3. PREVLI SYNTAX

The basic order of constituents in a Prevli sentence is: (adverbial modifier) + Verb + Subject + (noun Direct Object) + (Indirect Object) + (modifying prepositional phrase). We noted earlier that most adverbs precede the verb-- in particular single-word forms; adverbial prepositional phrases usually come either at the beginning or at the end of the sentence. Noun modifiers (adjectives, numbers/quantifiers, demonstratives) precede their noun. We will now examine Prevli sentence structure in more detail than has been possible in the short examples given in the Morphology section. We will look at--

1. simple declaratives with: (a) intrans.vb. (b) copula/existential heli/el (c) trans.vb.
2. multiple subject/object; conjoined sentences
3. comparatives and superlatives
4. "surprise" forms
5. relative clauses // active/pass.
6. questions
7. sentential/clausal complements and the aspectual forms
8. indirect speech/orders/questions

3.1. Simple declarative sentences-- V S (O) (IO), with optional modifiers

3.1.1. Intransitive V S: As has been mentioned before, if the subject is a pronoun, the verb must have either a pronominal suffix (identical with the possessive suffixes), or a full-form pronoun (usually but not always for emphasis); if the subject is a noun, no suffix is used:

--muko korvå zek ~ muko korvåz 'he runs well'
--korvåd yän goz o hupoz 'Yän ran toward the (his) longhouse'
--us afte ot malat 'the ropes are not tight'
--muhu neab ik peag 'this knife is very sharp'
--lavmalá zek diez näki 'his car is (sort of, rather) old'
--dolo galaŋ potsa 'it [its color] used to be red'

3.1.1a. Intrans. V S IO. Verbs meaning 'speak, say, tell, ask' etc. may have as IO the person addressed, e.g. bilda garo kitämni 'Garo spoke to Tä:mni'; monolka kitämni 'I'll ask Tämni'; additional material may of course follow: bilda garo keśké ut o halde 'Garo told him about the festival'; monolka kitämni pun mekovoz 'I'll ask Tämni if he wants to go'. Thus the IO must appear with the preposition ki- if it is a noun, or, if it is a pronoun, with k- +oblique form of the pronoun. (The full form ki may also be written as a separate word.)

It is also the preferred construction when both subject and IO are pronouns-- bildaz keśké 'he spoke to him', monolka keśké 'I'll ask him', kam tu semik kiblí uto 'I've already told you about it'. However, in everyday speech, in just this case, it is possible to use the compound pronoun forms, even though they normally indicate subject+direct object: thus, bilda kaz ut... 'I spoke to him about...', monor zehen pun... 'he asked him if....' (NB if a DO is present, this constructon is not possible; the sentence becomes transitive and must have V S O IO structure; see further below on Transitive Sentences.)

3.1.1b.1. Copular/existential sentences. Prevli has the copular verb /heli/ irrealis /ehli/, but it is rarely used, mainly when introducing oneself or another-- heli kan tämni 'I am Tämni', hel ik lüsi 'this is Lüsi'; also heli kan 'it is I, it's me'-- or identifying someone's role or profession-- heli ok bapak 'that is my father', heli garo menek 'Garo is my friend', heli zek sando 'he's a shaman'. These are considered rather formal; in colloquial usage they are usually rephrased-- heli is omitted and the word order reversed (the intonation changes, too)-- tämni kan, bapak oku, menka garo, sando zek. An alternative is to use el (see further below).

Irrealis /ehli/ (also rather formal) can also be used: ehlir kan 'it was (probably) me', ehli garo menez tämni 'Garo is probably a friend of Tämni's', ehliza sando 'he's probably a shaman'. But the colloquial versions of these must use the irrealis of /el/.

3.1.1b.2. Far more frequent in speech (though less so in formal settings) in such sentences is /el/ irrealis /on/ 'there is/are....', which can substitute for heli in the cases just mentioned-- thus, el kan 'it's me', el garo menek 'Garo is my friend', elda zek sando 'he was a shaman'.

Other uses of /el/ include

(1) its customary "existential" sense--
--el lopot i ratu 'there are lopas in the road'
--uselda duŋ nuvit i linu 'there were no clouds in the sky'
--el purán i ok tampig 'there's a restaurant in that town

(2) with adverbs and prepositional phrases--
--el epi zo daksa 'it's up on the roof'
--elda kant olta i köin 'we were all in the yard'
--irik usel zek 'he isn't here'

(3) to describe weather conditions--
--el yana 'it's sunny', el nunuvi 'it's very cloudy', dolo mitsä elda loo 'it rained earlier this morning'

(4) with a possessive or with a dative, to indicate possession (temporary, not ownership)--
--el kan diek toe ~el toe kagná 'I have money (with me, I can pay)'
--el zek diez peag ~el peag keśké 'he has a knife (that he can use)'

More rarely, this construction can be used of inalienable nouns, and indicates, of couse, some inherent quality. In this case the dative is unnecessary--

--el nam homnå erniz (lit. there-is one only his-arm) 'he has only one arm'
--el ultå doplåz huloz (lit. there-are five fingers-poss. hand-poss.) 'the hand has five fingers'

In all these cases, the irrealis /on/ may be used, as needed.

Finally, /el ~on/ is used when a noun or phrase is fronted for topicalization or emphasis-- but this usage will be discussed later.

3.1.2a. Transitive sentences V S O (IO). In active voice, these always require the presence of a subj/obj. pronoun, even when the subject is a noun (it immediately follows the pronoun):

--la:ter kaz a garo 'I saw Garo'
--la:ter zehen lüsi a garo 'Lüsi saw Garo'
--la:ter sehak garo 'Garo saw me'
--uśtuhi zeknät at uśänta 'we don't like outsiders' (i.e. usually, Kash people)
--zizer zento nuśno yadze 'they ate most of the food'

Trivalent transitive verbs, like "give", require an indirect object: donol ko toe keśké 'I will give him (some) money'; only in casual speech, may an IO be omitted if it has been mentioned or is understood--donol ko ultå truni 'I'll give 5 truni (the national currency)'. (There is a formal derivative of /dona/-- causative yabdono-- that means specifically 'to donate', and it may occur (correctly) without an IO.)

