to kor “have” plus causative -e. He explains its second-syllable accent through diminished awareness of that derivation and the ordinary accent pattern of the whole word Nakagawa (2024: 53) ‹speculative›. Shiraishi reports the same kind of lexicalization account Shiraishi (2022: §5.2).

Tamura lists korpare, analyzed as kor-pa-re, as the plural counterpart of kore Tamura (1996: s.v. korpare). The plural form retains the causative suffix after -pa.

The 2026-09-09 snapshot of 30 distinct ILCAA transcripts contains 8,088 aligned units. It gives the following written-form counts ILCAA Ainu materials (1976) ‹corpus-confirmed›:

The count matches complete written forms after normalizing Unicode spelling and letter case. An equals sign can separate a person marker from a counted form; adjacent letters prevent a match. Duplicate documents with identical Ainu text and publisher URL count once. Unaccented kore supplies no accent measurement, and the absence of written kóre is not evidence that a pronunciation never occurs. The counts describe this collection's spelling practice.

30.7 Possessed noun forms

Possessed forms involve grammatical morphology with both regular and lexically selected patterns. Nakagawa gives tek “hand” with teke(he), and sik “eye” with siki(hi). Parentheses mark an optional longer ending Nakagawa (2024: 226–227).

The final vowel of the base restricts some choices, but does not uniquely determine every possessive suffix. The suffix belongs to a morphological formation, rather than being a general vowel inserted whenever the noun combines with another word Nakagawa (2024: 227). The full patterns are described in Chapter 44 (Concept, Affiliative, and Personal Forms) and Chapter 45 (Affiliative Suffix Patterns).

30.8 Comparing the processes

Older records also differ in spelling; see Chapter 14 (Historical Orthographies and the Batchelor Tradition).