Kilda Kelen

Kilda Kelen
Spoken in:	Kamchatka Peninsula, Northeast Siberia
Timeline/Universe:	Possibly the League of Lost Languages
Total speakers:	At least 50,000.
Genealogical classification:	disputed Altaic  Tungusic   North Tungusic    Kïlda or Altaic  Mongolic   Kitan    Kïlda
Basic word order:	SOV AdjN GenN RelHead
Morphological type:	Agglutinative
Morphosyntactic alignment:	Accusative
Created by:
Kuroda	1996-

The language of the Kïldamnï (gentile noun; individually, Kïldaï) is likely a member of the League of Lost Languages that is spoken today on the Kamchatka Peninsula in northeastern Siberia. Foreign scholars universally consider Kïlda to be an Altaic language -- the family consisting of the Turkic, Mongolian, and Tungusic language groups, and according to some, Korean or Japanese. However, contemporary native Kïldamnï scholarship rejects the "Altaic Hypothesis" and considers these three (or five) groups to be related only through prolonged contact and mixing. In any event, it would be safe to say that the Kïlda language is closely related to both Tungusic and Mongolic languages, whether by descent or affiliation. Today, arguments on the genetic roots of Kïlda fall into one of two camps: Tungusic with heavy adstratum influence from Para-Mongolic or Pre-Classical Mongolian, vs. Para-Mongolic with heavy substratum influence and later borrowing from North Tungusic languages, specifically Evenki (or "Tungus") and Even (or "Lamut"). The position that Kïlda is a separate branch of Altaic was popular earlier in the 20th century but is now held only by a small, crackpot fringe.

Phonology

Kïlda has a regular system of so-called front/back vowel harmony, in which lexical stems determine the vowel quality of their affixes. "Front" vowels are /ə i u o/ (in standard Latin orthography {e i ü ö}) while "back" vowels are /a ɪ ʊ ɔ/ (Latin orthography: {a ï u o}). The latter series is described, by different researchers, as "tense", "pharyngialized", or "retracted tongue root (RTR, -ATR)". There is also limited "rounding harmony" where the vowels /a ə/ appear as /ɔ o/ (respectively) in suffixes when attached to stems containing only /ɔ o/. Vowel length is additionally distinguished (though not in many orthographies), which also affects vowel quality (e.g., /ə əː/ = [ə ɛ], /ʊ ʊː/ = [ʊ ɯ]) and stress allocation. Two additional phonemes are recognized by the Kïldamnï as diphthongs, in Latin orthography {ei, ê} (/eʲ, ɛj/) and {ai} (/aj, æʲ/).

Consonantal phonology is fairly unexceptional. The two series of obstruents are distinguished primarily by voicing: /b d ʤ g/ {b d j g} are voiced, while /p t ʧ k/ {p t c k} are voiceless and strongly aspirated in most positions. While /k/ has velar and uvular allophones, conditioned by front vs. back vowel harmony (respectively), these are not phonemic and are ignored in native orthographies (though not in most foreign transcriptions). /p/ is [ɸ] (orthographically {f}) before back vowels (and in some cases /ɪ/), though this is irregular due to many loanwords and ideophones - so some analyses present /ɸ/ as a separate phoneme.

Like most genetically and areally related languages, there is a phonemic palatal nasal /ɲ/ {ñ} or {ni-}. Unlike other Tungusic languages and neighbors (but like Manchu and Mongolic and Turkic languages), the velar nasal /ŋ/ {ng} cannot occur before a vowel. The hissing sibilant /s/ {s} is always pronounced "palatalized" (actually, retracted and grooved) before high front vowels (/i/ and /ı/) and thus merges with /ʃ/ {sh} or {š}. The phoneme /r/ {r} does not occur word-initially in native words or in loanwords from before the mid-19th century: /ɔrɔm/ "rum" (1820s) but /rɔkınrɔl/ "rock and roll" (1970s). Foreign descriptions cast /l/ {l} as a lateral fricative, voiceless in most but not all positions ([ɬ] and [ɮ]), but native accounts persist in calling it a lateral approximant. Likewise, according to Kilda sources /w/ {w, v} is not a labiovelar, but a bilabial approximant or fricative (realized as [β̞, β]), while many foreign scholars describe it as a velar approximant pronounced with lip compression, [ɰ]. The phoneme given here as /h/ has allophones ranging from voiceless velar and uvular fricatives [x, χ] to voiceless pharyngeal or epiglottal fricatives [ħ, ʜ] (identification is disputed) to voiceless and voiced glottal fricatives [h, ɦ] - depending on vowel harmony, syllable position, and dialect or idolect.

