In this section we will deal with the following topics:

  1. Common word orders
  2. Basic intonation
  3. Use of the cases.
  4. Uses of -ni
  5. Accidental (caka-) verbs
  6. Adjectival expressions
  7. Adverbial and sentential modifiers
  8. List of common prepositions and conjunctions
  9. Complex sentences
  10. Colloquial vs. "formal" usage

1. Word Order

1.1 Sentences. The basic word order in Kash sentences is subject - verb - (indirect object, if present) - direct object when the subject and objects are all nouns. Intransitive verbs, of course, will have no DO, but may have an IO. Verbs are marked with the personal prefixes to agree with their subject, whether noun or pronoun; subject pronouns are omitted, except for emphasis. Pronominal objects, however, precede the verb (in the order DO - (IO) - verb) in the short or clitic form, or in the full form when emphasized. (Note that the neuter singular object pronoun yu is usually omitted unless ambiguity would result.) The subject is always in the nominative case, IO in the dative; the DO may be dative or accusative (in a few cases, genitive). See §2 Use of the Cases.

This basic order may undergo various transformations, to place emphasis or focus on some particular part of the sentence. Principally these involve moving the DO (or IO) noun to the front of the sentence; the subject noun either remains in immediate pre-verbal position, or may be post-posed. (In colloquial speech, it is very common for a simple subject noun + verb sentence to be inverted-- verb + subject noun-- with little change in meaning or emphasis.) Topicalization also involves fronting some element, which is then set off with a change in intonation. (Basic intonation patterns will be discussed below in §1.4.)

1.1a. Relative clauses, indirect speech and sentence complements. These are all introduced by the particle (a)re and follow normal word order. In a relative clause, all pronominal forms are retained for semantic and grammatical clarity, since re is merely a connector, not a "relative pronoun", indicating only that its clause modifies the preceding noun in some way. As mentioned above, the neuter obj.pronoun yu may be omitted. Some examples:

Indirect speech: çenji yakota re mina yaporo ambal lalu 'Shenji says (that) Mina is going to have a party'
- erek yavirap re çenji mende yacosa 'Erek is certain (that) Shenji has gone'

Sentence complement: ama çenjiyi yamelo re yayukar kañulin 'Shenji's father wants him to become a carpenter'
- mileka-mandekasa lire lanimim lavi yapendoye 'we argued about which (of us) was wealthier'-- lire is a contraction of liri+are

Relative clauses: laluni re (yu) yambal mina, niya ilongo 'the party that mina gave was a lot of fun' (lit., the party RE (it) she-offered Mina...;)-- the two clauses could also be reversed
- taya sim toloti, re ne misorom punamim 'this is Mr. Toloti, to whom we're selling our house' (lit., ...RE to-him we-sell our-house)
- kaçuma ye re çenji yacetre yañen, iye tiçami 'the woman (that/whom) Shenji married is my cousin' (lit., ...RE Shenji married with-her, she (is) my cousin-- the lengthy subject noun phrase has been topicalized here)

1.1b. Questions. Yes-no questions, in correct usage, are simply declarative sentences with the question particle aka preposed, and rising intonation-- çenji yatimbat minaye 'Sheni knows Mina' > aka çenji yatimbat minaye? 'does Shenji know Mina?' (there are variants without aka in colloquial speech).

Information questions are introduced with a question word and have normal sentence intonation; in short sentences, the subject is frequently postposed: kandri çenji yamepu ~kandri yamepu çenji? 'what is Shenji doing? what did Shenji do?'; kariye havele toyeni? 'to whom did you give the money?' (Note that with fronted DO-- kariye (toyeni/yu) havele? 'to whom did you give it/the money?'-- "correct" DO-IO-verb order is violated by the requirement that the question word come first.)

1.1c. "Passive", impersonal and subjectless sentences; physical sensations.

i. There is no true passive voice in Kash; the stylistic equivalent (to avoid a repetitious sequence of SVO sentences) simply involves fronting the direct object (or the indirect object), e.g. tekangaya hangel yawurisa etengi yu 'Prof. Hangel wrote that book' > etengi yu yawurisa tekangaya hangel-- where the latter could be translated into English with a passive verb, 'that book was written by Prof. Hangel'. Sentences with a fronted IO cause problems if there is a noun subject: where should it go-- between the IO and the verb? after the verb+DO? Compare:

(?)minaye (~ne) amani yavele toye, or--
(?)minaye (~ne) yavele toye amani her father gave Mina (~her) money = Mina (She) was given money by her father.

Both sentences are felt to be awkward, and though written work shows examples of both, they are rare. The construction can be avoided by topicalizing the subject, though in that case a "passive" translation might not be appropriate. There is, of course, no problem with a pronoun subject--minaye yavele toye Mina was given money by him/her.

ii. Nor is there any problem with agentless or subjectless passive-type sentences, e.g. "Mina was given money", "The book was published in 748". In these cases, conventionally, the DO or IO is fronted, and the verb is in the 3rd pers. plural:

minaye ivele toye or toye ivele minaye lit., they gave Mina money
etengi yu irundengayi pehan 748 lit., they published the book in 748

iii. Weather verbs. (1) A few nouns that describe conditions require yale ..., e.g yale lero, yale yanga, yale sucup, yale tandi piyal it's sunny, it's windy, it's muddy, it's full moon. (2) If the condition can be described with an adjective/verb, there are two ways of expressing it: (a) as a verb, with 3rd pers. sing.-- yafasan, yaronek it's hot, it's cold. Or (b) with lero(ni), e.g. leroni (ya)fasan, leroni (ya)ronek (3) A few nouns may be used in all three ways, especially (leroni,) yaripa ~yale ripa, (leroni,) yaçoteru ~yale çoteru it's rainy/raining, it's foggy. The sentences with leroni might also be translated "it was a hot day,...a foggy day".

iv. For the English construction with "one", Kash prefers to use a personal pronoun--1st or 2nd pers., singular or plural, as best fits the situation. Thus, ta makayapi (~hakayapi etc.) kar pila! I (you)(=one) hardly know what to think! However, there is a way to express "one"-- with subject ka&ccedl; and bare verb stem-- kaç ta kayapi kar pila! kaç kuna çañuke... one (lit., person) hardly knows what to think! one might hypothesize....-- but it is considered evasive, condescending, affected and high-flown.

1.2a. Nouns and modifiers. Quantifiers-- numbers, and words like yuno 'all', pando 'much, many, a lot of', tapat 'some (of), any (of) (contrastive)'-- generally precede the noun they modify, though in certain cases they may follow. All other modifiers-- demonstratives, possessives, adjectivals, (noun) genitives and relative clauses-- follow their noun. Correct order before the noun is quantifier - number - noun; correct order after the noun is noun - demonstrative - adjective(s) - genitive-- demonstratives agree in gender and number but not case, adjectives are not declined at all. Presence of a genitive often leads to rephrasing, since speakers feel that possessed - possessor should occur as close together as possible (see further below). Some examples:

- yuno nim ana içangi 'all five children are sick'
- tapat anala inolit, liyani ipanip 'some (of the) children were reading, others were playing'
- andum ye 'the/that little girl'; andum virik 'a pretty little girl'; andum ye virik 'that pretty little girl'; in e.g. the genitive: andumi ye, andumi ye virik; plural andumila nila, andumila (nila) virik, dative andumile nila virik.
- puna yu 'that house'; punaç niç 'those houses'; punaç niç marok 'those old houses'; (case forms) gen. punayi yu, dat. punaçe niç marok etc. (Colloquially, many speakers ignore number agreement, saying e.g. andumila ye, punaçe yu marok, but it is somewhat substandard.)

With a genitive: ñaki çenjiyi 'Shenji's car'; ñaki velu çenjiyi 'Shenji's new car'; ñaki yu velu çenjiyi 'that new car of Shenji's' (many would prefer relativized ñaki (yu) çenjiyi re yavelu)

(Formerly, and still in poetry and "elegant" prose, the case and number endings could be attached to the last word in the noun phrase-- so, (acc.) andum yen or andum virikin for andumun ye, andumun virik, or puna niç, puna yu marokoçe for punaç niç, punaçe niç marok.)

Adjectival modifiers may themselves be modified with adverbials like niya 'very', uti 'rather, somewhat', lavi 'more', omban 'less', imik 'least', which are all postposed-- puna marok niya 'a very old house', payo mondes uti 'a rather shoddy cape'; also superlative krat 'most', which is preposed-- lero krat fasan 'the hottest day', payoni krat tevas 'his most elegant cape'-- though most speakers tend to relativize adjectivals with a preposed modifier.

lavi and omban may also be used as adjectivally, as in mipita añulin lavi 'we need more lumber'. Comparative sentences will be discussed in more detail in §6.2.

It should be noted that there is a growing tendency in colloquial speech to prepose all adverbial modifiers-- puna lavi velu, kaçut niya marok, payo uti mondes 'a newer house, a very old man, a rather shoddy cape'. However, such usage is still frowned upon by educated speakers, even though there is a reason for the change: in predicative sentences all modifiers precede the adjective form (e.g. puna yu lavi yavelu 'that house is newer') and this requirement has simply been extended to descriptive usage as well. See §6.

1.2b. No more than two adjectival modifiers may follow a noun, optionally connected with i 'and'-- puna marok (i) vanat 'an old white house'. When more than two modifiers occur, one or more of them must be rephrased as a relative clause or conjoined sentence, e.g. "he bought that old white stone house": yatraka puna yu cangar vanat i/re yamarok lit. he bought that white stone house and/that it is old. Presence of a genitive noun, also, frequently requires this rephrasing, since a genitive noun counts as a modifier.

Any adjectival modifier, in fact, may be phrased verbally as a relative clause, with little change in meaning or emphasis: thus, puna (yu) re yavanat 'house that is white' is equivalent to puna (yu) vanat; similarly, puna re yamarok (i) yavanat. This is also the preferred construction when the adjective has a preposed modifier like krat 'most' or sambat 'so..., such a...' -- so andi re krat yavital 'the tallest boy' (though NOUN krat ADJ is acceptable), puna re sambat yamarok 'such an old house' (NOUN sambat ADJ may occur colloquially but is considered substandard).

In certain cases a relative clause is required--

1. when the noun has more than two modifiers (not including a demonstrative), then the excess must be relativized; it does not matter which, although the relativized modifier can be interpreted as more important or more emphasized; so "that beautiful old white house" is puna yu marok (i) vanat re yavirik, or puna yu virik (i) vanat re yamarok, etc.-- or even, by choice, puna yu vanat re yamarok (i) yavirik.

2. when the modifier is a verbal base with some modifying material of its own; it is possible to say pumbik kumbor 'a collapsed barn' but pumbik re yakumbor pehan cosa 'barn that collapsed last year'; or sende çindi 'a/the spoken language' but sende re içindi keandolila 'the language spoken by Andolians', lit. '...that (it) they-speak Andolians'. The only exception is mepu 'make ~made' (and a few other verbs involving methods of manufacture) followed by a single noun denoting material, location, etc., as nimu mepu hece 'a pot made of iron', ñakiç volu holunda 'cars assembled in Holunda'; ...mepu pawu-ta '...made in Bau Da').

3. Relativization is also preferred for inchoative and causative modifiers-- better pumbik re yaçukumbor 'a barn (that is) about to collapse, in danger of collapsing' or pumbik re irungumbor 'a barn that is/was demolished ~pulled down', lit. '...that they demolished', than pumbik çukumbor or pumbik rungumbor (though these may occur, they are considered awkward and substandard.).

4. when the modifier is a phrase of some sort-- kaçut re mesa niyoñi 'a one-armed man, a man with one arm', lit., '...that (it is) one, his arm'; kaçut re aceni (ya)matra 'a man with a broken leg', lit., '...that his leg (it is) broken'; puna re nim atelni 'a five-room house, house with five rooms', lit., '...that five its room(s)'.

5. and usually when a genitive noun is present: for although puna virik çenjiyi 'Shenji's beautiful house' is acceptable, puna çenjiyi re yavirik is preferred by many. Relativization is required if more than two modifiers are present (and the genitive form counts as a modifier) puna virik çenjiyi re yamarok 'Shenji's beautiful old house' is acceptable, but puna çenjiyi re yavirik yamarok would be preferred.

1.2c. Nouns modified by other nouns. Such modifiers typically refer to the material of which something is made, its origin, brand-name, or some other characteristic. For example: nimu hece 'iron pot', puna cangar 'stone house', poren andoli 'Andolian wine', kofa -añol velu- 'New Star beer', prale toloti '(Mr.) Toloti's kiosk'. These may also be relativized, but then need some explanatory word: nimu re mepu hece, puna re mepu (or ningar 'built') cangar, poren re mepu andoli, kofa re mepu (or arañi 'its name') -añol velu-, prale re yapeña tolotiyi ~ re apeñani ('property of') toloti.

