Riktas

General information
Riktas (/ˈɾik.tas/) is a small language family or dialect continuum spoken primarily in the archipelago of ''Sumpa Rammay. The language of trade and governance in the islands, described here as the standard variety, is a dialect most often called Riktas Rammay. ''

The Speakers of Riktas are a culture of human beings who settled in Sumpa Rammay after crossing the ocean along a now-forgotten route from a continental landmass far to the west. Legend has it that they were forced to flee their original homeland by a nameless doom that was spreading across the continent, but ancient taboos against speaking of the time before the migration have led to most of the details being forgotten.

The people usually identify themselves by a variety of local place names, or else refer to themselves simply as nu, "people" or "human beings." When differentiating themselves and their fellow Riktas speakers from the members of other cultural and linguistic groups, particularly the indiginous ʔulili speakers (called Hikrakatrazmi, "those who do not have names"), they most commonly use the term Karikuwus, "people from the west" or "west wind people."

Consonants
Glottalized stops appear most commonly in loanwords, but have been "nativized" as well, mostly through sound symbolism inherited from an 'ulili substratum. The voiced palatal consonants /ʒ/ and /d͡ʒ/ are only marginally phonemic, occurring exclusively in diminuitive constructions.

Riktas also contrasts geminate stops (/pː, bː, tː, dː, tʲː, dʲː, kː, gː/), nasals (/mː, nː, ŋː/), and (more rarely) affricates (/t͡sː, t͡ʃː/) and the lateral approximant /lː/, with their "singleton" forms, as in the minimal pairs himi, "always" vs himmi, "still, yet" and kanan- "to be on the right" vs kannan- "to move or be moved to the right."

Allophony and Consonant Mutation
Many consonants in Riktas occur in regularly alternating pairs, as shown on the chart below.

In noun roots and particles, these phones normally occur only as the "primary" allophone, but may be realied as the corresponding "secondary" allophone under certain circumstanes. /k/ is a notable exception, as /k/ and /h/ frequently contrast in even unmodified noun roots. /t͡s/ and /t͡ʃ/ also contrast with /s/ and /ʃ/ in noun roots, but only in the root-final syllables /t͡si/ and /t͡ʃi/.

There are basically two circumstances under which primary and secondary allophones alternate in noun roots. Firstly, secondary allophones occur in medial consonant clusters following nasal consonants (/m/, /n/, or /ŋ/), fricatives (/ɸ/, /β/, /s/, /z/, /ʃ/, or /ʒ/), and liquids (/l/ or /ɾ/) they take on their "secondary" form. This rule can be considered a relatively simple form of sandhi, as it occurs most freqently at morpheme boundaries when affixes are applied to a stem. Consonants in word-initial and word-final position may also shift based on surrounding words, but these alternations tend to be more irregular and idiosyncratic.

Secondly, in roots that can be inflected as nouns or verbs, primary allophones undergo mutation into their secondary forms when the root is inflected as a verb. Consonants likewise mutate when a noun stem is modified into a verb root with a verbalizing suffix.

Consonants in verb roots behave somewhat less regularly, in that they may occur in primary or secondary form in the "primary" stem form. When verb stems are modified to become "derived" stems, however, root consonants uniformly take on their secondary forms. However, when verb stems take a nominalizing suffix the stem consonants usually shift uniformly to their primary allophone, subject to the "sandhi" rules above.

"Diminuitive" forms of nouns (and, occasionally, verbs), which may be used to express affection, contempt, insignificance, pitifulness, or smallness, depending on the context, involve a different form of consonant mutation. In "diminished" roots and stems, the approximants /l/ and /w/ become the nasals /n/ and /ŋ/ and /s/ and /z/ are palatalized to /ʃ/ and /ʒ/. Plosives and affricates are typically also "hardened," with voiced stops becoming voiceless and voiceless stops becoming glottalized. However, this latter rule is not always consistently applied, and generally applies more regularly to those plosives which most commonly occur in noun roots. /k/ is always hardened, and /g/, /t/, and /tʲ/ are hardened almost as consistenly, but the behavior of /p/, /b/, /d/ and the affricates is more irregular.

Diminuitive nouns are also typically marked with the suffix "-nis.", but most speakers consider this optional when the noun root is audibly modified.

Phonotactics
The canonical syllable structure of Riktas is CV(C), where V is any vowel and C is any consonant, though /h/ never occurs in coda position. As such, vowel sequences do not occur in Riktas, and consonant clusters are short and occur only between vowels.

