Waris Grammar Sketch

Grammar Sketch Final Draft

Bob Brown
June 1990

Final Draft (Corrected from consultant notes)

 Table of Contents
1  Introduction .......................................................................................................... 1 
1.1  The Waris Language ....................................................................................................... 1 
1.1.1  The Classification of Waris. ......................................................................................................... 1 
1.1.2  The name of the Waris language. ................................................................................................. 1 
1.1.3  The Orthography Used in this Paper ............................................................................................ 2 
1.1.4  The Salient Features of Waris Grammar ...................................................................................... 2 
1.1.5  This Grammar Sketch ................................................................................................................... 3 
1.2  Abbreviations .................................................................................................................. 3 

2  Word Classes and Word-level Constructions ...................................................... 4 

2.1  Particles ........................................................................................................................... 5 
2.2  Adverbs ........................................................................................................................... 7 
2.2.1  Time Adverbs ............................................................................................................................... 8 
2.2.1.1  Time Adverbial Phrases .............................................................................................................. 10 
2.2.2  Manner Adverbs ......................................................................................................................... 11 
2.2.3  Locative Adverbs ........................................................................................................................ 12 
2.2.4  Interrogative Adverbs ................................................................................................................. 13 
2.2.5  Derived Adverbs ......................................................................................................................... 14 
2.2.6  Adverbs Forming Derived Stems. .............................................................................................. 15 
2.3  Nominals ....................................................................................................................... 16 
2.3.1  Kin Terms ................................................................................................................................... 16 
2.3.2  Human Nouns ............................................................................................................................. 17 
2.3.3  Proper Nouns .............................................................................................................................. 17 
2.3.4  Personal Nouns ........................................................................................................................... 18 
2.3.5  Common Nouns .......................................................................................................................... 18 
2.3.6  Personal Pronouns ...................................................................................................................... 18 
2.3.7  Demonstrative Pronouns ............................................................................................................. 19 
2.3.8  Interrogative Pronouns ............................................................................................................... 20 
2.3.9  Quantifiers .................................................................................................................................. 21 
2.3.10  Adjectives ................................................................................................................................... 21 
2.3.11  Demonstratives ........................................................................................................................... 22 
2.3.12  Chart of Nominal Properties....................................................................................................... 23 
2.3.13  Noun Classification .................................................................................................................... 25 

ii
2.4  Compound and Derived Stems ..................................................................................... 25 
2.4.1  Compounding two noun stems ................................................................................................... 25 
2.4.2  Determiner plus Determined Constructions ............................................................................... 26 
2.4.3  Independent stem plus dependent stem ...................................................................................... 27 
2.4.4  The Function of -l ....................................................................................................................... 31 
2.5  Verbs ............................................................................................................................. 32 
2.5.1  Noun Incorporation in Verbs ...................................................................................................... 32 

3  Noun Phrases...................................................................................................... 33 

3.1  NP Constituents and Order ........................................................................................... 33 
3.1.1  The Minimal NP ......................................................................................................................... 34 
3.1.2  NP with Adjective ...................................................................................................................... 38 
3.2  Coordination of NPs...................................................................................................... 38 
3.2.1  Coordination by the Comitative Marker .................................................................................... 38 
3.2.2  Coordination by the Genitive Case Marker ............................................................................... 39 
3.2.3  Coordination by -e ...................................................................................................................... 39 
3.2.4  Lists with the Verb 'to do' .......................................................................................................... 39 
3.2.5  Coordination of Alternatives ...................................................................................................... 40 
3.2.6  Pronominal Copy ........................................................................................................................ 40 
3.3  Case Marking of NP...................................................................................................... 40 
3.3.1  Details of Case Marking ............................................................................................................. 44 
3.3.2  Wide Distribution of Goal Marking ........................................................................................... 45 
3.4  Topic Marking of NP .................................................................................................... 47 

4  The Verb Phrase................................................................................................. 48 

4.1  Pre-Core Positions ......................................................................................................... 49 
4.1.1  Subject Number .......................................................................................................................... 49 
4.1.2  Number of Things Carried ......................................................................................................... 50 
4.1.3  Number of People Accompanied................................................................................................ 50 
4.1.4  Noun-Classifying Verb Prefixes................................................................................................. 51 
4.1.4.1  Existential Verbs ......................................................................................................................... 52 
4.1.5  Location ...................................................................................................................................... 54 
4.2  Post-Core Positions ....................................................................................................... 55 
4.2.1  Non-singular Direct Object ........................................................................................................ 55 
4.2.2  Benefactive or Recipient Number .............................................................................................. 58 

iii
4.2.2.1  Subject Number Neutralization with the Benefactive.................................................................. 59 
4.2.3  Manner ........................................................................................................................................ 59 
4.2.4  Tense-Mode-Aspect .................................................................................................................... 62 
4.2.5  Irrealis Mood .............................................................................................................................. 64 
4.2.6  Question and Emphasis .............................................................................................................. 65 
4.2.7  Topic ........................................................................................................................................... 65 
4.3  Verb Serialization .......................................................................................................... 65 
4.3.1  Serial Finite Verbs ...................................................................................................................... 66 
4.3.2  Compound Verb Stems............................................................................................................... 66 
4.3.2.1  Compound Verb Stems with Aspectual Focus............................................................................. 67 
4.3.2.2  Compound Verb Stems with Additive Meaning........................................................................... 68 
4.3.2.3  Compound Verb Stems with Idiomatic Meaning......................................................................... 69 
4.3.2.4  Compound Verb Stems with Change of Actor ............................................................................. 69 

5  Clauses ............................................................................................................... 70 

5.1  Clause Classes ............................................................................................................... 70 
5.2  Clause Types with Examples ........................................................................................ 73 
5.2.1  Existential Clause ....................................................................................................................... 74 
5.2.2  Stative-Equative Clause .............................................................................................................. 74 
5.2.3  Motion Clause ............................................................................................................................. 75 
5.2.4  Orientation-Achieved Clause ...................................................................................................... 76 
5.2.5  Force-Motion Clause .................................................................................................................. 76 
5.2.6  Sensory-Inner State Clause ......................................................................................................... 77 
5.2.7  Sensory-Force Clause ................................................................................................................. 77 
5.2.8  Pseudo-Passive Clause................................................................................................................ 78 
5.2.9  Change of State Clause............................................................................................................... 78 
5.2.10  Telic Clause ................................................................................................................................ 79 
5.2.11  Burn-Light Clause ...................................................................................................................... 79 
5.2.12  Holding-Moving Clause ............................................................................................................. 80 
5.2.13  Ditransitive Giving Clause ......................................................................................................... 81 
5.2.14  Accompaniment Clause .............................................................................................................. 81 
5.2.15  Goal Clause ................................................................................................................................. 82 
5.2.16  Patient Clause ............................................................................................................................. 82 
5.2.17  Goal or Patient-Benefactive Clause ............................................................................................ 84 
5.2.18  Sound-Speech Clause ................................................................................................................. 84 

iv
5.2.19  Speech Benefactive Clause ......................................................................................................... 85 
5.3  Distribution of Word Classes in Clauses ...................................................................... 86 
5.4  Grammatical Relations in Waris ................................................................................... 86 

6  Discourse ............................................................................................................ 87 

6.1  Clause to Sentence and Discourse ................................................................................ 87 
6.1.1  Intonation .................................................................................................................................... 87 
6.2  Sentence Constructions ................................................................................................. 87 
6.2.1  Conditional Sentences ................................................................................................................ 88 
6.2.2  Conditionals and Topics ............................................................................................................. 89 
6.2.3  Simultaneous Action Sentence ................................................................................................... 89 
6.2.4  Counter-expectation Sentence .................................................................................................... 89 
6.2.5  Reason or Result Sentence ......................................................................................................... 90 
6.2.6  Alternative Sentence ................................................................................................................... 91 
6.2.7  Purpose Sentence ........................................................................................................................ 92 
6.2.8  Speech Quote Sentence .............................................................................................................. 93 
6.2.9  Comparison Sentence ................................................................................................................. 94 
6.2.10  Intention Sentence ...................................................................................................................... 94 
6.2.11  Syntactic Coordination and Subordination ................................................................................. 95 
6.2.11.1  Subordination.............................................................................................................................. 95 
6.2.11.2  Coordination ............................................................................................................................... 96 

6.3  Semantic Paragraphs ..................................................................................................... 96 

6.3.1  Clause Chaining and Semantic Paragraphs ................................................................................ 98 
6.3.2  Summary of Semantic Paragraph ............................................................................................. 101 
6.4  Topicalization .............................................................................................................. 101 
6.4.1  The Meaning of 'Topic' in Waris.............................................................................................. 101 
6.4.2  Topic Marker Allomorphs and Pronouns ................................................................................. 102 
6.4.3  Survey of Topic Marking ......................................................................................................... 103 
6.5  Topic Continuity ......................................................................................................... 106 
6.5.1  Introduction of Participants ...................................................................................................... 106 
6.5.2  Same Subject and Different Subject Pronouns ........................................................................ 107 
6.5.3  Deixis ........................................................................................................................................ 109 
6.5.4  Topic Continuity ....................................................................................................................... 110 
6.6  Information in Discourse............................................................................................. 111 
6.6.1  Quoted Speech .......................................................................................................................... 111 

v
6.6.1.1  Collateral and Evaluative Information ..................................................................................... 111 
6.6.1.2  Highlighted Elements ................................................................................................................ 112 
6.6.1.3  Non-Speech Act Quotes............................................................................................................. 113 
6.6.1.4  Argumentative Discourse .......................................................................................................... 114 
6.6.2  Negative and Positive Information ........................................................................................... 115 
6.6.3  Backgrounding and Foregrounding of Information ................................................................. 115 
6.6.4  Time Information ...................................................................................................................... 116 
6.7  Relative Clauses .......................................................................................................... 117 
6.7.1  Word Order and Syntactic Marking of Relative Clauses......................................................... 118 
6.7.2  Types of Relative Clause .......................................................................................................... 120 
6.8  Cohesion ...................................................................................................................... 120 
6.8.1  Items Having Diphoric Reference ............................................................................................ 121 
6.8.2  Items Having Mainly Anaphoric Reference ............................................................................. 122 
6.8.3  Items Having Mainly Cataphoric Reference ............................................................................ 124 
6.8.4  Locative Viewpoint .................................................................................................................. 125 
6.9  Types of Discourse ...................................................................................................... 125 
6.9.1  Narrative ................................................................................................................................... 126 
6.9.2  Folktales .................................................................................................................................... 126 
6.9.3  Descriptive ................................................................................................................................ 126 
6.9.4  Hortatory ................................................................................................................................... 127 
6.9.5  Argumentative .......................................................................................................................... 127 
6.9.6  Letters ....................................................................................................................................... 128 
6.9.7  Songs......................................................................................................................................... 128 
6.9.8  Conversation ............................................................................................................................. 129 

7  Bibliography ..................................................................................................... 130 

vi
 1

1 Introduction

This paper is an outline of the grammar of the Waris language intended to summarize all the
features of the language that have been found up to this point. It will serve as a basis for a
complete grammar of the language to be produced in the future. It presents knowledge gained
from 1973 to 1990.

I wish to thank the following people for their help in understanding the Waris language. Bob
Conrad, Cynthia Farr, Bob Litteral, Walter Seiler.

1.1 The Waris Language

1.1.1 The Classification of Waris.

Waris is a Papuan language of the Imonda Sub-district, Amanab District, Sandaun Province.
About 3000 people in about 24 villages speak one of the four major dialects of Waris, and an
additional number, perhaps as many as 1400, speak it in the adjacent Kecamatan Waris of
Indonesian Irian Jaya. Wurm (1982) classifies Waris as a member of the Border Stock, Trans-
New Guinea Phylum (1982, p.192ff). Waris shares the features Wurm posits as characteristic of
Trans-New Guinea Phylum languages except for the following.

1. it has a phonemic contrast between /r/ and /l/.

2. it has two fricatives, /s/ and /x/

3. it has a small class of human nouns, the basic forms of which are plural, and which are
affixed to form the dual and singular.

4. it distinguishes inclusive and exclusive in first person plural pronoun.

5. verbs do not mark person or gender of subject or object.

6. stem vowel raising is a commmon way of marking plural subject number of verbs.

7. there are no sentence-medial verb forms.

1.1.2 The name of the Waris language.

Waris speakers have no name for their language. Waris is the name of the administrative
center established by the Dutch in the early 1950s, and was applied by Loving and Bass (1964)
 2

in an early language survey. However the name Walsa seems to a genuine self-designation of
the people, including all the dialect areas. It seems to refer to them as the ones who successfully
overcame the previous people to live in the area. Vernacular publications now bear this name,
with the word Waris in parentheses; it does not seem worthwhile to try to change the name
Waris to Walsa in the linguistic literature.

1.1.3 The Orthography Used in this Paper

This paper employs the orthography used in vernacular publications, which is phonemic,
except that prenazalization is written. Here is the orthography, with phonemes indicated when
they differ:

b, d, g, p, t, k, v (/-b/), s, h (/x/), m, n, l, r, w, y,

a, e (/ɛ/), ei (/e/), ae (/æ/), é (/schwa/), u, o, ó (/b/).

Morphophonemics are the subject of another paper (in preparation) and are not dealt with in this
paper except in passing.

1.1.4 The Salient Features of Waris Grammar

Here is an overview of the topics which I attempt to describe in this grammar sketch because
they are salient features of Waris.

Although taken individually they are undoubtedly shared by one or other New Guinea
languages, taken in combination they distinguish it from all other languages.

1. three clear grammatical functions, S, O, IO, the number of which is marked on the verb. IO
may be distinguished as recipient or as benefactive. Number accompanied and number carried is
also marked on the verb.

Recipient, benefactor, accompanied, goal, and animate object are marked on the NP with -m,
which also marks absolutive subject (lack of control).

2. person and gender of NP arguments not marked on the verb.

3. basic order of the clause S O IO V.

4. a full set of case suffixes whose meanings are extended to include such things as goal, reason
and result.
 3

5. covert noun classes defined by existential verbs and by noun-class ('shape') markers prefixed
to verbs.

6. pronouns mark only person, and there are two sets, which are used in such a way that third-
person referents may be distinguished as to same-subject and different-subject.

7. a marker used extensively on NP and VP to indicate definiteness and topicality.

8. an extensive system for compounding and deriving word stems

9. extensive use of serial verbs of the type called by James (1982) 'lexicalized'. That is, two or
more verb stems are joined phonologically with only the rightmost bearing suffixation.

The meaning of a serial verb may be the sum of the meanings of the stems, or the meaning may
be idiomatic (not predictable).

10. extensive use of direct speech in narrative and argumentative discourse.

1.1.5 This Grammar Sketch

The thrust of this grammar sketch is semantic, in order to support translation and education
efforts in Waris, rather than to develop support for a particular grammar theory.

Waris verbs do not mark gender, so some of the examples used in this paper glossed as 'he' may
just as well have referred to a female in the context in which they were originally heard. Waris
cultural conventions about the division of labour between the sexes are reflected in some of the
glosses.

1.2 Abbreviations
1st first person IRR irrealis mood

2nd second person LOC locative

3rd third person MAN manner

INC first person plural inclusive NEG negation

ABL ablative case NP noun phrase

ABS absolutive case

 4

A agent NS non-singular

ACC accompanied O object

ALL allative case OPT optative

ALT alternative P patient

BEN benefactive PST past tense

CLAS noun-classifying verb prefix PL plural

COM comitative case PRS present tense

CON continuous verb mode Q question

DER derivation REA reason

DL dual number REC recipient

EMP emphatic RP recent past

EXP experiencer RT resumptive topic

FC force S subject

FUT future verb tense SG singular

GEN genitive TEL telic case

GER gerund TOP topic

GL goal V verb

IMP imperative verb mood

INST instrument

2 Word Classes and Word-level Constructions

Seiler (1984a) gives a detailed account of word classes and compound and derived stems in
Imonda, a language closely cognate with Waris. His presentation is more detailed than I feel is
needed to adequately describe Waris but the presentation here is based on his. Words in Waris
 5

fall into four types, verbs, adverbs, nominals, and particles. Particles do not take affixation; they
are described in section 2.1. Verbs take affixation of a particular kind and characteristically
encode events, states or processes; they are described in section 2.5.

The distinction between adverbs and nominals is less distinct, but nominals occur in the head
position of phrases bearing case marking, called noun phrases, while adverbs occur in clause-
level slots and generally indicate the time or manner or circumstances of what is predicated by
the verb. Adverbs may not take case marking; they may already contain affixation that shows
their derivation from nominals. Adverbs are like verbs and nominals and unlike particles in that
they may take clause-level topic or prominence marking. Adverbs are described in section 2.2
and nominals in section 2.3. The following chart summarizes these four word classes.

Waris Parts of Speech

---------------------------------------------------------------

| word class | affixed? | slot | semantics | case marked? |

---------------------------------------------------------------

| verb | yes | VP | predication | no |

| | | | | |

| particle | no | clause | manner, mode| no |

| | | | | |

| adverb | yes | clause | manner, time| no |

| | | | | |

| nominal | yes | NP | thing, | yes |

| | | | modifier, | |

| | | | abstract | |

---------------------------------------------------------------

2.1 Particles
Particles never take affixation. One small subgroup of particles occurs in a clause level slot
and there they express the attitude of a speaker towards what he or she is doing or saying.

(1) dihi ye-na kalel a

perhaps 2nd-GEN last.born tag.question
‘(This) is perhaps your youngest child?’
 6

(2) ka véthe-vai bavó

1st put.on-OPT attempt
‘I'll just try to put on (the shoe).’

(3) maim naloh-v

anyway exist.PL-PRS
‘(Regardless of what you think), there are things anyway (in the store, they are
not bought up).’

(4) baua hi-mba Kristus loh-v

intense 3rd-TOP K. exist-PRS
‘He must really be Kristus!’

Another member of this group is slei, 'for a long time, which is bad'

(5) slei ka-m-ba ye-inda-mba wusprómana

always 1st-GL-TOP 2nd-REA.PL-TOP
vi-ni-v-ma
be.angry do.PL-BEN.SG-PRS-Q
‘Do I always have to put up with you all?’

The other particles in this subgroup are:

aiv 'for pity's sake/poor thing'

mava 'sometimes'

sém 'likewise'

yunam 'the time (for something) is close'

mara 'reported information'

Another subgroup of particles includes interjections:

(6) pai mani men-ba

my! what this-TOP
‘My!, what is this?’

(7) mindanam naloh-v os i

many exist.PL-PRS thus intense
‘There are so many of them!’

(This i is nasalized, a phonetic feature not found elsewhere in the sound system of Waris.)
 7

Here are most of the remaining interjections:

léhra 'good, well-done'

oh 'here, take this'

érr expression of pleasure on meeting someone

The final subgroup of particles is the ideophones. The following example (8) contains two
ideophones, by which the speaker described a chicken running back and forth along the peak of
a house and then falling to the ground:
(8) os enggoa-vna-mba os
thus run-CON-TOP thus
engpéh-vna-mba gulungulungulu bi
went-CON-TOP sound sound
‘(The snake-bitten chicken) ran back and forth and fell and hit the ground
(dead).’

Here are a few more examples from this large class:

kolta 'sound of something hitting water'

pépépépé 'sound of someone going along stealthily'

kri 'sound of a frog singing'

puaaaav 'sound of a shotgun'

vó 'sound of someone limping'

vutu 'sound of crowned pigeon singing'

2.2 Adverbs
Adverbs in Waris occupy clause-level slots where they describe the location, manner or time
of the predication. They may bear the clause-level topic/definiteness marker but do not form
compound or derived stems like nominals. Rather, many adverbs are themselves derived forms
arising from nominals (section 2.2.5). Adverbs are divided into time, manner, locative and
interrogative adverbs, the latter which enquire into location, time and manner.
 8

2.2.1 Time Adverbs

Time adverbs are:

déti 'today'

om 'yesterday'

niv 'day before yesterday'

nivnémét 'day before day before yesterday'

simera 'tomorrow

iminisi 'day after tomorrow'

imininov 'day after the day after tomorrow'

daipota 'now'

dawonam 'after a while, for a while'

sewonam 'after a long time'

yunalm 'for a long time, forever'

daha 'after an intervening event'

doara 'previously'

doa 'completed'

mendah 'still, not yet'

elsus 'daily, always'

ata 'will, today'

si 'will, after today'

seilva 'at first'

seilnovra 'for the first time'

 9

némét 'newly'

boasna 'next, the first thing after now'

heva 'at a certain time'

nónam 'at that time'

honanam 'at this time'

okómbanam 'in the daylight'

okómbahóvranam 'in the middle of the day'

sinim 'in the night'

siindulnam 'in the middle of the night'

kuimbóvnam 'when the _kuimb_ bird starts singing' i.e. about 4 AM

makokoónam ‘when the _makoko_ frogs start singing’ i.e. about 4PM

okómbahul 'in the heat of the day'

The derivation of time adverbs ending in -nam is discussed in section 2.2.5. Here are some
examples of time adverbs:

(9) ka-m doara vraho-o

1st-GL previously give-IMP
‘Give it to me now not later.’

(10) hénanam heva hénga-va pró-vav

when? then again-TOP come-FUT
‘Just when will he come back?’

(11) simera heva hénga-va pró-vav nónam

tomorrow then again-TOP come-FUT at that time
‘He will come back tomorrow, at that time.’

(12) mendah pró-v-moa

still come-PRS-NEG
‘(He) is not coming yet.’
 10

2.2.1.1 Time Adverbial Phrases

Time adverbs occur in phrases, containing an obligatory Head plus one or two optional
determiners. Determiners are limited to the following words:

heva, nónam, doa, and days of the week.

Heads are any of the remaining time adverbs, and interrogative adverbials that enquire about
time (section 2.2.4).

Thus example (10) is analyzed as a sequence of

Head hénanam 'when?' plus determiner heva 'a certain time'.

Example (15) is analyzed as Head doara 'previously' plus determiner heva 'a certain time'. Here
the determiner not only specifies the time but adds emphasis, so that the combination means not
just 'previously' but 'a long time previously'.

Example (11) is analyzed as Head simera 'tomorrow' plus determiner heva 'a certain time' plus
second determiner nonam 'at that time'. This phrase is discontinuous, with the second determiner
moved to the last position in the clause, where it adds assurance to what the speaker is
promising.

Here are some more examples.

(13) simera heva ka-va ga-vav mingu-ram

tomorrow then 1st-TOP go-FUT Sunday-ALL
‘I will go tomorrow, Sunday.’

(14) mongo heva tendórini-sambla

some then man-two
ah-a-vna okómbahul
DL-sit-CON heat.of.day
‘One day two men were sitting in the heat of the day.’

(15) doara heva pró-na

previously then come-PST
‘(He/she) came a long time ago.’

(16) doa doara pró-i

complete previously come-RP
‘(He/she) came (earlier today).’
 11

2.2.2 Manner Adverbs

Here is a list of manner adverbs. Some end in -nam and their derivation is discussed in section
2.2.5.

besowonam 'well, nicely'

oiwonam 'slowly'

lélnam 'quickly'

hénga 'again

titnam 'ignorantly'

etetara 'unthinkingly, unintentionally'

nónora 'having knowlege'

ambo 'just'

hoahm 'gently'

esesm 'in reply; in payback'

bambanam 'with noise'

os 'thus'

Here are some examples of manner adverbs:

(17) ka-va nónora loh-v

1st-TOP knowing exist-PRS
I know (it/about it).’

(18) hi-mba etetara vra-na

3rd-TOP unintentionally got-PST
He took it by mistake.’

(19) ambo hi-mba ga-v

just 3rd-TOP go-PRS
He is just going (for no stated purpose).’

(20) bambanam ka-va vra-i

with.noise 1st-TOP take-RP
 12

I took it noisily (I did not steal it).’

2.2.3 Locative Adverbs

Here is a list of locative adverbs:

men 'here (location being established)'

ten 'there (location being established)'

hom 'to here (location already established)'

dom 'to there ( " " " )'

hui 'here ( " " " )'

di 'there ( " " " )'

mi 'in front'

sis 'behind'

hóvra 'in the middle'

Here are some examples of locative adverbs:

(21) hom pró-o

to.here come-IMP
Come here!’

(22) men dihel-v

here exist-PRS
It is here (first mention).’

(23) hui avha-o

here sit.down-IMP
Sit down here!’

(24) sis daha ye-oa pró-o

behind after 2nd-TOP come-IMP
As for you, come behind (us), after (you have done something freetranslation
lse).’
 13

The locative adverbs fall into two classes: cataphoric (location being established) and anaphoric
(location previously established).

The following example from conversation, (25) shows how they are used.In the reply, the first
deictic men points out to the questioner a location which is new information to him. The second
deictic hui reinforces that as now old information.

(25) hi ata ka-va vélaha-vav / men hui

where will 1st-TOP put-FUT / here here
Where should I put it (the load)? / Right here!’

The cataphoric locative adverbs men 'here' and ten there' have a wide distribution in clauses
with the topic suffix. This hanges them into demonstrative pronouns with the meanings 'this one
here (first reference)' and 'that one there (first reference)'. See section 2.3.7.

(26) an-na ten-ba deuv-pa loh-v

who-GEN there-TOP house-TOP exist-PRS
Whose house is that over there?’

2.2.4 Interrogative Adverbs

Here is a list of interrogative adverbs:

es 'how, how many'

hi 'where'

hém 'to where'

hénamini 'from where'

hénanam 'when'

manara 'why, what reason'

manaram 'why, what purpose'

mananam 'what time'

Here are some examples of interrogative adverbs:

(27) manara yi-mba owai ve-i

why 2nd-TOP no do-RP
For what reason did you not do it?’
 14

(28) ti-mba manaram naloh-v

wood-TOP why exist.stacked-PRS
For what purpose is the wood stacked here?’

(29) hém ga-v

where go-PRS
Where is he going?’

(30) es yi-mba ambe-i

how 2nd-TOP come.PL-RP
How did you get here?/How many of you have come?’

(31) Walsa-va manara hev-pa loh-v

name-TOP why 3rd-TOP exist-PRS
What does 'Walsa' mean?’