1.2d. Lengthy noun phrases. Kash speakers consider it awkward when too many modifiers or a lengthy relative clause come between a noun subject and its verb, or between an indirect and direct object; likewise in the case of a fronted direct or indirect object with many modifiers. Rephrasing of some sort is felt to be necessary, and the most common strategy is to convert the lengthy clause into a topic. (Of course, speakers vary as to what they consider "too lengthy"; most will allow two adjectives or a relativized adjective; some allow an adjective, but disallow anything longer. Compare:

Generally allowed: kaçut yu marok me yacura liri eçevani 'that old man told me about his trip'. Or: yavele anaye virik kreki 'she gave the pretty child a cookie'.

Disallowed: kaçut yu marok re ne manunji ri ñakrece me yacura.... 'that old man I met on the train told me....'; this must be rephrased, either as a topic-- kaçut.....ñakrece, me yacura....-- or as a conjoined sentence-- manunji kaçute marok ri ñakrece (i) me yacura... 'I met an old man on the train and he told me....'

Disallowed: yavele anaye virik re haloni çisu kreki 'she gave the pretty child in the red sarong a cookie'; the dative phrase would most likely be topicalized-- ana...çisu, ne yavele kreki-- or rephrased yavele anaye virik kreki, ine re haloni çisu lit. 'she gave the pretty child a cookie-- to her/the one in the red sarong'; or somewhat awkwardly fronted-- anaye...çisu yavele kreki (this last could be translated "the pretty child....was given a cookie by her" and would more likely occur in prose than in speech.)

1.3 Quantifiers with singular/plural nouns. As was discussed in the Morphology ("Numbers") section, numbers and quantifiers occur before the noun they modify (with one or two exceptions). The noun is not pluralized after a number (but a plural verb is required), except in the special cases mentioned-- ro kaçutila 'two of the men (out of several)' and kaçutila ro 'the two men (out of several)' etc. Some of the quantifiers, however, may take a plural noun, particularly yuno 'all', pando 'much, many', and angunjo 'half':

- yuno lero 'all day, the entire day' (in this case, yuno may be postposed for emphasis-- lero yuno(ni) 'the whole day, the entire day'; yuno leroç 'all (of the) days' (e.g. in a given time period; similarly here, leroç yuno 'every one of the days') -- and note formulaic ri yuno leroç amaraç 'for ever and ever, for evermore', lit. 'in all days (and) ages'. Similarly, yuno kumus ~kumus yuno 'all of a/the cake, an/the entire cake' vs. yuno kumuç 'all the cakes'
- pando lopa, ecutni yatevas 'many lopas have fine/delicate fur'-- a generic statement; versus pando lopala, ecutni yatevas 'many (of the) lopas [that we see here/have referred to] have fine fur'. Even clearer with a possessive or demonstrative-- pando lopala(ni, -mi etc.)... ~pando lopala nila.... Similarly, pando ñaki ivanat 'many cars are white' (generically) vs. pando ñakiç ivanat 'many of the cars (of a group) are white'
- angunjo lero 'half a day, half of the day'-- just 10 hours; or about 3-4 if working hours are meant; angunjo lelero 'half-days'-- i.e. 10-hour periods, or 3-4 working hours per day; versus angunjo leroç 'half of the days' (in a given time period); similarly with other time words-- angunjo aro 'half an hour', angunjo aro-aro 'half of every hour, by half-hour periods, every half-hour', angunjo aroç 'half the hours (of the day or other time period)'. Or, angunjo kumus 'half of a/the cake', angunjo pipinal kumus 'half of every cake', angunjo kumuç 'half of the cakes'.

2. Basic intonation

The variety of Kash being described here is a "syllable-timed" language, that is, every syllable is pronounced distinctly with approximately equal length. There is little rise and fall in the voice.

2.1 Declarative sentences and information questions are spoken with a generally level contour, with very slight rises in pitch on stressed syllables, up to the last stressed syllable in the sentence (the main stress), which rises more noticeably, followed by a drop-- this may be symbolized 2.......2^..1. (The rise on the main stress is less than that heard in English-- usually symbolized 2....31-- which would sound emphatic in Kash.) The demonstrative (i)yu carries main stress when it is actually functioning as a demonstrative, but not when it merely indicates definiteness; compare

- 2matraka puna yu2^1 'I bought that house'-- vs. 2matraka pu2^na yu1 'I bought the house'

Other examples of normal sentence intonation:

- 2veleka çenji2^ye1 'give it to Shenji'; 2yu ne ve2^leka1 'give it to him!'
- 2riyena mi2^na1? 'where is Mina?'
- 2kari ara2^ndi1? 'what (lit., who) is your name?'

2.2 Yes-no questions have a rising intonation, quite similar to that used in English; after the main stress, the voice continues to rise to the end of the question-- approximately, 2...2^..3. Thus, 2aka mende haçindi minje2^re3? 'have you spoken to the deputy?', or colloquial, 2haco2^saka 3? 'are you going?'

2.3 Emphatic intonation is characterized by increased stress (higher pitch) on the emphasized word, and can be symbolized as 2.....3..1 (or even 2....4..1 for extra emphasis, great surprise or dismay). In print, it is usually indicated by "italic" or bold type; in script by underlining. Typically, word order remains the same as in a normal declarative sentence, but emphasized subjects may be moved to the end, or objects may be fronted. Emphasized pronouns use their full form, and many speakers shift the stress to the final syllable. A few examples:

- mi3na2 yafilan po2^ren tayu1 'Mina [not someone else] brought this wine' (also possible: 2yafilan poren tayu mi3na1)

- 2mina yafilan poren ta3yu1 'Mina brought this wine' [not that one] (or, 2 poren ta3yu2 yafilan mi2^na1 -- this might be translated "this wine was brought by Mina".)

- 2te mama3ndumbo1! 'I only lent it to you!' [I didn't give it!]

-2kandri hat3 tikas1? ~ 2kandri tikas hat31? 'what did you see?' (With an emphasized subject, the verb's personal prefix is usually omitted.)

- 2iyu etengi ma3mi1 ~ 2 iyu etengi mamí31 'that's my book!'

- i3ne2 mave2^le1 ~ 2iné3 2mave2^le1 'I gave it to him'

2.4 Topic intonation is frequently heard. Topicalized and declarative sentences do not differ greatly in meaning, though the focus may differ. A noun topic is always fronted, and in the nominative case, regardless of its role in the main sentence, which will be indicated, if necessary, by a pronominal referent. In print, topics are usually indicated by a comma, in speech with the pattern main-stress...level or slight rise + a slight break (symbolized "...2^..2/,"), followed by the "comment" portion of the sentence with regular intonation. (Note: either level or slight rise may be used for topics; counting/listing intonation is quite similar, but uses the slight rise only, and has no break.) Compare:

Declarative: 2mina yafilan po2^reñi1 'Mina brought the wine'

Topicalized: mi2^na2/, 2 yafilan pore2^ñi1 '(as for) Mina, she brought the wine'; with the object as topic 2pore2^ñi2/, 2mina (yu) yafi2^lan1 '(as for) the wine, Mina brought it'.

Declarative: lopa ya yakici minan 'that lopa bit Mina' (same intonation as in the declarative sentence above)
Topicalized: mina, lopa ya yen yakici '(as for) Mina, that lopa bit her' (~mina, yen yakici lopa ya) (same intonation as in the topicalized sentence above)

Counting/listing: fiyan yale? eya,...ke2^li2/, sor2^/, fa2^nu1. 'How many are there? Let's see...six, seven, eight'. 2mimaru mina2^ye2/, 2çenji2^ye2/, 2(i) 2ere2^ke1 'we invited Mina, Shenji and Erek'.

3. Use of the cases

3.1 Nominative. This is the base form of the noun; there is no special case ending. It is used (a) as the subject of a sentence, and as the complement of the copular verbs ale 'to be' (whether present or not), ciyon 'to seem' and yukar 'to become'; and in topic position; (b) with some inflected titles, and in apposition with them; (c) in the so-called ni-construction, which has several uses: (i)as the (logical) object of compound prepositions and (ii) as "possessor" in the genitive construction noun+ni noun(nom.), used mostly with inanimate nouns, but in some cases with animates; (d) as direct object is certain verbal constructions; and (e) as subject (in certain cases for humans, all cases for non-human) of accidental caka-verb forms.

3.1a. The nominative is used as subject of all verbs (but see §3.1e), and as complement of the copular verbs ale 'to be', yukar 'to become', ciyon 'to seem, appear to be'; for example--

- (As subject/complement) çenji yale kandumbra 'Shenji is a doctor'. The copula ale is not usually omitted in sentences of the type NOUN is NOUN, but if it is, the subject then becomes a topic (see next).
- (As topic-- a very common construction) cenji, yale kandumbra '(As for) Shenji, he's a doctor'-- this sentence differs from the one above only in intonation, discussed above. Similarly with ale deleted-- çenji, kandumbra-- with topic intonation. Further: in çenji, ta ne matikassa 'As for Shenji, I didn't see him'-- "Shenji" is nominative, even though he is logically the object of tikas, as indicated by ne in the main clause.
- (Always without copula) riyena çenji? 'Where is Shenji?' lani çenji? 'Which one is Shenji?' arañi (aran+ni) çenji 'His name is Shenji'.
- (As complement) yamelo yukar kandumbra 'He wants to become a doctor'. sininji ya yaciyon kacivar 'That young man seems/appears to be a student'.

3.1b. The nominative is also used in the case of title + name where the title inflects for case, but the name does not:

- kandumbra alika (colloq. kandu alika) 'Doctor Alika', acc. kandumbran/kandun alika
- minjer pamban 'Deputy Pamban', genitive minjeri pamban

If, however, the name is in appostion with the title (in writing, set off with commas), then both nouns take case markers-- matikas kandumbraye, alikaye 'I saw the doctor, Alika;' maçindi pambane, minjere 'I spoke to Pamban, the Deputy'. Similarly with a pronoun, "we doctors..." mila kandumbrala..., (gen.) mili kandumbrali.

However, the common titles sim, lumi 'Mr., Ms/Mrs.' do not inflect for case, but the name following does-- (dat.) sim tolotiye 'to Mr. Toloti'. In direct address in very polite speech (overly polite or obsequious, in the opinion of many), the alternate forms simbi, lumbi (and other titles, too) are sometimes used in place of a 2nd person pronoun, and do inflect: maturo, ta matikassa lumbiye 'Excuse me, I didn't see you ~Madame'; puna minjeri niya yatevas 'Your ~Deputy's home is very elegant'; aka simbi yamelo yakuka? 'Would you ~Sir like to sit down?'

3.1c. In a very common, and very important usage, the nominative occurs instead of the genitive in what might be called a "pseudo-genitive" construction [noun/verb/adj.]a+ni + NOUNb(nom.). (Since our purpose in this section is to illustrate the uses of the nominative case, more detailed discussion of some of these "ni-constructions" will be deferred until §4.) Formerly, many such cases were considered colloquial, and were frowned upon in "proper" speech and writing; but they are now acceptable not only in speech but in all but very formal or technical written work.

3.1c.i. It occurs especially with compound prepositions, where it has always been considered correct. "Compound" prepositions are of the sort ri vaka/ni... 'under...', ri ciniye/ni... 'to the middle of..., into the midst of...', ri ondeni... 'in the interior of..., inside...' etc., alo ondeyi/ni... 'from the inside of..., out of...', and many others. Grammatically, vaka, cini, onde etc. are the object of their preposition, and carry any necessary case-markings, but the noun that follows (in the nom.) is the logical object.

3.1c.ii. Somwhat similar to the above, and also considered correct, the construction is used to express a genitive relation between inanimate nouns in cases where such nouns cannot be viewed as intrinsically "possessing" the quality named. Consider the following:

- ace lacayi yu ~aceni laca yu the leg of the table
- anduro voriyi ~ nuroni vori 'the depth of the river'
- lusni acivar yu the end of the lecture...
- kunini ecut angur the color of Gwr fur

There are cases where a human or other animate noun can occur in this construction, but they will be dealt with in §4. For example, similar to the "genitive" use, many verbs and adjectives can be "nominalized" or topicalized with -ni (and with the lst and 2nd person suffixes as well). Some of these are long-established and completely acceptable-- e.g. such things as vitalni çenji, vitalmi 'çenji's height (is...), my height (is...)'; others are considered somewhat colloquial (and subject to over-use), but have become acceptable in almost all situations.

Thus, a sentence like mapila ~çenji yapila re hanjayi hacosa 'I think ~Shenji thinks (that) you ought to go' can be transformed to: pilami ~pilani cenji, hanjayi hacosa with essentially the same meaning, except that pilani/mi is now technically a noun (though indeclinable)and has topic intonation-- we might interpret this as 'my ~Shenji's thought/thinking is...'or 'what I think ~Shenji thinks is, you ought to go'.

