səmnêk reference grammar

phonology

consonants

• the stops /p t k/ are unreleased as syllable codas; additionally, VOT (i.e. aspiration and voicedness) is not distinguished in this position.
• the voiced plosives /b d/ may be pronounced as implosives [ɓ ɗ]. these phonemes are somewhat rare but can occur in both native and loaned forms.
• /w/ is realized as [β̞], with no velar constriction.
• /h ɸ/ pattern as if they were /x xʷ/; that is, phonologically, they behave as though they were debuccalized from a previous (labio)velar realization.
• when adjacent to /u/, /w/ can dissimilate to [ꞵ] or even [v]. similarly, /j/ adjacent to /i/ may dissimilate to [ʝ].

vowels

syllable structure

the maximal syllable is of the form CɾVəF (see below); minimally, only the vowel nucleus V is required. in addition, there exist minor syllables of the form (C)ə(N) which cannot stand alone -- phonologically, they must be part of a longer word containing at least one major (i.e. non-minor) syllable.

• C stands for any consonant
• V stands for any vowel except for /ə/
• F stands for one of /m n ŋ p t k w j/

… subject to the following constraints:

• the only allowed onset clusters are /pɾ tɾ kɾ/.
• /ŋʷ kʰʷ kʷ ɸ/ cannot occur before /u/.
• /ij uw/ are forbidden within a syllable.
• the codas of syllables containing a medial /ə/ cannot be /w j/.

stress

səmnêk has phonemic lexical stress: exactly one major syllable in roots is considered stressed, which can be the difference between otherwise identical words. stressed syllables have a longer nuclear vowel and are pronounced with a higher pitch.

romanization

the standard romanization is listed below, with differences from the IPA highlighted. customarily, the glottal stop ⟨ʔ⟩ is left unwritten word-initially; no word makes such a distinction. stressed syllables are written with an acute ⟨◌́⟩, and stressed ⟨ê ô⟩ are written ⟨ế ố⟩. stress is always indicated on the nuclear vowel, and is obligatorily marked on all words containing more than one major syllable.

there is no standard romanization when restricted to ASCII, but possibilities include using ⟨v⟩ or ⟨r⟩ for /ə/, ⟨x⟩ or ⟨'⟩ for /ʔ/, and dropping all diacritics.

nominal morphosyntax

medium / noun class

the medium, the "substrate" that a noun is tied to, affects particles and pronouns used in reference to it. the three mediums distinguished in səmnêk are physical (glossed as IRL), online (ONL), and alternate (ALT). the physical and online mediums apply to entities tied to the physical world and online spaces, respectively. the alternative medium represents any world outside of the two specified above -- a fictional universe, a plural system's headspace / inner world, a parallel universe to this one, etc. səmnêk culture places emphasis on treating different mediums as equally as legitimate as each other, so the choice of class is purely a descriptive attribute and reflects no judgement of morality, worth, or correctness.

despite the name, the class of a given noun isn't tied to the particular phonological word used as the head of the noun phrase -- rather, it's a property of the referent thereof. a speaker living in brazil, for instance, may refer to the language toki pona as a phíəsa (language) in the online medium, as most of the communication in and about it happens online, and use the same term phíəsa in the physical medium to refer to brazilian portuguese.

some nouns are heavily biased in meaning towards one particular classifier; this doesn't mean that it's incorrect to use the "wrong" classifier with them -- just that there should be a good reason to do so.

pronouns

pronouns in səmnêk don't make a singular/plural distinction, but instead distinguish singulative/collective. the singulative (SGV) is used for an individual or individuals, while the collective (COL) is used for group(s) with any degree of cohesion. additionally, each pronoun differs based on the noun class of the referent; this ties into the cultural concept that one's irl self and online self could potentially be drastically different people.

demonstratives

correlatives

classifiers

noun phrases, especially when numerically quantified, are prefixed with a classifier (CLF) indicating the medium in which the noun exists. a bare classifier sans numeral can also be taken as an indefinite marker. the same noun or noun phrase can take different classifiers depending on context: for example, cou phidíəp might mean "a manual", while nêu phidíəp could mean "a reference website" like this one.

the choice of classifier doesn't necessarily correspond to the medium in which the noun exists: for example, someone meeting an internet friend in person may still refer to them as nêu lêəng (ONL internet_friend), despite being physically present with each other; a physical model of a weapon from a story may be ngai əcímte. in fact, it's incorrect to refer to it as cou əcímte, as it isn't a weapon in this world.

verbal morphosyntax

səmnêk verbal inflection alone is fairly simple, almost nonexistent, but the verbal complexes themselves can get quite complex due to productive use of serial verb constructions. optionally, verbs can also conjugate based on the medium in which the event itself occurred, forming a kind of evidential, and this marking doesn't necessarily need to agree with any of the involved arguments' classes.

serial verbs

negation

the negation particle nhê can appear anywhere within a sequence of serial verbs, applying to the immediately following verb.

medium / evidential

verbal complexes can optionally display a prefixed clitic indicating the medium of the event, known as the evidential (EVID). this prefix precedes and applies to the entire verbal complex, even if it consists of a complex multi-word structure such as a serial construction.

TAM

copular constructions

numerals

bytes

sentence structure

oblique arguments