Syllable structure is generally simple and of (C)V(C) form, and the language permits no internal clusters of more than two consonant phonemes. Word-initial or word-final consonant clusters are very rare even in loanwords. (E.g., /smiki/ "wandering tattler (a bird)", /tɔsk/ "riverbed sand".) By default, primary stress falls on the first syllable, but is attracted to syllables with long vowels.

Consonants
	Bilabial		Dental-Alveolar		Post-alv.		Palatal		Velar		Glottal
Nasal		m		n				ɲ		ŋ
Plosive	p	b	t	d					k	g
Fricative			s		ʃ						h
Affricate					tʃ	dʒ
Approximants		w		l	r			j

Vowels
	+FRONT		-FRONT
	+ATR	-ATR	+ATR	-ATR
+HIGH	ɪ	i	ʊ	u
-HIGH	a	ə	ɔ	o

Morphology and Syntax

Grammatically, Kïlda is highly agglutinative and almost exclusively suffixing. It has basic SOV word order, accusative syntax, and a wealth of non-finite verb forms, many tense and aspect distinctions, and a fairly elaborate system of local cases. In all these respects it is a quite typical North Tungusic language. Though it has a slightly reduced and simplified inflectional morphology compared to its closest relatives, most morphological and syntactic constructions can be directly matched to corresponding forms in Even and/or Evenki.

Inflectional morphology for verbal elements recognizes categories of finiteness, mood or modality, aspect, tense, evidentiality, person and number (of subject only), and switch-reference (in non-finite verbs only). There is a rich variety of verbal nouns (or "participles") which can serve as finite and non-finite verbs, attributive modifiers, and nominals. In addition there are a number of important periphrastic constructions to express certain TAM combinations (e.g., contact causation, pluperfect and future perfect) and in polite and honorific language.

Finite endings can loosely be defined as those verbal inflections which occur only in the heads of verb phrases serving predicative roles in the sentence, and which do not permit subordination to another verb phrase. However, not all such VP heads take finite endings; participial and converbal forms can and often do occur in that role. In fact, in Kilda finite verbs proper are largely limited to imperatives and a number of modal-epistemic forms. The finite endings are followed by pronominal suffixes (except as noted below).

Finite Verb Endings
Name	Suffix
Near Future	-jai-	Near or immediate future tense
Inferential Future	-mna-	More distant future tense, with less certainty about the action or state actually happening; also used to state probabilities or likelihoods in general or abstract times, or in customary behavior
Subjunctive	-mca-	Consequence of counterfactual conditionals
Near Probabilative	-na-	Possible event or action at a time close to the present
Distant Probabilative	-rka-	As above, but for events or actions in the more remote past
Mirative	-jug-	Actions or states which turn out to be the case, counter to expectation or in a narrative twist; usually with a pluperfect tense-aspect meaning.
Admonitive	-jik	Warning or reminding about the verbal action: "look out for..., be sure to..., beware of...". Only with second person reference.
Imperative	-kar (sg.), -karsu (pl.), -hun (polite)	Orders and commands to direct addressees.
Prescriptive	-tugin (sg.), -tuktin (pl.)	Instructions and remote commands to non-present third persons; in some cases, to direct addressees but concerning actions to take in the future or in a distant location.
Voluntative	-su (sg.), -ya (pl. incl.), -gat (pl. excl.)	Instructions, exhortations, and expressions of volition about the speaker or the speaker's group.

Nonfinite verb forms function as attributes, subordinated predicates, and finite or independent predicates. Nonfinite verbs never take modal markers, and rarely occur with certain categories of voice. They fall into three general categories: participles, or verbal nouns; inflecting converbs; and non-inflecting converbs.

As attributives, participles always agree in number with the head of their NP, but not in person or case. As predicates, they require the periphrastic verb bi- ("be"), except third person nonfuture participial predicates.