2.2.5 Derived Adverbs

As stated in the introduction to adverbs, many of them are derived from another class of stem.
In this section we discuss the derived adverbs that are listed above under the various classes of
time, location, manner, and interrogative. Here we show how some adverbs are derived from a
nominal, adjective or a verb. (The basic form _-nam_ means 'time or manner of action'.) See
section 2.5.1.

1. besel 'good' > besowonam 'well'

2. si 'darkness' > sinim 'at night'

3. si 'darkness' + indund 'straight up and down' > siindulnam 'in the middle of the night'

4. tit vev 'to not know' > titnam 'in ignorance'

It is a toss-up whether we call tit a nominal or verbal, since it has no wider distribution in the
language than as here, with the verb vev 'to do' and in titnam.

5. kuimb 'coucal' + óv 'to speak' > kuimbóvnam'4AM' (the time the coucal regularly starts
calling)

6. men 'here' + péhv 'to ascend' > mepéh 'up there nearby'

ten 'there' + péhv " > tepéh 'up there a long way'

7. mani 'what' + -ra 'reason' > manara 'for what reason?'

 15

" " + -ram 'result' > manaram 'for what purpose?'

" " + -m 'goal' > manam 'what goal?'

" " + -nam 'manner' > mananam 'when?'

Note that manara, manaram and mananam are classed as interrogative adverbs, but mani and
manam are classed as nominals.

8. néngv 'to think' + -ra 'reason' > néngara 'let me think, it is for the following reason' (a pause
or hesitation form)

9. eunumbul 'some, remainder' > eunumbulnam 'for the last bit'

si 'night' + eunumbulnam > sieunumbulnam 'during the rest of the night, early in the
morning'

10. hona 'this' + -nam > honanam 'at this time'

no 'that' + " > nonam 'at that time'

11. hona 'this' + -inda 'Locative' > honinda'because of this'

no 'that' + " " > noinda 'because of that, at that time, therefore'

2.2.6 Adverbs Forming Derived Stems.

Adverbs in general do not enter into constructions to form derived stems, but a few _locative_
adverbs are exceptions. In the first two examples below a case suffix changes a locative adverb
into a nominal:

1. sis 'behind' + -mini 'derived from' > sismini 'the one behind'

2. mi 'in front' + -rini 'derived from' > mimirini 'the one in front'

In the next two examples, a locative adverb men 'here' combines with either a verb of motion or
locative suffix to form other locative adverbs:

3. men 'here' + péhv ‘to ascend’ > mepéh ‘up there nearby’

4. men ‘here’ + -óngóhó 'lower' >móngóhó 'down there nearby'

 16

2.3 Nominals
This is the largest class of words, distinguished by occuring in the head slot of NPs and
bearing case marking. It is divided into a number of sub-classes: kin terms (2.3.1), human nouns
(2.3.2), proper nouns (2.3.3), personal nouns (2.3.4), common nouns (2.3.5), personal pronouns
(2.3.6), demonstrative pronouns (2.3.7), interrogative pronouns (2.3.8), quantifiers (2.3.9),
adjectives (2.3.10), demonstratives (2.3.11). After listing these types of nominals I go on to
discuss compound and derived nominal stems (2.4).

Clauses acting as nominals are discussed with relative clauses (6.7). Covert noun classes are
discussed in section (2.3.13).

In the discussion below I will use the following features to distinguish between the various
types of nominals:

1. number (is the basic form plural, or unmarked for number?)

2. type of case marking required (+Human or -Human)? (Described in section 3.3.)

3. can it be possessed, or not?

2.3.1 Kin Terms

Kin terms are unmarked for number, take [+human] case marking (refer to section 3.3 to find
a discussion of this), and can be possessed. The basic form is 1st person, with the 2nd person
requiring a suffixed second person emphatic pronoun, and 3rd person having suffixed -l:

ete 'my older brother'(reference)/'older brother!'(address)

etepev 'your older brother'

etel 'his/her older brother/an older brother'

(32) ete-pev-pa ye-na-mba hi

o.bro-2nd.EMP-TOP 2nd-GEN-TOP where
‘Where is your older brother?’

(For a summary discussion of the -l suffix see 2.4.4).

 17

2.3.2 Human Nouns

Human nouns are a small class with human referents, whose basic form is plural in number
and whose singular and dual are derived by affixation.Human nouns take [+Human] case
marking.

tendó 'men'

indhana 'people'

tuendis 'boys'

mutundis 'girls'

ungevli 'women'

ótól 'children'

The following examples show derivation of human nouns of non-plural number:

tendó + -rini 'ABL' > tendórini 'a man'

ungevli + -rini + sambla 'two' > ungevlirinisambla 'two women'

(33) tendó-mba aevul-v

men-TOP sit.PL-PRS
‘Men are sitting.’

(34) tendórini ombol-m wai-loh-v

man son-GL ACC-stand-PRS
‘A man is standing holding (his) son.’

2.3.3 Proper Nouns

Proper nouns are mainly the names of villages and areas of ground. They differ from the next
class, Personal nouns, in that the latter take [+Human] case marking but Proper nouns do not.

Daumoh 'village name' + -nind 'people' > _Daumohnind_'the people of Daumoh village'

(35) Daumoh-ra ka-va li-loh-v

D.-LOC 1st-TOP sleep-exist-PRS
‘I habitually sleep at Daumoh village.’
 18

2.3.4 Personal Nouns

Personal nouns are the names of humans. Like proper nouns they cannot be possessed, but
they take [+Human] case marking.

(36) Dand-ina ka-va li-loh-v

D.-LOC 1st-TOP sleep-exist-PRS
‘I habitually sleep at Dand's (house).’

2.3.5 Common Nouns

Common nouns are the names of things. Their basic form is unmarked for number, they
require [-Human] case marking, and they may be possessed.

(37) an-na deuv-ram vonga-v

who-GEN house-ALL ascend-PRS
‘Whose house is he entering?’

2.3.6 Personal Pronouns

The various forms of the personal pronouns are given in the following table. (The
irregularities introduced into this paradigm when the topic marker is suffixed are discussed in
section 6.5.).The use of normal and emphatic pronouns in discourse is discussed in section 6.5.
Suffice it to say at this point that emphatic pronouns are similar in use to the Tok Pisin
pronouns combined with 'yet', as: 'mi yet'.

------------------------------------------------------

| | AGENT | GOAL |

| | | |

| | normal emphatic | normal emphatic |

| | | |

|----------------------------------------------------|

| | | |

| 1st Person | ka kav | kam kavm |

| | | |
 19

| 2nd Person | ye pev | yem pevm |

| | | |

| 3rd Person | he hev | hem hevm |

| | | |

| 1st Person | pi piv | pim pivm |

| Inclusive | | |

| | | |

------------------------------------------------------

(The first three pronouns have no number, the 1st person inclusive pronoun means 'more than
one'.)

2.3.7 Demonstrative Pronouns

The demonstrative pronouns fall into two classes, the anaphoric deictics (existance and
location already established) and cataphoric deictics (existance and location being established):

Anaphoric Deictic Pronouns

honi 'this thing/person'

nói 'that thing/person'

honatóngó 'this something (indefinite)'

nótóngó 'that something (indefinite)'

manavema 'something (an action)'

manaema 'something (a thing)'

Cataphoric Deictic Pronouns

menba 'this one'

tenba 'that one'

 20

The derivation of the cataphoric deictics from locative adverbs men and ten by the suffixation of
-ba 'topic' was discussed in section 2.2.3. 'Cataphoric' was used there to refer to information
being introduced to discourse for the first time and 'anaphoric' was used to mean old
information that the speaker expects the hearer to be able to retrieve.

Noun phrases employing one of the two deictics _men-ba_ 'this one' and _ten-ba_ 'that one' fill
the subject of clauses types 1 and 2, sections 5.2.1 and 5.2.2.

The distinction in meaning between the anaphoric deictics _manavema_ and _manaema_ is
neutralized in some of the Waris dialects or village idiolects.

(38) mani ten-ba dihel-v

what that-TOP exist.inanimate
‘What is that (inanimate) thing over there?’

(39) manaema ka-m vraho-o pev-na toko-rini

something 1st-GL give-IMP 2nd.EMP-GEN store-ABL
‘Give me something from your store!’

(40) manavema vé-ne-na

something do-BEN.SG-PST
‘(He) did something to (him).’

(41) honi-na ve-na

this.thing-GEN do-PST
‘(He) did it by means of this.’

2.3.8 Interrogative Pronouns

Here are the interrogative pronouns followed by some examples:

an 'who?'

héni 'which?'

mani 'what?'

manam 'what goal?/why?'

(42) manam ga-v

what.goal go-PRS
‘What's he going for?’

In the following example (43) the Head ('food') and determiner ('what') of the NP occupying the
object slot have become discontinuous due to the fronting of the topic 'food' in the clause:
 21

(43) inne-mba yi-mba mani hala-wol ve-v

food-TOP 2nd-TOP what eat.PL.S-NS.O do-PRS
‘Food, what are you all eating?’

The following example (44) shows how manam plus future tense is the ordinary way of
expressing prohibition:

(44) manam yi-mba hui-va avha-vav

what.goal 3rd-TOP here-TOP sit.down-FUT
‘Don't sit down here!’

2.3.9 Quantifiers

Here are the quantifiers:

mongasal 'one'

sambla 'two'

móngó 'a, one, another; some [certain dialects]'

móngala 'some, some others'

móngao 'just one'

espeta 'a little, a few'

mindanam 'many, a lot'

indakonda 'each'

indkumbi 'all people'

To this list could be added the Indonesian and Tok Pisin numerals, which are used by different
age groups of Waris speaker according to their experience with primary education, whether
under the Dutch or Australians. Waris itself has only the numerals 'one' and 'two', which are
combined to form 'three' and 'four' and sometimes 'five'.

2.3.10 Adjectives
Adjectives are included in the word class 'nominal', section 2, because with topic suffixation
they can function as heads of NPs. See examples (81) and (83) and section 3.4.
Here are a few adjectives:
 22

besel 'good'

sahoklal 'bad'

némél 'new'

gingel 'white'

tokol 'long'

hutel 'short'

aembul 'red, ripe'

pundel 'unripe'

doaramini 'old'

Most adjectives have final -l, but they are not regular in loosing it in derivation processes. See
section 2.4.4 for the relation between some of these adjectives and verbs with the auxiliary vev
'to do'.

2.3.11 Demonstratives

Here are the demonstratives or deictics. Notice that they are not called 'demonstrative
pronouns', this term having already been used above in section 2.3.7 for substantives. The
demonstratives are included in the word class 'nominal', section 2, because, with topic
suffixation, they can function as heads of NPs.

hona 'this'

nó 'that'

snél 'like this'

(45) snél men

like this
‘(It's) like this (one) here.’

(46) nó-mba tendórini-va doa ga-na

that-TOP man-TOP complete go-PST
‘That man (previously referred to) has gone.’
 23

(47) hona ungevlirini an-na

this woman who-GEN
‘This woman is whose (wife)?’

(48) ka-va eva hona-mini-ma

1st-TOP ? this-DER-Q
‘I am not this person you refer to!’

Example (46) above contrasts with the following,(49) in which one slot in the NP is filled with
a personal pronoun rather than demonstrative:

(49) nói tendórini doa ga-na

that.one man complete go-PST
‘That person, the man, has gone.’

For a discussion of deixis and topicalization see section 6.5.

2.3.12 Chart of Nominal Properties

The following chart summarizes some of the syntactic properties of each type of nominal in
order to justify the classification used in this paper. Only Quantifiers (9) and Demonstratives
(11) are not clearly distinguished by this chart.

Numbers on the left correspond to the 11 types of nominals described above (2.3.1 - 2.3.11).
The first column indicates if the nominal in question takes +Human or

-Human case marking (see section 3.3), the second column if it can be possessed, and column
three indicates if it can determine a +Human noun (see 3.1.1). Column four indicates if final _-
l_ is important in this type of nominal. Column five indicates if it can occur with the Topic
suffix (3.4). The blank space means the answer is ambiguous, with not all stems in that class
behaving the same way.

Nom. +/-H 2 3 4 5

-------------------------------------

1 | + | + | + | + ** | + |

| | | | | |

2 | +*| + | + | - | + |

| | | | | |
 24

3 | - | - | - | - | + |

| | | | | |

4 | + | - | - | - | + |

| | | | | |

5 | - | + | - | + *** | + |

| | | | | |

6 | + | - | + | - | + |

| | | | | |

7 | - | - | + | - | |

| | | | | |

8 | - | - | + | - | - |

| | | | | |

9 | - | - | + | - | + |

| | | | | |

10 | - | - | + | + **** | + |

| | | | | |

11 | - | - | + | - | + |

-------------------------------------

Notes:

* Human nouns differ from all other stems in that they are basically plural in number. They all
have animate referents and so when marked for a case that distinguishes between +Human and
-Human they take the +Human allomorphs. However, when the Ablative case marker is
suffixed to them in its extended meaning of 'derivation', the -Human form is used (and the
resulting meaning is 'singular number'). (See 2.3.2).
 25

** See section 2.3.1.

*** Some common nouns are derived from verbs, section 2.4.4.
**** Almost diagnostic for this class; see section 2.3.10.

2.3.13 Noun Classification

Waris nouns are unmarked morphologically for class such as gender, but there are three
systems working in the language to delineate covert noun classes.

Certain case suffixes have allomorphs that collocate only with inanimate, animate, or body part
nouns. (Section 3.3).

There are existential verbs which collocate with certain noun subjects and predicate their
perceived mode of existence, such as sitting, standing, lying prone, hanging, etc. (Section 5.2.1)

Three classes of verb prefixes collocate with only certain nouns:

a. the accompaniment prefixes (Section 4.1.3) collocate only with animate subject or object.
b. the carrying prefixes collocate only with animate subject and inanimate object (Section 4.1.2)
c. the classificatory verb prefixes collocate only with semantically compatible nouns (Section
4.1.4)
The resulting noun classes are not rigid; a speaker may or may not choose to include the
classificatory verb prefix, thereby highlighting or not a semantic feature of what he or she is
talking about.

2.4 Compound and Derived Stems

Compound and derived stems have wide distribution in Waris. There are three ways to form
compound and derived stems, by compounding two noun stems with additive meaning (2.4.1),
by determiner plus determined constructions in which the meanings are not additive but
modified (2.4.2), and by constructions in which at least one of the stems is not capable of
standing alone and is only found in compounds (2.4.3). Some of these derived stems have
already been listed above.

2.4.1 Compounding two noun stems

One strategy used is to combine two noun stems to form one with additive meaning:

engla-mongala 'arms and legs'

ete-boasalel 'older and younger brothers = siblings'

yivuela-valongó 'arrow and bow'

 26

(Similar to this is the reduplication of a locative type adverbial hóvra 'in the middle':

hóvrahóvra 'in the very middle'

Reduplication plays a small role in Waris.)

2.4.2 Determiner plus Determined Constructions

In determiner plus determined constructions the meanings are not additive but modified. (The
order of constituents varies from example to example).

1. noun plus noun

keu 'chin' + ta 'hair' > keuta 'beard'

mona 'road' + hul 'big part' > monahul 'tractor road'

tata 'meat' + muemb 'saliva' > tatamomb 'desire for meat'

2. noun plus adjective

uvi 'banana' + aembul 'ripe' > uviumb 'ripe bananas'

The morphophonemic changes that take place in the formation of this compound stem are one
of the criteria by which such stems are distinguished from noun plus adjective constructions.
Another criterion is by contrasting behaviour in clauses. Noun and adjective may be interrupted
by other constitutents, but a compound stem may not:

(50) uvi doa aembul viló-v

banana complete ripe exist-PRS
‘Bananas are ripe.’

but
(51) *uvi doa umb viló-v
banana complete ripe exist.PL-PRS
‘Bananas are ripe.’

Another criterion is semantics: uvi aembul means 'ripe bananas', but uviumb means 'bananas
that are normally eaten only when they are ripe'

Here are more examples of determiner plus determined compound stems:

3. noun plus adverb:

nini 'armpit' + sengai 'under' > ninisingi 'under the arm'

 27

kembel 'village' + hóvra 'in the middle' >

kembelhóvra 'in the middle of the village'

4. noun plus verb:

The incorporation of verb stems into nominal stems is common in Waris but is not regularly
productive, being more idiosyncratic:

tand vev 'to be angry' + moa 'talk' > tandmo 'angry talk'

léngv 'to like a food a lot' + uvi 'banana' >

uviléng 'one who eats a lot of bananas'

mindil 'dead body' + phov 'to get up' + moa 'talk' >

mindphomo 'talk concerning dead people arising'

(In the first example, tand is not itself a verb, but depends on the verb vev 'to do', and has no
other distribution in the language.)

Here is an example in context of a nominal stem ('from just sitting') incorporating a verb stem
(_a_ 'to be sitting'):

(52) endatand vil-un-v a-ra-angas

sore.backside exist.PL-BEN.PL-PRS sit-LOC-only
‘(People) have sore backsides from just sitting.’

In example (53) the verb stem wehala 'eat' is incorporated in a nominal phrase meaning 'because
of sorcery'. (Someone chewed betel nut in order to work sorcery.)

(53) pul-na wehala-na he-m-ba daha-na

betel.nut-GEN eat-GEN 3rd-GL-TOP die.S-PST
‘(He/she) died because of sorcery.’

2.4.3 Independent stem plus dependent stem

These consist of one stem that can stand alone plus another which can only occur bound:

1. ka '1st' + -sna 'first' > kasna 'me first'

This derived stem occupies a time adverbial slot in a clause:

 28

(54) ka-va ka-sna pró-i

1st-TOP 1st-first come-RP
‘I arrived first.’

2. mie 'pig' + -wonga 'killer' > miewonga 'a good pig hunter'

3. weis 'moon' + -kómba 'light' > weiskamba 'moonlight'

4. nó 'that' + -hal 'area' > nóhal 'that area over there'

5. deuv 'house' + -nilm 'about?' > deuvnilm 'concerning the house'

6. ka '1st' + -mund 'people' > kamund 'the people with me'

7. he '3rd' + -angas 'the very one' > hengas 'this very person'

Here are some examples of the use of the above compound stems:

(55) he-angas kéknam mohvi-na

3rd-alone strongly take.PL-PST
‘These very people took it by force.’

(56) pi-na-hal-angas koasluh-un-v

1st.PL.INC-GEN-area-alone be.missing-BEN.PL-PRS
‘In our area alone (of all areas, power lines) are missing for (you and me).’

(57) hi-mba ka-mund-elm loh-v

3rd-TOP 1st-people-TEL exist-PRS
‘He is one of us.’

A small set of dependent stems always occurs with locative suffixation:

-ras-ra 'on the vines'

-mus-ra 'in the midst'

-s-ra 'on the surface'

-sha-ra 'in the crack'

Here are examples of these in constructions with independent noun stems:

heo-ras-ra 'on the vines clinging to a heo tree'

wohana-mus-ra 'in the midst of the wohana bamboo'

 29

sime-s-ra 'on the floor'

sime-sha-ra 'in the floor crack'

One dependent stem -sné 'like' is derived from the deictic snél 'like (this)'. It is peculiar in that it
triggers goal marking on the noun it is suffixed to:

po 'water' + -m 'goal' + -sné 'like' >

_pomsné_'a little water'

This occurs in the following clause:

(58) pomsné ka ne-mn-i

little.water 1st eat-BEN-RP
‘I want a little water to drink.’

Another derived dependent stem is the common noun mel 'hole', which sheds final _-l_ to
become a dependent stem meaning '(the) inside':

(59) ilsas-va péthe-me-ra loh-v

weed.roots-TOP ground-inside-LOC exist-PRS
‘The roots of weeds are under the ground.’

There is a small class of dependent stems that fall under the heading of 'intensifier' (positive or
negative):

-kumbi 'big'

-kola 'everyone; old; poor; endearment'

-peta 'little'

-nanoa 'true'

-ta 'baby'

-hui 'without'

-koa 'lucky; my!

Following are examples:

1. winde 'dog' > windeta 'puppy'

2. etel 'o.brother' > etelnanoa 'true older brother'

 30

3. Muk a name > Mukkumbi 'important man Muk'

4. Bov 'Bob' > Bovkola 'dear old Bob'

5. inne 'food' > innehui 'without food'

6. indhana 'people' > indkumbi 'everyone'

7. " " > indkola 'everyone'

8. ka '1st' > kakoa 'lucky me; poor me'

9. deumb 'tree type' > deumbkoa 'a huge deumb tree'

10. sumb 'bucket' > sumbkola 'old useless bucket'

Compounds with -koa and -kola could almost be classed with the interjections (section 2.1).
The suffix -mini looks a little like the Ablative case marker -namini and has the meaning
'derived from'. Here are two examples of its use to derive stems, the first with a time adverbial
to derive an adjective stem, and the second with a deictic to derive a locative stem:

doara 'previously' > doaramini 'old'

nó 'that' > nómini 'from there'

Finally I will mention some derived stems that are built from dependent noun or verb stems
alone. There are very few of these:

1. amba- 'outside?' + -ra 'at' > ambara 'located outside the house'

amba- " + -ram 'to' > ambaram 'going outside the house'

2. néng- 'to think' + -m 'goal' > néngam 'let me think, it's about this...' (a pause or hesitation
form).

(60) doa ka-va ga-na nénga-m bras-m

complete 1st-TOP go-PST think-GL rice-GL
‘OK I went, let me think for what, for rice.’

(61) doa ka-va ga-na nénga-ram misin-indam

complete 1st-TOP go-PST think-ALL mission-ALL
‘OK I went, let me think where, to the mission.’
 31

2.4.4 The Function of -l

Seiler (1984 and 1985) discusses thoroughly the function of -l in Imonda, a language closely
cognate with Waris. He concludes that it has the basic meaning of 'relational or part of whole'.
In my analysis of Waris I come to no such neat conclusion; instead here is merely a summary of
its functions. It is convenient to discuss -l with nominals and before discussing verbs because it
occurs with both.

1. It regularly occurs with kin terms: ete 'my older brother/older brother!' > etel 'his older
brother'

2. It occurs with -m 'goal' on any verb stem to form a gerund with the meaning 'intention'.

(62) doa ka-va ga-lm ve-v

complete 1st-TOP go-GER do-PRS
‘OK, I want to go!/I'm ready to go!’

(63) manara yi-mba ga-lm-ba owai ve-i

why 2nd-TOP go-GER-TOP fail
toko-ram-ba
do-RP store-ALL-TOP
‘Why did you fail to go to the store?’

This 'gerund' is the same as the Telic case marker that occurs on nominals:

(64) popoli andava-lm nilha-na

cocoon butterfly-TEL change.into-PST
‘The cocoon turned into a butterfly.’

3. Many adjectives end in -l but only a few loose it in a derivation process:

sahoklal 'bad' > _sahokla vev_ 'to do bad to'

besel 'good' > _bes vev_ 'to taste good' > _besowonam_'well'

4. The deictic snél 'like this' has removable -l, which must come off in the presence of Topic
marking:

sné-mba 'one like this'

5. The common noun mel 'hole' can shed final -l to become a dependent noun stem -me 'the
inside', as in péthe-me-ra 'under the ground'

6. An unpredictable but significant number of verbs can take -l to become nominals, the reverse
of the process in 3. above. A couple of time adverbials fit in this category, too:
 32

powalv 'to break' > powalal 'broken piece'

iav 'to possess' > ial 'possessions'

sahokla vev 'to do bad to' > sahoklal 'bad'

mendah 'still' > mendahal 'immature (animal)'

hesna 'he first' > hesnal 'first fruits'

2.5 Verbs
Verbs form a distinct class in Waris because of the characteristic affixation they bear. This is
described in section 4. Semantic classes of verbs based on case frames are described in section
5.1.

In this section I will give examples of the unpredictable way some verbs are related to other
parts of speech. (For verbs entering determiner plus determined constructions to form derived
stems see section 2.4.2. For verbs taking -l to become nominals see 2.4.4.6.)

2.5.1 Noun Incorporation in Verbs

This is a rare phenomenon in Waris. The only example I have heard is given below in (68).
The inverse process, verb incorporation in nominal or adverbial stems is common but
unproductive.

However, these is a large class of 'verbs' which make up verb phrases consisting of what might
be called gerund plus helping verb; the helping verb is always vev 'to do, make'. However, none
of these 'gerunds' has wider distribution in the language, and so they do not really fall under the
heading of noun incorporation. In the following examples I show how some of these gerunds
are idiosyncratically related to other parts of speech, such as adverbials or nominals, the latter
including nouns or adjectives. Refer to section 2.4.2 for more examples.

1. tit vev 'to be ignorant' > titnam 'unknowingly'

2. bes vev 'to taste good' > besel 'good' > besowonam 'well'

(65) hi-mba besel hev-m

3rd-TOP good 3rd-GL
‘He/she is good (in intrinsic moral character or appearance).’

(66) hi-mba besowonam a-v

3rd-TOP well sit-PRS
 33

‘He/she is well (not sick).’

(67) bes ne-o

good eat-IMP
‘Taste it!’

3. tand vev 'to be angry' + moa 'talk' > tandmo 'angry talk'

4. télp vev 'to urinate' > télpol 'urinary bladder'

5. sahokla vev 'to do bad to' > sahoklal 'bad' > sahonam 'badly'

6. iav 'to possess things' > ial 'possessions'

7. powalv 'to break' > powalal 'broken piece'

(68) ungevlirini-m wosepul-kovha-omana

woman-GL vine-cut-place.to.place
vé-ne-v
do-BEN.S-PRS
‘A woman continues to suffer from uncontrolled menstrual bleeding.’

In the above example (68) the noun stem wosepul 'vine type' is incorporated in a derived verb
stem with the verb stem kovha 'to cut'. The combination has an idiomatic meaning reflecting a
folk belief that cutting that vine can cause a woman to experience uncontrolable menstrual
bleeding.

3 Noun Phrases

In this section I first discuss the constituents of the NP and their order (3.1), then coordination
of NPs (3.2). Section (3.3) treats case marking and (3.4) briefly introduces topic marking.