Syntax
Sentence construction in Riktas is relatively flexible, as arguments are usually explicated through inflection rather than word order. In everyday usage, particularly in cases when nominal morphology leaves room for ambiguity, VSO order is the standard. Stative verbs may modify any of the nouns in a sentence, and are placed immediately before the nouns they modify in a manner similar to adjectives in English. They may only be separated from the corresponding noun by other stative verbs to which they are linked through conjunctions. For instance:

hawimilalu mistsiya yu hisuya huhiyasi ~ "I saw a beautiful and green tree"

Sentences normally require, at minimum, a verb and a subject noun, but simple equational sentences may be formed by pairing nouns in the nominative case without the use of a verb, expressing the meaning (noun a) is (noun b). For instance, yu=rakat Fatrik, "my name is Patrick."

Case
Nouns in Riktas are declined with suffixes marking case, although the nominative case (indicating the subject of a clause) is a null or unmarked morpheme.

The causal and attributive cases are used only infrequently. The former occurs primarily in the construction tin.-wa, "therefore" or "because of that," while the latter is used primarily in conjunction with the locative case marker -tak in classifying groups of people, as in rag-tak-wus, "borderlands people" or "shoreline people."

"Thematic" Suffixes
Thematic suffixes are special inflections that may be affixed to noun stems before case-marking suffixes in order to express various shades of meaning related to the meaning of the stem.


 * The thematic suffix -si becomes -sin when it is followed by a case-marking suffix other than the nominative -Ø. On verbs in the nominative case it is potentially ambiguous with the oblique case marker, and this will typically be clarified by context and word order.

Verbalizing Suffixes
Riktas includes two nominal inflections which may be applied to a noun stem in order to produce a related verb stem. The first is -ra, which functions similarly to the thematic suffix -si ~ -sin, creating verb stems meaning "to have" or "to be characterized by" the modified noun, as in nu-ra, "to contain or be associated with (many) people". The second verbalizing suffix is -ti, which creates verb stems meaning "to be" or "to behave like," as in nu-ti, "to be human."

Numerals
Cardinal Numerals in Riktas are treated as nouns and are inflected for case in agreement with any other noun they are "counting." Ordinal numerals ("first," "second," "third," etc) are treated as stative verbs formed with the verbalizing suffix -ti, as in kana-ti-, "to be first." Finally, numerals have a "distributive" form, corresponding to the English construction "apiece" or "by (number)s," which is marked by reduplication in the same manner as progressive verb stems (see below).

For instance, maka-mpi-ru wita-ru tin.ki kana-si wulu-si, "(I have heard that) they will come (in a while) and bring three birds", becomes maka-mpi-ru wita-ru tin.ki kanana-si wulu-si, "they will come and bring three birds each" or "they will come and bring birds three by three."

Pronouns
Pronouns in Riktas can take one of two forms. Most typically, they occur as enclitics following the first word in a clause, usually a verb. Pronominal enclitics take different forms marking voice and the nominative and oblique cases, the latter often directly following the former as in fafka-ti-ki=lu=mas, "I have heard you."

Pronouns may also take the form of independent words, usually as a way of marking emphasis or providing specificity. Unlike regular nouns, pronouns mark person and number as well as case, and decline irregularly in the oblique case. First person plural pronouns also take on distinct forms depending on whether or not they are intended to include the listener or reader of a statement.


 * Demonstratives (see below), declined in the appropriate case, are used interchangeably as 3rd-person independent pronouns.

Demonstratives
Demonstratives in Riktas (words analogous to the English "this," "that," "these," and "those") differentiate three degrees of distance, indicating whether the described object or person in closer to the speaker, closer to the listener, or distant from both. Like pronouns, demonstratives decline according to number as well as case, and as in English they may be used either pronominally or adnominally. Distal demonstratives are also used to mark definite articles; indefinite articles are never marked in Riktas.


 * ti becomes tin. when it takes a case-marking suffix other than the nominative -Ø. There is potential for ambiguity between the distal singular nominative and the listener-proximal singular oblique, as well as between the nominative and oblique forms of the distal plural. These ambiguities are usually resolved through context and word order.

Possessives
Possession is marked by a proclitic preceding the noun over which possession is claimed, which takes one of three forms based on person.

Verbs
Verbal morphology in Riktas tends to be a good deal more complex than noun declensions, as verbs are inflected (mostly though suffixation) for aspect, mood, voice, and evidentiality. Verb roots themselves can be divided into roughly two types: dynamic verbs, which describe an action, and stative verbs, which describe a stat of being and often function similarly to English adjectives.

