Substitution and assimilation

Further phonological constraints frequently impact the pronunciation and spelling of consonant clusters, triggered variously by reduplication and assimilation into neighboring sounds. Sound changes, especially in reduplication, are often reflected by a change in spelling. However, processes triggered by affixes as well as adjacent words are not indicated in spelling. In order to inflect, derive, and pronounce Pohnpeian words properly, the order of operations must generally begin with liquid assimilation, followed by nasal assimilation, and end with nasal substitution.^{[6]: 58–64}

First, liquid assimilation is seen most often in reduplication alongside spelling changes. By this process, liquids /l/ and /r/ are assimilated into the following alveolar (coronal) consonant: nur > nunnur ("contract").^[6]: 60

The second process, nasal assimilation, presents two varieties: partial and complete. In partial nasal assimilation, /n/ assimilates with a following stop consonant to produce [mp], [mʷpʷ], [mm], [mʷmʷ], or [ŋk]. For example, the prefix nan- ("in") produces: nanpar, pronounced [nampar] ("trade wind season"); nanpwungara, said [namʷpʷuŋara] ("between them"); and nankep, said [naŋkep] ("inlet"). Partial assimilation also occurs across word boundaries: kilin pwihk is pronounced [kilimʷ pʷiːk]. The allophone of /n/ is written "n" in these cases.^{[6]: 56–7}

In complete nasal assimilation, /n/ assimilates into adjacent liquid consonants to produce /ll/ or /rr/: lin + linenek > lillinenek ("oversexed," spelling change from reduplication); nanrek is said [narrɛk] ("season of plenty"). Complete nasal assimilation also occurs across word boundaries: pahn lingan is said [paːlliŋan] ("will be beautiful").^{[6]: 57, 60}

The third process, nasal substitution, also presents two varieties. Both varieties of nasal substitution affect adjacent consonants of the same type: alveolar (coronal), bilabial, or velar. The first variety is often triggered by reduplication, resulting in spelling changes: sel is reduplicated to sensel ("tired").^{[6]: 58–64}

The second variety of nasal substitution, limited to bilabial and velar consonants, occurs across word and morpheme boundaries: kalap pahn is pronounced as if it were kalam pahn ("always will be"); Soulik kin soupisek is pronounced as if it were souling kin soupisek ("Soulik is [habitually] busy"). This second variety of the nasal substitution process is phonemically more productive than the first: it includes all results possible in the first variety, as well as additional cluster combinations, indicated in green below. Some alveolar pairs produce an intervening vowel, represented as V below. Not all clusters are possible, and not all are assimilative, however.^{[6]: 58–64}

By following the order of operations, reduplication of the word sel ("tired") progresses thus: *selsel > *sessel (liquid assimilation) > sensel (nasal substitution).^[6]: 60 In this case, the same result is achieved by nasal substitution alone.

¹ In the Pohnpeic languages, geminate obstruents are realized as homorganic nasal-obstruent clusters.
² Often before /i/.
³ Before /a/.
⁴ The reflex is *∅ sporadically before PMc *e.

Determiners

Determiners in Pohnpeian may occurs as enclitics which are bound morphemes or independent words and occur in three basic types: demonstrative modifiers, pointing demonstratives, and demonstrative pronouns.^[6]: 143 All of the determiners have a three-way diectic distinction of proximal (near the speaker), medial (near the listener), and distal (away from both the speaker and listener), as well as an emphatic/non-emphatic distinction. Demonstratives are generally, suffixed to or following the last word of a noun phrase.^{[6]: 144–50} Orthographically singular clitics are suffixed to the word, while plurals are written as separate words.^[6]

Pronouns

The relative pronoun me means "one who is" or "which," and is used with adjectives and general verbs: Ih me kehlail (He one strong > He is the strong one); Ih me mwenge (He one eat > He is the one who ate).^[6]: 200

Possessive classifiers

Possessive classifiers are used frequently and differentiate among person, possessum, and honorific usage. Their personal forms appear below:

Further possessive classifiers include: sapwellime (third person honorific), were (vehicles, canoes), nime (drinkable things), imwe (buildings, homes), ulunge (pillows), sapwe (land), kie (things to sleep on), tie (earrings), mware (garlands, titles, names), ipe (covers, sheets), kene (edibles), and seike (catch of fish).^{[5][6]: 184}

Specialized kinship classifiers include: kiseh (relatives), sawi (clan members), rie (sibling in Crow kinship), wahwah (man's sister-relation's children), and toki (persons with whom one has had sexual intercourse).^{[6]: 116, 184}

Honorifics

Honorifics comprise a largely separate vocabulary.^[5]