3.1 NP Constituents and Order

The basic order of constituents in NPs is this:

+Head +/- determiner +/- adjective +/- coordination +/- case marking (+/- Topic)

(Topic is not a constituent of the NP but a pragmatic function of a whole utterance; it is

mentioned here to prepare the reader for recognizing NPs suffixed with Topic.)
 34

3.1.1 The Minimal NP

According to our definition of nominals (section 2.3), any nominal may function as the head
of a NP. Next, it may be determined by another nominal of one of the following types: pronoun,
kin term, quantifier, possessive NP, relative clause, or another noun. Below are minimal
examples, each labelled as to its derivation from the above constituents:

(Notes:1. the examples given are analyzed as Head + determiner except in the case of
possessive NP, which precede the head normally. This is because pronouns, which more often
refer to human beings, are therefore more 'topical' and are put at the front of the clause. 2. topic
marking is present in these examples not because it is under consideration at this point but
because the examples generally fit into a discourse context requiring Topic marking. See
sections 3.4 and 6.5. 3. a possessive NP consists of a head plus -na 'genitive' 4. relative clauses
are discussed in section 6.7.)

***kin term + kin term (does not occur)

***kin term + personal pronoun:

(69) etel ka-va loh-v

older.brother 1st-TOP exist-PRS
I am the older brother.’

***kin term + quantifier:

(70) etel-va sambla e-loh-v

.brother-TOP two DL-exist-PRS
There are two older brothers.’

***kin term + relative clause:

(71) etel-va [ten-ba mendah a-v-pa]

older.brother that-TOP still sit-PRS.TOP
hi-mba Poso-ra a-v
3rd-TOP name-LOC sit-PRS
‘The older brother [who is still alive] lives at Posoa.’

***noun + noun:

(72) pai sava ka-va ne-v

name taro 1st-TOP eat-PRS
‘I am eating taro (called) Pai.’
 35

***noun + pronoun:

(73) Luk hev-hó ga-v

L. 3rd.EMP-alone go-PRS
‘Luke is going alone.’

***noun + possessive NP:

(74) deuv-pa ka-na-mba loh-v

house-TOP 1st-GEN-TOP exist-PRS
‘The house is mine.’

In the above example (74), determiner follows Head, which is fronted in the clause for
topicalization. It contrasts in meaning with the following example (75):
(75) ka-na-mba deuv-pa loh-v
1st-GEN-TOP house-TOP exist-PRES
‘I have a house.’

***pronoun + pronoun:

(76) ka-va kav-hó ga-v

1st-TOP 1st.EMP-alone go-PRS
‘I am going alone.’

***pronoun + kin term:

(77) hi-mba aral loh-un-v

3rd-TOP father exist-BEN.PL-PRS
‘He is father to them.’

***pronoun + relative clause:

(78) hi-mba [no-mba ga-l-m-ba ve-v-ra-va]....

3rd-TOP that-TOP go-GER-GL-TOP do-PRS-IRR-TOP
‘He [that wants to go]....' = 'Whoever wants to go...’

***pronoun + quantifier:

(79) ka-va sambla e-loh-v

1st-TOP two DL-exist-PRS
‘There are two of us.’

***quantifier + determiner: (this order is not preferred, except for the case of quantifier +
interrogative pronoun, as follows)
 36

(80) sambla an
two who
‘Who are the two?’

***adjective + zero: (Head deleted)

(81) besel-va hi
good-TOP where
‘Where is the good one?’

***adjective + personal pronoun:

(82) besel hi-mba loh-v

good 3rd-TOP exist-PRS
‘He is well.’

***adjective + possessive NP:

(83) tokol-va ka-na-mba hi

long-TOP 1st-GEN-TOP where
‘Where is my long one?’

***adjective + relative clause:

(84) besel-va [ten-ba ka-va wulpró-i-va] hi

good-TOP that-TOP 1st-TOP bring-RP-TOP where
‘Where is the good one [which I brought]?’

***possessive NP + zero: (Head deleted)

(85) ka-na dihel-v

1st-GEN exist.inanimate-PRS
‘I have one.’

(86) Bov-na men

name-GEN here
‘Here is Bob's.’

***possessive NP + noun:

(87) he-na moa-mba novol ve-loh-v

3rd-GEN talk-TOP write do-exist-PRS
‘His talk is written.’
 37

In the following two examples of possessive NP plus noun, the Head follows the determiner in
the normal order (88) and is reversed in order to front the Head for topicalization (89).
(88) hev-na deuv-ram ga-v
3rd.EMP-GEN house-ALL go-PRS
‘He is going to his house (not to someone else's house).’

(89) deuv-ram hev-inam ga-v

house-ALL 3rd.EMP-ALL go-PRS
‘He is going home (not somewhere else).’

***possessive NP + pronoun: (this does not occur)

***possessive NP + kin term:

(90) ye-na mo-pev-pa hi

2nd-GEN daughter-2nd.EMP-TOP where
‘Where is your daughter?’

***possessive NP + possessive NP: (embedding of possessive NP)

(91) ka-na aral-na deuv-pa ten loh-v

1st-GEN father-GEN house-TOP there exist-PRS
‘My father's house is there.’

Here are two more examples of embedding. In (92) a demonstrative hona 'this' determines a
following Head, in which a possessive NP + noun are embedded:
(92) hona ye-na moa-mba besel loh-v
this 2nd-GEN talk-TOP good exist-PRS
‘This talk of yours is good.’

In the second example (93) a demonstative pronoun nói 'that one' determines a Head in which
possessive NP + kin term are embedded:

(93) nói ka-na eindil os ve-na

that.one 1st-GEN grandfather thus do-PST
‘That one, my grandfather, did thus.’

***relative clause + relative clause: (does not occur)

***relative clause + other determiners: (This combination occurs only in the reversed order,
that is, with the 'heavy' relative clause following the other, as in examples (71), (78), (84)
above.) Relative clauses are discussed in (6.7); the examples here are to show how they can be
embedded in the NP to a limited extent.
 38

3.1.2 NP with Adjective

Constructions in which an adjective follows a determiner plus embedded noun phrase were
first analyzed as highly embedded NPs but are now analyzed as examples of Topic-Comment
clauses, with one NP filling a topic slot and another filling a comment slot. These clauses are
discussed in 5.2.4. Here are a couple of examples to show how they contrast with the above
types of NP.

(94) etel ka-va besel

older.brother 1st-TOP good
‘I (am) a good older brother (to him).’

(95) deuv-pa ka-na-mba sambla sahoklal

house-TOP 1st-GEN-TOP two bad
‘My two houses (are) no good.’

3.2 Coordination of NPs

NPs may be coordinated by simple juxtaposition or by use of one or more of a selection of
several markers, with somewhat different meanings.

3.2.1 Coordination by the Comitative Marker

Human NPs may be coordinated in pairs by the Comitative case marker -i. Both or only the first
may be marked.

(96) Sak-i Luk-i hev e-nga-na

name-COM name-COM 3rd.EMP DL-go-PST
Sak and Luke went together.’

(97) Mona-i Das ah-a-v

name-COM name DL-sit-PRS
Mona and (her husband) Das are sitting together.’

Here is an example, (98), of such a coordinated NP filling a determiner slot:

(98) mie Nava-i-na Sowai-i-na Soahpeta-na
pig name-COM-GEN name-COM-GEN name-GEN
The pig belonged to Nava and Sowai, of Soahpeta village.’

Notice that the name of the village bears no coordination suffix and that with the embedded
Heads, the comitative suffix precedes the genitive suffix.

Coordination by means of the Comitative marker emphasizes the cooperative or interpersonal

aspect of the human NPs. Sometimes, no coordinator is used, resulting in a list:
 39

(99) Pita Sak Luk Néngai ka-ngas

name name name name 1st-alone
Pita, Sak, Luke, Néngai, just us’

3.2.2 Coordination by the Genitive Case Marker

Another coordinator of NPs is -na 'genitive'. It is used with non-human NPs, and with a few
lexicalized expressions with human NPs. The genitive marker can also mean 'close association'
(3.3).

(100) onga-na toandpol

wife-GEN offspring
‘wife and children’

(101) tendó-na ungevli-na

men-GEN women-GEN
‘husbands and wives’

(102) yes-na tata-na hi-mba hala-v

sago-GEN meat-GEN 3rd-TOP eat.PL-PRS
‘They are eating meat with their sago.’

3.2.3 Coordination by -e

The suffix -e has the syntactic function of marking direct speech, with the meaning of
'emotional involvement by the speaker'. It is also used to coordinate NPs and here I analyze it as
not having the same semantic function, but rather serving to slow down the rate at which
information is presented in the list and break it into more easily processed bits. It is used far
more extensively in translated Scripture than in myths, for example:

(103) honi amb-na tempel-andra-nind-e

these come.PL-PST temple-work-people-LIST
tempel-polisi-kuwul-e
temple-police-headman-LIST
‘These came, the temple workmen and the head of the temple police.’

3.2.4 Lists with the Verb 'to do'

NPs are sometimes coordinated into lists with the verb 'to do'. (For the use of this verb in
joining clauses, see section 6.2.x.)
 40

(104) ve-na pilotsambla Bov-i ka móngó

do-PST two.pilots Bob-COM 1st
duang-na mol-ombol-sambla eva guru nói
another white.man-GEN daughter-son-two too teacher
haiskul-na
that.one high.school-GEN
‘(They included) two pilots, Bob with me, another white man's’ daughter and
son the two of them, and too that high school teacher.

3.2.5 Coordination of Alternatives

The clitic -ka coordinates alternative NPs into one:

(105) suwembó tikla-ra-ka tombangla-ra-ka vonga-i

c.pigeon branch-LOC-ALT pandanus-LOC-ALT ascend-RP
‘The crowned pigeon (characteristically) flies up to a tree limb or to a
pandanus.’

3.2.6 Pronominal Copy

A special case of NP coordination is pronominal copy, in which the NPs are co-referential:

(106) etel [ten-ba mendah loh-v-pa] hi-mba mendekli

o.brother that-TOP still exist-PRS-TP 3rd-TOP big
‘The older brother [who is still alive], he is big.’

This brings 'older brother' back into conciousness of the hearer after he or she processes the
relative clause.

3.3 Case Marking of NP

The use of case marking in Waris is discussed in detail in Brown (1988). Here I merely
summarize the analysis given there. Waris employs eight basic markers to show semantic
relations within the NP (as with Comitative) and within the clause (as with Goal). These
markers occur as clitics on the right-most element of the NP, with a few exceptions. Here is a
summary of the case markers:

-------------------------------------------------------------

| Label | Clitic | Basic Meaning | Extended Meaning |

 41

-------------------------------------------------------------

| Ablative | -rini* | motion away | derivation |

| | | | |

| Allative | -ram* | motion toward | purpose |

| | | | |

| Comitative | -i | accompaniment | |

| | | | |

| Dative= | -m* | goal, animate | benefactor, recipi- |

| Goal | | object | ent, absolutive, |

| | | | accompaniment |

| | | | |

| Genitive | -na | possession, | close association |

| | | instrument | |

| | | | |

| Locative | -ra* | location | reason, inanimate |

| | | | cause |

| | | | |

| Manner | -nam | manner | |

| | | | |

| Telic | -lm | end of action | purpose (gerund) |

| | | or state | |

| | | | |
 42

-------------------------------------------------------------

Note: * marks those clitics which have allomorphs for non-human and human referents, and
allomorphs for singular and plural within [+human]. The following chart gives these details:

--------------------------------------------------------

| | Clitic |

| Label |--------------------------------------------

| | Non-human | Human singular | Human plural |

--------------------------------------------------------

| | | |

| Ablative | -rini* | -namini |

| | | |

| Allative | -ram | -inam** | -indam*** |

| | | | |

| Locative | -ra | -ina** | -inda |

| | | |

| Goal | -m | -val (body part) |

| | | |

--------------------------------------------------------

Notes: * when the ablative clitic -rini is functioning in its extended meaning of 'derivation from'
with a human noun (section 2.3.2), it is this base form which is used rather than one of the
[+human] allomorphs:

tendó 'men' + -rini > tendórini 'a man'

** In at least two dialects of Waris (Soah and Wainda) the distinction between singular and
plural [+human] Locative and Allative case is neutralized and only the plural form is used.
 43

*** The allative clitics can be analyzed into Locative plus Goal, -m. But it seems strained to
analyze the Manner clitic, -nam (first table), into Genitive plus Goal, because the meanings do
not add up. See (3.3.2), functions 1,3,11.

Following are a few examples (107 - 121) of case marking. Note that case clitics may occur
together.

(107) an-rini yi-mba

who-DER 2nd-TOP
‘What clan or village do you belong to?’

(108) ilpa-va ne-i-va doa péthe-rini-m

roach-TOP eat-RP-TOP complete ground-DER-GL
holope-i
descend-RP
‘(The chicken) having eaten roaches (on the roof of the house), it then came
down in order (to eat) what is found on the ground.’

(109) sambla-nam ne-wol-o

two-MAN eat-NS.O-IMP
‘Take (the pills) two at a time!’

(110) hi-mba sambla-lm veka ve-i

3rd-TOP two-TEL go.and.come do-RP
‘He went and came two times.’

(111) yi-mba besel-na-nam pró-na

2nd-TOP good-GEN-MAN come-PST
‘You came when it was a good time.’

(112) inne-mba wan-na dihel-v

food-TOP salt-GEN exist.inanimate-PRS
‘The food has salt on it.’

(113) he-m-ba daha-i

3rd-GL-TOP die-RP
‘He is dying.’

(114) indhana mina kapol-m dehu-v

people fuel plane-GL feed.PL-PRS
‘People are putting fuel in the plane.’

(115) Pita-i-m Yon-i-m pueil ninge-wul-in-na

P.-COM.GL Y.-COM.GL fence tie-PL.S.NS.O-BEN.PL-PST
‘They surrounded Peter and John.’
 44

(116) indhana-m-ba wóh-va ka-va

people-GL-TOP goods-TOP 1st-TOP
won-i-pró-i
ACC.PL-bring.PL-come-RP
‘I just came bringing goods and accompanying people.’

(117) popoli mana-lm nilha-vav

cocoon what-TEL change.into-FUT
‘What will the cocoon change into?’

(118) englis-na moa-na ka-va ishó-mana-vna

E.-GEN talk-GEN 1st-TOP speak-BEN.SG-CON
‘I was talking English to him/her.’

(119) obat-ra ka-m-ba kanandha-na

medicine-LOC 1st-GL-TOP get.well-PST
‘I got well because of medicine.’

(120) na-andra ka-va ti-ram ve-v

sago-work 1st-TOP tree-ALL do-PRS
‘I am working sago with the purpose of (later having sago to eat to be able to
cut down) trees (to make a garden).’

(121) aral ongal-m vra-na ombol-ina

father wife-GL get-PST son-LOC
‘The father got a wife for the reason of his son (needing one).’

For a classification of clauses based on their associated cases see section 5.2.

3.3.1 Details of Case Marking

Case clitics generally occur only on the right-most noun to which they apply:

(122) ka-na ara-na deuv-ra dihel-v

1st-GEN father-GEN house-LOC exist.inanimate-PRS
‘It is in my father's house.’

Sometimes, as a matter of style rather than denotative meaning, the clitic my be copied onto
preceding nouns:

(123) hi-mba deuv-ra ka-ina a-v

3rd-TOP house-LOC 1st-LOC sit-PRS
‘He is sitting in my house.’
 45

There is one construction which _requires_ a case clitic to be copied back on preceding elements,
an embedded purpose clause ending with -lm 'purpose'. It requires -m 'goal' to be copied on
the preceding subject and object:

(124) hi-mba pró-na [ye-na-m kél-m

3rd-TOP come-PST 2nd-GEN-GL bone-GL
wosapr-in-lm]
strengthen-BEN.PL-TEL
‘He came [in order to make you all strong].’

Here is another example of two case clitics occurring together:

(125) aral-va he-na-na popol a-v

father-TOP 3rd-GEN-GEN strength sit-PRS
‘The father lives/exists by means of his (son's) strength.’

3.3.2 Wide Distribution of Goal Marking

The case suffix -m, called Goal, has a wide distribution in Waris. Its function has been
generalized to cover a range of semantic case roles in cluding Benefactee, Recipient, and
Purpose. Here is a summary of its functions.

1. Combines with Locative marker to form Allative:

(126) deuv-ra a-v / deuv-ram ga-v

house-LOC sit-PRS / house-ALL go-PRS
‘He/she sits at home. / He/she goes home.’

2. Marks Goals:

(127) po-m ga-v

water-GL go-PRS
‘He/she is going for water.’

3. Combines with _-l_ to form Telic marker on nouns and verbs:

(128) pind-va winde-lm e-nilha-wol e-ve-na

possum-TOP dog-TEL DL-change.into-NS.O DL-do-PST
‘The two possums changed into dogs.’

(129) ka-va yimund-ra vonga-lm ve-v

1st-TOP ladder-LOC ascend-TEL do-PRS
‘I want to ascend the house ladder.’
 46

4. Marks other elements (S and O) in an embedded purpose clause. (Refer to example 124
above.)

5. Marks Benefactee:

(130) ka-m-ba os ishó-mna-vna

1st-GL-TOP thus converse-BEN-CNT
‘He was talking like that to me.’

6. Marks Recipient:

(131) dang ka-m li-ra-ho-o

pandanus 1st-GL CLAS-get-REC-IMP
‘Give me the pandanus!’

7. Marks incompletely-affected Goal:

(132) hi-mba ti-m he-tha-v

3rd-TOP tree-GL chop.down-action-PRS
‘He is chopping on the tree.’

8. Marks Goal of sensory verbs:

(133) tuawa-m nungl-u

bird-GL look.PL-IMP
‘You all look at the bird!’

(134) ka-m hill-u

1st-GL hear.PL-IMP
‘You all listen to me!’

9. Marks animate Patient:

(135) ungund-rini-m hélvakomandha-na

enemies-DER-GL kill-PST
‘He killed an enemy.’

10. Marks Absolutive subjects (lack of control):

(136) obat-ra ye-m kanandha-v

medicine-LOC 2nd-GL get.well-PRS
‘You are getting well because of the medicine.’

11. Combines with Genitive (-na)to form Manner marker:

 47

(137) pil sambla-nam ne-wol-o

pill two-MAN eat-NS.O-IMP
‘Take the pills two at a time.’

12. Marks Accompanied:

(138) kav-na boaslal-m ka-va won-pró-i

1st.EMP-GEN y.bro-GL 1st-TOP ACC.NS-come-RP
‘I have just come bringing my younger brothers.’

3.4 Topic Marking of NP

The last, right-most marking on NPs is topic. 'Topic' at this point in the paper is a label for a
series of related functions including topic, resumptive topic and definiteness. Generally, all
topics are given information, and topic marking is a pragmatic function relating to a whole
utterance or text, not limited to one NP. Topic functions are discussed in section 6.5, but for
convenience some examples are given here (139 - 144) to illustrate the uses of this marking.
The allomorphs of the Topic marker are discussed in section 6.5.

Topic is what is given as definite by the speaker.

An adjective filling the Head slot of a NP requires topic affixation, indicating a _definite_
referent.

(139) besel-va hi
good-TOP where
‘Where is a/the good one?’

In the following two examples (140) and (141), Topic marking has been translated as 'the' in
English to indicate that the speaker was referring to certain houses known to him and which he
assumed were known also to the hearer, either from a previous mention in the discourse, or by
the speaker pointing out the two houses in sight.

(140) deuv-pa sambla-va besel e-loh-v

house-TOP two-TOP good DL-exist-PRS
‘The two houses are OK.’

(141) deuv-pa besel-va sambla e-loh-v

house-TOP good-TOP two DL-exist-PRES
‘The good houses are two (in number) = There are two good houses.’

Resumptive topics have their own clitic, -oa:

 48

(142) pi-oa e-nga-vai

1st.PL.INC-RT DL-go-OPT
‘As for you and me, let's the two of us go.’

(143) he-oa sambla-oa deuv-ram e-nga-ra

3rd-RT two-RT house-ALL DL-go-IRR
‘As for the two of them, let them go home.’

In this example (143) the resumptive topic marker is not merely suffixed to 'two' but is copied
back onto the other element of the NP, the 3rd person pronoun. This seems to be a full, precise
style of speech but with the same denotative meaning.

The regular (non-resumptive) Topic marker is sometimes treated the same way, example (144),
where both elements of the NP, 'this' and 'woman' are suffixed.

(144) hona-mba ungevlirini-va manam ga-v

this-TOP woman-TOP why go-PRS
‘This woman, what is she going for?’

To conclude the description of NP I mention the fact that in the absence of a verb the question
clitic -ma my occur affixed to a NP:

(145) yi-mba tendórini loh-v-ma

2nd-TOP man exist-PRS-Q
‘Are you a (married) man?’

(146) yi-mba tendórini-ma

2nd-TOP man-Q
‘Are you a (married) man?’

4 The Verb Phrase

The verb phrase in Waris consists of three positions, a core, pre-core and post-core. The core
is filled by a stem or compound stem. The pre- and post-core each consist of a series of
positions of affixes. A minimal verb phrase consists of a core filled by a bare verb stem with no
affixation. Some constructions require that some of the affixation be placed on a helping verb
rather than on the core. The VP then becomes two phonological words which may be separated
in the clause by other constituents.This is frequently the case with habitual or continuous
predications. In this chapter I discuss the pre-core positions (4.1), post-core positions (4.2), and
verb serialization (4.3).
 49

4.1 Pre-Core Positions

The pre-core consists of five positions, numbered from right to left, with the following
information.

1. number of Subject

2. number of things being carried

3. number of people being accompanied

4. noun class

5. location (proximity to speaker)

4.1.1 Subject Number

Waris verbs mark Subject number in several ways. Stem vowel raising can indicate plural
(more than 2) subject and suppletive forms can indicate the same. Dual is marked by prefix e-
or by reduplication of stem-initial vowel. Sometimes plural is marked by prefix a-.

Subject number prefixes always go in the first pre-core position, next to the stem. The stem loh
'to stand, exist' takes dual prefix e- and, depending on dialect of Waris form the plural by prefix
a- or by a suppletive stem:

(147) loh / e-loh / a-loh / lovah

exist / DL-exist / PL-exist / PL.exist
‘One exists / two exist / many exist / many exist.’

The stem ishó 'speak' forms the dual by reduplication of the initial i with h interposed. Plural is
indicated by stem vowel raising:

(148) ishó / ihishó / ishu

say / two.say / many.say
‘One says / two say / many say.’

Another, relatively unimportant, way of marking plural subject is by use of the manner suffix -
pia 'completely' mentioned later in section 4.2.3. As discussed in that section it adds the
semantic component 'completness of action' to verb stems of either singular, dual or plural
subject number. But with certain verb stems it has come to take on the meaning of 'everyone', as
in example (149) and (150).

(149) pró-pia-o
come.SG.S-complete-IMP
 50

‘Everyone come!’

(150) doa ga-pia-na

complete go.SG.S-complete-PST
‘All have gone away.’

4.1.2 Number of Things Carried

Verbs of motion, sitting and standing can be prefixed for the number of things being held or
carried at the same time:

Number of Things Carried: 1 2 3+

---------------------------------------------

Prefix: wul- hai- i-

This prefix goes in the number two position left of the stem:

(151) ti hai-e-nga-v
wood two.carried-DL.S-go-PRS
‘Two are each carrying a piece of wood.’

The common form for one person carrying one thing has been simplified:

(152) hi-mba surat [*wul-nga-v > goa-v]

3rd-TOP letter one.carried-go-PRS
‘He is carrying a letter.’

(153) wan yi-mba wul-a-v-ma

salt 2nd-TOP one.held-sit-PRS-Q
‘Do you have any salt?’

4.1.3 Number of People Accompanied

Most verbs of action can be prefixed to indicate the number of people one is accompanying:

Number of People Accompanied: 1 More than 1

------------------------------------------------

Prefix wai- won-

This prefix occupies the next position to the left of the stem:
 51

(154) indhana-m-ba wóh-va ka-va

people-GL-TOP cargo-TOP 1st-TOP
won-i-pró-na
PL.ACC-PL.held-come-PST
‘I came bringing people and cargo.’

In general, there is a relationship of superior - inferior when accompanied is marked on a verb:

(155) aral ombol-m wai-a-v

father son-GL ACC-sit-PRS

‘A father sits with his son.’

(156) *ombol aral-m wai-a-v

son father-GL ACC-sit-PRS
A son sits with his father.

Accompaniment prefix can occur without an explicit Goal, as in the following example (157).
The nature of implicit Goal is retrieved from context or inferred from cultural knowledge.

(157) Apraham won-ga-na

name ACC.PL-go.SG-PST
Abraham went (and took his family).’

4.1.4 Noun-Classifying Verb Prefixes

Waris nouns fall into covert (non-morphological) classes based on semantic features perceived
by the native speakers. I first dealt with this in Brown (1981), and Seiler later clarified it and
provided much more data from the Imonda language (Seiler 1984b). The original label I used
was 'shape', and this is inadequate because semantic features other than shape are involved. In
this section I will merely give a few examples. The noun class of the NP argument of a verb,
either Subject or Object, is marked on the verb by a prefix, which goes in the left-most position
away from the verb.

Seiler's point about classificatory verb prefixes in Imonda also applies to Waris: there is a
close relationship between the classificatory verbs and serial verb constructions. In fact the
former presumably arose from the latter, because many of the classificatory prefixes can be
identified with verb stems which have a similar semantic content. Thus, the following
example (158) the verb stem vét- 'remove hot from the fire' occurs as a classifier prefixed to the
verb stem meaning 'get'. The total construction means not merely 'I got bread from the fire',
which is not true in the context of this example. What it does mean is 'I got bread that someone
originally cooked in a fire (and is, as a matter of fact, now cold)'.
 52

(158) plaua-sambla ka-va vét-rombo-t-na

bread-two 1st-TOP CLAS-get-DL.O-PST
‘I bought two pieces of bread.’

Here is another example (159), in which the total verb means 'fetch something round with a
stem'.

(159) lemo ka-m putil-ka-mn-o

lemon 1st-GL CLAS-fetch-BEN-IMP
‘Go and get me a lemon!’

In this example the noun classifier _putil_ refers to round things with stems, and to introduced
objects like balls. Another classifier refers to things in a container: aiwó:

(160) puemb aiwó-won-nongend-na

flood CLAS-ACC.NS-descend.PL-PST
‘The flood went down (the river) taking the two (children) inside (the netbag).’