The verbs "sell" and "feed" may be either bi- or tri-valent:

--tun akpo hen o zek diez näki 'he has sold his car'
--tun akpo hen o zek diez näki ki garo 'he has sold his car to Garo'
--padzer zehen lüsi a titiz 'Lüsi fed the (her) baby'
--padzer hen lilsi ki titiz 'she fed her baby lilis'

--l,ä:ter garo (nagná)'Garo was seen (by me)'
--z,izer nuśno yadze 'most of the food was eaten'
--d,önor toe keśké (nagná) 'some money was given to him (by me)'
--tun yäkpo zek (~tämni) diez näki (ki garo) 'his (~Tämni's) car has been sold (to Garo)'

An indirect object may also be promoted to subject in a passive sentence; it that case it is fronted, and the verb takes the special suffix -zi (for all persons):

--zek d,önorzi toe (nagná) 'he was given money (by me)'-- note kan d,önorzi... 'I was given...; min d,önorzi... 'you were given...'
--titiz p,ädzerzi lilsi (ni lüsi) 'the baby was fed lilis (by Lüsi)'

There are two cases where a passive verb is required-- (1) in certain questions and (2) in certain relative clauses. These will be discussed later §...

3.2. Multiple subjects/objects/verbs. Specifically, (1) VERB + multiple subjects or objects-- these are actually a reduction of conjoined sentences, e.g. "John ran, and Mary ran > John and Mary ran" or "John (and Henry) saw Mary, and John (and Henry) saw Lucy > John (and Henry) saw Mary and Lucy" or (2) multiple VERBS + same subject(s), e.g. "John and Mary laughed, and John and Mary ran > John and Mary laughed and ran". In this second case in Prevli, the verbs may only be intransitive; transitive verbs may not be conjoined (*"...saw Henry and beat him", nor may transitive + intransitives be conjoined (*"...beat Henry and ran away" or "he fell and broke his arm")-- in such cases, each verb must have its appropriate pronoun, and these would be in effect two full sentences conjoined.

3.2.l VERB + multiple subjects/objects:

--Intransitive verb: correctly, a subject pronoun should be used:
--lat korvåd zet yän te lüsi 'Yän and Lüsi ran away' (lit. away ran they Y. and L.)
--borkovo kanta min te kan 'you and I ought to go' (ought-go we you and I)

--Transitive verb: as usual, the pronoun must agree in number with the subjects/objects:
--la:ter zente yän te lüsi (~a garo) 'Yän and Lüsi saw him (~ saw Garo)'
--la:ter zehenta yän a lüsi te garo 'Yän saw Lüsi and Garo'
--läter kamelta min te garo 'I saw you and Garo' (lit., saw I-you.pl garo and you)

When multiple Direct Objects are passivized, a pronoun must still be used:

--(Single object) l,äter garo ni yän (te) ni lüsi 'Garo was seen by Yän and Lüsi', but contrast--
--(Mult. object) l,äter zet lüsi te garo ni yän 'Lüsi and Garo were seen by Yän'

When there are multiple indirect objects, the preposition ki ~k- must appear with each one (conjunction te is optional): donor hen toe kagná (te) kiblí (te) ki garo 'he gave money to me, to you, and (to) Garo'. Similarly, when IO's are passivized, a resumptive pronoun is used: kan (te) min (te) garo d,önorzi kanta toe (neśké) 'you and I and Garo were given money (by him)'.

With objects of prepositions (as with the dative forms just given) it is preferable to repeat the preposition, especially when pronominal forms are involved; with nouns, the rule may be ignored, but then te is required:

--kovolka dio enze baliblí (te) baleśké 'I'll go to the city with you and him'
--kovolka baliblí te (bal) garo (~...bal yän te garo) 'I'll go with you and Garo (~with Yän and Garo)
--tetkäz nifar peak te teta (~...nifar peak nifar teta) 'he carves with a knife and chisel'

3.2.2. MULTIPLE VERBS. As mentioned above, only intransitive verbs can be conjoined and the actions must be viewed as more or less simultaneous, not in sequnce (conjunction te is optional); there may be multiple subjects as well; a full pronoun form (not a suffix) must be used--

--saner (te) krandabra zek 'he sang and danced'
--mak:er (te) ŋalvar zet yän te garo i köin 'Yän and Garo laughed and played in the yard'
Since these actions are not taking place in sequence, the verbs could be in any order.

Another, rather formal/literary way involves the use of the unmetathesized gerund form, without /te/ (NB used only if the verbs are simple (non-derived) and in realis mode)--

--sane krandabra zet, approx., 'singing, they danced' or krandam, saner zet'dancing, they sang'
--ŋalav mak:er ot lakta ~ makeg ŋalvar ot lakta 'playing, the children laughed' ~ laughing, the children played'

More than two verbs may be so chained, as long as all are intransitive and same-subject: ŋalvar mak:er kokorvåd zet i köin 'they played and laughed and ran around in the yard', OR: ŋalav makek kokorvåd zet...

(On other uses of the unmet. base, see §.....)

Sequential events require the conjunction yate 'and then...' and require two (or more) full clauses (thus, true conjoined sentences)--

--imnagradza yate lat korvådza 'he got angry and ran away'
--idlugvuza yate doru-dormuz 'he gets drunk and goes to sleep'

These sentences, of course, have same-subject in each clause; if different subjects were involved, one would have to clarify with demonstratives, or use names: imnagradz ik (~yän) yate lat korvådz ok (~garo) 'this one (~Yän) got drunk and (then) that one (~Garo) ran away'.

The same stricture applies to verbs of different types, whether same-subject or not:

--(intrans.+trans.) dio enze zizerna i purán te marned eśkänto lagu kaśék 'in the city, we ate at a restaurant and saw (watched) two movies' (non-sequential events)
--(trans.+intrans, sequential) mitsir hen mahovnu arka yate nopsanerza 'he drank too much arka and (then) began to sing'

Sentences may also be joined with conjunctions of time (when, before, after etc.), purpose, or cause; generally the two (or more) clauses may be in any order:

--uśkam tu nomer o halde itus dirok kovorna 'the festival had not yet begun when we went there' ~itus dirok kovorn, uśkam tu nomer o halde 'when we went there, the festival had not yet begun'
--uśtuvŋalga garo pfuru saptaz 'Garo can't play because he's sick' ~ pfuru saptaz, uśtuvŋalga garo 'because he's sick, Garo can't play
--kovork di o enze untu nilmä ko näki 'I went to the city to buy a car' ~ untu nilmä ko näki, kovork di o enze 'In order to buy a car, I went to the city'
(Note: in formal usage, when the modifying clause is preposed--as in the second versions above--speakers/writers sometimes use the unmetathesized gerund form: e.g. itus dirok kovo,...; pfuru sapat...; untu niläm näki)

3.3. COMPARATIVES AND SUPERLATIVES. Comparatives are formed with prefixes tab- 'more' and das- 'less' (< /tabu/ and /dasi/ resp.); only /tabuk/ tavok tafko 'better' is irregular. Superlatives are formed with prefixes nus- 'most' and zil- 'least' (< /nusu/ and /zili/ resp.); only numuk 'best' and nuzala 'worst' are irregular.

