Kijeb pages: <DPL> category=Kijeb </DPL>
| Sohlob languages
|Basic word order:
|Agglutinating, with mild fusion due to sandhi
|Benct Philip Jonsson
Kijeb [cidʑɨb̥] 'the ancient language' is the oldest attested Sohlob language, and regarded by the native grammatical tradition as the mother language of the later languages or dialects Classical Sohlob, Heleb, Kidilib and Linjeb. In fact Kijeb may be the direct ancestor only of Kidilib, while Classical Sohlob, Heleb and Linjeb may be descended from other closely related dialects — dialects which in any case cannot have been very much different, given the great similarity between Classical Sohlob, Heleb and Kidilib, and the relatively long time (above a thousand years) which is estimated to separate Kijeb from the beginning of the Sohlçan grammatical tradition.
It is not known what the speakers of Kijeb called themselves or their language. The name Kijeb is a Classical Sohlob term, although its phonetic form has evidently been influenced by its unrecorded Kidilib cognate. It is in any case not likely that the speakers of Kijeb called their own contemporary language *kidipa 'ancient language'.
Kijeb is attested in the form of mostly short formal religious, commemorative, dedicative and funerary inscriptions on stone and metal written in a syllabic script. There is evidence that there existed other writings in Kijeb language and script, since Sohlçan grammarians mention finds of pottery with Kijeb inscriptions in ancient graves. It seems clear however that with few exceptions the preserved Kijeb texts derive from the area where Kidilib was spoken in classical times.
Despite the written attestation Kijeb is largely a reconstructed language, since the Kijeb syllabary is partly defective, ignoring the distinction between voiced and voiceless stops, partly between stops and fricatives, and in the oldest inscriptions also between palatalized and unpalatalized consonants, and between velar and labiovelar consonants, and since the vocabulary of the inscriptions is limited. The reconstruction of Kijeb was made by Sohlçan priest grammarians striving to understand ancient religious inscriptions in the hope of improving their theology and the wording of their liturgical texts. Fortunately some of them got hooked on the study of language and language history in its own right and delved deeper in the history of their languages than those who were merely motivated by religious conservatism and purism.
- 1 Introduction
- 2 Kijeb (Proto-Sohlob) grammar
- 2.1 Sound system
- 2.2 Root structure
- 2.3 Morphology
- 2.3.1 Nominal morphology
- 2.3.2 Adverbs
- 2.3.3 Pronouns
- 2.3.4 Verbs
- 2.3.5 Infinite verb forms
- 2.4 Syntax
Kijeb (Proto-Sohlob) grammar
|n ny [nʲ]
The absence of **xw is somewhat remarkable. Some Sohlçan grammarians note that Heleb and Kidilib x was in fact redundantly labialized, and suppose that this was the case also in Kijeb. Others propose a Pre-Kijeb merger with f; there are however no homonyms with f to support this theory.
The cardinal palatal nasal ŋy [ɲ] does not contrast with the dental/alveolar ny [nʲ] in roots, but appears in inflexion and derivation. This is shown by the fact that its reflex in the daughter languages is [j], e.g. tataŋya > CS: taday 'my own', while the reflex of ny is [ɲ], brunya > CS: brony [brɒɲ] 'hatchet'.
An y after a consonant or cluster is realized as palatalization of the consonant, or all consonants throughout the cluster.
Assimilation and dissimilation
There were some assimilatory and dissimilatory changes that probably applied synchronically in Kijeb:
|before voiced stops and nasals
kitbyu > *xidbyu > CS: xefd, K: sijd, L: xirz;
|between a vowel and an i
|between an i and an a or u
|between a vowel and an u
|between an u and an a or i
|before p, b, f, m
|before k, g, x, ŋ
|m, n, ŋ
|between a t, d or s and a consonant other than r, w, y
|between an s and a consonant other than r, w, y
|before voiced stops
|sasbinta > *sazbinta > CS: hæzbænd, K: hejdehn, L: hezbent.
The *rr > dr and *dd > zd rules conspire to produce *drr > zdr.
The very restricted range of consonant clusters in Kijeb arose through sweeping sandhi changes. Most of these changes were assimilatory, like the pervasive regressive voicing assimilation, but some were dissimilatory, like the change of geminate nasals into nasal + homorganic voiced stop. In the table below the first ('leftmost') consonants of input clusters are shown in the leftmost column, while the second ('rightmost') consonants of input clusters are shown in the top row. Outputs are shown in the intersecting cells. Where convenient mergers have been indicated with merged cells, and unchanged outputs are indicated with bold type. Character values are as IPA, except that kw, gw, ñw are probably true labiovelars, y is /j/ and consonant + y are realized as palatalized consonants. For ky, gy, ŋy, xy this probably means they were realized as [c, ɟ, ɲ, ç].