The verb stem nongend is not singular, to agree just with 'flood', but plural to agree with 'flood
+ two children', so a literal translation would be 'the two (children) and the flood went down
(the river) together, the children inside a container.' This is because in Waris human subjects are
more salient than non-human, and if they are involved in the action they must be referenced on
the verb. In this case the reference is by a plural subject number verb stem.

The classifier tuvul- is derived from the verb stem tuvul 'to exist, having been previously cut
off'. It is related to an verb stem tovhav 'to cut off', which collocates with certain garden objects
such as sugar cane and seed yams:

(161) wembtom-ba ka-va tuvul-piha-v

seed.yam-TOP 1st-TOP CLAS-go.down-PRS
‘I am bringing the (previously cut) seed yam down (to the new garden).’

It can also take on figurative meaning:

(162) he-na wevsa-va tuvul-in-v

3rd-GEN custom-TOP cut.off-BEN.PL-PRS
‘Their custom is too short for them = unhelpful.’

4.1.4.1 Existential Verbs

Not only do some Waris verbs take prefixes to indicate the class of the noun argument, there
is a set of existential verbs which collocate with only certain nouns. This subject was treated in
Brown (1981) and will only be mentioned here. Here is a partial list of the most common
existential verbs, their glosses and the nouns they commonly occur with.
 53

----------------------------------------------------------------

| Verb | Gloss | Noun Subject |

--------------------------------------------------------------

| av | sit | women, taro |

| | | |

| lohv | stand | men, house, mountain |

| | | |

| dihelv | inanimate | road, book |

| | exists | |

| | | |

| nalohv | exist.piled | firewood, food in a trade store |

| | | |

| diav | exist | stones, cut trees in a garden |

| | scattered | |

| | | |

| angavilv | bundle with | bundle of food from garden |

| | shoulder | |

| | strap exists | |

| | | |

| tungulv | egg-shape | egg, stone |

| | exists | |

| | | |
 54

| putilv | thing with | lemon, ball |

| | stem exists | |

| | | |

| liv | recline | human, water, kero, snake |

| | | |

| moanvilv | pliable | netbag, clothes |

| | thing exists | |

| | | |

| vilóv | exist in | trees in bush |

| | numbers | |

| | | |

| endv | hang | fruit, roof of house |

--------------------------------------------------------------

Here is an example,

(163) mani ten-ba moanvil-v

what that-TOP exist.soft-PRS
ten-ba ka-na bayu moanvil-v
that-TOP 1st-GEN shirt exist.soft-PRS
‘What is that lying crumpled there? That is my shirt lying crumpled.’

4.1.5 Location

Locative information can be prefixed to the verb. It is not a true locative prefix, but the
cataphoric adverbs men 'here' and ten 'there' normally occur just before the verb, and they can
be cliticized (phonologically attached) to the verb if it has the right phonological shape.
'Cataphoric adverb' means that the location being established is new information.

(164) Luk-va te-pró

name-TOP there-come-PRS
‘There's Luke coming now.’
 55

Example (164) contrasts with the following (165), in which the adverb is marked for topic,
making it a demonstrative pronoun modifying the head.

(165) Luk ten-ba pró-v

name that-TOP come-PRS
‘That's Luke coming there.’

4.2 Post-Core Positions

There are seven positions after the core numbered from left to right with the following
information: 1. non-singular Direct Object 2. number of Benefactor or Recipient 3. Manner of
action 4. tense-mode-aspect 5. irrealis mood 6. question and emphasis 7. topic

4.2.1 Non-singular Direct Object

Dual or plural direct object can be marked by suppletive stem but more commonly it is
marked by suffix -wol in the position immediately right of the verb stem. When this position
becomes occupied some verbs require that information that would be otherwise marked in the
following six positions instead be attached to a helping verb, usually ve 'to do'.

(166) hi-mba deuv-pa sambla-va dembro-wol ve-na

3rd-TOP house-TOP two-TOP put-NS.O do-PST
‘He built two houses.’

For many verbs there is no distinction between dual and plural DO. However, for the stem ve 'to
do', there is a full paradigm for singular, dual, and plural S and DO. (Each stem in the following
chart has present tense -v suffixed):

Paradigm of Verb vev 'to do'

--------------------------------------------------------

| |#S | |#DO 1 2 3+ |

| |--------------------------------------------------

| | | | |

| 1| ve-v | ve-wol-v | ve-wol ve-v |

| | | | |

| 2| e-ve-v | e-ve-wol-v | (e-)ve-wol e-ve-v |

 56

| | | | |

| 3+ | vi-v | ve-wul-v | ve-wol vi-v |

| | | | |

--------------------------------------------------------

So, for example, ve-wul-v means 'more than two people are doing something to two things or
people.' In this form, the common Waris strategy of stem vowel raising has been applied to the
vowel of -wol 'non-singular direct object' to mark plural subject, and that is the origin of the u
in vewulv. In the verbs marked for plural DO a helping verb (which happens to be the identical
stem) has been employed to carry some of the affixation.

The suffix -wol can also signal intensity of, or variety of location of, an action, and can
therefore be affixed to verbs of motion as well:

(167) ka-va hélve-na / hélvo-wol

1st-TOP hit-PST / hit-NS.O
ve-na
do-PST
‘I hit (it) once / many times.’

(168) ka-va ga-wol ve-v

1st-TOP go-NS.O do-PRS
‘I go (on that road) a lot.’

(169) indhana amb-wol ve-vna

people come.PL-NS.O do-CON
‘People were coming from various places.’

In the following example the stem is prefixed to agree in number with the dual subject. It is also
suffixed for non-singular direct object, which indicates that the death of the father and the death
of the mother were different events.

The subject NP is marked with GOAL indicating subject lack of control over the action of the
verb (Absolutive case). The verb is also suffixed for singular benefactive, in agreement with the
singular speaker.

(170) araraval-va ka-na-m-ba doara

fa.mo.-TOP 1st-GEN-GL-TOP
e-ndaha-wol-mana-na
previously DL.S-die-NS.O-BEN.SG-PST
‘My father and mother died already (at different times).’
 57

In order to further exemplify points from sections 4.1 and 4.2, here is a paradigm of the verb
lóv 'to shoot (with an arrow)'.

It illustrates the following:

***the formation of dual subject by e- prefix

***the formation of plural subject by stem vowel raising

***the formation of plural object by suppletive stem, and by -wol suffix

***the use of a helping verb to carry tense and number information when the stem
becomes long.

Paradigm of Verb lóv 'to shoot'

-----------------------------------------------

| |#S | |#DO 1 2 3+ |

| |-----------------------------------------

| | | | |

|1 | ló-v | ló-wol-v | welha-v |

| | | | |

| 2 | e-ló-v | ló-wol e-ve-v | e-wel-v |

| | | | |

| 3+ | lu-v | ló-wol vi-v | weilhi-v |

| | | | |

-----------------------------------------------

Here is the morphemic breakdown of two examples from the paradigm:

(171) ló-wol e-ve-v

shoot-NS.O DL-do-PRS
‘Two men shoot two (pigs)’
 58

(172) weilhi-v
shoot.PL.S.PL.O-PRS
‘Several men shoot several (pigs).’

Finally in this section I will give an example to show how the meaning of e- 'dual subject' can
be expanded to apply to two groups of people:

(173) kembelnind hi-mba hev-m e-wal-wol

village.people 3rd-TOP 3rd.EMP-GL
e-ve-vna
DL-shoot-NS.O DL-do-CON
‘People (from two clans) were shooting arrows at one another.’

4.2.2 Benefactive or Recipient Number

The next position to the right of the stem is that of the Benenefactive and Recipient number
markers, which indicate either singular or non-singular. The Benefactive and Recipient case
roles are discussed thoroughly in the paper Brown (1985) and just examples will be given here.

---------------------------------------

| | | |

| | Benefactee | Recipient |

| | | |

---------------------------------------

| | | |

| SG | -mana | -ho |

| | | |

| NON-SG | -in | -hun |

| | | |

---------------------------------------

Example (174) shows the verb stem marked for singular recipient, referring to the speaker:
 59

(174) pensil ka-m vra-ho-o

pencil 1st-GL get-REC-IMP
‘Get the pencil for me = give me the pencil!’

(175) mie-m ka-va sha-mna-na

pig-GL 1st-TOP sleep.PL-BEN-PST
‘We slept (waiting) for a pig (so we could shoot it).’

(176) Yosev he-m-ba inne-mba dembra-luh-un-na

name 3rd-GL-TOP food-TOP put-exist-REC-PST
‘Yosev was habitually giving food to them.’

(177) ka-namini-oa ge-in-u

1st-ABL-RT go-BEN.PL-IMP
‘As for us, go away from us = please go away!’

(178) ka-m-ba os ishu-mini-vna

1st-GL-TOP thus say.-BEN.SG-CON
‘They were talking like this to me.’

In example (178) the influence of the high stem vowel (indicating plural subject) is felt to the
right and raises the vowels of -mana (vowel harmony). In other words, our way of spelling
gives two clues to the reader that the verb is plural in subject number: u in the stem and i in the
benefactive suffix.

4.2.2.1 Subject Number Neutralization with the Benefactive

Verb subject number is neutralized when the plural benefactee is suffixed to some verb stems.

1. ishó- 'one person speaks' + -in 'NS.BEN' + -v 'PRS' > ishunv

2. ishu- 'people speak' + ' ' ' + ' ' > ishunv

Thus the form ishunv can mean either 'one person speaks to people' or 'people speak to people'.

4.2.3 Manner

The third position for suffixes is different from all the other positions in that it might be
analyzed to contain a wider or narrower selection of suffixes according to how many of the
potential 'suffixes' might occur as independent verb stems. Seiler (1984a) cites a number of
candidates for this slot that in Waris seem to be part of a particular verb stem rather than
occuring with a variety of stems. I first deal with the clear examples, then with the less clear
ones. Here is a list followed by examples:
 60

-laha 'with movement away'

-mana 'with movement from place to place'

-pia 'completely'

-ula 'intensely, or with good feeling'

-ungu 'located in various places'

-ó 'daily action, singular subject'

-uv 'daily action, plural subject'

simbaiha 'futile action'

(179) hi-mba andra ve-pia-na

3rd-TOP work do-comp-PST
‘He/she finished his/her work.’

(180) ka-na kinwonda kethe-laha-o

1st-GEN heavy.bilum hang-move-IMP
‘Relieve me of my burden!’

(181) ka-va a-nga-ula-vna bas-ra

1st-TOP PL-go-happy-CON bus-LOC
‘We were going along happily by bus.’

(182) hi-mba loh-mana ve-vna

3rd-TOP exist-move.around do-CON
‘He/she is taking a walk.’

(183) aiwohna loh-ungu ve-v

sanguma exist-around do-PRES
(Beware) there are sanguma about!’

(184) hari kamis ve-ó-vna

day Wednesday do-daily-CON
It was Wednesday.’

(185) hi-mba aev-simbaiha-i

OP sit.PL-futile-RP
They sat (with us) for just a little time (not really enough).’
 61

In example (182) -mana has triggered the use of a helping verb 'do' to carry the tense-aspect
information. There are two possible contributing reasons for this. One is simply length. The
other is to avoid confusion between -mana 'movement around' and -mana 'singular benefactee';
the latter is commonly followed by tense-aspect suffixes without the use of a helping verb while
the former generally triggers the use of a helping verb.

The suffix -simbaiha 'futile action' (185) lends itself generally to the formation of derived
adverbials as in the following example (186)

(186) hi-mba i-pra-simbaiha-nam

TOP carry.PL-come.SG-futile-MAN
pró-i
come.SG-RP
She brought (vegetables) futiley (you didn't buy them).’

The manner suffixes listed above may actually represent more than one position class, since the
following example (187) shows how at least two of them can co-occur.

(187) hi-mba andra ve-uv-mana ve-vna

3rd-TOP work do-daily.SG.S-place.to.place do-CON
They were working day after day.’

The following two morphemes fit into the third (manner) position. But they also act like verb
stems in that they can function alone (not in a compound stem) in the main slot of a verb word.
They are unlike other verb stems in that they cannot occur unprefixed for accompaniment or
things carried, example (190).

-ka 'one person goes and comes'

-hélvo 'more than one person goes and comes'

(188) ka-va kin i-ka-na

1st-TOP burden PL.held-go.and.come-PST
I fetched items of cargo.’

(189) kav-na séhél-m ka-va ó-mna-ka-i

1st.EMP-GEN friend 1st-TOP say-BEN-go.and.come-RP
I just went and got my friend.’

(190) *ka-va ka-na

1st-TOP go.and.come-PST
*I went and came.’
 62

(191) nongla-hélvo-o
see-go.and.come.PL-IMP
You people go and look (at it)!’

To end this section I list a few more morphemes which occur in this same position. They occur
with only a few verbs and the combinations may have taken on a specialized meaning.

-vha 'action towards an end (inceptive)'

-tha 'action towards an end (effect incomplete)'

-nda 'causative'

(192) ka-va a-v / a-vha-v

1st-TOP sit-PRS / sit-action-PRS
I am sitting / I assume a sitting position = sit down.’

(193) hi-mba pse-v / psi-vha-v

OP dig-PRS / dig-action-PRS
He digs (food) / he first digs (a little hole to see if’ the food is mature).'

(194) ka-va ti he-v / ti-m

1st-TOP tree chop.down-PRS / tree-GL
he-tha-v
chop.down-action-PRS
I chop down a tree / I chop on a tree.’

(195) suwe-mba vutha-i / hi-mba vutha-nda-i

fire-TOP die-RP / 3rd-TOP die-causative-RP
suwe-mba
fire-TOP
‘The fire has died / he put the fire out.’

4.2.4 Tense-Mode-Aspect
This position takes the suffixes of the Realis mood. They are summarized in the following
charts and examples are given. The Irrealis mode is signaled by suffixes in the next position,
discussed in (4.2.5). It may appear strange to assign irrealis and realis to different orders of
affixes, but it is done because irrealis markers, as well as occurring alone, may also co-occur
with certain realis markers, in which case irrealis follows (201, 202)).

Indicative Mode Verb Endings

 63

-------------------------------------------------------

| | | |

| Tense | Non-continuous Aspect | Continuous Aspect |

-------------------------------------------------------

| Past | -na | |

|---------------------------------| |

| Recent | -i | -vna |

|---------------------------------| |

| Present | -v | |

|-----------------------------------------------------|

| Future | -vav |

-------------------------------------------------------

Imperative Mode:

-o 'imperative'

-we 'polite imperative'

-vm 'polite imperative'

Optative Mode:

-vai 'optative'

For the use of the indicative tenses in sentences and discourse see sections 6.2 and 6.9.
Following are examples of imperative and optative.

(196) ka ga-vai
1st go-OPT
‘I want to go!’

(197) ga-o / ga-o-ta / ga-we

go-IMP / go-IMP-EMP / go-IMP
 64

‘Go! / go!! / go! (polite).’

(The form -we is perhaps the regular imperative marker -o followed by the -e used in
quotations and lists (3.2.3).)

Imperatives with the resumptive topic can take a special form of the imperative marker, made
from the present tense marker -v followed by Goal -m. The effect is politeness:

(198) ye-oa ga-v-m

3rd-RT go-PRS-GL
‘As for you, go ahead and leave.’

4.2.5 Irrealis Mood

Suffixes in this position signal a mood of the verb which is not 'real' indicating states that are
not part of history. They may mean 'anticipated and more or less desired', or 'unanticipated and
undesired' or 'did not happen'.' Therefore in sentences, there is overlap of meaning with the
future tense. The syntactic difference is that an irrealis mood verb can occur in the protasis of a
conditional sentence but a future tense realis mood verb cannot. There are two markers of
irrealis:

-ra 'anticipated and/or desired'

-ta 'unanticipated, undesired, or did not happen'

(199) ata e-ne-ta doara weha-ta

soon DL-eat-IRR previously cook-IRR
(You and I) would eat now if (you) had cooked food [but you didn't so we
won't].

(200) kapol-va ata pró-ra-va ata ka-va ga-vav

plane-TOP soon come-IRR-TOP soon 1st-TOP go-FUT
‘If/when the plane comes today, I will go [it may well come].’

In the protasis of this conditional sentence (200) the irrealis marker is followed by the topic
marker, and in the apodosis the future tense is used. See section 6.2.1 for a discussion of
conditional sentences.

The irrealis marker can co-occur with two of the indicative markers, -v 'present tense' and -vav
'future tense', which add the semantic component of continuous or repetetive action:

(201) hi-mba ga-v-ra-va si hi-mba pho-vav

3rd-TOP go-PRS-IRR-TOP future 3rd-TOP arrive-FUT
‘If he should keep going he will arrive (at his destination).’
 65

(202) yi-mba ka-m-ba und-va vi-vav-ta..

3rd-TOP 1st-GL-TOP love-TOP do.PL-FUT-IRR
‘If in the future you would keep on loving me (but perhaps you won't)..’

4.2.6 Question and Emphasis

This position is filled by the two mutually-exclusive question and reply or emphasis clitics:

-ma 'question'

-ta 'yes!'

(203) doa hi-mba ga-i-ma / doa ga-i-ta

complete 3rd-TOP go-RP-Q / complete go-RP-EMP
‘Has he gone? / yes, he has gone!’

In certain verb phrases meaning of -ta could mean either 'irrealis' (section 4.2.5) or 'yes'. Such
ambiguity would generally be resolved by context.

4.2.7 Topic

The final post-core position is that of the topic marker. I mention it here for convenience in
helping the reader understand the morphological form of verbs. Logically, it marks pragmatic
function of a whole utterance, of which the verb is merely the end. Topic marking on the verb
indicates that the action of the clause is given, and the stepping stone for the next event on the
event line. (Thus it is mutually-exclusive with the two fillers of position six, Question and
Emphasis). In the following example _doa_, a time adverbial with basic meaning 'completed'
marks the next event as being prominent on the event line. For a discussion of the function of
topic marking in discourse see section 6.5.

(204) ka-va pró-na pró-na-mba doa ka-va

1st-TOP come-PST come-PST-TOP complete 1st-TOP
avhó
sit.down
‘I came, having come OK I sat down.’

4.3 Verb Serialization

Verb serialization in Waris falls naturally into two categories due to the different semantic
function given them. (Both are analyzed as constituting one clause.) The first is the use of two
affixed verbs in series and the second is the formation of compound verb stems which become
 66

part of one phonological word. The first is used commonly to express intention, especially
frustrated intention or when something unanticipated happened. The second is extensively used
for the sake of precision of meaning. I describe the first (4.3.1) and the second (4.3.2).

4.3.1 Serial Finite Verbs

This occurs only with the use of the verb 'to say' in the first slot and has the meaning
'intention'. Here are a couple of examples:

(205) om ka-va ó-na ga-na

yesterday 1st-TOP say-PST go-PST
heva os owai po hasvo-na
but no rain fell-PST
‘Yesterday I intended to go but it rained and I didn't.’

(206) ka-va ó-na nongle-na pai mani hev-pa

1st-TOP say-PST see-PST my! what 3rd.EMP-TOP
‘I tried/started to look at (it), "my! what is it?" (I said).’

The following example (207) is with an inanimate subject and so the idea of intention does not
exactly fit, maybe 'inceptive' is better:

(207) po-mba ó-i mindiri te-ve-i

water-TOP say-RP boil there-do-RP
‘The water there is starting to boil.’

In the following example, the verb óv 'speak' occurs in a serial-like construction, but with a
function bordering on that of a conjunction since it clearly joins two clauses together into a
sentence:

(208) ó-i os yes-va ten-ba sahokla ve-v-pa

speak.RP thus sago-TOP that-TOP bad do-PRS-TOP
ó-i yes-va no-mba besowonam daha-i
speak-RP sago-TOP that-TOP well set-RP
‘It's like that sago is still bad, but that other sago has jelled well.’

4.3.2 Compound Verb Stems

The formation of compound verb stems to fill the core postion of the VP is an important
aspect of Waris. It is a mark of the speech of mature speakers that they use compound verb
stems rather than simple ones. (like children and expatriate Bible translators). The effect of this
is greater precision and vividness, since for each action they describe they add clarifying or
 67

modifying information, or aspectual information. Here I will give a semantic classification of

the compound verb stems into those that

1. give aspectual information

2. have additive meaning

3. have idiomatic meaning

4. involve the action of more than one actor.

Phonologically, all Waris compound verb stems are one word.

4.3.2.1 Compound Verb Stems with Aspectual Focus

These stems usually are built on helping verb stems, the verbs ve 'to do', loh 'exist, stand', and
a 'sit' being frequently used.

(209) ka-va di li-vna / li-loh-v misin-da

1st-TOP there recline-CON / recline-exist-PRS mission-LOC
‘I was sleeping there/habitually sleep there at the mission.’

(210) ka-va di li-loh loh-v misin-da

1st-TOP there sleep-exist exist-PRS mission-LOC
‘I am habitually sleeping there at the mission for the time being.’

(211) ka-va ve-siha ve-loh-v

1st-TOP do-enter do-exist-PRS
‘I habitually enter (it) from time to time.’

(212) indhana-mba dom ve-hél ve-loh-v

people-TOP to.there do-go.and.come.PL do-exist-PRS
taeks-ram
tax-ALL
‘People go there (yearly to pay) their tax.’

(213) indhana-mba nungulu-l-a-vna

people-TOP look.PL-?-sit-PRS
‘People sat and watched (it).’
 68

4.3.2.2 Compound Verb Stems with Additive Meaning

With these compound stems the meaning of the individual stems adds to the meaning; they are
not idiomatic and do not involve a change of actor:

(214) tuawa pho-laha-nga-i

bird get.up-movement-go-RP
‘A bird flew away.’

(215) okómba lovha-siha ve-v

sun shine-enter do-PRS
‘The sun shines into (the house).’

(216) po-ra ka-va peiha-vonga-na

water-LOC 1st-TOP descend-ascend-PST
‘I crossed the stream.’

(217) naka hulvo-holótho-loh-v

sago.frond break-cover.up-exist-PRS
‘A sago frond broke off (and fell down) and covered(something) and exists.’

(218) buku-va ka-va deuv-ra vélaha-pró-i

book-TOP 1st-TOP house-LOC put-come-RP
‘I left the book behind at the house.’

(219) ye-m ka-va dasvho-laha-nga-v

2nd-GL 1st-TOP put-movement-go-PRS
‘I put you and go away = I leave you and go away.’

Serial verbs of this type are very widely distributed in the language and they allow the speaker
to be very precise in describing their daily activities to other people. But they seem to be fixed
in number, not productive. For example the following sentence (220) cannot be simplified by
forming a serial stem as in (221):

(220) titung pilpe-na mona-m kélvha-na

rotten.tree fall-PST trail-GL block-PST
‘A rotten tree fell and blocked the trail.’

(221) *titung mona-m kélvhó-pilpe-na

rotten.tree trail-GL block-fall-PST
‘*A rotten tree fell and blocked the trail.’
 69

4.3.2.3 Compound Verb Stems with Idiomatic Meaning

The meaning of these stems is not the sum of the meanings of the constituent stems:

(222) hi-mba indharini-m hélva-komandha-v

3rd-TOP person-GL hit-hide-PRS
‘He kills a person.’

(223) moa-mba ka-va lévra-pi-vongo-v

talk-TOP 1st-TOP lift.up-up-ascend-PRS
‘I understand the talk.’

This example (223) is not merely of a figurative meaning applied to a verb which has another
literal meaning, rather it seems to have no other meaning than this.

4.3.2.4 Compound Verb Stems with Change of Actor

In these stems there is a change of actor between the two stems in the compound, and the
resulting semantics is one of causation:

(224) ka-va ye-m-ba nungulu-laha-v

1st-TOP 2nd-GL-TOP see.PL-movement-PRS
‘I look at you and you go away = I send you away.’

(225) nénv vélaiha-vongó ve-vna wemb sava

earthquake remove-ascend do-CON yam taro
The earthquake caused the yam and taro to come out of the ground.’

In this clause the earthquake is the subject and yam and taro the object. The meaning of vélaiha
from other contexts is 'remove'. From other contexts vongo means 'ascend' but not 'cause to
ascend', so there really is a sense in which there is a change of 'actor' within the clause.

Another interesting feature of compound stems of this class is the way in which number
information properly belonging to the first stem of a series is moved to the right and becomes
attached to the second stem. This becomes apparent due to the occurance of a suppletive stem
which in the example (227) is _suv_ 'people enter'.

(226) ka-va he-m bui-ra wolaiha-siha-v

1st-TOP 3rd-GL jail-LOC release-enter-PRS
I put him inside the jail = cause him to be in the jail.

(227) ka-va he-m wolaiha-suv ve-v

1st-TOP 3rd-GL release-enter.PL.S do-PRS
‘*I put them inside.’
 70

We put him inside.

In order to express the meaning in the starred gloss, Waris uses another suppletive stem, wola-,
which is obviously derived from the verb wolaihav 'one person releases another'. However, the
stem wola- is not distributed elsewhere in the data and so it is not possible to check its exact
meaning, whether it is singular or plural in isolation.

(228) ka-va he-m wola-siha ve-v

1st-TOP 3rd-GL release-enter do-PRS
‘I put them inside.’

I take this displacement of number information from one stem to another to be evidence of what
Bruce (1979) and James (1982) call lexicalization of serial verbs. In other words, serial verb
stems have become tightly bound so as not to be transparently analyzable semantically.

5 Clauses

Clauses are treated here from the standpoint of the cases that occur in them. First, I give a
classification of clause types based on Cook's matrix variety of case grammar (5.1). This was
done in detail in an earlier paper on cases and verb classification (Brown 1989), and the
treatment here is not as detailed. In section 5.2.1 describe the clause types with examples. In 5.3
I discuss distribution of word classes in clause slots. In 5.4 I briefly mention the subject of
defining grammatical relations in Waris.