Stem Formation
Before affixation, verb stems in Riktas generally take one of two forms derived from a basic lexical "root." In many cases verbal and nominal constructions will share a root form; for instance the root wiru can be analyzed as "ball" or "to play ball," depending on context and inflection. "Primary" verb stems are usually identical to this root form, though some nominal roots can or must be modified (for instance, though the use of the verbalizing suffixes -ra and -ti) in order to produce a grammatically correct verb stem. Prior to inflection, almost all primary stems have exactly two syllables.

Primary verb stems are used immediately preceding a suffix with the form "-CV," such as the reflexive -vu, the reciprocal -mu, and the active-causative -si. In all other verbal constructions, verb roots are modified to form a "derived" stem.

In cases where the primary stem form of a verb ends in a vowel, the corresponding derived stem will usually be identical. When the primary stem ends in a consonant, it will be modified to produce a derived stem according to the following pattern:

There are numerous exceptions to this general pattern which are noted in the lexicon, but a few alternative stem formations are regular enough to merit individual consideration.

Stem-forming suffixes produce derived verb stems when affixed to the corresponding primary stem. Most are idiosyncratic to the stems they modify, but two remain semantically productive. -sV marks the iterative-distributive aspect in dynamic verbs, signifying an action that is performed repeatedly by one subject or which is performed individually by each of a group of subjects. It is differentiated from regular suffixes in that it consistently exhibits harmony with the second vowel of the primary stem it modifies, here represented as "V."

The stem-forming suffix -wi actually replaces the second vowel in the primary stem, as well as the onset consonant in the second syllable if the first syllable of the primary stem has a coda. It signifies the reversal of the dynamic verb stem it modifies; for instance, pan.li-, "to tie," becomes pan.wi, "to untie."

The other major exception to the regular pattern of stem formation involves the continuitive or progressive aspect, which marks dynamic verbs describing actions that are ongoing at the referenced point in time. The continuitive form of a verb stem is formed through partial reduplication of the stem, according to the following pattern:

Primary Inflection
A verb stem may be followed by as many as two suffixes that form its primary inflection, marking voice, comparison, and direction. If none of these suffixes are applied, an unmarked dynamic verb stem denotes the active voice.


 * The passive suffix -n becomes -ni before the singular imperative suffix -y or any suffix of the form "-CCV," such as the perfective -kti

Nominalizing Suffixes
Four suffixes are regularly used in Riktas to transform a verb stem into a noun stem, each of which has distinct semantic properties. The first three are applied only to dynamic verb stems, while -zmi ~ -zmin* may be applied to a dynamic or stative verb stem or a verb stem with the passive suffix -ni.


 * -Ø becomes -n, -mi becomes -min, and -zmi becomes -zmin when the resulting noun stem takes a suffix other than the nominative -Ø.

Secondary Inflection
Secondary inflectional suffixes always follow the primary suffix(es) attached to a verb stem, if any. They mark aspect and mood.


 * -ti becomes -zti when it directly follows a stative verb stem or a stem modified with the passive suffix -ni

Evidential Suffixes
Unless they are conjugated in the imperative mood, all verbs in Riktas must take an evidential suffix following its primary and secondary inflectional suffixes (if any). These denote what kind of evidence has led the speaker to believe that an action took place or that the subject of a stative verb is in a particular state of being, and suggest how certain the speaker is in that belief. In general a noun may only take a single evidential suffix, but in the course of telling a story or reciting some other kind of narrative which the speaker learned from someone else, the "hearsay" suffix -ru may follow a sensory evidential in order to indicate something the protagonist of the narrative has sensed. In the same context, -ru may follow the inferential evidential -ka, which is analyzed as meaning that the speaker is unsure of the veracity events of the story or is extrapolating from what he or she has been told.

Another convention associated with storytelling is the use of the direct evidential -wa in place of the hearsay evidential -ru for emphasis at dramatically appropriate moments in the story. The judicious use of evidentials is considered part of the craft of storytelling, and overuse of the "emphatic" -wa is considered a mark of an amateurish performer.

Because all verb constructions must end with an evidential or imperative suffix, these inflections act as (relatively) unambiguous indicators of lexical class.

Tense
Tense-marking in Riktas is always optional, and may be accomplished paraphrastically in one of two ways. Firstly, a speaker may locate a clause at a particular point in time by including one of several temporal particles in the clause.