3.3.1. In comparative sentences, the item being compared is called the "target of comparison" (and will be the subject of the sentence), while the item it is being compared to is called the "standard of comparison". Thus in taftate lüsi man eŋka 'Lüsi is more beautiful than Enka', Lüsi is the target, Enka the standard; linking the two is the particle man, translated as 'than' (< /manad/ 'exceed, surpass'). When the target/subject is a pronoun, the full form must be used: tavlopku kan man garo 'I am older than Garo'; when the standard is a pronoun, the oblique form must be used: tavlopku garo man agná 'Garo is older than me'. If the standard is already present in context, it may be omitted: taftate lüsi 'Lüsi is prettier', tavlopku kan 'I am older'. Similarly, if the target is already known from context, it may be omitted-- onda tavu man nôm 'there were probably more than ten [whatever]', or elda tavu manok/manik 'there were more than that/this' or even elda tavu 'there were more'.

If the standard is a clause or sentence, then manu (man+u 'subordinate clause intro.part.') is used:

--tavlopku garo manu semiz 'Garo is older than he says (he is)' (following older usage, the verb in the standard may be in the passive tavlopku garo manu śe:mi neśké)
--el tavnu śuśtut manu kiridak 'there are more problems than I thought' (~ ...manu k,irid nagná)
--tusi el daśnu yana manu elda dolo mitsä 'now it is less sunny than it was this morning'

In a few cases, a passive verb is required, or at least preferable:
--tavneab kan diek peag manu ok nakp,ärne niblí 'my knife is sharper than the one you are using' (passive is required here, lit. ...manu that(one) PROG-PAss-use by-you)
--donor hen tavu (~tavno) kagnát manu l,e:rod (nagnát) 'he gave (us) more (~more of it) than was needed (by us)' ( ~'...more than we needed')

Included here are those cases where the sentence in the standard can be omitted because it is understood from context, but manu is still required:

--meŋgö o tafko manu dolo 'it looks better than before (~than it did)'
--zize zek tav manu kan 'he eats more than me (~than I do)'

Particle man(u) is also used with the verb /suyan/ sün,å 'different, other'-- sün,å zek diez ŋagat manu zek diez śemit 'his actions are different than (~differ from) his words', or, rather formal, sün,å ka n,äga neśké manu ka śe:mi neśké 'what he does differs from what he says' (the verbs are passive because ka 'what, that which...' is the (underlying) DO-- but in colloquial speech, that rule is ignored, and ka can be followed by an active verb: ...ka nagaz....ka semiz)

3.3.2. Superlatives. Superlative expressions may stand alone-- nuśtate lüsi 'Lüsi is the most beautiful' (~colloq. el lüsi o nuśtate 'Lüsi is the most beautiful one'); they may include a standard, of sorts, such as "Lüsi is the most beautiful of all" nuśtate lüsi man olta (lit., ...than all-pl), or el zek nuśloro gile manu t,örmo nagná 'he is the stupidest person I know' (passive required here, '...manu is-known by-me'). Or it may be a prepositional phrase, like "...in the world" etc.-- el lüsi nuśtate sipe ivo piglä 'Lüsi is the most beautiful woman in the village', ...o nuzala gile ivo pinen 'the worst person in the (our) world' (an older usage has bak 'from' instead of ive).

3.3.3. Colloquially, any comparison (comparative or superlative) may be phrased with introductory /el, on/-- el taftate lüsi (man eŋka) 'Lüsi is prettier (than Enka)'; on tavlopku kan (man garo) 'I am probably older (than Garo)'; el nuśloro zek 'he is stupidest', etc. Any comparison may also be nominalized with the definite article (and /el,on/): el ok o tafko 'that's the better one', el lüsi o taftate/~o nuśtate 'Lüsi is the prettier/the prettiest one', on kan o nuślopku 'I'm (probably) the oldest'. (When irrealis mode is used in a comparative sentence, /on/ must be used, since prefixed comparative adjective forms cannot use the irrealis stem, i.e olpúk is the irrealis of /lopku/ 'old', but *tavolpúk '(probably) older' is not a possible form.)

The various bases used in comparisons (tavu, dasi, nusu, zili) can also function as adjectives, usually along with partitive -nu (or contracted -no 'of it, of the...')--

--lero eśkänto tavnu diŋ 'we need more meat'
--lero eśkänto tavno 'we need more (of it)'(of something already mentioned)
--zizer hen nuśno lilsi 'he ate most of the lilis'
--zizer hen nuśno 'he ate most of it (already mentioned)'

They may also be nominalized, with the def.article o ~ot (nusu/zili more often than tavu/dasi):
--zizer hen o nuśno lilsi 'he ate the most lilis' (contrast zizer hen nuśno lilsi above) --zizer hen o nuśno/o zilno 'he ate the most/the least'
--nerki zetent yagnat o nusno loptaśte 'the Yagans (clan) own the most lopas'

When tavu/dasi (or their derived comparatives) are used in generic sense (i.e. with unknown or indefinite referent) they may function as adverbials:

--tavu pator ko manu min 'I liked it more than you (did)'
--taftaśpo korvåd zek manu tuvnür kan 'he ran faster than I could (run)'
--dasi zize kan manu dolo 'I eat less than I used to (eat)' (contrast zize ko daśnu diŋ manu dolo 'I eat less meat than I used to')

Finally, there are a few idiomatic uses and related forms:

--dasi-tavu 'more or less...'
--tafte-tavu 'more and more'; daśte-dasi 'less and less'
--tavu/dasi....tavu/dasi.... 'the more/the less....the more/the less...'
--/zizil/ zizli 'at least...'
--/nunus/ nutsu 'mostly, by and large, most of all' etc.