5.1 Clause Classes

Cook's variety of case grammar (1971) is useful in classifying Waris clauses. His basic idea is
that semantic case roles do not occur in random sets, but work in opposition to one another as
described by a matrix. On the left of the following matrix is the trinary opposition between
predications of State, Process, and Action types. This is taken directly from Cook. Across the
top is a binary opposition between predications in which Existence or Location is a salient
feature, and predications in which it is not. This opposition is not from Cook but is rather the
result of my reflection on Waris. Specifically it is an application of the notion of Fillmore
(1977) which he called the cognitive scene. According to Fillmore, the speaker's choice of a
particular verb allows him to bring into _perspective_ a certain set of the items of that scene by
making them nuclear elements, S, O, IO. Furthermore, certain saliency features favor the
inclusion of an item from the scene as a nuclear element:

humanness > change of state or location > definiteness/totality

 71

This idea of Fillmore led to the realization that in Waris there seems to be a clear division
between verbs with existential focus and those not. By 'existential focus' I mean that the
cognitive scene activated in the mind of a Waris speaker by one of these verbs involves either
location (including _change_ of same), or other mode of existence of items from the scene.
'Mode of existence' is exemplified by the Waris verbs which predicate the posture, shape or
perceived mode of existence of the items of the Waris environment. Refer to section 4.1.4 for
a discussion of these verbs.

This section is a classification of clause types and not just verb types because some verbs fill the
predicate of more than one clause. For example, the Existential verbs occur in both clause type
1 (5.2.1) and 2 (5.2.2). The clause classification presented in this section is a working one with
areas of the analysis unclear. For example, I call Benefactee a core argument (IO) in Waris, and
in line with this I separate clause types 15 and 16 from clause type 17 on the basis of the
presence of this argument on the verb filling the predicate of clause type 17. However in the
case of clause type 18 I have analyzed Benefactee as an optional case and its occurrance in
some cases of clause type 18 is not used to put them in an additional clause type.

The numbers in the cells agree with the numbering of the sections following.

Waris Clause Classification

-------------------------------------------------------------

| | Existential or | Non-Location Focus |

| | Location Focus | |

| |-----------------------|-------------------------|

| | | |

| State | 1 Existential Cl | 2 Stative-Equative Cl |

| | | |

| | | |

-------------------------------------------------------------

| | | |

| | 3 Motion Cl | 6 Sensory-Inner State Cl|

 72

| | | |

| | | |

| | 4 Orientation- | 7 Sensory-Force Cl |

| Process | Achieved Cl | |

| | | |

| | 5 Force-Motion Cl | 8 Pseudo-Passive Cl |

| | | |

| | | 9 Change of State Cl |

| | | |

| | | 10 Telic Cl |

| | | |

| | | 11 Burn-Light Cl |

-------------------------------------------------------------

| | | |

| | 12 Holding-Moving Cl | 15 Goal Cl |

| | | |

| | 13 Ditrans Giving Cl | 16 Patient Cl |

| Action | | |

| | 14 Accompaniment Cl | 17 Goal or Patient |

| | | Benefactive Cl |

| | | |

| | | 18 Sound-Speech Cl |
 73

| | | |

| | | 19 Speech Benefactive Cl|

-------------------------------------------------------------

5.2 Clause Types with Examples

There are 19 clause types based on the cases associated with the verb and fitting into the
above chart of oppositions. Following (5.2.1 - 5.2.19) are examples, preceded by information on
the types of transformation each clause can undergo.

But first here is a list of the semantic case roles used:

1. ABLative is the case of movement away from a source.

2. ABSolutive is the case indicating lack of control by the NP so marked over the action of the
verb.

3. Agent is the case of an animate instigator of a process or condition.

4. ALLative is the case of movement toward a goal.

5. BENefactive is the case of the benefactee of an action or state.

6. COMitative is the case of an animate entity that accompanies another of equal rank.

7. EXPeriencer is the case of the animate experiencer of a sense or emotion.

8. ForCe is the case of an inanimate non-volitional cause of an action.

9. GoaL is the case of the point of termination of an action, which does not experience any
change of state.

10. INSTrument is the case of means.

11. LOCative is the case of stationary location.

12. Object or Agent/Object is the animate or inanimate experiencer of a state.

13. Patient is the case of an entity that undergoes a change of state or location.

14. REAson is the case of a prior cause.

 74

15. RECipient is the case of a human receiver.

16. TELic is the case of the end of a process or purpose of a state.

5.2.1 Existential Clause

Cases [+O +/-LOC +/-BEN]

Existential clauses predicate a state of being, with focus on the _mode_ of existance, such as
sitting, standing, reclining, inanimate, round, etc. The grammatical subject is in the Object or
Agent/Object semantic case role; when this is animate the clause can undergo transformation to
take imperative. Optional case roles associated are Location and Benefactive. In contrast to
Stative-Equative clauses (5.2.2) Existential clauses must have a verb.

(229) po li-v mona-ra

water recline-PRS road-LOC
‘There is water on the road.’

(230) ye-oa deuv-ra a-v-m

3rd-RT house-LOC sit-PRS-GL
‘As for you, remain in the house.’

(231) mie-m ka-va sha-mana-vna

pig-GL 1st-TOP recline.PL-BEN.S-CON
‘We were sleeping (waiting) for a pig (to shoot it).’

(232) ti-mba manaram naloh-v

wood-TOP what.purpose exist.piled-PRS
‘For what purpose is the wood stacked up?’

Verbs occurring in this clause type are the existential verbs discussed in section 4.1.4.1. The
same verbs also fill Stative-Equative clauses in the next section, 5.2.2.

5.2.2 Stative-Equative Clause

Cases [+P +/-LOC +/-BEN]

A Stative-Equative clause identifies or gives information. The grammatical subject is Patient

(un-case marked) and the optional case roles are Locative and Benefactive. The verb is optional
in this kind of clause, which is one difference between this clause and Existential Clause
(5.2.1). Another difference is that this clause may not form an imperative, unlike Existential
clauses (5.2.1).
 75

(233) ka-va etel loh-mana-v

1st-TOP older.brother exist-BEN.SG-PRS
‘I am older brother to him.’

(234) he-na deuv-pa ten-ba (loh-v)

3rd-GEN house-TOP that-TOP (exist-PRS)
‘That is his house over there.’

(235) lemo ten-ba ti-ra (end-v)

lemon that-TOP tree-LOC (hang-PRS)
‘There is a lemon (hanging) in that tree.’

Stative-Equative clauses have the topic slot first and the comment slot next, (followed by the
verb). Thus there is a contrast in meaning in the following two examples:

(236) deuv-pa ka-na-mba sambla sahoklal

house-TOP 1st-GEN-TOP two bad
‘My two houses are bad (neither is good).’

(237) deuv-pa ka-na-mba sahoklal sambla

house-TOP 1st-GEN-TOP bad two
‘My bad houses are two in number (I don't just have_one_).’

The word 'topic' is used in two different senses above. The topic slot of the clause is what is
being talked about; Waris does not mark this directly. The topic suffix on certain NPs means
something like this: 'this item has been talked about or pointed out previously in this
conversation, and is now being reintroduced as what is being talked about.' I believe this is the
reason for the topic marking on 'my house' in example (237).

5.2.3 Motion Clause

Cases [+P +/-GL +/-COM +/-INST +/-ALL]

Motion clauses have their subject un-case marked. Optional cases are Goal, Comitative,
(Vehicle)Instrument, and Allative. This clause may form an imperative when the Patient
(grammatical subject) is animate.

(238) vuvi pró-v

wind come-PRS
‘The wind is blowing.’

(239) ye-i ka-va e-nga-v

2nd-COM 1st-TOP DL-go-PRS
‘I am going with you.’
 76

(240) po-m ga-v

water-GL go-PRS
‘He/she is going for water.’

(241) kav-na móngla-na ka-va pró-i

1st.EMP-GEN leg-GEN 1st-TOP come-RP
‘I came by foot (not car).’

The following clause (242) undergoes transformation by taking an animate Agent to become a
Holding- Moving clause. Compare example (272) in section 5.2.12.

(242) atomb lilve-vna

drops drip-CON
‘Rain or dew was dripping (from the trees).’

5.2.4 Orientation-Achieved Clause

Cases [+A/O +/-LOC +/-BEN]

This kind of clause describes motion taken in order to achieve a certain orientation. Verb
consists of an Existential verb stem (5.2.1) plus affixation meaning 'action towards an end'.
The subject, viewed as an Agent/Object, is un-case marked. Imperative can be formed. Optional
case roles are Location and Benefactive.

(243) hi-mba nin li-tha-v

3rd-TOP sleep recline-action-PRS
‘He/she is assuming a reclining position = going to bed.’

(244) yi-mba ka-m-ba loh-vha-mana-v-ma

2nd-TOP 1st-GL-TOP stand-action-BEN.SG-PRS-Q
‘Are you getting up for me?’

5.2.5 Force-Motion Clause

Cases [+P +Cause]

This clause predicates the motion of an inanimate object, un-case marked, acted on by an
inanimate force, marked with Locative (Cause).

(245) klal kinvehe-v móvól-la

branch weigh.down-PRS fruit-LOC
‘A tree branch is weighed down by fruit.’
 77

(246) tuv-pa vuvi-ra pétha-i

door-TOP wind-LOC close-RP
‘The door closed because of the wind.’

5.2.6 Sensory-Inner State Clause

Cases [+EXP +GL/BEN]

This clause predicates a sense or emotion. The subject as Experiencer is un-case marked, and
the object is marked as Goal or Benefactive, depending on the verb. Some of the predicates in
this class are volitional and can form an imperative: see, hear, think.

(247) tuawa-m-ba ka-va doa nongle-na

bird-GL-TOP 1st-TOP complete see-PST
‘I have seen the bird.’

(248) hi-mba pe-m vi-mini-vna

3rd-TOP fear-GL do.PL-BEN.SG-CON
ósah-mana-lm-ba
ask-BEN.SG-TEL-TOP
‘They were afraid to ask him.’

5.2.7 Sensory-Force Clause

Cases [+ABS +BEN +FC]

This clause predicates an undesireable bodily state such as pain or tiredness. Subject is marked
as Absolutive case with Goal, and cross-referenced on the verb as Benefactee. The cause, called
Force, is marked with Locative, or unmarked, idiosyncratically.

(249) ku ka-m-ba vé-ne-v

head 1st-GL-TOP do-BEN-PRS
‘My head aches.’

(250) ungevli-rini-m kélatha-mna-vna

women-DER-GL birth.pain-BEN-CON
‘A woman is having birth pains.’

(251) andra-ra he-m-ba nihóvóvra v-in-v

work-LOC 3rd-GL-TOP tiredness do-BEN.PL-PRS
‘They feel tired after having worked.’
 78

In this example (251) there are two causes or Forces, which some writers about case would call
'inner force' and 'outer force', tiredness and work, respectively. The former un-case marked and
the latter marked with Locative.

5.2.8 Pseudo-Passive Clause

Cases [+ABS +/-FC +/-INST +/-LOC +/-BEN]

This clause predicates a process involving a human being, who is experiencing something over
which he/she has no control. The subject is marked with the Goal case suffix and called
Absolutive case in our analysis. Optional co-occurring cases roles are Force, marked with
Locative and Instrument, marked with Genitive. Locative and Benefactive can also occur in this
clause. Pseudo-passive clauses cannot form imperatives, but they can take irrealis to form
optative as in example (253).

(252) obat-ra he-m-ba kanandha-na

medicine-LOC 3rd-GL-TOP get.well-PST
‘He/she got well because of medicine.’

(253) ye-oa kanandha-ra

2nd-RT get.well-IRR
‘As for you, may you get well.’

(254) pul-na weihala-na he-m-ba daha-na

betel-GEN eat-GEN 3rd-GL-TOP die-PST
‘He/she died of sorcery.' (Someone chewed betel nut and sorcerized
him/her.)’

(255) arapev-m-ba ye-na-m-ba daha-in-na-ma

your.father-GL-TOP 1st-GEN-GL-TOP die-BEN.PL-PST-Q
‘Has your father died for you? = Are you all fatherless?’

(256) ka-m takole-i mona-ra

1st-GL trip.and.fall-RP road-LOC
‘I tripped and fell on the road = something tripped me and I fell on the road.’

5.2.9 Change of State Clause

Cases [+P +/-FC]

This clause type predicates a process of change of state, for the worse, in an inanimate subject,
un-case marked and analyzed as Patient.A non-instigative cause may co-occur, marked with
Locative and analyzed as Force.
 79

(257) aomb-va doa óvóvó viló-i

greens-TOP complete wilt exist.PL-RP
‘The greens have already wilted.’

(258) ti brawol ve-v

tree rotten do-PRS
‘The tree is getting soft and rotten.’

(259) po-ra indkokla ve-v mona-mba

rain-LOC become.bad do-PRS road-TOP
‘Because of the rain, the road is being ruined.’

5.2.10 Telic Clause

Cases [+P +TEL]

The Telic clause describes a process of change in a subject, analyzed as Patient and unmarked
for case, with the end result of the change, marked with the Telic case marker. This clause
cannot form the imperative.

(260) popoli andava-lm nilha-vav

cocoon butterfly-TEL change.into-FUT
‘A cocoon will change into a butterfly.’

(261) Willie guru-elm loh-v

name teacher-Telic exist-PRS
‘Willie (became) a teacher and exists (like that).’

The former example (260) of Telic clause has no parallel in the other clause types, but the latter
example (261) is similar to the Stative-Equative type clause (262):

(262) Willie guru loh-v

name teacher exist-PRS
‘Wille is a teacher.’

By adding the Telic suffix to the goal the speaker emphasizes the process by which the subject
has become a teacher.

5.2.11 Burn-Light Clause

Cases [+P +/-LOC]

This clause predicates fire burning or light being produced.

 80

(263) suwe-mba ye-na-mba ta-v-ma

fire-TOP 2nd-GEN-TOP burn-PRS-Q
‘Is your fire burning (in your house)?’

(264) deuv-sowa ta-na

house-fire burn-PST
‘A house burned.’

(265) senter mindanam lovha-v

torch much light-PRS
‘A torch lights brightly.’

(266) wokala lavra-holótho-loh-v

torch light-put.up-exist-PRS
‘A bamboo torch is hanging up and burning.’

(267) okómba lovha-siha v-in-v

sun light-enter do-BEN.PL-PRS
‘The sun is shining in on them (people).’

Examples (265) and (266) above are instances of Burn-Light Clause that undergo transform by
taking an agent to become Patient clauses, section 5.2.16, examples (287) and (289).

5.2.12 Holding-Moving Clause

Cases [+A +P/GL +/-LOC +/-BEN]

This clause describes an animate Agent, unmarked for case, in the act of holding or carrying
something. If the Patient is inanimate it is unmarked for case, if animate it is marked as Goal.
Location can also occur as well as Benefactive, and the imperative can be formed.

(268) wan yi-mba wul-a-v-ma

salt 3rd-TOP carried.SG-sit-PRS-Q
‘Do you have any salt?’

(269) hi-mba surat goa-in-v

3rd-TOP burden carry-BEN.PL-PRS
‘He/she is carrying a letter for them.’

(270) tuv ka-m vélaiha-mn-o

door 1st-GL open-BEN-IMP
‘Open the door for me!’

(271) ye-m ka-va wolaihasiha-vav bui-ra

2nd-GL 1st-TOP put.inside-FUT jail-LOC
 81

‘I will put you in jail.’

The following example (272) is related by transformation to example (242) in section 5.2.3. The
former has no Agent but this does.

(272) winde lilve-i

dog drip-RP
‘A dog shook off water.’

5.2.13 Ditransitive Giving Clause

Cases [+A +P +REC +/-REA]

The Giving clause predicates the action of an Agent in getting and giving something to another
animate argument, marked with Goal and cross-referenced on the verb as Recipient. Imperative
can be formed and a noun-classifying verb prefix commonly occurs. The Patient is unmarked
and Reason can co-occur, marked with Allative.

(273) buku ka-m vra-ho-o

book 1st-GL get-REC-IMP
‘Get the book for me = give me the book!’

(274) nenas ka-m li-ra-ho-o

pineapple 1st-GL CLAS-get-REC-IMP
‘Get the pineapple for me = give me the pineapple.’

(In this example (274) the classifier _li-_ refers to elongated fruit like pandanus and pineapple.)

(275) ka-va ótól aral-m holvó-ra-ho-i

1st-TOP child father-GL CLAS-get-REC-RP
‘I gave the child to (its) father.’

(276) hi-mba sen ka-m dembra-ho-i inne-ram

3rd-TOP coin 1st-GL put.PL-REC-RP food-ALL
‘He gave me coins for food.’

5.2.14 Accompaniment Clause

Cases [+A/O +GL +/-LOC]

Accompaniment clauses predicate a state of orientation in which the grammatical subject, the
Agent/Object, accompanies another human, which is marked with the Goal -m. Locative is a
optional case role.As explained above, verbs filling such clauses include verbs of motion (5.2.3)
 82

and the existential-orientation verbs (5.2.1). The verb is prefixed marking either singular or non-
singular accompanied. Imperative can be formed.

(277) hi-mba hev-na séhél-m won-a-v

3rd-TOP 3rd.EMP-GEN friend-GL ACC.PL-sit-PRS
‘He is sitting with his friends = hosting them.’

(278) etel hev-na boasalel-m

o.brother 3rd.EMP-GEN
wai-pró-i
y.brother ACC.SG-come-RP
‘An elder brother has just brought his younger brother.’

5.2.15 Goal Clause

Cases [+A +GL +/-INST]

The Goal clause predicates the action of an Agent on a Goal, which is incompletely affected.
Certain verbs filling this kind of clause bear suffixation (-tha, section 4.2.3 end) to indicate the
incompleteness of the action, such as hev, example (279). The same verbs, unsuffixed, can
occur in the Patient clause (5.2.16), where the object is completely affected, example (283).

(279) ka-va ti-m he-tha-i

1st-TOP tree-GL chop.down-action-RP
‘I was chopping on a tree.’

(280) ungevli-va ósó-m lingi-vna

women-TOP garden-GL clean.bush-CON
‘The women were cleaning away undergrowth for a garden site.’

(281) hi-mba tuv-m hevra-v

3rd-TOP door-GL beat-PRS
‘He/she is beating on the door.’

(282) hi-mba wóng-m vénd-v

3rd-TOP drum-GL hit-PRS
‘He is beating on a drum (but it's not resounding well)’

5.2.16 Patient Clause

Cases [+A +P +/-INST]

Patient clauses predicate the action of an Agent on an object which is totally affected. If the
object is inanimate it is unmarked for case, if animate it is marked with Goal. Patient clauses
 83

may form the imperative. Some can undergo transformation by removal of verb suffixation to
form goal clauses (5.2.15). Another transformation is the addition of a Benefactee argument
and verb cross-referencing to become Goal/Patient Benefactive clauses (5.2.17).

(283) ka-va ti he-i

1st-TOP tree chop.down-RP
‘I just chopped down a tree.’

(284) hi-mba tata puis-v

3rd-TOP meat cut.up-PRS
‘He is cutting up the meat.’

(285) ka-va ye-m hélvakomandha-v

1st-TOP 2nd-GL kill-PRS
‘I kill you.’

(286) téh ka-va ha-wol ve-vav

firewood 1st-TOP split-NS.O do-FUT
‘I will split firewood.’

The following examples of Patient clause (287) and (289) are transforms of Burn-Light clauses,
section 5.2.11 examples (265) and (266), which have added an agent.

(287) ka-va senter lovha-v

1st-TOP torch light-PRS
‘I cause the torch to light.’

(288) ka-va doa suwe titoha-i

1st-TOP complete fire light-RP
‘I have already lit a fire.’

(289) ata ka-va wokala lavra-holótho-vav

will 1st-TOP torch light-put.up-FUT
‘I will hang up a (burning) torch.’

The following two examples (290) and (291) have no Agent. But it seems best to class them
here because the inanimate causes (stove and nettle) are not marked as Force using the Locative
suffix, as other causes are (5.2.5, 5.2.7, 5.2.8), but are unmarked like Agent.

(290) stov minak pél-ta-v

stove kero remove-burn-PRS
‘The stove burns up kero = uses kero.’

(291) endeumb ka-m ha-i

nettle 1st-GL scratch-RP
 84

‘I just got scratched by nettle.’

Waris has no true passive construction; pseudo-passive clauses (5.2.8) have some flavor of
passive. In Agent-Patient clauses, deletion of Agent allows the speaker to produce a passive-like
effect:

(292) ka-m-ba hélvo-wol vi-na os i

1st-GL-TOP hit-NS.O do.PL-PST intense
‘I really got stung (by wasps)!’

5.2.17 Goal or Patient-Benefactive Clause

Cases [+A +G/P +BEN]

In this class are grouped together Goal and Patient predications which have undergone
transformation by the addition of Benefactee to become ditransitive predications. In other words
this kind of clause is either like those of 5.2.15 or 5.2.16, with a Benefactee added. Compare
examples (281) and (293), and examples (286) and (294).

(293) hi-mba ka-m tuv-m hevra-mna-na

3rd-TOP 1st-GL door-GL hit-BEN-PST
‘He banged on the door for me (to wake me.)’

(294) ka-va ye-m-ba téh ha-wol

1st-TOP 2nd-GL-TOP firewood split-NS.O
vé-ne-vav
do-BEN.SG-FUT
‘I will split firewood for you.’

5.2.18 Sound-Speech Clause

Cases [+A/O +/-BEN +/-MEANS +/-RESULT +/-LOC]

This clause predicates the production of sound or speech. With animate agents it can form an
imperative. Optional case roles are the following:

-Benefactee, the one spoken to, marked with Goal and cross-referenced on the verb as
Benefactee

-Means, the language spoken, marked as Instrument with the Genitive

-Result, what is spoken, un-case marked

 85

-Locative

This clause undergoes transformation with the addition of another NP argument (the person
spoken about) to become a Speech Benefactive Clause (5.2.19).

(295) tuawa ó-v

bird speak-PRS
‘A bird is singing.’

(296) pawa ó-loh-vna sinim

generator speak-exist-CON night
‘A generator was running at night.’

(297) (moa) hi-mba ishó-v

words 3rd-TOP converse-PRS
‘He/she is speaking.’

(298) ye-m-ba ka-va óvó-mna-v nó timóv-pa

2nd-GL-TOP 1st-TOP forbid-BEN-PRS that fruit-TOP
‘I forbid you that fruit.’

(299) Englis-na moa-na ka-va he-m-ba

name-GEN words-GEN 1st-TOP
ishó-in-vna
3rd-GL-TOP converse-BEN.PL-CON
‘I was speaking to them in English.’

5.2.19 Speech Benefactive Clause

Cases [+A +BEN +GL]

This clause describes the speech of one person to another on behalf of a third party. It is closely
related to the Speech clause (5.2.18), but with one added argument, marked Goal, the person
about or to whom the speech is made.

(300) Néngai-putoa-m pi-mba Sesoa-m

name-aged-GL we.PL.INCL-TOP God-GL
óh-ósah-mana-vav
DL-ask-BEN.SG-FUT
‘You and I, the two, will pray to God concerning old man Néngai.’
 86

5.3 Distribution of Word Classes in Clauses

The distribution of word classes in clauses needs comment, since it helps define the word
classes and clause types.

1. The distribution of adverbs and particles is straightforward - they all modify or qualify the
predicate.

2. The distribution of verbs is straightforward - they all form the predicate. They obligatorily
occur in all clause types except Stative-Equative (5.2.2) and _can_ occur there, too.

3. The distribution of nominals is clear in light of the case analysis of the clause classes.

In an earlier draft of this sketch, it was proposed to specify the distribution of each nominal
class according to the types of clauses it can occur in. Now, it seems that is not necessary, in
light of the case analysis of clause types (5.1) according to the semantic roles of the nominals
involved, and in light of the semantic classification of 11 nominal classes (2.3). For example, in
a clause taken from one of the types which requires an Agent (5.1), only a nominal which is
semantically [+Human] may take this role.

5.4 Grammatical Relations in Waris

The definition of the grammatical relations in a language receives a lot of theoretical attention.
In this beginning analysis of Waris I take the position that since S, O and IO are all cross-
referenced on the verb for number, they clearly have the status of core grammatical functions.
Furthermore, it is a rare utterance in Waris where it is ambiguous which NP argument is to be
assigned to S, O, or IO (see below).

Seiler (1984a section 7.1ff) sees a problem in defining grammatical relations in the closely-
related Imonda language. For him, O and IO are not clearly core relations on the same level as
S. Interested readers should refer to his thesis.

Example (300) above is one of the few utterances heard in Waris in which the assignment of
NPs to core arguments might be ambiguous. One NP is un-case marked (pi 'we.INCL') and is
clearly the Subject. Two are marked with -m 'Goal' and therefore it might be ambiguous which
is the hearer of the speech (cross-referenced on the verb as the Benefactee), and which is the
person spoken about. However this ambiguity is resolved by other syntactic information, word
order. In my present understanding, the Benefactee must come right before the verb.
(Spontaneous (non-elicited) utterances like this are rarely heard.)
 87

6 Discourse

In this section I discuss Waris discourse features, with the emphasis, as in the preceding
sections, on semantics. First I discuss how clauses are joined together into sentences and
paragraphs. This includes types of sentences (6.2) and semantic paragraphs (6.3). Then I discuss
topicalization (6.4), topic continuity (6.5), information in discourse (6.6), relative clauses (6.7),
cohesion (6.8), and finally types of discourse used in Waris (6.9). The discussion here is
indebted to notes from Bob Litteral's grammar lectures.

6.1 Clause to Sentence and Discourse

The model followed in this discussion is as follows. Waris has no medial-final verb
distinction, most verbs being completely inflected. (The verbs that are not completely-inflected
do not come before an inflected final verb, but after one.)(6.3.1) In the case of a small number
of clauses, typically two, being joined together by conjunctions and other syntactic devices I call
this the sentence level. The processes of syntactic coordination and subordination that form
sentences I deal with in section

Sentences typically convey meanings such as conditional, simultaneous action, and counter-
expectation, but not sequence in past time. This latter meaning in Waris is conveyed by any
number of clauses chained together by certain syntactic devices, and I have chosen to call the
construction semantic paragraph or simply, discourse. The reason is that there are no syntactic
markers that serve to mark the boundaries of this unit; delineation of paragraphs is on the basis
of semantics. Thus, a typical Waris (narrative) discourse is analyzed partly into some sentences
(syntactic coordination and subordination) and some semantic paragraphs (semantic
topicalization and syntactic coordination).