3.1d. The nominative case occurs in direct object position with certain verbals expressions-- the so-called "double accusative"-- where (1) a number of causatives and basic verbs, and (2) a few unusual transitive verb constructions, have or appear to have two objects. Among the former are: rumaran 'to name/appoint s.o. as...'; lepes 'to call X Y', pilimen 'to consider X Y', tinje 'to elect s.o. to/as...'-- the person so named is, as expected, in the dative, or accusative if inanimate, and can be fronted for focus/emphasis, but the second object-- the position, what someone is called or considered-- is in the nominative and cannot usually be fronted. Examples of the first type::

- irumaran sim tolotiye tekangaya 'they named/appointed Mr. Toloti(dat.) Senior Professor(nom.)' ~sim tolotiye irumaran tekangaya but not *tekangaya irumaran sim tolotiye.
- mina yalepes çenjiye -çiçi- 'Mina calls Shenji "Shishi"
- milepes ele yu uku 'we call that plant(acc.) uku(nom.)' ~ iyu ilepes uku 'that's called uku' (lit., that(acc.) they-call uku(nom.)-- the nearest equivalent to a passive in Kash). Given the identity of the neuter nom/acc. cases, the case assignments here are based on analogy of sentences with animate objects.
- keyandolila ne ipilimen kataren 'The Andolians consider him (ne, dat.) a leader (kataren[nom.])'

(The causative verb rungarun 'to make/appoint/elevate s.o. (to the office of) karun', despite its Engl. translation, is an ordinary transitive verb in Kash usage, not a double accusative: minjerila irungaran akrowe sovandrula 'the deputies made Baron Sovandrula karun.')

Instances of the second type are made up of mepu or vele + noun/verb; the most frequently encountered is vele aran (lit. 'give name') 'to name'-- with reference to the ceremonial naming of children, as in

- mina yavele aran erek anayeni 'Mina named her child Erek.' Note that the apparent direct object aran (+erek) precedes the indirect object anaye (technically incorrect!); however, in this case vele aran is considered a unit, with the name 'Erek' in apposition (note too, only a proper name may follow aran in this construction). Thus, anayeni is the true direct object and as such may be fronted, while aran+erek may not be. It is possible to say, with normal word order, mina yavele anayeni aran erek as a somewhat literal translation of "Mina gave her child the name Erek" (with expected IO-DO order, though Erek in apposition remains in the nominative)-- acceptable but odd, since most Kash would have an immediate mental picture of Mina somehow handing the name over as if it were some tangible object. Still, it is correct to say yavele anayeni aran ehas/tarambik 'she gave her child a strange/unsuitable name,' or yavele anayeni aran otamayini 'she gave her child his grandfather's name ~she named him after his grandfather', but never *yavele aran ehas/otama anayeni which could only mean "she named her child 'Strange' or 'Grandfather'".

Some of the others like mepu honder 'to attack', mepu tenar 'to battle', vele tropa 'to warn', and irumale keti(ç) 'impose a condition/conditions on...'-- are also verb + noun, and like vele aran are considered "double accusatives", with honder, tenar, tropa viewed as accusatives. Similarly, (verb + verb), rumuwik mepu (lit., expel work) 'to fire, dismiss from a job', vele nanguça (lit., give consume-by-fire) 'to cremate', and vele/kota rapinda (lit., give/say welcome) 'to welcome'-- all these verbal expressions take a human object either in the dative (vele tropa, vele/kota rapinda, irumale keti(ç)) or accusative (the others, since they are seen as directly affecting their object), and non-human/neuter objects in the accusative, as expected. (mepu honder ri... and mepu tenar yam... can also occur with prepositional phrases, meaning more literally 'make an attack on...' and 'engage in/do battle with...').

3.1e. Finally, the nominative can occur as the animate (usually human) subject of many accidental (caka-) forms, although these more often take a dative human subject. (There are slight changes of meaning and/or connotations with a nom. subject.) Animate non-human subjects (rather infrequent), vary between nom. and dat. Inanimate subjects, however, are always in the nominative. These "accidental" verb forms are discussed more fully in §5; for now, just a few typical examples to illustrate the difference between nom. and dat. subjects:

- iya cakurum ~yacakurum (yam...) 'he(nom.) got into an argument (with...)' (< hurum 'argue')-- a subject prefix may be used here. Contrast (dative subject)--
- ne cakurum 'he's(dat.) argumentative, contradictory'-- the subject prefix is never used with a dative subject
- çenji (ya)cakanulañ 'Shenji(nom.) is depressed, feeling blue' (at the moment)-- contrast
- cenjiye cakanulañ 'Shenji(dat.) is melancholic/depressed' (by nature)
- ñakimi (ya)cakananjul 'my car(nom.) got stolen, has been stolen'
- me cakananjul 'I(dat.) got robbed'
- pumbikni cakumbor, nim lopaye (~lopa) cakombra 'the shed collapsed and 5 lopas got killed'

3.2 Accusative. Its principal use is as the direct object of transitive verbs-- mostly when the object is animate but non-human, or neuter. (Even though neuter nominative and accusative forms are identical, ambiguity hardly ever results, since neuter nouns rarely occur as subject of a transitive verb; or when they do, context or semantics usually clarify the subject/object relation.) Human objects are in the accusative only if the action of the verb directly affects them, especially in some physical way; otherwise, they are in the dative case (see further §3.3 Dative). A few verbs take their object in the genitive, see §3.4c. Examples of the accusative:

(Human objects:)
- lopa ya yakici minan 'The/that lopa bit Mina'. yen yakici 'It bit her'.
- erek yakepak çenjin 'Erek hit Shenji' [with his fist]. erek yan yakepak 'Erek hit him'.
- karun iya yatrakasa sila kinjin 'The duke bought 3 slaves' (archaic; slavery is long gone on Cindu.)
- man ikrahar 'they tortured me'.
- nin yarungombra ~nin yapakran. 'he killed them'. (He actually did something to cause their death; pakran implies "by violence" or "with a weapon".)

(Non-human animate, and neuter objects:)
- erek yasorom nim lopan 'Erek sold 5 lopas'. nin yasorom 'he sold them (anim.)'.
- mipole tikas tandin (~tandilan) loroni tayondre 'We can see both moons tonight'. hayi, nin matikas 'yes, I see them'. (The word 'moon', and the names of the two moons, are animate nouns.)
- anala isosir tukrayon 'the children heard a tukrayo' (a bird-like animal). aka yan hasorir? 'did you hear it?'
- yamelo titingas puna yu 'he wants to inspect that house'. (yu) matitingas koprat 'I looked it over yesterday'. (The neut. obj. pronoun yu is usually omitted if the context is clear.)
- mina yatarupat tingas cika 'Mina refuses to watch TV'. ta (yu) yalisam 'she doesn't like it'.
- çenji yarungombrato lopan 'Shenji will kill (slaughter) a lopa'. yan yarungombrato 'he will kill it'. (pakran would not ordinarily be used in this case, though one might pakran a wild animal.)
- pilami, harungombra angeç niç 'I think you've killed those trees'. (Deliberately, e.g. with a poison-- contrast niç harukorem 'you've killed them (~let them die)'-- by inattention, wrong planting, etc.)
- çenji yakaya (ya)çindi sende angur 'Shenji knows how to speak Gwr' (çindi can be transitive if its object is a language, "words" etc.)
- yanga yamace angeç taç 'the wind damaged those trees' ('wind' is animate)
- sunanja soket yanahan kanakinda 'the heavy surf eroded ("ate") the headland'-- although both 'surf' and 'headland' are inanimate, even if the object is fronted for focus or emphasis the meaning will still be clear since 'headland' cannot logically be the subject of 'erode'-- kanakinda yanahan sunanja soket would typically occur where Engl. uses a passive, 'the headland was eroded by the surf'. Similarly with topic intonation, kanakinda, yanahan sunanja soket
- ta milisam yanjakrumun 'we don't like a typhoon(anim.)'

The other use of the accusative case is with prepositions, principally ri and its compounds, indicating location in, at, or on: yale ri puna 'he's in the house' (or, 'at home'); (yu) matraka ri fale 'I bought it at the market'; hakukkapo ri laca yu 'just put it on that table'; kandri hatikas ri ondeni puna? 'what did you see in(side) the house?'; ne matikas ri kandini fale 'I saw him in front of the market'; mina (yu) yatromat ri vakani laca 'Mina found it under the table'.

Colloquially, the verb ale is often omitted in short sentences: çenji, ri puna(ni)/opor; indemi, ri fale 'Shenji's at home/outside; my mother's at the market.' Note also the question word riyena '(at) where?' (considered to be an accusative form)-- riyena çenji? 'where is Shenji'; riyena eyalaç poren? 'where are the wine-glasses?'.

Ri by itself is not used with human or living animate nouns; but it is used with that handful of inanimate nouns classed as animate (heavenly bodies, some forces of nature etc.), e.g. (r)i roçen, ri latondrelen 'on/in the sea, on the planet'. Compound prepositions may take human objects, of course-- ri kandini mina, ri çelumbi 'in front of Mina, behind me'. Note, however, that ambiguity may result since some of the location words, in particular vaka 'underside, bottom, below; rear end, buttocks' and çelum 'back of, behind; back (anat.)', also refer to body parts; if the context is not clear, ambiguity can be avoided by using the genitive, or another phrasing. For example:

- ri vakani kopak 'under the box' vs. ri vaka kopaki 'at the bottom of the box'
- yale kusimi (a cat-like animal) ri çelumbi 'there's a kusimi behind me', but possibly, 'there's a kusimi on my back'-- for clarity one could say yale kusimi ri nihiñi çelumbi '...on top of [on the surface of] my back'
- yale sucup ri vaka çenjiyi 'there's mud on Shenji's rear end'; and there is little likelihood of misinterpreting lopa yakici çenjin ri vakani 'a lopa bit Shenji on his/the butt'

The prepositions yam 'with (accompanying)' and liri 'about, concerning' also have their noun object in the accusative: yam çenjin, liri minan, liri acal yu 'with Shenji, about Mina, concerning that matter'. But with a pronoun object, yam combines with the accusative form--

yamban 'with me' yambin 'with us'
yakan 'with you' yakin 'with you (pl.)'
yañan/yañen 'with him/with her' yandin 'with them'

whereas liri adds the possessive suffixes--

lirimi 'about/concerning me' lirimim 'about/concerning us'
liriti 'about/concerning you' lirihi 'about/concerning you (pl.)'
lirini 'about/concerning him/her/it/them

3.3 Dative. The dative case has five important uses: (a) as animate indirect object-- recipient or sometimes beneficiary-- of many transitive, and some intransitive, verbs; (b) as human direct object of those transitive verbs that are viewed as not having a direct affect (usually physical) on their object; (c) as object of prepositions indicating motion to, toward, into, onto, etc.; (d) in a widely-used expression for "to have"; and (e) as human "subject" or experiencer of most accidental caka-verb forms. (Except in prepositional use, the dative is usually restricted to animate nouns.)

3.3a. Indirect object. The indirect object, if a noun, immediately follows the verb; if a pronoun, immediately before. Examples:

- mina yavele andine kreki 'Mina gave the little boy a cookie'; mina ne yavele kreki 'Mina gave him a cookie'-- an indirect object is generally required with di-transitive verbs like "give, lend" and may not be deleted.
- yavele lopaye ñuçuket 'he gave the lopa an injection'-- ñuçuket may also be used as a di-transitive verb yañuçuket lopaye andumbrani 'he injected the lopa with the medicine ~he injected the medicine (in)to the lopa'; or, its object may be implicit, as in kandumbra te poro yañuçuket tanju 'the doctor is going to inject you now'
- kacifes yasorom çenjiye ro lopan 'the shepherd sold two lopas to Shenji'-- but with this verb, and many others like it, an indirect object may be omitted (or simply implied); thus it is correct to say kacifes yasorom ro lopan ombi yapita toye 'the shepherd sold two lopas because he needed money' {where, in the discourse, it is unnecessary to know to whom he sold them)
- me yakota re... 'he said to me that..., he told me that...' versus yakota re... 'he said that...'