3.4. "SURPRISE" FORMS: Many verbs/adjectives as well as some nouns and adverbs can be used in these expressions, which involve (1) full reduplication, (2) final stress (no vowel harmony) and (3) addition of the stressed particle /há/. The implication is that the situation is not as expected (sometimes ironically so, and sometimes unpleasant)--

--nunutám há! 'it's really dark!'
--aharút há! '(it's/he's) really dead!'; of a person rescued from drowning one might exclaim lalakát ha 'he's alive!!'
--mamaŋaí há 'way too long (a time)'

Pronouns are seldom used, though a noun subject may be: teo nunutám há 'the day is really dark!'. Whether a "surprise" statement is intended seriously/admiringly, or ironically/sarcastically depends on tone of voice and facial expression: sasandó há '(he/she's) a real shaman! ~a really skillful shaman' or perhaps with a sneer,'some shaman, i.e. .a really inept/incompetent/phony shaman'.

These may take negative /usu/ (not prefixed)-- us aharút há '...not really dead !' They are assumed to be present tense, realis; to express irrealis, and/or past/future tense, one must use the appropriate form of /el ~on/ (NB with "surprise" statements only, they have irregular forms: of /el/, er past, la future; of /on/ da past, lo future)--

--on lalakát há 'he's/it's probably alive!'
--er us aharút há 'he/it wasn't really dead!'
--lo tatatí há 'it'll (probably) be really beautiful' ~'I'll bet she's really beautiful'

More rarely, these forms may be used (usually without /ha/) in ordinary sentences:

--user aharút ok talne, te papker zehen a garo 'that animal (surprisingly) wasn't really dead, and it bit Garo'
--uśtuvla:ter ko duglá pfuru er nunutåm (há) 'I couldn't see a thing because it was really dark.'

3.5. RELATIVE CLAUSES are introduced by /odi/ ore ~or, and directly follow the antecedent noun. However, if the NOUN + REL.CL. is the subject of the sentence, the entire phrase is usually fronted and the verb will have a resumptive pronoun, as Prevli does not allow so-called "heavy noun-phrases" in subject position--

--o lopo or tarted, uśtuft,üpirza 'the lopa that got loose could not be caught' (contrast the active version of this sentence: uśtuftupir zente o lopo or tarted 'they could not catch the lopa that got loose'
--o taro or bilir ki biZgoran, harkodza bak sün,å pilgä 'the man who spoke to our Council came from a different village'
--o gile or papkar kaz a tämni, lat korvådza 'the person who beat Tämni ran away'

3.5a. Further, /odi/ must be the SUBJECT of the relative clause. Thus, transitive verbs must be in passive voice; this does not affect intransitive verbs, of course--

--uśtormö kaz a taro or bilir 'I don't know the man who spoke'
--...o sipe or nirtü '...the woman who died'
--...o sipe or naktupkå 'the woman who [is grinding (grain)]' (i.e. using a tupa$#x014b;)
--nerki hen näki or galga 'he owns a car that is red'
(Note that no pronoun suffix is needed when the verb of the RC is intransitive.)

And when /odi/ is the agent/subject of a transitive verb in the RC, active voice is used (compound pronoun required):

--...o taro or nerki hen o galga näki '...the man who owns the red car'
--...o sipe or semir zehen u lat kovoz '...the woman who told him/her to go away'
--...o sipe or pamne zehen a garo '...the woman who married Garo'
--...o taro or davlar sehak dio enze '...the man who accompanied me to the city'
--dono hen toe kigidlá or lero hen 'he gives money to anyone who needs it'

But /odi/ must be followed by a passive verb in the following cases:
(a) as object of RC (use of the subject suffix is preferred here (though not required) if the subject is animate, but is usually omitted for inaminate subjects)--

--...o sipe or p,ämne(z) ni garo 'the woman who(m) Garo married' (...who was married by Garo)
--...o näki or n,e:rki nigná / ...ni ok taro '...the car that is owned by-me / ...by that man' = the car (that) I own / that man owns
--...o tupkå or p,ärned ni o sipe 'the tupaŋ that was used by the woman = ...that the woman used'
--...o taro or d,ävlar(za) nigná dio enze 'the man who was-accompanied-he by-me to the city.= the man whom I accompanied to the city'

(b) as IO of RC (requires the passive-dative suffix -zi)--

--...o taro or yäkporzi o näki nigná '...the man who was-sold(-to) the car by me' = the man to whom I sold the car
--...tämni or d,önorzi o toe (ni...) '...Tämni who was given the money (by...)' = ...to whom the money was given

(c) in a possessive construction--

--(alienable)...o taro or diez näki n,i:lmär nigná 'the man whose car was-bought by-me' = ...whose car I bought
--(inalienable)...gile or naroz tu d,üpku nigná '...a person whose name I've forgotten'
--o laŋ or erniz imnaŋkor 'the child whose arm got broken'

3.5b. Other relativizations on prepositional phrases require the prep. + resumptive pronoun following the relative or, and the verb of the rel.clause is preferably active, but in literary/formal usage may be passive; it is often possible to paraphrase such sentences to avoid the or + prep.phrase, which is considered awkward by many--

--"from" ...o taro or bakeśké nilmär ko o näki 'the man REL from-him bought I-it the car = the man from whom I bought the car' (OR passive ...or bakeśké n,i:lmär o näki nigná '...from whom was-bought the car by-me'

--"with" ...o taro or baleské kovorka dio enze kaftu lit. the man REL with-him went-I to city Kavatu = 'the man with whom I went to Kavatu-city' -- an equally good paraphrase, with /daval/ 'to accompany': o taro or davlar sehak dio enze 'the man who accompanied me to the city'.