6.1.1 Intonation

Intonation in Waris is a clause-level phenomenon and carries a low functional load even there.
Its role in joining clauses is low.

6.2 Sentence Constructions

In this section I discuss the conjunctions and other syntactic devices used to join clauses into
sentences. At the end I discuss coordination and subordination in general.
 88

6.2.1 Conditional Sentences

The co-occurance of irrealis mode and topic marking on the verb of the protasis are diagnostic
for conditional constructions in Waris:

(301) kapol ata pró-ra-va ata ka-va ga-vav

airplane will.today come-IRR-TOP will 1st-TOP go-FUT
‘If/when a plane comes today, I will go.’

The speaker's use of -ra in the protasis indicates he reasonably expects a plane to come. There is
an additional irrealis marker, -ta, which indicates the speaker's assessment of low probability or
undesireability of the action of the protasis (303). For past actions, it means Contrafactual - the
action did not occur:

(302) yi-mba doara weha-ta-va ata e-ne-ta

2nd-TOP before cook-IRR-TOP will.today DL-eat-IRR
‘If you had previously cooked food, (you and I, we) two would eat it today (but
you didn't and so we won't).’

Contrafactual is analyzed as a semantic alternation of the Conditional sentence since it shares

with Conditional the feature of having the protasis marked with Topic.

In the above two examples (301), (302)the apodosis exhibits future tense and irrealis,
respectively. In the following two examples (303), (304) the protasis and apodosis are both
filled by irrealis:

(303) yi-mba sala ve-ta-va ye-m-ba bui vi-ra

2nd-TOP sin do-IRR-TOP 2nd-GL-TOP jail do.PL-IRR
‘If you should sin, they could jail you.’

(304) hi-mba nó-mba ga-lm-ba ve-v-ra-va

3rd-TOP that-TOP go-TEL-TOP do-P S-IRR-RTOP
hi-mba nó-mba hui kusó-v-ra
3rd-TOP that-TOP here gather.PL-PRS-IRR
‘Whoever wants to go, they need to gather here.’

To express negative conditional ('if not exist', 'if not come'), the verb may be elided in the
protasis and replaced by 'no'. The topic marking remains on the noun argument:

(305) si pueil-va owai mie-mba hala-vav

future fence-TOP no pig-TOP eat.PL-FUT
‘If there is no _fence_, the pigs will eat (the garden).’

(306) kapol-va déti-va owai simera pró-vav

airplane-TOP today-TOP no tomorrow come-FUT
 89

‘If the airplane does not (come) _today_, it will come tomorrow.’

The topic marking on the words 'airplane' and 'today' means that the question had been raised
about the probability of a certain airplane arriving, and about it arriving on that particular day.
'Today' is an example of what Chafe (1976) calls 'point of contrast', and he discusses it along
with the notion of 'topic' in the same article.

Here is another example (307) of a conditional sentence, which also shows the use of topic
marking, both normal topic (_-mba_) and resumptive topic (_-oa_):

(307) henó-na-mba omó-mba kongala

whoever-GEN-TOP throat-TOP dry
v-in-v-ra-va henó-oa ka-indam amb-v-ra
do-BEN.PL-PRS-IRR-TOP whoever-RT 1st-ALL come.PL-PRS-IRR
‘Whoever(pl) is thirsty, as for them, let them come to me.’

6.2.2 Conditionals and Topics

Waris conditional sentences are a good example of Haiman's (1978) dictum 'conditionals are
topics' because that is exactly the way they are marked syntactically. In this analysis, I view
topic marking on the protasis of a conditional sentence as marking it as given information from
which the speaker is stepping off and making an assertion.

The basic meaning of topic in Waris seems to be givenness or definiteness, and so in many of
the examples so far, topic has been glossed as the.

The resumptive topic marker (307) is glossed as 'as for them'. In that example it refers back to
the topic of the preceding clause 'whoever'.

6.2.3 Simultaneous Action Sentence

This kind of sentence is recognized by the continuous action verb suffix -vna on both verbs.
The time adverb nónam 'at that time' may be present.

(308) ka-va pró-vna (nónam-ba) yi-mba a-vna

1st-TOP come-CON (then-TOP) 2nd-TOP sit-CON
‘When I came you were sitting (just then).’

6.2.4 Counter-expectation Sentence

The conjunction heva os 'thus; but' is diagnostic for this kind of sentence. The word 'no' may
occur with it.
 90

(309) ka-va ga-na-mba heva os

1st-TOP go-PST-TOP but no
owai tambkó-va naloh-v-moa
fish-TOP exist-PRS-NEG
‘I went (to the store), but no, there is no fish.’

In order to understand the use of this construction, one has to enter into the thinking of native
speakers about what is reasonable to expect to happen and what is not. Sometimes the same
conjunction joins clauses that are not in the semantic relationship 'expectation - counter-
expectation' but rather 'undesirable reason – undesirable result' as in the following example
(310):

(310) angumina heva os ishó-vna-moa

deaf thus speak-CON-NEG
besowonam moa-mba
well talk-TOP
‘(He/she is) deaf so does not talk well.’

6.2.5 Reason or Result Sentence

The same caveat applies to understanding the use of this kind of sentence in Waris as applies
to the preceding Counter-expectation sentence. One needs to enter into the thinking of the native
speaker about what constitutes cause and effect. The conjunction noinda 'at that = thus' is the
diagnostic feature of the Reason/Result sentence. The interrogative manara 'what reason?' may
co-occur before noinda with no change in meaning. Its effect is to slow down the rate of new
information in unfamiliar material, as in translated Scripture. The pause form néngara 'let me
think, for this reason..' can occur with manara to act as another conjunction for Reason/Result
sentences. It also has the effect of slowing down the rate of new information.

The conjunction noinda seems to merely imply that there is some logical relationship between
two clauses in the mind of the speaker, and the hearer needs to interpret. Thus, in the following
example (311), the conjunction occurs before the result. In the next example (312) the
conjunction occurs before the reason; it was taken from a discourse on nettle where it supplied a
reason for removing nettle carefully from one's garden.

(311) moa-mba he-na-mba besel

talk-TOP 3rd-GEN-TOP good
wo noinda he-na-mba aong-va elepe-wol-mini-vna
and thus 3rd-GEN-TOP ear-TOP put-NS.O-BEN.SG-CON
‘His talk was good so they listened to him intently.’
 91

(312) noinda endeumb-ra-va holvó-pétha-ra-va

thus nettle-LOC-TOP CLAS-fall-IRR-TOP
aval-va gumus vovhó
mother-TOP tired experience
‘Because if (a child) falls into the nettle, the mother will get tired (of having to
comfort its crying).’

The correct interpretation of this sentence depends on inside knowledge about what happens to a
child who falls into nettle.

A syntactic feature of this construction is the non-inflected verb stem vovhó 'to experience'. This
is not a feature of this kind of sentence but rather is a general way of backgrounding the
information in the verb (6.6.3). A cultural script is being followed at this point which contains
all the information about what happens when people contact nettle. The speaker does not want
to emphasize this more than is necessary, because his global theme was not 'the experience of
mothers' but 'nettle'.

6.2.6 Alternative Sentence

This sentence proposes two alternative actions that may occur.

(313) Sesoa si is-mana-vav-mi i

God will fill-BEN-FUT-perhaps or
óvó-mna-vav-mi
forbid-BEN-FUT-perhaps
‘God will perhaps fill (ideas) for him or perhaps deny him (them).’

(314) ga-v-mi owai-mi

go-PRS-perhaps no-perhaps
‘(He/she) will perhaps go.’

(315) óra ka-va ga-v óra ka-va a-v

or 1st-TOP go-PRS or 1st-TOP sit-PRS
‘I might go or I might stay.’

The conjunction óra 'or' is derived from the verb óv 'to speak', which was discussed in section
4.3.1 as the only verb that is used as the first member of a true serial verb (two finite verbs in
series, with nothing between, and manifesting one clause). The meaning in that construction is
'intention'. I would say that there is some sense of shared meaning between serial verbs using óv
'intend to', and alternative sentences where the irrealis marker -ra has been added to that verb to
produce the conjunction óra 'or'. Compare example (315) above with the following example of
an intention sentence using the verb 'to say', (316).
 92

(316) ka-va om ó-na ga-na

1st-TOP yesterday speak-PST go-PST
heva os owai po hasvo-na
but no rain fell-PST
‘I wanted to go yesterday but it rained.’

In the following example (317) a still third form of the verb óv 'to speak' is used to connect two
clauses of an alternative sentence. Compare section 6.2.9 Comparison Sentence, where example
(317) might just as well fit.

(317) ó-i os yes-va ten-ba sahokla ve-v-pa

say-RP thus sago-TOP that-TOP bad do-PRS-TOP
ó-i yes-va nó-mba besowonam daha-i
say-RP sago-TOP that-TOP well set-RP
‘It's like _that_ sago is still bad (but) _that other_ sago has jelled well.’

In the final example of alternative sentence (318) the particle mava 'sometimes, perhaps' is used
to coordinate the two clauses.

(318) mava posal e-nga-vna mava e-rombo-vna

perhaps futile DL-go-CON perhaps DL-get.PL.O-CON
Perhaps (the two) are going (home) empty-handed, perhaps they are getting
fish).’

6.2.7 Purpose Sentence

In this sentence the purpose is conveyed by an embedded clause marked either with irrealis or
with Telic. Embedded purpose clauses with irrealis may be introduced by the conjunction óra 'if'
(discussed in section 6.2.6 above).

(319) ka-va daswoto-mana ve-vna-moa indhana-m-ba

1st-TOP trick-around do-CON-NEG people-GL-TOP
[osm óra ka-m-oa wive-wol vi-ra-e] owai
so that 1st-GL-RT praise-NS.O do.PL-IRR-quote no
‘I wasn't (going) around tricking people so that they would praise me, no I
wasn't).’

The conjunction in this example consists of two words, osm, built of os 'thus' and -m Goal, plus
óra, built of óv and -ra 'irrealis'. I gloss it as 'so that' but paraphrase it as 'as if I were saying to
people'. The reason for this paraphrase is that the verb of the embedded clause has the quote
suffix -e meaning that the embedded clause is actually a quote. So I would paraphrase the entire
example this way:
 93

'I wasn't going around tricking people, saying to them "as for me, you praise me".'

Example (319) has irrealis on the embedded clause verb and the following example (320) has
Telic case marking on the embedded verb:

(320) hi-mba pró-na [ye-na-m kél-m

3rd-TOP come-PST 2nd-GEN-GL bone-GL
wosapr-in-ilm]
plant-BEN.PL-TEL
‘He came in order to give you (pl) strength.’

In this example the Goal marking on 'your' and 'bone' is a syntactic feature of embedded
purpose clauses marked with Telic on the verb. It does not seem to contribute to the meaning.
The plural Benefactee marked on the verb refers to the addressees of this utterance.

In the following example, the embedded purpose clause is also of the kind marked with Telic. It
is reduced to a lone verb, which is affixed for singular Subject and singular Benefactee. This
gives the meaning 'for one of them to ask him'.

(321) hi-mba pe-m vi-mini-vna

3rd-TOP fear-GL do.PL-BEN.SG-CON
ósah-mana-lm-ba
ask.SG.S-BEN.SG-TEL-TOP
‘They were afraid of him, for one of them to ask him (for something).’

6.2.8 Speech Quote Sentence

The significant thing about Waris speech quotations is that they can occur not only after verbs
of speech, but also of thought (325 -329). The suffix -e is used frequently to mark quoted
speech. It has the other functions of coordinating NPs in lists, and of forming interjections from
names: Luk-e 'hey, Luke!'. At present I understand this suffix as not merely marking quoted
speech but of marking only that quoted speech which the narrator feels strongly. This explains
the fact that it does not occur after all quotes.

(322) hi-mba indhana-m-ba os ishó-mana

3rd-TOP people-GL-TOP thus speak-around
v-in-vna ka-m patv-man-o-e
do-BEN.PL-CON 1st-GL follow.PL-BEN.SG-IMP-quote
‘He was telling various people "(you) (pl) follow me".’

(323) ótó-va nó-mba os sevri-v bas

auto-TOP that-TOP thus call.PL-PRS bus
‘(People) call that auto a "bus".’
 94

(324) os ka-va néng-na mani hev-pa dihi

thus 1st-TOP think-PST what 3rd-TOP perhaps
‘I thought "what could this be?".’

(325) ka-m-ba ind ve-uv-mana

1st-GL-TOP show do-daily.PL.S-around
v-in-vna os-m men-ba loh-v os-m
do.PL.S-BEN.PL-CON thus-GL this-TOP exist-PRS
men-ba loh-v
thus-GLthis-TOP exist-PRS
‘(People) were showing us around "this is for thus, this is for thus".’

(326) hi-mba und ve-vna besel hev-m-e

3rd-TOP love do-CON good 3rd.EMP-GL-quote
‘He loved her "she is beautiful".’

6.2.9 Comparison Sentence

This sentence is similar to the Speech Quote Sentence in that the verb óv 'to speak' is used and
the quote suffix -e may occur. However, the form of the verb must be recent past tense, the
tense used elsewhere in descriptive discourses (6.9.3). The function of the quote suffix may be
viewed as adding prominence to the content of the comparison or it may be viewed as putting
words in the mouth of the addressee or some third person when they see the comparison.

(327) ye-tindi-va si mindanam loh-vav

3rd-clan-TOP future many exist-FUT
ó-i engal di wan-da-e
speak-RP sand there salt-LOC-quote
‘Your descendants will be many in the future, like sand on the shore.’

or

'Your descendants will be many in the future; (you/someone will) say "(they
are) like sand on the shore".'

Compare example (317) above with this; it might belong in this category.

6.2.10 Intention Sentence

This sentence type also is based on the verb óv 'to speak'. As discussed in section 4.3.1 this is
the only verb that introduces genuine (African type) serial constructions: two fully inflected
verbs with nothing between and manifesting the same clause. The meaning of this sentence type
 95

is not just intention but of either _frustrated_ intention or unexpected result. In the following
example (328) the second clause gives a reason why the speaker's intention was frustrated. In
example (329) the second clause gives the content of what was seen, in the form of quoted
speech.

(328) ka-va om ó-na ga-na heva os owai po

1st-TOP yesterday say-PST go-PST but no rain
hasvo-na
fell-PST
‘I wanted to go yesterday but it rained.’

(329) ka-va ó-na nungeil-na men he-m-ba

1st-TOP speak-PST see.PL-PST here
pili-vonga-v
3rd-GL-TOP fill-ascend-PRS
‘We tried to see/began to see (what was happening) "(the elevator) is bringing
(people) up here!".’

In example (329) the number of actors is given by the plural (stem) form of the verb 'to see'; the
verb 'to speak' is in the singular. The speech quote gives what the speakers said to one another
when they went to look at an elevator at Yonki power station.

6.2.11 Syntactic Coordination and Subordination

In this section I deal with the above seven sentence types in view of the difference between
coordination and subordination. (Here I follow Ray Johnston's lecture notes.)

6.2.11.1 Subordination
This involves a non-equivalence relationship between clauses in which one is typically
sentence-initial, expresses a presupposition and is syntactically derivative. This is the
subordinate clause. The main clause typically comes second, has normal tense-mode-aspect
marking and is the figure of intended focus by the speaker. There are two Waris sentences that
fit this bill, the Conditional Sentence (6.2.1) and the Purpose Sentence (6.2.7). Speech Quote
Sentence (6.2.8) probably also fits in this category. The following example (330) is of a
Conditional Sentence.

(330) kapol ata pró-ra-va ata ka-va ga-vav

airplane will.today come-IRR-TOP will.today 1st-TOP go-FUT
‘If/when a plane comes today, I will go (on it).’
 96

The first clause expresses a presupposition and is derived in the sense that it is marked with
Irrealis. Being marked with Topic, it conveys the background or given information the speaker
steps off from. The second clause is the figure of intended focus. Thus, this typical Conditional
sentence fits in very well with Johnston's definition of subordination. The following example
(331) is of a Purpose Sentence.

(331) hi-mba pe-m vi-mini-vna

3rd-TOP fear-GL do.PL-BEN.SG-CON
ósah-mana-lm-ba
ask.SG.S-BEN.SG-TEL-TOP
‘They were afraid of him, to ask him (for something).’

In this example (331) the subordinate clause comes second. It expresses the goal or purpose
people were afraid of trying to attain, and so is marked with Telic case (section 3.3 chart). The
topic marking on the subordinate clause means it is the presupposition or background or context
from which the speaker departs to make their point, which is the first clause.

Referring back to example (322) of a Speech Quote Sentence, the definition of Johnston, that a
main clause expresses an illocutionary act while a subordinate clause expresses a
presupposition, fits that kind of sentence very well. It seems that Comparison Sentences (6.2.9)
probably fall under this same definition, too.

Thus in Waris I would presently maintain that conditional (including counterfactual), , quoted
speech and purpose clauses, and probably comparison clauses, are all subordinated to their main
clauses in the sentences in which they occur.

6.2.11.2 Coordination
As far as coordination is concerned, I would maintain that all the rest of the sentence types
above (6.2.1ff) fall in this category. Coordination is a whole-whole equivalence relationship
between propositions where the units are comparable in semantic and syntactic function. In
Waris it would include not only coordinated clauses which are not mutually-exclusive, such as
Reason or Result (6.2.5), but also coordinated _alternatives_ as in the Alternative Sentence
(6.2.6).

6.3 Semantic Paragraphs

In this section I discuss the way clauses and sentences are built into levels above sentence.
There are few syntactic markers of the paragraph or discourse level in Waris that I am aware of
now, and so I call it basically a _semantic_ unit. The syntactic marker that does stand out is the
topic slot, that can occur at the beginning, inside and at the end of a paragraph.
 97

(332) endeumb-va pepeh ve-wol vi-v

nettle-TOP fear do-NS.O do.PL-PRS
ósó-ra-va
garden-LOC-TOP
‘Nettle, in the garden (people) are afraid of it!’

In this beginning clause of a story about nettle, the word 'nettle' is not part of the syntactic
structure of the clause. It does not fit into my analysis of clause grammar but it does fit into a
pattern of discourse topics that occur both at the beginning and within a paragraph. The reaction
of one native speaker who edited a written version of this story was to put a comma after the
first word.

Topics not only occur at the beginning of a paragraph but also within it. Sometimes they are
marked with the Topic suffix, like 'nettle' in the above example, sometimes they are marked by
the suffix -e (322), which means something like 'emotional involvement of speaker'. (The same
form means 'quoted speech' when occurring after a quote, and 'hey, there' when suffixed to a
person's name.)

(333) a manda-mba heva hev ambo pho-wol ve-v

kwila-TOP then just come.up-NS.O do-PRS
‘Kwila trees just sprout up around of their own accord.’

b heva hev ambo loha-i-va móvól-e

then 3rd.EMP just exist-RP-TOP fruit-EMP
'They just exist of their own accord, (then) _fruit_.'

c móvól ketha
fruit hang
'There is fruit hanging (on the trees).'

d damba-ngas hala-wol ve-v manda-móv-pa

cockatoo-only eat.PL-NS.O do-PRS kwila-fruit-TOP
'Only cockatoos eat kwila seed.'

My present understanding of sematic paragraphs and topics in Waris would explain this
example as follows. 'Kwila' is the syntactic subject of the first clause. But, falling at the very
beginning of the paragraph, it can be a paragraph topic as well. When there is a change of topic
between the second and third clauses, the new topic, 'fruit' is introduced. In this transcription it
is attached to the second clause, where it seems to go phonologically, rather than being prefixed
to the third clause. In either place it is not part of the clause structure but rather a pargraph-level
topic. In the fourth clause, 'kwila fruit' is the object of the verb 'eat'. Its normal location would
be before the verb, but here it has been moved to clause-final position. There are other examples
of this in the data and I would maintain that, on both the clause level and paragraph level, 'kwila
 98

fruit' has been put in a position to indicate that it is the topic. This is supported by the fact that it
has the topic suffix. ('Kwila' and 'kwila fruit' overlap semantically enough to avoid a charge of
contradiction in saying this.)

The following example (334) is the first clause of a story about frogmouths [birds]. The first
word 'frogmouth' is followed by pronominal copy. I would maintain that, as the subject and
topic of its clause, it also falls in the paragraph topic slot and, through the added prominence of
the pronominal copy, fills the role of a paragraph topic.

(334) sut-va hi-mba besowonam

frogmouth-TOP 3rd-TOP well
na-ó-mana óvó-v
eat-daily-around exist.PL-PRS
‘The frogmouths, they (just go) around eating well.’

Once the topic of a semantic paragraph has been established, the structure of the paragraph
depends on the following relationships.

1. subordination of some clauses to others, which is dealt with in section 6.2.11.

2. coordination of clauses or _clause chaining_, which is discussed in section 6.3.1, the next
section.

3. topicalization and cohesive strategies. For a discussion of these see sections 6.4 and 6.8.

6.3.1 Clause Chaining and Semantic Paragraphs

In the above section I showed the main syntactic marker of a paragraph or discourse in Waris
is a slot for a semantic topic. It can occur initially, finally, or inside a paragraph, and 'fits' with
the clause-level topic, which can be fronted or backed within the clause. There is one other
syntactic feature that binds paragraphs together, and that is head-to-tail chaining of clauses.
Waris has no medial-final distinction of verbs which can be chained in order to bind paragraphs,
but it does the same thing by recapitulating the verb of one clause at the beginning of the next.
The preceding verb is suffixed with Topic 'given information'.

I analyze chaining of clauses as another example of coordination as first discussed in section

6.2.11. There, syntactic coordination was based on the use of conjunctions and verb endings.
Here, the coordination is based on the use of

1. topic marking

2. verb recapitulation
 99

3. two other conjunctions that mark, not logical relationships or time relationships as in section
6.2 , but relative prominence of the clause in the information structure of the discourse. The
following example (335) is the beginning of a story about a father-and-son hunting trip.

(335) a ka-va ve-na kav-na aral e-nga

1st-TOP do-PST 1st.EMP-GEN father DL-go
‘I did, I and my father went.’

b winde-m won-e-nga-na-mba
dog-GL PL.ACC-DL-go-PST-TOP
'(We) two took dogs.'

c doa mie-m e-wole

OK pig-GL DL-shoot
'OK (we) two shot a pig.'

d mie-m e-wola-na-mba e-puise

pig-GL DL-shoot-PST-TOP DL-cut.up
'(We) two having shot a pig, (we) two cut (it) up.'

e e-pusa-na-mba
DL-cut.up-PST-TOP
'(We) two having cut it up.'

f doa ara-i e-nga deuv-ram

OK father-COM DL-go house-ALL
'OK (I) and my father went home.'

This is a transcription of an oral story. Statistically, it is extreme in the frequency of

recapitulation of verbs, since this feature tends to be omitted almost completely in written texts
or edited oral texts.

Regarding the word _doa_, it means 'completed action' when in a verb phrase. But at the
beginning of a clause it means 'prominent event'. Thus, after non-event information it means
'return to the event line'. I gloss it as 'OK'

Regarding the use of Topic marking on the recapitulated verbs, I maintain that it has exactly the
same meaning here as it does on the verb (protasis) of a conditional sentence (6.2.1), namely
'the information in this clause is now given/presupposed and I depart from here to continue my
story'.

The following example (336) is the beginning of a written text by an educated Waris speaker. It
shows less recapitulation, a different distribution of Topic marking on the verbs, and shows
another conjunction _wo_ 'and', which means 'event, but not prominent'. (The Telic case suffix, _-
lm_, on a verb means 'purpose' as in section 3.3 chart; a syntactic label would be 'gerund').
 100

(336) a doa ka-va kav-na pró-lm mo-m ishó-lm

OK 1st-TOP 1st.EMP-GEN come-TEL talk-GL speak-TEL
ve-v
do-PRS
‘OK I want to tell my story about coming (here).’

doa seilva os di-va ka-va Amanav-ra Bov-i-va

OK first there-TOP 1st-TOP Amanab-LOC Bob-COM-TOP

b ah-a-na
DL-sit-PST
'OK first I stayed there at Amanab with Bob.'

c mingu-va móngal ka-va e-ve-na

week-TOP one 1st-TOP DL-do-PST
'We two spent a week.'

d wo ka-va rabu-ram e-pra-na hom-ba

and 1st-TOP Wed.-ALL DL-come-PST to.here-TOP
'And we two came on Wednesday, to here.'

e kapol-va ka-m-ba ónka-in-na Amanav-rini-va

plane-TOP 1st-GL-TOP get-BEN.NS Amanab-ABL-TOP
'The plane got us from Amanab.'

f doa nó-mini-va vi-na-mba

OK there-ABL-TOP do.PL-TOP
'OK (we pl) went away from there.'

g emb-na-mba
come.PL-TOP
'(We all) came.'

h Ambunti peive
Ambunti descend.PL
'(We all) descended at Ambunti.'

i minahénga kapol-m dahu-na-mba

fuel next plane-GL feed.PL-PST-TOP
'(we) refueled the plane.'

j doa hénga nó-mini-va pulvó

OK next there-ABL-TOP get.up.PL
'OK next we left from there.'
 101

One difference between the two above texts (335) and (336) is in the use of unaffixed verbs. In
the former most verbs are bare stems in the first mention. In the latter there are only two
unaffixed verbs. The bare stem of a verb following an inflected verb is a syntactic feature which
decreases the prominence of the semantic content of that verb in the information structure of the
discourse. It deletes tense-aspect information that can be supplied from context plus a
knowledge of the real world or cultural script being followed.

6.3.2 Summary of Semantic Paragraph

In summary, I would maintain the following about (narrative) paragraphs in Waris. They are
delineated semantically by the presence of topic slots. They are held together syntactically by
verb recapitulation, topic marking on the verb, and by conjunctions. The conjunctions also serve
to distinguish more prominent and less prominent clauses, as does the decision to use affixed
verbs or bare stems.

The main question about paragraph syntax I cannot answer now is the following: what is the
function of suffixing Topic marking to some verbs, but not others? (It does not seem to correlate
with the distinction between more prominent or foregrounded, and backgrounded information.)