Many verbs that are considered intransitive in Kash may also take an indirect object:

- yaçindi minaye; me yaçindi 'he spoke to Mina; he spoke to me'
- me yakinda; yaminda minaye 'she laughed at me; he smiled at Mina'
- icocon çenjiye liri añanguça 'they suspect Shenji of arson'
- (me) yamendro re hindani yamut 'he insisted (to me) that the Earth was flat'
- yataken çenjiye; ne yataken 'he pointed at Shenji; he pointed at him'
- mina me yaciyon yaçangi 'Mina (to me) seems/looks sick' (One could also say me yaciyon re mina yaçangi 'it seems/appears to me that Mina is sick')

Verbs of motion also may take a (human) dative, since ri cannot be used in such cases:
- yacosa amayeni yanuça toye 'he went to his father to ask for ~asking for money'
- a, sisati (te) yamamaçan 'oh, your lover will come back (to you)'-- cf. imamaçan ri holundaye 'they returned (came/went back) to Holunda'

BENEFACTIVE... [not yet!]

(In those cases where the IO phrase is considered to be "too lengthy", it is either topicalized, or rephrased-- as was exemplified in §1.2d.)

Finally, there is the unusual question word kañale 'how, in what condition..., like what? ', which takes the dative (animate only): kañale hate?; kañale çenjiye? 'how are you? how are you getting along (in a specific situation)?'; 'how is Shenji?'; kañale kusimileni 'what's the situation with the kusimis (that were sick, or were causing problems)?'-- this is a serious question, not formulaic as in English, to inquire after the state of someone's health or other situation, and expects a real answer. To inquire about inanimate things, the nominative is used: kañale eçevati?; kañale laluni? 'how was your trip ~what was your trip like?; 'how was the party?'; with a human nominative, it asks "what is ... like [as a person, or as to character etc.]"-- kañaleni ~kañale iya? kañale mina? 'what is/are he/she/they like? etc.; what's Mina like?'

3.3b. Direct object. Most transitive verbs that (in the Kash view) do not directly affect/change their object, put a human direct object in the dative case (non-human animates, and inanimates, are in the accusative, as we showed above, §3.2). Typical examples:

- matikas çenjiye 'I see Shenji'; ne matikas 'I see him/her'
- ta mavacanjan kaçute iya 'I don't trust that man'
- mina yasisa çenjiye 'Mina loves Shenji'; mina te yasisa 'Mina loves you'
- sovandrulaye yarumopor karun 'Sovandrula was exiled by the karun' (with fronted object, the sentence corresponds to an English passive, but structurally it is active; rumopor lit. cause to (go) out < opor 'out(side), away'). Note that similar rumuwik 'to banish, drive out' (lit., cause to flee/run away) by convention takes an accusative object; it refers mainly to the ancient punishment for taboo-violation.

In former times, speakers were more careful about distinguishing whether a verb took a dative or accusative object, and educated speakers and writers still are; but generally nowadays the rule of thumb is, if in doubt, use the dative. Thus examples like the following may occasionally still be heard with either accusative or dative:

- kandumbra ya yarundumbra minaye ~minan 'the doctor cured Mina'
- aka pilati, pole hatundru çenjiye ~çenjin? 'do you think you can lift Shenji?'
- yarungamon anayeni ~anañi kotrelo cosa 'she gave birth to her child a week ago ~last week' (this verb may also occur without a DO, "she gave birth last week")

In some cases it can be ambiguous whether a dative form is the direct or indirect object (if, for example, the neuter object pronoun yu has been omitted). This affects verbs like filan, copin, runjosa 'bring/take, go and get/fetch, send'. Thus:

- çenji yafilan minaye ri laluwe 'Shenji took Mina to the party' (or, 'Shenji took [it] to the party for Mina')
- yacopin mambreyeni 'he went and got/fetched his brother' (or, '... fetched [it] for his brother')

If the context is unclear, a confused listener may ask kandri filañi/copiñi? 'what did he take/fetch?', and the speaker will have to clarify.

It can happen that these same verbs, and the causative forms of some basic transitive verbs, can have a human direct object as well, in which case it must be in the accusative, since two dative objects are not possible--

- çenji yafilan indeyeni minan 'Shenji took/brought Mina to his mother' (perhaps to introduce her; note, however, that indeyeni without context is in fact ambiguous as to whose mother-- Shenji's, Mina's or someone else's)
- yacopin indeyeni mambreñi 'he fetched his brother for his mother'.

If we keep in mind that causative verbs have an underlying structure something like [SUBJ a cause [SUBJb verb OBJc]] which becomes [SUBJa rumcaus+verb DATb ACCc], the resulting sentence is clear; in effect, the subject of the original sentence becomes the indirect object of the causative verb. Thus basic sentence mina yanunji çenjiye 'Mina met Shenji' > [my father cause [mina met shenji]] > amami rundunji minaye çenjin 'my father introduced Mina to Shenji' (note that the case structure of the Kash sentence is the reverse of the English); and likewise if the objects are pronouns: ...ne rundunji çenjin; ...yan rundunji minaye; ...yan ne rundunji. It would also be possible to say amami nile rundunji (liya-liya) 'my father introduced them (to each other)'. It is also possible to omit the accusative form: amami rundunji minaye 'my father introduced Mina [e.g. to the group at large]'.

The synonymous verb rundimbat 'to introduce' (<timbat 'to know s.o., be acquainted with s.o.') has somewhat more formal connotations, and differs from rundunji in that it must always have both noun/pronoun objects-- i.e. neither *...nile rundimbat (liya-liya) nor *...ne rundimbat is possible.

The causative of copin, runjopin 'cause/send s.o. to fetch s.o/s.t.' behaves in the same way: yacopin mambreyeni 'he fetched his brother' > indeni ne yarunjopin mambreñi 'his mother had him/sent him to fetch his brother'. And likewise with rundikas 'to show, exhibit' (<tikas 'to see'): mile yarundikas memeñi 'she showed us her baby'; it may occur without indirect object in karun yarundikas anakasiñi 'the karun exhibited his first-born [understood: to the people]' (a formal ceremony), and of course in a sentence like ketrafun yarundikas kraçeteçni 'the artist exhibited his paintings'.

3.3c. Object of prepositions. With ri and its compounds, and usually with verbs of motion, the dative expresses motion to, toward, into, onto etc.:

- yacosa ri faleye 'he went to the market'; maharan ri vunuwe fale 'I walked in the direction of/toward the market'
- irata ri holundaye 'they came to/arrived at/in Holunda'; also with rapinda 'welcome!' as, rapinda ri holundaye/punayemi 'welcome to Holunda/my house!'

Colloquially, the verb cosa 'go' is often omitted in simple sentences: inde ri faleye 'mother [has gone] to the market'; yanda ri opore! 'don't [go] outside!'. Note also the question word riyene (with dative -e) 'to where?' e.g. riyene hafilan etengiç taç? '(to) where are you taking those books?'; riyene yacosa mina? ~riyene mina? '(to) where did Mina go?'

With the verbs cosa, rata, fosi 'go, come, sail' (and one or two others), instead of using ri + dat. the location may be treated as a "direct object", with the animate acc. ending -n, even though the noun is inanimate: yacosa falen 'he went to the market'; mirata holundan 'we arrived at Holunda'; ifosi kandrokon 'they sailed to Kadrok'; also in the expression fosi roçen 'to sail the seas, to go to sea'. Far from being colloquial, this is a survival of a much older and more widespread usage.

Other examples:

- yahakuk lipetni ri nimuweni 'she put the lid on(to) the pot'; yahakuk lipetni ri nihineni nimu 'she put the lid on top of the pot'
- hambos yamaç ri sutuwiye 'add the noodles (in)to the broth'
- kusimila iyama ri ondeyeni añange the kusimis ran into the forest'
- yapondam ri vakayeni kopak 'it fell (to)under the box'-- contrast yapondam ri vakaye kopaki 'it fell to the bottom of the box'.

Just as with the accusative usage, there can be ambiguity with dative çelum, vaka + human object; if the context is not clear, such expressions are clarified in the same way:

- kusimi ya yanecu ri çelumeni çenji 'the kusimi jumped (to) behind Shenji' or '...jumped onto his back'-- for this last, ...yanecu ri nihiñe çelum çenjiyi or ...ri çelume çenjiyi are unambiguous

3.3d. Dative of possession. "To have" is expressed in Kash with the construction X(dative) yale Y(nom.), 'X has Y' lit., to X there is Y (the grammatical subject); the dative form is always fronted, yale is always invariant (even if Y is plural). Most speakers restrict this to human possession, where, according to context, it may refer either to casual/temporary possession, or to intrinsic possession or ownership. For other animates, and inanimates, it may be used for intrinsic possession, e.g. body parts or integral component parts, but otherwise, alternative expressions are available and usually preferred. (In the Kash view, an inanimate object cannot truly "possess" another object.) Some examples:

- minaye yale etengi 'Mina has a book' (most likely interpretation: with her at the moment-- she may or may not own it). minaye yale pando etengiç 'Mina has a lot of books' (either 'with her' or 'she owns many'-- similarly in case of definite etengini ~etengi yu); but minaye yale pando etengiç ri pilusni 'Mina has a lot of books in her bag' quite clearly refers only to temporary possession.

- çenjiye yale puna ri hacu 'Shenji has a house at the beach' (he may own it, rent it, or simply have the use of it); similarly with çenjiye yale ñaki 'Shenji has a car'

It is also used with reference to parts of human and animal bodies--

- kaçute ya yale ecut kuniró 'that man has lightly-striped fur' (refers to the just-visible black/lighter black pattern in some fur)
- minaye yale çikunuç mimik 'Mina has small ears'
- ne yale ace cahorok 'he has a deformed leg' (a permanent condition; for a temporary condition, e.g. ne yale ace matra 'he has a broken leg' other phrasings would likely be preferred)
- lopaye/lopale yale ha ace 'the lopa has/lopas have 4 legs' (a generic statement); lopaye ya yale sila ace naponi 'that lopa has only 3 legs'

-- and in some cases to integral parts of inanimates: yuno ñakiye yale ro forici 'all cars have 2 headlights'; lacaye tayu yale ket ace 'this table has 6 legs'; punayeni yale nim atel 'his/the house has five rooms'. However, most speakers consider these cases odd, and prefer other phrasings, for example (1) topicalization (yuno ñaki, foricini ro lit., all cars, their headlights (are) two; laca tayu, aceni ket; (2) a ni-phrase (see §4.3), optionally topicalized, foricini yuno ñaki(,) ro; aceni laca tayu(,) ket; (3) a prepositional phrase, yale ro forici ri yuno ñaki lit., there are 2 headlights on all cars; or (4) (somewhat formal) a genitive, yale ket ace lacayi tayu, lit., there are 6 legs of this table. Generally, possession involving inaminates is best expressed with the ni-construction.

In fact, even the examples above relating to human possession may be phrased in these same ways, with the exception of (2), the ni-construction. There may be slight differences in their connotations, however; thus, "X(nom.,topic), yale Y-ni" is usually taken to mean "X has Y at the moment ~available for use", e.g. yale ñakini 'he has his/the car [with him]; mina, ta yale etengini 'Mina doesn't have her book', etc. This should be kept in mind especially when talking about money-- çenjiye/me yale toye can mean 'Shenji has/I have money' [i.e. we are rich]; it is better to say yale toyeni or yale toyemi 'he has money ~I have money' [at the moment, he/I will pay]. The Kash consider it gauche to discuss one's own, or others', financial status except with family, very close friends, or perhaps one's financial advisor.

Very often, rather than indicating possession with the dative construction, speakers will phrase their statement with other verbs: while me yale ñaki velu 'I have a new car' is acceptable, mende masorom ñaki velu 'I've bought a new car' could also be used. To express actual ownership, the verb apeña 'own, belong to' may be used; thus, amayeni yale çombala ri hutuñi tupos 'his father has land(s) in the Tupos region' could better be paraphrpased amani yapeña çombali... 'his father owns lands...' (apeña takes the genitive case, and will be discussed further in §3.4d.)

English and other familiar languages have many expressions with the verb "to have" that have nothing to do with possession; Kash has other vocabulary in such cases, e.g.--

- miyambal lalu 'we're having (lit. offering) a party'
- yaleka lopa ciki? (at a restaurant) 'is there (do you have) fried lopa?'; or mameloto ~malisando lopa ciki 'I will want/like (I'll have) fried lopa'
- yaçangi (~yakena) avos 'he has avos (a mange-like disease)', lit. he is sick ~he suffers avos; ne caçapalak 'she has a headache'-- but colloq. ne (~me etc.) çapalak is common
- ne cakakar 'he/she has hurt feelings, his/her feelings are hurt'
- niyombi (me) yahar ~yale niyombi ahar 'I have a sore/painful arm ~my arm hurts'; yale evine ri niyombi 'I have/there's a sore/wound on my arm'-- the Kash hold that one does not "have", i.e. possess, such physical problems, which are essentially transitory

3.3e. Subject of accidental caka- verbs. These verb forms most often have a human subject (or better, experiencer) in the dative; the subject is viewed as having little or no control over the action/state of the verb, which is essentially passive in meaning. (But inanimate subjects are always in the nominative; and human subjects may also be, though in that case the meaning and/or connotations of the form are slightly different-- cf. §3.1e above and especially the complete discussion of the accidental form in §5.) The following are typical of those forms that usually take a dative human subject:

- cakuça 'get caught/burned in a fire' (< huça 'fire')
- cakatevas 'overly refined, affected' (< tevas 'delicate, elegant')
- cakanuwak 'overly inquisitive, nosy' (< nuwak 'ask')
- cakrata 'show up unexpectedly/inopportunely' (< rata to come')
- cakafatap 'needlessly strict/demanding; to be a "pain in the ass"' (<fatap 'harsh, strict')
- cakuwis 'be/feel trapped (in a situation)' (< uwis 'tight, narrow')

3.4 Genitive case. The genitive case is used (1) to show possession, (2) as the object of prepositions indicating motion from, out of, away from, etc.; and as the "standard of comparison" in comparative sentences (3) as the direct object of certain verbs; (4) to indicate the "agent/cause" of caka-verbs; and (5) in colloquial speech, as a partitive.