Relativization on an animate agent, or (inanimate) instrument is considered very awkward and is generally avoided nowadays, and an active voice paraphrase is much preferred; but passives can occur in very formal or archaizing language. Consider the following:

--"by (agent)" ...ot tarot or niśtá n,i:gnir o hupon lit. the men REL by-them pass-built longhouse-our = 'the men by whom our longhouse was built'-- active would be much preferred: ...ot tarot or nignir zento o hupon 'the men who built our longhouse'
...o taro or niśké d,ävlarka di o enze the man REL by-him pass-accompanied-I to the city' = 'the man by whom I was accompanied...'. (Active: ...o taro or davlar sehak di o enze)

--"with/by (instrument)" passive, kernäd o peag or nifró t,e:pkir o diŋ (neśké) 'the knife with which the meat was cut (by him)--note: active ...o peag or tepki ozo o diŋ 'the knife that cut the meat' is disfavored (but not impossible)--because inanimates are not usually the subject of trans.vbs. The following, with nifro but an active verb, is simply wrong, a case of tangled syntax: kernäd o peag or nifró tepkir hen o diŋ, lit. was-dull the knife REL with-it cut(active!) he-it the meat = the knife with which he cut the meat was dull ~the knife he cut the meat with....' A possible paraphrase: kernäd o peag or p,ärned neśké untu tepig o diŋ 'the knife that was-used by him for cutting [gerund] the meat was dull'

3.6. QUESTIONS are of two main types: yes-no, and informational. (There are also yes-no tag questions-- "you're going, aren't you? are you going, or not?".) Since all questions, theoretically, are seeking information that one does not yet know, one might assume that they would be phrased with an irrealis verb. That is indeed the case with most informational questions, which begin with an interrogative pronoun or adverbial, as will be shown.

3.6.1. Yes-no questions are introduced with particle si, and can often be in realis, since the speaker expects only confirmation or denial--

--si kovom? 'are you going?'; si uśkovom? 'aren't you going?'
--si tu la:te miz a garo? 'have you seen Garo?'
--si hondöd mo ka semirza? 'did you hear what he said?'
--si tubhondö mek? 'can you hear me?' (by sandhi, /-b+h-/ > f:, [tuf:'ondö])

If the question has future tense, an irrealis verb is common-- si okfálam di o halde? 'will you go to the festival?', but any yes-no question, especially negative ones, may be phrased in irrealis to coŋote some doubt, disbelief or perhaps reticence--

--si okfám? 'are you (really) going? ~is it possible you're going? ~you're going despite...?
' --si usokfám? 'aren't you going? (I thought you were)'
--si uso:nórda mo? 'didn't you hear it? (I can't believe you didn't)'.

3.6.2. Tag questions append /mihi/ mi: 'yes', with rising intonation: si kovom, mi:? 'you're going, aren't you?', si uśkovom mi:? 'you aren't going, are you?'; si uśla:te mo, mi:? 'you didn't see it, did you?'. "Or not/or what?" is expressed by appending /ili/ 'or', also with rising intonation: si kovom, ili? 'are you going, or not?'; si el talne, ili? 'is it an animal, or what?' (These could also use irrealis mode.)

3.6.3. Information questions always begin with an interrogative pronoun/adjective (and associated nouns if any) or adverbial--e.g. siŋ:ä/sivi teŋgi 'what/which book...?'; siti nu toe... 'how much money...?'; sitsur kovormi? 'when did you go?' silvá min? 'where are you from?'. The verb should correctly be irrealis, though this is often overlooked. In simple "who/what/where etc. is..." questions, no verb is needed, although, colloquially, /el, on/ is often inserted: sigi (el, on) zek? 'who is he?', silmí (el. on) min diem peag? 'where is your knife?' Note idiomatic sigi naro(+poss)? 'what (lit., who) is (your, his, etc.) name?' Use of irrealis in these cases conveys a sense of doubt: sigi on ok? 'who do you suppose that is?', silmí on o peag? 'where can the knife be (where did I leave it?)'. One can also use sigilá 'who ever...? who on earth...?', siŋalá 'what ever..., what on earth...?', siliblá 'where ever..., where on earth...?' and the other similarly suffixed interrogatives.

In transitive sentences, when an interrog. pronoun or adjective is the direct or indirect object of the verb, or a possessive (e.g. "who(m) did you see? what are you eating? which car will you buy? to whom did you give it? whose knife did you use?" etc.), correct usage requires a passive (irrealis) verb. In colloquial speech, again, the irrealis rule is sometimes overlooked, though it is considered a mistake.

--sigi yälhítra niblí? 'who(m) did you see?'-- colloq. (active verb) sigi lä:ter miz, or (a definite error) sigi lä:termi?
--siŋ:ä (~siŋä) nagyi:zze niśké? 'what is he eating?'-- colloq. siŋä nagzize hen?
--sivi näki yi:dlábla niblí? 'which car will you buy?'-- colloq. sivi näki nilmäl mo?
--siŋ:ä (~siŋä) yödnád niblí keśké? 'what did you give him?' (colloq. ...d,önor...)
--sigi yödnádzi niblí? 'to whom did you give it?' (colloq. ...d,önorzi...)
--sigi yödnádzi o toe? 'who was given the money?'
--sigi diez näki yäpridla nagnát? 'whose car shall we use? (alienable noun)
--sigi lagza tun inaśpát? 'whose child has gotten sick? (inalienable noun, comparable to formal zek titiz 'his/her/the baby')-- some speakers say sigiz lagza... but that is considered incorrect.

If the interrog. pronoun is the object of a preposition, the construction used is similar to that seen in relative clauses, and transitive verbs may be either active or passive--

--sigi baleśké a:mírmi? 'with whom did you travel? (who did you travel with?)'
--sigi suśké ik toe? 'for whom is this money (who is this money intended for?)'-- contrast
--siŋ::ä unto ik toe? '(for) what (purpose) is this money, what is this money (intended) for?'

Interrogative sentences like the following, with instrumental nifro, are considered awkward, and are best rephrased--

--siŋ:ä nifro tuvetpík eśkänto? 'with what can we cut it? (what can we cut it with?)' or
--siŋ:ä nifro tuvye:tpík o (nagnát)? 'with what can it be cut (by us)?'--

And as with sentences involving animate agents, it is not possible to use sigi ni+pron. as animate agent of a passive verb in a question-- thus, e.g. "by whom was this food cooked?" must be phrased as an active sentence: "who cooked this food?"

If something has been mentioned in the discourse, an information-question about it may be in realis, e.g. "--When I visited Holunda-city..." --sitsur kovormi dirók? 'when did you go there?' or (verbless) sitsur dirók min? 'when were you there?'

3.7 Aspectual forms and sentential/clausal complements.

3.7.1. (Same subject) Many clausal complements are handled with the various aspectual forms-- when the subject of the clause is the same as the subject of the main verb, for example--

--desiderative: mekovok 'I want to go', irreal. meokfák 'I may want to go'
--inceptive: tu nobzizeza 'he has beguan to eat'; nobzizer hen 'he began to eat it'
--intentive: diśnuzo ko o kurkän 'I'm going to fill the water-jar' (irr. disunzá...)
--obligative: tetseminta (tembilinta) kitämni (ut...) 'we have to/must talk (speak) to Tämni (about...)'
(The full list of aspectual derivations and their prefixes is given in the Morphology §2.8.6.