Labiovelars as the first element of an underlying cluster behave like velars, and palatalized sounds as the first element of an underlying cluster behave like their non-palatalized counterparts. Second elements retain their palatalization, and non-palatalized second elements don't get palatalized by palatalized first elements, i.e. palatalization is subject to regressive assimilation, just as voice.
Maybe I could have a distinction so that labiovelars behave like labials before non-palatalized sounds but like velars before palatalized sounds, but I doubt there is any sound phonetic basis for that. Maybe a dialectal difference between labiovelar first element > velar and labiovelar first element > labial?
I'm not sure what happens to underlying combinations of palatalized velar + labial (*ky+p etc.), velar + palatalized labial (*k+py etc.) and palatalized velar + palatalized labial (*ky+py etc.), but my hunch is that at least the two latter become palatalized velars (ky etc.)
- Whenever final **y or **w arose it colored a preceding **a to **i or u before disappearing itself. **-uy and **-iw both resulted in *i, possibly via [y].
- Pre-Kijeb *ww possibly gives b in Linjeb. Alternatively or simultaneously it is possible that underlying *uwwV/*wwuV give ugV/Vgu while *uwV/*Vwu give ubV/Vbu.
A Kijeb nominal or verbal root has the following structure:
- (s) (C) (r) (y) V (D) C (r) (y)
- slots in parentheses are optional
- V is any vowel (a, i, u)
- C is any consonant
- D is any dental (t/d, n, s, r)
There are some restrictions on possible phoneme sequences, including:
- Geminates do not occur. Thus e.g. if there is an s in a C slot there can be no s in the preceding s or D slot, resulting in an *ss, and similarly no *rr, *ww, *yy, *tt. *dd, *nn.
- n + a nasal does not occur (no *nn, *nm, *nŋ).
- Labial + w does not occur.
- None of *yi, *iy, *wu, *uw can occur.
- Palatalization or y may occur either in the consonant(s) before the V or in the consonant(s) after the V or not at all, but not both before and after the V.
- A cluster is either palatalized throughout or not at all.
- In nominal and verbal roots
- the same consonant may not occur twice,
- no two stops or fricatives (other than s) at the same point of articulation may occur within a single root,
- no two nasals may occur within a single root.
- The nominal stems mama 'mother' and papa 'father' are the only known exceptions to these restrictions among nominal and verbal roots. In pronoun and determiner stems, on the other hand, shapes like tat, kak, nan, sas are allowed and even favored. (There is also the numeral stem sas 'one', but this may be a determiner in origin.) It may be that these stems were originally CV stems that became reduplicated.
- Roots consisting only of VC are extremely rare.
- Roots with a heavy cluster both before and after the V are rare.
In order to function as a stem such a root must be followed by a vowel. In nouns (including adjectives) this second vowel is a random extension, while in verbs it is an inflectional morpheme.
(In the daughter languages the accusative is derived from *-si and the dative from *-tu.)
It is not certain that the inanimate plural accusative and dative were distinguished from the singular. In any case the Kijeb syllabary could not distinguish word-final -rs, -rt from word-final -s, -t, while in the animate plural the diacritic for syllable-final -n could be added to CVs and CVt characters to indicate word-final -ns, -nt.
The ergative of the daughter languages derives from *-rya and is certainly derived from the instrumental.
Local adverbs are formed by adding local case endings to the three deictic roots proximal *i 'this', medial *a 'that' and distal *u 'yon'.
Kijeb pronouns have several unique morphological characteristics. Most basic pronoun roots have the shape CV. Such a root cannot stand on its own, but attaches clitically to a noun or verb stem. A clitic pronoun functions as a possessive marker when attached to a noun and as a person agreement marker when attached to a verb.
Free standing pronouns are emphatic, and are formed by reduplicating the CV root.
Plural pronouns are formed either suppletively or by infixation of the plural marker -n or -r between the syllables of the reduplicated CV root. For this reason plural emphatic pronouns take simple case markers -s, -t etc., thus the accusative of anda is andas rather than **anans or **andans.