When referring to an event taking place at an indefinite point in the (usually near) future, a speaker may instead use one of three auxiliary verbs in conjunction with the main verb in a clause.

wita- means roughly "to wait a while" and is used in much the same way as the English "later," as in wiru-nga-ka=lu wita-ka, "I will go to play ball later," literally "I wait a while and go play ball."

wila is an irregular verb that can be used on its own without suffixation as an imperative meaning "wait!" If paired with a noun or pronoun in the nominative case, however, wila- is treated as a regular noun meaning "to wait," usually with the connotation "to wait until something happens" or "to wait while something happens." In either its imperative or indicative form, wila can be used in conjunction with another verb to indicate that something will happen or enter a particular state of being in the immediate future, or that an ongoing action or state of being will reach its completion or end. The "wait" described by wila is usually understood to be shorter in duration than that described by wita-.

Finally, gisa is an inherently imperative verb similar to the uninflected wila which is used to request that another verb be allowed to reach completion, as in hurkuku-wila=lu gisa, "let me swallow," literally "I am swallowing, let me finish." In conversation, gisa is often used on its own when context renders it obvious to the listener what activity or state the speaker is referring to.

Particles
In addition to the temporal particles described above, Riktas employs particles to serve a variety of other grammatical functions

Negation
Verbs in Riktas can be negated through the application of the proclitic hik=, as in hik=huruk-ni-mila=lu, "I don't (regularly) swallow." On its own, hik is also used as a generic negative particle which can function like the English "no," "don't," or "there is/are none/nothing," depending on the context.

Negative imperatives are conjugated irregularly, being constructed from the uninflected primary verb stem and the proclitic hik=, without any further affixation, as in hik=makki, "don't go (away)!"

Quantification
Indefinite quantities in Riktas are usually identified with one of three particles. tullu can be used as an independent particle, paired with a another noun to mean "many" or used on its own to mean "there are many" or "there is much." It can also be applied as a proclitic before verbal constructions, in which case it becomes tullun= and is more or less synonymous with the comparative suffix -ksi, as in Tullun=mukki-n-du=ta, "he/she runs a lot." In this case tullun= is generally regarded as more emphatic than -ksi, and the two are occasionally combined in the same construction for further emphasis still.

The other quantification particles function identically to tullu in terms of grammar; they are kitis, "few/there are few," and yasdi, "enough/there are enough."

Location
Relative location in Riktas is marked by locative particles, many of which carry a complex set of meanings related to the cosmological beliefs of the inhabitants of Sumpa Rammay. Indeed, the conventional name for the archipelago itself comes from Sumpa, "mountain" or "high island" and the locative particle Rammay, and can be validly translated as "Inside Islands," "Inhabited Islands," or "Known Islands," among other possibilities.

For purposes of navigation and orientation, Riktas defines cardinal directions relative the winds, using the nouns tafdaku, north wind, rammaku, east wind, and kariku, west wind, declined in the locative and/or ablative cases. For instance, "south" or "southward" is expressed as tafdaku-ru, "(away) from the north wind," while "west" or "westward" is kariku-tak, "toward the west wind."

Conjunctions
Riktas uses conjunctions to link clauses in coordinate and subordinate constructions, but relies on inflection to link words. For instance, the English phrase "sheep and horses" would be rendered as va tagarak-duk, literally "sheep with horses."

Adverbial Clauses
Riktas has several particles that are used to modify the meaning of verbs in the manner of an adverbial clause.

Interrogatives
Riktas uses a variety or words to form interrogative clauses, not all of which are particles. In the context of a question, evidential suffixes are reinterpreted as denoting the type of information with which the listener is expected to answer the question. For instance, the hearsay evidential -ru is used when asking if the listener has been told something or heard something said, while the direct evidential -wa is used when asking about something the subject can presently see or has some other direct evidence of.

Dialects and Sister Languages
The rugged islands of Sumpa Rammay host countless local varieties of Riktas sharing various degrees of mutual intelligibility. Riktas Hulildu Karikuru ("western Hulil Riktas"), for instance, originates from roughly the same region as Riktas Rammay, and is differentiated from the standard variety only by a slightly different history of sound changes (a rather extreme example of this would be the Sumpa Rammay demonstrative ti ~ ting, which in Riktas Hulildu Karikuru uniformly occurs as tsingə). On the other end of the spectrum, Riktasnis Tafdakururu ("lesser southern Riktas," known it its speakers as Huku) is almost certainly different enough to qualify as a separate language.