Other verbs that can take same-subject complements require a separate clause or sentence, usually introduced with the particle u, equivalent to Engl. 'that', though the particle is often omitted in colloquial speech. This construction is the same as that required for full sentential complements (see next, §3.7.2). The complement verb, if intrasitive, does not have a subject suffix--

--patoz u bovon or colloq. patoz bovon 'he likes to do the bon dance'
--kutnårk (u) kovo 'I asked(requested) to go'
--kör,ådza (u) korvå 'he refused/was unwilling to run'
--kör,åz (u) zize 'he refuses to eat'

But if the complement verb is transitive, with a direct object, then u and a subj+obj pronoun are required:

--kör,åz u zize hen (~yotsu) 'he refuses to eat it (~yotus)'
--ŋarodk u utpí ko o popo, ŋok uśtuvnür ko 'I tried to catch the popo, but (I) couldn't [do it]' (ŋaro 't o try' usually takes its complement in the irrealis, especially in present/future tense.)
--kör,ådz u zize hen o diŋ or tu k,äśred niśtá 'he refused to eat the meat (that) had been cooked by them (that they had cooked)'

The same u construction is used in the case of an auxiliary modal verb + a derived aspectual form, e.g.--

--tean 'have to/must' + causative apkovo 'to send s.t.' (cause s.t to go): teamni u apkovo mo toe keśké 'you have to/must send money to him'-- realis, because the speaker expects the action to occur. If the mood is irrealis, both verbs must be in that mood (/tian/ tean > irreal. /it:án/ itsán): itsámni u apokfá mo toe keśké 'you may have to (probably) send money to him'-- speaker is uncertain whether the action is truly required. Especially so in the case of an aspect, like debitive bor- that usually takes the irrealis:
--borok,ärm u aŋgá mo 'you (probably) ought to (irreal.) refuse to do (irr.) it' (debitive bor+ok,är < /okyár/ irr. of /koyar/ 'refuse').

A clause may also be introduced by untu 'in order to ~ so that'-- dirók kovont untu kapka baleśté 'we went there (in order) to fight with them'

Some same-subject complements can begin with the particle ka 'what (i.e. that which...); when it is the object of the complement's verb, the verb must be passive, though as mentioned above, that rule is often ignored:

--uśakmek ka inehlir 'I don't know what happened' (intheli is intrans.irreal., therefore ka is subj.)
--yakmez ka n,äga 'he knows what to do' (lit. what is done)-- colloq. and very common with active voice, yakmez ka naga
--yakmez ka boryägná 'he knows what ought to be done' This could also be translated as 'he knows what he (same subject) ought to do' (prefix bor- usually requires an irrealis verb)-- see further &#l67;8c below regarding sentences with multiple 3rd person pronouns and avoidance of ambiguity.

3.7.2. (Different subject) Sentential complements. These include both clausal complements with different subjects ("I want you to go") as well as full sentences ("I think [that] you should go"). Many of these latter may also be categorized as "indirect speech" (see §3.8), since the main verb will often be one of "telling, ordering, asking, thinking, knowing, hoping" etc. Aside from u, other particles that can begin the complement sentence are pu 'if/whether', ka 'what, that which...', untu 'in order to, so that...', plus a few others, and the various interrogatives.

3.7.2a. Clausal. The particle u is required--

--meent u kovom 'we want you to go' (emphatic: meent u kovo min 'we want you to go')
--meek u zize imlitso ok ôl yadze 'I want you(pl) to eat all that food'
--kutnårk u apkovo zento 'I asked (requested) them to send it'
--kör,åk u ŋalvam i ratu 'I forbid you to play in the road'

3.7.2b. Sentential. Many verbs-- kiri 'to think' and uśakme < /us+yakme/ 'not know', in particular-- require an irrealis verb in their complement:

--kirik u aśpátsa 'I think (that) he's sick' ~...u aśpatsa tämni '...that Tämni is sick'
--yakmek u patoz (u) bovon 'I know (that) he likes to do the bon-dance'
--uśakmek pu aptúz (u) bovon 'I don't know if/whether he likes to do the bon-dance'
--hondödka ka śe:mir niblí ut garo 'I heard what you said about Garo' (...what was said by you).
--meyakmek ka śe:mir neśké 'I want to know what he said'
--nagar hen untu uśteandam 'he did it so that you didn't have to' (~...uśteanda mo)
--mono kaŋka sitsul indútsa 'I wonder when he will die' (lit., I ask myself...)
--taŋok u uśpítmi 'I'm afraid/I fear that you'll fall (/taŋu/ 'afraid' usu. takes irrealis)

3.8. Indirect speech/orders/questions. Indirect speech will always involve the verb semi 'to say, to tell' or a synonym; indirect orders likewise may use semi or /kutan/ 'request', or more specific verbs like 'to order, demand' etc.; indirect questions will usually involve /mona/ mono 'to ask, inquire'. These verbs may also take an indirect object (the person spoken to, ordered, asked), and in colloquial speech may sometimes use a compound pronoun to express a pronominal IO.

3.8a. In indirect speech and questions, the complement may be in any tense, realis or irrealis:

--semirk u nayá kovolmi 'I said that you will(would) go tomorrow' (realis)
--semirk u nayá okfálmi 'I said that you will(would) (probably) go tomorrow' (irrealis)
--semirza kagnát u nayá okfálmi 'he said to us (told us) that you will (probably) go tomorrow'
--omnálmi (kagná) siprü aŋar ko 'you will (probably) ask (me) why I did it'(irrealis)
--semilka keśtá u tun itsaptam 'I'll tell them (that) you've gotten sick'
--monorka (keśké) siŋä nagz,izer niśké 'I asked (him) what he was eating' (more correct if irrealis: ...siŋ:ä nagyizzer [na'džidzer] niśké)

The last four examples are cases where the compound pronoun could be used: semir zehänt u..., omnál mek siprü..., semil kazent u...; monor kaz siŋä.... Correct and/or formal speech, however, avoids this usage, reserving the compound pronouns only for subject+direct object.