Personal and demonstrative pronouns
First and second person pronouns
Third person clitic pronouns
|'they (mixed male and female)'
|'it (not living'
Third person emphatic pronouns
These are formed by compounding three deictic roots — proximal *i 'this', medial *a 'that' and distal *u 'yon' — with the third person pronominal stems. These deictic roots do not occur independently, although they are also used in local adverbs, and neither are there any non-clitic third person pronouns not including these deictic stems. The plural forms are most readily explained as the result of reduplication of the root followed by syncope of the vowel of the second syllable and then dissimilation of the resulting geminate according to regular Kijeb sandhi, e.g.:
- *ina > *inana > *in'na > inda, *aru > *aruru > *ar'ru > adru.
There are no special clitic forms for possessive pronouns. The clitics listed above have a possessive meaning when attached to nouns.
First and second person emphatic
Third person emphatic
|Proximal 'of this'
|Medial 'of that'
|Distal 'of yon'
Reflexive pronouns are formed by adding the clitic -ŋi to the personal and demonstrative pronouns, e.g. tataŋi 'myself' anaŋi 'himself'. There is also a possessive version -ŋya, e.g. tataŋya 'my own', anaŋya 'his own'. These clitics can also be attached to nouns, e.g. snupuŋi 'the man himself', snupuŋya 'the man's own'.
'Each other' is -ŋgi and 'each other's' is -ŋgya.
|'of what?, of which?'
Note that the clitic singular possessive is the same for animate and inanimate. There is no trace of them ever having been differentiated.
Indefinite pronouns are formed by adding the clitic -fu to the interrogatives: gwigifu 'someone' gugufu 'something'. This clitic can also be attached to nouns, e.g. snupufu 'some man'.
This entire section needs reworking in view of my improved understanding of inverse marking and hierarchical alignment. In the meantime the following is largely obsolete!
Finite verb forms
- root - aspect - (direct-inverse) - (tense/mood/voice) - subject
Items in parentheses meqn that that marker may be empty. In those cases the absence of a marker functions as a marker in itself.
The inverse marker -s-/-z- signals that the subject of the verb is lower on the animacy hierarchy than the direct object. A verb that is not inverse is said to be direct, but there is no overt directness marker in Kijeb.
Tense and mood and voice
There are two tenses, present and past, and three moods, realis and irrealis and imperative. However marking of tense and mood overlap, so that any given verb form is marked either for present tense -n-/-na-, past tense -r-/-ri- or irrealis mood -t-/-s-/-z-.
A verb form marked for tense cannot be marked for mood, but is always realis, and a verb form marked for irrealis mood cannot be marked for tense, but can function as any tense. The irrealis performs such functions as future, subjunctive, optative and hortative.
The only voice distinction in Kijeb is between non-imperative and imperative. The imperative is signalled by the absence of inverseness and tense/mood markers, and that the only person marking is an animate plural marker -n if the injunction is adressed to a group. An imperative can only be addressed to animate adressees. To express a 'command' or wish concerning inanimates the irrealis must be used.
Infinite verb forms
The infinitive or gerund
There is also an infinitive or gerund stem formed by adding -a- to the verb root. By taking different case endings this stem can express various infinite functions. The infinitive/gerund marker itself is in complementary distribution with the aspect markers, but by adding case endings aspect relative to the main verb may be expressed.
Participles are verb forms inflected for aspect and tense/mood, but which take nominal number and case endings rather than person suffixes. Pariciples are used chiefly to form relative clauses, as there are no relative pronouns in Kijeb, and each sentence may have only one finite verb. Examples are sabura 'Having given', pantina 'hitting' and basnura 'gone out'.
(Scattered notes so far)
Word order is SOV, with everything that entails.
The 'rational' plural gender is used for mixed masculine/feminine groups. The reason it isn't called 'human' is that it is used for gods, ghosts, demons and the like as well as for humans. Besides I'm not sure yet if the Sohlçan are human!
There is no genitive case. The possessive construction is like this:
- giwri-t tyazgu-na-n king-DAT elephant-his-PL 'The king's elephants'.
Negation is made with the auxiliary paxya 'not do', e.g. sazba paxintaŋka 'I don't see them'. This is quite possibly the main use of the infinitive/gerund!
Do predicate nominals/adjectives require a copular verb or not?
- If not will the verb una only mean 'be' or also 'exist', or not exist! ☺ It will probably be needed for other aspects/tenses than imperfective present.
- Note the conflict with the pronoun una!
- Will ufa 'not be' be in complementary or partially alternating distribution with paxya 'not do'?