Numbers and measure words

Numbers normally follow the nouns they count, however they may be pre-posed in certain situations. Numbers and measure words depend on the grammatical class and physical characteristics of the object being counted. The several number systems are grouped by linguists into three sets, reflecting their term for "ten." When naming numbers in order, natives most often use the –u class. Ngoul is an alternate word for "ten" for -pak and -sou classifiers.^{[6]: 127, 135, 141–2}

Higher numerals such as pwiki "hundred", kid "thousand", do not inflect for noun class. The ehd system, above is likewise not class-based.^{[6]: 137–40}

Ordinals are formed with the prefix ka–, pronounced as ke– in certain words.^{[6]: 141–42, 215–218, 318}

Intransitive verbs

Pohnpeian intransitive verbs can be divided into the following types:

There are five verbal prefixes, which appear as bound morphemes: the causative ka-, the negatives sa- and sou-, and two other semantic modifiers ak- and li-.

Ka-, the causative prefix, makes intransitive verbs into transitive ones. It is the most productive prefix, as it is the only that can precede the other four above. It often occurs in conjunction with a reduplicative vowel suffix. For example, with luwak, "be jealous", an adjective:^{[6]: 215–218, 221}

• Liho luwak: That woman is jealous
• Liho kaluwak: That woman was made jealous
• Liho kaluwaka lihet: That woman made the [other] woman jealous
• Pisek, idle
• Soupisek, busy (i.e., un-idle)
• Kasoupisek, to make busy

The majority of intransitive verbs have only a transitive causative form: pweipwei > kapweipwei, "to be stupid." Among verbs where ka- is productive, only adjectives and a few resultative intransitive verbs have both intransitive and transitive causative forms. Though the prefix is productive in many active and resultative verbs, it is not productive with neutral intransitive verbs, nor for a handful of intransitives denoting bodily functions such as "sneeze" (asi), "frown" (lolok), "be full" (tip), and "be smelly" (ingirek). The prefix ka- often has assimilative allophones depending on the stem, for example soai (to tell a tale) becomes koasoia (to talk), dou (to climb) becomes kodoudou (to trace one's ancestry), and rir (to be hidden) becomes kerir (secret sweetheart). As illustrated in these examples, the prefix often causes semantic differentiation, necessitating different constructions for literally causative meanings; karirala, a different form employing ka-, is used to mean "to make hidden."^{[6]: 216–218}

Sa- and sou- negate verbs, however sou- is less productive than sa-, which itself varies in productivity according to regional dialect. The general meaning of sa- appears to be "not," while sou- apparently means "un-," thus:^{[6]: 218–219}

• wehwe, to understand; sawehwe, to not understand
• pwung, correct; sapwung, incorrect
• nsenoh, concerned; sounsenoh, careless (i.e., un-concerned)

Like ka-, sa- displays assimilative allophony: ese, "to know" > sehse, "to not know;" loalekeng, "intelligent" > soaloalekeng, "not intelligent." Only a single example has been found of sa- preceding ka-: the word koasoakoahiek means "inappropriate," deriving from the verb koahiek, "be competent."^[6]: 220

Ak- adds a semantic meaning of demonstration or display when combined with adjectives. When preceded by ka-, it becomes kahk-. Li- generally means "may," or "predisposed, given to" some quality or action.^{[6]: 221–3}

Transitive verbs

Transitive verbs consist of single roots and various suffixes upon modern intransitive verbs. Historically, intransitive verbs probably developed by dropping these transitive suffixes and ablauting.

Some transitive verbs end in -VC on intransitive forms, appearing as unablauted or without reduplication; as intransitives were likely products of final syllable dropping, the endings are rather unpredictable: poad > poadok, "to plant," id > iding, "to make fire," pek > pakad, "to defecate," and dapadap > daper, "to catch."^[6]: 203

Several transitive verbs end in -ih on intransitive roots, sometimes also with vowel changes: malen > mahlenih, "to draw," sel > salih, "to tie," and erier > arih, "to stir, probe." This form is the most productive and is used with loanwords.^{[6]: 204–5} For example: mahlenih, deriving from German mahlen, means "to paint, draw."^[5]: xv

Some transitive verbs ending in short final vowels have intransitive counterparts that lack those endings; again, ablaut and reduplication often differentiate. Examples include langa > lang, "to hang up," doakoa > dok, "to spear," and rese > rasaras, "to sharpen." The short vowel ending -i appears only in -ki.

Transitive verbal suffixes include the perfective -ehr, -ki (which derives verbs from nouns; different from the noun instrumental suffix -ki and short vowel suffix), object pronoun suffixes, and a host of directional suffixes. These include -ehng (towards) and -sang (away, without).^{[6]: 106–7, 222–52, 246–9, 279}