3.4a.Possession by humans. This is the principal use of the genitive; It is also used in certain cases for non-human animates, but much less often in cases involving inanimate possessors-- the restrictions are similar to those governing use of the "dative of possession", §3.3d above. The genitive form follows the thing possessed (with a single exception: the question-word kariyi 'whose?' which normally is the first word of its sentence).

Examples of human possessives:
- niluç çenjiyi 'Shenji's hands'
- çaka karuni 'the karun's power'
- ñaki mambreyi minayi 'Mina's brother's car' (colloq. ñakini mambre minayi might be heard, but is frowned upon)
- puna yu simi toloti re yavelu 'that new house of Mr. Toloti'-- (puna yu velu simi toloti is acceptable but not preferred-- speakers feel the possessive should not be separated from its noun)

- nilusmi, nilusti etc., 'my hand, your hand'.
- ñaki mambreyini 'his/her brother's car'. Note that the 3rd person suffix -ni can be ambiguous-- "his, her, their", even "the"; if necessary, the ambiguity can be avoided by using a name, or the alternate forms iyani(tu) 'his (own)', iyeni(tu) 'her (own)', nili(tu) 'their (own)', or a demonstrative pronoun for "the"; thus, ...mambreyi iyeni 'of her brother'.

The genitive of the personal pronouns can be used for emphasis or contrast--
- puna mili 'our house'
- lopala ini (~iyani) 'his lopas'

and also if the noun is modified by a demonstrative--
- puna yu mami ~punami yu 'that house of mine'

Noun and pronoun forms in the genitive can also be used as possessive adjectives: yale çenjiyi 'it's Shenji's'; niç, mami 'those are mine'; and, preceded by a demonstrative, they function anaphorically as nouns--

amami kañulin, i eme ya çenjiyi (~...i ya çenjiyi eme) 'my father is a carpenter, and so is Shenji's (~and Shenji's is, too)'
- punati yavirik, ta mowa yu nili 'your house is pretty, but not theirs'
- riyena niç ini ~iyani 'where are his (things)?

Note that the demonstrative agrees in case:
- matimbat amayeni, ta mowa ine hati 'I know his father, but not yours [i.e., I don't know your father (subj+verb deleted)]'-- contrast matimbat amayeni, ta mowa ya hati 'I know his father, but yours doesn't [i.e., your father doesn't know his father (verb+object deleted)]; this would more likely occur with emphasis on "I" and/or "yours", mam timbat amayeni, ta mowa ya hati.

3.4b. Non-human/inanimate genitives. In current usage such nouns occur in the genitive only when referring to inherently possessed qualities or characteristics-- for example, body parts of animals, component parts of objects-- or to avoid ambiguity. As with the dative of possession, there is variation here as to what speakers consider inherent; animate nouns are more likely to be in the genitive in questionable cases than inanimates. Some speakers hold that inanimate objects cannot truly "possess" anything, hence can never occur in the genitive, and always use alternate constructions-- this is the excuse used especially by colloquial speakers. Examples:

Animate, non-human:
- ace lopayi; nafeç nundiveki; anala kusimiyi 'the lopa's leg; scales of the long-fish; the kusimi's kits (babies)'-- most speakers agree that these are appropriate uses of the genitive
- acangi lopali; andraka nundiveki; angamon anali 'a disease of lopas; the size of the long-fish; the birth of the kits'-- most speakers would consider these better with different phrasing, most likely with -ni: çangini lopala, nafeçni nundivek, kamoñi anala

Note that edible animals and their parts undergo grammatical changes; first of all, animals as food are inanimate. Thus, at the slaughterhouse there are aceçni lopa(la) 'lopa legs'; with further processing, by the time it reaches the market or one's oven, it is ace lopa 'lopa leg, leg of lopa'.

- ace/nihin lacayi; fundoñ/ñeraç punayi; anduro voriyi; hañon sovareyi; foriciç ñakiyi 'the leg/top of the table; the roof/walls of the house; the depth of the river; the sheath of the sword; the headlights of the car'-- here, some speakers might use the genitive, since these describe part-to-whole relationships. But many others would prefer some other construction, usually the ni-phrase: aceni laca, nihiñi laca, ñeraçni puna, nuroni ~anduroni vori, foriciçi ñaki, etc. (See further §4.)

3.4c. Object of prepositions. Three prepositions take the genitive case: principally alo 'from,' indicating origin or motion away from, out of, etc. Its compound forms and usage are comparable to those that take dative or accusative-- typical examples:

- yarata alo holundayi 'he comes from Holunda'; yacosa alo punayi 'he went (away) from the house'
- çeka yapondam alo selimini fundoñ 'a tile fell off/from the edge of the roof
- yapotikas alo vakayini roçe 'it emerged from under the sea'-- note that "from the bottom of the sea" would be alo vakayi roçeyi
- matundru alo nihinini laca 'I picked it up off (the top of) the table'

In comparative sentences, alo + genitive expresses the "standard of comparison", that is, "than" in 'more...than X (alo X)', "of/in" in 'most...in the X (alo X)', and "as" in 'as...as X (alo X)'. The preposition may be deleted if 'X' is a noun, but not if it is a pronoun or indeclinable form-- see §6.2 for further discussion and examples.

Note also the question word riyeni '(from) where?', with genitive -i: riyeni iya? 'where is he from?'; riyeni irundata yawunduni? 'from where do they import their oil?'.

Two other prepositions also take the genitive: inga 'without' and uçoñi (uçom+ni) 'for, intended for, on behalf of; for the use/purpose of'. (Formerly, umit and yambit 'with, by means of' took the genitive, but now take the accusative.) Colloquially, uçoñi is often heard with a nominative, as if it were a ni-phrase-- considered uneducated usage. Examples:

- yanda ri opore inga payowi! 'don't go out without a cape!'
- etengi inga andipeti 'a book without a cover/binding'
- icosa ingami 'they left without me'-- here we see that a pronominal object is expressed with the possessive suffix: ingati, ingani, ingamim, etc. 'without you, without him/her/it/them, without us'

A sentence complement may follow inga, with re:
- icosa inga re mile ikota kinunji 'they left without saying good-bye to us'
- yaçindi inga re yapila-pila 'he spoke without thinking'-- colloquially but equally acceptable: yaçindi inga pila-pila(ni)

- yanda imu poren tayu, uçoñi laluwi sapat 'don't drink this wine-- it's for the party tomorrow' (better than ...uçoñi lalu...)
- maçindi karune uçoñi çenjiyi 'I spoke to the karun for/on behalf of Shenji' (without context, it is ambiguous whether I interceded for Shenji, or Shenji asked me to)(better than ...uçoñi çenji)
- mina yafilan andahan uçoñi keçangili 'Mina brought food for the invalids'-- implied: for their eventual use, or, she merely delivered it to their house. Contrast with the dative: ...yafilan keçangile andahan implying rather, for their immediate use, or, she served it to them. A fine distinction, clearer perhaps in: sim anekri mile yasorom punani 'Mr. Anekri sold (to) us the house' vs. ...yasorom punani uçombim '...sold the house for us' (i.e. as agent)

3.4d. Genitive direct objects. A few verbs take their direct object in the genitive case: kendra 'to worship, pay (religious) homage to' (somewhat archaic), teçayu 'to yearn, long for', apeña 'to own; to belong to', inga 'to lack, be without, not have' and orana 'be like, similar to; resemble'. Two others take either genitive or dative/accusative, with slight change in meaning: cuta 'forget' and nimbur 'remember'. Note that a genitive object pronoun does not precede the verb (as dat/acc. pronouns do), but rather, follows it. Examples:

- kañulin yakendra çehamali angeç 'carpenters worship/pay homage to the tree-spirits'
- iteçayu yuno nuraki hindakale 'they all yearn for world peace'
- ama çenjiyi yapeña landiñi 'Shenji's father owns a bank'; landiñ yu yapeña amayi çenjiyi 'that bank belongs to Shenji's father'
- ukat/niya-niya yayinga añiñayi 'he is quite lacking in intelligence'; miyinga toyeyi vara mitraka 'we lack the money to buy it'-- colloquially, most speakers substitute ta yale... for inga, which is considered rather formal
- çenji yaworana amayini 'Shenji resembles/looks like his father'; andotri yaworana ini re matikas ri yundrafun 'your vase resembles/looks like/is similar to (the) one I saw at the museum'. This verb may occur without an object: na, yaworana 'well, it's similar'; iyorana 'they're similar; they look alike'. Note that the causative rundana 'to copy, to imitate' behaves normally, with dative/acc. object: yarundana añuri yu he copied the document'; yarundana kacipe 'he imitated the teacher' (implied: disrespectfully)
- mina yateçayu ini (~iyani) 'Mina longs for him'; çenji yaworana hati 'Shenji looks like you'

With a genitive object, nimbur connotes long-term, permanent memory, cuta total forgetting, erasure from memory; with a dative/accusative object they refer to temporary or sudden states, for example:

With the genitive:
- anje manimbur amayimi, kambun yahorem anju male ana 'I still remember my father, even though he died when I was a child'; manimbur ini (~iyani) 'I remember him'
- kacivar yanimbur acivari 'the student remembered the lesson'
- mina mende yacuta çenjiyi 'Mina has forgotten Shenji ~put him out of her mind'; mende yacuta ini 'she has forgotten him/her/it'
- talunda micuta lerowi tayu 'we will never forget this day'

With the accusative/dative:
- maturo, ta manimbu(r) arandi 'I'm sorry, I don't remember your name'
- aka hanimbur etengiti 'did you remember [to bring] your book?'
- mina yacuta ereke 'Mina forgot [to include] Erek'
- ayi, macuta poreñi 'oh dear, I forgot the wine'; ayi, (yu) macuta 'oh, I forgot it'

The causative forms of these two verbs behave normally, taking dat/acc. objects:

- suwaves iyu me rundimbur indembi 'that perfume reminds me of my mother'
- angundi ri hakuli mende ne runjuta pando-pando 'an injury to (in) the brain has made him forget all sorts of things'

3.4e. As "agent/cause" of caka-verbs. Although caka-forms tend to be agentless, it is possible to express the agent or cause of the experience with a noun in the genitive (also, with a prepositional phrase or subordinate clause), for example:

- ne cakapak pirowi 'he got hit by/with the ball'
- ne cakepak ereki 'he got hit by Erek'
- çenji cakasisa minayi 'Shenji is desperately/inappropiately in love with Mina', or ...liri minan lit., with respect to Mina
- me cakaçangi endaki rus 'I got sick from mouldy meat', or, ...cakaçangi ombi manahan endak rus '...got sick from eating (because I ate) mouldy meat'

3.4f. Partitive. Mainly with reference to food items, some speakers use the genitive in a partitive sense: meloka ukuwi? '(do you) want some uku?'; pitani samosi 'it needs salt'; tambelo poreni 'I don't want (any) wine'. However, as these examples show, it is a very colloquial and informal usage, generally viewed as inappropriate outside one's intimate circle.

3.4g. English and other languages have many expressions with "of" either with a partitive or some other sense-- these usually have a different structure in Kash. For example, londo sawu 'full of water', mepu hece 'made of iron'; me yarungaya kracalni 'he informed/told me of the problem', lit., caused me to know the problem; mapori lirini 'I'm tired of it', lit., with respect to it.

Likewise, Kash does not distinguish "objective vs. subjective" genitive; sisa minayi 'love of Mina' does not mean *'(someone's) love of Mina'-- that would be e.g. sisa çenjiyi liri Minan 'Shenji's love of Mina'-- it means only "Mina's love (of someone)"; angamon minayi means only 'Mina's birth (Mina's being born)', not *Mina's (act of) giving birth'-- that would require rephrasing, perhaps mina rakop yarungamon 'Mina gave birth with difficulty'. In other words, Kash has only the subjective use.