In section 6.9 I give an inventory of the types of discourse I have recorded in Waris. Each type
is made up of a different type of semantic paragraph or paragraphs. For example, a narrative
paragraph consists of a location or locations (part of a trajectory) and events that happened at
each. In contrast, an argumentative paragraph consists of at least two speech acts joined by an
adversative; the effect is like this: 'my opponents say ''we should do so and so'' but I say ''we
should do so and so''.' For more information about semantic paragraph types see section 6.9.

6.4 Topicalization
Topic marking can occur on any word except a particle. The reader of this paper is already
aware of the wide distribution of topic marking, and its function has been discussed at a few
places. In this section I will first discuss the meaning of 'topic' as applied to Waris. Then I
discuss the way topic marking occurs on the pronouns, including the allomorphs. Then I discuss
how it functions within clauses, then within sentences and paragaphs.

6.4.1 The Meaning of 'Topic' in Waris

In dealing with topic marking I Waris I have decided to follow the definition of Comrie (1981
pp.56-59), that topic is what the speaker is talking about. As Bob Conrad likes to emphasize,
that is usually given information. (It contrasts with focus, which is the essential piece of new
information.) I choose to use his definition not for theoretical reasons but merely because this
general definition serves as a good heading under which to arrange the various functions of
 102

topic in this language. Seiler (1985) deals with topicalization in Imonda, a closely related
language, from the standpoint of Haiman (1978) and Dik (1978). My discussion here is
indebted to his.

6.4.2 Topic Marker Allomorphs and Pronouns

Following is a chart of Waris pronouns with Topic marking suffixed.

----------------------------------------------------------

| | AGENT | GOAL |

| | | |

| | normal emphatic | normal emphatic |

| | | |

|---------------------------------------------------------

| | | |

| 1st person | ka-va kav-pa | ka-m-ba kav-m-ba |

| | | |

| 2nd person | yi-mba pev-pa | ye-m-ba pev-m-ba |

| | | |

| 3rd person | hi-mba hev-pa | he-m-ba hev-m-ba |

| | | |

| 1st inclusive | pi-mba piv-pa | pi-m-ba piv-m-ba |

| | | |

----------------------------------------------------------

Regarding this chart, the following should be noted. The practical orthography we have used so
far (with adults literate in another language) includes non-phonemic prenasalization on voiced
 103

stops. For convenience this orthography has been used in this paper, which introduces ambiguity
of spelling between the Agent and Goal in 1st person inclusive.

In spoken Waris, 1st person inclusive normal Agent pronoun is homophonous with 1st person
inclusive Goal pronoun. This does not seem to cause a major problem of interpreting utterences.
But in the other three persons of the pronoun, Waris speakers appear to have made an
adjustment in the phonemic form of the normal Agent pronouns to avoid ambiguity. Here is the
reasoning behind this claim:

The rules regarding distribution of the forms of the Topic suffix are as follows, with minor
exceptions:

1. -pa follows nouns ending in /v, p/ and verbs ending in /v/.

2. -mba follows nouns ending in /m, n, a, e, é, ó/; it follows verbs ending in /a/.

3. -va follows nouns ending with any other segement, plus verbs ending in /i/.

On the basis of these rules, the normal Agent pronouns would be [kamba, yemba, hemba]. This
would make them homophonous with the corresponding Goal pronouns. So it seems that the
rule that assigns -mba to ka has been ignored and ka-va is used instead. Furthermore, the vowels
of the expected forms yemba and hemba have been modified phonemically to yield yimba and
himba. Thus ambiguity between Agent and Goal in the first three persons is avoided.

I believe that this process is another demonstration of the tendency in the language to
distinguish grammatical subject and object clearly (since most times Agent is a subject and Goal
an object). I would take this as another evidence of the core status of O as well as S in Waris.

6.4.3 Survey of Topic Marking

In this section I will review the uses of Topic marking in Waris beginning within the clause,
and working up.

1. Definiteness

Topic marking on adjectives gives definiteness.

(337) besel-va hi
good-TOP where
‘Where is a/the good one? (this one is broken)’

2. Pronominal Subjects and Objects

 104

Pronominal subjects generally have topic marking, pronominal objects less frequently so.
Exceptions are some relative clauses and optatives, and imperatives. (Imperatives may take
resumptive topic.) Negative prohibitions take topic on the subject.

(338) ka-va ga-v

1st-TOP go-PRS
‘I'm leaving.’

(339) ka ga-vai
1st go-OPT
I want to go!’

(340) ga-o
go-IMP
Go!’

(341) ye-oa ga-v-m

2nd-RT go-PRS-GL
As for you, go (I'm staying).’

(342) ye-m ka-va ishó-in-v hona-mba

2nd-GL 1st-TOP converse-BEN.PL-PRS this-TOP
plal-la-va [ye avul-v]
root-LOC-TOP 2nd sit.PL-PRS
‘I am telling you (pl) the basis by which [you exist].’

(343) a an-m yi-mba nongle-v

who-GL 2nd-TOP look-PRS
‘Whom do you see?’

b ye-m-ba ka-va nongle-v

2nd-GL-TOP 1st-TOP see-PRS
'I see _you_'.

In the last example above, the topicalized object is also fronted in the clause rather than
occupying its normal slot before the verb.

I maintain that the use of the topic suffix on the above examples falls within the scope of
Comrie's definition of what is being taked about. I also believe that it is possible to formulate a
rule based on the topic of a given discourse and its continuity through the discourse which
predicts the occurance of topic suffixes on NP subjects and objects in the discourse. See section
6.5.4.

3. Topicalization of Deictics
 105

Topicalization of deictics adds prominence, as in an English cleft sentence.

(344) a tendórini ten a-v

man here sit-PRS
‘A man is sitting there.’

b tendórini ten-ba a-v

man here-TOP sit-PRS
'That there, it is a man. = That is a man sitting there.'

4. Point of Contrast

What Chafe (1976) calls contrasts can be marked as topic in Waris.

(345) manam yi-mba hui-va a-vav

why 2nd-TOP here-TOP sit-FUT
‘Don't sit down here (sit elsewhere)!’

Sometimes the topic marking is accompanied by a change in word order that further conditions
topicalization. Contrast the next four examples (346- 349).

(346) ka-va ata nelus ne-vav

1st-TOP will green eat-FUT
‘I will eat greens.’

(347) ka-va ata nelus-va ne-vav

1st-TOP will greens-TOP eat-FUT
‘I will eat the greens (you just gave me).’

(348) nelus-va ata ka-va ne-vav

greens-TOP will 1st-TOP eat-FUT
‘The greens (you just gave me), I will eat them.’

(349) nelus-va mani ata ka-va ne-vav

green-TOP what will 1st-TOP eat-FUT
‘Greens, what am I going to eat for greens?’

5. Conditional Clause

A verb in the protasis of a conditional sentence are affixed with topic. See section 6.2.1.

(350) kapol-va ata pró-ra-va ata ka-va ga-vav

plane-TOP will.today come-IRR-TOP will 1st.TOP go-FUT
‘If the plane comes today, I will go.’

6. Purpose Clause
 106

A purpose clause of the type marked with Telic may have topic marking.

(351) hi-mba pe-m vi-mini-vna

3rd-TOP fear-GL do.PL-BEN.SG-CON
ósah-mana-lm-ba
ask-BEN.SG-TEL-TOP
‘They were afraid to ask him.’

7. Given Information in General

A whole clause may be suffixed for topic when it is used as given information in context.

(352) hém yi-mba ga-i-va

where 2nd-TOP go-RP-TOP
‘Where did you go (now that I see you are returning)?’

8. Clause Chaining

Coordination of clauses into paragraphs depends on head-to-tail linkage. Topic marking seems
to make the first clause a topic from which the speaker departs to make the next point. For an
example of this refer to example (336) clauses 6,7,9.

6.5 Topic Continuity

In this section I describe the rules followed by Waris speakers in introducing participants,
marking topics on the clause and discourse level, and maintaining continuity. This includes
deixis, and subject change pronouns.

6.5.1 Introduction of Participants

Waris commonly introduces participants by means of an introductory sentence. This also

establishes the topic of the discourse and sometimes the time. In the following examples (353)
and (355), the 1st person pronoun is marked with topic, meaning, I believe, old information
being reintroduced into the story as topic of this clause'. The respective discourse topics 'story of
my coming' and 'house' are new information and not marked with topic. In example (354), the
narrator is introducing a traditional story and I suspect the topic marking on 'two men' means
both 'topic of this clause' and old information you know from having heard this story many
times before'. The topic marking on the word 'some' in the time phrase 'some then' I believe is
an example of what Chafe (1976) calls the 'contrastive' function of topic. It marks what is old
information to the hearers as being prominent - a certain time, like the formulaic introduction to
English fairy tales 'once upon a time'.
 107

(353) doa ka-va kav-na pró-lm mo-m ishó-lm

OK 1st-TOP 1st.EMP-GEN come-TEL talk-GL converse-TEL
ve-v
do-PRS
‘OK I want to tell (the) story of my coming (here).’

(354) doa móngó-va heva tendórini-sambla-va

OK some-TOP then men-two-TOP
ah-a-vna okómbahul
DL-sit-CON sun.heat
‘OK once there were two men sitting in the heat of the day.’

(355) deuv-m ka-va ó-v

house-GL 1st-TOP speak-PRS
‘I'm talking about (the) village.’

6.5.2 Same Subject and Different Subject Pronouns

Third person referents in Waris can be distinguished as to 'same subject as the preceding
clause' and 'different subject from the preceding clause'. This is done by alternating the normal
he '3rd' and emphatic hev '3rd' pronouns. (It is unlike the system of obviation in NA Indian
languages where a specifically fourth person form is used.). Third person participants are
introduced into discourse by means of the normal pronoun. After being identified, reference
shifts to hev 'em yet'. Reference to another 3rd person referent triggers the use of he again.
Conjunctions and the word owai 'no' upset this neat pattern; they require the use of he in the
clause following. This applies not only to subject pronouns but also to possessive pronouns, and
not only to third person pronouns but also to first person pronouns. In the following example
(356), the referent is first person.

(356) doa ka-va os-va hil-ula-na-mba

OK 1st-TOP thus-TOP hear.PL-happy-PST-TOP
doa kav u-li-na besel-angas...
OK 1st.EMP say.PL-happy-PST good-alone
doa nó-si-va sha-ula-na-mba
OK that-night-TOP sleep.PL-happy-PST-TOP
lingilliha-na-mba
dawn-PST-TOP
doa kav po-ra ahal-ungu ve-vna
OK 1st.EMP water-LOC jump-around do-CON
 108

‘OK we heard that gladly, OK we said (to one another) ''very good...', OK (we)
slept happily that night, it dawned, OK we bathed...’.

In the style that native speakers prefer for precise speech and translated Scripture, pronominal
possessors in a clause that are co-referential with the Subject must made explicit, and they are
made explicit throught the use of the emphatic pronoun. Thus, in the following example (357)
omission of the word hev-na can be acceptable in oral speech but is unacceptable written.

(357) hi-mba hev-na móngala-na pró-i

3rd-TOP 3rd.EMP-GEN leg-GEN come-RP
‘He came by foot. = Em i wokabaut i kam long lek bilong em yet’.

Having seen this use of the emphatic pronouns the reader can now understand the following
examples.

(358) moa-mba he-na-mba besel

talk-TOP 3rd-GEN-TOP good
wo noinda he-na-mba aong-va elepe-wol-mini-vna
and thus 3rd-GEN-TOP ear-TOP put-NS.O-BEN.PL-PRS
‘His talk was good and so as a result they listened intently.’

The two pronouns he have different referents, a person who was speaking, and other people who
were listening to him. If the second clause is considered in isolation, the pronoun he should be
hev, because it was the actor's own ears they were 'putting'. (Compare example (357). However,
because of subject change between the first and second clause and probably because of the
conjunction, too, he is used.

In the following example (359), the shift from hev to he indicates no change of subject because
of the intervening owai.

(359) wo hev ga-na-moa dom-ba

and 3rd.EMP go-PST-NEG to.there-TOP
owai hi-mba hiliwal-nga-na Masetonia-hóv-mona-m hénga
no 3rd-TOP straight-go-PST name-middle-road-GL next
‘..and he did not go there, no, he next went straight on the road through
Macedonia.’

Another deviation from the pattern of hev = SS and he = DS is when speech is being reported.
Then, all references are with hev.

(360) doa sis-va Sesoa-va Apraham-na-m-ba

OK later-TOP God-TOP Abraham-GEN-GL-TOP
néngavnong-m léh-mana-vna óra ka-m-angas loh-mana-v
thought-GL test-BEN.SG-CON if 1st-GL-alone stand-BEN-PRS
 109

hev ó-mna-na Apraham-e

3rd.EMP speak-BEN.SG-PST name-quote
wo hev ó-mna-na me-ka-va a-v
and 3rd.EMP speak-BEN.SG-PST here-1st-TOP sit-PRS
‘OK later God wanted to test Abraham's thoughts ''does he believe in me
alone?'' (and) he said to him ''Abraham!'' and he said to him ''here I am''.’

6.5.3 Deixis

In Waris the far deictics nói 'that one' and nó 'that' are used for unmarked anaphoric reference.
The near deictics honi 'this one' and hona 'this' are used for marked anaphoric reference,
meaning a referent that is not only recoverable from the context, but is of special relevance or
prominence. For cataphoric (new information) reference the deictics memba 'this one' and
temba 'that one' are used. Their unmarked meaning is 'near or far distance' and their marked
meaning is something like 'near or far in relevance to the discourse'. When two kinds of deictics
are used, as in a Relative Clause (6.7), I would maintain that that relative clause is marked for
prominence. See example (386), where the Head of such a clause is marked with Topic suffix.
Also see example (362) below. In the following example (361) a deictic and pronoun are used
together to convey the meaning 'whoever'.

(361) hi-mba nó-mba ga-lm-ba ve-v-ra-va

3rd-TOP that-TOP go-TEL-TOP do-PRS-IRR-TOP
hi-mba nó-mba hui kusó-v-ra
3rd-TOP that-TOP here gather.PL-PRS-IRR
‘Whoever wants to go, they must gather here.’

In this example the deictics are pointing not at something recoverable from the discourse but at
something outside it, namely a group of people who, in the mind of the speaker, want to go. In
the following example (362) a personal pronoun himba 'they' is strengthened by a deictic nómba
'those'.

(362) doa hev Yesus-nind

OK 3rd.EMP name-people
[ten di Antiok-kembel-la avul-vna-m]
there there name-village-LOC sit.PL-CON-GL
hi-mba nó-mba indku vithi-vna Sesoa-m-ba
3rd-TOP that-TOP glad do.PL-CON God-GL-TOP
‘OK the Christians [who lived there in Antioch] _those_ were glad about God.’

In this example the personal pronoun maintains continuity of reference with preceding clauses.
The deictic points back to this group of people having been mentioned in the discourse and
 110

gives a meaning something like 'those people and not others'. Notice also that two deictics
introduce the relative clause [brackets]. Di has anaphoric reference to the town of Antioch; ten
is the cataphoric deictic, which I believe is reintroducing that old information into the discourse
as if it were new.

6.5.4 Topic Continuity

The normal pattern for clauses that do not have a noun subject is for a pronoun to occur. With
my present understanding I interpret absence of a pronoun as linking the action of that clause
more closely with that of the preceding clause than otherwise, as when a culurally predictable
script is being followed. Next, I would maintain that the topicality hierarchy of Givón (1976)
applies to Waris. First person subject pronouns are almost invariably suffixed with the Topic
marker (indicatives and not non-indicatives, that is), as are second and third person pronoun
subjects. However, for personal pronouns marking Benefactive, Goal or Patient, the presence or
absence of Topic suffix is conditioned by the role those referents have had in the preceding
discourse.

(363) ku ka-m-ba vé-ne-v

head 1st-GL-TOP do-BEN.SG-PRS
‘My head aches.’

In this example the first person pronoun is not Subject but Benefactive ('my head is doing for
me'). The high saliency of first person means that it still receives Topic marking. Even pronoun
Patients, a low category in Givón's hierarchy can be Topic marked. But this reflects their role in
the discourse.

(364) ka-va ata ye-m-ba hélvakomandha-vav

1st-TOP will.today 2nd-GL-TOP kill-FUT
‘I will kill you (that's what I'm going to do today).’

(365) ye-m-ba ata ka-va hélvakomandha-vav

2nd-GL-TOP will.today 1st-TOP kill-FUT
‘I will kill you (that's what's going to happen to _you_ today).’

Givón's hierarchy also applies to noun topics.

(366) ka-va ti he-v

1st-TOP tree chop.down.PRS
‘I am chopping down a tree.’

(367) ti-mba ka-va he-v

tree-TOP 1st-TOP chop.down-PRS
‘I am chopping down the tree.’
 111

In the latter example (367), Topic marking on 'tree' means that it has been mentioned before and
is being reintroduced. Furthermore it is being introduced in a 'topic' role, what the speaker is
talking about. Furtheremore it has been moved to the front of its clause to indicate the function
that it plays not in the whole discourse but in just that clause. I call this 'topic' also, which is
probably more confusing than helpful, but I don't know any better term.

(368) ka-va ótól aral-m holvó-ra-ho-i

1st-TOP child father-GL CLASS-get-REC.RP
‘I gave a child to (its) father.’

6.6 Information in Discourse

In this section I survey the way information is organized in discourse. In general it is possible
to identify collateral, evaluative and explanatory information in texts. In Waris such non-event
information is not highly marked syntactically as it is in Bukiyip, for example (Conrad 1988).
What is interesting about Waris is that information like the above can be identified in quoted
speech.

6.6.1 Quoted Speech

Verbs of speaking are widely distributed in Waris discourse. Quoted speech may realize
speech acts, or it may serve other functions. In the following discussion quoted speech realizing
speech acts is treated first, then non-speech acts. All examples are of direct speech unless
otherwise labelled.

6.6.1.1 Collateral and Evaluative Information

In the following example (369) the quoted speech gives collateral information - what might
happen. Example (370) gives evaluative information - how the speakers felt.

(369) ka-va kav-na duang-m u-mni-na

1st-TOP 1st.EMP-GEN whiteman say.PL-BEN.SG-PST
ka-m won-ga-o-e ka Yonki-m nongle-hélvo-i-e
1st-GL ACC.PL-go-IMP-quote 1st name-GL look-go.come.PL-RP-quote
‘We said to our whiteman ''Take us to Yonki; we want to go and see Yonki.''.’

(370) doa kav u-li-na

OK 1st.EMP say.PL-happy-PST
beselangas léhra ka-m-ba u-n-na-ta
very.good good 1st-GL-TOP say.SG-BEN.PL-PST-EMP
 112

‘OK we said happily (to one another) ''Very good!, (it is) good(that) he said
(that) to us!''.’

The following example (371) also gives evaluative or reason information. Note that there is no
verb of speaking, but rather a verb of emotion.

(371) Markus-tatana-mba noinda belhat vi-na

name-group-TOP thus angry do.PL-PST
os-va ai ka-ngas-va sém vi-na-e
thus-TOP my! 1st-alone-TOP how? do.PL-PST-quote
‘Markus's group was angry because of that, like _this_: ''My!, how can (they)
do (this) to us?''.’

The following examples (372) and (373) give evaluative information after verbs of emotion or
sense.

(372) indkumbi-va hilli-vna

everyone-TOP hear.PL-CON
tanam ó-i-e Pita-va
true speak-RP-quote name-TOP
‘Everyone listened (to him) ''Peter has spoken the truth''.’

(373) noinda ó-na Resel-m-ba und ve-vna

because say-PST name-GL-TOP love do-CON
besel hev-m-e
good 3rd.EMP-GL-quote
‘(He) said (that) because he loved Rachael ''(She) is beautiful''.’

6.6.1.2 Highlighted Elements

The following example (374) shows how quoted speech is used to highlight elements in a
discourse about Yonki power station, where the highlighted elements are not people but the
machinery.

(374) ka-m-ba ind ve-uv-mana v-in-vna

1st-GL-TOP show do-CON.PL-around do.PL-BEN.PL-CON
os-m men-ba loh-v os-m men-ba loh-v
thus-GL this-TOP exist-PRES thus-GL this-TOP exist-PRS
‘(They) were showing us all over ''This is for thus, this is for thus''.’
 113

6.6.1.3 Non-Speech Act Quotes

The following examples show how quoted speech serves other purposes than the realization of
speech acts. First is _identification_, example (375).

(375) ótó-va nó-mba os sevri-v bas

auto-TOP that-TOP thus call.PL-PRS bus
‘That auto is called a ''bus''.’

The next example (376) is of _purpose_.

(376) ka-va daswot-ó-mana ve-vna-moa

1st-TOP trick-daily-around do-CON-NEG
indhana-m-ba os-m ó-ra ka-m-oa wive-wol
people-GL-TOP thus-GL say-IRR 1st-GL-RT
vi-ra-e
praise.NS.Odo.PL-IRR-quote
‘I wasn't going around tricking people, saying to them ''As for me, you praise
me''.’

The following example (377) also gives a purpose. But it is typical of similar clauses in that is
not marked with the quote suffix, which leaves its interpretation open: it could be a direct quote
or thought of the person being talked about, or it could be the thoughts of the person speaking.

(377) Yesus-va os-m ve-vna he-oa tau vi-ra

name-TOP thus-GL do-CON 3rd-RT know do.PL-IRR
‘Jesus was doing (something) so that ''As for them, they will know''.’ or 'Jesus
was doing (something) so that as for those (people) they would know.'

The following example (378) shows how comparisons can be made using quoted speech. This is
done by putting the content of the comparison in someone's mouth, although it may not be clear
and is probably irrelevant to the discourse whose mouth. The main syntactic feature of this kind
of quote is the use of recent past tense on the verb of speaking, the verb regularly used in
descriptive discourse. Furthermore, the particle maim 'anyway' may occur, as in example (379),
meaning something like 'people will go so far as to point out the likeness'.

(378) ye-na mundklal-va inungpunung-va vrei-vav

2nd-GEN old.men-TOP dream-TOP get.PL-FUT
ó-i Sesoa-namini-e
say-RP God-ABL-quote
‘Your old people will have dreams ''(They're) like (dreams) from God!''.’
 114

(379) indhana-mba u-vav pai os pho-v

people-TOP say.PL-FUT my! thus
suwul-e maim ó-i moamba-nén-e
arrive-PRS smoke-quote anyway say-RP frog-eggs-quote
‘People will say ''My! (something) is coming like smoke'', (they will even say)
''It's (thick and opaque) as frog eggs!''.’

The following example (380) shows how quoted speech can serve to advance the event line in
Waris narrative.

(380) indhana-mba nungli-v-ra-va u-vav

people-TOP see.PL-PRS-IRR-TOP say.PL-FUT
pai os pho-v tovol-e suwe-e suwul-e
my! thus arrive-PRS blood-quote fire-quote smoke-quote
‘People will see (those things) and say ''My!, (things) are coming up like blood
and fire and smoke''.’

6.6.1.4 Argumentative Discourse

Argumentative discourse is characterized by quoted speech, both that of the speaker and the
speech or thoughts he puts into the mouth of his opponents. The following example (381) is the
beginning of an argument for people to leave their traditional hamlet and settle permanently in
an administrative village.

(381) a deuv-m ka-va ó-v os ka-va néng-vna

house-GL 1st-TOP speak-PRS thus 1st-TOP think-CON
‘I'm talking about the village. This is what I have been thinking.’

b kampong-putoa-va wuleiha-wol ve-v

village-old-TOP leave-NS.O do-PRS
'(That we) leave the old village.'

c ambo Po-ra hui hoanavhó vri-v

just name-LOC here clear do.PL-PRS
'(That we) just clear a site here at Po (ground).'

d [hi-mba os u-v-m]
3rd-TOP thus say.PL-PRS-GL
'They say...'

e ówówos-e ka-m-ba desans v-in-v-e

famine-quote 1st-GL-TOP irritation do-BEN.PL-PRS-quote
''(We will have) no food, we (will) not live well.''.'
 115

f owai pon-ba ve-v-moa

no hunger-TOP do-PRS-NEG
'No (they are wrong), (we will) not experience hunger.'

g Po-ra besowonam aev-v

name-LOC well sit.PL-PRS
'(We) live well (here) at Po ground (the administrative village).'

Notice that the arguments of the opposing side are given as direct quotes just like the thought of
the speaker. Note also that the quote formula of the opponents is marked with Goal suffix on the
verb, like some relative clauses. I would maintain that this indicates given or background
information, with the important new information being the actual content of the speech.

6.6.2 Negative and Positive Information

Native speakers like to have negative information come before positive. Furthermore they like
to use litotes, and there is a standard formula for it, employing the word mani 'what' and the
suffix -ma 'interrogative. Thus besel 'good' forms mani beselma 'very bad'. Sometimes the mani
is omitted and the standard form for 'so distant!' is built on mura 'near': murama.

Rhetorical questions are not a prominent part of discourse.The following example (382) is taken
from translated Scripture.

(382) Yesus hénga u-n-na

name back speak-BEN.PL-PST
hi hev-pa manaema mendekli ka-va ve-na-ma
where 3rd-TOP something big 1st-TOP do-PST-INT
sné-mba an ó-na
such-TOP who speak-PST
‘Jesus said to them in reply ''where is something big I might have done, who
said anything about such a thing?''.’

6.6.3 Backgrounding and Foregrounding of Information

Waris makes extensive use of the word doa in all kinds of discourses. In a typical written
narrative about an unfamiliar subject (a visit to a power station), out of 42 events 27 were
marked with doa. (This reflects the large amount of non-event explanatory information,
including a lot of quoted speech.) In the verb phrase, this word means 'completed action'. But at
the beginning of a clause it means 'return to the event line' after intervening non-events. After
another event, doa seems to mean 'prominent event'. The connective wo 'and' is much less
frequent in texts and seems to have a very neutral effect on prominence. My present
 116

understanding of Waris is that when the speaker wants to put an event in a clearly non-
prominent position, he/she uses the strategy of using only a bare verb stem. This strategy was
first noticed in a narrative about a totally-unfamiliar topic (visit to a power station) which had
many motion verbs. But it was also found in a folktale with many action verbs but little
movement from place to place. So it seems that I need to be looking for another strategy by
which the speaker can background events in general, not just motion events. This may be
connected with the role of the Topic suffix, which occurred on 18 of the 42 events in the above-
mentioned narrative. In my analysis so far, occurrance of the Topic suffix has been assigned the
meaning 'the information in this clause is now viewed as given or presupposed, and the speaker
is using it as the starting point for presenting new information'. Assuming that this definition has
an element of truth in it, it is not clear how the Topic suffix is connected with the idea of
prominent and less-prominent events in the narrative.