Participles (Verbal Nouns/Nominalized Verbs)
Name	Suffix
Nonfuture	-ra-, -da-, -sa-, -na- (pred.), -ri-, -ri-, -si-, -di- (attrib./dependent)	Complicated range of meanings and functions; very widely used -- possibly the most frequent verb form in the language. Can refer to the recent past (of single perfective events or actions), simultaneity or present tense (of prolonged actions), imperfective past (of a small number of activity verbs); regardless, if it refers to any situation in the past, it must be one that was directly observed or experienced by the speaker. It thus has an evidential force as well. It also functions as the coreferential equivalent of the Purposive converb -da- (see below), and is sometimes referred to as "the infinitive form". (It is also the traditional citation form of verbs in Kilda dictionaries.) The forms with the low front vowels occur when used in predicative roles; those with high front vowels when the participle serves as a modifier or nominal. Verb stems fall into four classes, distinguished by the form of their nonfuture participle; they are not phonologically or semantically predictable. The -ra-/-ri- form is by far the most common.
Perfect	-ca-, -ka-	Describes events of longer duration or further in the past than the Nonfuture participle; also those events or states which the speaker did not directly witness or experience -- i.e., with inferential or quotative evidential force. Despite the name "perfect", it does not always imply completed action. The alternation of -ca- and -ka- is again lexically determined; the former is more frequent.
Hypothetical	-jiŋga-	Possible or future indefinite events; to the degree it implies uncertainty, it conveys categorical probability or likelihood and not the speaker's subjective opinion.
Past Iterative	-jaŋki-	Actions that occurred many times in the past or states that frequently or on many occasions were true. (But not habitual or natural conditions.)
Contact Posterior	-ltak-	Events that are just about to happen, on the cusp of transpiring. Rather rare, but does occur in both predicative and attributive roles.
Optative	-(ŋ)gati-	Describes circumstances which the speaker would like to happen or would consider favorably. Usually only as a predicate.
Necessitative	-nna-	Actions or events that are necessary or required from an impersonal or universal perspective, or that are logically necessary.
Debitive	-mcin-	Like the Necessitative, but those situations that are personally or subjectively obligatory

Converbs have only a predicative function; they never serve in attributive or substantive roles. As the heads of VPs, they are limited to non-finite clauses. They come in two varieties: inflecting (switch-reference, SR) and non-inflecting (coreferential, CR). The former are used where the subject of the converb is different from the subject of the main VP; the latter, where the subject is the same in both the subordinate and main VPs.

Converbs
Name	Suffix
Conditional	-mi (CR), -raki-, -daki-, -saki-, -naki- (SR)	Covers both temporal and causal conditions: "if she Xes, (then she Ys)" and "when she Xes, (then she Ys)".
Imperfective	-mnak (CR), -ŋsi- (SR)	Past events that were simultaneous with the action or state described by the main verb. "While she was Xing (she was Ying)."
Perfective	-psa (CR), -cala- (SR)	Events or states that were completed or achieved before the main verb.
Terminative-Resultative	-(ŋ)kan (CR only)	Actions or states up until which the main verb continues, or at the point of which the main verb begins.
Negated	-r (CR only)	Serves as the periphrastic complement of the auxiliary negative verb.
Purposive	-ri, -ri, -si, -di (CR), -da- (SR)	The reason, goal, or purported outcome of the main verb.

Inflectional morphology for nominals is less complicated, but includes categories of number (singular vs. plural, with a fair number of suppletive forms to indicate plurality; also collectives and an obsolete/nonproductive dual form), possessor, inalienable possession, and case (around ten in number, depending on whether one categorizes certain local case forms as proper cases or as clitics).

Noun Case System
Name	Suffix
Nominative	-0	Subject; nonreferential/indefinite objects; citation form
Accusative	-wa	Specific direct object; distances of travel; durations of time; reference of communication
Designative	-ga +[personal possessive suffix]	Direct object with beneficiary function assigned to possessor; can function as subject with some intransitive verbs with benefactive function; designation of direct object in double-accusative verb phrases
Instrumental	-ji/-ic	Tool or means of action, function of use, mode of transport, material or source of creation, source of emotional reaction. (-ji with stems ending in vowels or nasal consonants; -ic with stems ending in non-nasal consonants)
Sociative	-g(a)li	Along with, together with, in the company of, bringing along (usu. with non-animate nouns)
Comitative	-nan / -ñun	Together with, along with (usu. with animate nouns)
Ablative	-dok	Motion from source; time after which; general source; object of comparison; source of taking; (raw) material of creation or making; location of action of certain verbs; partitives of numbers
Allative/Directional	-t(a)ki	Motion to point/location at which action is completed, but does not begin; motion in direction of which but not up to; in many possession constructions; object of some verbs; addressee of communication; dative of giving; object of active perception verbs; source of emotional reactions when not controlled or intentional
Essive/Dative	-du	Location at rest; location of action/event/topic with animate subjects; end point of motion with emphasis on final position (rather than motion); causee of some derived causatives/adversatives; dative of purpose or benefactive; ending usually applied to spatial postpositions; time of completed/general/discrete action; subject and object complements
Prolative	-li	Motion through, within, or across; route of process or event; punctual future times, time past or through which event will occur; prices or values of exchange; reference of communication verbs
Illative	-la	Motion to point with emphasis on course of motion; location within which action takes place (atelic verbs only); time up to which; used to create possession predicates
Terminative	-gla	Motion past or along edge of feature; benefactive of destination or work; time up to which; substitution ("in place of...", "instead of...")