3.8b. There are two ways of expressing indirect orders: one--more often used colloquially, and where the context is clear--is the same as any indirect speech after semi, kutnå etc. + u except that the verb of the complement (the order) will be in present tense, realis; further, it can be used only for positive, not negative, orders--

--semirk u kovom 'I told you to go' (or semirka kiblí u... or semir kame u...
--semirz u delmök u zize ko diŋ 'he told me to stop eating meat' (semirz kagná u... or semir sehak u...)
--semirz u korvåz 'he(A) told him(B) to run'

The problem with this usage is that, without context, many such sentences could be interpreted as meaning simply "X said that....". Using /tenad/ 'to order' would avoid that ambiguity:
--tendädza u kovom 'he ordered you to go' (or tendädza kiblí u... ~ tendäd zehim u...)
--tu tendä o karmo u delmö ko (u) zize diŋ 'the doctor has ordered me to stop eating meat'

The second way, somewhat more formal but unambiguously an order, is to put the complement verb in the imperative mode:

--semirka keśtá u lat kəvá (more formal kəvaʔá) 'I told him to go away' (colloq. semir kaz...)
--semilka keśtá u zəzá (~zřzaʔá) hen 'I'll tell him to eat it'
--tendälzet kiblí u uZvilá 'they will order you to be quiet/keep silent' (~tendäl zetla...)

Negative orders/prohibitions must use this construction:
--semirka keśtá u daśkəvá 'I told him not to go' (< /daza kəva-ʔá/)
--tendädzet kagná u dazvəliʔá 'they ordered me not to speak' (/daza bəli-ʔá/, colloq. dazavliʔá)

Direct/quoted orders differ from indirect orders in that (1) particle u is absent, and (2) there is a break in the intonation preceding the quoted material-- semirka keśtá (~semir kaz,...), lat kəvá 'I told him (said to him), go away; tendädzet kiblí, uZvilá 'they ordered you, be quiet' (tendäd zetla,..)

The compound pronouns are often used in direct quotes:

--semir kaz, siŋä n,aga niblí? 'I said to him, what are you doing?
--semir kame, daz nəgá mo ok 'I told you, don't do that'
--garmed kaz ki höiz, uśnabrad kaz a lagza 'I replied to the judge, I did not harm the child'
--monor kame, si tun zizem? 'I asked you, have you eaten?'

Note that in these case, the compound pronoun references subject+indirect object--it is colloquial usage, and technically incorrect but widespread. In very correct speech, the compound pronoun is correctly used only for S+DO, and in the sentences above there would have to be a subject suffix, and the person addressed would be in the dative: semirka kestá,... / semirka kiblí,... / garmedka ki höiz,... / monorka kiblí,... etc.

In a few cases the reflexive pronoun is used:

--semir (~kirid) kaŋka, uśmaśte ok 'I said (~thought) to myself, that is not correct/proper'
--monor kaŋka, siprü dirók kovormi? 'I wondered, why did you go there?' (Note that without the break in intonation--indicated by the comma--this sentence would simply be the indirect question, 'I wondered why you went there'.)

3.8c. Avoidance of ambiguity (same/different 3rd pers. subjects). We have already established, in §3.7.1,2 above, that in cases of Main Verb + u Complement, when both the main verb and the complement verb (intransitive) have same-subject pronouns, no subject suffix is used on the complement verb; but when there are different subjects, then the suffix is required. Thus consider the following, with same subject--

--kirik u inapsát 'I think I'm getting sick' (not *inapsátka)
--si kirim u okfál? 'do you think you'll go? (not *okfálmi)
--semir yän u tu zizer 'Yän said that he (i.e. Yän) had eaten' (not *zizerza)

Or with different subject:

--kirik u inapsátmi 'I think you're getting sick'
--si kirim u okfálza 'do you think he'll go?'
--semir yän u tu zizerza 'Yän said that he (other, not Yän) had eaten'

This is also the case if (colloquially) the main verb uses a compound pronoun to express an indirect object:

--semir kame u itsapatra 'I told you I was getting sick' (not *sapatrak)
--semir zehim yän u tu zizer 'Yän told you that he (Yän) had eaten (not *zizerza)
--semir zehim (yän) u tu zizerza 'he (~Yän) told you that he (other, not Yän) had eaten'

If the main verb has a noun subject, the same interpretations apply:

--a) semi tämni u kovol_ 'Tämni says he (Tämni) will go'
--a') semi tämni u kovolza 'Tämni says he (other) will go'-- and of course change of subject is clear if the complement verb has a noun subject--
--a'') semiz u kovol garo 'he says Garo will go'

Clearly, when the complement verb is intransitive, the absence or presence of a pronoun suffix indicates, resp., whether same or different subject in intended. Of course, if one starts to pile up complements, ambiguity may arise:

--a) semiz u kiri_ u aśpát_ 'he(A) says that he(A) thinks he(A) is sick'--no subject markers, same subject in all three clauses, no ambiguity
--b) semiz u kiri_ u aśpátsa 'he(A) says that he(A) thinks he(B) is sick', no ambiguity as to who is sick
--c) semiz u kiriz u aśpát_ 'he(A) says that he(B) thinks he(B) is sick'-- again, no ambiguity; but--
--d) semiz u kiriz u aśpátsa 'he(A) says that he(B) thinks he(A? C?) is sick'-- now there is ambiguity: aśpátsa might refer either to subject A or someone else (C) previously mentioned in context (and this could even be a "mistake" for sentence (b) above). The most likely interpretation is that aśpatsa refers to A, and to clarify that C is meant, one could say ...aśpát zek ~ok, or insert a noun subject-- then the full-form pronoun, demonstrative or noun is taken to mean change of subject. Probably, if kiri has a noun subject, the interpretation would then be that aśpátsa most likely refers to A, as in-- semiz(A) u kiri (garo, o karmo) u aśpátsa 'he(A) says that (Garo, the doctor) thinks he (A) is sick', but even this remains potentially ambiguous and could refer to some other sick person (C).