(383) doa kav siha-pia-ula-na-mba

OK 1st.EMP enter-complete-happy-PST-TOP
doa Ukarumpa-rini-mba ka-m-ba dambla-na-mba
OK name-ALL-TOP 1st-GL-TOP get.PL.O-PST-TOP
doa pilivoha-ula
OK fill-happy
doa kav a-nga-ula-na-mba
OK 1st.EMP PL-go-happy-PST-TOP
móngó-m kota-m kuvhu
one-GL town-GL pass.PL
‘OK we all happily got into (a bus), OK (it) got us from Ukarumpa, OK (it)
filled up, OK we went along happily, (we) passed a town.’

6.6.4 Time Information

Waris narrative is highly iconic and deviations such as flashbacks are clearly marked. In my
present understanding I view all events as overlapping minimally unless marked with the
Continuous verb aspect, as in the Simultaneous Action Sentence (6.2.3). However there is still
an area to be explained in the Waris view of time, and that is the native speaker view of the
essential duration of events. In other words, there are two tense-mode suffixes used in past
narrative, one more or less punctiliar and the other more or less continuous. I would maintain
now with Litteral that each verb has as part of its meaning the category of 'intrinsic duration'
which basically determines which suffix it will bear. However, at least with some verbs, the
speaker has some liberty to play with the system and impress his/her own meaning. Here are a
few of the verbs that have been found to fall in each category.

Essentially Punctiliar Essentially Continuous

 117

-----------------------------------------------------------

| | |

| _ishóv_ 'to speak' | _andra vev_ 'to work' |

| | |

| _héllev_ 'to listen' | _dallohv_ 'to bathe' |

| | |

| _nin liv_ 'to sleep' | _sangal sevriv_ 'to sing' |

| | |

| _sihav_ 'to enter | _lingillihav_ 'day dawns' |

| (a house)' | |

| | |

| _pihav_ 'to go down' | _indtohamnav_ 'to show |

| | something' |

| | |

| _nonglev_ 'to see' | _vev_ 'day of week was..' |

| | |

-----------------------------------------------------------

The following verbs are essentially punctiliar but the speaker may use them with the Continuous
aspect marker to emphasize the extended duration of the event which they are describing: to
come, to go, people gather together.

6.7 Relative Clauses

In this section I describe the ways relative clauses are marked syntactically and their semantic
functions. Pronouns and deictics can be used to tie the relative clause to the main clause and this
is described.
 118

6.7.1 Word Order and Syntactic Marking of Relative Clauses

Relative clauses, herein abbreviated RCl and written in brackets, almost always follow the
head noun. The following example (384) shows that feature as well as the common strategy of
suffixing the non-present-tense verb with Goal. (An exception which precedes the head noun is
given in (387).)

(384) aral-va hev-na winde di lukuhpiha

father-TOP 3rd.EMP-GEN dog there
ve-na mel-la
push.inside do-PST hole-LOC
[saiuh di li-vna-m]
snake there recline-CON-GL
‘The owner pushed his dog into the hole [where a snake lived].’

The function of the above relative clause is to specify a location and so it copies the locative
adverb there of the matrix clause.

The RCl in the following example (385) is in the present tense and is marked not with Goal but
with Topic on the verb.

(385) yi-mba tau ve-v-moa hona-mba

2nd-TOP know do-PRS-NEG this-TOP
[men-ba ka-va ve-v-pa]
here-TOP 1st-TOP do-PRS-TOP
‘You do not know this (thing) [which I am doing here].’

In the above example (385) every element of the RCl is marked with Topic. This reflects the
Topic marking on the head which it follows, this, which in turn reflects the prominence of this
in the discourse. It might be paraphrased as 'this thing I am doing here (in your sight), you do
not know what it is.' The example before, (384), is taken from a text about a hunter and his dog.
The head noun hole in the matrix clause is not Topic marked and so the RCl identifying the hole
is not marked with Topic either.

The above example (385) also exemplifies the common strategy of introducing a RCl with a
demonstrative pronoun men-ba 'this thing'.This is a cataphoric deictic and I explain its use in
RCls as bringing something to the attention of the hearer which they were not thinking about in
the way the speaker intends them to think. In other words it is introducing new information. In
the following example (386) a combination of anaphoric deictic and cataphoric deictic are used.

(386) yi-mba tit ve-v

2nd-TOP ignorance do-PRS
 119

[hona-mba men-ba yi-mba ka-m-ba ósah-man-i-va]

this-TOP this-TOP 3rd-TOP 1st-GL-TOP ask-BEN.SG-RP-TOP
‘You don't understand what you have just asked me for.’

The following sentence (387) is the only case where the RCl has been found to precede the head
noun.

(387) [wondoa-va péthemona-m ga-i-va]

bird-TOP ground.path-GL go-RP-TOP
[winde men wa-wol vi-v-m]
dog here cry-NS.O do.PL-PRS-GL
os moa-mba hi-mba ó-wol vi-v
thus talk-TOP 3rd-TOP speak-NS.O do.PL-PRS
‘[When the _wondoa_ bird goes along the ground], [(like) a dog cries], thus he
cries.’

In (387) the last clause is the matrix clause of the preceding RCl. The first RCl is marked with
Topic on the recent past tense verb and gives the characteristic circumstances under which this
bird sings. It is translated with 'when'. The last (matrix) clause is the speaker's comment about
what the bird does as it runs along the ground. The head in the last clause which the second
RCl precedes and refers to is probably not the noun talk but rather the manner adverb os 'thus'
meaning 'in the manner a dog cries, thus the wondoa cries'.

The following example (388) is a fragment of a list of people and contains two RCls one
embedded in the other.

(388) eva móngó Simon

and.too another Simon
[hi-mba seilm doara móngawulm lovah-vna-m nó-mba
3rd-TOP first before together stand.PL-CON-GL those-TP
indhana-mba
people-TOP
[ten-ba Rom duang-m engilpingi-lm ve-v-m]]
there-TOP Rome official chase.out-TEL do-PRS-GL
‘and too another Simon, who previously lived with those people who want to
throw out the Rome officials.’

My present understanding of Waris topic marking and RCls leads to the following analysis. The
subject of the first RCl is immediately recoverable from the context and so a personal pronoun
is used rather than a deictic. It is marked as Topic because it is old information and the topic of
its clause. In the second (embedded) RCl the subject is a cataphoric deictic pronoun _ten-ba_
'that one' which is gives prominence to the information that those people were characterized by
 120

a desire to throw out the Romans. It is Topic marked because it is old information, its referent
'those people' having been mentioned in the preceding clause, and it is the topic of its clause.

The above RCls have exemplified identification, location and circumstance. Following is an
example (389) with two RCls giving _reason_ information.

(389) ata kuskus-m nongla-ka-vav

will clerk-GL look-go.and.come-FUT
[téh ka-na ten dambo-na-mba]
firewood 1st-GEN there get-PST-TOP
[téh-ram ka-m di-m dembra-ho-lm ó-na-mba]
firewood-ALL 1st-GL money-GL get-REC-TEL say-PST-TOP
‘(I) will visit the clerk, who got my firewood there, who told (me) he would
pay me for the firewood.’

Embedded purpose clauses are marked differently from other RCls.

1. The are not introduced with a deictic.

2. The verb is marked with irrealis or with Telic.

Clauses like this have been discussed in section 6.2.7.

6.7.2 Types of Relative Clause

Subject and Locative are commonly relativized on in Waris. Relativization on Object is less
common, and no example has yet been found of relativization on Indirect Object. Following
(390) is an example of Instrument being relativized, but note that within the RCl there is no
Instrument marking, the semantic relationship has to be inferred.

(390) doa pholahanga-na kain-nahonam

OK escape-PST cloth-with
[ten némét ninge-wol vi-mni-na-m englamóngla-m]
there first tie-NS.O do.PL-BEN.SG-PST-GL arms.and.legs]
‘OK (he) came out (of the tomb) together with the cloth, (with which) (people)
had previously tied (his) arms and legs.’

6.8 Cohesion
In this section I will give an overview of the the topic of cohesion in Waris. I will discuss
mainly those features that strike me as being peculiar to Waris.
 121

6.8.1 Items Having Diphoric Reference

The grammatical items that join clauses into sentences fall in this category, namely
conjunctions like _heva os_ 'thus, but', _noinda_ 'at that, because' and the continuous aspect verb
suffix -vna. These items join the clauses that make up Simultaneous Action, Counter-
expectation, Reason or Result and Alternative Sentences (6.2.3, 6.2.4, 6.2.5, 6.2.6). They do this
by standing between the clauses and pointing both backwards and forward. Since the verb suffix
is part of one clause it may be stretching the definition of cataphora to say that they point 'back'
to the clause in which they occur; but the above conjunctions clearly belong to neither the
preceding or following clause and point to both.

The word doa 'next prominent event' also points both forward and backward, (except when it
occurs in the beginning clause of a discourse).

The manner adverb os 'thus' seems to have exclusively the function of cohesion in Waris
discourse. The semantic meaning it adds is very little. It commonly occurs after verbs of
speaking and thinking and points forward to the content. Other times it points back. Both uses
are exemplified in (391).

(391) dawonam ka-sna boasna telifon-da saho-wol

later 1st-first first telephone-LOC ask-NS.O
ve-nga-v-ra
do-go-PRS-IRR
wo ye-m-ba os ka-va u-n-vav os-va
and 3rd-GL-TOP thus 1st-TOP speak-BEN.PL-FUT thus-TOP
nósim pi-mba a-nga-vav
certain.day 1st.INC-TOP PL-go-FUT
‘Wait, first I will telephone and ask (them) and I will thus tell you(pl) thus:
''We(incl) will go on a certain day''.’

In this example the first occurrance of os (second clause) points back to the first clause. Maybe
it means something like 'information in this clause is in temporal sequence to information in the
preceding clause'.

The second occurrance of os is marked with Topic suffix and points forward to the content of
the quote. I would maintain that the Topic marking points to the content of the quote as being a
restatement of the theme of the whole paragraph, which is 'we asked (Jeff Bailey) to take us to
Yonki'.

Since analysis of the word os is still somewhat unclear I will give a few more examples.

(392) indhana-mba moa-mba i-lvah-vna-moa

people-TOP talk-TOP CARRY-stand.PL-CON-NEG thus-TOP
 122

os-va noinda
so that
ye-m-ba koasromb vi-v-ra
3rd-GL-TOP cross do.PL-PRS-IRR
‘People have not had any grievence so that thus they could be angry with you.’

Again, I would maintain that the Topic marking on os points to the discourse role of the content
of the result clause, which happens to be the topic of other people being angry with the
addressees. In the following example (393), os is case marked with Goal, as the Object of the
verb 'to do'.

(393) os-m men-ba yi-mba ve-i

thus-GL here-TOP 3rd-TOP do-RP
‘This thing you have just done thus.’

In context 'thus' points back to an action of the addressee, which the speaker is reintroducing,
using the cataphoric deictic 'this thing'. By using the cataphoric deictic, which is normally used
to introduce new information, the speaker indicates that he is using the action of the addressee
as a point of departure for further discourse [God pointing out Adam's sin to him]. In the
following example _os_ is marked with the Resumptive topic marker:

(394) a ambo hi-mba ó-wó-vna

just 3rd-TOP speak-daily-CON
‘(He) just kept saying..’

b sen-peta ka-m vrei-tuk-e

coin-little 1st-GL get.PL-IMP-quote
''(You people) give me some coins!''.

c os-oa dombo-na
thus-RT get.S.SG.PL.O-PST
'Thus he got them. = He got them just that way.'

In example (394) the os points back to the action of the first clause, a beggar asking for money.
The use of Resumptive Topic marking means something like 'as for that previously-given
action, it had the following sequel'.

6.8.2 Items Having Mainly Anaphoric Reference

In narrative discourse, a bare verb stem points back to the preceding finite verb, which usually
is also marked for Topic as well.

(395) tendórini pró-na-mba avhó

man come-PST-TOP sit.down
 123

‘A man came and sat down.’

My present understanding is that the bare stem encodes an action that is predictable and
probably contemporaneous with the action of the preceding verb. and to which the speaker
wants to give low prominence.

A recapitulated verb also points back to the preceding clause.

(396) doa mie-m e-wole

complete pig-GL DL-shoot
mie-m e-wola-na-mba
pig-GL DL-shoot-PST-TOP
e-puise
DL-cut.up
e-pusa-na-mba
DL-cut.up-PST-TOP
‘OK (we) two shot a pig. Having shot a pig (we) cut it up. Having cut it up...’

The inverse of the above anaphoric verb reference is the cataphoric reference of the topic
marking on the verbs, which point forward to the action of the next clause. As in many Papuan
languages a dummy verb may be employed. I think this is done to create a varied style or
perhaps slow down the rate of information.

(397) doa dawonam indowonam andra ve-uv-vna

OK later happily work do-daily.PL.S-CON
ve-na-mba
do-PST-TOP
doa kav-na duang hénga mungnasim ishu-n-na...
OK 1st.EMP-GEN white.man next one.day said-PL.BEN-PST
‘OK (we) worked along happily for a period, having done so, OK next our
white man one day said to us ''..’

The third person emphatic pronoun hev 'em yet' generally points back in a discourse to an
antecedent.

Deictics including demonstative pronouns like nói 'that one' and demonstratives like nó 'that'
point back in a discourse to previously-given information.

(398) nó-mba tendórini-va doara pró-na

that-TOP man-TOP previously come-PST
That man already came.’

In the above example (398) the use of nó means that the man being referred to is given
information in the discourse and potentially retrievable. The use of 'that man' rather than the
 124

shorter form nói 'that one' means that the speaker feels the need to stir the memory of the
hearer, there being the possiblity that 'that one' will not be enough of a stimulus. Furthermore,
the Topic suffix on both elements of the NP 'that man' means that this item is not merely old
information being repeated, it is being reintroduced into the discourse as the topic or what the
speaker wants to talk about again/still. Thus the Topic suffix is exhibiting anaphoric reference.

When the Topic suffix is being used to mark Chafe's point of contrast (5.4.3.4), it also exhibits
anaphoric reference.

(399) kapol-va déti-va owai simera pró-vav

plane-TOP today-TOP no tomorrow come-FUT
If the airplane doesn't come today, it will come tomorrow.’

Topic marking on 'today' points back in the discourse to the question that had been raised, if a
certain airplane would come on that day.

6.8.3 Items Having Mainly Cataphoric Reference

As mentioned above, the Topic suffix on the first verb of a head-to-tail linkage points forward
to the next clause. It means something like 'information in this clause is given or presupposed
now' and I usually translate it as 'having...'

The Topic suffix on the protasis of a conditional sentence (6.2.1) also means 'presupposed
information' and points forward to the next clause. (See sections 6.2.2 and 6.2.11.1.) The irrealis
verb suffix on the protasis likewise performs the same function.

The locative adverbs men 'here' and ten 'there' (2.2.3) and their related demonstrative pronouns
menba 'this one' and tenba 'that one' (2.3.7) precede new information in the same clause.

(400) men-ba tendórini

here-TOP man
This is a married man.’

(401) men-ba tendórini mie-m ló-i

this-TOP man pig-GL shoot-RP
This is a man who shot a pig (today).’

Other deictics such as himba nómba 'whoever' also point forward to another clause.

(402) hi-mba nó-mba ga-lm-ba ve-v-ra-va

3rd-TOP that-TOP go-TEL-TOP do-PRS-IRR-TOP
hi-mba nó-mba hui kusó-v-ra
3rd-TOP that-TOP here gather.PL-PRS-IRR
‘Whoever wants to go, they need to gather here.’
 125

The filler of a semantic topic slot at the beginning of a discourse can be conceived of as
pointing forward to what is coming up in the story.

6.8.4 Locative Viewpoint

An important rule in Waris conversation which is needed to interpret discourse is that of

locative viewpoint. The main rule is that the viewpoint of the _addressee_ is taken by the
speaker, not the viewpoint from the speakers own location. This determines the form of verbs of
motion in discourse as is seen in the following conversation, in which a group of people leaving
a village for a lower place state their intention to a person who intends to follow them later,
example (403). Examples (404), (405), (406) are replies that person might make to them.

(403) ka-va nongend-v

1st-TOP descend.valley.PL-PRS
‘We are going down (to the mission).’

(404) ye-oa mi nongend-vm *ata ka-va sis

2nd-RT first desc.val.PL-IMP will 1st-TOP behind
nonga-vav
desc.val-FUT
‘You(pl) go down first, *I will go down later after you.’

(405) ye-oa mi nongend-vm ata ka-va sis

2nd-RT first desc.val.PL-IMP will 1st-TOP behind
pró-vav
come-FUT
‘You(pl) go down first, I will come later after (you).’

(406) ye-oa mi nongend-vm

2nd-RT first desc.val.PL-IMP
ata ka-va sis honó-vav
will 1st-TOP later arrive.from.above-FUT
‘You(pl) go down first, I will arrive there afterwards, from above’

The starred clause is unacceptable because the speaker, in using the verb (I) will go down, is
maintaining his own viewpoint. The next two examples are acceptable, in which the speaker
takes the viewpoint of the addressees, who will later be below, watching him arrive from above.

6.9 Types of Discourse

In this section I discuss the kinds of discourse found in Waris and their major features.
 126

6.9.1 Narrative

Narrative falls into two categories, oral and written, with significant differences. Oral narrative
may be either first or third person and both classes show considerable use of head-to-tail verb
recapitulation. Written narrative in Waris is so far restricted to first person texts by native
authors, or translated Scripture. In both of these verb recapitulation is greatly reduced.
Furthermore, use of pronouns and nouns is increased over that in oral texts, to increase
precision of reference. (Oral narrative so far recorded is mainly about familiar subjects.)

6.9.2 Folktales

Many folktales have been recorded and edited to use in literacy materials. By this I mean
traditional stories that are typically told to children. Each village seems to have its own
repertoire, with some overlap. These typically have non-human actors in them, such as the moon
or a cassowary, and are viewed mainly as entertainment. They typically embody cultural values,
however, such as generosity, and in that sense are very 'true'. The only grammatical difference
between this genre and oral narrative is that folktales each have a formulaic title. Like most
other oral narrative they are familiar to the hearers and so tellers use less precision of reference:
fewer pronouns and nouns.

6.9.3 Descriptive

This genre is grammatically obvious because main events are in the third person and marked
with the recent past tense (or occaisionally with the present tense). In narrative, the recent past
tense refers to events that took place in the past few hours, typically since the preceding night.
But in a cultural description, as of a ritual, the effect is one of vividness and can be compared to
the use of present-tense verbs in an English first or third person narrative, or to the use of
present-tense verbs in Koiné third person narrative. When used in descriptive text the recent
past tense conveys the meaning 'what is customarily done'. Descriptive discourse typically deals
with the way rituals are performed or artifacts are produced. Verb recapitulation is a major
syntactic feature of this genre. Following is an example taken from a text describing the fertility
ritual called wevti:

(407) a ai doa nénga-m wevti-néngand-m

well OK think-GL wevti-decoration-GL
‘Well let me think, yes, about the wevti decorations.’

b si-mba eunumbul-va wul-vhoavo-i

night-TOP part-TOP carry-leave.PL-RP
'When it is still dark (men) (habitually) leave (the village).'
 127

c doa néngand-o-mana ve-v

OK get.decoration.PL.S-CONT-around do-PRS
OK they (habitually) get decorations from around (in the bush).

Descriptive text is found in another context too, as paragraphs set in another discourse type,
narrative. In a narrative decribing a totally unfamiliar subject, a trip to Yonki power station,
descriptive paragraphs of what was seen there are marked by a mixture of present, recent past
and past tenses. Some sentences are without agents, the number just being marked on the verb.
This means that in a descriptive text people are deemphasized and processes are emphasized.

(408) wo nó-mba po-mba kleihal-saiv-na-mba

and that-TOP water-TOP join-enter.PL-PST-TOP
paiv mendekli-na
pipe big-GEN
ve-na-mba di me-ra [masin-da me-ra-va
do-PST-TOP there inside-LOC machine-LOC inside-LOC-TOP
di-va sambla-móngó wosei-na]
there-LOC two-one put.PL.S.PL.O-PST
nó-mba po-mba umharavhó vré-v-pa
that-TOP water-TOP turn do-PRS-TOP
‘That water, (workmen) brought it in by big pipes. (They) having done so,
inside there, inside the machines of which (they) have installed three there, that
water turns.’

6.9.4 Hortatory

This type of discourse has been infrequently recorded. (Culturally, the 'pep talk' is not
important.) Grammatically it is not distinctive except that irrealis verbs are used to convey the
meaning of optative or polite imperative: 'may people do this, let people do this'. Vernacular
sermons fit into this pattern, and contain many inclusive references 'we (incl) should do this'
When the appeal is based on Scripture, it generally goes like this: 'since Scripture says this, we
(incl) should do so and so.'

6.9.5 Argumentative

This type of discourse was recorded only once, but it has a very distinctive semantic content
and pragmatic organization. Each paragraph has the following speech acts and adversative: 'I
say ''we (incl) should do so and so'' but (my opponents say ''we (incl) must do such and such'',
but no, they are wrong for the following reason'. See example (381).
 128

6.9.6 Letters

This is not a distinctive discourse genre in Waris since almost all letters are written in Tok
Pisin, and because translation of New Testament letters has not produced any distinct form as
yet. The few vernacular letters seen are, like their Tok Pisin counterparts, invariably asking for
the reader to do something for the writer, typically give money.

6.9.7 Songs

The Waris song repertoire is severly restricted. One kind is wóngsangal 'drum song', in which
Tok Pisin songs of the type heard on the radio are sung to the accompaniment of hand drums.

In the healing ritual wó lóv , wó (spirit?) shoots (someone?)', magical incantations are sung in
the vernacular to the beat of hand drums.

In the final category wevti isv ,a fertility ritual, songs are sung by the spectators while men
impersonating cassowaries and bush spirits dance to the tune of wooden horns. Some of these
songs have been transcribed. Grammatically and sematically they fall into none of the above
categories of discourse. Instead they are group expressions of pleasure, as in 'let the airplane
come, then we'll go to the store and buy tinned meat', or expressions of sadness, as in 'Wuse
friend, we are sorry you are dead'. Wevti songs seem to be about the only part of Waris culture
where people can exhibit self-expression (by composing new ones).

All the wevti songs recorded so far are sung to the same tune and display the same structure:
line A line B line A line C line A. Here is an example.

(409) kukul owo-v-e, kukul owo-v-e,

grunt say-PRS-EMP grunt say-PRS-EMP
mama mesala o-v-e
? squeal say-PRS-EMP
Line A: 'A pig grunts loudly, pig grunts loudly, pig squeals.’

(410) ye-m-ba Moses kovha-i-ma,

2nd-GL-TOP name shoot-RP-Q
ye-m-ba Muesp hlivha-i-ma
2nd-GL-TOP name shoot-RP-Q
Line B: 'Did Moses just shoot you? Did Muesp just shoot you?

(411) ye-na nambsa ting-va wola-v-ra,

2nd-GEN sago.type stump-TOP exist-PRS-IRR
ye-na yua ting-va pil-v-ra
2nd-GEN sago.type stump-TOP exist-PRS-IRR
 129

Line C: 'Your _nambsa_ sago stump exists, your yua sago stump exists (where
pigs like to feed and can be hunted).

6.9.8 Conversation

Waris dialogue has not been studied, except the topic of locative viewpoint, which is discussed
in section 6.8.4.
 130

7 Bibliography

Brown, Robert 1981 Semantic Aspects of Some Waris Predications in Karl Franklin, ed. Syntax
and Semantics in PNG Languages, SIL Ukarumpa

" " 1985 Waris Case System and Verb Classification later published in LLM 1988
19.37-80

Bruce, Leslie P., Jr. 1979 A Grammar of Alamblak, Ph.D. dissertation Australian National
University, later published in PL 1984 C-81

Chafe, Wallace 1976 Givenness, Contrastiveness, Definiteness, Subjects, Topics, and Point of
View. In Charles N. Li, ed.

Comrie, Bernard 1981 Language Universals and Linguistic Typology

Conrad, Robert J. 1987 Kinds of information in Bukiyip oral narrative discourse, LLM 16:23-40

Cook, Frances 1971 Case Grammar Georgetown University Press, Georgetown

Dik, Simon C. 1978 Functional grammar, Amsterdam

Fillmore, Charles 1977, The case for case reopened, in Syntax and Semantics vol. 8 ed. Cole
and Sadock, Academic Press, NY

Foley, William A. 1986 Papuan Languages of New Guinea, Cambridge University Press,
Cambridge

Givón, Talmy 1976 Topic, Pronoun, and Grammatical Agreement. In Charles N. Li, ed.

Haiman, John 1978 Conditionals are topics, Language 54:564-589

Haiman, John 1987 Review of Seiler [Imonda, a Papuan Language, PL B-93] in LLM vol
16 no 1-2 pp. 136-142

James, Dorothy 1982 Verb serialization in Siane. Manuscript, SIL Ukarumpa, PNG

##" 1983 LLM vol 14 no 1-2 pp. 26-75

Li, Charles N., editor Subject and Topic. New York, Academic Press, 1976
 131

Loving, Richard and Jack Bass 1964 Languages of the Amanab Sub-District, DIES Port
Moresby and SIL Ukarumpa

Seiler, Walter 1984a The main structures of Imonda-a Papuan language. Ph.D thesis. Australian
National University

###" 1984b Imonda part-of-whole marking, Pacific Linguistics, Series A No. 69, pp. 273-283

###" 1985 Imonda, a Papuan Language, PL B-93

Wurm, Stephan A. 1982 Papuan langauges of Oceania, Tubingen, Narr (Ars Linguistica)