4. Uses of -ni

Aside from its use as 3rd person possessive suffix, it has three other important and very frequent functions. (Far too frequent, in the opinion of many grammar-school teachers, who have coined the term nini-nini, cakanini to point up and criticize their students' overuse and/or misuse of -ni). Its use is somewhat colloquial, often idiomatic-- it is one of those aspects of the language (to evade the issue!) that is best acquired by experience.

4.1 As definite marker. Although forms of (i)ya 'that, the' mark definiteness in all cases, it is required only with human nouns, merely preferred with other animates. Inanimate nouns, however, may use either (i)yu or -ni; some speakers make a distinction, using yu only in its demonstrative sense 'that (specific or contrastive)', as opposed to -ni simply to refer to previously mentioned, known or presupposed material in the discourse. (From the translator's point of view, it can often be ambiguous whether 'his/hers/its/their' is meant, or simply 'the'.) The demonstrative may also imply a certain emotional, even physical, distance from the object, while -ni implies a more intimate, more familiar connection. Many speakers also make these distinctions with animate non-human nouns, though sometimes inconsistently; still, while such use of animate+ni is acceptable or at least tolerated in casual speech, more careful speakers frown on it. Examples:

- numu ya, traponahan 'that fish is inedible' (referring perhaps to a particular [living] species), similarly numu yu, traponahan 'that fish is inedible' (i.e., the one offered for sale-- it's not fresh, it's discolored)-- versus numuni, traponahan 'the fish is/was inedible' (i.e., the one we were served; it was poorly cooked, or tasted a bit off)

- lopala nila ikavan 'the/those lopas are healthy'-- perhaps a prospective buyer's comment; versus lopalani ikavan 'the lopas are healthy'-- the owner's opinion of his own flock

- atel yu sut yaworamba 'the/that room is too dark' (as viewed from a distance) versus atelni sut yaworamba 'the room is too dark' (speaker/hearer are in the room; or, it was previously mentioned in the discourse)

Nouns referring to humans are never made definite with -ni (except incorrectly); forms of iya are used, and even they can be omitted if the context is clear. Personal names are inherently definite. The handful of "animate" inanimates (the sea, heavenly bodies, forces of nature), being in a sense personifications, generally behave like personal names-- inherently definite, at least insofar as the world of Cindu is concerned. (In a discussion of our Earth, the Kash might refer to tandini 'the moon'-- though it is equally likely they mean 'its moon'.)

4.2 As a nominalizer. Not unrelated is the use of -ni (or one of the other possessive suffixes) to nominalize verbs/adjectives. With respect to verbs, we mentioned this briefly in §3.1c.ii above-- viz., the transformation of sentences like mapila re [Sentence]; çenji yapila re [Sentence] 'I think that..., Shenji thinks that...' to pilami, [Sentence]; pilani çenji, [Sentence] with essentially the same meaning, except that topicalized pila+poss. is now technically a noun (as shown by the possessive suffix). Note that the actual nominalized form ambila refers more to the process or ability of thinking, and would not be appropriate here-- *ambila(-mi,-ni) [Sentence].

Only the bare verb base can be nominalized with -ni-- inchoative and causative forms in particular are not. Colloquially, some potential and (quite often) accidental forms may be, but it is not considered good usage. For example, ne cakacut i niya me yakraku lit., he is forgetful and it very much annoys me, or 'his forgetfulness is very annoying to me' is more "correct" than colloq. cakacutni niya me yakraku

Many other verbs can be nominalized with -ni (or another possessive suffix)-- in many cases these are a less formal version of the moribund "gerundive/infinitive" with -ale (its usage is a mystery to all but the most pedantic grammarians)-- a verbal noun meaning roughly 'the doing ~the state of [VERB]'-- versus the añ- nominal which usually means 'thing that [VERBS]' or 'thing that [IS VERBED]'. For some verbs, the ni-form and añ-form are synonymous, or nearly so, and in such cases the añ- form is falling into disuse, at least in the colloquial language; for some verbs that has already happened, and only the ni- form occurs-- especially true of adjectives. Some examples:

- yamani takrivus ~trivus 'his running is clumsy, he runs clumsily' (the word order may be reversed; both are colloquial, trivus moreso than takrivus;) cf. ultra-correct yamaleni yatakrivus). There are other, less colloquial, ways of saying this: yayama (ya)trakivus ~-trivus lit., he runs he is clumsy; trakivus ~trivus yayama, where trakivus would be considered an adverbial usage. (Note: yamami ~yamati etc. are also possible.)

- tengeni yapita ro aro 'the washing [i.e., his/her/the act of washing (something)] took two hours' (note that the true nominal andenge refers to the things to be washed, the laundry). Also topicalized (colloq.) tengeni, pitani ro aro. We could view these as derived from yapita ro aro vara (ma-, ya-...)tenge 'it took two hours in-order-that (I, he...)wash (something)' ~more freely, 'it took (me, him...) two hours to do the washing'.

- sakundri ne yayukar yarakop 'swallowing is becoming difficult for him' (sakurale 'to swallow, swallowing' would be ultra-correct)-- there is no form atukar attested. Alternatively, less colloquial: yayukar yarakop re yasakur lit., it becomes it is difficult that he swallows.

The ni-forms of adjectives are also nominalizations, and may occur by themselves-- e.g. navani ri holundaye, nipola cili 'the distance to Holunda is 50 cili ("km")' (<nava 'far')-- this is colloquial, however, and synonymous andava(ni) could be used. (Note that these ni-forms, while nouns, are indeclinable, and their exact meaning with prepositions must depend on context: ri navani 'in the distance, at a distance' or 'into the distance'; alo navani 'from a/the distance'.) Or, kunironi niya katrayi 'his/her/the [subtly bi-colored (fur)] is very attractive'. More often, however, nouns and adjectives with -ni occur with a noun complement, in what we have been calling "the ni-phrase or construction".

4.3 The ni-phrase. We have already seen examples of this usage in compound prepositions with accusative, dative or genitive cases, of the sort ri vakani...; ri vunuweni...; alo ondeyini... 'under ~at the bottom of; toward ~in the direction of; from the inside of...'-- these are cases of NOUN-ni NOUN-- where the complement (in these cases, the true object of the preposition, which may be human) is in the nominative case. And in §3.4b above, and elsewhere, we have alluded to its use in expressing (in a very general way) a possessive relationship between two inanimate nouns-- these may involve NOUN/ADJ/VERB-ni NOUN; as a rule, the complement may not be human (the genitive case is used then). We have also pointed out that this usage is largely colloquial, though not always-- (1) sometimes there is no corresponding synonymous añ- form of the adjective or verb, or (2) the añ- form has a different meaning. Consequently in these cases, a human complement may occur. Some typical examples:

- puhi yamaceto nihiñi lacani 'alcohol will damage the top of the table'
- vakani ekuk yu yavecut 'the seat [bottom] of that chair is dirty'
- landruni surañ ambo, sila amba harongo li 'the height of Mt. Ambo is 3400 li' (the nominal andandru means 'heights, high place(s)')
- lusni etengi 'the end (last part) of the book; the ending of the book'; livekni acivandri 'the length of his lecture'
- kaçetni... 'a picture of...' may take a human complement, kaçetni mina 'a picture of Mina'-- contrast kaçet minayi 'Mina's picture' (she owns it, or she drew it)

Other ni-phrases that may take a human complement are those which describe physical attributes-- height, weight, age, etc. The ni-phrase, in fact, is the only way to express these:

- vitalni (~amitalni) çenji, mesa kik fanu li 'Shenji is 1.8 li tall, Shenji's height is 1.8 li'. Colloquial vitalni is used far more often than the nominal; even in official documents. Alternatively, the name may be topicalized: çenji, vitalni mesa kik fanu li. Note that while çenji yavital means 'Shenji is tall'-- there is no way to include the number in that sentence.
- kicatni (~angicatni) veveni, sila cipem ~veveni, kicatni... 'the/her baby weighs 3 cipem [about 9 lbs.]'

In dealing with age, the most neutral term is pehañi (some prefer the plural pehanaçni) 'year(s) of...': pehañi erek, rofola ~erek, pehañi rofola 'Erek is 20'. When discussing people who are known to be, or can be assumed to be, of relatively advanced age (say, 40 and up-- parents with adult children, grandparents, senior professionals, etc.), marokni... 'oldness of...' may be used-- older people are highly respected, and marok has connotations of dignity, learning or knowledge of the world, gravitas in short, that younger people simply have not achieved. Note, however, that marok is used in relation to non-humans (often implying they are past their prime) and inanimates; it is also used in comparisons, though careful speakers will observe the pehañi/marokni distinction with humans:

- marokni puna yu, nim pehan 'that house is five years old'
- pehañi/marokni çenji, lavi alo minayi ~çenji, pehañi/marokni lavi alo minayi 'Shenji is older than Mina'; also possible çenji yalamarok alo minayi (lamarok is an irregular comparative)

Similarly with height, weight etc.:
- kicatni amami, lavi alo amayiti (~ini hati) 'my father weighs more than your father (~yours)'; also possible: amami lavi yakicat alo... 'my father is heavier than yours'
- vitalni çenji, lavi alo ereki ~çenji lavi yavital alo ereki 'Shenji is taller than Erek' (See further §6.2)

These constructions are also used with the interrogative fiyan 'how much, how...', as in--

- fiyan vitalni... 'how tall is...?'; fiyan vitalti 'how tall are you?'; also possible fiyan yavital 'how tall is he ~are you?' etc.
- fiyan kicatni... 'how heavy is..., what does ...weigh?'; also fiyan yakicat
- fiyan pehañi..., fiyan marokni... 'how old is...?' (fiyan yamarok is possible)

Note also:
- fiyan çarekni (~fiyan yaçarek) 'how strong is he?'
- fiyan livekni (~fiyan yalivek) 'how long is it?'
- fiyan çangini (~fiyan yaçangi) 'how sick is he?''
- fiyan toyeti 'how much money do you have [with you]?'
and many similar uses, especially in colloquial speech.

4.4 Idiomatic usages. The ni-form of some words may take other meanings; many of these established forms show an older but now irregular final-consonant+ni sandhi. For example:

- ciyoñi (<ciyon 'to seem') 'apparently, seemingly'; but also 'it seems...'
- ñupindi (<ñupit 'accustomed', with irreg. sandhi) 'usually'
- nulañi (<nulañ 'sad') 'unfortunately'; but also 'the sad part is...', and in some cases (colloq.) 'sadness' (syn. andulañ)
- vihatni (<vihat 'specific, definite') 'specifically; i.e., that is, to wit'
- ombi (<om 'reason', irreg. sandhi) 'because' (conj.); also ombini 'because; the reason is...'
- livengi <<livek 'long', irreg.) 'along' (as a preposition, ri... is optional-- miharakaran (ri) livengi vori 'we strolled along the river'; it can also mean 'front/side' as in livengi hacu, livengi vori 'beachfront, riverfront' and in such cases is sometimes used as a street designation)
- ritandi; riyandi 'right here; right there' (<ritan, riyan)
- oranani ~orandi 'just like..., just as...' (<orana 'like, similar')
- kunangi (~kuna, kunani) 'about, approximately...' (<kunak 'maybe, perhaps')

5. Accidental (caka-) verbs

As was mentioned in the Morphology, this prefix has the following forms:

These verbal derivatives have an essentially passive meaning; they describe actions or states that affect or characterize their subject, usually without any control on the subject's part-- "SUBJ. is overcome by..., afflicted with..., affected by..., obsessed with..., is...by nature" etc., and the experience is viewed as unexpected, unpleasant (broadly speaking) or culturally negative. Most often, the base is an adjective or intransitive verb; but transitive verbs, and even nouns, may occur. Since the base generally indicates the cause of the experience, there is no need to indicate an actual "agent/cause", but it is possible to do so with e.g. a noun in the genitive case, a prepositional phrase or sentence complement.