Other problems arise if the complement verb is transitive, since it will require a compound pronoun that cannot be omitted. There is no chance of ambiguity for most subjects+object forms when subject and object are different persons (and a complement verb's noun subject clearly indicates different-subject, e.g. semirz (semi zehim) u tu zizer hento garo ot aplet 'He said (he told you) that Garo had eaten the apels'). But when both main verb and complement have 3d pers. pronoun subjects there can be a problem:

--a) semirz (semi zehim) u tu zizer hento ot aplet 'he said (~told you) that he had eaten the apels', or semi garo (semi zehim garo) u tu zizer hento ot aplet 'Garo said (~told you) that he had eaten the apels'-- who ate the apels? The subject of semi (he, Garo) or someone else (the subject of zizer)?
-b) semir zet u kapkal eśtím 'they said they will (=would) fight you (pl.)'-- who will fight you? whichever group zet represents, or some different group represented in eśtím?
--c) kiriz (~kiri yän) u izzer hen derenu matla diŋ 'he (~Yän) thinks he ate a little raw meat.' Who ate the raw meat-- Yän or someone else?
--d) kiriz u alhítra zehen 'he thinks he saw him'--this has several possible readings: A thinks A saw B, A thinks B saw A, A thinks B saw C. In context, such a sentence could stand alone, with any of those meanings. One could, if necessary, insert a demonstrative or names/nouns in the subject or object slots.

Since the Lañ-Lañ do not ordinarily write their language, it follows that ambiguities like these will occur mainly in conversation, where it will usually be clear from context who the referents are. If there should be a misunderstanding, the hearer might ask sigi zizer hen? 'who ate it?', and the speaker would then clarify with o kan (zek) ~el kan (zek) 'me (him) ~it was me (him)', or o garo ~el garo 'Garo ~it was Garo', or even ya, lepätka, o tämni '(my intention/meaning=) oh, I meant Tämni'.

More formally, and more precisely, it is possible to disambiguate the pronoun of a transitive verb with the use of the particle hô ~å to indicate "same subject/object" (see the discussion of hô ~å in the Morphology, Particles §7.3). Thus one could clarify the sentences above as follows:

--a) semirz (semi zehim) u tu zizer å hento ot aplet 'he(A) said (~told you) that [same] he(A) had eaten the apels'. Without å the reading is that "A said that B had eaten..."
--b) semir zet u kapkal å eśtím 'they(A) said [same] they(A) will (=would) fight you (pl.)'-- now it is clear that the group referenced by zet will fight you; without å, a different group.
--c) kiriz (~kiri yän) u izzer å hen derenu matla diŋ 'he (~Yän)(A) thinks [same] he(A) ate a little raw meat'; without å, someone else ate it.
--d) kiriz u alhítra å zehen 'he(A) thinks [same] he(A) saw him(B)' versus kiriz u alhítra zehen an å zek 'he(A) thinks he(B) saw [same] him(A)'; and without å at all, (sentence d above) it will mean 'A thinks B saw him(C)' which could be further clarified as kiriz u alhítra zehen an zek (~an ok, ~a garo) 'A thinks B saw him(C) (~saw that one, ~saw Garo)'.

Thus, generally in such cases, å refers back to the primary (or "speaker") pronominal referent. A further example:

--a) kiriz u abdád zehen 'he(A) thinks he(A? B?) injured him(A? C?)' The default readings are A--(B or A)--C (note "that B injured B(himself)" is not a possible reading since that would require the reflexive, ...abdád ośko); but it can be clarified further:
--b) kiriz u abdád å zehen 'he(A) thinks he(A) wounded him (B)'
--c) kiriz u abdád zehen an å zek 'he(A) thinks he(B) wounded him(A)'
--d) kiriz u abdád zehen a zek (~an ok) 'he(A) thinks(B) wounded him(C)'

What if sentences like those above are themselves the complements of a verb like semi 'say, tell'? Intransitive verbs can indicate same/different subject by absence/presence of subject markers, and there in no ambiguity in the following sentences:

--semiz u kiri_ u okfál_ 'he(A) says he(A) thinks he'll(A) go'
--semiz u kiri_ u okfálza (~okfál zek) 'he(A) says he(A) thinks he'll(B) go'
--semiz u kiriz u okfál_ 'he(A) says he(B) thinks he'll(B) go'
--semiz u kiriz u okfálza (~okfál zek) 'he(A) says he(B) thinks he'll(C) go'
--monorza pu ikrídz u borokfá_ 'he(A) asked if he(B) thought he(B) ought to go'

But the following has potential ambiguity: monorza pu ikrídz u borokfáz 'he(A) asked if he(B) thought he(A? C?) ought to go'-- default is that '...he(C) ought to go', but for more clarity one could say ...borokfá å zek '...[same] he(A) ought to go' or ...borokfá zek (~ok) '...he(C) (~that one) ought to go'

An additional problem arises if an inalienable noun is involved.

--a) mepadze zehen a titiz 'she(A) wants to feed the(possessed) baby (A's? B's)'. The default reading is that A's own baby is meant; use of å-- ...an å titiz-- would (redundantly) make that clear, while mepadze zehen an ok/ik (~a zek) titiz ('that/this ~her(B) baby') would indicate that someone else's baby was to be fed. (Or, one could add a name--...a titiz eŋka '(feed)...Enka's baby'.)
--b) meez u padze zehen a titiz 'she(A) wants her(B) to feed the/her baby (A's, B's, C's?)'. In this case, default now is that B's baby is meant, and ...an å titiz would make that clear; to clarify that A's baby is meant, ...an å zek titiz; to clarify "someone else's baby" one could say ...an ok/ik titiz, or add a name ...a titiz eŋka.

With multiple complements:

--a) semiz u kutnår_ u padze zehen a titiz 'she(A) says that she(A) asked to feed the/her baby'-- by default here, A's baby is meant, and ...an å titiz would (redundantly) state this; ...a zek (~an ok) titiz or ...a titiz eŋka would clarify that someone else's baby was meant.
--b) semiz u kutnår_ keśké u padze zehen a titiz 'she(A) says that she(A) asked her(B) to feed the baby (A's? B's? C's?)-- the baby's ambiguity can be resolved as in the previous example.
--c) semiz u kutnårza keśké u padze zehen a titiz 'she(A) says that she(B) asked her(A? C?) to feed the/her baby'-- here, in addition to the baby problem, it is unclear who is to do the feeding-- as given, the interpretation should be 'A says that B asked C to feed...', but the other reading could be resolved with semiz u kutnårz å keśké u padze zehen a titiz where it is now clear that 'A says that she(B) asked [same] her(A) to feed...' (with "baby" to be resolved as above).

But we should reiterate that the use of å is rather formal, and rarely occurs in everyday conversation, where other strategies exist to disambiguate such sentences.

Plus more yet to be determined.....

Return to main page