There is a large residual class of word forms including "adverbs" and/or "adverbials", "clitics", "conjunctions", "particles", etc. - and about as many different schemes of classification as there are scholars of the language.

Lexicon

See Neo-Khitanese_Lexicon

Genetic Affiliations and History

Despite its morphosyntactic affinity with the North Tungusic languages, the Tungusic lexicon of Kïlda shows signs of being more archaic than either Evenki or Even. Most noticeable is its preservation of initial *p- as p- or f-, which has gone > h- > 0- in the other North Tungusic languages. E.g., PTg *pesin "handle" > Evenki, Even, Negidal, Oroch hesin, Udihe hehi vs. Ulcha, Orok, Nanai pesi(n), Manchu fesin, but Kilda pesi(n); or before back vowels, PTg *po:si- "glowing ember" > Evenki ho:sin, Negidal xosinca "hot coal" vs. Ulcha and Orok posi, Nanai posi(n) "spark, burning coal", Manchu foson "sunlight, sun ray, glow of fire", but Kilda fo:sïn "ember, glowing coal". Likewise, PTg *ü in the first syllable of a word typically becomes i in North Tungusic but u in South Tungusic and also in Kïlda: PTg *tügde "rain" > Evenki/Negidal/Oroch/Udighe tigde, Solon tiki- "to rain", Even tīd vs. Ulcha/Nanai/Kilda tugde, Orok tugde/tugje, Manchu tuhe- id. Moreover, Kilda has many lexemes that are found only in the southern Tungusic languages and absent entirely from the northern ones.

These two factors give Kilda the superficial appearance of being more closely related to South (or "Amuric") Tungusic languages than the North Tungusic languages. All Tungusic etymologies of Kilda lexemes are, moreover, complicated by a sizable number of relatively loanwords from Even and Evenki into Kilda, in some cases supplanting an earlier or inherited lexeme.

Additionally, there is a large quantity of what are either inherited Para-Mongolic terms or early borrowings from Preclassical or Middle Mongolian, depending on individual scholars' interpretations. These, combined with quantities of "high culture" vocabulary items borrowed from Old Turkic, are nowadays generally accepted as evidence in support of the traditional Kïldamnï belief that they are descendants of the Khitans who ruled northern China, Manchuria, and parts of modern Mongolia under the Liao dynasty (907-1125 CE). While of restricted use, there even exists a sizable body of Indian and Iranian loanwords.

The Kïlda language has furthermore absorbed very large numbers of lexical items from languages indigenous to Kamchatka (Kurile Ainu, Southern and Eastern "dialects" of Itelmen) and from languages neighboring Kamchatka: the Koryak and Alutor languages, and later from Aleut and even Pacific Yupik -- confusingly now called Alutiiq. (Kïlda thus has been influenced by "Alutor", "Aleut", and "Alutiiq", actually three different languages belonging to three different language families or branches thereof.) There are also a smaller number of foreign loans from the early modern period (primarily Portuguese, French, English, Japanese, and Russian, but also Chinese and Chinook Jargon). Perhaps as much as 1/4 or even 1/3 of the total lexicon is of non-Altaic or non-Tungusic origin. In the 19th and 20th century, it proved much more conservative towards foreign influence than in the past, and has (relatively) few "global" or "international" items of vocabulary taken from English or Russian.

Writing Systems

The Kïldamnï in Kamchatka used forms of the Khitan logosyllabic writing system until the early 1800s, when Latin orthographies were adopted. The original Khitan "Large Script" and "Small Script" writing systems of the Liao dynasty seem to have died out by approximately 1300 and are documented in Kamchatka only by a few short and poorly-preserved inscriptions in stone. They did, however, give rise to a third related script first attested around the same time; this writing system consisted of 1200-1600 graphemes and during its period of use (c. 1325-1825) became gradually more and more regularized and rationalized.

Since the 1820s the official and generally used orthrography for Kïlda has been one of several Latinate systems (also successively modified and modernized over the decades). The language has also been written in the Russian Cyrillic alphabet (primarily by Orthodox missionaries, priests, and converts beginning in the early 1700s), Japanese katakana (during 1941-1945), and to a very limited extent in a "runic" syllabary devised at the turn of the twentieth century that was inspired by the recently-discovered Old Turkic "runic" script from the Orkhon and Yenesei valleys in Central Asia.

Ethnographic Background Information

Kildamni Ethnographic Questionnaire