There are just a few forms with mostly positive connotations, but even here, the idea of suddenness/unexpectedness is present, or sometimes perhaps a slight tinge of envy or resentment. Consider:

- cakaminda 'overjoyed, elated' (<minda 'smile, be happy'-- no negative connotations at all!)
- cakatanja˝ 'to intuit; gain insight into' (<tanja˝ 'understand')
- cakatikas with dative 'to see/understand suddenly, to have a flash of insight'; with nominative 'to appear suddently/unexpectedly (perhaps inopportunely, as well)' (<tikas 'see')
- cakupan said of s.o. who seems always to get what he wants, for whom everything always goes right, lucky-- usually admiring, but sometimes envious or resentful (<upan 'receive, get')
- cakiyo˝ (of a surface) 'highly polished', (of a person's manners/behavior) 'highly refined, elegant' (<iyo˝ 'glossy, shiny'--perhaps just a bit too much...-- contrast (<tevas 'fine, delicate') cakatevas 'overly refined, affected behavior')
- cakapanip 'be tricked, fooled; be the victim of a practical joke' (in a playful, non-malicious way, though the victim might be embarassed; <panip 'to play')-- contrast (<hura 'to mock, laugh at') cakura 'be/get mocked (maliciously)' or (<onji 'feign, pretend') cakonji 'fooled, taken in, cheated, swindled')
- cakatuša 'to be a perfectionist' (<tuša 'exact, precise'-- not necessarily a bad thing, although sometimes annoying to others)

As has been mentioned above, the human subject of a caka-form is either in the dative case (see §3.3e above)-- if the state is viewed as inherent, chronic, serious or permanent-- or in the nominative (see §3.1e above)-- if it is only temporary, or less serious. Dative subjects are more common, and over time, have come to be used with some "temporary" conditions-- one might speculate that the "when in doubt, use the dative" rule has operated here; but the converse is true, too.

Non-human animate subjects may also be either dative or nominative, with the same conditioning, but there is a tendency in colloquial speech to use the nominative most of the time. Inanimate subjects are always in the nominative case. Nominative subjects may in all cases be indicated with a pronominal prefix in addition to the noun or pronoun subject.

Caka-forms do not generally use tense markers, nor do they generally occur with the modal auxiliaries. They may, however, be modified with adverbials like niya, kundak, kunak, uti etc. 'very, not very, perhaps, rather...'; and some may have a noun complement, e.g. cakonji toye means specifically 'swindled financially, swindled out of one's money'. They may sometimes be translated as nouns, e.g. cakambo 'acting superior, snobbish': yanda sambat cakambo! 'don't be so snobbish ~don't be such a snob!'.

The prefix is very productive; in addition to the well-established forms listed in the dictionary, whose meanings are familiar through long usage, forms are often created on the spur of the moment-- for example, from cika 'TV' one could form cakacika to describe someone who spends altogether too much time watching TV; or from fanu 'eight' perhaps cakafanu 'obsessed with the number eight'; or cakanini(-nini) 'to over-use/mis-use the ni suffix', mentioned above, coined by schoolteachers. Such nonce-forms may or may not find wider usage. Thus it can happen that the actual meaning of a caka-form sometimes seems quite far removed from the meaning of its base.

In addition to examples already given, we will list a few more typical ones here:

- cakaya 'blinded by a bright light' (<haya 'light')
- cakinja 'oppressed by; oppressive, repressive' (<kinja 'to lie on')
- cašendek 'stubborn, inflexible' (<šendek 'strong-willed, determined')
- cakacoyok 'very complicated; completely confused, at a loss' (< coyok 'complex, complicated')
- catrasosir 'go one's own way, refuse to take advice; wrong-headed' (<trasosir 'disobedient' <sosir 'to hear')
- cakatovar 'to hope in vain' (<tovar 'to hope')
- cakanava 'unattainable; beyond all expectations; extraordinary' (<nava 'far')
- cakanuro (of a ship) 'lost at sea' (< nuro 'deep')
- cakripa 'get caught in the rain' (< ripa 'rain')
- cakalero (dat.) 'suffer a sun-stroke', (nom.) 'stay out in the sun too long' (<lero 'sun')
- cakafonjip 'touched or fondled in an inappropriate way' (<fonjip 'touch, feel')
- cakamola (dat.) 'overly conscious of/impressed by wealth', (nom.) 'overly concerned with the price of things; cheap, a cheapskate' (<mola 'value')
- cakambik 'prudish, priggish; overly concerned with appearances' (<ambik ''fitting, proper')
- cakuri 'to write badly, scribble, scrawl' (<uri 'to write')

6. Adjectival expressions.

6.1a. In §1.2 above, we discussed the usage of adjectives as modifiers (attributes) of nouns; the correct word order is noun + adjective. Adjectives themselves may be further modified with qualifiers like niya, uti 'very, rather', most of which follow the adjective. And adjectives may be used nominally ("the...one") in combination with forms of the demonstrative pronouns, e.g. yu vanat 'the white one (inanim.)', taye vital 'this tall one (fem.)'; and nominalized with the -ni suffix or a˝- prefix, e.g. nuroni, anduro 'depth'. In many cases, the a˝-form may have acquired a specific meaning-- e.g. from vanat 'white' we find amanat 'whitewash; (modern usage) white paint'.

6.1b. When used predicatively, adjectives behave like any other verb-- they take the personal prefixes, tense markers, agree in number with their subject, and may have inchoative and causative derivatives. A typical example is paha 'open':

- findu yapaha 'the door is open' (state- intransitive)
- findu yaçupaha 'the door opened (i.e. became open)' (inchoative- intransitive); note also findu yatata/yarakop (ya)çupaha 'the door opened easily/with difficulty'
- šenji yarumbaha findu 'Shenji opened the door' (causative- transitive. note that an inanimate object may also serve as subject, 'key, crowbar; magic word'); and note yatata/yarakop re marumbaha findu 'it was easy/hard for me to open the door ~I opened the door easily/with difficulty'
- cf. also the accidental findu cakapaha 'the door [suddenly/inexplicably] flew/burst open'

or vošil 'wide':

- ritan, vorini ta yavošil 'here, the river is not wide'
- vorini yaçuvošil rimorani ecak yu 'the river widens (become wider) near that village'
- poro irumošil ratuni 'they're going to widen the street'

6.2 Comparison of adjectives.

6.2a. The comparative degree is expressed with lavi 'more' and omban 'less'-- in descriptive use these follow the adjective: puna velu lavi 'a newer house', ametrita pešaka omban 'a less powerful government'. (Colloquially, lavi/omban often precedes the adjective, following usage when relativized-- puna re lavi (ya)velu-- or when used predicatively.) The superlative uses krat 'most' and imik 'least', which precede the adjective, and the adjective phrase is often relativized-- puna krat velu ~puna re krat (ya)velu 'the newest house', ana imik katrayi ~ana re imik (ya)katrayi 'the least attractive child'. (Note: the deletion of the personal prefix in these relative clauses is usual in colloquial speech.)

A handful of common adjectives have irregular comparatives, with prefix la- for the comparative 'more', ca- for superlative 'most'. In the past, there were more such forms, but only these are in current use:

Note also the verb lalisam 'to prefer'-- it is an old comparative based on lisam 'to like' (though lalisam and lavi lisam...alo 'like X more than...' are not quite synonymous); when used in the sense "prefer X to/over/rather than Y" or, "would rather X than Y" it takes its complement with alo in careful usage (but more commonly with liri):--

- yalalisam poren pambara alo poreni troše (~...aloni troše) 'he prefers black [i.e. red] wine over green [i.e. white] wine (~...over white)' versus yalalisam poren pambara liri poreni troše (~liri trošeni)
- malalisam manolit alo re (~lire) matingas cika 'I'd rather read than watch TV'

It may of course be used without a complement: yalalisam poren pambara; malalisam manolit 'he prefers red wine; I'd rather read'

When modifying predicates, lavi, omban, krat, imik all precede the verbal form, as do most other modifiers. Thus parallel with sentences like šenji niya yavital 'Shenji is very tall', we find:

- šenji lavi yavital aloni (~alomi, [alo] ereki, alo yunoni) 'Shenji is taller than him (~me, Erek, than everyone/all of them)'-- without the alo phrase, šenji lavi yavital is colloquial for 'Shenji is taller [than someone, understood]' and also for šenji yale kaš/ya vital lavi 'Shenji is the taller one', e.g. when pointing out Shenji in a group of people.

- šenji krat yavital alo yunoni 'Shenji is the tallest of all'
- šenji krat yavital (alo) ecaki 'Shenji is the tallest in the village'-- again, without the alo phrase, šenji krat yavital means 'Shenji is the tallest [of an understood group]' or a paraphrase of šenji yale kaš/ya krat vital '...is the tallest one [pointing him out]'

Note the various formations of the alo phrase (in comparative sentences):

(Colloquial speech does not always observe the distinction alo + suffix vs. alo re + pronoun, relying more on context or the underlying logic of the situation. Thus we find šenji lavi yasisa minaye aloti with both possible interpretations out of context-- or lavi yanahan aloti 'he [tends to] eat more than you [do]' or yanahan lavi(ni) aloti 'he ate more [of it] than you [did]' where, clearly, the interpretation "he ate more [of it] than [he ate of] you" is situationally impossible, except perhaps in a story or joke about cannibals.)

6.2b. Comparatives used adverbially. This topic will be discussed more fully in §§7,9 below. For now, we give a few examples of the usage:

- šenji yayama (picikni) lavi yahulu˝ alomi 'Shenji ran (a little bit) faster than me'-- note that technically this should be ...alo re mam, but colloquial usage has prevailed in these constructions.
- mina yarinju (pando/niya) yalale˝ alo taniyi 'Mina sings (far ~much) better than Tani'; also possible: colloq. mina yarinju lale˝/cale˝ 'Mina sings better/best'

As the examples just above showed, lavi by itself may also modify a main verb, as in lavi yasisa... 'love (X) more (than Y)', lavi yapila-pila alo re yamepu 'he thinks (tends to think) more than act'

6.2c. Ambiguity of -ni. When all referents in a comparative sentence (and in complex sentences as well) are 3rd person, some verbal prefixes and aloni may be ambiguous without context. Thus šenji lavi yavital aloni may mean '...is taller than him/her/it, that/them'; for clarity one would use alo iyani/iyeni/nili (or a noun or personal name) '...than him/her/them', reserving aloni for 'than it, than that'. This last would be used mostly to correct someone's mis-statement, e.g.:

A: pilami, vitalni šenji kuna mesa li angunjo 'I think Shenji's about 1.5 li tall'
B: tayi, lavi (vitalni) aloni 'No, he's taller than that'; with surprise and emphasis, nandayi, lavi (vitalni) alo iyu! 'Certainly not, he's taller than that! (alo iyu is idiomatic)

Because two finite verbs in a row always refer to the same subject, only aloni would be ambiguous in:

- šenji yapila lavi yavital aloni (~alo iyani/iyeni etc.) 'Shenji thinks he is taller than (...)'; similarly šenji lavi yavital alo pilani 'Shenji is taller than he (Shenji) thinks'

- šenji yapila krat yavital is unambiguous: 'Shenji thinks he (Shenji) is the tallest'

However, because re generally indicates a change of subject, šenji lavi yavital alo re yapila 'Shenji is taller than he thinks' will generally be taken to mean '...taller than he (someone else) thinks', though it could be misread as '...than he (Shenji) thinks'. Unambiguous would be: ...alo re (iya/iye/nila/amani etc.) yapila '...than he/she/they/his father thinks', as well as ...alo pilani amani (but not *pilani iyani etc.-- these pronoun forms do not co-occur with ni-forms).

A sentence like šenji yapila re lavi yavital aloni clearly indicates that the subjects of pila and vital are different, but aloni is now multiply ambiguous: 'Shenjia thinks that heb is taller than [?]' or:

  1. Shenjia thinks (b) is taller than [Shenji himself], or
  2. Shenjia thinks (b) is taller than [someone elsec], or
  3. Shenji thinks (b) is taller than that [correcting a mis-statement], or even
  4. Shenji thinks (b) is taller than [(b) actually is]

Aside from using nouns or personal names, these can be disambiguated as follows:

  1. šenji yapila re lavi yavital alonitu ~alo initu ('...than he (Shenji) himself')
  2. šenji yapila re lavi yavital alo iyani/iyeni ('than he/shec'-- in most contexts, the usual interpretation)
  3. (no change is necessary; ~...alo iyu for emphasis)
  4. šenji yapila re lavi yavital aloni kundrini ('than heb [is] in fact') or ...alo vitalni kundri '...than his true/real height'

6.2d. Comparison of equality. To indicate that two or more comparanda possess some quality in equal degrees (i.e. "X is as...as Y") Kash uses the verb kuwa 'equal, same' in a compound (serial) verb construction:

- šenji yakuwa yavital alo ereki 'Shenji is as tall as Erek'-- Or a plural subject may be used:
- šenji (i) erek ikuwa ivital 'Shenji and Erek are the same height, are equally tall'.
- šenji (yapila) yapole yayama lavi yahulu˝ alo ereki 'Shenji (thinks he) can run faster than Erek'-- technically correct, but most speakers would probably delete some of the ya- prefixes, or rephrase-- šenji (yapila) pole yama lavi hulu˝ alo...

It also possible to substitute ni-forms: ...yakuwa vitalni alo...; pilani šenji, pole yayama lavi (ya)hulu˝ alo...; vitalni šenji i erek, yakuwa etc.

