This post will detail some of the irregular inflection found in Namari. Like any natural language, not all words in Namari inflect regularly. This is mainly prominent in verbs, due to the nominal declension system being relatively new, but some nouns do show irregularlities.
I mentioned in my post on verbs that there are several categories of verbs that can be considered “irregular”. Two of those categories only contain one verb each, and are exceptional with regards to conjugation. These categories are represented by kū (“to come”) and sū (“to do”):
- Imperfective: ko-, se-
- Infinitive: ki, shi
- Conclusive: kū, sū
- Attributive: kuyo, suyo
- Perfective: kue, sue
- Imperative: koyo, seyo
- Participle: kite, shite
- Polite: kebei, säbei
The above two verbs have the following aspect forms:
- Imperfective: kyoi, sëi
- Perfect: kitei, shitei
- Progressive: kitoi, shitoi
- Perfect Progressive: kiteitoi, shiteitoi
The other two categories are the r-irregular and the monograde categories. The r-irregular category only consists of three verbs, of which two have irregularly formed negatives. The term “r-irregular” is a misnomer, as all three r-irregular verb stems actually end in -y (historically, the forms were written with r-row kana). The monograde category is a small, closed class of verbs with monomoraic stems ending in /e/ or /i/. A typical monomoraic verb, mī (“to see”), is given below:
- Imperfective: mi-
- Infinitive: mi
- Conclusive: mī
- Attributive: miyo/moyo
- Perfective: mie/mue
- Imperative: miyo
- Participle: mite
- Polite: miyabei
The alternate forms listed are dialectal. One may notice that the attributive and imperative are identical; while this may theoretically cause confusion with regard to kakari-musubi clauses, often context and agreement are enough to distinguish between the two (in dialects with agreement, miyo is exclusively third person singular in kakari-musubi clauses and second person in imperative clauses).
This verb has the following aspect forms:
- Imperfective: miyoi
- Perfect: mitei
- Progressive: mitoi
- Perfect Progressive: miteitoi
The r-irregular category is a special category in that it is actually defective with respect to tense-aspect combinations; the imperfective aspect can only occur with the past tense in this category, and in all three cases it is irregular. Additionally, the conclusive is marked with -i in two of the three verbs, and omitted entirely (along with the stem-final -y) in the third.
The following lists the forms of the r-irregular verbs, ai (inanimate existential), oi (animate existential) and ya (copula):
- Imperfective: aya-, oya-, yaya-
- Infinitive: ai, oi, yai
- Conclusive: ai, oi, ya
- Attributive: ayo, oyo, yayo
- Perfective: ae, oe, yae
- Imperative: *ae, oe, yae
- Participle: atte, otte, yatte
- Polite: ayabei, oyabei, –
The copula does not have a proper polite form; instead the inanimate existential is used (regardless of animacy), and the noun to which the copula would normally be attached instead has the clitic -ya.
These three verbs have the following negative and aspect forms:
- Negative aorist: nān, oyan, yān
- Imperfect: aita, oita, yauta
- Negative imperfect: ända, önda, yånda
- Perfect: attei, ottei, yattei
The verb oi happens to have a morphological progressive (ottoi); however, it is identical to the aorist or imperfective depending on tense.
Now looking to the regular verb classes, the quinquegrade and the two bigrade classes. The bigrade verbs are completely regular with one exception (despite being non-productive), and will not be mentioned here any further (save the exception). I also will not give most of the forms of the quinquegrade verbs, as I’ve mentioned them previously (although I will mention that if an imperfective form would give *ua, an epenthetic /w/ is added to give uwa). However, there are some irregularities to keep in mind.
In principle, regular verbs form their participles by combining the infinitive with -te. However, many quinquegrade verbs actually alter this form depending on the stem-final consonant:
- Null + -te → *-ite → -ute
- ki + -te → *-kite → -ite
- gi + -te → *-gite → -ide
- shi + -te → -shite (→ -ite)
- chi + -te → *-chite → -tte
- n(i) + -te → *-nite → -nde
- bi + -te → *-bite → -nde
- mi + -te → *-mite → -nde
- (y)i + -te → *ite → -tte
- rri + -te → *rrite → -rte (this is rare)
While the above list may seem incomplete, quinquegrade verbs ending in other consonants do not exist (all former p-stem verbs became null-stem quinquegrade verbs, while all former r-stem verbs either became y-stem or, very occasionally, rr-stem). One kk-stem verb exists (akku, an alternate form of aiku “to walk”), but it is considered to have an irregular participle.
Even then, some verbs do not follow these rules. The following are a sample of irregular participles:
- eku/yuku (“to go”) → itte
- iu (“to say”) → yutte
- aiku/akku (“to walk”) → aiete (the aie actually represents an overlong vowel)
Technically, I could mention the polite/imperfective forms here, but apparent irregularities in quinquegrade verbs are the result of the romanisation used (-tu becomes -chabei in one romanisation system, showing apparent irregularity, but another gives -tyabei, which is a completely regular form).
Finally, there is one bigrade verb I said was irregular. This is tabu (“to eat”, lower bigrade), which is otherwise regular, but has a suppletive passive kuwayu (also lower bigrade). This is actually the passive of a now-obsolete verb kū (“to eat”, quinquegrade).
That will be it for verbs. There isn’t much to say about nouns, but some irregularities include:
- kami (“god, spirit”) → kanna (genitive form)
- kami (ditto) → kan (short vocative form)
- mezu (“water”) → mena (genitive form)
- Sino-Namari nouns which derived from words ending in a velar nasal will induce rendaku in the non-focus comitative and ablative cases, even though case markers typically do not undergo rendaku.
- Some nouns ending in a diphthong change their final -e or -o to -in or -un respectively in the dative (as well as the allative, locative and abessive, whose suffixes begin with a moraic nasal), to reflect a closer retention of the diphthong pronunciation than usual (the addition of the nasal splits the diphthong regardless of the existence of this behaviour). Typically this behaviour is found in loanwords, but some native names (such as Pimae, cognate with Japanese Himari) also display this behaviour (in this case becoming Pimain).
That will be all.