There are rather idiomatic expressions for "as...as possible" (or similar translations): ADJ krat ADJ. These generally follow the main verb. Thus we find:

- hulu˝ krat hulu˝ 'as fast/quickly as possible, as fast as can be, as fast as one can'
- kumo(r) krat kumor 'as silently as possible'
- le˝ krat le˝ 'as well/good as possible; the best one can'
- muko krat muko 'as badly as possible'
- tanju krat tanju 'as soon as possible' (tanju 'now' for 'soon', used only in this construction, and in lavi/krat tanju) 'sooner, soonest'; 'soon' itself is cumicu 'in a little while, shortly')

and many others; see also §7 Adverbial modifiers.

6.3. Miscellaneous uses of the comparative.

6.3a.lavi, krat 'more, less', and less commonly omban, imik 'less, least' may be used by themselves as adjectival modifiers, and (usually with -ni) as nouns in their own right; e.g.

- mapita toye lavi 'I need more money'
- yameloto uku omban 'he wants/would like less uku' (often paraphrased ta yameloto sambat uku 'he doesn't want so much uku')
- erek yanahan krat kumus 'Erek ate the most cake' (cf. ...yanahan lavini kumus and ...yanahan kratni kumus both '...ate most of the cake', but lavini implies "a good part of ~more than he ought" while kratni implies "a very large part of, almost all of...")
- minuša rofola, omo ifilan omba˝i 'we ordered 20, but they sent less/fewer' (or paraphrased ...omo tak ifilan timbani '...but they didn't send enough'
- yatraka lavini/kratni 'he bought more (of it)/most (of it)'

As we saw in §6.2b, lavi, krat, preposed, may modify a main verb; similarly, lavini, kratni may serve as sentential modifiers, with rather variable meanings:

- lavini: 'more often, by and large, in a majority of cases' etc.
- kratni: 'mostly, most often, for the most part' etc.

See further §7.

6.3b.lavi and omban are also used in a construction meaning "the more [X]...the more [Y]", as follows:

- lavi yayama šenji, lavi yaçupori 'the more Shenji ran, the more tired he became'-- colloq. lavi yamani (šenji), lavi porini
- lavi mile yarungaya mina, omban ne mivacan 'the more Mina told us [about it], the less we believed her'-- colloq. only a paraphrase is possible, since the causative verb cannot take -ni: lavi kotani, omban vacambim 'the more she said, the less we believed' or lavi kayamim, omban vacambim 'the more we knew, the less we believed'
- lavi yaçumarok, lavi ne cakacut 'the older he gets, the more forgetful [he is/becomes]' or equivalent lavi yaçumarok, omban yanimbur '..., the less he remembers'-- colloq. lavi marokni (~çumarok), lavi cakacutni
- lavi kašni, lavi ilongoni (purely colloq. and idiomatic) 'the more people, the more fun ~the more, the merrier'
- omban kotani, cale˝i 'the less said, the better'

7. Adverbial and sentential modifiers.

These include independent lexical items (usually classed as "hambiyaš" 'pieces', or particles); compounds, reduplications and derived forms; and prepositional phrases. They modifiy the verb (or in some cases the entire sentence) with respect to quality or quantity, time, location, or manner; some can function in more than one category.

7.1 Adverbs of quality/quantity. These most often modify the verb, and are usually placed directly before it (or before the verb phrase if there are object pronouns present), and include such words as--

- niya 'very', niya-niya 'extremely'
- ukat 'quite; very'
- yuno-yuno 'completely, totally; absolutely'
- kundak 'not very' (a polite, indirect 'no, not')
- uti 'rather, somewhat'
- sut 'too'
- mora(ni) 'almost'
(The above are restricted to verbal/adjectival modification. The following may modify verbs/adjectives; as noun modifiers they act as adjectives (in these two cases they are preposed); and they may sometimes function as nouns.)

- supat(ni) 'too much'
- sambat(ni) 'so much, so'
- pando(ni) 'much, a lot, greatly; by far' (usually in comparative sentences)
- napo ~ -po 'only, just' (the full form is more emphatic)
- napi ~ -pi 'barely, scarcely' (the full form is more emphatic)
- picik(ni) 'a little bit, slightly'
- timbani 'enough'
- tušani 'exactly, precisely' - lavi(ni) 'more', krat(ni) 'most, mostly'; lavi-omban 'more or less'
- imikni 'at least'
- kuna˝ ~kuna(ni) 'about, approximately'
- cinyonjiyon ~ciyo˝i 'apparently'

7.2 Adverbs of time. Included here are a number of particles, and some derived forms; prepositional phrases also function as time-adverbs. In simple sentences, the particles and derivatives tend to occur pre-verbal; in more complex sentences their placement is freer, as is the placement of prepositional phrases in all cases.

- letrayu 'today'; koprat 'yesterday', kopramés 'day before yesterday'; sapat ~letrata 'tomorrow', sapamés 'day after tomorrow'
- tanju (~formal tayanju) 'now', tiyanju 'then', and various derivatives: tatanju 'right now, just now', titanju 'just then'
- tikulu˝ 'straightaway, immediately'
- cumicu 'soon, in a little while'-- but note lavi/krat tanju 'sooner, soonest'
- tayondre ~condre 'tonight', onjos 'last night'
- cis 'again', cicis 'again and again, over and over'
- anje 'still, yet'
- tamende ~tamét 'not yet'
- nunu (free form of -nu) 'often'
- yunda 'always'
- lunda 'at some time, ever; once, formerly (in the past); used to...'; talunda 'at no time, never'
- pandašu '(for a) long time'; pandacosa 'a long time ago'
- kariyanju 'sometime, someday (vague)'; katrelo 'someday' is considered more definite

- angasini 'initially, first'; less formal, mesa-mesani; counting in series, one may use kumés
- šelu˝i 'afterwards, next'; šelucelum 'later on'; tayanju šelu˝i 'from now on'
- kakandi 'beforehand, earlier'
- lulusni 'finally, at last'
- veluni 'recently'

All numbers and time-words take on adverbial meaning when reduplicated (full or partial):
- mesa-mesa 'one by one, singly', loro 'two by two, by twos' etc.
- nasa-nasa ~nanasa 'minute by minute, every minute'
- lero-lero ~lelero 'daily, every day'
- sote-sotrelo ~tetrelo 'weekly'
and so forth

With measures of time, cosa ('go') means 'last..., the past..., ...ago'; rata ('come') means 'next..., the coming..., ...from now'

Prepositional phrases, of course, can also indicate time, as for instance, ri lusni pehan yu... 'at the end of that year...', riyanjuni ašuprót 'during 10th Month'.

7.3 Adverbs of location. Mostly particles, a few derivatives, and prepositional phrases.

- ritan 'here', riyan 'there (but nearby), riyun 'yonder, far away'
- ritandi 'right here', riyandi 'right there' (irreg. sandhi with -ni)
- karinena 'somewhere'
- vitani 'up, upwards', vahani 'down, downwards'
- ri kandini..., ri šelu˝i... etc. 'in front of, in back of'

7.4 Manner adverbs. Aside from lembo 'well' and sa˝ 'thus, like this/that', these are derived from the corresponding adjectival form in an unpredictable variety of ways. Most correctly, one uses a compound verb structure-- yašindi yakret 'he spoke bitterly' (lit., he spoke he was bitter); yapole yayama yahulu˝ 'he can run fast'. More colloquially, one might use a reduplication-- yašindi kret-kret; yapole yama hulukulu˝; many such forms are established from long usage, others may be nonce-forms. Many reduplicated adjectives imply some modification of the base-- e.g. hahaya 'brightly lit' (<haya 'light'), mara-marak 'very angry'. In some cases, a ni- or caka-form might be appropriate-- hakasni yašufat ~šufatni (ya)hakas 'he is behaving strangely ~his behavior is strange', cakacut yarundayo etengini ri ˝akanga 'he absent-mindedly left his book in the taxi'. And in a few cases, the adjectival form is used without any modification: fatap tarambu 'strictly forbidden'

8. List of common prepositions and conjunctions

8.1 Prepositions. Bear in mind that ri and its compounds vary in meaning depending on the case of the following noun-- accusative for location, dative for movement.

- ri (acc.) 'in, at, on'; (dat.) 'to, into, onto'
- alo (gen.) 'from, out of, away from'

These combine with location nouns (in the appropriate case form and usually with -ni) to form compound prepositions; the "object" of these prepositional expressions is in the nominative case.
- (ri, alo) +hambokni..., mu˝okni... 'at/to/from the right ~left of...'
- +kandi-ni... '...front of..., before...' (in expressions of time, kandi˝, e.g. kandi˝ aro folanim; kandi˝ ašurak rata 'before 15 o'clock; before next month')
- +šelum-ni... '...back of, behind, after' (in time expressions, šelu˝i alone is used, e.g. šelu˝i aro mepola; šelu˝i cindero 'after 10 o'clock; after Mid-month day')
- +cini-ni... 'middle of..., midst of...'; with two objects, it may mean 'between'
- +nihin-ni... 'top of..., on (top of)..., above...'
- +vita-ni... 'up..., upwards..., upper surface of...'
- +vaka-ni... 'bottom of..., below..., under...'
- +vaha-ni... 'down..., downwards...'
- +andopra-ni... 'other side of..., across...'
- +livengi... 'along (the length of)...'
- +mora-ni... 'nearby..., near..., vicinity of...'
- +vora-ni... 'side of..., beside, next (to)...'
- +vunu-ni... 'direction of..., toward...'
- +onde-ni... 'interior of..., in, within; into, inside; out of...'
- +ambra-ni/umbrik-ni/aros-ni/ures-ni... (compass points) 'north/south/east/west of...'

Bear in mind, too, that many Kash verbs have a prepositional sense as part of their basic meaning, e.g. numba 'to ride (on)', menjar 'to climb, go up (s.t.).'

Other prepositions and specialized usages:

- yam (acc.) 'with (accompanying)'
- ri anjuni... (acc.)--often written as one word, riyanjuni-- 'at the time of..., while..., during...'
- alo anjuwi(ni) 'since [+time expression]' (colloq. alanji +acc.)
- yambit, umit (acc.) 'with (by means of)
- liri (acc.) 'concerning, about, with respect to'
- ušo˝i (gen.) 'for, intended for, for the purpose/use of...'
- inga (gen.) 'without'
- ombi alo... (gen.) 'because of...'
- ri + (person, gen.) 'at (person's) home, place'
- ri arani (trade name, nom.) 'at the (X) shop ~place of business'
- ri (dat.) ri (dat) (formal) 'from (noun) to (noun)' e.g. ri angeye ri angeye 'from tree to tree'; in less formal speech/writing, the second ri is omitted, and -ni added: ri angeye angeyeni 'from tree to tree'; the construction implies random movement. When the movement is from one specific place to another, or involves a specific distance, alo (+gen.)...ri (+dat.)... would be used: yanecu alo angeyi yu ri ine liyani 'it jumped from that tree to the/that other one'; alo punayimi ri (ine ~punaye) šenjiyi, navani kuna mesa cili 'from my house to Shenji's (house) is about 1 cili ("km.".)' (There are some colloquial usages, in which ri may be omitted altogether, such as (ri) mesani liyani 'from (the) one to the other ~from one to another' and (ri) kasini lusni 'from beginning to end').

8.2 Conjunctions and other sentence connectors.

- i 'and'-- it is frequently omitted between nouns, adjectives, and clauses with the same subject
- omo (~mowa; colloq. mo) 'but'; usually first in its sentence or clause, but it may follow the subject/topic where it corresponds more to 'however'
- mowa-mowani 'but rather, on the contrary, instead'
- iti 'or'; iti...iti... 'either...or...'; also used colloquially as a question tag, e.g. cosaka, iti? 'are you going, or not (~or what)?'
- (i) tati 'neither, nor..., not...either'; tati...tati... 'neither...nor...'
- pun 'if, whether'; tatipun 'if not, or else', colloquially 'whether or not, regardless'
- oranani ~orandi 'like, just as'; orambun 'as if'
- kambun 'even if, although'
- ombi (~ombini) 'because, since'
- aloni 'therefore'
- vara 'in order to, so that'
- eme 'also'; eme...eme... 'both...and...'
- tambo...(i) eme... 'not only...but also...'
- anju 'when'; anju-anju 'whenever'; riyanju(ni) re... 'while [+sentence]'
- kandikre (<kandi˝ re) 'before [+sentence]'
- šelu˝i re 'after [+sentence]'
- alo anjuwi re... 'since [+sentence]'
- yangeti re 'on condition that..., provided that...'
- kar 'what, that which...', pl. karaš 'those things which...'
- are ~re 'that...; who, which' introducing subordinate and relative clauses

Common sentence introducers, though not, strictly speaking, conjunctions:
- takanda 'nevertheless'
- lavi-lavini 'moreover, in addition'


Return to:

Eventually...Forward to: