A8llc18 Ol llNoO okAmmAk

by
lALl LNolk
A 1ll8l8 8Llml11lu lN lAk1lAl lLlllllmlN1 Ol
1ll klQLlklmlN18 lOk 1ll ulokll Ol mA81lk Ol Ak18
in
1ll lAcLl1Y Ol okAuLA1l 81Lull8
llNoLl81lc8
we accept this thesis as conlorming to the required standard
...............................................................................
}. kandolph kadney, lh.u., 1hesis 8upervisor
................................................................................
lmma lavey, lh.u., 8econd keader
...............................................................................
lonnie lenson, lh.u., 1hird keader
1klNl1Y wl81lkN LNlVlk8l1Y
9 April 2008
C laul Lnger
Al81kAc1
lengo (language code [lgr]), an Austronesian Oceanic language, is a member ol
the oela branch ol the 8outheast 8olomonic subgroup. lengo is spoken by some 14,000
people ol north and east-central ouadalcanal province, 8olomon lslands. while there
are briel relerences to lengo in the literature, there is no comprehensive treatment ol
the grammar. 1his thesis begins to lill the void ol studies ol lengo in particular and ol
ouadalcanal languages in general.
1his thesis presents an overview ol some ol the more readily evident leatures ol
lengo grammar. lollowing a section presenting the phonology ol lengo, there are
sections covering word, phrase, clause, and sentence syntax. ltems ol particular interest
in this grammar are the case ol vowel raising and harmony in certain third person
singular inalienably possessed nouns, a distinct set ol cardinal numerals lor custom"
(lijin kastom) stories, tense J aspect J mood prelixes which join to subject relerence
pronouns (including one lor a mood described as 'apprehensive'), processes by which
verbs derive lrom nouns and nouns derive lrom verbs, a particle analysed as a 'realis
locative', and a discussion ol lengo's various serial verb constructions.
Along with a presentation ol lengo's geographic and lamilial situation in the
8outheast 8olomonic language lamily and a briel discussion ol orthography issues as
appendices, a selection ol texts lrom my lieldwork in ouadalcanal (c. 2002-2006) are
included. lour ol those included here were recorded and transcribed with the help ol
lengo speakers. 1he lilth was given to me in hand-written lorm.
ii
AclNOwlluolmlN18
A project like this is never the product ol one person. 1hose who helped deserve
credit and thanks.
l thank my lengo lriends who took the time to talk, had the patience to explain,
and had the grace to gently correct as l learned their language.
l thank my supervisor and readers who saved me lrom many mistakes and
misunderstandings, and opened up ways ol looking at and thinking about lengo that l
had not previously explored.
l thank my lriends catherine, }ong-loon, and }onathan lor their example ol
quiet perseverance and many encouragements.
l thank my lamily lor helping me throughout: my children lannah (who came
home with her six-year-old share ol lengo words learned while playing with her
lriends), lsaac (whose boundless energy restored mine at the end ol many long days),
Nathan (l hope you have a sale ride to school on the motorcycle, uaddy"), Aaron (who
was sorting out lnglish grammar while l was sorting out the lengo), and kachael (who
was brought into the world about when this thesis was), and my wile catharine
without whom not.
iii
AllklVlA1lON8 ANu O1llk cONVlN1lON8
1he abbreviations used in this grammar are given below.
A agent llm limiter (po)
Alll abilitive (tangomana) lOc locative (i, tena)
AuV adverb N noun
Al1 alternative (pa, p-) Nlo negative
Allk apprehensive (b-) Nk nominaliser
Ak1 article (a, na) OJO object
cAL8 causative (-caghini) lAL paucal (tu-, -tu-, -tu)
cll classilier llc1 perlective (g-)
cOm comitative (kolu) ll plural
cOml complementiser (thi, tha) lOc lrotoOceanic
cONu conditional (ghua) lkll preposition
cON} conjunction (ma, m-) l8 possessor pronoun J person
uA1 dative (vani) lLkl purpose (ne)
ulm demonstrative (deni, deri, dini) k recipient
ulO deontic (ba) kul reduplication
ul8 desiderative (ngao) klcl reciprocal (vi-)
ulk directional klll rellexive (tibo)
uO direct object kll relative pronoun (thi, tha)
uL dual (ko-, -ko-, -ko) kl realis (t-)
lmll emphatic (ne) 8 subject
ll emphatic pronoun 8o singular
lX exclusive s.o. someone
lL1 luture (ba) sp. species
olN genitive (ni) 8kl subject relerence pronoun
lmll imperlective (bo) s.t. something
lN inclusive 1 theme
lN81 instrumental (ghini) 1Am tense J aspect J mood
lN1 interrogative 1k transitiviser (-ci)
lN1} interjection V verb
lN18 intensilier (koto, ne) V
i
intransitive verb
lO indirect object V
t
transitive verb
lkk irrealis (k-)
1his thesis was typed using the oentium lont.
1
1he text ol languages other than
lnglish is rendered in italics. loldlace type is used in lengo text in an attempt to
highlight the leature(s) under consideration in a given example. orammatical glosses
1
oentium was used lor the simple reason that 1imes New koman does not have a single-story 'u' nor
an open-bowl 'q', both ol which are needed to express phonetic data in lengo. 1he oentium lont is
lreely available at: http:JJscripts.sil.orgJoentium.
iv
are presented in 8mAll cAl8 lollowing, as nearly as possible, the abbreviations lor
interlinear morpheme translation appended to the leipzig olossing kules.
2
lnterlinear
examples have a lengo line (with morpheme breaks indicated with the appropriate
punctuation), a gloss line, and a lree translation line.
lengo ma k-ami soni iti-i vati tuthivo-gu inau
oloss cON} lkk-1lX.ll throw up-O:!ll journey's.end waterlront.landing-l8:18o ll:18o
lree 'at journey's end we may throw [the sticks] up at my own water-lront landing'
lt is important to note that while emphatic (ll), object (O), and possessor (l8) pronouns
are indicated with abbreviations belore the appropriate person and number (e.g., inau
ll:18o), the subject relerence pronoun is indicated with person and number only. lrom
time to time a lourth line is included. 1his is to indicate a morphophonemic process, as
with teme 'lather-l8:!8o' below, or elision (e.g., teigha na te'na, deni de). ln these
examples, the morpheme breaks andJor lull words are indicated on the second line.
lengo a J. te dea kolua a teme i leo ni pono
morpheme a } t-e dea kolu-a a tama-a i leo ni pono
oloss Ak1 } kl-!8o go cOm-O:!8o Ak1 lather-l8:!8o lOc inside olN bush
lree '}. went with his lather to the bush'
when necessary square brackets [ ] are used in the lree translation line to indicate non-
overt items (such as 28o 'you' in imperative sentences) and other extraneous
inlormation. An asterisk (*), unless otherwise noted, marks ungrammatical sentences.
linally, lor reasons ol privacy, the proper names ol persons and places are represented
by a capital letter in lengo examples. l hope that this helps, and doesn't distract, the
reader.
2
1he leipzig glossing rules are available at http:JJwww.eva.mpg.deJlinguaJresourcesJglossing-
rules.php, while http:JJwww.unm.eduJ-wcroltJlapersJ1ypAbbrev.pdl provides a somewhat larger
set ol abbreviations. l have created some ol my own abbreviations as they were not in either list.
v
1abl e ol contents
Al81kAc1..........................................................................................................................................ii
AclNOwlluolmlN18..................................................................................................................iii
AllklVlA1lON8 ANu O1llk cONVlN1lON8...........................................................................iv
1 lN1kOuLc1lON.............................................................................................................................1
2 llONOlOoY...................................................................................................................................4
2.1 lhonemes...............................................................................................................................4
2.1.1 consonants....................................................................................................................4
2.1.2 Vowels.............................................................................................................................
2.2 lhonotactics..........................................................................................................................6
2.2.1 8yllable structure.........................................................................................................6
2.2.2 llision and contraction...............................................................................................8
2.! morphophonemics...............................................................................................................9
! NOLN8, lkONOLN8 ANu NOLN llkA8l8.............................................................................12
!.1 Nouns....................................................................................................................................12
!.1.1 uerivation and inllection..........................................................................................1!
!.1.2 Noun classes and articles..........................................................................................1
!.2 lronouns..............................................................................................................................2
!.2.1 lmphatic......................................................................................................................29
!.2.2 8ubject relerence........................................................................................................!!
!.2.! Object............................................................................................................................!8
!.2.4 uirect possessor..........................................................................................................40
!.2. lndirect possessor.......................................................................................................41
!.2.6 kellexive.......................................................................................................................4!
!.2. lnterrogative J relative..............................................................................................44
!.! lasic noun phrase structure.............................................................................................4
!.!.1 lossession....................................................................................................................46
!.!.2 Numerals and number-marking..............................................................................6
!.!.! Nominal modiliers......................................................................................................64
!.!.4 uemonstrative.............................................................................................................66
4 Vlkl8 ANu Vlkl llkA8l8......................................................................................................4
4.1 A-type and L-type verbs....................................................................................................
4.2 Verb derivation and inllection.........................................................................................
4.2.1 uerivation....................................................................................................................
4.2.2 lnllection......................................................................................................................81
4.! Valence changing constructions......................................................................................84
4.!.1 causative -caghini.......................................................................................................84
4.!.2 keciprocal vi-...............................................................................................................8
4.4 Verb phrase structure........................................................................................................86
4. lre-verb particles...............................................................................................................8
4..1 1Am prelixes..................................................................................................................8
4..1.1 lerlective g-........................................................................................................9!
4..1.2 kealis t-.................................................................................................................9
4..1.! Lnmarked ........................................................................................................98
vi
4..1.4 lrrealis k-............................................................................................................100
4..1. Apprehensive b-...............................................................................................10!
4..2 1Am auxiliaries............................................................................................................104
4..2.1 luture ba............................................................................................................104
4..2.2 lmpossible luture boro.....................................................................................10
4..2.! ueontic ba..........................................................................................................109
4..2.4 lmperlective bo.................................................................................................110
4..2. continuous ghe.................................................................................................11!
4..! conditional ghua.......................................................................................................114
4..4 lmphatic ne...............................................................................................................116
4.6 lost-verb particles............................................................................................................116
4.6.1 uegree koto J po.........................................................................................................11
4.6.2 kealis locative ti........................................................................................................119
4. Verb negation....................................................................................................................1!!
4.8 Verb serialisation..............................................................................................................1!
4.8.1 uirectional.................................................................................................................1!8
4.8.2 8equential..................................................................................................................14!
4.8.! causative....................................................................................................................14
4.8.4 manner.......................................................................................................................146
4.8. Ambient......................................................................................................................14
4.8.6 lrepositional verbs..................................................................................................148
4.8.6.1 comitative kolu.................................................................................................149
4.8.6.2 uative vani.........................................................................................................10
4.8.6.! lnstrumental ghini............................................................................................11
4.8. modal..........................................................................................................................14
4.8..1 Abilitive tangomana..........................................................................................1
4.8..2 uesiderative ngao..............................................................................................16
4.8..! Negative teigha..................................................................................................16
4.8..4 lrohibitive tabu................................................................................................18
4.8.. Non-volitive kou...............................................................................................19
clAL8l 81kLc1Lkl.................................................................................................................161
.1 Verbless clauses.................................................................................................................162
.1.1 lquative......................................................................................................................162
.1.2 lossessive...................................................................................................................16!
.1.! lnterrogative.............................................................................................................16!
.2 Verb clauses: core arguments........................................................................................164
.2.1 lntransitive................................................................................................................16
.2.2 1ransitive...................................................................................................................166
.! Verb clauses: peripheral arguments.............................................................................1
6 lmllkA1lVl ANu lN1lkkOoA1lVl 8lN1lNcl8................................................................181
6.1 lmperative and cohortative sentences.........................................................................181
6.2 lnterrogative sentences...................................................................................................18!
6.2.1 lolar questions.........................................................................................................184
6.2.2 leading questions.....................................................................................................18
6.2.! content questions....................................................................................................18
vii
cOmlllX 8lN1lNcl8..............................................................................................................189
.1 coordinate clauses............................................................................................................189
.1.1 8imultaneous.............................................................................................................189
.1.2 8equential..................................................................................................................191
.1.! Alternate....................................................................................................................19
.2 8ubordinate clauses..........................................................................................................19
.2.1 complement..............................................................................................................19
.2.2 kelative.......................................................................................................................198
.2.! lurpose.......................................................................................................................200
.2.4 keason.........................................................................................................................202
.2. conditional................................................................................................................20!
8 kllOk1lu 8lllcl....................................................................................................................206
9 cONclL8lON..............................................................................................................................209
AlllNulX A: 8outheast 8olomonic language lamily............................................................212
AlllNulX l: Orthography.........................................................................................................214
AlllNulX c: 8ample texts.........................................................................................................216
8OLkcl8........................................................................................................................................22
viii
li gures
ligure 2.1: consonants....................................................................................................................4
ligure 2.2: Vowels.............................................................................................................................
ligure !.1: lronouns......................................................................................................................2
ligure 4.1: 1Am continuum.............................................................................................................92
ligure 4.2: 1Am tree..........................................................................................................................9!
ligure 4.!: 1Am indications ol k-..................................................................................................101
ligure 4.4: 1ypes ol lengo serial verb constructions............................................................1!
ligure 9.1: Orthography options...............................................................................................214
lll ustrati ons
lllustration 1: lengo language area...............................................................................................1
lllustration 2: lolar question pitch track.................................................................................184
lllustration !: leading question pitch track............................................................................18
lllustration 4: lengo language neighbours..............................................................................21!
ix
1 lN1kOuLc1lON
lengo (language code [lgr]) is an Oceanic language that belongs to the central-
lastern 8outheast 8olomonic branch ol the vast Oceanic language lamily tree. lengo is
spoken by some 14,000 people on north and east central ouadalcanal, 8olomon lslands.
lllustration 1: lengo language area shows the region ol ouadalcanal that is
predominantly lengo speaking.
Although the socio-political unrest ol mid-2000 and the altermath
!
have changed the
situation somewhat, there is a signilicant population ol other language speakers
employed at the various palm oil plantations spread throughout the lertile ouadalcanal
!
A helplul overview ol the ethnic tension" is provided by lraenkel, }on. 2004. The Manipulation of
Custom: From uprising to intervention in the olomon !slands. canberra: landanus looks.
1
!llustration ": #engo language area
8OlOmON l 8lANu8
= lengo
ouadalcanal
plain. 1he long term ellect this will have on lengo remains to be seen. 1here are lour
recognised dialects ol lengo: Aola, laripao, keko, and 1asimboko (oordon 200:6!). ln
my experience, the dillerences are in terms ol pronunciation and not morphological or
syntactic variation.
According to the lthnologue map ol 8olomon lsland languages (oordon
200:884),
4
lengo is bordered by ohari to the west, malango to the south west, lirao to
the south, and longgu to the east. lnterestingly, lughotu (southern 8anta lsabel) is held
to be more closely related linguistically to lengo than the immediate neighbour longgu
(a malaitan isolate). lengo's closest linguistic neighbour is oela, some !0 km across
8ealark channel to the north.

language examples, except as noted, are lrom my data. 1his is comprised ol

textsoral and writtenand some additional elicited material. 1he texts were collected
in Aola ward ol ouadalcanal, though some samples came lrom last 1asimboko and
laripao wards, over a lour year period lrom 2002-06. lt should be noted that the data
were not collected with the express purpose ol constructing a grammar, in lact
somewhat dillerent data would have been collected il such was the case. laving said
that, l have tried to represent the data lairly, not purposely excluding examples that
might contradict a generalisation l have identilied. contradictory examplestroubling
though they may be lor someone sorting out a grammarhave been included at various
points in this grammar.
much ol my lieldwork was done using 8olomon lslands lijin, the lingua lranca ol
4
Available online at: http:JJwww.ethnologue.comJshow_map.asp7name=8l.

8ee Appendix A: 8outheast 8olomonic language lamily. Lnlortunately, there are no published
descriptions available lor lirao, ohari, or malango. 1he nature ol the similarities and dillerences
between these languages and lengo is, at present, dillicult to establish.
2
8olomon lslands, as a metalanguage. As many descriptive linguists will well understand,
coming up with a suitable gloss when working with two languages is no mean leat. 1he
truth ol this is not minimised when working with three. 1here are instances in which
lijin seems to capture the essence ol a lengo word or construction more ably than
lnglish. ln those instances lijin is provided lor those who may benelit lrom its
inclusion. lor those who do not reap these sparse benelits, the lnglish gloss will have to
do.
1his grammar is organised according to the general layout lound in lynch, koss,
and crowley (2002). 1heir work, over the years and in that volume, represents
somewhat ol a standard lor Oceanic languages. where there is a standard it seems
prudent to lollow it. Ol course, there are points at which this grammar will, out ol
necessity, depart lrom their layout, but it is hoped that this grammar will 'lit' into the
larger Oceanic linguistic world.
!
2 llONOlOoY
As is olten the case with Oceanic languages, the inventories ol both consonants
and vowels in lengo is lairly small, there are lew (il any) allophones, and complex
articulations are generally avoided in carelul speech. 8tress is, with lew exceptions in
my lexicon, lound on the penultimate syllable. 1he exceptions, in which stress is on the
linal syllable, are: [
m
be'
m
be] 'grassland' (compare ['
m
be
m
be] 'butterlly'), [i'nuu] 'll:18o',
and [su'kui] 'one'. 1one is not contrastive. 8yllable structure is (c)V(V).
2. 1 lhonemes
2. 1. 1 consonants
lengo has 1 consonants.
p t k
b d q
v s
m n p
r
l
Figure $.": Consonants
Voiced stops are prenasalised ([
m
b], [
n
d], [
p
q]). 1he voiced 'labial' lricative varies
regionally between bilabial (JJ) and labiodental (JvJ), with the latter being lar more
common. 1he voiced dental lricative JJ also varies regionally with apical JzJ. 1he
phone JzJ is especially prevalent in the region surrounding kuavatu, but is rare
elsewhere. Note the lack ol both a voiceless 'labial' and 'velar' lricative counterpart (JlJ
and JxJ) lor those that are voiced (JvJ and JJ).
6
8ome minimal pairs are shown in (1).
6
maddieson (1984:46) observes, oenerally, the existence ol a given voiced lricative in the inventory ol
a language implies the presence ol the voiceless counterpart in the inventory."
4
(1) a. [pu] 'or'
[
m
bu] 'lL1'
b. [ute] 'lung'
[u
n
de] 'thatch rib'
c. [siki] 'any'
[si
p
qi] 'dry'
d. [sisi
m
bo] 'splash'
[sisivo] 'breeze'
e. [poto] 'rotten'
[poso] 'straight'
l. [tu
p
qu] 'story'
[tuu] 'overlap'
g. [tu
m
bu] 'relative'
[tumu] 'lall'
h. [mu
n
de] 'right side'
[mune] 'male'
i. [
p
qutu] 'pound'
[putu] 'strong'
j. [mumu] 'bait'
[munu] 'bird'
k. [
n
dui] 'child'
[rui] 'awaken'
l. [
n
deni] 'this'
[
n
deri] 'that'
m. [eno] 'lay down'
[epo] 'lly (noun)'
n. [uni] 'eat'
[uli] 'build'
o. [tu
n
du] 'drip'
[u
n
du] 'sit'
p. [tuve] 'llood'
[ue] 'give'
q. [uru] 'least row'
[suru] 'arrive'
r. [u
n
du] 'lriend'
[uru] 'dive'
s. [u
n
du] 'lriend'
[ulu] 'head'
t. [uru] 'dive'
[ulu] 'head'
2. 1. 2 Vowel s
1here are six vowels in lengo.
i u
e o
c
u
Figure $.$: %o&els
1he lront open mid vowel JcJ is quite rare, appearing in live words in the data:
[kc
n
de] 'crisp potato', [kc
n
di] 'bottle', [mcmcle] 'choose', [tctc] 'chop', and [tctcte] 'hill'.
8ome minimal pairs are given in (2).
(2) a. [ni] 'olN'
[ne] 'l8:!8o'
b. [vuti] 'pull'
[voti] 'break'
c. [kc
n
de] 'crisp potato'
[e
n
de] 'yellow-leal greens'
d. [
n
deni] 'this'
[
n
duni] 'day'
e. [uro] 'shoulder'
[uru] 'they'
Vowel sequences, examples ol which are shown in (!), are not uncommon.

(!) a. [mui] 's.o. comes' b. [uoka] 'lemale, woman' c. [uu] 'knile'

d. [deu] 's.o. goes' e. [kei] 'basket' l. [leo] 'inside'
g. [leu] 's.o. lights' h. [buliu] 'scar' i. [vutie] 'lourth'
j. [lio] 'neck' k. [niu] 'coconut' l. [dou] 's.o. is blind'
m. [ioe] 'll:28o' n. [loi] 'lrog' o. [dou] 'heart'
p. [buu] 'betelnut' q. [tue] 'one shell' r. [kui] 'skink'
Aside lrom JcJ, which is never lound lollowing or lollowed by another vowel, there are
two vowel sequences, JueJ and JuoJ, which do not occur. lvery other vowel sequence is
lound in the language.
1here are what seem, on lirst hearing, to be long vowels. lurther analysis
reveals that these are in lact two vowels ol the same quality that happen to be one
lollowing the other across a morpheme boundary, as in (4)a. and b.
(4) a. ara pi tu-u
!ll wait-O:18o
'1hey await me.'
b. me garaa tine
m-e gara-a tina-a
cON}-!8o pull-O:!8o mother-l8:!8o
'and it [the crocodile] pulled his mother'
1hese are interpreted as VV sequences.

1here is a single word in my lexicon ol approximately 200 words, that appears

to have a long vowel that does not occur across a morpheme boundary: [koo] 'deadlall,
a tree rotting on the ground'. lt is distinct lrom the subject relerence pronoun [k-o]
'lkk-28o' and the modal verb [kou] 'unwilling'.
2. 2 lhonotacti cs
2. 2. 1 8yll abl e structure
lengo syllables are (c)V(V). 1here are no closed syllables in lengo, that is there

Note that when -a 'l8:!8o' joins to a word ending in JuJ as in [tinu-u], the two JuJs raise to an JeJ
resulting in [tine]. 1his is discussed below beginning on page 9.
6
are no syllable linal consonants, nor are there consonant clusters.
8
cV syllables are by
lar the most common, though V and cVV syllables are also prevalent. 8ome examples
are provided in ().
() a. [u] '18o'
b. [tu] 's.o. stands'
c. [tuu] 'spouse'
d. [e.o] 'yes'
e. [i.nuu] 'll:18o'
Vowel sequences have predictable syllablilication. ll V
1
V
2
are high-mid, high-
low, mid-mid, or mid-low they are members ol dillerent syllables. ll, however, V
1
V
2
are
low-mid, low-high, mid-high, or high-high they are members ol the same syllable.
linally, il V
1
V
2
are the same they are members ol the same syllable. 1he data lrom (!)
above are presented again, this time with syllable breaks indicated.
(6) a. [mui] 's.o. comes' b. [uo.ka] 'lemale, woman' c. [uu] 'knile'
d. [de.u] 's.o. goes' e. [kei] 'basket' l. [le.o] 'inside'
g. [le.u] 's.o. lights' h. [bu.li.u] 'scar' i. [vu.ti.e] 'lourth'
j. [li.o] 'neck' k. [niu] 'coconut' l. [do.u] 's.o. is blind'
m. [i.o.e] 'll:28o' n. [loi] 'lrog' o. [dou] 'heart'
p. [bu.u] 'betelnut' q. [tu.e] 'one shell' r. [kui] 'skink'
1he one exception to the syllabilication rules above is that the vowels in the mid-high
vowel sequence JeuJ are members ol dillerent syllables, as is shown in (6)g.
8yllable structure at morpheme boundaries lollows these syllabilication rules.
consider the data in ().
() a. [pi.tu-u] 'wait-O:18o'
b. [tuu.-u] 'spouse-l8:!8o'
ln ()a. the syllable structure is cV.cVV. 1he VV sequence that results when the two
morphemes come together is heard as a double JuuJ. 1he situation in ()b. is somewhat
8
loan words such as [kliniki] 'clinic' and [spedi] 'spade, shovel', which have consonant clusters, are not
considered in this analysis.

more complex. 1he low-high vowel sequence JuuJ is part ol a cVV syllable. 1he
morpheme J-uJ, lollowing as it does the high vowel JuJ, is a member ol a dillerent
syllable. 1he resulting syllable structure is cVV.V.
2. 2. 2 ll i si on and contracti on
lengo speakers are given to collapsing certain words in what amounts to elision.
8ome examples lrom natural speech are given in (8).
(8) a. [sukui] [skui] 'one'
b. [siki] [ski] 'any'
c. [siki pileu] [skpileu] 'around, approximately'
when slowed down, however, the words are restored to their lexical lorms. what is
evident here is that consonant clusters, which are not lound when words are uttered in
isolation, are present in connected speech.
1here are some lengo words that are shortened by most speakers. when
shortened, these words take on a 'heavy sound' (lengo speaker description) on the lirst
syllable. 8o mamala 'lightweight' becomes mala, with a somewhat heavy emphasis on
the lirst syllable. 1his is to dillerentiate mala 'lightweight' lrom mala 'naked'. l learned
this during a volleyball game. 8omeone was described as being mala tupa 'light lor
jumping', l didn't get it quite right and called him mala tupa 'naked nut' . . . l had been
perplexed lor some time by what appeared to be an aspirated JkJ on kau 'crab' that
distinguished it lrom kau 'dog', as it was the only instance ol the apparent aspiration. A
somewhat wider distribution might be expected il aspirated JkJ was indeed a lengo
phoneme. At the volleyball game l was given a partial list, given in (9), ol shortened
words that take on this 'heavy sounding' quality.
8
(9) a. [kukuu] [kuu] 'crab', cl. [kuu] 'dog'
b. [kikilu] [kilu] 's.o. demands more', cl. [kilu] 'plant sp.'
c. [mumulu] [mulu] 's.t. is lightweight', cl. [mulu] 'naked', [mumulu] 'tree sp.'
d. [pupuru] [puru] 's.o. accuses another', cl. [puru] 's.t. is hot'
e. [poporo] [poro] 's.o. is clothed'
l. [tutupu] [tupu] 's.o. jumps', cl. [tupu] 'nut sp.'
g. [vuvupu] [vupu] 's.t. is sharp', cl. [vupu] 's.o. eats'
lt should be said that the dillerence is not one ol tonehigh versus lowbut emphasis.
with some speakers the head actually moves lorward when uttering the lirst syllable ol
some ol these words.
9
llision is widespread in normal speech (that is, not slowed down lor 'lengo as a
second language' speakers). ln lnglish, lor example, one linds: do not don't, would
not wouldn't. 8ome examples ol elision, marked here with an apostrophe ('), are:
(10) a. [puru nu uo] [puru n'uo] 'it's hot (lit., hot Ak1 sun)'
b. [e teiu siki tuve] [e te'si'tuve] 'it [the river] is not llooded'
c. [ke sukui ni] [kes'k'i] 'belore'
d. [teiu nu V] [te'nu V] 'Nlo Ak1 V'
1he disappearance ol JuJ lrom the article na is quite regular and predictable belore
words that begin with JuJ. lt is not dropped belore the other vowels. 1he negative
teigha, which is lollowed by na 'Ak1', is lrequently shortened to te' na. 1his is not the
same as tena 'lOc'.
2. ! morphophonemi cs
1here is an interesting two-step morphophonemic process in lengo. lt involves
vowel raising and vowel harmony in third person singular directly possessed nouns.
when -a 'l8:!8o' joins to a noun that ends in low vowel JuJ, the two JuJs are raised to a
single mid vowel JeJ. ln addition, any JuJ lurther lelt in the word harmonises with the
9
8pectral analysis ol these 'heavy sounding' words, though beyond the scope ol this paper, would make
an interesting topic lor lurther investigation.
9
word-linal JeJ. 1his continues until a vowel other than JuJ is lound, in which case
any JuJ to the lelt ol the intervening mid or high vowel remains JuJ. 1his vowel raising
does not allect mid (JeJ, JoJ) or high (JiJ, JuJ) vowels. lxample (11) demonstrates -a
'l8:!8o' joining to nouns that do not end in JuJ.
(11) a. kuli 'ear' - -a 'l8:!8o' kulia 'ear-l8:!8o, her ear'
b. abe 'armpit' - -a 'l8:!8o' abea 'armpit-l8:!8o, his armpit'
c. livo 'tooth' - -a 'l8:!8o' livoa 'tooth-l8:!8o, its tooth'
d. ulu 'head' - -a 'l8:!8o' ulua 'head-l8:!8o, her head'
lxample (12) demonstrates -a 'l8:!8o' joining to nouns ending in the low vowel JuJ.
(12) a. gha 'oral.cll' - -a 'l8:!8o' ghe 'oral.cll-l8:!8o'
b. lima 'hand' - -a 'l8:!8o' lime 'hand-l8:!8o'
c. longa 'landward' - -a 'l8:!8o' longe 'landward-l8:!8o'
d. vunga 'hair' - -a 'l8:!8o' vunge 'hair-l8:!8o'
e. atha 'name' - -a 'l8:!8o' ethe 'name-l8:!8o'
l. ghalagha 'up-coast' - -a 'l8:!8o' gheleghe 's.t.'s up-coast side'
g. ghanaghana 'thinking' - -a 'l8:!8o' gheneghene 'thinking-l8:!8o'
lxamples ol an intervening non-low vowel that disrupts JuJ harmony are given in (1!).
(1!) a. kabula 'inner thigh' - -a 'l8:!8o' kabule 'inner thigh-l8:!8o' (not *kebule'
b. vanua 'village' - -a 'l8:!8o' vanue 'village-l8:!8o' (not *venue'
1he object pronoun sullix -a 'O:!8o' has the same shape as the possessor person
sullix -a 'l8:!8o'. lowever, when -a 'O:!8o' is joined to a verb that ends in low vowel JuJ,
the two JuJs are not raised to an JeJ but become a double JuuJ. 1his is evident in (14).
(14) a. ghu-vi dea-a 'shout-1k go-O:!8o'
b. gara-a 'pull-O:!8o'
1he examples in (1) nicely demonstrate the dillerence in vowel harmony
between possessive pronoun -a 'l8:!8o' and object pronoun -a 'O:!8o'. ln (1)a. sapa
'seaward' is used as a noun ol direction possessed by vae 'house', while in (1)b. sapa is
used as a verb indicating motion toward the sea.
10
(1) a. pala sepe ni vae
pala sapa-a ni vae
side seaward-l8:!8o olN house
'seaward side ol the house'
b. sapaa na thinaghe
sapa-a na thinaghe
seaward-O:!8o Ak1 canoe
'launch the canoe' (lit. '[lilt it] seaward the canoe')
ln (1)a. -a 'l8:!8o' sets oll a process ol vowel raising and harmonisation. Note that the
process ol vowel harmony is restricted to the word, that is, the JuJs in the preceding
word pala 'side' are not allected by the process which sapa 'seaward' undergoes. ln
(1)b. -a 'O:!8o' does not set oll the vowel raising and harmonisation process.
11
! NOLN8, lkONOLN8 ANu NOLN llkA8l8
1o the category 'noun' belong words lor persons, places, and things. Nouns can
be inllected as directly possessed nouns
10
, they are preceded by an article, they are
sometimes modilied by adjectives J stative verbs, and they can act as modiliers in
attributive constructions. lengo has an extensive set ol pronouns that distinguishes
lour persons (lirst person inclusive and exclusive, second person, and third person),
and lour numbers (singular, plural, dual, and paucal), but no gender.
11
ln addition to the
paradigmatic pronouns, there is a rellexive pronoun stem to which direct possessor
pronoun sullixes are joined to distinguish person and number and an interrogative J
relative pronoun pair that distinguishes between human and non-human.
uemonstratives distinguish between singular and plural proximal, medial, and distal
relerence.
!. 1 Nouns
ln lengo words lor persons, places, and things (including concepts and
abstractions) are nouns. Nouns can be derived lrom verbs, can be inllected as directly
possessed nouns,
12
are most olten preceded by an article, are sometimes modilied by
adjectives J stative verbs, and can act as modiliers in attributive constructions (e.g., gari
ghaoka 'child lemale, girl', vae ato 'house leal, i.e., a house with a sago palm leal rool' [as
distinct lrom vae kava 'house copper, i.e., a house with a tin rool']).
10
ln keeping with the larger Oceanic tradition, possession is discussed under the labels direct and
indirect possession. 1his is intended as a means to distinguish the syntax ol possession lrom the
semantics ol possession. 8emantically, possession is either alienable or inalienable. 8yntactically, this
is expressed with lree possessor person pronouns (indirect J alienable) or possessor person sullixes
joined to the possessed noun (direct J inalienable).
11
1here are two words that distinguish gender in lengo: thogho 'same-sex sibling' and vavine 'opposite-
sex sibling'. Otherwise the general terms ghaoka 'lemale' and mane 'male' are used.
12
1hat is, semantically inalienable.
12
!. 1. 1 ueri vati on and i nll ecti on
Verbs are derived lrom nouns by a process ol stem-initial cV reduplication.
(16) noun verb
a. vothe 'paddle' vo(vothe 's.o. paddles a canoe'
b. vugho 'lishing net' vu(vugho 's.o. nets lish'
ln these instances cV reduplication is a process ol verbilication (although compare verb
derivation (page ) where nouns are derived lrom verbs by a similar process ol stem-
initial cV reduplication).
lntensilication ol a noun is also achieved by stem-initial (V)cV reduplication.
1!
(1) )-ami sake longa-i na leoni nimami na agu(agutu.
lkk-1lX.ll carry landward-O:!ll Ak1 all l8:1lX.ll Ak1 kul-thing
'we take all our cargo" (lit., thing.thing) ashore with us.'
larlier in this textone concerning my lamily going by canoe to the capital citythe
non-reduplicated lorm ol agutu is used.
(18) k-ami oni nimami na agutu
lkk-1lX.ll prepare l8:1lX.ll Ak1 thing
'we prepare our cargo"'
lere, at the outset ol the story, our cargo is considered somewhat manageable, that is,
the author hasn't lully considered (or recalled) the lot ol cargo that we tend to take
with us when we travel. Alter describing the process ol choosing what we will take with
us and what we will leave behind, carrying it all to the seaside, and loading it into the
canoe the magnitude ol the agutu has grown in the speaker's mind. 8o when it comes
time to describe the unloading ol our cargo, it has grown to agu(agutu. 1he noun is
'intensilied' through reduplication ol stem-initial VcV.
1!
with relerence to reduplication 8apir (1921:6) comments that Nothing is more natural than the
prevalence ol reduplication, in other words, the repetition ol all or part ol the radical element. 1he
process is generally employed, with sell-evident symbolism, to indicate such concepts as distribution,
plurality, repetition, customary activity, increase ol size, added intensity, continuance".
1!
ln another example the noun dae 'child [ollspring]' is lound reduplicated.
(19) ga(gara na vua ga(gara na da(dae m-e ga(gara ngata na dadae
kul-pull Ak1 crocodile kul-pull Ak1 kul-child cON}-!8o kul-pull strong na da-dae
me gara tine me gagara laka na vua
m-e gara tina-a m-e ga-gara laka na vua
cON}-!8o pull mother-l8:!8o cON}-!8o kul-pull also Ak1 crocodile
'the crocodile pulled, the child pulledand the child pulled hard, and his
mother pulled, and the crocodile also pulled'
ln this example, the child ol the woman (tina 'mother') attempts to pull her out ol a
crocodile's grasp. larlier in this text the non-reduplicated lorm dae 'child' is used twice
lor the same relerent.
14
child is reduplicated at this point in the text because, while he
is indeed the woman's ollspring, he is not the small boy the term might seem to imply.
1

le is a grown man with the strength and courage to engage in such a lile and death
struggle. 1o rellect the lact that while this man is indeed a 'child' he is also a man, that
the noun is larger than the hearer might expect, the noun is reduplicated. 1here is an
intensity, a sense ol increased magnitude, with reduplicated nouns.
16
1he single inllectional process to which nouns are subject is that directly
possessed nouns are inllected lor possession person, as in (20).
(20) na vae-gu
Ak1 house-l8:18o
'my house'
1he syntactic category ol direct possession indicates the semantic notion ol inalienable
possession.
1
14
loth instanceste ghuvia a daea 'she shouted to her son' and me rongoa a daea 'and her son heard
her'are used prior to the struggle with the crocodile.
1
1here are two words lor 'child' in lengo: dae and gari. 1he lormer indicates a child with particular
relerence to his or her lineage (i.e., 'descendant ol') while the latter indicates a pre-pubescent person.
16
lt should be noted that agu(agutu and da(dae are not anomalies such as stuttering or hesitation.
1
8ee page 46ll lor more discussion.
14
!. 1. 2 Noun cl asses and arti cl es
1here are two particles in lengoa and naeither ol which are lound belore
nearly every noun.
18
1he temptation is simply to assign them to the category 'article'.
1he situation, however, is not that simple. 1he category 'article' is usually thought to
have something to do with a noun's deliniteness.
19
ln lengo, the particles a and na do
not play any apparent role in marking relerentiality, that is, an item's status as new or
given, relerential or non-relerential in a context. lurther, a and na co-occur with the
demonstratives, which indicates that they probably belong to a dillerent class lrom
'determiners'. lut the more the widely accepted delinition ol 'article' is stretched to try
to lit the lunction ol these particles, the less satislying it is. 1he dilliculty in terms ol
these lengo particles is with the delinite J indelinite connotations that the label
'article' carries.
1hree options lor naming these particles present themselves: 1. come up with
unique (and opaque) labels such as noun class marker one" and noun class marker
two", 2. lollowing laspelmath (200), capitalise the label article" and render it
language specilic (e.g., lengo Article) to gain the reader's understanding that it is
18
1here are some instances ol nouns that do not take an article: personal names in seriesother than
the lirstdo not have one, neither do nouns lollowing the locative i, nor does the second noun ol
certain possessive J genitive constructions (an N possessed by a possessed N and N ni N). 1he articles
are also lound with other classes ol words than nouns: interrogative and relative pronouns (thi J tha),
demonstratives, and verbs in certain contexts. 1hese exceptions are discussed below.
19
1wo sample delinitions lor 'article' are: a particle that indicates a noun's delinite or indelinite
relerence, and new or given status" (lrom 8ll's glossary ol linguistic terms, accessed }uly 200 at
http:JJwww.sil.orgJlinguisticsJolossaryOllinguistic1ermsJwhatlsAnArticle.htm), and a subclass ol
determiners which displays a primary role in dillerentiating the uses ol nouns, . . . a distinction is
usually made into delinite and indelinite types, partly on semantic and partly on grammatical
grounds" (crystal 200!:!!). loth make relerence to a delinite J indelinite distinction. uryer
(200a:18) removes this limitation lrom his delinition ol article when he writes, Although these
words do not vary lor deliniteness, what they share with articles in luropean languages is the lact
that they are a set ol words which occur with high lrequency in noun phrases and which vary lor
certain grammatical leatures ol the noun phrase . . . Lnder this notion ol 'article', the coding ol
deliniteness is not a delining leature, but simply one ol the many grammatical leatures ol noun
phrases that articles olten code."
1
relerring to something article-like but that doesn't lit exactly the widely held
delinitions ol 'article',
20
or, !. call them 'articles', describe their uniqueness, and
demonstrate their use. 1he latter approach is lollowed here. uixon (1988:114) appears
to lean in a similar direction when he writes ol loumaa lijian, another Oceanic
language: lt is an established part ol the lijian grammatical tradition to use the term
'article' here, although it is being used in an unusual sense." Lnlortunately, uixon does
not describe what is unusual about it. koss (2004:00) notes ol Oceanic languages that,
typologically, Noun phrases in canonic languages usually consist minimally ol article -
noun, with a distinction between a common article and a personal article that is used
with a proper personal name, a kin term used to name someone, or a disjunctive
pronoun."
21
what lollows will demonstrate that although the term 'article' is used in
this grammar it indicates an article ol a particular type.
lt is dillicult to draw any kind ol line between delinite and indelinite, new and
given, when it comes to a and na. 1hey do not help distinguish between delinite J
indelinite, relerential J non-relerential, identiliable J non-identiliable, specilic J non-
specilic, or activated J non-activated inlormation about an item in context. 1he lirst
and all lollowing relerences to a noun are invariably lound with the same lorm ol the
appropriate article: a lor named humans and na lor unnamed humans and non-humans.
20
laspelmath (200:6) opines, l would not recommend the use ol opaque names to descriptive
linguists. Opaque names may be justilied by theoretical considerations, but they are not practical
because they are very hard to remember. 1he best solution is to use lamiliar terms lor mnemonic
reasons, but to capitalize them in order to emphasize that the categories are 'proper names'".
lollowing this logic, it seems that every grammatical category term would need to be capitalised and
preceded by the particular language name ol the one under study as every grammatical category is
realised somewhat dillerently in dillerent languages. ll the uniqueness ol the category's realisation in
the language being described is adequately explained, the capitalisation ol terms seems superlluous.
21
crowley (2002a:2) identilies two articles in related oela: a 'proper' and na 'common delinite subject
and object'.
16
New" inlormation in a discourse is olten marked with the same particle as those
instances in which it is given" inlormation. compare (21)a. and b., which are
sequential sentences in a text.
(21) a. u dea m-u tu-ri thai-a t-i na aliva.
18o go cON}-18o step-1k arrive.at-O:!8o kl-lOc Ak1 centipede
'l went and l stepped on a centipede.'
b. T-e mai na aliva e mai pasu i riki-gu.
kl-!8o come Ak1 centipede !8o come bite lOc digit-l8:18o
'1he centipede came, it came and bit my toe.'
ln the text, aliva 'centipede' is introduced as new inlormation in what is here example
(21)a. 1he noun aliva 'centipede', lor this lirst and all lollowing instances in the text, is
lound with na 'Ak1', as in (21)b. 1his demonstrates, at least in terms ol relerential
marking, that there is not properly a distinction between delinite and indelinite
relerence with a J nathe same particle does both jobs. 1he particles a and na on their
own are unmarked as to deliniteness.
22
1he only distinguishing parameter with relerence to a and na is that ol
humanity, and even this must be qualilied. 1here are two distinct noun classes with
which a and na are used. 1he lirst class is made up ol named" humans. lroper names
and terms ol close kinship (e.g., dae 'child, descendant', tama 'lather', tau 'spouse', and
tina 'mother') are lound with the article a.
2!
All other nounshuman and non-human,
22
1here are, however, larger constructions involving both articles that do provide specilic J non-
specilic relerence. 1hese are a ulm na 'Ak1 ulm Ak1' and a sakai na 'Ak1 one Ak1' respectively. 8ee below.
2!
Names have a special place in lengo society. ln lact, there are strictly held name taboos"persons ol
certain relationship to the speaker whose names are never spoken. 1his holds true lor people ol
higher status than the speaker: one's parents and one's spouse's parents are never relerred to by
name. Nuclear lamily terms are used lor lather and mother and the term tabu [tu
m
bu] 'lorbidden' is
used lor lather- and mother-in-law. lor those ol same or lesser status (i.e., spouse and children), their
name is not strictly tabu. 1wo less strictly held conventions are to reler to uncles and aunts as 'lather'
and 'mother', and to people with the same name as the speaker as simply 'name'. 1hese titles are, lor
all intents and purposes, the person's name.
1
animate and inanimateare lound with the article na 'Ak1' (e.g., gari 'child, youth (with
no relerence to descent)', ghaoka 'woman', mane 'man', tinoni 'person', vae 'house', vua
'crocodile').
1he article a is lound with a class ol nouns comprised ol named humans. lroper
names and kin terms that lunction as names populate this class. lxample (22) serves to
illustrate.
(22) a J. te dea kolua a teme i leo ni pono
a } t-e dea kolu-a a tama-a i leo ni pono
Ak1 } kl-!8o go cOm-O:!8o Ak1 lather-l8:!8o lOc inside olN bush
'}. went with his lather to the bush'
1he name }. and the kin term tama 'lather' are both preceded by the article a.
lnterestingly, tama 'lather' can be used as a proper name (as in the preceding example)
and as a common noun. consider its use in (2!).
(2!) ma !. igeia na tama-di ra na tinoni ni *.
cON} l ll:!8o Ak1 lather-l8:!ll Ak1 person olN A
'and l. he is their lather, the people ol [tribe] A.'
lere tama 'lather' is preceded by na, the common article. what dillerentiates the two
uses ol tama 'lather' in (22) and (2!), and inlorms the hearer which class ol noun is
being invoked, is the lorm ol the article.
lxample (24) demonstrates the use ol the article a as a pronoun. 1he lirst
sentence shows the article a belore the proper name l.
(24) a. + tavu-a a ,. b. * nga.
18o lind-O:!8o Ak1 l Ak1 ulm:8o
'l'm looking lor l.' '8he's over there.'
1he second sentence ol example (24) has the same article a acting as a pronoun belore a
demonstrative. ln this instance a 'Ak1' stands lor the proper noun 'l.' where one might
18
expect one ol the subject pronouns e '!8o' or igeia 'll:!8o'.
larly on it was dillicult to hear the article a in sentences like the lirst ol (24), as
the object marker -a 'O:!8o' got lused together with a 'Ak1'. lt was only by identilying a
'Ak1' lollowing emphatic pronouns, as in (2), that it was possible to distinguish a 'Ak1'
lollowing singular object marker -a.
(2) !-ko.ghami a /. ami-ko dea i nughu.
uL-ll:1lX.ll Ak1 l 1lX.ll-uL go lOc river
'l. and l, we two go to the river.'
while in (24) the article was dillicult to hear, it was there. lelore directly possessed
nouns, however, the article a is optional.
24
most times, as in (26)a. and b., a is present.
(26) a. 0 ghani-a na vua a tina-mu1
!8o bite-O:!8o Ak1 crocodile Ak1 mother-l8:28o
'A crocodile is biting your mother!'
b. e ghu-vi-a a dae-a
!8o shout-1k-O:!8o Ak1 child-l8:!8o
'she shouted to her son'
when heard in a stream ol connected speech the article a is lused with the
preceding JuJs ol, in these examples, vua 'crocodile' and ghu-vi-a 'shout-1k-O:!8o'.
lowever, during transcription with lengo speakers, it olten emerges. lut sometimes, as
24
crowley (2002a:2) notes ol oela that, ln object position, personal names are obligatorily preceded
by the article JuJ, while the article is optional belore sullixed nouns." As evidence ol this he provides
two examples:
and,
lere ( ) indicates optionality and 2 is the voiced velar lricative. lecause ol orthography issues,
crowley decided to represent lorms in phonemic rather than orthographic script" (26).
19
e inu beti a tephen
!8o:lA81 drink water Ak1 8tephen
'8tephen drank water'
geva ku ri2i-a 3a' tama-gu
later 18o:lL1 see-!8o Ak1 lather-18o
'later l will see my lather'
with dae-miu 'child-l8:2ll' in (2) below, a 'Ak1' is not there.
(2) ighamu so(soko na ti(tina ma na ta(tama a thi t-e logho
ll:2ll kul-linish Ak1 kul-mother and Ak1 kul-lather Ak1 kll kl-!8o have
dae-mi u i 4. skul
child-l8:2ll lOc l school
'all you mothers and lathers who have your children at l. school'
1his is particularly interesting because the article a is not lused with a preceding JuJ. ln
lact, the expected object marker on logho (-a 'O:!8o' or -i 'O:!ll') is also absent.
1he other article, na 'Ak1', is lound belore common nouns. ln (28) two animate
non-human nouns are lound with the article na.
(28) ! vi ga t-e dea ga na igha e tumuri po na vonu.
lOc where there kl-!8o go there Ak1 lish !8o lollow llm Ak1 turtle
'wherever the lish went the turtle just lollowed.'
ln (29) na 'Ak1' is lound with the animate non-human noun kau 'dog' and the already
noted human exception mane 'man'.
(29) * deni na ghai t-e ne abu ghini-a na kau na mane.
Ak1 ulm:8o Ak1 stick kl-!8o lmll hit lN81-O:!8o Ak1 dog Ak1 man
'1his is the stick with which the man really hit the dog.'
lxample (29) also demonstrates the 'specilic' construction a deni na 'Ak1 ulm Ak1' used to
relerence ghai 'stick'.
ln all the examples ol the article na presented to this point the article has
directly preceded the noun. 1his is not always the case. 1here are two classes ol words
quantiliers and indirect possessor pronounsthat can be lound between na 'Ak1' and
the noun. ln (!0) an intervening quantilier leoni 'all, every' is present.
(!0) !ghamu na l eoni tinoni k-amu rongo mai1
ll:2ll Ak1 all person lkk-2ll listen come
'lveryonelisten up!'
20
ln (!1) both a quantilier (balu 'some') and an indirect possessor pronoun (nigua 'l8:18o')
are lound between na and the noun.
(!1) + ngao-a k-ami-tu dea tura-i na bal u ni gua koga
18o want-O:!8o lkk-1lX.ll-lAL go carry-O:!ll Ak1 some l8:18o stick
'l want we lew to go carry my sticks'
1he situation lound in (!1) is only the case when a quantilier is involved. without a
quantilier the article lollows the indirect possessor pronoun and directly precedes the
noun as in (!2).
(!2) m-u dea i vae ni talu-a nigua na 4ible
cON}-18o go lOc house olN leave-O:!8o my Ak1 bible
'and l went to my house to leave my lible'
lxample (!!) nicely demonstrates the dillerent contexts in which the two
articles are used.
(!!) * thi na ethe a daemu ia5
a thi na atha-a a dae-mu ia
Ak1 kll Ak1 name-l8:!8o Ak1 child-l8:28o ulm
'what is your child's name7'
1he human article a 'Ak1' is lound twice: once belore the human relative pronoun thi
and again belore the close kinship term dae 'child', while na 'Ak1' is lound belore the
common, and in this case possessed, noun ethe 'name-l8:!8o'.
1here are a number ol contexts in which a noun does not take an article. ln
addition, there are a number ol word classes besides nouns that are lound with an
article.
An article does not appear belore nouns lollowing the locative i.
2
One could well
say i leghai 'lOc garden' but not *i na leghai, i ,oniara 'lOc loniara' but not *i a ,oniara 'lOc
Ak1 loniara'. ln lact, proper place names, when they take an article, take na 'Ak1'.
2
uixon (1988:11) reports the same phenomenon in loumaa lijian.
21
(!4) nimiu na thara tibo-miu t-i ighamu na ).
l8:2ll Ak1 least.row klll-l8:2ll kl-lOc ll:2ll Ak1 l
'your least.row yourselves, you [community] l.'
Another context in which the article is not used where it might be expected is
belore the possessed noun ol a possessed noun as in (!). 1his type ol construction is
considered a single unit and takes a single article.
(!) e dea otia na lime tine
e dea oti-a na lima-a tina-a
!8o go grab-O:!8o Ak1 hand-l8:!8o mother-l8:!8o
'he went and grabbed his mother's hand'
1his could be less-lreely glossed as 'he went and grabbed it the hand ol her [the]
mother ol him'.
oenitive constructions ol the type N ni N are also berelt ol an article on the
subordinate noun, an article appears on the lirst noun only.
(!6) na vua ni niu
Ak1 lruit olN coconut
'the lruit ol the coconut'
Again, this is considered a single unit and takes a single article.
linally, the article a 'Ak1' does not appear lollowing the conjunction ma 'and'
when more than one proper noun is listed.
(!) !tu-ghami a 4. ma 6. ma /. ma #. t-ami-tu uru.
uL-ll:1lX.ll Ak1 l cON} k cON} l cON} l kl-1lX.ll-lAL dive
'we lewl., k., l., l., [and l]we lew went diving.'
1he article a is lound belore the lirst name only.
1he particles a J na are lound with a variety ol other word classes other than
nouns. 1he marked-lor-human particle a is lound belore the human interrogative J
relative pronoun thi, and demonstratives (e.g., a nga inau 'Ak1 ulm ll:18o, l choose that
22
one'). 1he marked-lor-common particle na is lound belore the non-human animate and
inanimate interrogative J relative pronoun tha, and verbs lound lollowing certain
modal verbs ol aspect and negation. 1he particle a is never lound in the place ol na
belore verbs and while na does co-occur with demonstratives it is never lound directly
belore one (e.g., na mane deni 'Ak1 man ulm' and not 7na deni8 compare a deni, which is
grammatical, see (29)). 1he dillerent environments in which a and na are lound are
displayed in (!8).
(!8) a na
noun __ N (named" humans) __ N (common)
interrogative J relative pronoun __ thi (human) __ tha (non-human)
demonstrative __ ulm __ N (common) ulm
verb __ V
1he human interrogative J relative pronoun thi is preceded by the article a. ln
(!9) thi lunctions as the interrogative pronoun 'who'.
(!9) * thi na atha-mu ighoe5
Ak1 lN1 Ak1 name-l8:28o ll:28o
'who is your name7'
ln (40) thi has the role ol relative pronoun 'who'.
(40) ighamu so(soko na ti(tina ma na ta(tama a thi t-e logho
ll:2ll kul-linish Ak1 kul-mother and Ak1 kul-lather Ak1 kll kl-!8o have
dae-miu i 4. skul
child-l8:2ll lOc l school
'all you mothers and lathers who have your children at l. school'
ln both instances thi is preceded by a 'Ak1'. Note in (40) that the intensilied lorms ol tina
'mother' and tama 'lather' do not have the expected article a, but na. 1hat is because
here the terms are not being used to name kin (as in (22) above) but to reler in a more
general sense to all mothers and lathers.
2!
ln (24) and (29) examples ol the article a belore demonstratives were given. 1he
same is evident in (41). ln (29) and (41) the demonstrative relers to an inanimate object.
(41) * nga inau.
Ak1 ulm:8o ll:18o
'l [choose] that one.'
lxample (41) was spoken by a person indicating which lruit lrom among many she
wanted at a local market. lere the designation named human" lor the noun class
marked by a 'Ak1' is clearly stretched, but the article na is never used belore a
demonstrative.
1he article na is belore the non-human interrogative pronoun tha in (42).
(42) 9a tha t-o ghoni-a5
Ak1 lN1 kl-28o do-O:!8o
'what are you doing7'
1he 'what' ol (42) is the thing being done. when tha lunctions as a non-human
complementiser it also takes na 'Ak1'.
(4!) !nau ba k-u tugu vani-ghamu na tha ba k-e sara
ll:18o lL1 lkk-18o tell uA1-O:2ll Ak1 cOml lL1 lkk-!8o arrive
'l will tell to you what is going to happen'
1he 'what' ol (4!) is that which is going to happen.
1he article na can also be lound belore a verb in certain contexts. 1his kind ol
construction is olten translated by lengo speakers as an inlinitive, as in (44).
(44) 6uke na agutu nga, na tavu niu.
second Ak1 something ulm:8o Ak1 lind coconut
'1he second step is to lind some coconut.'
ln the examples in (4) the article na is lound between the verb and preceding modal
verbs ol desire and ability (ngao, tangomana) or negation (teigha, tabu, kou). 1hese modal
verbs are discussed in more depth below (see page 14ll), but lor now observe the
24
position ol the article na in (4).
(4) a. o ghua ngao-a na ghali-a siki 'e:tra flavour'
28o il want-O:!8o Ak1 make-O:!8o any extra llavour
'il you want to make any extra llavour"'
b. t-amu tangomana na ta(tagho
kl-2ll able Ak1 kul-go.lish
'you are able to go lishing'
c. ami-ko teigha na ta(tavu thai-a na kei
1lX.ll-uL Nlo Ak1 kul-lind arrive.at-O:!8o Ak1 basket
'we couldn't lind the basket'
d. Tabu na lavi-a na ghau: b-e ghado-gho.
Nlo Ak1 grab-O:!8o Ak1 knile Allk-!8o pierce-O:28o
'uon't grab the knile: no good it cuts you!'
e. )ou na vanga na igha.
Nlo Ak1 eat Ak1 lish
'1he lish are unwilling to eat [the bait].'
1he presence ol the article na in these instances treats the verb lollowing a modal verb
somewhat like a noun while retaining its verbal properties (notably the lact that it still
takes the object sullix when required).
8tative verbs can be nominalised with the article.
26
(46) k-a ne rongo-ni-a: na sule ma na pile, ghaoka ma na mane,
lkk-1lN.ll lmll hear-1k-O:!8o Ak1 big cON} Ak1 small woman cON} Ak1 man
gari mokoni, gari ghaoka
child post-pubescent girl child woman
'we all will listen to her: the big and the small, [the] women and the men, post-
pubescent girls, pre-pubescent girls'
Note that in (46) neither ghaoka 'woman, lemale', nor the serial nouns gari mokoni 'child
[lemale] post-pubescent' and gari ghaoka 'child lemale [pre-pubescent]', have an article
26
1he phrase na sule ma na pile here relers to peoplepart ol the 1lN.ll ka. 1he speaker is delineating
inclusive opposites: important (cl. mane sule 'big [i.e., important, respected] man') and not so
important, women and men, and so on. 1his reading is substantiated in other texts.
2
preceding.
2
As with (2) above, it is dillicult to explain this.
uespite the lact that the articles in lengo do not distinguish delinite and
indelinite relerence, there are constructions that provide specilic and non-specilic
relerence. 8pecilic relerence is signalled by a ulm na N while non-specilic is by a sakai na
N 'Ak1 one Ak1 N'. lxample (4) demonstrates specilic relerence with the construction a
ulm na N.
(4) a deni na ghai t-e ne abu ghini-a na kau na mane
Ak1 ulm:8o Ak1 stick kl-!8o lmll hit lN81-O:!8o Ak1 dog Ak1 man
'1his is the stick with which the man really hit the dog.'
1he noun ghai 'stick' is given specilic status with this construction.
lxample (48) serves to demonstrate non-specilic relerence.
(48) m-ara-ko dea thai-a a sakai na kokomu kikiki
cON}-!ll-uL go arrive.at-O:!8o Ak1 one Ak1 island small
'and they two arrived at a small island'
1he turtle and the lish were trying to outdo one another: diving deep, going under
rocks, and so on. 1hey were each unable to best the other until marako dea thaia a sakai
na kokomu kikiki 'they arrived at a small island'. 1his could have been any ol the
approximately 990 islands that make up 8olomon lslands (or, indeed, any other island in
the vast lacilic). 1he construction a sakai na 'Ak1 one Ak1' marks this as non-specilic,
new inlormation in the context. 1he next mention ol the island is marked with a
demonstrative.
(49) m-e tagu iti tena maone tena kokomu i a
cON}-!8o crawl up lOc sand lOc island ulm
'and it [the turtle] crawled up on the sand on that island'
2
Another example ol this involving inanimate objects is: ara kuki rais ara kuki viuvi ma na ti 'they cooked
rice, they cooked potato and tea'. while there is an article belore the linal noun ti 'tea', neither rais
'rice'nor viuvi 'potato' have an article.
26
lere the relerence is clearly 'specilic'.
!. 2 lronouns
lengo has a rich set ol pronouns with over 90 lorms in the categories ol
emphatic, subject relerence, object, and direct and indirect possessor pronouns (see
ligure !.1: lronouns). lengo distinguishes lour persons (1 inclusive (lN), 1 exclusive (lX),
2, !) and lour numbers (singular (8o), plural (ll), dual (uL), and paucal (lAL)), but no
gender. ln addition to these, there are human (thi) and non-human (tha) pronouns that
have both interrogative and relative lunctions, and a rellexive pronoun stem (tibo) to
which direct possessor pronouns join to indicate person and number.
emphatic subject
relerence
object
(e.g., V-u)
possessor
(direct)
(e.g., N-gu)
possessor
(indirect
general)
possessor
(indirect oral
consumable)
singular
1 inau u -u -gu ni-gu-a gha-gu-a
2 ighoe o -gho -mu ni-mo-a gha-mu-a
! igeia e -a -a, -na ne ghe
plural
1 lN ighita a -ghita -da no-da gha-da
1 lX ighami ami -ghami -mami ni-mami gha-mami
2 ighamu amu -ghamu -miu ni-miu gha-miu
! igeira ara ira, -ra, -i -dira no-dira gha-dira
dual (ko)
1 lN iko-ghita a-ko -ko-ghita ko- -da ko-no-da ko-gha-da
1 lX iko-ghami ami-ko -ko-ghami ko- -mami ko-ni-mami ko-gha-mami
2 iko-ghamu amu-ko -ko-ghamu ko- -miu ko-ni-miu ko-gha-miu
! iko-ira ara-ko -ko-ira ko- -dira ko-no-dira ko-gha-dira
paucal (tu)
1 lN itu-ghita a-tu -tu-ghita tu- -da tu-no-da tu-gha-da
1 lX itu-ghami ami-tu -tu-ghami tu- -mami tu-ni-mami tu-gha-mami
2 itu-ghamu amu-tu -tu-ghamu tu- -miu tu-ni-miu tu-gha-miu
! itu-ira ara-tu -tu-ira tu- -dira tu-no-dira tu-gha-dira
Figure ;.": /ronouns
lirst person non-singular (i.e., ll, uL, lAL) has inclusive (lN) and exclusive (lX)
lorms. lnclusive has within its scope the addressee(s) (e.g., iko-ghita 'uL-ll:1lN.ll, we
2
twoyou and l') while exclusive does not include the addressee(s) within its scope (e.g.,
iko-ghami 'uL-ll:1lX.ll, we twoshe and l but not you').
Number has a lour-way distinction in lengo. 1he lirst distinction is between
singular (8o) and non-singular (ll, uL, lAL). llural is non-singular non-specilic, that is to
say it can be used to mark two to an inlinite number ol relerents. within non-singular
there is a lurther distinction between specilic and non-specilic. uual (uL) is used lor two
and only two relerents, while paucal (lAL) is used to indicate three or more.
28
while dual
stands out as quite distinct (two and only two), the distinction between plural and
paucal is not always immediately apparent. what it comes down to is a matter ol
specilicity: paucal is used when speakers want to delimit the range ol relerents. 8o, lor
example, when someone is addressing a large crowd with instructions lor an upcoming
least and bases their speech on what happened at the last least, the paucal lorm may be
used to indicate those lew" who were at the last least (despite the lact that they
numbered over one hundred people).
1he single instance ol a distinction based on animacy among personal pronouns
is -i 'O:!ll' which is used lor inanimate objects only (e.g., talu-i na koga 'put.down-O:!ll Ak1
stick, put the sticks down'). All other third person pronouns are used lor both animate
and inanimate relerence (e.g., na koga ira 'Ak1 stick ll:!ll, the sticks' and na mane ira 'Ak1
man ll:!ll, the men'). As mentioned above, the relative J interrogative pronouns thi and
tha distinguish animate lrom inanimate.
28
koss (2004:498) notes, in some languages there is also a trial or a paucal or occasionally both, again
marked by a morpheme derived lrom a numeral." lt is possibe that the lengo paucal lorm is derived
lrom tolu 'three', something along the lines ol: tolu 'three' t[ol]u tu 'lAL'.
28
!. 2. 1 lmphati c
1he lengo emphatic pronoun provides participant relerence outside ol the verb
phrase. 1he emphatic pronoun is used lor subjects, objects and possessors. lt can stand
on its own, but is olten used together with a noun phrase. 1he set ol emphatic
pronouns is given in (0).
(0) 1lN 1lX 2 !
8o 3i'nau 3i'ghoe i3gei'a
ll 3i'ghita 3i'ghami 3i'ghamu i3gei'ra
uL 3i'-ko-ghita 3i'-ko-ghami 3i'-ko-ghamu 3i'-ko-ira
lAL 3i'-tu-ghita 3i'-tu-ghami 3i'-tu-ghamu 3i'-tu-ira
1he elements in parentheses are not always spoken when the emphatic pronoun is
used.
29
1his can lead to conlusion when attempting to sort out participant relerence,
especially in light ol the lact that, when the parenthetic parts are dropped, the non-
singular object pronouns are identical to their emphatic pronoun counterparts.
!0
when
emphatic and object pronouns appear similar, the subject relerence pronoun plays an
important role in sorting out participant relerence (see page 168), especially given the
variable nature ol lengo subject - object word order.
ln (1) both clauses employ the emphatic pronoun. 1he actors in each clause are
given emphasis by means ol this optional pronoun even though both are indicated with
subject relerence pronouns in the verb phrases.
(1) ara-ko gara iti-a na thinaghe i -ko.i ra m-u ghe tapa i nau
!ll-uL pull up-O:!8o Ak1 canoe uL-ll:!ll cON}-18o continue run ll:18o
'they two pulled up the canoe and l continued to run'
A more literal" translation is: 'they two pulled up the canoe they two and l continued
29
lor most ol these l have not been able to determine the conditions that give rise to the shortened
lorms. 1he one that seems consistent is the emphatic use ol ira 'll:!ll' with objects. 8ee (6) and ().
!0
1his is not unique to lengothe lnglish pronoun <ou presents its own challenges, serving as it does
as second person singular - plural, subject - object pronoun.
29
to run l'. Note that both instances ol the emphatic pronoun are lound lollowing the
verb in their respective clauses, while their corresponding subject relerence pronouns
(ara-ko and m-u) are lound belore the verbs. 1he role ol the emphatic pronoun here is to
reiterate the subjects and, by so doing, help dillerentiate the actors one lrom the other.
1hat is to say, that the 18o actor m-u = inau was not involved in pulling up the canoe
(even though he rode in it and would normally be expected to help), nor did the !ll
actors ara-ko run at this point in the narrative (they did run alter their lriend once the
canoe was ashore). 1he lact that the 18o actor had been bitten by a centipede has a lot to
do with the distinction ol actors and actions: the one bitten hightailed it home while
his companions took care ol other details belore they lollowed.
when a lirst person non-singular lorm ol the emphatic pronoun is used, the
speaker does not receive explicit mention, although his companion(s) do.
!1
(2) a. ! -ko.ghami a /. ami-ko dea i nughu.
uL-ll:1lX.ll Ak1 l 1lX.ll-uL go lOc river
'we two, l. [and l], we two went to the river.'
b. ! -tu.ghami a 4. ma 6. ma /. ma #. t-ami-tu uru.
lAL-ll:1lX.ll Ak1 l cON} k cON} l cON} l kl-1lX.ll-lAL dive
'we lewl., k., l., l. [and l]we lew went diving.'
c. m-ami-tu dea talu iti i -tu.ghami 4. ma 9.
cON}-1lX.ll-lAL go put up lAL-ll:1lX.ll l cON} N
'we lew put them up: we lew l. and N. [and l].'
1he narrator does not receive specilic mention in these examples, though he is
delinitely included among the participants by virtue ol the lirst person plural emphatic
pronoun lorm. Note in these examples that the subject noun phrase lollows the
emphatic pronoun. 1his is always the case when they co-occur.
!1
compare lengo 'we lew X and Y and 2' to lnglish 'X, Y, 2, and !'.
!0
when a non-singular emphatic pronoun is not used lor the speaker and her
companions, the speaker can be made explicit along with her companions by using a
singular lorm ol the emphatic pronoun, such as inau 'll:18o' in (!).
(!) a. Tena ade rodo i nau ma . kol u-a T. ami-tu thivo
lOc 8unday evening ll:18o cON} 8 cOm-O:!8o 1 1lX.ll-lAL go.down
'8unday evening l and 8. with 1. we lew went down'
b. ! nau, . ma T. ami-tu dea
ll:18o 8 cON} 1 1lX.ll-lAL go
'l, 8., and 1., we lew went'
1here is no evidence that this type ol construction lor naming multiple participants
raises the prolile J status ol the speaker over against her companions in contrast to a
non-singular lorm ol the emphatic pronoun (as in (2)). lt is worth noting ol (1) above
that it comes in the middle ol the source text. lxample (!)a. comes at the beginning ol
the same text, and in (!)a. the emphatic pronoun inau is lound belore the verb. 8o
while the emphatic pronouns in (1) help contrast the respective actors, the emphatic
pronoun coming as it does at the beginning ol (!)a., belore the subject relerence
pronoun - V, serves an enumerating lunction lor the main actors ol the storythe
narrator inau 'l' and his lriends.
lxample (4) presents evidence ol the emphatic pronoun being used to indicate
possessor person. ln (4) the possession type is direct as indicated by the nouns allixed
by the lirst person possessor pronoun sullix -gu.
(4) a. ma k-ami soni iti-i vati tuthivo-gu i nau
cON} lkk-1lX.ll throw up-O:!ll journey's.end waterlront.landing-l8:18o ll:18o
'at journey's end we may throw them up at my own water-lront landing'
b. ara gito-a t-i m-ara lavi dea-a na kei-gu i nau
!ll steal-O:!8o kl-lOc cON}-!ll take go-O:!8o Ak1 basket-l8:18o ll:18o
'they stole it and they took it away my basket mine"'
!1
1he emphatic pronoun may also indicate indirect possessor person, as in the case ol
noda 'our' in ().
() t-e vonu noda na commitment i ghi ta
kl-!8o lull l8:1lN.ll Ak1 commitment ll:1lN.ll
'it is lull [i.e., big to overllowing], our own commitment'
uespite that lact that possessor person is the syntactic category in view (and not
subject), the emphatic nature ol this pronoun is evident. 1he owner ol the possessed
noun is emphasised by the emphatic pronoun. ln (4)a., b., and () the emphatic
pronoun is a dillerent person and number lrom the subject ol the sentence (k-ami
'lkk-1lX.ll' in (4)a., ara '!ll' in (4)b., and t-e 'kl-8lk:!8o' in ()). lt is, however, the same
person and number as the possessor. 1his means that the emphatic pronoun is not
standing pronominally lor the subject but lor the possessor. 1here is no evidence in
lengo ol a repeated possessor pronoun, no instance ol, lor example, *noda na > noda.
1he emphatic pronoun lills the role ol emphasising possessor person.
ln (6) the object, the noun phrase na koga 'the sticks', is marked by the plural
inanimate object pronoun sullix (-i) on the verb and emphasised with the pronoun ira.
(6) )-ami-tu longa dea ma k-ami-tu dea tura-i na koga i ra.
lkk-1lX.ll-lAL landward go cON} lkk-1lX.ll-lAL go carry-O:!ll Ak1 stick ll:!ll
'we will go landward and we will go carry the sticks.'
8ince the retrieval ol the sticks is the goal ol the author's story in (6), they are called
to the hearer's attention at this pointthe outsetol the text. 1he use ol the emphatic
pronoun is a signilicant means lor doing this.
Again, in () both the third person plural versions ol object pronoun and
emphatic pronoun are used.
!2
() t-e rongo-ni-ra na kau ma na i ra
kl-!8o hear-1k-O:!ll Ak1 dog cON} Ak1 ll:!ll
'it [the crocodile] hears the dogs and the others [boys]'
1he sullix -ra indicates that the verb rongo 'hear' has an animate !ll object. 1he dogs
and the boys who imitate dogs are, taken together, the plural object ol the transitivised
verb rongo-ni-ra 'hear-1k-O:!ll'. 1he lree lorm ira 'll:!ll' is used pronominally lor the
young boys who tease the crocodiles by barking like dogs.
An intriguing use ol the lirst person singular emphatic pronoun is shown in (8).
(8) 4. k-o dea tugu vani-a nga i nau
l lkk-28o go tell uA1-O:!8o ulm:8o ll:18o
'l., go tell herwho is actually me . . . '
lere inau 'll:18o' is used as part ol a third person relerence to the storyteller in the
story she is telling that involves her.
!2
!. 2. 2 8ubj ect relerence
1he subject relerence pronoun is a verb phrase constituent which indexes
subject arguments. 1he subject relerence pronoun:
1. is, with lew exceptions, the lirst constituent ol the verb phrase.
!!
ln these
rare instances when a subject relerence pronoun is not present, the verb
lacking a subject relerence pronoun is part ol a larger chain ol clauses in
which the subject is clearly understood. lt is also optional in imperative
sentences,
2. is allixed lor 1Am (e.g., ba k-u dea 'l will go'),
!4
and dual or paucal number,
!. has word-level stress,
4. is juxtaposed to noun phrases in predicate nominal constructions (e.g., e na
ghaoka 'she [is] a woman').
!2
1his is discussed in more depth in relation to demonstratives on page 1.
!!
1he auxiliaries ba 'luture' and boro 'impossible luture' precede the subject relerence pronoun.
!4
1he subject relerence pronoun can remain unmarked (, indicative), or be prelixed by a 1Am marker:
g- 'perlective', t- 'realis', k- 'irrealis', or b- 'apprehensive'. 1he prelixes m- 'cON}' and p- 'Al1' also join to
the subject relerence pronoun, but as conjunctions and not as 1Am markers. 8ee page 8ll lor more
discussion.
!!
oiven this particle's role ol indexing subject person and number lor the verb as well as
indicating the 1Am signature ol the clause, it is closely related to the verb. ll this was the
limit ol the subject relerence pronoun's role its classilication would not pose such a
problem. lt is when the subject relerence pronoun is lound as part ol predicate nominal
constructions that classilicatory dilliculties arise. According to layne (199:114-19), a
predicate nominal will consist, at a minimum, ol two juxtaposed noun phrases with
variants ol this including the addition ol a copula as verb, pronoun, an invariant
particle, or a derived lorm. 1he available options lor describing this morpheme are
either a subject indexing particle that belongs to the verb phrase or a pronoun. 1he
dilliculty is that this class ol morphemes is both at the same time.
ln the pronoun section ol their typological overview ol Oceanic languages,
lynch et al. (2002) observe that, 1here are generally several separate paradigms ol
pronominal lorms." 1he third pronoun paradigm they note, and the one ol interest
here, is,
8ubject: most languages have one or more sets ol preverbal morphemes
usually clitics, but sometimes prelixes or lree lormswhich indicate
the person and number ol the subject. ln melanesian languages these are
olten portmanteau lorms which combine with the expression ol the
tenseJaspectJmood categories ol the verb (s!.2, in the grammar sketches
these are olten handled under the verb phrase) (!).
ln section !.2 they explain lurther that,
lt is in the area ol verbal morphology and verb phrase syntax that
Oceanic languages generally exhibit the greatest complexity. Oceanic
languages generally have preposed verbal morphemes, lalling into two
basic types, according to whether these morphemes are lree or prelixed .
. . 8traddling the divide between lree and prelixing systems are those in
which a morphologically complex preverbal marker expresses a
combination ol subject and 1Am categories (4).
!4
lt is into this class that lengo's pre-verbal subject relerence pronouns lall. leesing
(1988) provides much the same analysis when he writes,
l have argued that at the very 'core' ol Oceanic syntax lies a clause
structure minimally comprising a Vl incorporating a subject relerence
pronoun and (in transitive clauses) an object-marking pronoun. 1he
subject Nl is an optional expansion on this Vl, as is an object Nl. Once
subject relerence is established, with an explicit noun or contextually,
relerence is maintained in subsequent clauses with subject relerence
pronouns within the Vls: there are no subject Nls in such clauses (14).
!
lollowing leesing l use the term subject relerence pronoun lor this particle.
One dilliculty presented by this analysis is that these subject relerence
pronouns do not pattern alter nouns in that they occupy dillerentand cannot occupy
the samespaces in a clause. According to layne (199:4!) pronouns are, lree lorms
(as opposed to allixes) that lunction alone to lill the position ol a noun phrase in a
clause. 1hey normally have all the distributional properties ol noun phrases." 1his is
not true ol the subject relerence pronoun: it does not share the distributional
properties ol noun phrases.
Another dilliculty this analysis presents is that ol the subject relerence
pronoun's role in marking time. 1he subject relerence pronoun is discussed as part ol
the verb phrase because ol the range ol 1Am markers which allix to it. 1he subject
relerence pronoun's role in carrying 1Am marking is unmistakable. lt is dillicult to lind
relerence in the literature to the marking ol time on a subject marker, much ol what
has been written seems to assume that time is marked on the verb or with some sort ol
auxiliary to the verb.
!6
Nordlinger and 8adler (2004), however, have evidence that
!
1his accurately describes the situation in lengo with the exception that subject noun phrases
continue to be lound alter subject relerence has been established, as demonstrated above in the
discussion ol the emphatic pronoun.
!6
layne (199:2!) at least makes allowance lor 1Am marking on the verb phrase: All 1Am operations
are most olten associated with the verbal word (lor polysynthetic languages) or verb phrase (lor more
!
contradicts those assumptions.
A standard assumption in linguistic theory is that leatures relevant to
the clause as a whole are associated with a clausal head. ln more
concrete terms, this translates into the assumption that clausal leatures
such as tenseJaspectJmood (hencelorth 1Am) are encoded by verbs, verb-
like auxiliary elements or particles . . . lowever, this assumption is
incorrect. ln many languages dependent nominals and nominal
modiliers may also be inllected lor tense, aspect and mood interpreted
with respect to the clausal predication (1).
1his provides the space required to analyse the lengo subject marking particle as a
pronoun with 1Am marking as part ol the verb phrase.
1he lengo subject relerence pronouns are shown in (9).
!
(9) 1lN 1lX 2 !
8o u o e
ll a ami amu ara
uL a-ko ami-ko amu-ko ara-ko
lAL a-tu ami-tu amu-tu ara-tu
1hat the subject relerence pronoun is best considered a lree standing pronoun, and not
an allix or a (pro)clitic to the verb, is conlirmed by the lact that a number ol words
including the 1Am auxiliaries ba 'ulO' and bo 'lmll', and adverbs ne 'lmll', sakai ni 'belore',
soko 'linish', and tighi 'lirstly'can come between the subject relerence pronoun and
the verb.
!8
1he subject relerence pronoun, and especially its role in marking 1Am, receives a
more detailed treatment in the discussion ol the verb phrase. lrielly, however, consider
example (2) again, re-presented here as (60).
isolating languages)."
!
koss (1988:!6), in 1able !8: 8ubject lronominal clitics ol 8elected Oceanic languages, gives lengo
lorms ol atu '1ll' (1 inclusive plural) and amo '2l' (2 plural). l analyse a-tu as '1lN.ll-lAL' and l hear '2ll'
as [umu].
!8
8ee the relevant sections lor examples.
!6
(60) a. !-ko.ghami a /. ami -ko dea i nughu.
uL-ll:1lX.ll Ak1 l 1lX.ll-uL go lOc river
'we two, l. [and l], we two went to the river.'
b. !-tu.ghami a 4. ma 6. ma /. ma #. t-ami -tu uru.
lAL-ll:1lX.ll Ak1 l cON} k cON} l cON} l kl-1lX.ll-lAL dive
'we lewl., k., l., l. [and l]we lew went diving.'
c. m-ami -tu dea talu iti i-tu.ghami 4. ma 9.
cON}-1lX.ll-lAL go put up lAL-ll:1lX.ll l cON} N
'we put them up: we lew l. and N. [and l].'
1he subject relerence pronoun agrees with the subject noun phrase in person and
number, it is always lound belore the verb, and it is marked lor 1Am.
most olten, there is a single subject relerence pronoun per clause with a single
1Am argument.
!9
Only rarely, and in a limited number ol lorms, are serial subject
relerence pronouns lound. 1hese are m- k- lorms with single syllable subject relerence
pronouns, specilically m-u k-u 'cON}-18o lkk-18o', m-o k-o 'cON}-28o lkk-28o', m-e k-e 'cON}-!8o
lkk-!8o', and m-a k-a 'cON}-1lN.ll lkk-1lN.ll' as in (61).
40
(61) k-o dea ta(tagho m-o k-o lavi mai-a
lkk-28o go kul-lish cON}-28o lkk-28o take come-O:!8o
'you go lishing and bring what you catch'
A more literal translation ol this is, '[imperative] you go lishing and you [imperative]
you take come it [i.e., what you catch]'. with the multi-syllable lorms a subject
relerence pronoun marked lor 1Am (g-, t-, k-, or ) is lound at the beginning ol the lirst
clause and either subject relerence pronouns marked with the conjunction m- in
consecutive clauses (e.g., t-ara . . . m-ara . . . m-ara) or the lull conjunction ma belore k-
marked subject relerence pronouns (e.g., ma k-ara) in consecutive clauses. what is not
!9
while there may be more than one subject relerence pronoun per clause, this latter assertion is
invariable: each subject relerence pronoun can take only one prelix.
40
1he prelixes m- 'cON}' and p- 'Al1' do double duty as clause and 1Am conjunctions. lor a discussion ol p-
k- lorms see page 196.
!
lound is, lor example, *m-ara k-ara.
!. 2. ! Obj ect
Object pronoun sullixes, shown in (62), index object arguments on the verb
constituent(s) ol the verb phrase.
41
8erial verb constructions may have more than one
indexing object sullixone per verbwhen there is more than one object-type
argument.
(62) 1lN 1lX 2 !
8o -u -gho -a
ll -ghita -ghami -ghamu -ra (animate)
-i (inanimate)
uL -ko-ghita -ko-ghami -ko-ghamu -ko-ira
lAL -tu-ghita -tu-ghami -tu-ghamu -tu-ira
1he object pronoun sullix joins to verbs to index the object.
(6!) a. /i tu-u1
wait-O:18o
'wait lor me!"
b. *ra pi tu-u.
!ll wait-O:18o
'1hey await me.'
c. ? pi tu-a na thinaghe5
28o wait-O:!8o Ak1 canoe
'[Are] you awaiting a canoe7'
1he object is not alwaysindeed, not oltenindicated with a lull noun phrase.
One thing ol note concerning the object pronouns is that there is a distinction
between animate and inanimate in third person plural. compare the examples in (64).
(64) a. t-e rongo-ni-ra na kau ma na ira
kl-!8o hear-1k-O:!ll Ak1 dog cON} Ak1 ll:!ll
'it [the crocodile] hears the dogs and the others [boys]'
41
8imons (19:12-1!) makes the case that in ohaimuta, 1he one syllable object pronouns [-u,'O:18o', -go
'O:28o', -a 'O:!8o', -ra 'O:!ll'] are written as sullixes to the verb" (8imons 19:12). 1he multi-syllable
lorms [(ko-=tu-)gita J gami J gamu J ira] are, according to 8imons, lree lorms. cl. layne (199:42-44).
!8
b. e pulu-i na ghola na ghaoka
!8o roll.up-O:!ll Ak1 pudding Ak1 woman
'the woman rolls up the puddings'
1he animate objectsdogs and boysare marked on the verb with -ra in (64)a., while in
(64)b. the inanimate objectspuddingsare marked with -i.
when there is a single verb and two objects in a sentenceone direct and the
other indirectthe indirect object is marked on the verb with an object pronoun sullix,
while the direct object is not. 8o in (6) a single object is indexed on the verb despite
the lact that there are both recipient (-gho 'O:28o') and theme (na buka 'Ak1 book' and na
igha 'Ak1 lish') arguments in the sentences.
(6) a. u athe-gho na buka
18o give-O:28o Ak1 book
'l gave you the book'
b. * thi t-e athe-gho na igha5
Ak1 who kl-!8o give-O:28o Ak1 lish
'who gave you the lish7'
ln lengo there are never two objects marked on the verb: *u athe-gho-a na buka and *u
athe-a-gho na buka are ungrammatical. ly way ol contrast, consider (66), which is a
serial verb construction with two object-type arguments. lere the book remains the
direct object J theme and is marked on the lirst verb with an object pronoun sullix
while the indirect object J recipient is marked on the second verb.
(66) u athe-a vani -gho na buka de
18o give-O:!8o uA1-O:28o Ak1 book ulm:8o
'l'm giving you this book'
1he introduction ol a second verb allows the indexing ol a second object argument.
1he object pronoun can also be used in genitive constructions, as in (6).
!9
(6) 0 bo ghilaghana t-i na sanga ni ghi ta.
!8o lmll know kl-lOc Ak1 talk olN O:1lN.ll
'le almost knows our language.' (lit., 'the language ol us')
lere there is no object pronoun marked on the verbit is part ol the larger object
genitive marked by ni 'olN'.
linally, it is worth pointing out the similarity ol lorm but the signilicant
dillerence ol meaning between -a 'O:!8o' and -a 'l8:!8o'. compare the role ol -a in (68).
(68) a. m-e rongo-a a dae-a t-e ghu-vi dea-a
cON}-!8o hear-O:!8o Ak1 child-l8:!8o kl-!8o shout-1k go-O:!8o
'and her son heard her, the one she shouted to'
b. M-e ribe sakai na ghaoka m-e ghu-ghu8 e ghu-vi -a
cON}-!8o surprise one Ak1 woman cON}-!8o lN18-shout !8o shout-1k-O:!8o
a dae-a.
Ak1 child-l8:!8o
'8urprised, one ol the women started shouting, she shouted to her son.'
lt is important to remember that -a 'O:!8o' joins to verbs while -a 'l8:!8o' joins to nouns,
possessive classiliers ni 'olN' and gha 'oral', and rellexive tibo. ln (68)a., -a 'O:!8o' is joined
to the directional dea 'go' lollowing ghu-vi 'shout-1k'. ln (68)b. the -a attached to ghu-vi
'shout-1k' is 'O:!8o' whereas the -a joined to dae 'child' is 'l8:!8o'.
!. 2. 4 ui rect possessor
1he direct possessor pronouns, used with inalienably possessed nouns, are
presented in (69).
(69) 1lN 1lX 2 !
8o -gu -mu -a, -na
ll -da -mami -miu -dira
uL ko- -da ko- -mami ko- -miu ko- -dira
lAL tu- -da tu- -mami tu- -miu tu- -dira
1here are two lorms ol the !8o bound possessor pronoun: -a and -na. Ol these, -a is the
40
more common (though recall the vowel raising involved when -a 'l8:!8o' joins to a noun
that has a linal JuJ). 1he bound possessor pronoun -na is lound only with a lew words:
tua-na 'leg-l8:!8o, his leg' (compare vanua-a which becomes vanu-e 'village-l8:!8o, her
village' with vowel raising), vua-na 'lruit-l8:!8o, its lruit', and de-dea-na 'Nk-go-l8:!8o, her
going'. 1here are also cases ol wordslor example dae 'child' and vae 'house'that are
lound with either -a or -na, depending on the speaker (i.e., some say dae-a 'child-l8:!8o,
his child' while others say dae-na 'child-l8:!8o, his child').
1hese sullixes attach to the directly possessed noun, as in (0).
(0) a. na vae-gu
Ak1 house-l8:18o
'my house'
b. ko-ghanaghana-di ra na ghaoka dini
uL-thinking-l8:!ll Ak1 woman ulm:ll
'those two women's thinking'
8ee page 4 lor a more complete discussion ol direct possession.
!. 2. lndirect possessor
1here are two classes ol indirectly possessed nouns: the oral consumable" class
indicated with the classilier gha and another general" class indicated with the
classilier ni. 1o each ol these classiliers is joined the set ol direct possessor pronoun
sullixes to indicate person and number. 1he general" indirect possessor pronouns,
used with alienably possessed nouns, are in (1).
(1) 1lN 1lX 2 !
8o ni-gu-a ni-mo-a ne
ll no-da ni-mami ni-miu no-dira
uL ko-no-da ko-ni-mami ko-ni-miu ko-no-dira
lAL tu-no-da tu-ni-mami tu-ni-miu tu-no-dira
1here are a lew things to note about the indirect possessor pronoun set. lirst, that the
41
lorm ol the classilier ni is dillerent in 1lN and ! person (no and ne). 8econd, that the 18o
and 28o lorms appear to have a !8o object pronoun sullix -a that indexes the possessed
noun (see (2)a. and (!)b.). 1he singular lorm is used even when the possessed noun is
plural and -i 'O:!ll' would be expected in terms ol number agreement. 1he plural lorms
ol the general possessor pronouns do not have an object pronoun sullix (see (2)b., c.
and (!)c.). linally, the 28o lorm has a slight lowering ol the vowel in the direct
possessor sullixlrom J-muJ to J-moJ (compare -mu 'l8:28o' in (69)). morpheme breaks
ol a lew representative indirect possessor pronouns are shown in (2).
(2) a. ni-gu-a 'cll-l8:18o-O:!8o'
b. ko-ni-mami 'uL-cll-l8:1lX.ll'
c. no-da 'cll-l8:1lN.ll'
1he syntax and semantics ol possession is discussed in more depth below. lor
now, note that the lree indirect possessor pronoun precedes the possessed noun with
its article. 1he possessor lor whom the pronoun stands may be indicated by a noun
phrase lollowing the possessed noun, as in (!)a.
(!) a. ne na be ?.
l8:!8o Ak1 pig O
'O.'s pig'
b. ni -gu-a na 4ible
cll-l8:18o-O:!8o Ak1 lible
'my lible'
c. k-ami oni ni -mami na agutu
lkk-1lX.ll prepare cll-l8:1lX.ll Ak1 thing
'we prepare our things' [i.e., cargo]
ln (!)a. the possessor, O., is made explicit with a noun phrase lollowing the possessed
noun (na be), while in (!)b. and c. only the possessor pronoun and the possessed noun
are lound. ln all three examples the indirect possessor pronoun is at the beginning ol
42
the noun phrase.
1he oral consumable" indirect possessor pronouns are used lor things
consumed via the mouth (e.g., lood and tobacco). 1hese are presented in (4).
(4) 1lN 1lX 2 !
8o gha-gu-a gha-mu-a ghe
ll gha-da gha-mami gha-miu gha-dira
uL ko-gh-ada ko-gha-mami ko-gha-miu ko-gha-dira
lAL tu-gha-da tu-gha-mami tu-gha-miu tu-gha-dira
while this set dillers lrom the general" class ol indirect possessor pronouns in that the
lorm ol the classilier remains the same lor all persons and numbers,
42
this set is the
same in that the 1 and 28o lorms retain the !8o object sullix -a. Additionally, the direct
possessor sullix lor 28o retains its high vowel quality J-muJ.
1he lood in () is specilied with a noun phrase na vudi lepa 'Ak1 ripe bananas'.
() @ha-mu-a na vudi lepa.
oral.cll-l8:28o-O:!8o Ak1 banana ripe
'[lere is a] ripe banana lor you to eat.'
1hat the banana is the possession ol a 28o participant is indicated by gha-mu-a 'oral-
l8,28o-O:!8o'.
!. 2. 6 kell exi ve
1he lengo rellexive is lormed by adding direct possessor pronoun sullixes to the
stem tibo 'klll'. lt decreases the valence ol a verb, as in (6).
(6) a. ba k-o ne dea thudu t-i tena ovu ni vanua-mu8
lL1 lkk-28o lmll go sit kl-lOc lOc group olN village-l8:28o
thara ti bo-mi u t-i
least klll-l8:2ll kl-lOc
'you will go sit in your village group, eat among yourselves'
42
1he !8o lorm is a case ol vowel raising: gha - -a 'l8:!8o' ghe.
4!
b. + toka ti bo-gu.
18o cut klll-l8:18o
'l cut mysell.'
ln (6)b., the transitive verb toka 'cut' would normally have a sullix to indicate the
object. lere the object sullix is dropped and the verb is lollowed by a valence-reducing
rellexive.
1he rellexive also serves as a kind ol emphatic possessive pronoun, as in ().
() na tha t-o lavi mai-a, na vanga tha t-o ghali mai-a
Ak1 kll kl-28o take come-O:!8o Ak1 lood kll kl-28o make come-O:!8o
tena arere, ba k-e oli tena nimoa na thara ti bo-mu
lOc 8aturday lL1 lkk-!8o return lOc l8:28o Ak1 least.row klll-l8:28o
'what you bring, the lood that you make on 8aturday, it will return to your own
least row'
1he same 'emphatic' behaviour in relation to possession is lound with the emphatic
pronoun as mentioned in the discussion on page !1.
!. 2. lnterrogati ve J rel ati ve
lengo has two interrogative J relative pronouns: thi (human) and tha (non-
human). 1hey are not marked lor number. 1he interrogative J relative pronoun thi
always appears with a 'Ak1' preceding, while tha is lound with na 'Ak1'.
4!
ln (8) examples ol thi and tha as relative pronouns are presented.
(8) a. na tinoni ketha a thi ga deni ba k-e mai lau-a
Ak1 person dillerent Ak1 kll there ulm lL1 lkk-!8o come take-O:!8o
pile-a na vanga de
little.bit-O:!8o Ak1 lood ulm
'a dillerent person who is over there will come take a bit ol this lood'
b. na vanga tha t-o ghali mai-a tena arere, ba k-e oli
Ak1 lood kll kl-28o make come-O:!8o lOc 8aturday lL1 lkk-!8o return
4!
1his gives some indication that a human J non-human distinction between the lengo articles a and
na may have been stronger at some point in the past. 8ee page 1 lor more discussion.
44
tena nimoa na thara tibo-mu
lOc l8:28o Ak1 least.row klll-l8:28o
'the lood that you make [and] bring on 8aturday will return to you in your
least row'
1he lunction ol thi and tha within relative clauses is discussed below (page 198).
ln (9) examples ol thi and tha as interrogative pronouns are lound.
(9) a. * thi t-e athe-gho na igha5
Ak1 lN1 kl-!8o give-O:28o Ak1 lish
'who gave you the lish7'
b. 9a tha t-o ghoni-a5
Ak1 lN1 kl-28o do-O:!8o
'what are you doing7'
1he pronouns thi and tha do double duty as interrogative pronouns 'who' and 'what'
respectively. 1he role ol thi and tha in posing content questions is discussed below
(page 18ll).
!. ! lasi c noun phrase structure
1he lengo noun phrase is minimally comprised ol a noun. while the article is
rarely omitted, it is absent lrom time to time and must be considered optional. leyond
this there are optional indicators ol possessor (direct and indirect) person and number,
along with adjectives and demonstratives.
(80) (l8) (NLm) (ll) (Ak1) N (-l8) (Au}) (ulm)
1he direct possessor pronoun (-l8) is the only allix lound in the noun phrase.
8ome coordinate noun phrases are joined by the conjunction ma 'cON}', as na sule
ma na pile 'the big and the small' are in (81), while others are simply juxtaposed, as are
gari mokoni, gari ghaoka'young teen girls, young girls'.
(81) k-a ne rongo-ni-a: na sule ma na pile, ghaoka ma na mane,
lkk-1lN.ll lmll hear-1k-O:!8o Ak1 big cON} Ak1 small woman cON} Ak1 man
4
gari mokoni, gari ghaoka
child post-pubescent girl child lemale
'we all will listen to her: the big and the small, [the] women and the men, post-
pubescent girls, pre-pubescent girls'
uisjunct noun phrases are joined by the conjunction pa 'or'.
(82) @ari mane pa na gari ghaoka5
child boy or Ak1 child girl
'ls it a boy or a girl7'
!. !. 1 lossessi on
lossession is a somewhat complex subject in the Oceanic lamily ol languages.
1here is an interplay between semantics and syntax that results in distinct possessive
constructions lor nouns ol particular classes. According to koss,
lossessive construction systems in canonic languages are delined along
lour parameters (lichtenberk 198, koss 1998c): a. whether the
possessed noun is alienable or inalienable, b. how many subtypes ol
alienable possession occur, c. whether the possessor is a pronoun, a
common noun phrase, or a personal noun phrase, d. whether a noun
phrase possessor is specilic or non-specilic (2004:11).
oiven the interplay between semantics and syntax, there is overlap among these
parameters, especially with points c. and d. lengo also has a verb ol possession, logho
'have', that operates both with and independently ol the possessive constructions lor
nouns. 1he discussion that lollows will address these parameters in turn in an attempt
to demonstrate possession constructions in lengo.
lengo nouns lall into one ol three classes: inalienable or alienable. linship
terms, body parts, and certain locations J directions are considered inalienably
possessed. 1here is another set ol nouns which, although they do not belong to the
inalienable categories attested in many other Oceanic languages, nevertheless are used
with the inalienable possession construction type. lor the purposes ol this grammar, l
46
consider these inalienable.
44
most other nouns are alienable. within the alienable set ol
nouns there are two classes: an oral consumable class (lood and tobacco) and a general
class.
lelore going lurther, and in keeping with the larger Oceanic tradition, it is
necessary to distinguish between semantic and syntactic categories when discussing
possession. lalmer states the case succinctly:
Oceanic languages typically have two distinct constructions lor marking
nominals to index their possessor. ln one, the 'direct' construction,
possessor-indexing sullixes attach directly to the possessum noun, while
in the other 'indirect' construction the sullixes attach to one ol several
possessive classiliers or bases, rather than the possessum noun itsell.
1his is olten relerred to as a distinction between an inalienable
construction and an alienable one. lowever, although those terms do
capture the prototypical lunctions ol the constructions, they reler to
semantics, not morphology or syntax (200:1).
lor the purposes ol this description the terms direct and indirect possession are used to
describe the bound and lree lorm possessor constructions with the understanding that
these represent the semantic categories ol inalienable and alienable possession
respectively. 1hese parameters are represented on the chart in (8!).
(8!) lossession constructions
possession type direct indirect
sematics inalienable alienable
class item location J
direction
oral general
syntax bound pronoun lree pronoun
uirect possession is the construction type used lor the class ol nouns to which
belong body parts, kin terms, and certain locations J directions. with these nouns,
possession is expressed with a possessor pronoun sullix attached directly to the
44
lynch (19!:) proposes the term semi-alienable lor such a set, that is, nouns that, may be thought
ol as being normally part ol the possessor but somehow removable under certain ill-delined
circumstances."
4
possessed noun. 1he examples in (84) demonstrate direct possession constructions lor a
body part ((84)a.), kin term ((84)b.), and locations ((84)c. and d.).
(84) a. na aro-gu mauli t-e rasa
Ak1 shoulder-l8:18o lelt kl-!8o tear
'my lelt shoulder tore'
b. *i1 0 ghani-a na vua a ti na-mu1
hey !8o bite-O:!8o Ak1 crocodile Ak1 mother-l8:28o
'ley! 1he crocodile is biting your mother!'
c. 4a ku kabia na gilu i pala sepe na vaegu.
ba k-u kabi-a na gilu i pala sapa-a na vae-gu
lL1 lkk-18o dig-O:!8o Ak1 hole lOc side seaward-l8:!8o Ak1 house-l8:18o
'l am going to dig a hole to my house's seaward side.'
d. m-e eno ti ko-levugha-mami t-i a J. g-i na gari
cON}-!8o lay kl-lOc uL-between-l8:1lX.ll kl-lOc Ak1 } llc1-lOc Ak1 child
'and he lay down between }. [and l], the boy did'
ln (84)c. the direction sapa 'to seaward' is inalienably possessed by the side ol the
house.
4
ln (84)d. the location levugha 'between' is directly possessed by }. and the
storyteller.
uual and paucal lorms appear as prelixes to the possessed noun with a direct
possessor pronoun sullix, as in (84)d. and (8).
(8) * ko-dae-mami e belo.
Ak1 uL-child-l8:1lX.ll !8o ring.bell
'Our two's child is ringing the bell.'
ln addition to the possessor person sullix joined to the directly possessed noun,
the possessor can also be specilied by an emphatic pronoun lollowing the possessed
noun, as in (86).
4
Ol course, the respective sides ol the house also possess 'upcoast', 'downcoast', and 'to landward'
directions. ln lact, depending on the orientation ol a piece ol timber, a person can cut the end that is
pointing in the 'upcoast', 'downcoast', 'to landward' or 'to seaward' direction.
48
(86) na kei-gu i nau kolu-a na batere i leo
Ak1 basket-l8:18o ll:18o with-O:!8o Ak1 battery lOc inside
'my basketmine, with the battery inside'
1his is not the only context in which lengo uses two distinct lorms to indicate aspects
ol the same thing, emphatic pronoun and subject relerence pronoun are used to
indicate subject, and dual pronouns and the numeral ruka 'two' are used together in the
data to specily number (see example (114)).
A directly possessed noun may itsell be directly possessed, as (8) demonstrates.
lere e '!8o', the actor who went and grabbed, possesses the mother who in turn
possesses the hand. 1he direct possessor pronoun sullix is applied to both inalienable
nouns. Note that the second noun, tine 'mother', does not have an article preceding.
(8) e dea otia na l i me ti ne
e dea oti-a na lima-a tina-a
!8o go grab-O:!8o Ak1 hand-l8:!8o mother-l8:!8o
'he went and grabbed his mother's hand'
1his could be less-lreely glossed as 'he grabbed it the hand ol her [the] mother ol him'.
1he chart provided in (88) shows constructions lor the direct J inalienable
possession constructions discussed thus lar.
(88) lossession constructions
possession type direct indirect
sematics inalienable alienable
class item location J
direction
oral general
syntax N-l8 lOc-l8
1his will be lilled out lurther as each aspect ol possession in lengo is discussed.
1here is a signilicantly large set ol nouns which do not lall into the widely
attested inalienably possessed noun category ol Oceanic languages (i.e., kin, body parts,
and certain directions), but which in lengo take the direct possessor construction.
49
8ome lengo words that do not lall within the commonly attested category ol
inalienable possession in Oceanic languages, but take the direct possession lorm, are:
atha 'name', ghanaghana 'thinking, opinion', kei 'basket', leghai 'garden', leo 'want,
desire', mauri 'someone lives' (e.g., mauri-dira 'their living'), pera 'basket', thinaghe
'canoe', tuthivo 'waterlront landing', vae 'house', vanga 'lood', vanua 'village', and vua
'lruit (as belonging to a tree)'. lt may well be that this set has always existed in lengo,
though lurther research might show whether the changes wrought by the relatively
recent introduction ol a cash economy have altered the terms ol possession in lengo
society.
46
1he examples in (89) show some ol these in context.
(89) a. @ari k-amu mai thanga-u na turu-vaghini-a na vae-gu.
boy lkk-2ll come help-O:18o Ak1 erect-cAL8-O:!8o Ak1 house-l8:18o
'loys, come help me stand up my house [posts].'
b. * thi na atha-mu ighoe5
Ak1 lN1 Ak1 name-l8:28o ll:28o
'what is your name7'
c. *ra gito-a na kei -gu1
!ll steal-O:!8o Ak1 basket-l8:18o
'1hey stole my basket!'
d. 0 l eo-mu na inu beti bithi5
!8o want-l8:28o Ak1 drink water cool
'uo you want a cold drink7' (lit., 'ls your desire a cold drink7')
uespite the lact that in (89)c. the speaker was alienated lrom his kei 'basket' when
someone stole it, it is marked with the direct J inalienable possessor construction.
1here are two classes ol indirectly possessed nouns: a class ol oral consumables
(i.e., lood and tobacco) and a general class. 1he oral consumable class is marked by the
classilier gha together with the appropriate direct possession sullix lor possessor
46
lt is dillicult, however, to see how items such as 'name', 'thinking', and 'desire' would be so allected.
0
person and number.
4
1he general class is marked by a lorm ol the classilier ni 'olN'
together with the appropriate person and number direct possession sullix. 1hese
classilier pronouns are lound belore the noun phrase ol which the possessed noun is a
part.
1he set ol pronouns lor the class ol possessed oral consumables is given on page
4!. 1hese have a classilier stem gha 'oral.cll' to which the direct possessor pronoun
sullixes join.
48
lxample (90) shows how this possessive classilier is used.
(90) gha-mu-a na vudi lepa
oral.cll-l8:28o-O:!8o Ak1 banana ripe
'[lere is a] ripe banana lor you to eat.'
while the more literal translation 'your oral-comsumable ripe banana' is rather bulky,
it is an accurate rellection ol the meaning ol (90).
while a number ol Oceanic languages have separate possession classes lor lood
and drink, lengo does not seem to distinguish the two. ln lact, the oral consumable
class may well include anything that is taken into the body via the mouth. ln (91) this
lorm is used lor a cigarette.
(91) + pulu-a gha-gu-a na piala.
18o roll-O:!8o oral.cll-l8:18o-O:!8o Ak1 cigarette
'l'm rolling my oral.consumable cigarette.'
while this is the only example in the data ol something being taken into the body via
the mouth that is not 'lood' per se, it is widely attested.
1he oral consumable possession construction is lilled in on the chart below.
4
1he oral consumable classilier gha is not the lengo generic noun lor lood, vanga 'lood' is that. oiven
that gha is not the generic 'lood' noun, it is considered a classilier in this description.
48
As previously noted, the 18o and 28o lorms also have the !8o object pronoun sullix attached.
1
(92) lossession constructions
possession type direct indirect
sematics inalienable alienable
class item location J
direction
oral general
syntax N-l8 lOc-l8 gha-l8 Nl
lt remains to discuss the general possession construction.
1he general class ol indirect possession is marked by a lorm ol the genitive ni
along with a direct possessor pronoun sullix. As with possessed oral consumables, in
the cases ol the 18o and 28o alienable specilic lorms, there is the addition ol the !8o
object pronoun sullix -a.
49
1he set ol pronouns lor the class ol general possession is
given on page 41. 1he general possessor pronoun appears belore the possessed noun
with its article, as in (9!).
(9!) m-u dea i vae ni talu-a ni -gu-a na 4ible
cON}-18o go lOc house olN leave-O:!8o olN-l8:18o-O:!8o Ak1 bible
'and l went to my house to leave my lible'
1he noun lor whom the possessor pronoun stands can be made explicit with a noun
phrase lollowing. 1his is shown in (94).
(94) ni -mi u na thara tibo-miu t-i i ghamu na ).
olN-l8:2ll Ak1 least.row klll-l8:2ll kl-lOc ll:2ll Ak1 l
'your least.row yourselves, you [community group] l.'
lere, while nimiu adequately indicates the possessor persons, their identity is more
lully specilied by the emphatic pronoun ighamu and the noun phrase na ). (the name ol
the community group being discussed).
1he addition ol the general indirect J alienable possession construction
completes the chart below.
49
lt is dillicult to say whether this has been added to the 1 and 28o lorms or dropped lrom the others.
2
(9) lossession constructions
possession type direct indirect
sematics inalienable alienable
class item location J
direction
oral general
syntax N-l8 lOc-l8 gha-l8 Nl ni -l8(-O) Nl
lt also completes the discussion ol direct and indirect possession. lut it does not
exhaust the subject ol possession in lengo. 1wo topics ol discussion remain. lirst, the
possessive verb logho 'have', and second, the use ol the general indirect possession
classilier ni in an array ol possessive constructions detailed below.
1he possessive verb logho 'have' sometimes supplants and sometime works
together with the possessed noun constructions discussed above. consider (96).
(96) a. leoni ghai doku ara l ogho vua-di ra doku
all tree good !ll have lruit-l8:!ll good
'every good tree has good lruit'
b. e lavi-a a tau-na ma ne na tinoni thairo ma na leoni
!8o take-O:!8o Ak1 spouse-l8:!8o cON} l8:!8o Ak1 person work cON} Ak1 all
agutu t-e l ogho-i
thing kl-!8o have-O:!ll
'he took his spouse and his servants and everything he had'
ln (96)a. the possessive verb is lollowed by a possessed noun. 1he vua 'lruit' are marked
lor possession twice as indicated by the !ll agreement between the subject relerence
pronoun ara and the direct possessor -dira. 1his is rellected in the more literal
translation 'all good trees they have lruit belonging to them [that is] good'. lxample
(96)b. is quite dillerent. ln the lirst two clauses possession is marked directly (tau-na
'spouse-l8:!8o, his spouse') and indirectly (ne na tinoni thairo 'l8:!8o Ak1 person work, his
servants') on the nouns. 1he third clause marks possession with a verb. 1he linal clause
in (96)b. does not mark possession indirectly on the noun as, lor example, ma ne na leoni
!
agutu 'cON} l8:!8o Ak1 all thing, and everything [else] ol his', though such a possession
construction could well have been used (compare kami sake longai na leoni ni mami na
aguagutu 'we carried to landward all our things').
lxample (9) demonstrates the use ol logho as a nominal modilier.
(9) a J. a sakai na mane l ogho8 ara vonu na dae-na ma ne na
Ak1 } Ak1 one Ak1 man have !ll many Ak1 child-l8:!8o cON} l8:!8o Ak1
tinoni thairo
person work
'}. [was] a rich man, many [were] his children and his servants'
layne (199:112) notes the semantically empty" nature ol verbs such as logho that
lunction as predicate adjectives and J or copular verbs in possession clauses.
1he discussion ol lengo possession concludes with the prepositional use ol the
general possession classilier ni. 1his possession construction is treated separately lrom
the others because it is ol a dillerent syntactic and semantic type. when ni is used as a
preposition it can be lound between two nouns (N ni N), between two verbs (V ni V),
and between a preposition ol location and a noun (i lkll ni N). when lound between two
nouns, ni indicates a purpose, characteristic, or part-whole relationship between the
two nouns.
0
ln V ni V constructions, ni contributes a sense ol purpose. And in i lkll ni N
constructions, the relationship is ol a possessed location. A lurther distinction ol these
constructions lrom those discussed previously is that the possession person sullixes do
not appear in this type ol construction.
when in the company ol two nouns (N ni N), ni can indicate purpose,
characteristic, or part-whole relationships: vae ni vanga 'house lor lood [preparation],
0
crowley (2002a:29), notes that, lurposive, characteristic and part-whole noun phrases are lormed
by means ol the preposition JniJ, which appears between two nouns."
4
kitchen' indicates the purpose ol the house, ghaoka ni . '8. woman, woman lrom 8.'
indicates an identilying characteristic ol the woman, and rau ni simiu 'leal ol simiu
plant' indicates a part-whole relationship between leal and plant.
when in the company ol two verbs (i.e., V ni V), ni indicates purpose. lxamples
ol this include pitu ni sukeri 'wait lor the purpose ol removing [hot stones lrom boiling
water]', and thudu ni saluva 'sit lor the purpose ol talking'. 1he verbs can be separated
by an adverbial phrase in addition to ni. ln the sentence m-u dea i vae ni talu-a nigua na
4ible 'and l went to the house to put my lible' the location, i vae 'lOc house', is an
adverbial phrase modilying the verb dea 'go'. 1he genitive preposition ni still indicates a
relationship ol purpose between the two verbs dea 'go' and talu 'put'.
8patial and temporal prepositional phrases must be in a genitival relationship to
a noun as indicated by ni. ln i leo ni pono 'lOc inside olN bush' the spaceinsideis a
location possessed by the bush. 8imilarly, in i levugha ni rodo 'lOc middle olN night' the
timemiddleis a part ol the larger night. 1he genitive preposition ni marks the
relationship ol the locative preposition to the noun.
while the discussion above covers the more common uses ol the genitive
preposition ni, there are other, more idiomatic, uses ol the particle. ln the sentence
thairo ni vae ni kolivuti 'work on behall ol house ol prayer" [i.e., church]' the
relationship ol vae 'house' to koluvuti 'prayer' as indicated by ni is one ol purpose, as
discussed above. 1he relationship ol the verb thairo 'work' to the remainder ol the
clause, however, is not covered by purpose, character, nor part-whole relationships.
1his is, rather, a genitive ol benelit or advantage. 1hough the dative is a more common
means to express this, the genitive ni is also pressed into service to express this.

linally, there is the idiom ke sakai ni V 'belore [ol a luture] V'.

1
ll one was to try
to break it down one might come up with k-e sakai ni 'lkk-!8o one olN'. ln the end, it is
probably best treated as irreducible. 1here is, however, a sense ol 'genesis' in the idiom:
'belore' is the antecedent of an ensuing event. ln that light ni plays a genitival role.
!. !. 2 Numeral s and number-marki ng
lengo has a decimal number system with cardinal and ordinal variants. 1he
cardinal numbers are given in (98). ln addition to the 'common' cardinal numbers,
which are used in day-to-day interactions, there is also a set ol 'kastom' cardinal
numbers,
2
that is, a special set ol numerals that are used in the telling ol custom
stories.
(98) common cardinal numbers
!
')astom' cardinal numbers
sakai 'one, 1' kea 'one, 1'
ruka 'two, 2' ura 'two, 2'
tolu 'three, !' lotu 'three, !'
vati 'lour, 4' tave 'lour, 4'
lima 'live, ' mila 'live, '
ono 'six, 6' no 'six, 6'
vitu 'seven, ' tivu 'seven, '
alu 'eight, 8' rau 'eight, 8'
thiua 'nine, 9' beta 'nine, 9'
thangavulu 'ten, 10' taleru 'ten, 10'
thangetu 'hundred, 100' thangetu 'hundred, 100'
togha 'thousand, 1 000' togha 'thousand, 1 000'
mola 1 000 0007
maua 1 000 000 0007
vuthia 1 000 000 000 0007
uki 1 000 000 000 000 0007
lega 'countless'
1
A variation ol this is g-8kl sakai ni V which means 'belore [ol a past] V'. 8ee (102)c.
2
)astom is the body ol customary practises, stories, and beliels ol a 8olomon lslands people. custom
story (lijin kastom stori) is the name lor the myths and legends ol 8olomon lslands peoples. Among
other things, they attempt to describe why things are the way they are (origins, practises, etc.).
!
1he numbers one through ten rellect, almost exactly, the lOc numerals lynch et al. (2002:2) have
reconstructed.
6
1here is lairly widespread knowledge ol the existence and members ol the set ol kastom
cardinal numbers (i.e., those who report it give a consistent set), but they do not appear
in any ol my texts. 1here is some conjecture among lengo speakers as to whether this
set ol numbers is borrowed lrom lirao, a neighbouring language, but l was unable to
conlirm it. what is evident is that these numbers are, by and large, variations ol the
lengo cardinal numbers. consider the 'kastom' number lotu 'three'. 8witch the
consonants (Jc
1
Vc
2
VJ Jc
2
Vc
1
VJ) and the result is cardinal number tolu 'three'. 1he
same process works lor tave (vati) 'lour' (although the quality ol the linal vowel is
slightly dillerent), mila (lima) 'live', and tivu (vitu) 'seven'. lt also works lor ura il one
allows lor a dropped initial JkJ ([k]ura ruka), lor no il one allows lor a dropped
initial JoJ ([o]no no), and lor rau il one allows that JrJ and JlJ are allophones (rau
alu).
creating compounds with the twelve 'common' cardinal numbers yields higher
numbers. lxamples are given in (99).
(99) a. thangavulu sakai 'ten one, 11'
b. thangavulu ruka 'ten two, 12'
c. ruka thangavulu 'two ten, 20'
d. ruka thangavulu sakai 'two ten one, 21'
e. sakai thangetu ruka thangavulu ruka 'one hundred two ten two, 122'
l. ruka thangetu ruka thangavulu ruka 'two hundred two ten two, 222'
multiples ol ten are lormed with a number lrom two to nine lollowed by thangavulu
'ten', as in (99)c. and d. ln the case ol the teens", thangavulu is not preceded by sakai
'one'. 1he same does not hold true lor the hundreds" and thousands": one hundred
and one thousand are preceded by sakai 'one' (as in (99)e.).
1here are lexical items lor certain groups ol ten. lncluded are ivolo 'ten garden

rows', kua 'ten eggs', panga 'ten live animals', pigu 'ten shells' (e.g., dry coconuts, live
crabs, seashells), and voragha 'ten eel lish'. 1hese can be compounded with other
cardinal numbers, as in (100):
(100) a. pigu 'ten shells'
b. pigu ruka 'ten shells two, 12 shells'
c. ruka pigu ruka 'two ten shells two, 22 shells'
Other exact 'tens' can be marked with an emphatic matapono.
(101) ruka thangavulu matapono
two ten even"
'twenty even"'
Numerals belong to the class ol stative verbs. Numbers may stand as the head ol
a predicate in lengo.
4
oiven that there is no existential J copula in lengo, and that they
lollow the subject relerence pronoun, the numerals are the predicate heads in the
sentences ol (102).
(102) a. Aoku t-i t-e vati 1
good kl-lOc kl-!8o lour
'lour [children] is enough!' (lit., 'it is good it's only lour [children]')
b. 0 ngitha na ivolo ni pana5
!8o how.many Ak1 ten.garden.row olN pana
'low many ten.garden.rows ol pana [are there in your garden]7'
#i ma teigha m-e ono na ivolo.
live Nlo cON}-!8o six Ak1 ten.garden.row
'[1here are] live, or il not, six ten-garden-rows.' (i.e., lilty or sixty rows)
c. ami-ko tumuri-a na theutu g-ami-ko sakai ni sara dea ga8
1lX.ll-uL lollow-O:!8o Ak1 road llc1-1lX.ll-uL one olN arrive go ulm
e ruka toni thevu tena nughu
!8o two times across lOc river
'we two lollowed the road belore we arrived there, two times [we had to go]
across the river'
4
uixon (1988:8) notes ol lijian that, numbers constitute a distinct word class, with greatest similarity
to verbs'.
8
d. e utha e vati thangavul u na dani ma na rodo m-e obo
!8o rain !8o lour ten Ak1 day and Ak1 night cON}-!8o llood
na maramana udolu
Ak1 earth whole
'it rained lourty days and nights and it llooded the whole earth'
1he examples in (102) pattern the same as the quantiliers avo 'many' and vonu 'lull,
many', which are also stative verbs (see discussion and examples on page 64).

ln addition to lilling the role ol predicate head as stative verbs, numbers can
also be nominal modiliers, as in (10!).
(10!) a. akai tena ovu sakai , ruka, tolu, vati, m-e dea tena lima.
one lOc group one two three lour cON}-!8o go lOc live
'One to group one, two, three, lour, and going [all the way] to live.'
b. ighami na ovu l i ma
ll:1lX.ll Ak1 group live
'we, group live'
ln (10!)a., group one, sakai, is the location indicated by tena 'lOc'. 1he speaker, discussing
the seating arrangement ol an upcoming least, maintained that there would be a least
row lor each village groupnone would be lelt out or combined with another because
they were smaller than the others. lach groupidentilied by numberwould have its
own place.
Ordinal numbers, lor the most part, are variations ol the 'common' cardinal
numbers given in (98) above. 1he set ol ordinal numbers is given in (104).

lynch et al. (2002:2) observe that lOc numerals had dual status ol adjectival verbs and nouns.
9
(104) tighi 'lirst, 1
st'
ruke 'second, 2
nd'
tolue 'third, !
rd'
vatie 'lourth, 4
th'
lime 'lilth,
th'
onoe 'sixth, 6
th'
vitue 'seventh,
th'
alue 'eighth, 8
th'
thiue 'ninth, 9
th'
thangavulue 'tenth, 10
th'
with the exception ol tighi 'lirst', and the numbers that end with JuJ (ruka 'two' and
lima 'live'), the ordinal numbers are the cardinal counterparts with an -e sullix. ln the
cases ol ruka and lima the linal JuJ is deleted belore the addition ol -e.
Number, that is singular and plural, is not marked on the lengo noun. compare
examples (10)a. and b. below. ln (10)a. all the nounsghai 'stick', kau 'dog', and mane
'man' are 'singular' while in (10)b. they are all 'plural'.
(10) a. * deni na ghai t-e ne abu ghini-a na kau na mane.
Ak1 ulm:8o Ak1 stick kl-!8o lmll hit lN81-O:!8o Ak1 dog Ak1 man
'1his is the stick with which the man really hit the dog.'
b. * dini na ghai t-ara ne abu ghini-ra na kau na mane.
Ak1 ulm:ll Ak1 stick kl-!ll lmll hit lN81-O:!ll Ak1 dog Ak1 man
'1hese are the sticks with which the men hit the dogs.'
As is evident, the nouns ghai 'stick', kau 'dog', and mane 'man' are not inllected lor
number. All indication ol number is handled by singular and plural pronouns:
demonstrative pronouns (deni = dini), subject relerence pronouns (e = ara), and object
pronouns (-a = -ra). A common means lor marking singular and plural number lor
objects is with third person object pronouns allixed to the verb phrase. Numerals are
also used to mark number: sakai 'one' lor singular and others lor plural. A linal means
lor marking number are the quantiliers siki 'a, any', balu 'some', and leoni 'all, every'
60
expressing the poles ol little and much as well as a medial number.
lerhaps the most common means lor marking number is with subject and
object pronouns which stand in relation to the verb. oiven the near obligatory nature ol
the subject relerence pronoun, subject number is olten indicated by it. 8imilarly, object
number is most olten marked by the indexing object pronoun sullix. 8ingular is marked
by the object pronoun sullix -a 'O:!8o' on the verb phrase, as in (106).
(106) m-u dea i vae ni tal u-a nigua na 4ible
cON}-18o go lOc house olN leave-O:!8o l8:18o Ak1 bible
'and l went to my house to leave my lible'
Only one lible was lelt at the speaker's house.
llural can be marked by the object pronoun sullix -i J -ra 'O:!ll'. ln (10) -i is
lound on the intransitive verb lebo 's.o. J s.t. lloats'. 1he addition ol the object pronoun
-i makes the verb transitive (i.e., 's.o. lloats s.t.') and the noun (present elsewhere in the
text but only indexed with a pronoun here) plural.
(10) m-ami-tu dea l ebo-i laka t-i
cON}-1lX.ll-lAL go lloat-O:!ll also kl-lOc
'we'll also lloat them [sticks lor ralters]'
ln (108) the plural object pronoun sullix is attached to the directional longa 'landward'.
(108) )-ami sake l onga-i na leoni nimami na agu(agutu.
lkk-1lX.ll carry landward-O:!ll Ak1 all l8:1lX.ll Ak1 kul-thing
'we take all our cargo ashore with us.'
1he 'plural' markers are multiplied in (108). Note the agreement between longa-i
'landward-O:!ll' and leoni 'all, every'. oiven that reduplication can indicate
intensilication ol a noun, the reduplication ol agutu 'thing' to agu-agutu 'cargo' is yet
another way to mark plurality.
1he quantilier siki 'one, any' is also lound belore nouns (with the exception ol
61
proper nouns) to indicate number. ll the articles a and na could be neatly described as
'delinite', siki would be a good candidate lor an 'indelinite' article. As the discussion ol
articles shows, however, a and na are used lor both delinite and indelinite relerences
(see page 1ll). lxample (109) is lairly straightlorward: one house is being torn down
(because it has begun to lall apart) to make room lor 'any' house to be built in its place.
(109) rosi-a na vae ni M. ni ghoni tughu-a ga si ki vae vaolu
remove-O:!8o Ak1 house olN m olN build change-O:!8o there any house new
'tear down m.'s house in preparation lor making a new one in its place'
1he type ol house to be builtleal or 'copper', lor sleep, lood preparation, or storage ol
an outboard engineis not specilied, it is just time to tear the old one down to make
room lor a new one. 8imilarly, in (110) siki points to 'any' bowlwhich one doesn't
really matter, the substance in question just has to go in something lest it make a mess.
(110) k-o talu dea i tena si ki popo
lkk-28o put go lOc lOc any bowl
'you put [it] inside a bowl'
ln (111) siki 'any' quantilies the number ol people the crocodile might eat.
(111) k-ara masi vana-thi mate-a t-i k-e sakai ni ghani mate-a si ki tinoni.
lkk-!ll must shoot-1k die-O:!8o kl-lOc lkk-!8o one olN bite die-O:!8o any person
'they must shoot it [a crocodile] dead belore it bites dead a person.'
1he speaker could have said balu 'some' orheaven lorbidleoni 'all, every', but
presenting the best case scenario (which in this case is still pretty bad), she chose siki
'one, any'.
As seen earlier in (108), plural can also be marked by the plural particle leoni 'all,
every'. 1he quantilier leoni can also emphasise the plurality ol the subject, as in (112).
(112) !ghamu na l eoni tinoni k-amu rongo mai1
ll:2ll Ak1 all person lkk-2ll listen come
'lveryonelisten up!'
62
oiven the presence ol the plural pronouns, ighamu 'll:2ll' and k-amu 'lL1-2ll', (112) has
abundant 'plural' marking. ln light ol this, it might seem sullicient to leave the marking
ol plurality to the pronouns, but the quantilier leoni 'all' is added lor emphasis.
1he medial quantilier balu 'some' is also used to mark plurality. ln (11!) balu
'some' is in agreement with vani-ra 'uA1-O:!ll' and modilies gari 'child'.
(11!) ami-tu dea bosa vani-ra bal u gari laka BMai . . . C
1lX.ll-lAL go tell uA1-O:!ll some child more come
'we went and told some more boys, come . . . "'
Although not all (leoni) the boys are addressed, balu 'some' indicates that there were
more than one.
1hough sometimes apparently redundant, plural can also be marked by
numerals. 1he subject relerence pronoun in (114) already indicates two persons with
the dual marker -ko 'uL', but that does not prevent a speaker lrom lurther specilying
that there are two people with a numeral.
(114) ara-ko ruka na ghaoka
!ll-uL two Ak1 woman
'they two two women'
linally, plural is marked by demonstratives:
(11) *ra-ko g-ena na ghaoka di ni
!ll-uL llc1-!8o:8AY Ak1 woman ulm:ll
'1hese two women said . . . '
As (11) shows, there is not always agreement between person and number. Quote
margins and relative J interrogative pronouns have only singular lorms. lere the
subject relerence pronoun and demonstrative are both plural, but the quote margin g-
ena 'llkl-!8o.say' is singular.
1here is another example ol lack ol agreement ol number in (116). 1he
6!
unwanted things (ngao-i 'want-O:!ll') are the same as the staying thing, which is
relerred to by the third person singular subject relerence pronoun (t-e mono = k-e mono).
(116) 9a tha t-e mono t-ami teigha na ngao-i ba k-e mono.
Ak1 what kl-!8o stay kl-1lX.ll Nlo Ak1 want-O:!ll lL1 lkk-!8o stay
'what stayswhat we don't wantwill stay.'
1he interrogative J relative pronouns thi J tha are unmarked lor number. 1hey do,
however, always take a !8o subject relerence pronoun, even when the relativised noun
phrase is plural, as in (116).
Agreement ol number with leoni 'all' is evident in the examples in (11).
(11) a. ami-ko kere-ra na l eoni igha t-ami-ko lavi
1lX.ll-uL burn-O:!ll Ak1 all lish kl-1lX.ll-uL caught
'we cook the lishes that we caught'
b. l eoni ghai doku ara logho vua-dira doku
all tree good !ll have lruit-l8:!ll good
'all good trees have good lruit(ll)'
ln (11)a. the !ll object pronoun sullix -ra is in agreement with plural leoni 'all'. ln
(11)b. leoni 'all' is in agreement with the !ll subject relerence pronoun ara.
Another way to mark everything without exception is (so)-soko '(kul)-linish', as
in (118).
(118) i gei ra so(soko na be i leo ni pilu
ll:!ll kul-linish Ak1 pig lOc inside olN lence
'all the pigs inside the lence'
keduplicated sosoko'linish' intensiliedis used with pronominal relerence while leoni
'all' modilies nouns.
!. !. ! Nomi nal modi li ers
1here are an interesting variety ol nominal modiliers, sometimes behaving as
adjectives and at other times as stative verbs. 1here are instances ol nominal modiliers
64
that appear rather adjectival lound lollowing the noun they modily. lxamples are given
in (119).
(119) a. e mai na kau asi
!8o come Ak1 dog wild
'a wild dog is coming'
b. siki vae sul e
any house big
'a big house'
c. na vunga vul ogo
Ak1 hair black
'black hair'
ln these examples the nominal modiliers asi 'wild', sule 'big', and vulogo 'black' lollow
the nouns they modily.
Olten, the same nominal modilier can be used as a stative verb. consider the
examples in (120).
(120) a. e asi na kau
!8o wild Ak1 dog
'wild dog' (lit., 'it is wild the dog')
b. @honi sul e-a.
make big-O:!8o
'make a big one.'
c. e vonu t-i na tagi
!8o lull kl-lOc Ak1 tank
'the tank is lull'
d. e bi thi na beti
!8o cool Ak1 water
'the water is cool'
e. e vonu t-i na niulu t-e mo-mono i beti na vua
!8o many kl-lOc Ak1 year kl-!8o kul-stay lOc river Ak1 crocodile
'the crocodile stayed many years in the river'
compare (119)a. to (120)a. ln (119)a. asi 'wild' lunctions as a direct nominal modilier
6
while in (120)a. the same word lunctions as head ol a verb phrase.
6
1he nominal
modilier sule 'big' in (119)b. takes on verb-like properties when it is inllected with an
object indexing sullix in the serial verb construction in (120)b. 1he one class ol words
that lunctions only as nominal modiliersand not as stative verbsis colours.
ln (121) examples ol compound nouns modilied by adjectives" are given.
(121) a. gari ghaoka ki ki
child lemale small
'small girl'
b. 0 leo-mu na inu beti bi thi 5
!8o want-l8:28o Ak1 drink water cool
'would you like a drink ol cool water7' (lit., 'ls your desire a cold drink7')
oiven the variable nature ol these wordssometimes behaving as verbs and at
other times as adjectives (compare (120)d. to (121)b.)they are grouped under the
banner 'nominal modiliers'.
!. !. 4 uemonstrati ve
1he demonstrative olten co-occurs with some lorm ol the article.

1here are
two basic constructions in which demonstratives occur. 1he lirst is na N ulm (e.g., na
mane deni 'Ak1 man ulm, this man'). 1he second is a ulm na N (e.g., a deni na ghai 'Ak1 ulm Ak1
stick, this is the stick'). 1here are proximal (near), medial (unmarked lor distance,
neither here nor there"), distal (lar), singular, and plural demonstratives.
8
1he lengo
demonstratives provide spatial and temporal relerence. 1he set ol demonstratives is
provided in (122).
6
8ee koss (2004:0) where he calls these stative or 'adjectival' verbs a subclass ol intransitive verbs.

uryer (200a:162) notes ln such languages in which delinite articles co-occur with demonstratives or
occur in dillerent positions in the noun phrase, there is little justilication lor a grammatical notion ol
determiner." lt is lor this reason that 'determiners' are not discussed in this grammar.
8
lynch, koss, crowley (2002b:2) make a three-way distinction among demonstratives between near
speaker, near hearer, and near neither.
66
(122) 8ingular llural
lroximal de(ni) di(ni)
medial nga ngi
uistal deri
Note that deni and dini are sometimes shortened to de (see (124)) and di (see (126))
respectively. Also note that there is no plural lorm ol the distal demonstrative.
ln (12!) the narrator is keen to point out that it is the 8aturday close at hand,
the 8aturday that is a day away (this text having originated on the lriday preceding),
on which the least will take place.
(12!) arere deni ba k-a ghali-a po na thara.
8aturday ulm:8o lL1 lkk-1lN.ll make-O:!8o llm Ak1 least
'1his 8aturday we will make a least.'
1he phrase arere ke mai '8aturday, the one that will come' was available to the speaker,
but she chose to say arere deni 'this 8aturday' to indicate temporal deixis.
ln (124) the narrator is pointing at a photograph ol people loading copra bags
into a canoe, explaining what the dillerent participants were doing when the picture
was taken. 1he demonstratives de J deni specily which person in the picture is being
relerred to. 1his was accompanied by pointing at each 'man' in the picture.
(124) 9a mane de e sake vani-a sakai na mane deni ni tura.
Ak1 man ulm:8o !8o lilt uA1-O:!8o one Ak1 man ulm:8o olN carry
'1his man lilts one lor this other man to carry.'
8ince the photo was within pointing distance ol the speaker, the images ol the men
were 'proximate' to the speaker, as such the same lorm ol the demonstrative is used lor
both. lt was the physical pointing that accompanied the spoken demonstratives that
distinguished the relerents, not the demonstratives alone.
lxample (12) demonstrates the plural lorm ol the proximal demonstrative and
6
its agreement with the third person dual subject relerence pronoun ara-ko.
9
(12) *ra-ko g-ena na ghaoka di ni , . . . "
!ll-uL llc1-!8o:say Ak1 woman ulm:ll
'1hey two, these women, said, . . . "'
1he shortened lorm ol the plural proximal demonstrative dini is shown in (126).
(126) iko-ira di a sakai na udu la(laka
uL-ll:!ll ulm Ak1 one Ak1 lriend kul-also
'these two were best lriends'
while the distal counterpart ol denithat is, deriis attested by lengo speakers,
it appears only once in the data. 1his occurrence is given in (12).
(12) ! vi ga na vae ni minista5 ! deri .
lOc where there Ak1 house olN minister lOc ulm:8o
'where is the minister's house7' '1hat way.' or '1hat [lar] location.'
1he house was out ol sight to both speakers. Accordingly, its location was indicated
with the distal demonstrative.
1he nga = ngi pair ol demonstratives present a special case. lt can be used ol
something or someone unmarked lor proximity. kather, nga = ngi relers to something or
someone medial. lxample (128) presents the case ol a person speaking on behall ol a
group ol people (the community ol l.) who were responsible lor the preparation and
distribution ol least lood to lour other communities.
(128) !ghami nga ba k-ami ne athe na tha t-ami tangomana vani-ghamu.
ll:1lX.ll ulm:8o lL1 lkk-1lX.ll lmll give Ak1 kll kl-1lX.ll able uA1-O:2ll
'we will give what we are able to you.'
None ol the other members ol community l. were with the narrator at the time she
was speaking (she being alone on a stage), but the demonstrative nga includes them
9
Note, however, that the third person singular quote margin g-ena does not agree with the third person
dual subject relerence pronoun ara-ko in number. As is discussed on page 206, there are no plural
quote margins in lengo.
68
with the speaker. lt would not have been appropriate to use either deni nor deri in this
context. 1he speaker is not relerring to 'this we' as opposed to 'that we', she is relerring
to a group scattered throughout her audience (therelore 'medial') identilied by their
membership in a community (specilic and demonstrable).
ln (129) the narrator was with another person, but only the narrator awoke, his
companion remained asleep. 1hus the use ol the demonstrative nga to specily himsell J
single himsell out.
(129) 0 roropo u rai mai inau nga m-u bere dea
!8o morning 18o awaken come ll:18o ulm:8o cON}-18o look go
'ln the morning l woke upjust meand looked around'
1o say inau deni would not work: the speaker is not relerring to 'this me' as opposed to
'that me'. lut he is relerring to himsell and not his companion. what nga does here is
denote a specilic 'me', but 'me' at a distance, in this case, distant in time and space lrom
the speech event,
60
therelore 'medial'. consider (41) again (a nga inau 'l choose that
one'), where the speaker was selecting lood at the market. 8he chose one specilic lruit
lrom among many, but none ol them were in her hand (it is tabu 'lorbidden' to handle
lood at market belore it is purchased). ln that sense nga is 'medial': not so lar away as to
be out ol sight (deri), but not in hand either (deni).
lased on these examples, one might well ask whether nga is a demonstrative
reserved lor people. lxample (1!0)b. indicates otherwise. 1he lirst sentence, (1!0)a., is
included to provide some context lor (1!0)b. which shows nga specilying an idea, a
possible ('medial') course ol action, rather than a person.
60
1his example comes lrom a text in which the narrator described the events ol the previous day in the
capital city, some 80km distant.
69
(1!0) a. Ma k-ami-tu mai soni ga i T. t-ami-tu ghua vulo.
cON} lkk-1lX.ll-lAL come throw there lOc 1 kl-1lX.ll-lAL il willing
'we will come throw them [sticks lor ralters] ashore at 1. il we are willing.'
b. Teigha nga ba k-ami-tu lebo saliu
Nlo ulm:8o lL1 lkk-1lX.ll-lAL lloat overshoot
'larring that, we will lloat larther downstream'
1he speaker was not sure which course they were going to take. An in between, neither
here nor there, demonstrative is most litting in such a case. lurther research could
explore whether the use ol deni or deri would have increased or decreased the likelihood
ol the second option in (1!0)b., though the modal verb teigha 'Nlo' is never heard with a
demonstrative other than nga.
lxample (1!1) comes as the second thing" to do in a recipe lor making tapioca
pudding.
(1!1) 6uke na agutu nga, na tavu niu.
second Ak1 thing ulm:8o Ak1 lind coconut
'1he second thing is to lind some coconut.'
1he interesting thing is thatlrom an outsider's perspectivethis is not precisely the
second step ol the recipe: the author indicates that alter you go get some tapioca lrom
the garden you bring it home, peel the skin, wash it, scrape it, and put it inside a bowl.
1here are live things to do with the tapioca, live sub-steps between the lirst" thing ol
getting tapioca and the second" thing ol linding coconut. And once a coconut is lound,
there is another set ol sub-steps that lollow: husking it, breaking it, scraping out the
meat, and squeezing out the milk. 8o, again, lrom a non-lengo speaker's perspective,
linding a coconut may well be the second main thing to do, but it is not precisely the
second thing done. A question lor lurther research is whether the medial
demonstrative nga downplays the precision ol the thing demonstrated, and whether
0
the proximal demonstrative deni would be used il the text was organised dillerently.
1here is some signilicant specilication in (1!2). 1he speaker, alter recounting
how he and his lriends had gone to the bush that morning to get some sticks to build a
kitchen, pointed to them lrom the place where he was telling the storyabout meters
awayand said:
(1!2) Talu-i na koga nga ga tena puku ni gavigha nga
leave-O:!ll Ak1 stick ulm:8o there lkll underneath olN apple ulm:8o
'we lelt those sticksthe ones here under that apple tree'
1he sentence continues:
(1!!) m-ami-tu mai atheathe po t-i de ga.
cON}-1lX.ll-lAL come rest llm kl-lOc ulm:8o here
'and we came to rest a bit here.'
lt seems, based on (1!2) and (1!!), that de3ni' ga and nga ga are somewhat distinct, the
'here' ol the sticks is not so lar lrom the 'here' ol the immediate location ol the
conversation, but lar enough to be expressed dillerently. lxample (128) and, to a lesser
extent, (1!0) indicate a similar thingdeni would not work in those examples.
As (1!4) shows, nga can also indicate 'there' in addition to 'here' (which is
normally indicated by deni).
(1!4) B+ tavu-a a ,.C B* nga.C
18o lind-O:!8o Ak1 l Ak1 ulm:8o
'l'm looking lor l.' '8he's over there.'
1he person who responded a nga did not know exactly where l. was, but indicated that
she was 'over in that general direction'.
lxamples (1!) and (1!6) merit some lurther explanation. 1he construction nga
inau is a device lor the subject to 'hide' behind (this is dillerent lrom a nga inau in (41)
[by virtue ol the article a] and inau nga in (129)). Notice that the medial demonstrative
1
comes belore the pronoun it specilies (only inau 'll:18o'), which is opposite to the
normal noun J pronoun - demonstrative order. ln a culture where speaking certain
people's names is tabooin lact, where these people (relatives by marriage) are
relerred to as tabu 'lorbidden'there are also situations in which one leels the need to
'hide' onesell. 8o in (1!) the speaker (8) is requesting l. to go ask l. lor some betelnut
lor himsell (8).
(1!) #. k-o dea thuge-a na bua i te a /. ena nga i nau.
l lkk-28o go ask-O:!8o Ak1 betelnut lOc lOc Ak1 l !8o:say ulm:8o ll:18o
'l., go ask l. lor some betelnut, so says the over-here-l.'
Now, 8 doesn't want l. to go to l. and say, lor example, ley l., 8 wants some betelnut,
have you got any lor him7" 1hat would blow 8's cover. lnstead, 8 uses nga inau 'the-
over-here-l'and expects l. to do the same
61
to preserve his 'hidden' identity.
lxample (1!6) is somewhat more complicated. lt comes lrom a text in which the
speaker (the object ol the dative vani-a) was being spoken ol by a third party. 8o A (the
speakers ol the reported speech ['these two women' in the text]) was telling l (l. in the
text) to go tell something to c (the person telling the story).
(1!6) [*ra-ko g-ena na ghaoka dini,] B/lis, ngata [vani-gho 4., k-o dea
!ll-uL llc1-say:!8o Ak1 woman ulm:ll please strong uA1-O:28o l lkk-28o go
A l
tugu vani-a] [nga i nau] veghe na thara [t-e bosa] ia
story uA1-O:!8o ulm:8o ll:18o like Ak1 least kl-!8o tell ulm
l
con't
c c
[ghami tena agri ighami,C g-i geia ena.]
ighami teigha na agri ighami g-i geia ena
ll:1lX.ll Nlo Ak1 agree ll:1lX.ll llc1-lOc ll:!8o say:!8o
A
'1hese two women said, llease, we strongly ask you l, go tell me-over-here
that the least she's talking aboutwe don't agree with it," so they said.'
61
8 expects l. to say something like, ley l., nga inau wants some betelnut, have you got any lor him7"
2
1he construction nga inau specilies a participant who does not want to be named in a
story she is telling in which she hersell is involved. 1his squares with the medial,
neither-here-nor-there nature ol the demonstrative nga.
A linal pronoun ia is included in the discussion ol demonstratives because, even
though it may be an incomplete lorm ol i3gei'a 'll:!8o', it is used as a demonstrative.
62

1he demonstrative ia in (1!) is being used to reler to a specilic childthe one playing
on the ground just over there.
(1!) * thi na ethe a daemu i a5
a thi na atha-a a dae-mu ia
Ak1 kll Ak1 name-l8:!8o Ak1 child-l8:28o ulm
'what is your child's name7'
1his could be more woodenly translated, 'who is the name ol your child, that one7'
62
!a is also a demonstrative in 8olomon lslands lijin.
!
4 Vlkl8 ANu Vlkl llkA8l8
1he verb stands at the heart ol the lengo clause. ln lact, a clause can be made up
ol a single verb (e.g., imperative Mai! 'come!' and prohibitive Tabu! 'uon't!').
A bare intransitive verb is the basic lorm ol the lengo verb. keduplication ol a
verb's stem-initial cV can have one ol two ellects: it can turn the verb into a noun or
signal intense J continuous action. 1ransitive verbs are derived lrom intransitive verbs
either with a -ci sullix (where c is a non-predictable consonant) plus an object pronoun
sullix, or with just an object pronoun sullix. 1he set ol verbs that take the -ci sullix is
quite small. 1he object pronoun sullix joins to the linal verb ol the verb phrase.
lengo verbs are distinguished as either being A-type or L-type (koss 2004:04)
based on the role ol the subject in corresponding transitive J intransitive lorms. with
A-type verbs the subject ol the intransitive lorm corresponds to the agent ol the same
verb's transitive lorm. with L-type verbs the subject ol the intransitive lorm
corresponds to the undergoer ol the verb's transitive lorm as a new agent is supplied.
1ense J aspect J mood (1Am) is not indicated on the verb but with a set ol
prelixes on the near obligatory subject relerence pronoun. oiven the close connection
between 1Am and the verb, the 1Am - subject relerence pronoun construction will be
discussed in this section. ln conjunction with the 1Am-marked subject relerence
pronouns, there is a set ol semi-auxiliary verbs that contribute to the varieties ol 1Am.
8erial verb constructions play a signilicant role in lengo verb lormation. 8erial
verb constructions express meanings ol:
1. direction (V
1
(movement) V
2
(direction)),
2. sequence (V
1
(movement) V
2
(action), V
1
(become) V
2
(state)),
!. cause (V
1
(action) V
2
(result)),
4
4. manner (V
1
(action) V
2
(manner)),
. ambiance (V
1
(action) V
2
(sub-event)),
6. comitative (V
1
kolu-object),
. dative (V
1
vani-object),
8. instrumental (V
1
ghini-object),
9. modal (V
modal
na V(-O)).
linally, there are two means available to the lengo speaker to change the
valence ol a verb: the causative sullix -caghini (again, where c is a non-predictable
consonant), and the reciprocal prelix vi-.
4. 1 A-type and L-type verbs
lengo verbs are distinguished as either being A-type or L-type based on the role
ol the subject in corresponding transitive J intransitive lorms.
6!
ln the case ol A-type
verbs, verbs that are inherentl< d<namic (koss 1998:21), the subject ol the intransitive
lorm corresponds to the agent ol the verb's transitive lorm. L-type verbs, verbs that are
inherentl< stative or that when intransitive imply some unmentioned agent" (koss
1998:22), diller in that the subject ol the intransitive lorm becomes the undergoer ol
the verb's transitive lorm as a new agent is supplied. compare (1!8)a. and b.,
intransitive and transitive examples ol a lengo A-type verb.
64
(1!8) a. k-ami-tu l ebo tave mai i nughu
lkk-1lX.ll-lAL lloat current come lOc river
8 V
'we will lloat downstream in the river'
b. m-ami-tu dea l ebo-i laka t-i
cON}-1lX.ll-lAL go lloat-O:!ll also kl-lOc
8 V-O
'and we'll also lloat them [sticks]'
6!
koss (2004:04) uses the terms A- and L-type to describe these verbs (lor 'actor' and 'undergoer'
respectively). Other Oceanists (cl. uixon 1988:4, 200) call them A and O verbs ('transitive subject' and
'object'). 1he phenomenon they are describing is one and the same.
64
8" and O" in (1!8) mark the subject and object indexing elements ol the sentences. loth subject and
object are indicated with lull noun phrases elsewhere in the text.

1he verb lebo 'lloat' in (1!8)a. is intransitive as there is no object argument in the
sentence. when a second argument like -i 'O:!ll' is added, as in (1!8)b., the subject ol
the intransitive remains the agent ol the transitive verb lorm and the added argument
becomes the undergoer.
Now consider (1!9)a. and b., examples ol a L-type verb in intransitive and
transitive sentences.
(1!9) a. e pul u [na ghola]
6
!8o roll.up Ak1 pudding
8 V Nl
8
'the pudding is [being] rolled up'
b. e pul u-i [na ghola] [na ghaoka]
!8o roll.up-O:!ll Ak1 pudding Ak1 woman
8 V-O Nl
O
Nl
8
'the woman rolls up the puddings'
ln (1!9) na ghola 'the pudding', the subject-as-undergoer ol the intransitive lorm ol pulu
'roll.up' in (1!9)a., becomes the object-as-undergoer ol the transitive lorm pulu-i 'roll
them up' when the agent na ghaoka 'the woman' is added in (1!9)b. lt is important to
note that (1!9)a. does not constitute a passive lormit is active in every sense. lynch et
al. (2002:4) note that, lassive constructions are only very rarely encountered in the
languages ol melanesia" and they are absent lrom lengo. Nor, it should also be said, is
(1!9)a. rellexive. 1here is no second argument present in (1!9)a., the subject noun
phrase na ghola 'the pudding' is relerenced by e '!8o'. lxample (1!9)a. contains a single
argument: na ghola 'the pudding'. ly way ol contrast, (1!9)b. presents two arguments:
the subject na ghaoka 'the woman' relerenced by e '!8o' and the object na ghola 'the
pudding' relerenced by -i 'O:!ll'. 1he noun phrase object na ghola 'the pudding' is
6
[8quare brackets] are used here to demarcate the subject and object noun phrases.
6
indexed by -i 'O:!ll' and not e '!8o'. koss (2004:04) notes that, ln some canonic
[lrotoOceanic] languages there is a subclass ol intransitive verbs which . . . denote a
one-participant semantic relation with a potential second participant, but the subject
ol the verb is the undergoer." lt is the lack ol such a construction in lnglish that
requires the misleading passive gloss in (1!9)a.
4. 2 Verb deri vati on and i nll ecti on
1he lengo verb is prone to a variety ol derivational processes while only a single
inllectional process is possible.
4. 2. 1 ueri vati on
Verb are subject to two derivational processes. 1he lirst process derives nouns
lrom intransitive verbs by stem-initial cV reduplication. A second process involves the
derivation ol transitive verbs lrom intransitive verbs with a transitivising sullix andJor
an object pronoun sullix.
1he derivational process ol stem-initial cV reduplication ol a verb changes it
into a noun. 8ome examples ol this are given in (140).
(140) verb noun
a. digi 's.t. is closed' di(digi 'door'
b. tughu 's.o. changes' tu(tughu 'change (monetary)'
c. dea 's.o. goes' de(dea 'program, event, proceedings, goings on'
d. rongo 's.o. hears' ro(rongo 'news'
ln (141) digi 'close' is shown in a number ol distinct lorms.
(141) a. e di gi na vae
!8o close Ak1 house
'the house is closed'
b. ba k-a di gi -a noda na de(dea
lL1 lkk-1lN.ll close-O:!8o l8:1lN.ll Ak1 Nk-go
'we will close our program'

c. na di (di gi ni vae
Ak1 Nk-close olN house
'the house door' (lit., 'the closer ol the house')
ln (141)a. and b. digi 'close' is used as a verb in intransitive and transitive clauses
respectively. ln (141)c., however, digi 'close' is reduplicated and nominalised. ln
instances such as this, reduplicated cV is considered a nominaliser ('Nk', although
compare verb derivation (page 1!) where stem-initial cV reduplication is used in a
process ol verbilication).
1he second derivational process involves intransitive verbs becoming transitive.
8ome transitive verbs are derived lrom intransitive lorms by means ol a transitivising
sullix -ci ('1k', where the consonant c is non-predictable).
66
lollowing the transitiviser is
some lorm ol the object pronoun sullix. 1hose verbs that undergo this derivational
process do not appear to constitute a large set. 8ome examples include:
(142) a. bere 'see' bere-ngi-a 'see it'
b. ghu 'shout' ghu-vi-a 'shout at J to s.o.'
c. mataghu 'lear' mataghu-ni-a 'lear it'
d. rongo 'hear' rongo-ni-a 'hear it'
6
e. sara 'arrive' sara-vi-u 'arrive at me'
l. sivo 'breeze' sivo-li-a 'the wind blows something'
g. tavu 'search' tavu-ti-a 'lind s.t.'
h. uru 'dive' uru-vi-a 'dive lor it'
66
lawley (19!:114) remarks, ln passing it should be noted that in many daughter languages the linal
consonant ol the lOc [lroto Oceanic] verb stem has been reassigned to the transitive sullix . . . 1he
morpheme cuts in example sentences thus olten isolate sullixes with the shapes -Ci and -Caki3ni'
rather than [lOc] -i or -aki3ni'."
6
1he verb rongo 's.o. hears' has two transitive lorms: rongo-ni-a and rongo-a. 1he lorm with the
transitivising sullix is lar more common. ln lact, the lorm without the transitivising sullix, rongo-a,
occurs only once in the data, in the sentence:
1his is a complex sentence involving three relerents (mother ol child [grown, married], [male] child J
spouse (the one who heard his wile call him), and [lemale] spouse), and a relative clause. oiven the
complexity and that this is the only example ol this lorm ol transitive rongo 'hear', it is dillicult to say
with any certainty why in this instance rongo 'hear' does not take the transitivising sullix.
8
m-e rongo-a dae-a t-e ghu-vi dea-a a tau-a nga
cON}-!8o hear-O:!8o child-l8:!8o kl-!8o shout-1k go-O:!8o Ak1 spouse-l8:!8o ulm
'and he, the [other woman's] child, heard her, she who shouted to him, his spouse'
when it is said that the consonant ol the transitivising sullix -ci is non-predictable, this
can be so even lor the same stem. when bere 'see' is derived to a second person singular
transitive, the lorm ol the -ci sullix is -ni (i.e., bere-ni-gho 'see you') as in (14!)a.
(compare (142)a.). 8ome examples ol these in context are given in (14!).
(14!) a. 4a k-u ghe bere-ni -gho.
lL1 lkk-18o continue see-1k-O:28o
'l'll see you later.'
b. m-u ghoni-a na pou t-ami uru-vi -a tena rodo
cON}-18o prepare-O:!8o Ak1 b.d.m. kl-1lX.ll dive-1k-O:!8o lOc night
'l prepared the beche-de-mer that we dove lor during the night'
lor contrast, consider the examples in (144).
(144) a. m-u bere dea i vunga ni bake raisi
cON}-18o look go lOc on.top olN bag rice
'l looked away on top ol the rice bag'
b. m-ami-ko dea uru i T.
cON}-1lX.ll-uL go dive lOc 1
'we two go dive at 1.'
lere bere 'look, see' and uru 'dive' are intransitive.
Other verbs do not take the transitivising sullix -ci, but are nonetheless
rendered transitive with an object sullix. lirst consider intransitive lebo 's.o. J s.t. lloats'
in (14).
(14) )-ami-tu l ebo tave mai i nughu.
lkk-1lX.ll-lAL lloat current come lOc river
'we'll lloat with the river current.'
ln (146), however, lebo 'lloat' becomes transitive with the addition ol the inanimate
object pronoun sullix -i.
9
(146) m-ami-tu dea l ebo-i laka t-i
cON}-1lX.ll-lAL go lloat-O:!ll also kl-lOc
'we'll also lloat them [sticks lor ralters]'
lere there is no evidence ol the transitivising sullix -ci.
lt was mentioned above that the object pronoun sullix lollows the transitivising
sullix. 1his is not always true. 1here are instances ol verb serialisation in which the
lirst verb is a derived transitive (as evinced by the -ci transitivising sullix) and the
second verb carries the object sullix. 1he examples in (14) demonstrate this.
(14) a. t-e ghu-vi dea-a a tau-a
kl-!8o shout-1k go-O:!8o Ak1 spouse-l8:!8o
'she shouted to her spouse'
b. ba k-amu bere-ngi ghi l aghana-ra tena agutu t-ara ghali
lL1 lkk-2ll see-1k understand-O:!ll lOc thing kl-!ll do
'you will understand them by seeing the things they do'
c. k-ara masi vana-thi mate-a t-i k-e sakai ni ghani mate-a
lkk-!ll must shoot-1k die-O:!8o kl-lOc lkk-!8o one olN bite die-O:!8o
siki tinoni
any person
'they must shoot it dead belore it bites a person to death'
1his shows that, in lact, the object pronoun sullix joins to the linal verb ol the verb
phrase but that the lirst verb still has a transitive lorm.
Ol some interest are the examples in (148) ol nouns that appear to derive
directly to transitive verb lorms (lollowing the -ci sullix pattern).
(148) a. ghavi 'tongs' ghavi-thi-a 's.o. pinches s.t.'
b. ithu 'nose' ithu-ri-a 's.o. snills s.t.'
1hey do not go through the stem-initial cV reduplication noun-to-verb derivation
process belore undergoing the -ci sullix intransitive-to-transitive derivation process.
80
4. 2. 2 lnll ecti on
1he single inllectional process lengo verbs undergo is that ol reduplication to
indicate intense andJor continuous action.
68
keduplication takes two lorms with a two
syllable word (i.e., c
1
V
1
.c
2
V
2
). 1he lirst involves reduplication ol the lirst syllable, c
1
V
1
-.
1he second and more complex lorm involves reduplication ol both syllables but with
the omission ol the second consonant, c
1
V
1
V
2
-. lt is not clear what conditions prompt
the use ol either the lirst or the second type ol reduplication with a given word.
1he lirst type ol reduplication, c
1
V
1
-, is shown in example (149).
(149) soko m-ami-tu biku de(dea m-e dea roropo
linish cON}-1lX.ll-lAL sleep kul-go cON}-!8o go morning
'then we lew slept on and on and it went to morning'
1his example comes lrom a story in which two men who were sleeping were joined by a
thirdwho laid down right between themin the middle ol the night. ln the morning
the third man was gone, along with one ol the other two men's baskets. keduplicated
dedea 'go go' here has the dual sense ol continuous and intense action: the men slept
until morning (continuous) but they also slept so deeply (intense) that they didn't
awaken when the third man got up and stole the basket (which is surprising given that
he was right between them). 1he duration has inherent within it a measure ol intensity.
1he non-reduplicated lorm ol dea 'go' is used ol time" going to morning. 1his does not
68
lvens (19!!:166) states ol lughotu that, 1he general idea conveyed by reduplication is that ol
intensilication ol meaning: hiohiro 'to search earnestly', but this is not always the case, and many
verbs occur only in reduplicated lorm", and ol the language ol 'llorida' (oela) that, 1he ellect ol
reduplication is to increase the sense ol lrequency ol the action or to intensily the meaning".
(19!:1102). more recently, llevins (200!:499) reports that in lughotu and cheke lolo (a Northwest
8olomonic language lrom 8anta lsabel) keduplication ol a verb stem marks emphasis, intensity,
lrequency, or duration", and lynch et al. (2002:44) remark that, keduplication is almost universally
used in Oceanic verbal morphology, as well as in noun derivation. lt expresses a wide range ol
meanings, including randomness ol action, repetition, and plurality ol actors and patients." 8ee also
8apir (1921) in lootnote 1!.
81
have the continuous or intense sense since it is part ol the normal, expected order ol
things.
1he same continuous and intense action ol a reduplicated verb is in (10).
(10) ga(gara na vua ga(gara na da(dae m-e ga(gara ngata na
kul-pull Ak1 crocodile kul-pull Ak1 kul-child cON}-!8o kul-pull strong Ak1
dadae me gara tine me gagara laka na vua
da-dae m-e gara tina-a m-e ga-gara laka na vua
kul-child cON}-!8o pull mother-l8:!8o cON}-!8o kul-pull also Ak1 crocodile
'the crocodile pulled, the child pulledand the child pulled strong, and his
mother pulled, and the crocodile also pulled'
ln this story a mother going to bathe in the river was attacked by a 4-metre-long
crocodile. ler adult child came to help her and a tug-ol-war between child and
crocodile ensued, with the woman in between. Note that every time the crocodile or the
child 'pulls' that a reduplicated lorm ol the verb is used but that when the mother pulls
the verb is not reduplicated. 1he crocodile and the child are pulling continuously and
vigourously, the mother, losing courage and strength (not to mention a lair amount ol
blood), is able to pull but on a dillerent level ol magnitude lrom the others. 1he
duration ol the crocodile and the child's pulling has inherent within it the intensity ol a
lile and death struggle. 1he story continues:
(11) de(dea m-e ngata na da(dae m-e gara-a lithi soni-a
kul-go cON}-!8o strong Ak1 kul-child cON}-!8o pull-O:!8o out lrom-O:!8o
na vua me gara iti maia mo a tine
na vua m-e gara iti mai-a mo a tina-a
Ak1 crocodile cON}-!8o pull up come-O:!8o more Ak1 mother-l8:!8o
'as things went the child was strong and he pulled her out lrom the crocodile
and he pulled her up come more his mother'
lere the non-reduplicated lorms ol gara 'pull' indicates more conclusive, non-
continuous action.
82
1he second type ol reduplication, c
1
V
1
V
2
-, is shown in (12).
(12) M-e thai-a mo na dani t-ara-ko ghunughunu-a m-ara-ko
cON}-!8o arrive-O:!8o more Ak1 day kl-!ll-uL plan-O:!8o cON}-!ll-uL
vui (vuni -a na visaghiri. Aoku na igha t-i t-e vui (vuni -a.
kul-start-O:!8o Ak1 competition good Ak1 lish kl-lOc kl-!8o kul-start-O:!8o
'1he day arrived that they two had planned lor and they started the
competition. Ol, the lish actually started it.'
later in the story, when it was the turtle's turn to take the lead, vuni 'start' is not
reduplicated.
(1!) m-e atheathe ki(kiki po na igha ga m-ara-ko g-e sakai ni
cON}-!8o rest kul-small llm Ak1 lish there cON}-!ll-uL llc1-!8o one olN
vuni -a po t-i na visaghiri
start-O:!8o llm kl-lOc Ak1 competition
'and the lish rested a bit there belore they two started the competition'
1he dillerence in meaning between the two lormsreduplicated and notis subtle but
signilicant. 1he turtle and the lish engaged in a game ol lollow-the-leader. 1he lish got
lirst shot at trying to outdo the turtle. 1he competition was the lish's to lose, it had
only to lind one thing that the turtle couldn't do and it was the winner. lut try as it
might, the more agile lish couldn't outdo the turtle. 1he lish got the lavoured starting
position in the competition, but surprisingly it couldn't capitalise on it. 1he
reduplicated lorm ol vuni 'start' indicates the measure ol importance given to the lirst
start. 1his is in contrast to the non-reduplicated lorm ol vuni 'start' used with the turtle
in (1!). while it is true that the turtle got its turn to lead in the competition, its start
was anticlimactic in comparison to the lish'sit does not have the intensity ol the lirst
'start'. 1he point here is that there is a contrast between the two 'starts', that ol the lish
and the turtle, and the lish's start is intensilied, given greater importance, by means ol
8!
the reduplication.
4. ! Val ence changi ng constructi ons
1he number ol arguments a verb can take may be altered with valence changing
mechanisms lor: causative, rellexive, and reciprocal. while the causative and reciprocal
involve verb inllection, the rellexive makes use ol a lorm ol the direct possessor
pronoun joined to a lree stem, tibo 'klll'.
69
4. !. 1 causati ve -caghi ni
1he causative in lengo is lormed by adding the sullix -caghini, lollowed by an
object pronoun sullix, to an intransitive verb. 1he consonant c dillers with dillerent
roots, c is not predictable phonologically.
0
1he causative may be best viewed in
comparison to an intransitive lorm ol the same verb stem. lirst, consider an example ol
intransitive tumu 'lall' in (14).
(14) e tumu na vua ni niu
!8o lall Ak1 lruit olN coconut
'the coconut lell'
lere the subject-as-undergoer na vua ni niu 'the coconut lruit' is relerenced by the
subject relerence pronoun e '!8o'. compare this to example (1).
(1) na ara sivo-li-a m-e tumu-l aghi ni -a na vua ni niu
Ak1 wind breeze-1k-O:!8o cON}-!8o lall-cAL8-O:!8o Ak1 lruit olN coconut
'the wind blew it and it [the wind] caused the coconut lruit to lall'
ln (1) the valence ol the intransitive verb tumu 'lall' is increased with the causative
sullix -caghini and the object indexing sullix -a 'O:!8o'. As this example shows, when a
69
1he discussion ol the rellexive on page 4! covers the subject sulliciently and it will not be discussed
lurther here.
0
1he non-predictable nature ol c in -caghini is also the case with the transitivising sullix -ci.
Lnlortunately, the data do not provide an instance which would indicate whether c is the same lor
causative and transitive allixes lor a particular verb root.
84
causer (na ara 'the wind') is added as an argument to the L-type verb tumu 'lall', the
valence ol the verb is increased with the addition ol -caghini. with the addition ol a
subject-as-causer, the causee, in this case the coconut, is demoted to direct object.
1he similarity between the instrumental ghini and the causative -caghini is
intriguing. laspelmath and mller-lardey (2001:9) comment that, 1he instrumental
applicative exhibits a certain tendency to acquire a causal (reason) and a stimulus
lunction (like the lnglish preposition through), the latter in turn having allinities to
the directive applicative (e.g. in Oceanic languages, compare koss 1988:!-!)." lt
certainly seems to be the case in lengo that the instrumental has, in a more
grammaticalised lorm, taken on a causative lunction.
4. !. 2 keci procal vi -
ln lengo, reciprocal is marked by the prelix vi- 'klcl' and the plural object sullix
-i. loth na kau 'the dog' and na be 'the pig' are agent and patient ol ghani 'bite' in (16).
(16) na kau ma na be t-ara-ko vi -ghani -i
Ak1 dog cON} Ak1 pig kl-!ll-uL klcl-bite-O:!ll
'the dog and the pig bit each other'
1he agent is dual (tarako), the patient is plural (-i), and the prelix vi- indicates that
subject and object are the same relerent. 1he agent J patient relationship is more
dillicult to see in (1) as they are both subsumed in the pronouns ira 'll:!ll' and ara
'!ll', but vi- 'klcl' with plural subject and object pronouns makes the point clear without
needing to delineate specilic participants with lexical noun phrases. 1he reciprocity ol
action, that the members ol ara 'they' helped each other, is marked by vi-.
(1) *ra vi -thanga-i na mane ira i tena leghai.
!ll klcl-help-O:!ll Ak1 man ll:!ll lOc lOc garden
'1hose men help each other in the garden.'
8
ln (18) a curious instance ol the plural object sullix is lound in vi-leu-thi.
(18) *ra-ko leta t-i vini vi -l eu-thi 8 na leoni bona
!ll-uL not.know kl-lOc actually klcl-light-O:!ll Ak1 all time
ara-ko vi -doku-i po t-i i levu m-i levu.
!ll-uL klcl-good-O:!ll llm kl-lOc lOc side cON}-lOc side
'1hey didn't actually know how to light each other, they were always just good
to each otherlrom one side and the other side.'
1hat this is a reciprocal construction is conlirmed by the second reciprocal ol the
sentencethe goodness comes lrom both sides (i levu m-i levu 'lOc side cON}-lOc side')but
this is the only instance ol the plural object sullix -thi in the data. whether this lorm ol
the plural object sullix is unique to reciprocals remains a question lor lurther research.
4. 4 Verb phrase structure
1he basic lengo verb phrase is comprised ol a number ol constituent classes,
some obligatory and others not. 1hese are:
1. verb: obligatory, can stand alone in imperative clauses, is either dynamic or
stative, can carry the optional transitivising sullix (-ci) andJor the object
indexing sullix (in serial verb constructions the linal verb carries the object
sullix)
2. subject relerence pronoun: nearly obligatory (optionally present in
imperative clauses), lound belore the verb, carries 1Am marking and provides
a subject indexing component (in addition to subject Nl and emphatic
pronoun)
!. auxiliary: optional, lound belore the verb, either belore or alter the subject
relerence pronoun, contributes aspects ol 1Am
4. adverb: optional, lound alter the subject relerence pronoun, either belore or
alter the verb, indicates time, manner, and degree
. conditional: optional, lound between the subject relerence pronoun and the
verb.
1he position ol these constituents in the verb phrase is illustrated in (19).
(19) (ALX) 8kl (cONu) (ALX) (AuV) V(-ciJ-O) (AuV)
1his is the basic lengo verb phrase. 8erial verb constructions present an array ol
86
variants lrom the basic verb phrase. 1hese are discussed in section 4.8 (see page 1!ll).
lt is important to maintain a distinction between phrase level and clause level
constituents, especially in terms ol determining something such as a basic word
order". 1he phrase level subject and object indexing constituentsthe subject
relerence pronoun and the object sullixare considered as part ol the verb phrase and
are, lor the intents and purposes ol deciding on a basic word order", opaque at the
clause level.
1
4. lre-verb parti cl es
ln addition to the subject relerence pronoun with its 1Am prelixes, there are a
number ol 1Am auxiliaries that appear belore the verb. 1hese are ba 'lL1', boro
'impossible lL1', and bo 'lmll'. 1he conditional ghua 'cONu' and the adverb ne 'lmll' also
precede the verb. 1he auxiliaries ba 'lL1' and boro 'impossible luture' are lound belore
the subject relerence pronoun while the remainder are lound lollowing it (i.e., between
the subject relerence pronoun and the verb). 1he order ol constituents is given in (160).
(160) (ALX tense) 1Am prelix- 8kl (cONu) (ALX aspect) (AuV) V
{
ba
boro
}
{
g D
t D
E
k D
b D
}
ghua bo ne
4. . 1 1Am preli xes
1he role ol the subject relerence pronoun in relation to 1Am brings the discussion
to that part ol the grammar which can give some insight into how lengo speakers
conceive ol time: their attitude toward it and their dealings with it. lerhaps a lew
1
8ee page 166ll lor more discussion ol argument structure at the clause level.
8
episodes lrom my experience ol lengo daily lile will help shed some light on time and
the lengo attitude toward time.
1he lengo people belong to an island nation, 8olomon lslands. 1here are over
900 islands in 8olomon lslands and much travel is done by 6 to metre, outboard engine
powered, libreglass boats. l vividly remember the lirst time my lamily and l rode as
passengers in one ol these boats on a journey lrom the capital city loniara to the
community in which we were invited to live. we consulted with the driver belorehand
and, alter receiving permission to travel with him, set ourselves to the task ol packing
and preparing to be at the beach lor the scheduled" departure time: 1 p.m.
2
we
arrived at hall past twelve and were pleased to see a calm sea and a lew others prepared
to leave. oreat," we thought. A light load and line seasit should be a quick trip!" we
had hoped lor that, especially since it was to be our lirst time on the open ocean with
our two small children. Ae(de(dea 'kul-kul-go, much much later' (compare lijin go go
go), we were still waiting at the beach, only now we had two cranky children, the
daylight remaining was last slipping away, and the sea was being tossed by late
alternoon winds and appeared to be getting rougher by the minute. we seemed to be
the only ones agitated by the apparent delay", and more than once wondered aloud il
maybe we should try another day. we were assured by the others who were waiting that
we would be leaving soon. ln due time one last person arrived with his cargo, it was
quickly loaded, the rest ol the passengers climbed on board, and we were oll (three
hours 'late').
lt took some time lor us to realise that things don't necessarily happen
2
we were told, ba k-ami dea i tena sakai ni kiloko 'we'll go at one o'clock'.
88
according to schedule", things happenleasts, meetings, departures, and so onwhen
those who have been invited (especially the big men = &omen, i.e., 'important people')
are present and ready lor the event to begin. leople, and especially maintaining
relationship with them, are more important than the clock and keeping a schedule. 1his
practise ol putting people belore clocks is relerred to locally as 8olomon time" and it is
markedly dillerent lrom time as l conceive it (i.e., clock-governed-time), 8olomon
time" is people-governed, and people tend to be less predictable (though more
interesting!) than clocks.
ln another episode, we were chatting with an older gentleman on his veranda
and were oll-handedly made aware that he had intended to go to his garden to work
that day. Amid much apology, we began to excuse ourselves so he could go do his work,
at which he waved his hand dismissively and said, uon't worry! what l don't get done
in my liletime will be there lor my children to do, and what they don't get done will be
there lor their children. 8o stay and talkthe work will always be there, we're not."
loth ol these episodes, lrom my perspective, have to do with the 'luture': a boat
scheduled to depart at a certain time in the luture, an unlinished task looming in the
luture. 1he lengo perspective is dillerent lrom mine. ln the boat story there were some
things realis (not all ol which were known to us) and others irrealis. 1he scheduled
departure time was, lrom my perspective, something lixed and certain (though still
luture), what could be more 'real' than 1 o'clock p.m.7 ln a lengo speaker's perspective,
however, it was clearly irrealis. what was realis to the driver (and probably the other
passengers as well) was that passengers X, Y, and 2 had come to town with him in the
morning and had committed to return to the village with him that alternoon. while
89
passengers X and Y were ready and waiting at the beach, passenger 2lor whatever
reasonwas not, and the boat was not going to leave until 2 was ready. 1he people
involved and their relationships to one another were on a level ol reality J actuality
that outweighed an unrealised time in the luture. 8imilarly, in the second story, there
were certain realities that outweighed work unrealised, people, and especially those
people on your veranda, are realis while work in the lutureeven very necessary, lile-
sustaining workis irrealis.
1hese two episodes rellect an understanding ol time that contrasts with my
sense ol time. And herein lies an opportunity to make an important point ol
clarilication: l do not claim to lully understand lengo 1Am. lt seems that lengo 1Am is
broadly divided into realis and irrealis moods, and that all other shades ol tense, aspect,
and mood lall within one or the other ol these major categories.
!
1hinking in terms ol
realis J irrealis is a stretch lor a mother-tongue lnglish speaker like mysell. l tend to
think in the supposed 'hard and last' categories ol tensepast, present, and luture
along with a variety ol aspects. mood is limited to subjunctive and imperative, while a
realis J irrealis distinction is rarely considered. 1he situation seems to be the complete
reverse with lengo speakers. lt is not the case that past, present, and luture are
unknown categories, nor that aspect is absent, but they are not the major categories
and they are not expressible outside ol the major categories ol realis J irrealis.
4
!
koss (2004:00) notes that in Oceanic languages typologically, 1here is usually a distinction between
realis (k) and irrealis (lkk) mood and olten between various aspects." ln a lootnote to this he adds
that, 1here is typically no tense as such in canonic languages. 1he irrealis by delault denotes the
luture, though it also has other uses. 1he realis denotes present and lA81. lowever, there are
exceptions to these generalisations."
4
oivn (1984:22) maintains that, 1he division within the 1Am notional space into tense, aspect and
modality is lar lrom spurious. ln one way or another, these three represent three dillerent points of
departure in our experience ol time . . . ln describing the three major categories and their sub-
components or variants, we will initially maintain the pretense that each lorms a separate, sell-
90
1here are six prelixes that attach to the subject relerence pronoun: g-
'perlective', t- 'realis', b- 'apprehensive', k- 'irrealis', m- 'and', and p- 'or'. 1hese latter,
m- and p-, do not play a 1Am marking role.

when the unmarked lorm is included, there

are live lorms ol the subject relerence pronoun to account lor with relerence to 1Am.
8orting these lorms out has been a challenge. 1here is no single neat category that they
all lit into. 1here are a number ol 8outheast 8olomonic languages that show evidence ol
a distinction between realis and irrealis mood. 1he situation in lengo seems more
complex than these two categories are able to account lor. According to crowley
(2002a:!2), oela shows evidence ol past, present and luture tense.
6
Again, lengo
appears to have more going on with the subject relerence pronoun than three tenses.
lndeed, lengo seems to have an overriding savour ol mood mixed with subtle tones ol
aspect, and, to maintain balance, a dash ol tense. 1hese may be best viewed as residing
on a continuum as in ligure 4.1.
contained lunctional domain. 8uch pretense is convenient lor the purpose ol exposition, but is
probably ultimately not warranted. 8ynchronically, diachronically and ontogenetically, 1Am
categories are interconnected, as well as connected to other regions ol our conceptual map" (italics
his). with relerence to this last point, it is intriguing that lengo 1Am is directly linked, in the lorm ol a
prelix, to a subject relerence pronoun, time is understood in relation to person.

1hey do, however, play a 1Am maintaining role. 1hat is, the 1Am signature ol the subject relerence
pronoun preceding applies to an m- or p- marked subject relerence pronoun.
6
crowley's analysis ol oela -8kl as 'past' and t-8kl as 'present' is curious. ln lact, there is some
inconsistency in the glosses he provides lor certain examples: -8kl is glossed both 'luture' and 'past',
t-8kl is variously glossed as 'non.luture', 'past', and 'present'. le is, however, consistent in glossing k-
8kl as 'luture'. l suspect that, given the lamilial relationship between oela and lengo and also the
similarity ol the lorms t- and k-, this is a mere typological errorthat t- is meant to be 'past' and
'present'. ooing lurther, one might be lorgiven lor speculating whether t- is in lact 'realis' and k-
'irrealis'. Lnlortunately, l was unable to discuss this with crowley belore his untimely passing.
Another description ol oela (coombe 2002 [unpublished m.s.]) claims that k- marks luture and that t-
marks past continuous or completive while marks present continuous.
91
realis irrealis
immutable mutable
past luture
g- t- b- k-
realis irrealis
indicative apprehensive imperativeJcohortative
perlective imperlective
(8kl - bo)
possible
luture
(ba - k-)
impossible
luture
(boro - k-)
Figure F.": T*M continuum
1his is one way to represent the lengo 1Am markers, it is not the only way. lut
what this representation does is show those 1Am markers that are polar counterparts ol
one another: realis J irrealis modality, perlective J imperlective aspect, and indicative J
imperative mood (b- 'apprehensive' does not have a polar counterpart), and possible J
impossible luture tense. 1he 1Am prelixes t- and k- represent the middle grounds ol
realis and irrealis moods respectively in lengo. 1o either side ol these 'centres' are 1Am
markers that tend toward more lirmly entrenched or more luzzily realised
manilestations ol the centres.

1he unmarked lorm occupies the conceptual space

between realis and irrealis: it indicates realis at some times and irrealis at others, and
when combined with bo becomes imperlective (i.e., an event or state that is moving out
ol irrealis and into realis but remains, at the moment ol utterance, with one loot in
each).
An alternate representation ol lengo's 1Am notional space is given in ligure 4.2.

1ense, aspect, and mode are sometimes dillicult to tease apart . . . Operators that occur in the 1Am
areas ol the verb or verb phrase are likely to have indistinct semantic ranges" (layne 199:2!4).
92
lere the separation ol 1Am into the broad categories ol realis and irrealis is represented.
4. . 1. 1 lerlecti ve g-
lerlective aspect, an event or state viewed as a whole, is marked on the subject
relerence pronoun with the prelix g-.
8
1he 1Am prelix g- is a prototypical example ol the
realis mood. consider (161).
(161) Ma na bona t-u gari vaolu m-e doku na thuli-gu ni bebeu g-u
cON} Ak1 time kl-18o boy young cON}-!8o good Ak1 body-l8:18o olN play llc1-18o
tangomana na bebeu, mena deni u ghua thairo kikiki ba k-e viti.
able Ak1 play lOc ulm:8o 18o cONu work little lL1 lkk-!8o pain
'And when l was a young boy my body [lelt] good lor play and l was able to play,
but now il l work a little bit it hurts.'
1he 'event' is the realis time 'when l was a young boy'. 1he ability to play, however
here marked with perlective g-is viewed as a whole. lt is not broken down into
8
lerlectivity indicates the view ol a situation as a single whole, without distinction ol the various
separate phases that make up that situation" (comrie 196:16). 1his is distinct lrom perlect aspect
which normally describes a currently relevant state brought about by the situation (normally an
event) expressed by the verb" (layne 199:2!9).
9!
Figure F.$: T*M tree
1ense J Aspect J mood
realis
t-
irrealis
k-
perlective
g-
indicative

imperlective
- bo
apprehensive
b-
imperative
k- (- ba)
luture
impossible
boro - k-
possible
ba - k-
individual games or periods ol play, that is, it is not broken into separate phases (as
indeed the period 'when l was a young boy' is a separate phase lrom 'but now' as the
narrative continues). kather it is the whole ol the boy's ability to play that is in view.
ln (162) three instances ol the perlective marker g- are lound, including one
instance ol it allixed to the temporal J spatial locative i.
(162) 4alu bona ba k-e teigha na tavu-ti-a na be, m-e teigha na
some time lL1 lkk-!8o Nlo Ak1 lind-1k-O:!8o Ak1 pig cON}-!8o Nlo Ak1
tavu-ti-a na kau, ma na mamanu veghe t-e ngao-a na ghani-ra
lind-1k-O:!8o Ak1 dog cON} Ak1 animal like kl-!8o like-O:!8o Ak1 eat-O:!ll
m-e teigha na tavu-ti-ra g-e gora g-e ne tangomana
cON}-!8o Nlo Ak1 lind-1k-O:!ll llc1-!8o hungry llc1-!8o purpose able
na ghani tinoni ia, g-i na vua
Ak1 bite person ulm llc1-lOc Ak1 crocodile
'8ome times it will not lind a pig, and it will not lind a dog, and the animals like
it wants to eat and it doesn't lind them [so] it's hungry, [that's why] it was able
to bite that person, that's the way ol a crocodile'
lere, alter describing the scarcity ol suitable lood (pigs, dogs, and so on), the villagers'
inability to deal with problem crocodiles since the surrender ol their lirearms (during
the national 'lirearms amnesty'), and the young boys' riverside tauntings ol the beasts,
the interlocutor responds to the question, 'low is it possible that a crocodile bit
someone7!' 1he response is punctuated by the perlective aspect. 1hat is, because when
a crocodile can't lind pigs, dogs, or any other kind ol lood it lavours, it is hungry. 1he
entire state ol the crocodile's hunger is in view here. And when a crocodile is hungry it
is possible lor it to eat a person. lere, the main event ol the conversationa recent
crocodile attack in the villageis in view. 1he individual events ol the attackthe
stealth leading up to the attack, the struggle, the beginning or the end ol the attack
94
are not so much in view here as the event, viewed as a whole (lrom causes to
consequences), ol a crocodile biting a person.
1he summary statement gi na vua 'that's the way ol a crocodile' is, again,
perlective. 1he particle gi is an instance ol the perlective realis marker g- attached to
the locative i.
9
oiven the situation as described in (162), the speaker resigns himsell to
the lact that in the conceptual space surrounding crocodiles, that's the way things are:
they are ruled by their stomachs. crocodiles have behaved this way since the beginning
ol time and there is no indication that they will change any time soon, gi na vua . . .
80
4. . 1. 2 keali s t-
ll g- is perlective, a prototypical realis, then t- represents the middle ground ol
the realis mood. oiven the conceptual overlap between past and present with both
being under the umbrella ol realis mood, it is easy to mistake t- 'realis' as a marker ol
past tense. lxample (16!) certainly appears to be past tense, as it describes an event
that happened the day belore.
(16!) !tu-ghami a 4. ma 6. ma /. ma #. t-ami -tu uru.
lAL-ll:1lX.ll Ak1 l cON} k cON} l cON} l kl-1lX.ll-lAL dive
'we lewl., k., l., l., [and l]we lew went diving.'
lowever, to understand t- as marking past tense would introduce dilliculties in (164).
(164) * thi ighoe t-o mono i buriti dea k-o gharasu mai.
Ak1 kll ll:28o kl-28o stay lOc back go lkk-28o move come
'You who are at the back [ol the house], move this way.'
ln (164) the speaker is exhorting those who, at the time ol utterance, are (present) at
the back ol the house. 1o understand it as 'you who were at the back' would not make
sense in light ol the imperative ko gharasu mai 'lkk-28o move come'. ll they were (past) at
9
1he particle g-i is discussed in the section on t-i 'realis locative' (page 122).
80
compare lijin hem nao krokodael.
9
the back they no longer are (present) there to respond to the command to move toward
the speaker lrom there. Additionally, to understand mono 'exist, stay' as 'situated' or
some equivalent (leaving a reading with the continuing result ol a past event like 'those
ol you who situated [past] yourselves at the back ol the house') does not do justice to
the meaning ol mono 'exist, stay' (compare lijin stap). 1he word mono 'exist, stay' has to
do with the current state, not the events that led up to the current state. An
understanding ol t- as realis, however, removes these dilliculties. Lnderstanding t- as a
marker ol realis mood makes sense ol both (16!) and (164), the one past and the other
present. 1he 1Am marker t- marks realis, but is unmarked lor past or present tense.
ln (16) two realis statements about the people ol village l. are given: that there
are many people in l. and that they are able to lish.
81
(16) !ghamu na ). t-ara vonu koto na tinoni t-amu tangomana na ta-tagho.
ll:2ll Ak1 l kl-!ll lull very Ak1 person kl-2ll able Ak1 lN18-lish
)-o dea ta-tagho m-o k-o lavi mai-a.
lkk-28o go lN18-lish cON}-28o lkk-28o take come-O:!8o
'You [village] l.there are very many people [in l. village]you are able to
lish. You go lishing and you bring what you catch.'
lecause the lirst two clauses are realis, the speaker is able to command them to go
lishing and to bring what they catch.
ln (166) a good example ol the contrast between realis and irrealis is lound.
(166) na tha t-ami ngao-a na lavi dea-i ba k-ami lavi dea-i
Ak1 what kl-1lX.ll want-O Ak1 take go-O:!ll lL1 lkk-1lX.ll take go-O:!ll
'what we want to take we will take'
81
Note the three dillerent 'persons' used to address the hearer(s) (i.e., the people ol village l.) in (16):
second person singular, second person plural, and third person plural. while switching lrom second
person plural to singular may be explained as the speaker moving lrom general to specilic relerence
in order to make the message more applicable to individuals, it is dillicult to explain the shilt lrom
second person ighamu to third person tara and back again to tamu.
96
1he wanting-to-take, marked by tami, is realis while the actual taking, the kami clause,
remains, as yet, unrealised. 1he distinction here is not one ol tensepast and luture
82

but ol realis and irrealis.

1his contrast is also evident in (16). lere, looking ahead to a least, the hosts
promised to give what they were able to those who came.
(16) !ghami nga ba k-ami ne athe na tha t-ami tangomana vani-ghamu.
ll:1lX.ll ulm:8o lL1 lkk-1lX.ll lmll give Ak1 kll kl-1lX.ll able uA1-O:2ll
'we here, we will give what we are able to you.'
1he ability is not past tense, it is the ability realised at the yet-to-arrive time ol the
giving.
An example ol luture realis is given in (168).
(168) na vanga tha t-o ghali mai-a tena arere ba k-e oli tena
Ak1 lood kll kl-28o make come-O:!8o lOc 8aturday lL1 lkk-!8o return lOc
nimoa na thara tibo-mu
l8:28o Ak1 least.row klll-l8:28o
'the lood that you make [and] bring on 8aturday, it will return to your own least
row'
8aturday was the next day with relerence to the time this was spoken. lut the speaker
was able to say, with some certainty using the realis mood, that the lood you prepare
and bring (realis) to the least will be (irrealis luture) available to you in your least row
it won't end up in another row.
linally, in (169) some examples ol present realis are lound. 1he temporal time
space is identilied as 'now'the locative demonstrative pair tena deni meaning
something like 'this location [in time]'.
(169) tena deni u thudu le po ni bungu-ti-ra po na gari vaolu
lOc ulm 18o sit still llm olN watch-1k-O:!ll llm Ak1 child new
82
lengo does have a luture marking morpheme ba 'lL1'. 8ee page 104 lor discussion.
9
t-ara ghali-a na thairo pukua t-e viti na aro-gu
kl-!ll do-O:!8o Ak1 work because kl-!8o hurt Ak1 shoulder-l8:18o
'now l just sit still to just watch the youths that do the work because my
shoulder hurts'
1he speaker, lirmly situated in the present 'now' (note the unmarked lirst person
subject relerence pronoun u), comments that he is currently watching the young boys
working because his shoulder still hurts. 1he youths working and the shoulder hurting
are not past events. lndeed, it is the ongoing pain in the shoulder that prevents the
speaker lrom working himsell and lorces him to sit and watch. 1he realis mood, and not
past tense, make the most sense ol these kinds ol situations.
4. . 1. ! Lnmarked
1he unmarked subject relerence pronoun is just that: unmarked. On its own it
gives no indication ol 1Am. 1he unmarked subject relerence pronoun is the lorm most
commonly used with stative verbs, and when used with dynamic verbs is ambiguous in
its 1Am signature. 1he label 'indicative' springs to mind with relerence to the unmarked
subject relerence pronoun.
8!
1he unmarked subject relerence pronoun can be lollowed
by the particles bo 'lmll' and ti 'kl-lOc'both ol which alter its 'time signal'.
lxample (10) below begins to shed some light on the nature ol the unmarked
subject relerence pronoun.
(10) e vonu na tagi
!8o lull Ak1 tank
'the tank is lull'
lere, as statives olten are, the verb vonu 'lull' is lound with the unmarked subject
relerence pronoun. 1he speaker in (10) makes a simple declaration as to the amount ol
water in the tank. Now, the tank in question wasn't lull to the brim, the speaker having
8!
crystal (200!:299) notes that the indicative is olten the unmarked lorm ol a verb paradigm.
98
just run his hand up the side ol the tank as it was beginning to catch the morning sun,
was able to tell lrom the temperature dillerential that the tank was about two thirds
lull. compare the sentence e vonu ti na tagi 'the tank is lull [to capacity]'. 1he addition
ol ti 'kl-lOc' lends a realis element to the declaration ol the tank's lullness.
84
1his second
sentence was used to describe a water tank with water llowing out the top during a
tropical downpourthere was no question as to the amount ol water in the tank: it was
lullreally lull!
Another example, this time involving a dynamic verb, is given in (11). 1his
sentence is the opening sentence ol a text describing an injury that prevented the
speaker lrom participating in the day's work. At the time ol utterance the work was
ongoing and the scene was being set. 1he unmarked subject relerence pronoun, the
indicative mood, is most appropriate in this context.
(11) ! ngeni deni ami thairo ni vae ni kolivuti.
lOc today this 1lX.ll work olN house olN pray
'1oday we work on behall ol the church.'
1he next sentence in the text, however, is marked realis.
(12) na bona t-u thairo
Ak1 time kl-18o work
'while l worked'
lere, the speaker begins to recount what happened minutes previously to aggravate an
old shoulder injury, leaving him unable to continue working.
when bo 'lmll' combines with the unmarked subject relerence pronounthe
only lorm ol the subject relerence pronoun with which bo is loundthe combination
indicates imperlective aspect. Nuances ol the particle bo are discussed below, but lor
84
1he realis locative ti is discussed on page 119ll.
99
now consider examples (1!) and (14). lxample (1!)a. came in response to the
question asked ol someone walking past a veranda: o bo dea i vi7 '28o lmll go lOc where,
where are you going7'.
(1!) a. + bo dea po i vae-a V.
18o lmll go llm lOc house-l8:!8o V
'l'm just going to V.'s house.'
b. + dea po i vae-a V.
18o go llm lOc house-l8:!8o V
'l just went to V.'s house.' or 'l just go [habitual] to V.'s house.'
lxample (1!)a. has a clear sense ol underway-ness, that the speaker is in the act ol
going. lxample (1!)b., lacking bo as it does, may be punctual, iterative, or habitual, but
it is not imperlective.
ln (14) note that, even though the event happened i bongi 'last night', the
subject relerence pronoun is unmarked lor 1Am. lt is, however, lound together with bo.
(14) ! bongi ara-ko bo dea pogho ara-ko ruka na ghaoka.
lOc last.night !ll-uL lmll go bathe !ll-uL two Ak1 woman
'last night they two were going to bathe they two two women.'
1he speaker was intent on communicating the imperlective nature ol the women's
'going' by avoiding a subject relerence pronoun marked lor realis mood (i.e., t-, in
which the imperlective is not possible).
4. . 1. 4 lrreal i s k-
1he 1Am prelix k- 'lkk' is lound in lour environments representing live lacets ol
1Am: irrealis, possible and impossible luture, and imperative and deontic mood.
100
irrealis k-8kl
possible luture ba G k-8kl
impossible luture boro G k-8kl
imperativeJcohortative k-8kl
deontic k-8kl - ba
Figure F.;: T*M indications of k-
As was the case with t- 'kl', which is easily mistaken lor past tense, k- is easily
mistaken lor luture tense. lurther analysis shows, however, that k- is best understood
as irrealis mood. 1hough in certain environments k- does combine with other elements
to indicate luture tense or imperative or deontic mood, none ol these is the overarching
role ol k-. 1he 1Am marker k- indicates irrealis mood, within which are lound the
imperative and, with the auxiliaries ba 'lL1' and boro 'impossible lL1', varieties ol luture
tense and deontic mood.
1hat luture, imperative and deontic are marked, in whole or in part, by k- 'lkk' is
not surprising. 1here is a sense ol irrealis in the imperative and deontic moodsthe
action has not happened yet, though it is strongly expected that it will happen. luture
also has a signilicant level ol irrealis, as noted by oivn: the luture is a clear irrealis
tense, dealing with hypothetical, possible, uncertain states or events that have not yet
occurred" (1984:28). ln this analysis, luture tense is a subset ol irrealis mood. At the
same time luture has a measure ol certainty with regards to actually happening. 1his is
marked by ba 'lL1' in lengo to distinguish it lrom irrealis, the latter having no overt
indication (or expectation) ol ever being realised. 8o in terms ol expectation ol
happening (i.e., becoming realis) lengo has, lrom high to low: imperative J
cohortative J deontic luture irrealis impossible luture. Again, however, it must
be said that imperative and deontic mood and luture tense are best understood as being
101
within the scope ol irrealis. luture tense is irrealis, but irrealis is not necessarily luture
tense. likewise, imperative and deontic moods are irrealis, but irrealis is not imperative
or deontic.
1he basic meaning ol k- 'lkk' is demonstrated in (1). 1he thundukolu 'meeting'
is a planned, though as yet unrealised, event. 1he meeting can only be talked about in
the irrealis mood.
(1) + ngao-a maghe tangomana na bosa i ropo k-e sakai ni sara
18o want-O:!8o il possible Ak1 tell lOc tomorrow lkk-!8o one olN arrive
na bona veghe k-a ghe thudukolu.
Ak1 time like lkk-1lN.ll continue sit.together
'l want, il possible, to talk tomorrow belore the time arrives, the time we meet.'
A sentence such as (1)one that remains at the level ol possibility on many levelsis
most appropriately marked as irrealis. lndeed, the speaker lurther downplays the status
ol the meeting with the use ol veghe 'like' (lijin olsem). At this point he can speak ol it
only as a semblance ol a meeting.
1here is an interesting case ol irrealis in (16).
(16) 6oropo rukana k-o pelu vani-u ruka na vugho
morning day.alter.tomorrow lkk-28o purchase uA1-O:18o two Ak1 net
k-u ne vu(vugho ko-gha-da igha.
lkk-18o so.that kul-net uL-oral.cll-l8:1lN.ll lish
'Next tomorrow morning you buy lor me two nets so that l can net some lish
lor us to eat.'
1here is a sense ol indeterminacy right lrom the outset. 1he time signature is roropo
rukana 'day alter tomorrow [in the] morning'. why the delay7 why not tomorrow
morning7 why not now7! 1he reason is that this was not a serious request, it was more
ol an 'il-the-planets-line-up-and-you-are-willing' kind ol request, a testing ol the
102
waters, so to speak. lt was put oll lor two days to give the hearer time to lorget" or
come up with some other excuse lor not satislying the suggestion. 'lt would be nice il
you would buy me a couple ol lishing nets, and il you did l would even catch some lish
lor you!', might better capture the sense ol (16). 1his is prototypical irrealis: there is
no commitment to the realityor even the potential realityol getting some lishing
nets. lt would be exciting il it happened, but there is no real expectation that it will.
Aside lrom marking irrealis, there are lour other 1Am leatures within the scope ol
k-: possible and impossible luture, and imperative and deontic mood. 1hese will be
discussed in more depth in subsequent sections.
4. . 1. Apprehensi ve b-
1he 1Am prelix b- is used to warn the addressee ol the predictablethough as yet
unrealisednegative outcome ol his actions in order to help him avoid such an
outcome. 1he label l use to describe b- is 'apprehensive'. 1his, lollowing lichtenberk,
8

captures both the ability ol the speaker to loresee (i.e., apprehend) the danger ol the
situation and the lear (i.e., apprehension) ol the speaker. As such, b- marks another
aspect ol irrealis mood. 1here is an element ol irrealis conditional mood with b-: il the
listener does not heed the speaker's warning and continues with the current course ol
action the outcome is lixed and certainunpleasant, to be sure, but certain
nonetheless. 8o at the same time as there is a sense ol irrealis conditionality to b-, there
is also a sense ol declarative certainty. Overarching this mix is a concern best described
as preventative.
86
consider (1).
8
lichtenberk (199:29) discusses apprehensional-epistemic modality in 1o'abaita (8outheast
8olomonic) noting, there is apprehensiontypically on the part ol the speakerthat a potential
(undesirable) situation may turn out to be so."
86
8ee lockett (198:2!) lor relerence to a mood meaning 'ood lorbid that X should happen' in lox
10!
(1) 4ere-ngi-ko-ghamu: b-amu-ko tumu1
look-1k-uL-O:2ll Allk-2ll-uL lall.down
'look out, you two: you're going to lall!'
1he speakerloreseeing a dire outcomeissues a clear warning: change your course ol
action or lace the consequences! lut it's more than just a declarative 'you're going to
lall down', as the lnglish gloss leebly conveys. lt's a 'you're going to lall down and hurt
yourselves [and l don't want to have to take you to the clinic so get down lrom the top
ol that coconut tree]!' kind ol warning. ln lact, lengo speakers gloss b- phrases with
lijin nogud. 8o (1) would be glossed in lijin as nogud iutufala foldaon 'it would be a
shame (i.e., no good) lor you two to lall'. 1he speaker is keen to help the hearer avoid
the negative consequences.
1o the same end (18) is, lrom time to time, spoken to young children.
(18) Tabu na lavi-a na ghau: b-e ghado-gho1
Nlo Ak1 grab-O:!8o Ak1 knile Allk-!8o pierce-O:28o
'uon't grab the knile: no good it cuts you!'
1he speaker's interest is in helping the child avoid the inevitable cut that would come il
she was to pick up a sharp knile.
4. . 2 1Am auxi li ari es
4. . 2. 1 luture ba
koss (1988:10!), in one ol lew relerences to lengo in the literature, notes that,
it is possible that lOc [lroto Oceanic] had a luture-marking morpheme 7ba . . . which
has retained its phonological independence in lengo."
8
lis accompanying example
(4.10) is included here as (19).
88
(mesquakie, Algonquian) language.
8
keaders not lamiliar with the conventions ol historical and comparative linguistics should note that
koss uses the *asterisk to mark the proto-lorm and not non-grammaticality.
88
Note that lcl = possessive classilier and l = possessive. Also, the addition ol 'tomorrow' to the lree
104
(19) a para ba k-e-dea i leo-na na ne-na etea
Ak1 lara lL1 lL1-8:!8-go.up lkll inside-l:!8 Ak1 lcl-l:!8 canoe
'lara will get into his canoe tomorrow.'
According to many lengo speakers, ba is the short lorm ol nganiba 'soon, later,
sometime (today7)'. lt seems that nga.ni.ba is a compound word, composed ol nga
'medial demonstrative', ni 'olN', and ba 'lL1', a neither-here-nor-there sometime ol the
luture. lxample (180) demonstrates the use ol nganiba.
(180) 0 ghua logho bo(bona vata i ngani ba na nulavi m-ami-tu dea
!8o cONu have kul-time later lOc later Ak1 alternoon cON}-1lX.ll-lAL go
lebo-i laka ti.
lloat-O:!ll also kl-lOc
'll there is still time, later in the alternoon, we lew will also lloat them [down
the river].'
1here is a sense ol specilicity introduced with i nganiba na nulavi 'later [in] the
alternoon'. 1he reduplicated bo(bona vata 'kul-time later' is much more general than i
nganiba na nulavi by means ol which the time is narrowed down to 'sometime in the
alternoon' (though it remains imprecise).
luture ba is consistently used in conjunction with the subject relerence
pronoun modilying 1Am marker k- 'lkk'. 1he particle ba is lound both belore and alter k-
8kl. ln the overwhelming majority ol instances in my texts, ba is lound preceding k-8kl.
89

ln these, ba is clearly a luture-marking morpheme, as koss observed. As such, luture is
the only tense that is marked overtly in lengo,
90
but even so it works in conjunction
translation is open to question as later koss glosses the same data as simply 'lara will get into his
canoe' (!2). 1he luturity ol the event is not indicated here with the lexical time word ropo
'tomorrow'. Another dillerence koss introduces in the latter citation ol the same data is that k-e-dea is
glossed as '1A-8:!8-go.up', where 1A indicates 'tenseJaspect marker'. my analysis ol k- as 'irrealis' and
the lack ol ropo 'tomorrow' lrom (19) makes the role ol ba 'lL1' somewhat clearer.
89
when ba lollows the subject relerence pronoun it marks deontic mood. 8ee below.
90
lengo makes a two-way distinction in terms ol tense: luture and non-luture. within luture tense
there is a lurther distinction between possible and impossible (see below).
10
with irrealis mood. 1hat (181) has yet to happen, is expected to happen in the luture, is
clear lrom the locative phrase i ropo 'tomorrow'.
(181) ! ropo ba k-u dea.
lOc tomorrow lL1 lkk-18o go
'1omorrow l will go.'
1he weatherand especially predictions concerning weather to comeis a
lrequent topic ol discussion. And so even though one might not hear (182) every
morning, it is heard quite olten.
(182) 4a ke para n' atho.
ba k-e para na atho
lL1 lkk-!8o hot Ak1 sun
'1he sun will be hot [today].'
And it is olten true: the sun is hot at some point during the day (as, at 9 degrees south
ol the equator, one would expect). Although just as oltenand olten on the same day
it would rain, belore which, ol course, one would hear ba ke utha 'it's going to rain'.
ln (18!) a statement is being made by someone who was about to give
instructions to a large group ol people concerning a least that was being planned lor
the next day.
(18!) !nau ba k-u tugu vani-ghamu na tha ba k-e sara tena arere.
ll:18o lL1 lkk-18o tell uA1-O:2ll Ak1 cOml lL1 lkk-!8o arrive lOc 8aturday
'l will tell you what will happen on 8aturday.'
loth the near luture telling-ol-the-plans and the more distant luture yet-to-arrive
events ol 8aturday are presented with the luture marking ba.
likewise, the second clause ol (184) has a level ol luturity: you" have to throw it
belore l" can catch it. Note that the lirst clause (ko bila mai 'you throw come') is
imperative while the second (ba ku dala 'l will catch') is indicative.
106
(184) )-o bila mai8 ba k-u dala.
lkk-28o throw come lL1 lkk-18o catch
'You throw [it] here, l'll catch [it].'
1he imperative clauseko bila maiis not preceded by ba 'lL1', but the second non-
imperative clauseba ku dalais preceded by ba and is clearly in the luture.
4. . 2. 2 lmpossibl e luture boro
1he auxiliary boro 'impossible' is the polar counterpart ol ba 'lL1'. while ba
marks (possible) luturethat 'X has a good chance ol happening', boro indicates
impossible luturethat 'X has no chance ol happening because it is not possible lor X
to happen.' 1he auxiliary boro 'impossible' precedes the subject relerence pronoun and
occupies the same position as ba 'lL1', that is to say, the luture is either possible (ba) or
impossible (boro) but never both. consider example (18). 1his is part ol a discussion
between two people who were trying to decide which ol them would tell to me a story.
1he response to the question ba ke gere thivoi7 'will he write them [i.e., the speaker's
words] down7' is given in (18).
(18) Teigha1 4oro k-e gere thivo-i8 ba k-e rikordi po.
Nlo impossible lkk-!8o write down-O:!ll lL1 lkk-!8o record llm
'No! lt's impossible lor him to write them [words] down [as you're speaking], he
will just record [them].'
1he negative teigha is the retort to the polar question that had been posed, while boro
'impossible' provides some rationale: it is not physically possible to write as quickly as
someone speaks (especially, in my case, in an acquired language). comparing the
question to the answer, it certainly stands out that boro and ba occupy the same space
in the clause belore the subject relerence pronoun: [ba = boro] ke gere thivoi 'he [will J
will not (because it is impossible)] write them down'. lt is also important to note that
10
both boro and ba are always in the context ol the irrealis-marked k-8kl.
lxample (186) arose as part ol an address regarding the preparations lor a least.
leople were encouraged to bring lood lor their 'group' (there are live groups in the
community) to supplement what the hosts would provide. A concern was that lood
people brought as a supplement lor their group would be incorporated into the host's
lood and distributed to the other groups. lxample (186) addressed this concern.
(186) 4oro k-e ghe dea tena ovu sakai.
impossible lkk-!8o continue go lOc group one
'lt is impossible lor it [the lish you caught] to go to group one.'
lven though it might conceivably have been possible lor a group's supplementary lood
to go to another group, it was considered so undesirable as to be impossible.
ln (18), sitting separate is unrealised and, as lar as this speaker is concerned,
will dely the laws ol possibility il it is realised.
(18) pukua na thara deni boro k-a tovothi thudu
because Ak1 least ulm impossible lkk-1lN.ll separate sit
'because at this least it will be impossible lor us to sit separate'
1o reinlorce this, alter telling the members ol village group live that they cannot sit, lor
example, with village group one, the speaker continues:
(188) na ovu ni vanua-mu ba k-o ne dea thudu t-i tena ovu
Ak1 group olN village-l8:2ll lL1 lkk-28o lmll go sit kl-lOc lOc group
ni vanua-mu thara tibo-mu t-i
olN village-l8:2ll least klll-l8:2ll kl-lOc
'you will really go sit in your groupleast among yourselves'
with the repetition ol ovu ni vanuamu 'your village group' and the rellexive tibomu, the
speaker is intent on making her point that sitting separate lrom one's group is not what
she wants to happen at the least. 1he use ol boro 'impossible' in (18) does this with one
108
word, (188) reinlorces it.
4. . 2. ! ueonti c ba
1here are lour instances in the data in which ba lollows k-8kl.
91
ln each instance
there is a sense ol obligation, ol ought-ness best described as deontic. consider the
examples in (189).
(189) a. B0 ghani-u na aliva1 )-a-tu oli ba t-i,C una m-u tapa.
!8o bite-O:18o Ak1 centipede lkk-1lN.ll-lAL return ulO kl-lOc 18o:say cON}-18o run
'A centipede bit me! we need to go back [home]," l said and l ran.
b. Mai k-a ba turu-vaghini-a na vae-a T.
come lkk-1lN.ll ulO stand.up-cAL8-O:!8o Ak1 house-l8:!8o 1
'come, we all need to stand up 1.'s house.'
c. 9., k-o ba iti mai.
N lkk-28o ulO up come
'N., you come up [on stage to address the audience].'
d. ighoe a ,. k-o tura nimoa na matau k-o oli dea i vanua
ll:28o Ak1 l lkk-28o carry l8:28o Ak1 axe lkk-28o return go lOc village
inau ba k-u tura na kei ni ghole dini ma k-a ba i vanua.
ll:18o lL1 lkk-18o carry Ak1 basket olN greens ulm:ll cON} lkk-1lN.ll ulO lOc village
'l., you carry your knile to return to the village, l will carry the basket ol
these greens and let's all go to the village [belore night lalls].'
lach ol these sentences is irrealis, as indicated by k-. And while the presence ol ba could
lead to an analysis ol 'luture', the sense in each ol these is not adequately captured by a
simple luture. 1hese sentences are just short ol commands, but somewhat stronger
than suggestions, that is, somewhere between 'must' and 'should'. 1he deontic mode,
according to layne (199:246), expresses the subject's duty or obligation to perlorm
the irrealis act expressed by the verb." 1he companions ol the person bitten by the
91
ln (189)a. ba is lound lollowing the subject relerence pronoun and the verb. while this does not lollow
the ordering ol the other three examples, it clearly does not lollow the ordering ol a luture
construction, in which ba is belore the subject relerence pronoun - verb.
109
centipede in (189)a. have a duty to accompany him home to ensure he gets there salely
and to help him il he cannot. lengo society holds it as a high duty to assist others to
build their houses, as in (189)b. 1he N. ol (189)c. is obliged to take the stage so that the
program can continue uninterrupted. And the speaker and his companions in (189)d.
need to get on their way back to the village lest darkness overtake them on the trail.
ln summary, the luture must have ba belore irrealis k-8kl but when ba lollows
irrealis k-8kl (and sometimes the verb as well) the result is a deontic mode.
4. . 2. 4 lmperlecti ve bo
As was discussed above, ongoing action is indicated by bo 'lmll'. 1he auxiliary bo
'lmll' appears only with a subject relerence pronoun unmarked lor 1Am as in (190).
(190) B? bo dea i vi5C B?leole po.C
28o lmll go lOc where stroll llm
'where are you going7' '}ust walking around.'
1he question was asked ol someone on the go, someone walking past the questioner (as
the answer indicates). 1he 'on-the-go-ness' is marked by imperlective bo. 1he subject
relerence pronoun o '28o' is unmarked lor 1Am.
1he sense ol underway-ness is also evident in (191).
(191) a. i bongi ara-ko bo dea pogho ara-ko ruka na ghaoka
lOc last.night !ll-uL lmll go bathe !ll-uL two Ak1 woman
'last night they two [were] going to bathe they two women'
b. ara-ko dea pogho m-e tighi thivo sakai na ghaoka m-e tumu-ri-a
!ll-uL go bathe cON}-!8o lirst down one Ak1 woman cON}-!8o lollow-1k-O:!8o
a sakai e thivo m-e theo-a na beti m-e vuli iti ma
Ak1 one !8o down cON}-!8o scoop-O:!8o Ak1 water cON}-!8o pour up cON}
110
na ruke e theo-a na beti m-e vuli dea i beti
Ak1 second !8o scoop-O:!8o Ak1 water cON}-!8o pour go lOc water
'they two go bathe and lirst one woman [went] down and the other one
lollowed, she [went] down and she scooped some water and poured it
upstream and a second [time] she scooped some water and she poured it
across stream'
c. m-e thivo ni bo dea pogho m-e mai na vua
cON}-!8o down olN lmll go bathe cON}-!8o come Ak1 crocodile
'and she [went] down lor going bathing and a crocodile came'
ln this extended example there are two women going to the river to bathe in the
evening. 8ince crocodiles are known to inhabit the river, local practise is to scoop water
in a bucket and pour some upstream, across stream, and downstream to trick a
crocodile into thinking that perhaps a dog or a pig has lallen into the water. 1he hope is
that any crocodile lurking in the dark on the riverbank will go alter the prey" and
thereby alert the bathers to its presence. 1he particle bo 'lmll' is lound at the opening ol
the story and again alter the parenthetical explanation ol the local river bathing
practise. ln both instances it is in the context ol dea pogho 'go bathe', the women were
not going to pour water on the river or taunt crocodilesthey just wanted to clean up
alter a long day ol working in the garden and sweating over a cooking lire. 1heir
primary purpose, their main intent, was to go bathe and the storyteller makes this
point by putting it in the imperlective mode. lt is as il to say, while activity A (bo dea
pogho 'going to bathe') was underway activity l (theoa na beti 'scoop the water') took
place, and both A and l were interrupted by c (mai na vua 'come a crocodile'), but A is
what was in process when l and c happened."
1he subject relerence pronoun in (191)a. is unmarked lor 1Am, as is consistent
with the context in which bo 'lmll' can appear. while in (191)c. it appears that bo
111
appears with a marked subject relerence pronoun, in actual lact the conjunction m-
carries the 1Am signature ol the preceding subject relerence pronoun not marked with
conjunction m-, which in the case ol (191)a. is, indeed, unmarked (ma na ruke e theoa na
beti 'and the second [time] she scooped the water').
considering the lunction ol imperlective bo and the realis locative ti together is
instructive.
92
1he examples in (192) show bo preceding and ti lollowing the verb while
the subject relerence pronoun is unmarked lor 1Am.
(192) a. 0 bo ghilaghana t-i na sanga ni ghita.
!8o lmll know kl-lOc Ak1 language olN O:1lN.ll
'le almost knows our language.'
b. 0 bo ghaghare t-i na vonu.
!8o lmll near kl-lOc Ak1 lull
'lt's almost near lull.'
1he realis locative ti indicates the realisation ol an event J state, but imperlective bo
indicates that the realisation remains ongoing. what results is an 'almost'the lull
realisation ol the event J state is held oll as it remains in process. 1he ellect ol these
two particles on an indicative sentence is to indicate something on the cusp ol
realisation, an imperlective realis. 1his is best rendered by lnglish 'almost' (lijin
kolsap). Obviously, 'knowing' a language (as in (192)a.) is not something that happens in
an instant. lndeed, it may well be impossible to pinpoint just when one knows a
language. lt is a progressive act: while some parts are known well, others are not known
at all, and still others remain only partially known. 1here are elements ol realis mood
and imperlective aspect involved in this kind ol knowing. ln lengo this is expressed
with an indicative subject relerence pronoun in the company ol imperlective bo 'lmll'
92
1he realis locative ti is discussed in more detail beginning on page 119.
112
and the realis locative ti.
4. . 2. conti nuous ghe
1he auxiliary verb ghe 'continue' is lound belore the main verb and indicates
continuation ol a previously initiated event. 1he auxiliary nature ol ghe is conlirmed by
the lact that it never occurs on its own.
(19!) a. m-u ghe tapa inau
cON}-18o continue run ll:18o
'and l continued to run'
b. m-ami-tu ghe mai talu thivo-i po t-i na koga
cON}-1lX.ll-lAL continue come put down-O:!ll just kl-lOc Ak1 stick
m-ami-tu ghe thanga-ra po t-i
cON}-1lX.ll-lAL continue help-O:!ll just kl-lOc
'we lew continued coming, just put the sticks down, and we lew just
continued to help them'
c. k-o ne teigha t-i na tangomana na ghe thuge na tha
lkk-28o lmll Nlo kl-lOc Ak1 able Ak1 continue ask Ak1 what
ba k-a ghali-a tena arere
lL1 lkk-1lN.ll do-O:!8o lOc 8aturday
'[alter receiving these instructions] you are not able to continue to ask what
we will do on 8aturday'
ln each ol the examples in (19!) there is the sense that the verb which ghe 'continue'
precedes had already been undertaken and is merely being continued. ln (19!)a., lor
example, a lew sentences earlier the author, having just been bitten by a centipede,
dropped his things and, as he recounted it, mu tapa po ti tapa mu olimai 'l just ran, ran
and l returned [home]'. Alter his lriends ask where he is going he responds that a
centipede has bitten him and then mu ghe tapa inau 'l continued to run.' 1he lirst
instances ol tapa 'run' represent the initiation ol the event whereas the instance ol tapa
'run' modilied by ghe 'continue' picks up where the running lelt oll belore being
11!
interrupted by the lriends' question.
lt is important to distinguish ghe 'continue' lrom bo 'imperlective'. loth ghe
'continue' and bo 'imperlective' have a share ol progressive aspect, but they represent
dillerent kinds ol progress. whereas ghe 'continue' allows an author to pick up where
an activity lelt oll, there is no 'leaving oll' with bo 'imperlective'. with ghe 'continue'
the author is able to jump lrom instance to instance ol an activity over the course ol a
text, resuming the activity alter 'interruptions', whereas with bo 'imperlective' the
activity in its entiretyinterruptions notwithstandingis understood. ln this regard,
the imperlective and the perlective are polar counterparts.
4. . ! condi ti onal ghua
1he lunction ol the conditional ghua 'il' is discussed in the section concerning
complex clauses, beginning on page 20!. lere it is to be noted that a conditional clause
in lengo has ghua 'il' as part ol the protasis (condition). 1he particle ghua 'il' is never in
lirst position in the clause, rather, it is postpositive, that is, most olten in second
position (lollowing the subject relerence pronoun) and sometimes third position
(lollowing the subject relerence pronoun and the verb, see (196)) in the clause. 1he
protasis (condition) may precede or lollow the apodosis (result). 1he apodosis is most
olten, though not always, in irrealis mood. consider example (194).
(194) o ghua ngao-a na ghali-a siki 'e:tra flavour' t-i i leo-a
28o il want-O:!8o Ak1 make-O:!8o any extra llavour kl-lOc lOc inside-O:!8o
t-i nimoa na thara k-o ghali mai-a
kl-lOc l8:28o Ak1 least.row lkk-28o make come-O:!8o
'il you want to make any extra llavour" inside your least row you make [and]
bring it'
lere the protasis (o ghua ngaoa na ghalia 'il you want to make') precedes the apodosis
114
(ko ghali maia 'you make [and] bring it'), the conditional ghua is in second position
lollowing the subject relerence pronoun, and the apodosis is in irrealis mood. 1he
presence or absence ol any 'extra llavour' is conditional on the listener and their
culinary skills.
ln (19) there are two instances ol ghua 'il': one at the beginning and the other
at the end ol a plan to retrieve some building materials lrom the bush. 1here are two
conditions in this example: the lirst concerns time and having enough ol it and the
second people's willingness. 1he result ol the lirst is that they would lloat with the
river current, the result ol the second is that they would throw the sticks ashore at
village 1.
(19) e ghua logho bo(bona vata i nganiba na nulavi m-ami-tu
!8o il have kul-time continue lOc later Ak1 alternoon cON}-1lX.ll-lAL
dea lebo-i laka t-i k-ami-tu lebo tave mai i nughu ma
go lloat-O:!ll also kl-lOc lkk-1lX.ll-lAL lloat current come lOc river cON}
k-ami-tu mai soni iti-i ga i T. t-ami-tu ghua mavulo
lkk-1lX.ll-lAL come throw up-O:!ll there lOc 1 kl-1lX.ll-lAL il willing
'il time remains later in the alternoon and we also lloat them [sticks] we will
lloat with the river current and we will throw them up at [village] 1. il we are
willing'
1he lirst condition precedes its apodosis. 1he second condition lollows its apodosis. 1he
apodosis lor both the lirst and the second conditions is in irrealis mood. what is
interesting is that the second protasis is stated in realis mood. 1ense J aspect J mood in
relation to conditionals is discussed in more depth on page 20!.
lxample (196) demonstrates that ghua can be lound in third position in a clause.
(196) e dea ghua i namo na vonu m-e tumuri po na igha
!8o go il lOc deep Ak1 turtle cON}-!8o lollow just Ak1 lish
'il the turtle went deep the lish just lollowed'
11
1he conditional ghua is lound alter the subject relerence pronoun and the verb. 1hough
this is not common, it is not impossible. A second thing to note lrom this example is
that the apodosis is not in irrealis mood. 1his is because in this example the condition
was met at the time ol telling.
9!
4. . 4 lmphati c ne
1he emphatic particle ne 'lmll' adds a dimension ol intensity to a sentence. 1he
particle ne 'lmll' is lound between the subject relerence pronoun and the verb.
(19) a. m-u ne tapa inau
cON}-18o lmll run ll:18o
'and l really ran'
b. m-u ne vasangi viti koto
cON}-18o lmll leel pain very
'and l really leel a lot ol pain'
c. k-amu ne rongo-ni-a
lkk-2ll lmll hear-1k-O:!8o
'you delinitely [need to] hear this'
1he examples in (19) could conceivably do without ne 'lmll' but they would be missing
the urgency andJor intensity that this particle provides.
4. 6 lost-verb parti cl es
1here is a smaller set ol verbal modiliers that are lound lollowing the verb than
those that are lound preceding. 1hese are adverbs ol degree: intensive (koto 'very') and
diminutive (po 'llm, just'). ln addition to these there is the llexible particle ti 'kl-lOc'.
(198) V (AuV) (ti)
{
koto
po
}
Ol the pre-verb particles discussed above there is little restriction between
9!
ln (194) and (19) the results remained, at the time ol the utterance, unrealised.
116
combinations with the post-verb particles. 1he 1Am prelixes and auxiliaries co-occur
lreely with the post-verb particles. 1here are restrictions, however, with the pre-verb
adverbs. 1he only evidence ol a pre-verb particle co-occurring with a post-verb one is
ne 'lmll' and koto 'very'. 1hese together mark a heightened intensity.
4. 6. 1 uegree koto J po
1here are two post-verbal markers ol degree in lengo: an intensive particle koto
'very' and its diminutive counterpart po 'just'. 1hese are in addition to the pre-verb
degree marker ne 'lmll'. ln (199)a. the leeling ol pain is marked twice lor intensity: once
with ne 'lmll' preceding the verb and again with koto 'very' lollowing.
(199) a. /isa na aro-gu m-u ne vasangi viti koto tena de.
dislocate Ak1 shoulder-l8:18o cON}-18o lmll leel pain very lOc ulm
'my shoulder dislocated and now l really leel a lot ol pain.'
b. t-ara vonu koto na tinoni
kl-!ll many very Ak1 person
'they are very many the people'
ln (199)b. the verb has a single modilier ol intensity. 1hough not quite as intense in
degree as (199)a., it is more than il koto 'very' was absent.
1he degree word po 'just' is used to minimize or lessen the ellect ol a word. 1he
limiter po is always lound alter the word it modilies. when used with nouns or
pronouns po can serve to decrease the magnitude ol the item. 8o the subject geia 'll:!8o'
in (200) is reduced, despite the lact that it was a rather lengthy story (i.e., there is a
measure ol lalse humility present).
(200) @eia po-a na tugu t-u tangomana na tugu-a.
ll:!8o llm-O:!8o Ak1 story kl-18o able Ak1 story-O:!8o
'lt's all the story l'm able to tell.'
Note the !8o object sullix joined to po in this example. ln the absence ol an existential
11
verb to which to join, it is here indexing na tugu 'the story'. 8imilarly, the relative
pronoun tha in (201) is reduced so as not to appear to be putting too onerous a demand
on the groups.
(201) na tha po t-ami ngao-a vani-ghamu na leoni group
Ak1 kll llm kl-1lX.ll want-O:!8o uA1-O:2ll Ak1 every group
'what we want lrom every group is just . . . '
1he limiter po may also accentuate the limited scope ol a noun.
(202) Ma 9. po ma 4. t-ara-ko iti dea.
cON} N llm cON} l kl-!ll-uL up go
'And just N. and l., they two went up.'
ln (202) po 'just' is used to indicate that, ol those people present, only N. and l. went up.
1he numeral sakai 'one' is limited in (20!) to indicate that the event was, most
unlortunately, going to happen lor only one day as opposed to two or three days.
(20!) sakai po na dani
one llm Ak1 day
'just one day [no more]'
lut it is with verbs that po plays its part most olten. lere, as an adverb, po is
comparable to lnglish 'just' or lijin nomoa, as in the examples in (204).
(204) a. 4a k-u tu po.
lL1 lkk-18o stand llm
'l'll just stand [i.e., l don't need a chair].'
b. Teigha1 4oro k-e gere thivo-i8 ba k-e rikordi po.
Nlo Nlo lkk-!8o write down-O:!ll lL1 lkk-!8o record llm
'No! lt's impossible lor him to write them [words] down [as you're speaking],
he will just record [them].'
c. ! vi ga t-e dea ga na igha e tumuri po na vonu.
lOc where there kl-!8o go there Ak1 lish !8o lollow llm Ak1 turtle
'wherever the lish went the turtle just lollowed.'
ln (204)c. the turtle didn't do any more or any less than the lishit just lollowed, doing
118
exactly what the lish did.
4. 6. 2 keal i s l ocati ve ti
1he particle t-i 'kl-lOc' provides realis mood with a subject relerence pronoun
unmarked lor 1Am. ln (20) ti indicates something about the stealing ol a basket.
(20) ara gito-a t-i na kei
!ll steal-O:!8o kl-lOc Ak1 basket
'they stole the basket'
}ust what ti indicates here will require some explanation since, as 8ymons 198
(unpublished ms) observes, 1he use ol ti does not succumb to easy analysis." 1he
particle ti, and what seem to be its companions, mi and gi, is one ol those particles that
are lound in a wide variety ol contexts. Analysis ol something so apparently llexible is
no easy task. 1he lact that there does not seem to be anything similar in related
languages (neither oela nor lughotu), nor in lroto Oceanic makes the analysis that
much more dillicult.
94
Add to that the complexity ol working with 8olomon lslands lijin
as a metalanguage (and its inherent paucity ol vocabulary and expression) and analysis
becomes dillicult indeed. lt was only by becoming more conlident in my analysis ol 1Am
markers and the locative i that a solution began to emerge.
lt would be instructive to consider the distribution ol ti belore delving into its
analysis. 1he particle ti occurs: alter dynamic verbs (transitive and intransitive), stative
verbs, adverbs, negative modal verbs, numerals, and quote margins, alter nouns,
pronouns, rellexive pronouns, prepositions, the degree word po 'just' (this is actually
quite lrequent), and interrogative words.
1his leaves a lairly complex situation to untangle. what could possibly account
94
l have a copy ol an unpublished oari dictionary, and there is a word ti in it, but it is a conditional with
Ti kau vano. 'll l go.' as an example sentence. 1his does not resemble the lunction ol ti in lengo.
119
lor the wide variety ol contexts ti appears in7 what grammatical particle can modily
verbs, nouns, pronouns, and other modiliers7 lengo speakers, when transcribing texts,
will olten render ti as lijin finis nau or hem nau, which could, in turn, be translated as
lnglish 'that's it' (lit., 'linish now') or 'so be it' (lit., 'it [is] now').
9
when trying to
describe what ti might mean, they ollered suggestions such as everything is linished"
and a satislaction word", that ti represents some sort ol conclusion. ln short, realis
locative ti serves to declare the word, phrase, or clause it modilies as temporally or
spatially located in the reality ol the context ol the utterance.
oiven that lengo has a realis marker t-, and that native speaker intuition
consistently points toward things like 'linis', 'satislaction', and 'conclusion', it seemed
natural to begin looking in that direction. lut is realis marking t- really joining itsell to
the locative i7 lt certainly seems to be, and lollowing are some reasons why.
As an entry point into the discussion ol ti, consider (206)a. and b. 1hey are quite
similar in some respects, but markedly dillerent in others.
(206) a. t-ara mono i lighi-a na tetete i +.
kl-!ll stay lOc beside-O:!8o Ak1 hill lOc L.
'they are beside [at the base ol] the hill at L.'
b. m-ami thudu t-i lighi ni vae-a M.*.
cON}-1lX.ll sit kl-lOc beside olN house-l8:!8o m.A.
'we sat beside m.A.'s house'
1he phrase structure ol both examples is quite similar: a subject relerence pronoun is
lollowed by a verb, what looks like a lorm ol the locative i, and the preposition lighi
'beside'. 1he preposition lighi 'beside' is obligatorily preceded by i 'lOc'.
96
lntriguingly,
9
Note the temporal component (nau 'now') common to both ol the lijin renderings. lengo i 'lOc' marks
both spatial and temporal locations.
96
1he relationship between prepositions and the locative i is discussed on page 19.
120
however, in (206)b. it is lound with ti. 1he question that needs to be answered is is this a
prelixed locative or something else7 1he bound nature ol prepositions is a good reason
to lavour the lormer: there should be some lorm ol the locative i preceding lighi. 1he
reasons why this analysis is prelerred are discussed in what lollows.
ln lengo, one means lor marking 1Am is with a set ol prelixes on the subject
relerence pronoun. 1he lengo 1Am prelixes are: perlective (g-), realis (t-), irrealis (k-),
and apprehensive (b-). 1he subject relerence pronoun is also lound in unmarked lorm,
that is, not marked with a 1Am prelix and therelore not marked with a 1Am signature. 1he
conjunctions m- 'and' and p- 'or' may also join to the subject relerence pronoun but
they do not indicate 1Am, rather, they maintain the 1Am signature ol the previously
marked subject relerence pronoun. oiven this, compare the dillerence between the
prelixes on the subject relerence pronouns in (206)a. and b.: t- 'kl' and m- 'cON}'
respectively. lxample (206)a. is clearly in realis mood while, apart lrom the larger
context, it is not clear what is the 1Am signature ol (206)b.
9
ln the source text lrom
which (206)b. comes, one has to go back ten consecutive subject relerence pronouns
all marked with the conjunction m-belore coming to one marked dillerently. As it
turns out, it happens to be t- 'kl'.
As was pointed out above, (206)a. and b. also diller in the lorm ol the locative
preceding the preposition lighi 'beside': i lighi and ti lighi respectively. what could
account lor this dillerence7 oiven the 'distance' lrom the nearest subject relerence
pronoun overtly marked lor 1Am in (206)b.ten pronouns previousit appears that the
9
oiven the conjunction m-'s position in the 1Am marking slot (belore the subject relerence pronoun), it
preempts 1Am marking and is compelled to carry the 1Am signature ol the previous subject relerence
pronoun.
121
speaker re-establishes the realis 'tone' in his narrative with a locative marked with
realis mood: ti.
loth clauses are in realis mood: (206)a. by virtue ol t-ara and (206)b. by the
presence ol the realis locative ti. 1his despite the lact thator, better, precisely because
the subject relerence pronoun in (206)b. is marked with the conjunction m-. 1hat is,
because the subject relerence pronoun is marked as conjoined it cannot take the realis
1Am marker t-. kealis locative ti allows realis mood to be marked in a clause when the
realis 1Am marking t- is unable to lill the 1Am slot on the subject relerence pronoun.
1he prelixes g- 'perlective' and m- 'conjunction' also allix to the locative i and
lorm temporal and spatial aspectual and conjoined constructions respectively.
(20) a. Bg-ami tena agri ighami,C g-i gea ena
g-ami teigha na agree ighami g-i gea e-na
llc1-1lX.ll Nlo Ak1 agree ll:1lX.ll llc1-lOc ll:!8o !8o:say
'we don't agree," they said'
b. g-i na vua
llc1-lOc Ak1 crocodile
'that's the way crocodiles are'
c. ara-ko leta t-i vini vi-leu-th-i8 na leoni bona
!ll-uL 'no save' kl-lOc actually klcl-light-1k-O:!ll Ak1 all time
ara-ko vi-doku-i po t-i i levu m-i levu
!ll-uL klcl-good-O:!ll llm kl-lOc lOc side cON}-lOc side
'they two never lought each other, they two were always just good to each
otherlrom [one] side and [the other] side'
d. t-ara be(bere-ngi-a dea i beti m-i tathi
kl-!ll kul-see-1k-O:!8o go lOc river cON}-lOc ocean
'they watched it [a crocodile] in the river and in the sea'
ln the vast majority ol cases ti appears with the unmarked and m- prelixed
subject relerence pronoun. ln these instances ti marks realis mood. compare the
122
lollowing.
(208) a. e mai
!8o come
'he comes' or 'he came'
b. t-e mai
kl-!8o come
'he comes' or 'he came'
c. e mai t-i
!8o come kl-lOc
'he came linis"'
lxample (208)a. presents a simple (i.e., unmarked lor 1Am) sentence. uue to the
unmarked subject relerence pronoun, the status ol his coming is not clearwhether
realis or irrealis, just begun, ongoing or completed, repeatedly or just once, past,
present, or luture. lt could be any ol the above. lxample (208)b., in contrast, is marked
realis with te 'kl-!8o', the lact ol the subject's coming is established. Again, however, the
hearer does not know enough to say whether he is coming or he has come, whether he
comes repeatedly or just this once. 1he addition ol ti in (208)c. adds considerably to the
understanding ol the nature ol the coming as compared to (208)a. and b. uespite the
lact that the subject relerence pronoun is not marked lor 1Am, (208)c. is understood to
mean that his coming is realis, completed, not repeated, and past.
lt is rare lor ti to appear in clauses leaturing k- 'irrealis' and t- 'realis' prelixed
subject relerence pronouns. ln the case ol the latter, realis mood is already marked in
sentences where a t- prelixed subject relerence pronoun is used, and with relerence to
the lormer, mixing realis and irrealis mood is, most olten, contradictory.
98
ln my
collection ol texts there is a single sentence (with two instances) ol ti occurring with a
98
1he realis luture is very rarely encountered in lengo.
12!
subject relerence pronoun marked lor realis mood (t-8kl). ln the same sentence there is
also one ol two examples ol ti appearing with the irrealis marked subject relerence
pronoun (k-8kl, the other instance is in (214)). lxample (209) is part ol a strongly
worded announcement by a village chiel. 1his is the way it is," he is saying in the
larger context. 1hese things are established and they will not change. uon't come later
on saying that you didn't understand my announcement."
(209) na vanga t-e mai t-i , vanga vuru t-e mai t-i , tha ba k-amu
Ak1 lood kl-!8o come kl-lOc meat dish kl-!8o come kl-lOc kll lL1 lkk-2ll
ghali-a ighamu tovothi na rurutu ni vanua ba k-amu rongo-ni -a t-i
make-O:!8o ll:2ll separate Ak1 group olN village lL1 lkk-2ll hear-1k-O:!8o kl-lOc
ngeni de
today ulm
'the lood and meat dishes are already prepared, what you dillerent groups will
prepare you will hear today, at this time'
ln (209) ti is used together with realis and irrealis marked subject relerence pronouns to
give strong emphasis to the lixed and established nature ol the lactsthose that
already exist (that some lood is ready in l. village) and those that have yet to be heard
but, with a strong measure ol certainty, will be (that you are responsible to prepare
some lood yoursell, which you willwithout doubthear about shortly). 1he linal
clause ol (209), ba kamu rongonia ti ngeni de 'you will hear it today', has a 1Am signature ol
luture realis. 1he certainty ol the people's hearing in the luture is established by ti.
contrast this with the sentences in (210).
(210) a. ara rongo-ni -a po teigha na ghani-ra na kau
!ll hear-1k-O:!8o llm Nlo Ak1 eat-O:!ll Ak1 dog
'they just heard it [a crocodile] only ate dogs'
b. pukua na mamanu veghe ia na vua t-e rongo-ni -ra na kau ma
because Ak1 animal like ulm Ak1 crocodile kl-!8o hear-1k-O:!ll Ak1 dog cON}
124
na ira ba k-e tapa dea ni ghuru-ra
Ak1 ll:!ll lL1 lkk-!8o run go olN chase-O:!ll
'because an animal like thata crocodile, it hears the dogs and the others
[and] will run go to chase them'
1here is no indication ol 1Am on the third person plural subject relerence pronoun ara in
(210)a. As was the case with (208)a. above, the status ol their hearing is unclear
whether realis or irrealis, just begun, ongoing or completed, repeatedly or just once,
past, present, or luture. And (210)b., as was the case with (208)b. above, is marked realis
with te '!8o', the lact ol the subject's hearing is established. Again, however, the hearer
does not know enough to say whether the crocodile is hearing or it has heard, whether
it hears repeatedly or just this once. Verbs ol motion and verbs ol perception behave
similarly with relerence to the realis marking properties ol ti.
A question remains: why not just use t-8kl to indicate realis moodwhy
introduce ti5 1he answer is that there are instances where it is prelerable, and others
where it is necessary, to mark the subject relerence pronoun other than realis (t-), but
that at the same time there is a realis sense that needs to be included at some point ol
the utterance. ln these instances ti is used. lor example, in (211) and (212), ti marks
realis mood with a temporal locative where the verb phrase is otherwise unmarked lor
1Am (by virtue ol the use ol an unmarked subject relerence pronoun).
(211) *mi-tu dea t-i i pono.
1lX.ll-lAL go kl-lOc lOc bush
'we lew went to the bush.'
lxample (211) presents an interesting case. ll ti is really 'kl-lOc', there seems to be a
locative lollowing a locative. lowever, it is worth bearing in mind that the locative i
marks both temporal and spatial location. 1he location i pono 'lOc bush' is spatial while
12
dea ti 'go kl-lOc' is temporal. 1he temporal J spatial nature ol the locative i will come up
again when instances ol ti being used with nouns are discussed below.
Again, in (212) where the subject relerence pronoun is unmarked lor 1Am, the
sentence is placed in realis mood by virtue ol ti.
(212) 0h, ara gito-a t-i na kei1
lN1} !ll steal-O:!8o kl-lOc Ak1 basket
'ley, they stole the basket!'
1he thiel and the basket were gone. 1he statement is in realis mood, despite the lack ol
the expected marking on the subject relerence pronoun.
1he situation with stative verbs, as in (21!), is much the same: ti marks realis
mood in an otherwise unmarked-lor-1Am clause.
(21!) 0 vonu t-i na tagi.
!8o lull kl-lOc Ak1 tank
'1he tank is lull [and overllowing].'
compare e vonu na tagi 'the tank is lull'. ln this latter case the tank is merely 'not
empty'. ln (21!), the lalling rain water is spilling out the top ol a tank lilled to capacity.
1he negative modal verb teigha also undergoes the inlluence ol ti, as in (214).
(214) 4a k-u teigha t-i na mono varongo i vanua.
lL1 lkk-18o Nlo kl-lOc Ak1 stay quiet lOc village
'l really won't be sitting around in the village.'
1he speaker, alter enumerating a list ol things he was responsible to do in the capital
city and other provinces during the upcoming year, was overcome with the realisation
ol (214), it crossed his lace like a cloud on a sunny day. lt was not a realisation with
some 'ludge lactor' built into it. 1he negative did not leave any 'room to move', it was a
sure, a lixed, a complete negative. 1here was no way out ol the responsibilities and he
would really not be spending much time in his village.
126
lt is worth asking ol (21) whether ti is modilying the verb leboi 'lloat' or the
adverb laka 'also'. 1hat is, is it the 'lloating' that the speaker wishes to establish as
realis, or the 'also-ness' ol the lloating7
(21) e ghua logho bo(bona vata i nganiba na nulavi m-ami-tu
!8o il have kul-time continue lOc later Ak1 alternoon cON}-1lX.ll-lAL
dea lebo-i laka t-i
go lloat-O:!ll also kl-lOc
'il there's still time later in the alternoon, we'll also lloat them (down-river)'
8ome more inlormation about the situation will help clarily. 1he speaker had been
looking all over the village lor his boys and linally lound them near our house. le
wanted to take them into the bush to retrieve some house-building materials that he
had set aside previously. laving spent considerable time looking lor his boys, the time
lelt to get the job done was short (and telling me the story lelt him even less!). 1hey
needed to walk some two hours, gather the timber, carry it to the river and then, il
there was still time, they would also lloat it down the river to the village. lut the
lloating, as the text continues to make clear, had two options: to a hall-way point where
they would throw it up on the riverbank and carry it the rest ol the way the next day
(to avoid travelling during darkness in crocodile inlested waters), or all the way to the
ocean and along the beach to the waterlront by his house. 1his either J or nature ol the
'lloating' makes it unlikely that ti is modilying leboi 'lloat them'. consider the case ol
laka 'also'. lxample (21) opens with a conditional (e ghua '!8o cONu'), which lends a
measure ol uncertainty (in the lorm ol a conditional) to the sentence. 1he speaker
really wants the 'also' to happen, though, he doesn't want to have to go back to the
bush again tomorrow. le indicates this with laka ti.
12
Numerals can also be modilied with ti, as in (216).
(216) B0 ngitha t-i niulu-a5C BThangavulu tolu t-i na vula.C
!8o how.much kl-lOc year-l8:!8o ten three kl-lOc Ak1 moon
'low old is he7' '1hirteen months.'
1he situation involved one woman asking another how old her toddler was. 1hey
hadn't seen one another lor some time so while the woman who asked could tell by
looking at the child that he was approximately one year old she couldn't recall exactly
how old he was. And this is what she asked by lraming the question with the realis
locative ti. 1he addition ol ti indicates a desire to 'locate' the child's actual age with a
certain amount ol precision. 8o even though the question was asked in niulu 'years', the
answerin response to the request lor a more precise responsecame in vula 'months'
(lit., 'moons'). 1he realis temporal locative works well in the mother's response to
establish and conlirm the precision ol the number ol months.
Quote margins are also verbal and may be modilied by ti.
(21) B? doku,C ena t-i na vonu.
lN1} good say:!8o kl-lOc Ak1 turtle
'Oh good," said the turtle.'
lnterestingly, quote margins are rarely marked lor 1Am. 1he lew instances there are in
my texts are ol the type mena 'cON}-say:!8o' and gena 'llc1-say:!8o'. 1he absence ol tena 'kl-
say:!8o' may be to avoid conlusion with similar wordlorms tena 'lOc' and teHighaI na 'Nlo
Ak1'.
99
Ol course, realis mood can be readily, and unambiguously, marked on quote
margins with ti.
As one might expect, ti appears with prepositions in a spatial locative way with,
ol course, realis overtones. ln (218) the lirst instance ol ti modilies the verb phrase o
99
Although mena 'and J but now' could easily be conlused with mena 'cON}-say:!8o'.
128
ghua ngaoa na ghalia 'suppose you want to make' and its object complement siki Be:tra
flavourC 'any extra llavour'. lt lorms a conditional presumed true lor the sake ol
argument. lt is the second ti, however, the one that modilies the preposition i leoa 'lOc
inside', that is ol interest here.
(218) ? ghua ngao-a na ghali-a siki 'e.f.' t-i i leo-a t-i
28o il want-O:!8o Ak1 make-O:!8o any e.l. kl-lOc lOc inside-O:!8o kl-lOc
nimoa na thara k-o ghali mai-a.
l8:28o Ak1 least.row lkk-28o make come-O:!8o
'll you want any 'extra llavour' in your least row, you make it and bring it.'
1he second ti gives emphasis to the location ol the 'extra llavour'inside your least
row, not outside (i.e., distributed to a dillerent least row). 1he particle ti establishes
spatially where the 'extra llavour' is to appear. ll this was i leoa nimoa na thara ti, with ti
modilying the noun thara, there would be some place other than a dillerent thara 'least
row' that the extra llavour could end upin someone's house, lor exampleand
therelore the thara 'least row' was being specilied with the spatial realis locative.
A second example ol ti with a preposition is in (219). Again, it is the second
instance ol ti in the sentence that is ol interest at this point.
(219) 4iku inau m-u rai mai inau m-e eno t-i
sleep ll:18o cON}-18o wake come ll:18o cON}-!8o lay.down kl-lOc
ko-levugha-mami t-i a J. g-i na gari mane.
uL-between-l8:1lX.ll kl-lOc Ak1 } llc1-lOc Ak1 child man
'l was sleeping and l woke up and laying right between }. and me there was a
youth [who wasn't there when l went to sleep].'
1here is an element ol shock in (219). lt comes lrom a text recounting the night that the
narrator's basket had been stolen. le was in town with a lriend, }., sleeping in a semi-
public place when, in the middle ol the night, he awoke to lind a third person asleep
129
with them. And not only with them but between themright between them! lt is ti, the
realis locative, that narrows the locus to 'right between', kolevughamami 'between us
two' is one thing, but kolevughamami ti 'right between us two' is quite another. 1he
particle ti highlights the unexpected nature ol the sleeping youth's location.
what began as a pesky little particle
100
is somewhat more tractable with the
analysis provided. 1hough a 'realis locative' is, as lar as l know, not attested in other
Oceanic languages, it is a surprisingly llexible particle in lengo.
1he situation becomes somewhat more complex outside the verb phrase. low is
it possible to account lor a noun or a pronoun with realis mood7 consider (220).
(220) Mo(mono t-i sakai na dani na vonu t-i ma na igha.
kul-stay kl-lOc one Ak1 day Ak1 turtle kl-lOc cON} Ak1 lish
'Once upon a time there was a turtle and a lish.'
1he stative verb mono 'stay, exist, remain' is directly modilied by tipart ol a story-
telling lormula along the lines ol 'once upon a time'. lere the temporal sense ol the
realis locative at work. lut na vonu ti 'Ak1 turtle kl-lOc' is lound later in the sentence. lt is
not the case that the second ti is merely dislocated lrom the verb, while reduplication
lor emphasis is a leature ol lengo, *ti ti is never lound. 8o it seems to be the case that ti
here is modilying the noun vonu. And in case the reader might think (220) is an
anomaly, later in the same story an identical situation is lound as shown in (221).
(221) Aoku na igha t-i t-e vuivuni-a.
ok Ak1 lish kl-lOc kl-!8o start-O:!8o
'Ok, the lish started it [the competition].'
ln (220) and (221), ti lunctions almost as an emphatic demonstrative. 1hese sentences
100
A pesky little particle is a morpheme that is either lexically contentless or behaves in a text in a
manner that is anomalous with regard to its usual lexical content, and is intractable to grammatical
analysis"
(http:JJwww.sil.orgJlinguisticsJolossaryOllinguistic1ermsJwhatlsAleskylittlelarticle.htm).
1!0
come lrom a 'custom story'a story about why things are the way they are. while both
vonu 'turtle' and igha 'lish' are modilied by ti, it happens at dillerent points in the story
they are never both modilied by ti at the same time. lxample (220) comes as the
opening line ol the story. lere at the outset the turtle is highlighted by ti. 1he turtle, as
the 'winner' ol the competition, is the hero ol the story. ln lact, due to the nature ol the
competitiona game ol lollow-the-leaderthe lish dies: it lollows the turtle onto the
shore but cannot make it back to the water. 8o in (220) ti serves to establish the turtle as
the protagonist. ln order to wrest the spotlight away lrom the turtle, when the lish sets
out as the 'leader' in (221) it is marked with ti.
1he 'demonstrative' lunction ol ti is apparent when it lollows an emphatic
pronoun.
(222) ma inau t-i na tighi mane8 t-u kabi
cON} ll:18o kl-lOc Ak1 lirst man kl-18o dig
'and l was the lirst man, l dug'
lere inau ti 'll:18o kl-lOc' is also, in a way, an emphatic. 1here were a number ol young
men at the job site, a number who used the shovel to dig. later in the text, one young
man even grabbed it lrom the hands ol another deemed to be digging too slowly. lut
the narrator, inau ti 'll:18o kl-lOc'and not anotherwas lirst. 1he lact that he was lirst
among many is marked emphatically by ti.
1he rellexive pronoun can also be modilied by ti to specily more exactly. ln (22!)
group l. is being used as an example lor the rest ol the village groups ol what will
happen during the upcoming least.
(22!) 4oro k-e ghe dea tena group sakai, boro tena group ruka, ba k-e
Nlo lkk-!8o continue go lOc group one Nlo lOc group two lL1 lkk-!8o
1!1
masi ba oli ba tena nimiu na thara tibo-miu t-i ighamu na ).
must lL1 return lL1 lOc l8:2ll Ak1 least.row klll-l8:2ll kl-lOc ll:2ll Ak1 l
'lt cannot go to group one, it cannot go to group two, it must return to your
least rowyourselves [group] l.'
A main point ol concern in the text is that the larger village groups will not have
enough lood allotted to them by the sponsoring group. 1he narrator goes to great
lengths to assure people that the lood that belongs to them will get to their group and
to no other. lt is dillicult to break the lengthy lengo possessive phrase nimiu na thara
tibo-miu ti ighamu na ). into something equivalent in lnglish, but 'your least row you
yourselves alone l.' comes close. 1he role ol ti is, again, as an emphatic demonstrative,
a realis substantiator.
1he particles po 'llm' and ti work so closely together that at times they seem to
be a single word: poti. lndeed, many lengo speakers consider it one. lven il it is a single
word, however, it is surely a compound ol the two. lor the purposes ol this grammar po
and ti are analysed as two words. 1he construction po ti is most commonly present
lollowing a verb, as in (224).
(224) m-ami-tu mai atheathe po t-i de ga
cON}-1lX.ll-lAL come rest llm kl-lOc ulm there
'we came just to rest here' or 'we came to just rest here'
1he construction po ti serves to diminish and establish, to limit and lend substance, to
the verb. what po takes away lrom the verb ti gives back, so that the action (or state, in
the case ol statives) is not given too much emphasis, on the one hand, nor too little, on
the other. ln (224) it was not excruciating labour that the people were resting lrom
(thus po), but they had worked, the sun was hot, and they were resting nonetheless
(thus ti), it was not undeserved rest (ti), but there was work lelt to do and they couldn't
1!2
remain idle lor long (po).
lxample (22) is an equative clause in which the subject pronoun is juxtaposed
to a subject noun phrase, and in this case, interposed with po ti.
(22) @eia po t-i na theutu.
ll:!8o llm kl-lOc Ak1 way
'[1hat's] just the way [we need to do it].'
1his example is lrom a text in which two people were discussing the implications ol a
recent crocodile attack at the river near the village. 1he conversation quickly turned to
how to eliminate the threat. Lpon coming up with a plan ol action, the originator ol the
plan uttered (22). le acknowledged that it was all they needed to do (po), that it
needed to be implemented as soon as possible (ti).
1he analysis ol ti is complex, all the more so in light ol its wide distribution.
lowever, the analysis provided herethat ol a realis locative, a substantiator in time
and space (as context requires)makes some sense in all the contexts in which ti is
lound. oiven that lengo separates the 1Am notional space into the broad categories ol
realis and irrealis, the presence ol an independent and highly llexible realis marking
particle is not surprising. what is unexpected is its shape, that is, a 'realis locative'.
4. Verb negati on
1here are a number ol ways to negate a verb in lengo. 1here is a grammatical
negator, a discontinuous morpheme, that surrounds the verb being negated. 1his is
supplemented by a variety ol modal verbs with negative meanings that are part ol
serial verb constructions (see page 14ll). 1hese range in meaning lrom a strong
prohibitive to a lack ol willingness.
lengo's most grammaticalised negator is the discontinuous morpheme mo . . .
1!!
mo 'Nlo . . . Nlo'. 1his negator is the only 'discontinuous' morpheme in lengo. 1he
negative construction mo . . . mo surrounds the verb being negated as in (226).
(226) a. )-amu mo sapono mo1
lkk-2ll Nlo loud Nlo
'le quiet!' (lit., 'You(ll)not so loud!')
b. Mo ole mo.
Nlo walk Nlo
'uon't walk.'
c. Mo thaghata leo mo.
Nlo bad inside Nlo
'uon't worry.'
An intriguing use ol the mo . . . mo negative construction is lound in the
lollowing quote lrom a letter to the editor ol the national newspaper.
moreover, a lot ol lengo people, young and old, nowadays increasingly
incorporate lnglish words with language when conversing. lor instance,
'you must come' in lnglish, is interpreted as, ko mo ghe teigha mo na mai in
lengo. lut many people now increasingly use 'must' instead ol ko mo ghe
teigha mo. As a result today in lengo language people will say something
like ko must mai ti, inserting 'must' (the lnglish word, 'must') instead ol
saying ko mo ghe teigha mo na mai' (Allred mane, lave kide [sic, pride] in
Your language", 8olomon 8tar, 9 August 2004).
uuring some lour years in a lengo speaking context, l never heard anything but lnglish
'must' (or lijin masi) in lengo conversation (see examples (111) and (22!)). 1his is not
surprising, given that the substance ol the above letter was an encouragement lor
lengo speakers to strive to maintain their language in the lace ol signilicant pressure
lrom other languages (notably lnglish and 8olomon lslands lijin).
uespite its inlrequent use, this example ol the negative mo . . . mo deserves a
closer look. 1he segment ol text is presented in (22).
1!4
(22) k-o mo ghe tei gha mo na mai
lkk-28o Nlo continue Nlo Nlo Ak1 come
'you must come' (lit., 'you must not not continue to come')
what lengo does here is apply a double negative to allect a strong allirmative with
deontic modality. 1he verb mai 'come' is negated by teigha 'Nlo'. As is the case with verbs
negated by the modal teigha 'Nlo', mai 'come' is nominalised and becomes na mai 'Ak1
come'. 1he negative modal teigha, being head ol the verb phrase, is in turn negated by
mo . . . mo. 1he auxiliary ol aspect, ghe 'continue', is within the discontinuous negative
mo . . . mo. A demonstration ol the possible progression lrom imperative through
negative imperative to strong imperative is shown in (228)a.-c.
(228) a. k-o ghe mai
lkk-28o continue come
'you, continue coming'
b. k-o ghe tei gha na mai
lkk-28o continue Nlo Ak1 come
'you, don't continue coming'
c. k-o mo ghe teigha mo na mai
lkk-28o Nlo continue Nlo Nlo Ak1 come
'you must come'
lt is impossible to say whether this is a one-step or a two-step process, that is, whether
a speaker actually goes through the second step represented in (228)b. or whether
(228)c. lollows directly on (228)a. in moving lrom an imperative to a strong imperative.
what is certain, however, is that the complexity ol this construction is avoided by
borrowing the lijin lorm masi 'must'.
4. 8 Verb seri al i sati on
8erial verb constructions are prevalent in lengo. As evidence ol this lirst
consider two verbs, tapa 'run' and tumuri 'lollow', in separate examples in (229). ln
1!
(229)a. and b. tapa 'run' and tumuri 'lollow' are the solitary verbal heads ol their
respective clauses.
(229) a. J)-a-tu oli ba t-i,C una m-u tapa.
lkk-1lN.ll-lAL return lL1 kl-lOc 18o:say cON}-18o run
'we need to go back now," l said and l ran.'
b. ! vi ga t-e dea ga na igha e tumuri po na vonu.
lOc where there kl-!8o go there Ak1 lish !8o lollow llm Ak1 turtle
'wherever the lish went the turtle just lollowed.'
ln (2!0), however, they together lorm the verbal head ol the clause.
(2!0) m-ara-ko tapa tumuri -u
cON}-!ll-uL run lollow-O:18o
'and they two ran alter me'
1he serial verb construction tapa tumuri 'run.alter' expresses a single action, has a
single subject (relerenced by arako 'they two') and a single object (-u 'O:18o'), does not
have a connective word between the verbs (i.e., they belong to the same clause), and
has a single 1Am signature (m- 'cON}', which carries the 1Am signature ol the preceding 1Am
marker). 1hese characteristics are consistent with the criteria lor analysing it as a serial
verb.
101
lynch et al. (2002) identily live semantic types ol verb serialisation in Oceanic
languages.
1. directional J positional: the lirst verb expresses movement, the second the
direction ol that movement or the position reached as a result ol that
movement,
2. sequential: the lirst verb expresses movement, the second the action that
lollows the movement. A purposive relationship between the actions is
usually implied,
!. causative: the lirst verb is transitive, the second expresses the result ol the
action ol the lirst. 1he object ol the lirst verb is subject ol the second,
101
8ee crowley 2002b:19.
1!6
4. manner: the second verb expresses how the action ol the lirst verb was
perlormed,
. ambient: the implicit (third person singular) subject ol the second verb is
the sub-event expressed by the lirst. 1he second verb is olten 'linish',
expressing completive aspect or sequence ('and then') (4ll).
102
lvidence ol each ol these types is lound in lengo. ln addition to these, lawley (19!)
and later crowley (2002b) make mention ol prepositional verbs - verbal prepositions
lor such semantic categories as comitative, dative and instrumental (among others).
1hese, too, are common in lengo. A serial verb construction indicating purpose has the
genitive ni between the verbs. linally, there are modal serial verb constructions in
lengo in which the second verb is preceded by an article. 1he lollowing chart provides
an overview ol lengo's various serial verb constructions.
8erial verb constructions Non-serialised alternatives
uirectional V - V
dir
(-O)
8equential V
dir
- V
t
-O
V
i
- V
t
-O
causative V
t
- V
t
-O V
i
-caghini-O
10!
manner V
i
- V(-O)
Ambient V - soko(-O)
comitative V - kolu-O
uative V - vani-O V
t
-lO uO
104
lnstrumental V - ghini-O
modal V - na V(-O)
Figure F.F: T<pes of #engo serial verb constructions
lach ol these is discussed in turn below.
102
1he term 'ambient' derives lrom chale (190:101-2), and an ambient verb makes a general
predication about the world without relerring to any particular participants, such as: lt pours in 8uva
most ol the time. Ambient serial constructions are those in which a serialized verb makes some kind
ol qualilication about the manner in which an action is perlormed, with the manner being expressed
by means ol a serialized stative verb" (crowley 2002b:41).
10!
8ee the discussion ol causative on page 84.
104
8ee the discussion ol transitive verbs with two objects on page 1.
1!
4. 8. 1 ui recti onal
1here are a number ol words that indicate direction in lengo. 1hese include mai
'come, toward', dea 'go, away lrom', thai 'arrive at',
10
iti 'up' and thivo 'down', as well as
cardinal" directions boko 'down.coast', ghalagha 'up.coast', longa 'landward', and sapa
'seaward'. 1hese are lound lollowing the verb they supply a directional component to.
when directional words are lound in a serial verb construction with a transitive verb,
the directional takes the object pronoun sullix that would otherwise attach to the
transitive verb.
106
1he directional words are used with relerence to the narrator's perspective:
10

one can athe dea 'give go' to another or ask another to athe mai 'give come' to hersell,
one can sara mai 'arrive come' or sara dea 'arrive go'. 1he sentences in example (2!1)
illustrate this. 1hey are the second and the linal sentences respectively in a story
recounting a trip to the river, the location ol the narrator's garden. 1he narrator's
deictic centre is his home village. 1he story opens with: ikoghami a /aul amiko dea i
nughu 'we two laul [and l] we two go to the river'.
108
lt continues:
(2!1) a. ami-ko tumuri-a na theutu g-ami-ko sakai ni sara dea ga
1lX.ll-uL lollow-O:!8o Ak1 path 1lX.ll-uL one olN arrive go ulm
'we two lollow the path belore we arrive there'
b. mamiko sara mai te nulavi ti i vanua
m-ami-ko sara mai tena nulavi t-i i vanua
cON}-1lX.ll-uL arrive come lOc alternoon kl-lOc lOc village
'and we two arrive in the alternoon in the village'
Arrival at the destination is marked with sara dea 'arrive go' and arrival back in his
10
1he directional thai 'arrive at' expresses arriving at the goal ol an action: il looking, lastening one's
gaze on an object, il searching, linding what is sought, and so on. cl. lijin kasem.
106
1his behaviour is seen in (2!!)a. and b., (2!4), (2!)a, (2!6), (2!)a., and (2!8) in this section.
10
while a good number ol the texts in my collection are lirst person narrative, there are some that
indicate that this applies to the more general deictic centre ol the text, as in (2!6) below.
108
Note the use ol dea 'go' as head ol the verb phrase.
1!8
village is rendered as sara mai 'arrive come'. oiven that the narrator is situated in the
village while recounting the trip, the directionals dea 'go, away lrom' and mai 'come,
toward' rellect the relationship ol the two points ol arrival to the place where the
narrator is at the time he tells the story.
lt needs to be said near the outset ol the discussion that the verbs that lunction
as directionals can also appear as the head ol a verb phrase. 1he sentences in (2!2)
demonstrate this.
(2!2) a. Mai na vothe.
come Ak1 paddle
'[land me] the paddle.'
b. m-e mai na ara sule m-e kovo-thi-a na vae
cON}-!8o come Ak1 wind big cON}-!8o collapse-1k-O:!8o Ak1 house
'and a big wind came and it collapsed the house'
c. 4a k-o dea i leghai5
lL1 lkk-28o go lOc garden
'will you be going to the garden7'
d. m-e thai -a mo na dani t-ara-ko ghunughunu-a
cON}-!8o arrive.at-O:!8o 7 Ak1 day kl-!ll-uL plan-O:!8o
'it arrived, the day they had planned lor'
e. t-ara i ti -a na thinaghe
kl-!ll up-O:!8o Ak1 canoe
'they beached (lit., upped-it) the canoe'
l. m-ami-tu sapa-a na thinaghe
cON}-1lX.ll-lAL seaward-O:!8o Ak1 canoe
'we lew [carry] the canoe seaward' (lit., seaward-it the canoe)
ln each ol these examplesin imperative and indicative sentencesthe head verb is a
word that also, in other contexts, can lunction as a second verb in a directional serial
verb construction.
lt is the phenomenon ol these words combining with other verbs as directional
1!9
verbs that is ol interest here. uirectional verbs are lound lollowing verbs ol motion,
speech, and perception.
uirectional verbs combine with verbs ol motion, as in example (2!!).
109
(2!!) a. ami-ko teigha na ta(tavu thai -a na kei
1lX.ll-uL Nlo Ak1 kul-lind arrive.at-O:!8o Ak1 basket
'we couldn't lind the basket'
b. k-o tal u dea-a i te siki popo
k-o talu dea-a i tena siki popo
lkk-28o put go-O:!8o lOc lOc any bowl
'you put it inside a bowl'
c. m-e pasu mai i riki-gu na aliva
cON}-!8o sting come lOc digit-l8:!8o Ak1 centipede
'and it bit [me] on my toe the centipede'
ln the lirst two ol these examples the object sullix joins to the directional verb.
1he context lrom which example (2!!)c. comes is helplul in demonstrating the
nature ol the dillerence between the dynamic and directional uses ol these verbs.
consider the use ol mai 'come' in (2!4).
(2!4) u dea m-u tu-ri thai-a t-i na aliva, tu-ri
18o go cON}:18o stand-1k arrive.at-O:!8o kl-lOc Ak1 centipede stand-1k
thai-a t-i na aliva t-e mai pasu i riki-gu, m-e
arrive.at-O:!8o kl-lOc Ak1 centipede kl-!8o come sting lOc digit-l8:18o cON}-!8o
pasu mai i riki-gu na aliva m-u . . . 8 na bona t-e mai ghani-u
sting come lOc digit-l8:18o Ak1 centipede cON}:18o Ak1 time kl-!8o come bite-O:18o
laka po na koba po8 na bona t-e mai ghani-u m-u kimu-a
also llm Ak1 hermit.crab llm Ak1 time kl-!8o come bite-O:18o cON}:18o kick-O:!8o
109
1his set includes: alo 'beckon', athe 'give', bila 'throw', gara 'pull', ghali 'make, do', gharasu 'move', lavi
'take', lebo 'lloat, lusa 'load', oli 'return', rai 'awaken', sara 'arrive', sovo 'join', talu 'put', tapa 'run', thevu
'cross', tovothi 'separate', and vuli 'pour', among others.
140
m-e tumu lithi
cON}-!8o lall beside
'l went and l stepped right on a centipede, [l] stepped right on a centipede that
came stung my toe, and it stung come my toe the centipede and l . . . , when it
came bit me [it lelt like] just a hermit crab, when it came bit me l kicked it and
it lell beside'
Ol the lour times that mai 'come' is lound in this section ol text, three times it is lirst in
a serial verb construction and once it is lound last. when mai 'come' is lirst in a serial
verb construction it is used as a dynamic verb, that is, a verb ol motion as part ol a
sequential serial verb construction (see below). while it is true that the storyteller
stepped on the centipede, he lets the hearer know that the centipede played its part in
stinging him. 1he centipede had to orient itsell so as to sting the speaker, it had toin a
dynamic sense'come' sting him. when mai 'come' is last in a serial verb construction,
as it is in the second instance in (2!4), it is not being used in a dynamic sense but a
directional one. 1here is an important shilt ol the narrative spotlight due to this shilt
lrom dynamic to directional mai 'come'. 1he narrator begins with his role ol stepping
on the centipede (u dea mu turi thai ti na aliva 'l went and l stepped right on a
centipede'). le repeats his role (turi thai ti na aliva 'stepped right on a centipede') but
and signilicantlydrops any relerence to himsell, the expected subject relerence
pronoun is absent. At this point the spotlight swings to the centipede and its active role
in stinging him, expressed by the relative clause te mai pasu i rikigu 'that came stung my
toe'. with the next clause, however, the spotlight begins to swing back to the narrator.
1he narrator begins to set up lor some sort ol statement, possibly about his response to
the sting (me pasu mai i rikigu mu . . . 'it bit [me] on my toe and l . . . '), but he delays this
to comment on how the sting lelt. lecause there is no passive in lengo, the centipede
141
retains its active role in stinging the narrator, but with the use ol directional mai 'come'
the orientation ol the spotlight has changed. 1he listener is invited to take the
narrator's perspective on the centipede's stingits bite was directed at the narrator.
lrom the centipede's perspective it 'stung go' whereas lrom the narrator's perspective
the centipede 'stung come'. lt was a sting directed toward the narrator's perspective.
ln (2!)a. and b. there is a three verb serial verb constructiondea 'go' -
V(motion) - V(direction).
(2!) a. m-u ghoni soko-i m-ami-tu dea tal u i ti -i
cON}-18o make linish-O:!ll cON}-1lX.ll-lAL go put up-O:!ll
itu-ghami 4. ma 9.
lAL-ll:1lX.ll l and N
'l linished making them and we went and put them up, l. and N. [and l]'
b. k-ami-tu dea thua sapa i mata ni beti
lkk-1lX.ll-lAL go pole seaward lOc eye olN river
'we lew will pole seaward to the river mouth'
ln these examples dea 'go' is not lunctioning as the directional complement but as
another dynamic verb, the directionals are iti 'up' and sapa 'seaward'.
uirectionals, especially dea 'go' and mai 'come', also combine with the verb ol
speech J address ghu 'shout'. whether one addresses another (dea) or is addressed by
another (mai) is indicated by the directionals.
(2!6) t-e ghu-vi dea-a a tau-a
kl-!8o shout-1k go-O:!8o Ak1 spouse-l8:!8o
'she shouted to her spouse'
1his does not conlorm to the subtype, based on the transitivity ol the lirst verb, as
presented in lynch et al. (2002:4). 1hey claim that il the lirst verb is transitive that the
moving (or, more generally, acting) thing or person is object ol the lirst verb and
142
subject ol the second. ln (2!6) the acting person is subject ol both verbs.
linally, directionals are seen to play a role in relation to verbs ol perception, as
with bere 'see' and rongo 'hear' in (2!).
(2!) a. teigha siki agutu k-ami-ko bere-ngi thai -a iko-ghami
Nlo any something lkk-1lX.ll-uL see-1k arrive.at-O:!8o uL-ll:1lX.ll
'we two didn't lasten our gaze on anything [i.e., what we were looking lor]'
b. !ghamu na leoni tinoni k-amu rongo mai .
ll:2ll Ak1 all person lkk-2ll listen come
'You, all the peoplelisten here [i.e., in the speaker's direction]!'
1he situation in (2!8) is quite dillerent lrom that lound in (2!)b.
(2!8) k-amu mai rongo-ni -a na tha ba k-a ghali-a
lkk-2ll come hear-1k-O:!8o Ak1 cOml lL1 lkk-1lN.ll do-O:!8o
'come listen to what we're going to do'
ln (2!8) the speaker is coaxing people to move closer to him so that they can hear what
he has to say without him having to shout or strain his voice, whereas (2!)b. is an
attention-getting introduction. 1he people are already nearby, but distracted and
talking among themselves, they just need to be encouraged to direct their listening
attention toward the speaker. lxamples (2!)b. and (2!8) reinlorce the dillerence that
word order makes with, lor example, mai 'come'. when mai 'come' is at the beginning ol
a serial verb construction it is dynamic, that is, a verb ol motion. when it lollows
another verb in a serial verb construction, as in (2!)b., it indicates direction.
4. 8. 2 8equenti al
A second type ol serial verb construction is, according to lynch et al., sequential.
ln this type the lirst verb expresses movement while the second verb indicates an
action lollowing the movement. 1he sequential ordering ol the verbs suggests a
relationship ol purpose between the verbs. consider example (2!9).
14!
(2!9) m-ara-ko tapa tumuri -u
cON}-!ll-uL run lollow-O:18o
'and they two ran alter me'
1he verb ol motion is the lirst verb: tapa 'run'. 1he second verb, tumuri 'lollow' is the
action that lollows the movement. 1he lirst verb necessarily precedes the second. '1hey
two' began to run and their running was with the purpose ol lollowing 'me'. 1hey did
not just run aimlesslythey ran alter someone, here with the purpose ol coming to the
storyteller's aid.
lxample (240) provides more evidence ol a sequential type serial verb
construction.
(240) ! tena sanga-dira, muri na gilu-a na sipsip nga, t-e bere doku
lOc lOc speech-l8:!ll behind Ak1 skin-l8:!8o Ak1 sheep ulm kl-!8o look good
m-i leoa na dou-dira nga t-e veghe na kau asi. 4a k-amu
cON}-lOc inside Ak1 heart-l8:!ll ulm kl-!8o like Ak1 dog wild lL1 lkk-2ll
bere-ngi ghi l aghana-ra tena agutu t-ara ghali.
see-1k understand-O:!ll lOc thing kl-!ll do
'ln their speech, behind a sheep skin, it looks good but inside their heart it is
like a wild dog. You will understand them by seeing the things they do.'
ln this example the lirst verb, while not a verb ol 'motion' per se, is necessarily
preliminary to the second verb. Lnderstanding 'them' is contingent on seeing the
things 'they' do. uespite the lact that there are two verbs, however, they are so closely
related, so much ol a piece, that the two together lorm a serial verb construction with a
single meaning: 'see understand'.
Another common type ol sequential serial verb construction involves the verbs
ol directiondea 'go' and mai 'come'in lirst position. when in this position dea and
mai are dynamic and not directional.
144
(241) a. m-u dea thuge-a a J. BJ. o teigha na bere na kei-gu5C
cON}-18o go ask-O:!8o Ak1 }. }. 28o Nlo Ak1 see Ak1 basket-l8:18o
'and l went and asked }., }., you haven't seen my basket7"'
b. m-ara ghe dea tura-a niu tena thinaghe
cON}-!ll continue go load-O:!8o coconut lOc canoe
'and they continue going loading coconut in the canoe'
c. e mai biku kolu-ko-ghami
!8o come sleep accompany-uL-O:1lX.ll
'he came sleep with us two'
d. k-amu mai rongo-ni-a na tha ba k-a ghali-a
lkk-2ll come hear-1k-O:!8o Ak1 cOml lL1 lkk-1lN.ll do-O:!8o
'you come hear what we will do'
ln the examples in (241) the lirst verbs in the serial verb constructionsdea and mai
are verbs ol physical motion with a subsequent action expressed by a second verb.
A linal type ol sequential serial verb constructions also involves dea 'go' in lirst
position. ll the preceding sequential serial verb constructions are typilied as 'motion-
action', this type might be best characterised as 'become-state'. consider (242).
(242) m-e dea kola po t-i na igha
cON}-!8o go tired llm kl-lOc Ak1 lish
'and the lish just became tired'
ln this example dea 'go' does not express physical motion, rather it indicates the sense
ol moving lrom the state ol having energy to tiredness.
4. 8. ! causati ve
A third type ol serial verb construction is one that lorms a causative. 1he lirst
verb is the action while the second indicates the outcome ol the lirst. 1his is shown in
example (24!).
(24!) k-ara masi vana-thi mate-a t-i k-e sakai ni ghani mate-a
lkk-!ll must shoot-1k die-O:!8o kl-lOc lkk-!8o one olN bite die-O:!8o
14
siki tinoni
any person
'they must shoot it to death belore it bites a person to death'
1he narrator's concern was with a crocodile in the river that had recently mauled
someone. lt needed to be killed belore it killed someone! 1he two serial verb
constructions are necessary to indicate the ultimate ends the narrator was relerring to.
1hey could well shoot the crocodile, but to really take care ol the problem they needed
to 'shoot die' it. 1he crocodile could bite someone (as it had, lortunately, the person
escaped with serious but non-lile threatening injuries), but it needed to be dealt with
belore it 'bite died' someone. 1he second verb, mate 'die', is the result ol the lirst verb,
the verbs are in a causal relationship.
4. 8. 4 manner
1he lourth type ol serial verb construction that lynch et al. mention is one that
indicates manner. ln manner serial verb constructions the second verb expresses how
the action ol the lirst verb was perlormed" (2002:4). ln (244) the stative verb seghi
'quick' describes the manner in which the hearer is meant to mai 'come'.
(244) )-o dea bosa vani-a . k-e mai se(seghi 1
lkk-28o go tell uA1-O:!8o 8 lkk-!8o come kul-quick
'oo tell 8. he must come quickly!'
1he cV reduplication here indicates intensityi.e., very quickly, it is dillicult to
understand the lunction ol the reduplication here as continuous or repetitive. ln (24)
the non-reduplicated lorm ol the verb seghi is lound.
(24) kabi 4. m-e teigha na sara soko seghi -a m-e dea 4.%.
dig l cON}-!8o Nlo Ak1 arrive linish quick-O:!8o cON}-!8o go l.V.
146
me laua tulithia na spedi lime me kabi sokoa
m-e lau-a tulithi-a na spedi lima-a me kabi soko-a
cON}-!8o grab-O:!8o take.out-O:!8o Ak1 spade hand-l8:!8o cON}-!8o dig linish-O:!8o
'l. was digging but he did not arrive quickly at linishing and l.V. went and
grabbed the spade lrom his hand and he linished digging it'
lere seghi 'quick' describes how l. lailed to dig. ln lact, he took so long to dig the post
hole he was working on that someone actually grabbed the shovel out ol his hand to
linish the job. Note that in this example seghi 'quick', the third and linal verb in the
series, is inllected with an object pronoun sullix.
1he speaker in (246) uses a serial verb construction ol manner to indicate that
he would not have much quiet time in his village during the upcoming year.
(246) 4a k-u teigha t-i na mono varongo i vanua.
lL1 lkk-18o Nlo kl-lOc Ak1 stay quiet lOc village
'l really won't be sitting around in the village.'
1he second verb, varongo 'quiet', describes the manner in which the lirst verb, mono
'stay', is carried out. Note here that the entire serial verb construction is negated by
teigha 'Nlo'.
4. 8. Ambi ent
lynch et al. mention a linal type ol serial verb constructionambient. 1his type
is delined by crowley (2002b:61) as a construction in which a verb is serialized to
another verb, but in which there is no specilic relerent associated with the subject ol
the serialized verb, and the verb simply describes a general predication." lynch et al.
(2002:49) note that the second verb is olten 'linish', expressing completive aspect or
sequence ('and then')". 1his type ol serial verb construction is evident in (24).
(24) m-u ghoni-a t-i na tughuru ni tu i levugha ni enoana
cON}-18o make-O:!8o kl-lOc Ak1 post olN stand lOc between olN jackpost
14
ma na ghai-be-tina. M-u ghoni soko-i m-ami-tu dea
and Ak1 main.crossbeam cON}-18o make linish-O:!ll cON}-1lX.ll-lAL go
talu iti tu-ghami 4. ma 9. Aea m-ami-tu tal u i ti soko-i
put up lAL-ll:1lX.ll l and N go cON}-1lX.ll-lAL put up linish-O:!ll
'l made the posts lor standing between the jackposts and the main crossbeam. l
linished making them and we lew went and put them up, we lew l. and N. [and
l]. [1ime] went and we lew linished putting them up'
1he completive aspect supplied to the serial verb construction by the verb soko 'linish'
is clearly evident.
110
1he second instance ol soko 'linish' in (24) could be seen as part ol
a three verb serial verb construction or as a directional serial verb construction
embedded within an ambient serial verb construction ([talu iti] sokoi '[put up] linish').
lxample (248) is another example ol the completive aspect soko 'linish' brings to
serial verb constructions.
(248) [+] pogho soko m-u dea pogho mami. oko m-u dea tughu
18o bathe linish cON}-18o go bathe lresh.water linish cON}-18o go change
m-u belo nulavi i nea.
cON}-18o ring.bell alternoon lOc yesterday
'Alter swimming in the ocean l went and bathed in lresh water. [laving]
linished l went and changed [my clothes] and rang the [church] bell yesterday
alternoon.'
ln a serial verb construction the verb soko is only ever lound lollowing another verb. lt
is, however, also lound clause initially, as in the second sentence ol (248). ln such
instances soko joins clauses at a discourse level along the lines ol: '[having] linished [X],
l did Y.'
4. 8. 6 lreposi ti onal verbs
1here are a handlul ol devices available in lengo to bring oblique arguments
110
1he verb soko 'linish' is a lexical perlect (that is, a grammaticalization ol the current relevance, at the
moment ol utterance, ol an event or state that occurred prior to the moment ol utterance"
[http:JJwww.sil.orgJlinguisticsJolossaryOllinguistic1ermsJwhatlsAlerlect.htm]).
148
into closer relationship with the head verb. 1hese are comitative kolu, dative vani, and
instrumental ghini prepositional verbs.
111
1o each ol these an object pronoun sullix is
joined to indicate person and number.
4. 8. 6. 1 comi tati ve kol u
1he comitative prepositional verb, which indicates the notion ol
accompaniment (someone acting or being with another), is expressed by kolu.
112
1he
comitative kolu is cross relerenced to its object noun phrase with an object pronoun
sullix which agrees in person and number. 1he prepositional verb kolu - object pronoun
sullix lollows the verb it complements and is lollowed by the object noun phrase, as in
(249).
(249) * J. te dea kol ua a teme i leo ni pono.
a } t-e dea kolu-a a tama-a i leo ni pono
Ak1 } kl-!8o go cOm-O:!8o Ak1 lather-l8:!8o lOc inside olN bush
'}. went with his lather to the bush.'
1he object noun phrase can be dropped as in (20)a. and b.
(20) a. m-e biku kol u-ko-ghami
cON}-!8o sleep cOm-uL-O:1lX.ll
'and he slept with us two'
b. ami talu kol u-i
1lX.ll put cOm-O:!ll
'we put them [the sticks] together'
when the Nl
O
is dropped, the larger context is the only place to access the relerenced
object.
111
lawley (19!:142) explains, 1he term 'prepositional verb' was used by codrington and other early
grammarians ol certain disyllabic lorms which connect a verb with its grammatical object. like
transitive verbs and transitive sullixes (and unlike true prepositions), such lorms are always lollowed
by an object pronominal sullix. A verb which precedes a prepositional verb is olten lormally
intransitive, i.e., does not carry a transitive sullix or pronominal sullix."
112
8hades ol kolu's meaning are lound in the compound words thudu.kolu 'meeting (lit., sit together)' and
vanga.kolu 'least (lit., eat together)'.
149
4. 8. 6. 2 uati ve vani
1he dative prepositional verb vani, again with an indexing object pronoun sullix,
behaves at dillerent times as dative and benelactive, expressing meanings ol 'to', 'lor',
and 'with relerence to'.
11!
lxample (21) demonstrates the most common use ol the lengo dative: to
point" to the indirect object, here the semantic 'goal'.
(21) a. inau ba k-u tugu vani -ghamu na tha ba k-e sara
ll:18o lL1 lkk-18o tell uA1-O:2ll Ak1 cOml lL1 lkk-!8o arrive
'l will tell to you what's going to happen'
b. u athe-a vani -gho na buka de
18o give-O:!8o uA1-O:28o Ak1 book ulm:8o
'l am giving you this book'
114
1he dative vani can be dislocated lrom a serial verb construction, as in (22).
(22) m-ami-tu dea toka-a na koga vani -a na kisini C
cON}-1lX.ll-lAL go cut-O:!8o Ak1 stick uA1-O:!8o Ak1 kitchen c
'we lew went to cut sticks lor c.'s kitchen'
lere there is a subject switch. 1he subject ol the lirst part ol the serial verb
constructiondea tokaais mamitu 'we lew'. 1he subject ol vania is na koga 'the sticks',
which were the object ol dea tokaa.
A somewhat dillerent shade ol the lengo dative is seen in (2!). 1he preposition
vani here is acting as a benelactive with the meaning 'lor (the benelit ol)'.
(2!) 6oropo rukana k-o pelu vani -u ruka na vugho
morning day.alter.tomorrow lkk-28o buy uA1-O:18o two Ak1 net
11!
cl. lOc *pani, concerning which lawley (19!:14!) observes, is reconstructible as an independent
verb meaning 'to give'. ln languages which rellect this lorm as a prepositional verb its lunction is
most olten as a dative."
114
compare transitive clauses with two objects in (!0!).
10
k-u ne vu(vugho ko-gha-da igha.
lkk-18o so.that kul-net uL-oral.cll-l8:1lN.ll lish
'Next tomorrow morning you buy lor me two nets so that l can net some lish
lor us to eat.'
Note in (2!) that the object ol the dative is expressed (-u 'O:18o) while the object ol the
verb pelu 'buy' is not. lt seems that the object ol the prepositional verb vani has more
topical presence than the object ol the head verb pelu, that is, the (potential) recipient
is more 'allected' than what he may receive.
4. 8. 6. ! lnstrumental ghi ni
1he instrumental is indicated by the preposition ghini - an object pronoun
sullix. koss (1988:!) makes some interesting observations about the supposed 'rellex'
ol lroto Oceanic *-aki[ni]
11
'instrumental' in lengo. ln lengo (ouadalcanal) (i) we lind
many rellexes ol the variant *-akini, (ii) some ol these appear to be separable lrom the
verb (not as rellexes ol *akini but ol *kini) and used as a prepositional verb, and (iii) the
uses ol *-akini go lar beyond the conlective, relective, or instrumental ol western
Oceanic".
116
1he observation ol particular interest here is his second. le provides lour
examples ol *-akini in lengo, two ol which are included below as (24) and (2).
(24) e-pusi-aghini-a na rave
8:!8o-chop-All-O:!8o Ak1 axe
'le chopped it with an axe.'
11
(2) e-pusi-a na ghai ghini-a na rave
8:!8o-chop-O:!8o Ak1 tree lkllV-O:!8o Ak1 axe
'le chopped the tree with an axe.'
118
11
ln (2004:0ll) koss uses a variant ol this lroto Oceanic lorm: *-akinHiI.
116
koss (1988:41) explains, l adopt the terms conlective and relective lrom larrison (1982), who in
turn borrows them lrom Arms (194). A conlective participant is typically a concomitant, occasionally
an instrument, with a verb ol motion (e.g. the coconuts in he came with the coconuts). A relective
participant is 'typically a stimulus (source, cause, reason or beneliciary)' (larrison 1982:189-190),
usually with a verb expressing a psychological state or action (e.g. }ohn in le thought about }ohn)."
11
where All = lOc lorm *-aki[ni].
118
where lkllV = prepositional verb.
11
1he only apparent dillerence between the two examples is the presence ol the Nl
O
na
ghai 'Ak1 tree' in (2). lt seems that ghini is doing the same thing in both examples:
indicating an object as instrument. lt also seems that, given the analysis ol pusi-a 'chop-
O:!8o' in (2), it would make sense to look lor the same kind ol thing in (24). lndeed,
(24) could be parsed as l have done in (26).
(26) e pusi-a ghini-a na rave
!8o chop-O:!8o lN81-O:!8o Ak1 axe
'he chopped-it with-it an axe'
when compared to (2), na ghai 'the tree' is the understood (i.e., dropped) Nl
O
ol pusi-a
'cut-O:!8o'.
lxample (26) has dillerent morpheme cuts lrom koss. le, arguing lor an
underlying lOc lorm *-aki[ni] 'instrument', views aghini in (24) as an allix to the verb
pusi 'chop'. l, on the other hand, understand the initial a ol koss' aghini as an object
sullix on the verb pusi and ghini 'lN81' as a new word. lnterestingly, koss presents
precisely this analysis in his second example ((2) above)apparently an alternate ol
the lirstwith the one dillerence being that ghini is analysed as a prepositional verb.
1his begs the question ol (24) and (2): which analysis is to be prelerred7 And why
two such dillerent analyses ol such similar data7 l am inclined to say that, in lengo, a
single analysis ol ghini as a prepositional verb marking 'instrument' is more coherent
with the data.
lxamples (2) and (28) are provided as corroborating examples ol this
analysis. 1hese are lrom my own data. ln (2) the object ol bila-a 'shoot-O:!8o' is not
made explicit with a lexical noun phrase, but the context identilied 'it' as a pig. 1he
object indexed by the instrumental ghini-a is na vatu 'a stone'.
12
(2) + bila-a ghi ni -a na vatu m-e sipu.
18o shoot-O:!8o lN81-O:!8o Ak1 stone cON}-!8o llee
'l shot it [a pig] with a stone and it lled.'
1his is similar to koss' lirst example, (24) above. 1he stone is the instrument used to
shoot the pig", the omitted object ol bila-a 'shoot-O:!8o'.
1he indexing ol objects in (28) is straightlorward, the noun phrase objects are
lound immediately lollowing the respective object pronoun sullixes. 1his is remarkably
similar to koss' example in (2).
(28) e ghoti-a na ghai ghi ni -a na ghau
!8o prune-O:!8o Ak1 tree lN81-O:!8o art knile
'he pruned the tree with a knile'
1he knile is the instrument used to prune the tree. 8o while object omission is possible
in lengo, it need not allect the analysis ol the instrumental ghini. lxamples where the
object is present should be used to interpret those in which it is absent.
lxample (29) also presents the instrumental in the lorm ghini. lt does, however,
raise another question: where is the object marker on the verb abu 'hit'7 1he object
marker -a on the instrumental ghini points to na ghai 'Ak1 stick', and one would expect
there to be an object marker on the verb abu 'hit' to point to the direct object na kau 'Ak1
dog'.
(29) * deni na ghai t-e ne abu ghi ni -a na kau na mane.
Ak1 ulm:8o Ak1 stick kl-!8o lmll hit lN81-O:!8o Ak1 dog Ak1 man
'1his is the stick with which the man really hit the dog.'
1he object ol abu is marked in (261), but not here in (29).
lnterestingly, the object noun phrases ol the head verb and the instrumental can
switch places with no additional marking, as in (260), though there is a corresponding
change in meaning.
1!
(260) * deni na kau t-e ne abu ghini-a na ghai na mane.
Ak1 ulm:8o Ak1 dog kl-!8o lmll hit lN81-O:!8o Ak1 stick Ak1 man
'1his is the dog which the man hit with the stick.'
Note that despite the lact that na kau 'the dog' and na ghai 'the stick' have 'switched'
places, na ghai 'the stick' remains the object ol the instrumental ghini.
lxample (261), with its much reworked word order (and omission ol the
demonstrative), provides somewhat ol a corrective lest one began to think that abu 'hit'
didn't take an object marker.
(261) T-e abu-a na kau ghi ni -a na ghai na mane.
kl-!8o hit-O:!8o Ak1 dog lN81-O:!8o Ak1 stick Ak1 man
'1he man hit the dog with the stick.'
ln light ol (261), how is the lack ol object marking on the verb in (29) and (260)
explained7 oiven that the role ol the instrumental is to increase the valence ol the verb,
it makes some sense that the object ol the instrument is indexed while the object ol the
verb is not. 1he valence-increased item is given a marked place in the clause while
other objects are optionally lelt unmarked.
4. 8. modal
lengo expresses live modes with serial verb constructions. 1hese are abilitive
(tangomana) , desiderative (ngao), negative (teigha), prohibitive (tabu), and non-volitive
(kou). 1he modal verb comes lirst in the serial verb construction while the second verb
is preceded by an article. 1he presence ol an article belore the second verb does not
make it a deverbal noun (i.e., derived noun) since it does not lollow the derivational
pattern lound in lengo (stem-initial cV reduplication, see page ) and it retains the
verb properties ol intransitive-to-transitive derivation with the -ci sullix and argument
indexing with an object pronoun sullix where applicable. 1he second verb may best be
14
considered an inlinitive lorm that is treated like a noun but retains verbal properties.
119
4. 8. . 1 Abi li ti ve tangomana
1he actual or supposed ability to undertake an action is expressed by the modal
verb tangomana. lt says nothing about actually perlorming the action, merely being able
to do so. lor example, being able to lish, as in (262)a., does not entail that one actually
lishes, just that the ability to do so exists.
(262) a. t-amu tangomana na ta(tagho
kl-2ll able Ak1 kul-lish.with.a.hook.and.line
'you are able to [go] lishing'
b. t-o tangomana na mate-a
kl-28o able Ak1 die-O:!8o
'you are able to kill it'
c. * thi ighoe t-o mono i buriti dea k-o gharasu mai k-o
Ak1 rel ll:28o kl-28o stay lOc back go lkk-28o move come lkk-28o
ne rongo-ni-a k-o ne teigha t-i na tangomana na ghe thuge
lmll hear-1k-O:!8o lkk-28o lmll Nlo kl-lOc Ak1 able Ak1 continue ask
na tha ba k-a ghali-a tena arere
Ak1 cOml lL1 lkk-1lN.ll do-O:!8o lOc 8aturday
'You who are at the back you move this way [so that] you hear [and] you are
not able to continue asking what we will do on 8aturday.'
ln (262)b. the second verb ol the serial verb construction, matea 'kill it', indexes the
object na vua 'the crocodile' which is understood lrom the larger context. 1he second
verb maintains its verbal properties. likewise, the linal verb ol the serial verb
construction in (262)c., thuge 'ask' is modilied by the 1Am auxiliary ghe 'continue'. Nouns
do not take auxiliaries ol aspect in lengo, despite being preceded by an article, thuge
retains its verbal properties.
119
1hese modal serial verb constructions are consistently translated by lengo speakers with a lijin (fo)
or an lnglish (to) inlinitive.
1
4. 8. . 2 uesi derati ve ngao
1he desire to do something is expressed with the modal verb ngao. 1his is
evident in (26!).
(26!) a. 9a bona t-u thairo u ngao-a po na tugu tauni-a
Ak1 time kl-18o work 18o want-O:!8o llm Ak1 tell around-O:!8o
na aro-gu t-e viti.
Ak1 shoulder-l8:18o kl-!8o hurt
'while l worked l just wanted to tell about my shoulder that hurts.'
b. o ghua ngao-a na ghali-a siki 'e:tra flavour' t-i i
28o il want-O:!8o Ak1 make-O:!8o any extra llavour kl-lOc lOc
leo-a t-i nimoa na thara k-o ghali mai-a
inside-O:!8o kl-lOc l8:28o Ak1 least.row lkk-28o make come-O:!8o
'il you want to make any extra llavour" inside your least row you make and
bring it'
ln (26!)a. a serial verb construction is lound lollowing the article lollowing ngao. 1he
position tauni 'around' is a member ol the direction J position class ol verbs discussed
in directional serial verb constructions on page 1!8.
Another means lor expressing desire is with leo 'want, desire'.
(264) e l eo-gu na inu beti bithi
!8o want-l8:18o Ak1 drink water cool
'l want a cold water drink' (lit. 'it [is] my desire [to have] a cool water drink')
1he 'desire' is the subject ol the clause. 1he lirst person singular agent enters the clause
as a direct possession sullix on the verb. 1he 'cold water drink' is the object, but the
slot where it should be marked on the verb is taken by the direct possession sullix.
4. 8. . ! Negati ve tei gha
ly lar the most common and versatile negator in lengo is the modal verb teigha
'Nlo, no, none'.
120
when teigha is used, the verb lollowing is preceded by an article. 1he
120
8ee uixon (1988:40) lor a discussion ol lijian semi-auxiliary verb" sega 'Nlo'.
16
resulting construction teigha na V is oltenand sometimes conlusingly (compare tena
'lOc')shortened to tena (i.e., teHighaI na) as (26) demonstrates.
(26) Bgami tena agri ighami,C gi gea ena
g-ami teigha na agri ighami g-i gea ena
llc1-1lX.ll Nlo Ak1 agree ll:1lX.ll llc1-lOc ll:!8o !8o:say
'we don't agree," they said'
8imilarly teigha siki is olten shortened to te si, as in te[igha] si[ki] tave '[the river] is not
llooded' (lijin no eni flood).
lxample (266) serves to demonstrate the lull lorm ol teigha and the article.
(266) a. na tha t-e mono t-ami tei gha na ngao-i ba k-e mono
Ak1 what kl-!8o stay kl-1lX.ll Nlo Ak1 want-O:!ll lL1 lkk-!8o stay
'what stayswhat we don't wantwill stay'
b. ami-ko tei gha na ta(tavu thai-a na kei
1lX.ll-uL Nlo Ak1 kul-lind 'arrive.at'-O:!8o Ak1 basket
'we couldn't lind the basket'
words can be interposed between teigha and na, as ti is in (26).
121
(26) l tei gha t-i na mono varongo1
!8o Nlo kl-lOc Ak1 stay quiet
'le simply cannot be still!'
A polar question can be asked:
(268) 4a k-o dea o tei gha5
lL1 lkk-28o go or Nlo
'Are you going or not7'
and answered:
(269) ? bo dea i leghai5 Tei gha.
28o lmll go lOc garden Nlo
'Are you going to the garden7' 'No.'
with teigha.
likewise, a content question can be answered with teigha. ln (20) the negative
121
cl. (22) where mo 'Nlo' comes between, and (26) where laka ti 'also kl-lOc' is lound in this position.
1
verb constitutes a one word answer.
(20) 0 ngitha na igha t-o lavi5 Tei gha.
!8o how.many Ak1 lish kl-28o take Nlo
'low many lish did you catch7' 'None.'
lxample (21) is somewhat more complex. oiven that numbers are stative verbs
in lengo it is not surprising that they can also be negated. ln this example, the
respondent did not know the exact number ol rows ol the root crop pana that had been
planted. le lramed his estimate as a partial negative, as il to say, uon't hold me to this,
but it's somewhere around live or six ten-garden-rows."
(21) Q: 0 ngitha na ivolo ni pana5
!8o how.many Ak1 ten.garden.row olN pana
'low many ten.garden.rows ol pana [are there in your garden]7'
A: #ima tei gha m-e ono na ivolo.
live Nlo cON}-!8o six Ak1 ten.garden.row
'live, or il not, six ten.garden.rows [i.e., 0 or 60].'
Note that in this example the negated number is not preceded by an article.
lurther evidence lor the verbal nature ol teigha is its use in the response to a
polar question as in (22).
(22) * /. t-e mai5 Tei gha vata.
Ak1 l kl-!8o come Nlo continue
'las l. come7' 'Not yet.'
lere it is modilied by the adverbial vata 'stop, start, continue'.
122
4. 8. . 4 lrohi bi ti ve tabu
Another common negative is the word tabu 'prohibitive (with consequences),
lorbidden'. 1he word tabu is especially common to hear lrom parents correcting their
122
1he word vata might be best described as a valency changing J 'change (and sometimes maintenance)
ol state' word: il something is started vata stops" it, il something is stopped vata starts" it, and il
something is expected to start J stop vata continues" it.
18
children. larents will blurt tabu to mean anything lrom '8top!' to 'uon't!' to 'Quit it!' to
'leave it alone!' and so on. ln short, tabu can be a quick corrective (to those children
who heed it). 1he word is by no means, however, reserved lor parent J child
interactions. lt has signilicant religious and social uses. lurial sites and the place(s)
where the spirits ol the dead go are tabu, gardens, houses, and canoes (among other
things), can be made tabu by assigning specilic consequences (by means ol a curse) to
the item that will allect any unauthorised person who might access the item, speaking
the name ol a close relative is tabu (indeed, such persons are relerred to as tabuthat's
their 'name', lor all intents and purposes), ood, and anything associated with ood, is
tabu (though the meaning in such contexts is closer to 'sacred, holy' than 'lorbidden',
see lender and leller 200!).
1he consequence associated with tabu is sometimes explicitly stated, as in (2!).
(2!) Tabu na lavi-a na ghau: b-e ghado-gho.
Nlo Ak1 grab-O:!8o Ak1 knile Allk-!8o pierce-O:28o
'uon't grab the knile: no good it cuts you!'
Again, notice the article preceding the verb lollowing tabu: na lavi-a. At other times the
consequence is an unspoken 'or else . . . ', as in (24).
(24) Tabu le(leu1
Nlo kul-light
'uon't light!'
ln this example the verb lollowing tabu is not preceded by an article.
4. 8. . Non-voli ti ve kou
1he negator kou 'unwilling', an modal verb like teigha and tabu, also comes
belore the verb it negates. And, like teigha and tabu, the verb it negates is preceded by
an article.
19
(2) )ou na vanga na igha.
Nlo Ak1 eat Ak1 lish
'1he lish are unwilling to eat [the bait].'
while it is true that the lish are not eating, what is in view here is their unwillingness
to eat. 1he modal verb kou is a negative with a reason: unwillingness.
Note in (26) that the article na is separated lrom the negator kou by laka ti 'also
kl-lOc'. linding these adverbs here is consistent with the analysis ol kou as a modal verb.
(26) Ma na tha laka e kou laka t-i na lighu-ni-a ghini-a igha deni
cON} Ak1 kll also !8o Nlo also kl-lOc Ak1 pass-1k-O:!8o lN81-O:!8o lish ulm
m-e ghe laka po t-i tena maone.
cON}-!8o continue also llm kl-lOc lOc sand
'And what's more, he [the lish] was unwilling to be passed by him [the turtle] so
this lish also just continued onto the sand.'
1he lish's actions would ultimately lead to its untimely demise, but its unwillingness to
be beaten by the turtle was more important than continuing to live. clearly there is a
strength to the unwillingness ol kou that is greater than mere ambivalence.
160
clAL8l 81kLc1Lkl
ln lengo there are clauses both with and without verbs. 1he three types ol
clauses that do not require a verb are equatives (where, in lact, verbs are prohibited),
possessives, and interrogatives. A stative or dynamic verb is lound in the remainder ol
lengo clauses. Verb clauses are ol the types intransitive and transitive. lach ol these
clause types may be lound in the imperative or the indicative mood.
lengo's basic word order is verb initial. ldentilying a basic word order with the
three traditional constituentsverb and lexical subject and object noun phraseshas
proven elusive.
12!
Noun phrase subjects and objects are conspicuous by their absence in
discourse once initial relerence has been established, and a subject or object noun
phrase candepending on the pragmatic situationbe placed belore or alter the verb.
crowley (2002b:!6) notes ol Oceanic languages in general that, clauses in which both
subject and object are realized as noun phrases are relatively inlrequent in discourse."
1his is certainly the case in lengo, where very olten the only subject-like constituent in
a clause is the verb phrase's subject relerence pronoun. lowever, conllating phrase and
clause level constituents, that is mistaking this subject indexing constituent with the
subject ol the clause, only leads to conlusion. when a clause contains a subject noun
12!
As uryer (200b:80) points out, 1he normal understanding ol what we mean when we talk about the
basic order ol subject, object, and verb, or ol just subject and verb, or ol just object and verb, is that ol
the order when the subject or object is a noun, rather than a pronoun, or more accurately, a lexical
noun phrase, i.e. a noun phrase headed by a noun, rather than a noun phrase consisting ol just a
pronoun. ln some languages, like lnglish, pronouns exhibit a distribution that dillers very little lrom
that ol lexical noun phrases, so that it makes little dillerence whether one includes pronouns or not
in discussing the position ol subjects and objects. lut in many other languages, pronouns exhibit
word order properties that diller considerably lrom lexical noun phrases, either because the syntactic
rules ol the language treat them dillerently, or because the pragmatic rules are such that their
distribution is rather dillerent . . . lt should be emphasized that although the position ol lexical
subjects and objects is crucial in determining basic order at the clause level, the position ol
pronominal subjects and objects, il dillerent lrom that ol lexical subjects and objects, is just as
important in giving a complete description ol word order in the language."
161
phrase, an emphatic pronoun as subject, and a subject relerence pronoun, deciding
which is the subject ol the clause is somewhat clearer when the subject relerence
pronoun is analysed as an indexer at the level ol the verb phrase (see page 86). oiven
the relative rarity ol subject and object noun phrases, their positional lreedom within
the clause, and the analysis ol the subject relerence pronoun as a verb phrase
constituent, the most that can be said about basic word order is that lengo is verb
initial.
ln terms ol Nichols's (1986) head J dependent marking criteria, lengo is best
described as a split-marking language, that is, that a roughly equal number ol heads
and dependents are marked in the language. uirectly possessed nouns are marked lor
possessor person and number and verbs are marked lor object. On the other hand,
indirect possession is not marked on the head noun but on a dependent classilier and
1Am is marked on an indexer within the verb phrase and not on the verb itsell.
. 1 Verbl ess cl auses
Verbless clauses are ol three types: equative, possessive, and interrogative. 1hey
are all characterised by the juxtaposition ol a pronoun with a noun phrase. 1he lirst
two are discussed below, while interrogative clauses are discussed in depth in section 6.
. 1. 1 lquati ve
As is true typologically ol Oceanic languages, an equative verbless clause in
lengo is expressed by simple juxtaposition ol noun phrases with no intervening verb"
(lynch et al. 2002:49). An emphatic pronoun is used in (2):
162
(2) !nau a /.
ll:18o Ak1 l
'l [am] l.'
and a subject relerence pronoun in (28):
(28) 0 na ghaoka.
!8o Ak1 woman
'8he [is] a woman.'
lxamples (2) and (28) represent predicate nominal constructions with no copular
verb between the two Nls.
. 1. 2 lossessi ve
A verbless possessive clause involving a directly possessed noun is shown in the
second clause ol (29). 1he possessed noun is juxtaposed to the possessor with no
intervening verb.
(29) *. te loghoa a sakai na tubue na ethe a #.
A t-e logho-a a sakai na tubu-e na atha-a a l
A kl-!8o have-O:!8o Ak1 one Ak1 relative-l8:!8o Ak1 name-l8:!8o Ak1 l
'A. had a relative [with him], his name [was] l.'
ly way ol contrast, the lirst clause ol (29) has a verb ol possession (logho 'have')
between the possessor and the possessed noun.
. 1. ! lnterrogati ve
A verbless interrogative clause involves the juxtaposition ol a question phrase to
a noun phrase.
(280) a. * thi na atha-mu ighoe5
Ak1 lN1 Ak1 name-l8:28o ll:28o
'what is your name7'
b. 9a vula tha5
Ak1 moon lN1
'which month [is it]7'
16!
1here is no copular element between the question phrase and noun phrase.
. 2 Verb cl auses: core arguments
Verbal clauses are ol two types: intransitive and transitive. ln the vast majority
ol lengo sentences the verb has a subject indexer (in the lorm ol a subject relerence
pronoun) preceding the verb and, in the case ol transitive clauses, an object indexer (in
the lorm ol an object pronoun sullix) joined to the verb.
124
1he near obligatory nature
ol the subject relerence pronoun has much to do with the lact that 1Am is marked on it.
A subject noun phrase, at times in the lorm ol an emphatic pronoun and at other times
as a lexical noun phrase, is optional. ln transitive clauses an object noun phrase is also
optional. loth subject and object noun phrases, when included, are normally lound
lollowing the verb phrase. An optional prepositional phrase can come belore or alter
the optional noun phrase subject andJor noun phrase object, as is shown in (281).
(281) Vl (ll)
{
Nl
8

Nl
O
}
(ll)
lither subject or object noun phrase maylor pragmatic reasonsprecede the verb
phrase, though never both at the same time. An optional prepositional phrase may also
precede the verb phrase.
1he valence ol a lengo verb is increased lor causatives (see page 84), but
prepositional verbs as part ol serial verb constructions are used to introduce
comitative, dative, and instrumental arguments. 1wo valence decreasing constructions
used in lengo are rellexive and reciprocal (see page 4! and 8 respectively).
124
l compare the situation in lengo to uixon's (1988:242) concise statement ol the core arguments ol
loumaa lijian: A predicate (l) must involve pronominal relerence to the subject and, il it is
transitive, to the object, these are the core constituents".
164
. 2. 1 lntransi ti ve
An intransitive clause is, at a minimum, comprised ol a verb phrase, that is a
subject relerence pronoun and a verb.
(282) u biku
18o sleep
Vl
'l sleep'
8ince there is no subject noun phrase it is impossible to comment on basic constituent
order at this point.
1here are also instances ol both a Vlwith its subject relerence pronounand
an emphatic pronoun in indicative intransitive clauses. As is seen in (28!), the subject
relerence pronoun precedes the verb while the emphatic pronoun lollows it.
(28!) a. u biku [i nau]
12
18o sleep ll:18o
Vl 8
'l sleep'
b. m-u ghe tapa [i nau]
cON}-18o continue run ll:18o
Vl 8
'and l continued to run'
lere it can be said that basic constituent order is V8.
A subject can be more lully expressed with a Vl subject relerence pronoun, an
emphatic pronoun, and a subject noun phrase as in (284).
(284) k-ami thairo [i ghami na ghaoka ma na mane]
lkk-1lX.ll work ll:1lX.ll Ak1 woman and Ak1 man
Vl 8
'we will work we women and men'
oiven the presence ol a subject noun phrase lollowing the Vl, the basic constituent
12
[8quare brackets] are used in the source text line in this and the lollowing section to demarcate
subject and object noun phrases.
16
order is still V8.
A Nl subject lollowing the Vl can be separated lrom it by a prepositional
phrase, as in (28).
(28) e dea i rughu ni vatu [na igha]
!8o go lOc beneath olN stone Ak1 lish
Vl ll 8
'the lish went beneath a stone'
1here are very lew instances ol a subject noun phrase preceding the Vl in an
intransitive clause. 8ome examples lollow.
(286) a. [! tu-ghami a 4. ma 6. ma /. ma #. ] t-ami-tu uru.
lAL-ll:1lX.ll Ak1 l cON} k cON} l cON} l kl-1lX.ll-lAL dive
8 Vl
'we lewl., k., l., l. [and l]we lew went diving.'
b. [a *. ] e thaghataleo vata po pukua t-e teigha na
Ak1 A !8o bad.inside continue llm because kl-!8o Nlo Ak1
8 Vl
logho-a siki dae-na
have-O:!8o any child-l8:!8o
'A. just continued to worry because he didn't have any children'
Note that in (286)a. the emphatic pronoun 'stays with' the subject noun phrase when it
is moved belore the Vl. lmphatic pronouns and subject noun phrases are in
appositional relationship in lengo, that is, when both emphatic pronoun and subject
noun phrase are used they either both precede or both lollow the Vl, they are never
separated by a Vl.
. 2. 2 1ransi ti ve
1he inllection ol transitive verbs has already been discussed in the section
dealing with the verb and verb phrase in section 4. lere the concern is with their
relation to the rest ol the clause, especially how the multiple arguments ol a transitive
166
verb are, or are not as the case may be, realised in the clause.
within the verb phrase there is a lixed, invariable order ol argument indexing
constituents (see page 86). 1here is, however, no lixed, invariable order ol constituents
at the clause level. while verb-initial orders (that is, V8 and VO) are most basic, it is
impossible to state that either V8O or VO8 is basic. loth orders are equally common in
the data. Non-basic orders include either subject or object noun phrases belore the Vl,
but never both at the same time.
A simple indicative transitive clause is given in (28).
(28) m-u kimu-a
cON}-18o kick-O:!8o
Vl
'and l kicked it'
lere the clause consists ol a verb phrase in which the agent is relerenced by a subject
relerence pronoun (having been established by a noun phrase previously in the text)
while the patient is marked by the third person object sullix -a. 1here is no lexical noun
phrase subject or object in (28).
lxample (288) has a noun phrase subject, na aliva 'the centipede', alter the Vl.
(288) 0 ghani-u [na al i va]1
!8o bite-O:18o Ak1 centipede
Vl 8
'A centipede bit me!'
lere evidence begins to unlold lor which constituent is what. ln (288) the !8o subject
relerence pronoun e can be identilied as the biting centipede and -u as the 18o one who
was bit.
126
ln (289) a lexical object noun phrase, na thara 'Ak1 least', lollows the verb.
126
ln lengo society a person biting a centipede would be considered oll norm. ln lact, nothing that can
do harm to a person is eaten lor lood (e.g., shark, snake, crocodile, centipede, etc.).
16
(289) arere deni ba k-a ghali-a po [na thara].
8aturday ulm lL1 lkk-1lN.ll make-O:!8o llm Ak1 least
Vl O
'1his 8aturday we will make a least.'
lere again subject and object can be sorted out based on person and number. 1he
object noun phrase na thara 'the least' is !8o and is indexed by the !8o object pronoun
sullix -a. 1here is no noun phrase corresponding to the 1lN.ll subject relerence pronoun
in person or number, leaving a noun phrase object, but no noun phrase subject. 8o to
summarise, thus lar there is evidence ol a single noun phrase in a sentence: either a
noun phrase subject or a noun phrase object lollowing the Vl.
loth subject and object noun phrases may be lound lollowing the Vl in an
indicative transitive clause, as in (290).
(290) t-ara iti-a [na thinaghe] [iko-ira]
kl-!ll up-O:!8o Ak1 canoe uL-ll:!ll
Vl O 8
'they two beached (lit., upped it) the canoe'
lere the third person dual emphatic pronoun ikoira is relerenced by the third person
subject relerence pronoun tara in the Vl.
12
1he object noun phrase, na thinaghe 'the
canoe', is marked in the Vl with the !8o object pronoun sullix -a. lt is more dillicult to
tell that the object is singular, as nouns are not inllected to show number. ln this case,
though, na thinaghe 'the canoe' is assumed to be singular as there is no indication
elsewhere in the clause lor number. 8o in (290) word order remains 8VO. 1he same
constituent order is evident in (291).
12
1his is curious, as normally the dual subject relerence pronoun tarako (rather than tara) would have
been expected to stand in agreement with ikoira. while both tara and ikoira are non-singular, this
provides some evidence that while the dual lorm is always and only two the more general plural lorm
indicates two and more (see page 28 lor more discussion).
168
(291) M-e ghani mate-a [na koivo] [na kau].
!8 eat die-!8 Ak1 snail Ak1 dog
Vl O 8
'1he dog ate the snail and killed it.' (8ymons s4.2.4, (18))
Ol course, it would be a strange situation indeed lor a snail to bite a dog. 8o while
semantically (291) is quite clear, syntactically it is part ol the larger puzzle. 1he larger
puzzle is apparent in (292) where the order ol the noun phrases lollowing the Vl is
reversed, with the noun phrase subject preceding the noun phrase object.
(292) a. *i1 0 ghani-a [na vua] [a tina-mu]1
hey !8o bite-O:!8o Ak1 croc Ak1 mother-l8:28o
Vl 8 O
'ley! 1he crocodile is biting your mother!'
b. 0 ghani-a [na kau] [na be].
!8o bite-O:!8o Ak1 dog Ak1 pig
Vl 8 O
'1he dog bit the pig.'
ln these examples the object is no longer adjacent to the object indexing pronoun on
the Vl. constituent order alone is insullicient to sort out the arguments ol a transitive
verb. Again, based on semantics, it is highly improbable that a woman would bite a
crocodile, but given the possibility ol variable noun phrase subject - object ordering
lollowing the Vl, the syntax ol constituent order does not allow the hearer to entirely
rule it out. lxample (292)b. has the potential to clarily the situation considerably as
semantically, both the dog and the pig are capable ol biting the other. Lnlortunately,
the data does not include a lengo phrase equivalent to 'the pig bit the dog'.
1he examples in (291) and (292) indicate that, in lengo, the location ol noun
phrase subject and object is 'lree'. 1his raises the issue ol how to determine which noun
phrase is subject and which is object when there are two noun phrases ol the same
169
person and number lollowing the Vl.
128
consider (29!) in which both 8 and O are !8o.
(29!) 0 ghani-a [na kau] [a M.]
!8o bite-O:!8o Ak1 dog Ak1 m
Vl 8 O
'1he dog bit m.'
ln this example there is little hope ol sorting out which noun phrase is lilling what
syntactic role based on person or number. Nor can participants be sorted out based on
their position in the clause. lased on the evidence ol (290) and (291) it might be
expected that the noun phrase na kau 'Ak1 dog' is object, given that it is adjacent to
object indexing sullix on the Vl. lxample (292), however, presents the opposite case. ln
lact, in (29!) na kau 'Ak1 dog', the noun phrase immediately lollowing the verb, is 8.
129
1his example was presented to me as something ol a puzzle, something lor the
'language learner' to ligure out. who bit whom7" l was asked. l had a good guess, but l
gamely played along. changing the order ol the two sentence linal noun phrases, as in
(294), brought some scattered laughter.
(294) 70 ghani-a [a M.] [na kau].
!8o bite-O:!8o Ak1 m Ak1 dog
'm. bit the dog.'
m. wouldn't bite a dog," people said. while this would seem to imply a basic word
order ol V8O, a lew minutes ol restatement by moving constituents around (e.g., na kau
e ghania a M. and a M. e ghania na kau) revealed that they had no way ol knowing 'who bit
whom' apart lrom a larger context.
1!0
oiven the admissibility ol (290)-(29!) and the
128
compare uixon (1988:8) who says ol lijian that, A predicate will generally come lirst in any clause,
lollowed by noun phrases. where both subject and object Nls occur, subject-object and object-subject
orders are equally common in texts, although object-subject is generally prelerred in elicitation . . .
Any Nl (sometimes even two Nls) can be topicalised and moved to the lront ol a clause, but there
should normally be some 'marker' ol the Nl(s) in the predicate."
129
Ol course, this would be most likely semantically.
1!0
ll this seems surprising, see (!01) below where the object is lound belore the V and the subject alter.
10
rejection ol (294) as laughable by lengo speakers, there appears to be something
operating at a semantic level that determines the order ol subject and object noun
phrases lollowing a transitive verb. 1hat is, il na kau 'Ak1 dog' is permissible as subject
lollowing the Vl in the syntactic lrame presented in (29!) and a M. 'Ak1 m.' is not
allowed in the same position in (294) (or is allowed but results in a laughable meaning),
there must be something other than syntax that is at work to allow the one syntactic
ordering and disallow the other. A possible explanation involving a hierarchy ol
animacy is explored below.
kecall that lengo dillerentiates between animate and inanimate objects in !ll
(see page !8). lengo also dillerentiates between human and non-human animate lorms
with interrogative J relative pronouns (page 44) and articles (page 1). 8ince
dillerentiation according to animacy is at work elsewhere in lengo, it is not entirely
novel to suggest that a hierarchy ol animacy may also be at work here.
1!1
l propose the
lollowing as a hierarchy ol animacy lor lengo:
human animate non-human inanimate
with a more linely-grained hierarchy ol animate non-humans than is shown. my
observation ol lengo society is that pigsas an indicator ol wealthrank near the top
ol animate non-humans. uogs, on the other hand, seem to be quite low among the
1!1
uonohue (1999) pursues the idea ol animacy as a means lor sorting out the order ol arguments in
Asmat, a language ol lapua New ouinea. le states, Appealing to the animacy ol the arguments in
Asmat as a means ol modeling the orders ol the sullixes is not without its merits. consider these two
examples: (1) /or-n-Km. 'see-18o.O-28o.A, You saw me', and (2) /or-L-n. 'see-18o.A-28o.O, l saw you.' uespite
the sullixes showing the syntactic roles ol the arguments that they index, we can see that there is a
requirement operating that makes the actual order ol the two allixes independent ol the syntactic
roles, and depends on the animacy ol the arguments, with the 18o sullix in both cases being placed
closer to the verb, despite its marking subject in one case, and object in the other" (414). Lltimately, a
hierarchy ol animacy does not account lor all instances in Asmat, but he notes an example ol one
language where it does: 1he prelixal system ol verbal agreement in Yimas (lapua Newouinea, loley
1991, similar lacts hold in Abinomn ol west lapua) orders two prelixes according to animacy" (41).
11
ranks ol animate non-humans, they are, at best, neglected (the exception would be a
dog kept lor hunting andJor guard purposes). lt is not clear whether crocodiles and
sharks would be higher or lower than pigs, but given that there are crocodile and shark
priests who interact with them on the people's behall, they must surely rank quite
high. lt is also not clear where reptiles and lish would reside on the scale. lnvertebrates,
such as snails and beche-de-mer, are at the bottom ol the animate non-human category.
with that suggested animacy scale in mind, consider examples (290), (291), (292)a., and
(29!), in which both subject and object noun phrases are lound lollowing the Vl. ln
each ol these examples, the member highest on the animacy scale is lound at the end ol
the clause, regardless ol whether the noun phrase in question has the role ol agent or
undergoer in the clause. ln (29), an overview ol the preceding examples, the most
animate noun phrase, the one larthest lrom the Vl, is agent.
(29) lxample Nl
1
(undergoer) Nl
2
(agent)
(290) Vl na thinaghe 'the canoe' ikoira 'they two [men]'
(291) Vl na koivo 'the snail' na kau 'the dog'
lowever, in (296) the most animate noun phrase, the one larthest lrom the Vl, is
undergoer.
(296) lxample Nl
1
(agent) Nl
2
(undergoer)
(292)a. Vl na vua 'the crocodile' a tinamu 'your mother'
(29!) Vl na kau 'the dog' a M. 'm.'
8o in (292)a., lor example, a tinamu 'your mother' is lound at the end ol the clause
lollowing na vua 'the crocodile' because human mothers are higher on the animacy
scale than crocodiles. 1his hierarchy ol animacy also helps to explain (294) which, by all
appearances, is syntactically correct but lor some reason is not acceptable semantically.
Ol course, this explanation requires an understanding ol lengo culture when it might
12
seem to a non-lengo person that both noun phrases are at the same place on the
animacy hierarchy, as in (291) and (292)b. ln cases in which both noun phrase subject
and object lollow the verb and their level ol animacy is demonstrably dillerent, this is a
plausible explanation lor an apparent llexibility in post-verb noun phrase order.
laving said that, it is not common to lind both subject and object noun phrases
(lexical or pronominal) lollowing the Vl.
1!2
lt is, l think, signilicant that there are no
instances in my data ol both agent and undergoer noun phrases at the same level on
the hierarchy ol animacy lollowing the Vl: the use ol the animacy scale would lail in
such instances. kather, subject relerence pronouns and object pronoun sullixes olten
carry the subject J object indexing load which leaves room lor a single subject or object
noun phrase lollowing the Vl or, alternately, either subject or object noun phrase
belore the Vl and the other lollowing.
1he discussion now turns to non-basic word orders. 1he emphatic pronoun is
lound at the beginning ol a clause when the subject is in locus, as in (29).
(29) i nau ba k-u tugu vani-ghamu [na tha ba k-e sara]
ll:18o lL1 lkk-18o tell uA1-O:2ll Ak1 cOml lL1 lkk-!8o arrive
8 Vl cOml
O
'l will tell you what will happen'
lynch et al. (2002:49) conlirm that, the presence ol an independent pronoun as subject
or object marks contrast or locus," and, most Oceanic languages have lairly lixed basic
constituent orders, but generally allow the movement ol constituents to clause-initial
position in order to express topicalisation" (0). 1he object complement (introduced by
na tha 'Ak1 cOml') is not marked on the verb with an object pronoun sullix.
1!2
lynch et al. (2002:49) maintain that, in Oceanic languages, clauses in which both subject and object
are realised as noun phrases are rare in discourse. clauses olten consist only ol a verb phrase, with its
clitics or allixes corelerencing subject and object. "
1!
loth a subject noun phrase and an emphatic pronoun may appear at the
beginning ol a transitive clause, as in (298).
(298) [*. ma #. i ko-i ra ru(ruka] ara-ko logho-ra ara vonu na mamanu
A and l uL-ll:!ll kul-two !ll-uL have-O:!ll !ll many Ak1 animal
8 Vl cOml
O
'A. and l. they two, [the] two [ol them], they had many animals'
1he order ol subject noun phrase and emphatic pronoun can be reversed, as in (299).
(299) [i ghoe a ,. ] k-o tura-a [nimoa na matau]
ll:28o Ak1 l lkk-28o carry-O:!8o l8:28o Ak1 axe
8 Vl O
'you, l., you carry your axe'
ln both (298) and (299) the subject is placed belore the verb complex to put it in a place
ol prominence. ln (!00) the emphatic pronoun is lurther emphasised with a
demonstrative.
(!00) ! ghami nga ba k-ami ne athe na tha t-ami tangomana vani-ghamu.
ll:1lX.ll ulm:8o lL1 lkk-1lX.ll lmll give Ak1 kll kl-1lX.ll able uA1-O:2ll
'we here will give what we are able to you.'
An object can also be placed at the beginning ol a transitive clause, as in (!01)
and (!02). Again, the ellect is one ol pragmatic locus: pointing out the importance ol
the item lound lirst in the clause.
(!01) [#i ma na thara] ba k-a ghali-a ighita tena arere deni.
live Ak1 least.row lL1 lkk-1lN.ll make-O:!8o ll:1lN.ll lOc 8aturday ulm:8o
O Vl 8
'live least-rows we will make this 8aturday.'
ln (!01) the live least rows are the object ol ghali-a 'make-O:!8o'. 1he speaker, intent on
making clear the lact that there is to be one least row per village groupthat no one
would be lelt outplaces the crucial object at the lront ol the sentence.
lxample (!02) is the 'title' ol a recorded texta short preamble that the speaker
14
gave to provide the hearer with an overview ol the story.
(!02) 9a tugu-a [na ghaoka] t-e ghani-a [na vua].
Ak1 story-O:!8o Ak1 woman kl-!8o bite-O:!8o Ak1 crocodile
O Vl 8
'1he story ol the woman the crocodile bit.'
lere again, the participant ol prominence is placed at the beginning ol the clause. 1hat
na ghaoka 'the woman' is the object makes the lact that she is lound at the beginning ol
the clause that much more striking. Note that in (!02) there is no syntactic means to
dillerentiate subject lrom object. 1hey cannot be sorted by means ol subject or object
indexing constituents since they are both third person singular. 1he animacy scale also
does not help since one noun phrase precedes and the other lollows the verb phrase.
1he listener is lelt to rely on her understanding ol the way the world is (that is, that
women don't bite crocodiles) and either have it conlirmed or challenged during the
course ol the story. ln this case it is conlirmed: the crocodile bit the woman.
1here is a type ol transitive clause that includes two objects but is not properly a
ditransitive clause. 1he data do not contain any convincing examples ol prototypical
ditransitive constructions, that is a verb with three noun phrase arguments. lndeed this
seems to be somewhat ol a rarity in Oceanic languages as koss comments that,
lOc verbs had a valency ol either one or two, that is, they were either
intransitive or transitive. 1here were probably no trivalentJditransitive
verbs, i.e., verbs whose role structures required or allowed three noun
phrases without case marking, but we cannot be certain about this, as
some modern languages do have trivalent verbs (manam, loava)
(1988:20),
and again (2004:10) that, lijian, like other canonic languages, has no ditransitive
verbs."
1!!
while lengo argument indexing does allow lor three nominal clause
1!!
l'm not sure, based on koss' description (2004:492), whether lengo is a 'canonic language'. lengo is
within the 8outheast 8olomonic lamily ol the Oceanic subgroup (.a ol 1able 2 (koss 2004:494)).
1
constituents with a combination ol noun phrases and pronouns, there are never three
noun phrases as arguments.
ln (!0!) there are three arguments present: a subject as agent (A, relerenced by
u '18o'), an object noun phrase as theme (1), and an indirect object -gho 'O:28o' indexing
the recipient (k).
(!0!) a. u athe-gho na buka
18o give-O:28o Ak1 book
A V-k 1
'l gave you the book'
b. * thi t-e athe-gho na igha5
Ak1 lN1 kl-!8o give-O:28o Ak1 lish
A V-k 1
'who gave you the lish7'
lnterestingly, the verb is allixed with the object pronoun -gho 'O:28o' (here indexing the
assumed recipient argument 'you'), while there is no trace ol the object pronoun -a
'O:!8o' to relerence the theme argumentsna buka 'the book' and na igha 'the lish'. oiven
that the verb is able to index a single argument, it is recipient that is relerenced in
these instances and not theme.
1!4
1wo arguments can be indexed in a dative serial verb construction as there are
two verbs, each able to index an argument. consider example (!04).
(!04) u athe-a vani -gho na buka de
18o give-O:!8o uA1-O:28o Ak1 book ulm:8o
A V-1 V-k Nl
1
'l am giving you this book'
ln this example the recipient is indexed on the dative verbal preposition vani, while the
theme is indexed on the dynamic verb athe 'give'. A serial verb construction such as this
1!4
One area ol lurther research at this point is whether the animacy hierarchy l proposed earlier applies
here, that is, whether the item highest on the scale is marked with an object pronoun.
16
is the only way to index both recipient and theme arguments.
. ! Verb cl auses: peripheral arguments
locatives are expressed with prepositional phrases in lengo. 1he locative
markers i 'lOc' andJor tena 'lOc' come at the beginning ol the location phrase. when both
locatives are present i always precedes tena. 1he locative prepositions i and tena mark
both temporal and spatial locations. 1hey each cover a comprehensive range ol
prepositional meanings: to, at, in, during, onto, lrom, and so on (compare lijin long).
ln (!0) there are lour instances ol the two lengo locatives.
(!0) [tena ade rodo t-i] [i M.] iko-ghami a J. t-ami-ko biku8
lOc 8unday night kl-lOc lOc w uL-ll:1lX.ll Ak1 } kl-1lX.ll-uL sleep
lOc
time
lOc
space
biku [tena vae ni vanga] [i M.]
sleep lOc house olN lood lOc w
lOc
space
lOc
space
'On 8unday night, at w., we two}. [and l]we slept, slept in the kitchen at w.'
1here are two prepositional locative phrases in each clause ol (!0): one temporal and
one spatial in the lirst clause, and two spatial in the second. loth locations are lound
belore the verb in the lirst clause while in the second they are lound alter the verb.
1here is no subject relerence pronoun in the second clause, though it is understood
who the subject is in the second clause by the verb's tail - head linkage to that ol the
lirst clause. 1he second clause is a recapitulation ol the lirst, with the addition ol a
more specilic location (tena vae ni vanga 'in the kitchen').
1he locatives i and tena operate independently ol one other and also in
conjunction with one another to mark temporal and spatial locations. 1emporal
locatives can be introduced with any ol the lollowing constructions:
1
(!06) a. i Nl
time
b. i tena Nl
time
c. tena Nl
time
1he order *tena i Nl
time
is not grammatical.
ln the lirst prepositional phrase ol (!0) i alone is lollowed by a noun phrase ol
time, while in the second i tena is lollowed by a noun phrase ol time.
(!0) ! ropo ba i tena thangavulu kiloko u ngao-a ighamu
lOc tomorrow lL1 lOc lOc ten o'clock 18o want-O:!8o ll:2ll
'1omorrow at ten o'clock l want you to . . . '
lt is examples like this that have caused me to struggle to lind any distinction between
the two locatives. lndeed, a quick glance at (!11) below, the continuation ol example
(!0) in the source text, reveals both locatives indicating a spatial location.
ln (!08) tena 'lOc' is lound alone belore arere '8aturday', a noun phrase ol time.
(!08) #ima na thara ba k-a ghali-a ighita tena arere
live Ak1 least-row lL1 lkk-1lN.ll make-O:!8o ll:1lN.ll lOc 8aturday
'live least-rows we'll make on 8aturday.'
A temporal locative phrase can be lound preceding (as in (!0)) or lollowing (as in
(!08)) the verb.
1he temporal locative tena lollowed by a demonstrative has the meaning 'now'
(lit., 'lOc ulm', i.e., this point in time).
1!
ln (!09) tena ulm 'now' helps to describe the time
ol the verb vasangi viti.
(!09) pisa na aro-gu m-u ne vasangi viti koto tena de
dislocate Ak1 shoulder-l8:18o cON}-18o lmll leel pain very lOc ulm
'my shoulder dislocated and now l leel a lot ol pain'
lt is a very real now. 1he conjunctive counterpart ol tena, that is mena, does not appear
in post-verb position, its use is reserved lor clause conjoining purposes.
1!6
1!
kecall that the demonstrative deni can be shortened to de. 8ee page 66.
1!6
1he discussion ol the conjunctive locative mena is lound on page 19!.
18
when marking spatial locations, i and tena are again lound together and on their
own. 8patial locatives can be introduced with any ol the lollowing constructions.
(!10) a. i Nl
space
b. i tena Nl
space
c. tena Nl
space
Again, the order *tena i Nl
space
is not grammatical.
ln (!11) (which is the continuation ol (!0)) i and tena are lound together
marking the spatial locative.
(!11) ba k-amu mai i tena vae ni kolivuti
lL1 lkk-2ll come lOc lOc house olN pray
'you will come to church'
ln (!12) i alone marks the location.
(!12) gharasu mai ighamu i vae ni vanga
move come ll:2ll lOc house olN eat
'move this way, you in the lood-house'
ln (!1!) tena alone carries the locative-marking load.
(!1!) m-e tagu iti tena maone tena kokomu ia
cON}-!8o crawl up lOc sand lOc island ulm
'and it crawled up on the sand at that island'
ll there is any way to account lor two locative markers working sometimes together
and at other times alone in the same environments, it is not evident to me.
1he only apparent distinction between i and tena is in relation to a lairly large
set prepositions. with these the locative marker i 'lOc' is obligatorily present while tena
'lOc' is obligatorily absent. 1hese prepositions ol both time and space include: i leo
'inside', i muri 'behind', i nagho 'in lront', i rughu 'underneath', i teghea 'atop', i levugha
'between', i ropo 'morning', i nea 'yesterday', i nganiba 'later', i boko 'downcoast', i
19
ghalagha 'upcoast', i longa 'landward', and i sapa 'seaward',
1!
among others. 1hese
locative constructions are used to indicate spatial or temporal location in relation to a
noun.
(!14) a. 4a ku kabia na gilu i pal a sepe na vaegu.
ba k-u kabi-a na gilu i pala sapa-a na vae-gu
lL1 lkk-18o dig-O:!8o Ak1 hole lOc side seaward-l8:!8o Ak1 house-l8:18o
'l am going to dig a hole to my house's seaward side.'
b. e ghua logho bobona vata i ngani ba na nulavi m-ami-tu
!8o il have time continue lOc later Ak1 alternoon cON}-1lX.ll-lAL
dea lebo-i laka t-i
go lloat-O:!ll also kl-lOc
'il there is still time later in the alternoon, we will also lloat them [down the
river]'
1he locations are sometimes in a genitival relationshipas marked by ni 'olN'Nto the
nouns they describe.
(!1) a. e dea i rughu ni vatu na igha
!8o go lOc underneath olN stone Ak1 lish
'the lish went underneath ol the stone'
b. talu-i i l eo ni popo
put-O:!ll lOc inside olN bowl
'put them inside ol a bowl'
8o in (!1)a., lor example, rughu 'underneath' is possessed by vatu 'stone'.
1!
1hough there are certainly times that boko 'downcoast', ghalagha 'upcoast', longa 'landward' and sapa
'seaward' correspond quite closely to west, east, south and north, one could become very disoriented
indeed il one merely substituted compass" directions lor the more lluid orientations the lengo
cardinal directions indicate. As lor distinguishing 'upcoast' lrom 'downcoast', it has been explained to
me that marau 8ound (what l would call the eastern tip ol ouadalcanal) is the 'top' ol the island. lt is
widely acknowledged as the place where the spirits ol the dead reside.
180
6 lmllkA1lVl ANu lN1lkkOoA1lVl 8lN1lNcl8
6. 1 lmperati ve and cohortati ve sentences
1he imperative can, at a minimum, be indicated by a bare verb as in (!16).
(!16) a. Mai 1
come
'[You,] come here!'
b. 4i ku!
sleep
'[You, go to] sleep!'
ln these examples the irrealis marked subject relerence pronoun is absent.
more commonlyand more completelya verb in the imperative is preceded by
a subject relerence pronoun with irrealis marking (k-8kl), as in (!1).
(!1) )-o mai1
lkk-28o come
'You, come [here]!'
1he use ol the irrealis marked subject relerence pronoun (k-) lor the imperative mood
indicates that there is no commitment to the reality ol the event in the imperative.
1here is a signilicantly high desire J expectation lor the event to become a reality, but
at the time ol utterance the event is not yet a reality.
A subject may be more lully expressed in an intransitive imperative with an
emphatic pronoun lollowing the verb (here with the directional mai 'come'), as in (!18).
(!18) k-o gharasu mai i ghoe
lkk-28o move come ll:28o
'you move this way, you'
1his can be expanded to include a lexical noun phrase lollowing the verb, as in (!19).
(!19) k-ami thairo i ghami [na ghaoka ma na mane]
1!8
lkk-1lX.ll work ll:1lX.ll Ak1 woman cON} Ak1 man
'we must work, we women and men'
181
ln (!19) three constituents describe the subject: the subject relerence pronoun ami
'1lX.ll' allixed lor 1Am, the emphatic pronoun ighami 'we' lollowing the verb, and the
conjoined noun phrases na ghaoka ma na mane 'the women and the men' lollowing the
emphatic pronoun. kecall that when an emphatic pronoun and a subject noun phrase
are both present, the subject noun phrase always lollows the emphatic pronoun.
1hese same constituentsemphatic pronoun and lexical noun phrase (and only
in that order)may also be lound belore the subject relerence pronoun - verb, as in
(!20).
(!20) ! ghamu [na l eoni ti noni ] k-amu rongo mai1
ll:2ll Ak1 all person lkk-2ll listen come
'lveryonelisten here!'
1he ellect ol this is to topicalise the subject.
what sets the imperative apart lrom the luture is the use (or not) ol ba 'lL1' and
boro 'impossible lL1'. 1he presence and placement ol the luture marking particle ba or
boro is important as a means to distinguish luture lrom imperative, as in (!21).
(!21) )-o bila mai8 ba k-u dala.
lkk-28o throw come lL1 lkk-18o catch
'You throw [it] here, l'll catch [it].'
At the time ol utterance, neither the throw nor the catch have taken placethey both
remain unrealised. lut they are unrealised in dillerent ways. 1he lirst clause, ko bila mai
'you throw [it] here', is imperative and awaits the hearer's compliance. 1he second, ba
ku dala 'l will catch [it]', is distinguished lrom the lirst with luture-marking ba, and
awaits an object to catch. while either ba or boro is obligatory with the luture, it is not
1!8
[8quare brackets] are used in the source text line in this section to demarcate subject and object noun
phrases.
182
with the imperative. 8yntactically, what sets the imperative apart lrom irrealis is
unclear. lurther studies in this area might concentrate on whether ba is in the process
ol being dropped lrom imperative sentences, that is, whether it was previously more
obligatorily present lollowing k-8kl than it is now.
Object arguments are added to imperative verbs as they are with indicative
lormswith an object pronoun sullix or noun phrase.
(!22) a. Tal u-a1
put.down-O:!8o
'[You,] put it down!'
b. Mai na vothe.
come Ak1 paddle
'[You,] hand [me] the paddle.'
Note that there is no object pronoun sullix on the verb in (!22)b. lt is dillicult to explain
this departure lrom the norm.
1here is evidence ol what appears to be a third person imperative in (!2!).
(!2!) )-o dea bosa vani-a . k-e mai se(seghi.
lkk-28o go tell uA1-O:!8o 8 lkk-!8o come kul-quickly
'You, go tell 8. he [must] come quickly.'
while the lirst clause is second person imperative, the second clauseke mai seseghi 'he
[must7] come quickly'is a command to a third person. 8ignilicantly, the second clause
is also in irrealis mood, so it seems to lullill all the criteria ol an imperative.
6. 2 lnterrogati ve sentences
lnterrogative sentences are ol three types in lengo: polar, leading, and content.
1wo typespolar and leading questionsare indicative sentences marked by
intonation, while the third typecontent questionis marked by the presence ol
interrogative pronouns.
18!
6. 2. 1 lol ar questi ons
lolar, or yes J no, questions are in the lorm ol indicative sentences. what sets
them apart is their intonation. ln (!24) intonation lalls progressively lrom teigha to the
second occurrence ol na, and then rises over the linal word kei-gu.
(!24) o teigha na bere na kei-gu5
28o Nlo Ak1 see Ak1 basket-l8:18o
'you haven't seen my basket7'
1he intonation contour is shown in lllustration 2: lolar question pitch track.
kesponses to negative polar questions are interesting. ll one answered eo 'yes' to (!24) it
would mean that 'yes, l have not seen your basket'. conversely, a negative answer teigha
'no' would mean, 'no, l have seen it'.
when the alternative o teigha 'or Nlo' is added at the end ol the sentence,
intonation rises over the main part ol the sentence and then lollows a low-high-low
contour over the linal segments o teigha. 8o in (!2) intonation rises lrom ba to dea, then
starts low at o, and then goes high to low over teigha.
(!2) 4a k-o dea o teigha5
lL1 lkk-28o go or Nlo
'Are you going or not7'
184
!llustration $: /olar Ouestion pitch track
6. 2. 2 leadi ng questi ons
leading questions
1!9
, or polar questions lor which an answer is assumed, are
possible in lengo. 8yntactically, a leading question in lengo has the lorm ol an
indicative sentence, as polar questions do above, but the rise in intonation is reserved
lor the additional 'leading question marking' particle na at the end ol the sentence.
consider (!26).
(!26) t-amu tangomana na ta(tagho i leo-a na ovu ia, na5
kl-2ll able Ak1 kul-lish lOc inside-O:!8o Ak1 group ulm lN1}
'you in that group are able to go lishing, right7'
1he pitch track ol (!26) is given in lllustration !: leading question pitch track.
lven though (!26) is posed as a question, it is assumed to be true that the group ol
people being addressed are in lact able to go lishing. 8o while a question is asked, the
answer is assumed by the speaker.
6. 2. ! content questi ons
content questions, with the exception ol tha 'which', are similar in lorm to each
other. 1here are ways to ask who, what, where, when, and how much. lxcept in the case
ol tha 'which', the interrogative pronoun is always lound at the beginning ol a content
1!9
uryer (200b:9!) delines a leading question as one in which, the speaker makes an assumption as to
what the answer will be."
18
!llustration ;: #eading Ouestion pitch track
question.
1he construction a thi, the human" article a and interrogative pronoun thi, is
used to ask 'who7'
(!2) a. * thi t-e athe-gho na igha5
Ak1 lN1 kl-!8o give-O:28o Ak1 lish
'who gave you the lish7'
b. * thi na atha-mu ighoe5
Ak1 lN1 Ak1 name-l8:28o ll:28o
'what is your name7'
1he construction na tha, the non-human" article na and interrogative pronoun
tha, is used to pose the question 'what7'
(!28) 9a tha t-o ghoni-a5
Ak1 lN1 kl-28o do-O:!8o
'what are you doing7'
lxample (!28) demonstrates the inanimate use ol na tha. 1he animate, non-human use
ol the interrogative tha is shown in (!29).
(!29) m-ara-ko thuge-u, B9a tha t-e ghani-gho5C
cON}-1lX.ll-uL ask-O:18o Ak1 lN1 kl-!8o bite-O:28o
'and they two asked me, what bit you7"'
Lnlortunately, a centipede had bitten the narrator.
1he construction na N tha is used to ask 'which - what kind7' types ol questions.
1he noun under question is interposed between the article na and the interrogative
pronoun tha.
(!!0) a. B9a beti tha5C BMerasine.C
Ak1 liquid lN1 medicine
'what kind ol liquid is that7' 'medicine.'
b. 9a vula tha7
Ak1 moon lN1
'which month [is it]7'
186
with the na tha construction there is an unlimited range ol possible answers whereas
with na N tha the range ol answers is limited to the range ol possibilities presented by
the noun.
1he question 'where7' is asked with i vi 'lOc where'.
(!!1) B! vi t-o bo dea5C B?leole po.C
lOc where kl-28o lmll go stroll llm
'where are you going7' '}ust walking around.'
ln the case ol a verbless question, vi can take the object pronoun -a as in (!!2).
(!!2) ! vi -a na vae-mu5
lOc where-O:!8o Ak1 house-l8:28o
'where is your house7'
1here is a certain level ol specilicity introduced with the use ol ga 'there'.
(!!!) ! vi ga t-o bo dea5
lOc where there kl-28o lmll go
'where are you going7'
1he person who asks i vi 'lOc where' with ga 'there' is looking lor a specilic location as
an answer. lxample (!!!) is asked ol a passerby, wondering where, specilically, he's oll
to.
1he construction e ngitha '!8o how.much' is used to ask questions concerning
quantity ('how many - much7').
(!!4) a. B0 ngi tha t-i niulu-a5C BThangavulu tolu t-i na vula.C
!8o how.much kl-lOc year-l8:!8o ten three kl-lOc Ak1 moon
'lt is how much, his years7' '1hirteen months.'
b. 0 ngi tha na matea5
!8o how.much Ak1 price
'lt is how much the price7'
'when' questions are also lormed with ngitha 'how.much' as part ol the question
lormula, however, in the place ol e '!8o' is lound i 'lOc'.
18
(!!) ! ngi tha t-o mai5
lOc how.much kl-28o come
'when did you come7'
1he locative, in its temporal sense, lends the dimension ol time to the question.
188
cOmlllX 8lN1lNcl8
clauses can be joined as either coordinate or subordinate. coordinate clauses
are ol the types simultaneous, sequential, and alternate. 8ubordinate clause types in
lengo are relative, purpose, reason, and conditional. kelative clauses can be restrictive
or non-restrictive with distinct lorms lor each.
. 1 coordi nate cl auses
coordinate clauses are ol the lollowing types in lengo: simultaneous ('while'),
sequential ('and, but'), and alternate ('or'). 8imultaneous and sequential clauses are
joined by a lorm ol the conjunction ma 'and, but' while alternate clauses are joined by a
lorm ol the conjunction pa 'or'. loth ma and pa can drop their JuJ and join to the
subject relerence pronoun as m- and p-. ln this regardthat is, that they are joined as
prelixes to the subject relerence pronounthey behave like the 1Am marking prelixes.
oiven that they supplant other 1Am markers by occupying their slot at the lront ol the
subject relerence pronoun, m- and p- carry the 1Am signature ol the preceding subject
relerence pronoun. 1here are instances ol up to ten sequential clauses, that is, ten
clauses marked with m- 'cON}', between clauses whose subject relerence pronouns are
marked lor 1Am.
140
. 1. 1 8i mul taneous
when two clauses describe simultaneous actions or states one ol them is
introduced with the time phrase na bona 'Ak1 time, while, when'. 1he clause introduced
with na bona is a dependent clause. lt is always in realis mood (marked by t-8kl) while
the independent clause can be unmarked, marked perlective (by g-8kl), or joined by a
140
1here is an example ol this in Yesterday", the lourth text in Appendix c.
189
subject relerence pronoun marked with the conjunction m-. 1he na bona clause may be
the initial clause, as in (!!6)a. and (!!)a., though it may also be non-initial, as in (!!6)b.
(!!6) a. 9a bona t-ami thudu m-ara ghoni 'engini'.
Ak1 time kl-1lX.ll sit cON}-!ll lix engine
'while we sit they lix the engine.'
b. e thudu na bona t-e gholi
!8o sit Ak1 time kl-!8o scrape.coconut
'she sits while she scrapes coconut'
1he dillerences between (!!6)a. and b. deserve lurther comment. ln (!!6)a. the
dependent na bona 'while' clause comes lirst whereas in (!!6)b. the na bona clause
comes second. ln (!!6)b. the lirst clause is lully independent but in (!!6)a. there is a
certain level ol interdependence between the clauses. clearly the na bona 'while' clause
is dependent, setting the scene lor an independent clause describing the nuclear action
or state. lut the second clausemara ghoni engini 'they lix the engine'is dependent on
the preceding clause by virtue ol the obligatory clause conjunction m-. As is discussed
below in the section on sequential clauses, m- is prelixed to the subject relerence
pronoun ol a clause that lollows sequentially on the preceding clause. when a na bona
clause lollows another clause, as in (!!6)b., the independence ol the lirst clause is
unmistakable. when, however, a na bona clause precedes another clause, there is an
interdependent relationship between them.
ln (!!)a., the speaker's 'good body' was simultaneous with being a 'young boy'
and being 'able to play'. ln (!!)b., however, the clause beginning with m-e na deni
'cON}-!8o Ak1 ulm' marks a shilt in time lrom 'when' (i.e., some time in the past) to 'now'.
(!!) a. Ma na bona t-u gari vaolu m-e doku na thuli-gu ni bebeu
cON} Ak1 time kl-18o boy young cON}-!8o good Ak1 body-l8:18o olN play
190
g-u tangomana na bebeu
llc1-18o able Ak1 play
'And when l was a young boy my body lelt good lor play and l was able to
play'
b. m-e na deni u ghua thairo kikiki ba k-e viti.
cON}-!8o Ak1 ulm:8o 18o cONu work little lL1 lkk-!8o pain
'but now il l work a little bit it will hurt.'
lt is dillicult to understand (!!)a. and b. as being simultaneous. while in (!!)a. the
clauses are simultaneous to one another, there is a signilicant time lapse between
(!!)a. and b.
. 1. 2 8equenti al
8equential clauses are joined with a lorm ol the conjunction ma 'and, but', as in
(!!8) and (!!9). ln (!!8) two separate events are loundlandward going and carrying
joined by the conjunction ma.
(!!8) )-ami-tu longa dea ma k-ami-tu dea tura-i na koga ira.
lkk-1lX.ll-lAL landward go cON} lkk-1lX.ll-lAL go carry-ll Ak1 stick ll:!ll
'we will go landward and we will go carry them, the sticks.'
1he speaker was at a seaside village describing his plan lor the day: to go retrieve some
house building materials he had set aside at the base ol a hill some distance inland. le
had to go landward to get to the sticks belore he could carry them.
lxample (!!9) demonstrates two other means used to join sequential clauses.
(!!9) 6osi-a na ato, m-e rosi-a na tiba, m-ami kere,
tear-O:!8o Ak1 leal.panel cON}-!8o tear-O:!8o Ak1 timber cON}-1lX.ll burn
vuti na tughuru, rosi talaghini-a ruka na vae t-e sovo.
pull Ak1 post tear completely-O:!8o two Ak1 house kl-!8o join
'[we] pulled down the leal-rool panels, took down the timber, we burned [the
rubbish], pulled the posts [out ol the ground], [we] tore down completely the
two houses that were joined.
1he lirst means used to mark sequential ordering in (!!9) is with the conjunction m-.
191
lurning the rubbish (kere) lollows temporally on taking down the timber (rosi-a na tiba)
which in turn lollows temporally on taking down the leal-rool panels (rosi-a na ato).
lach ol the 'lollowing' clauses are introduced with a subject relerence pronoun marked
with the prelix lorm ol the conjunction m-.
141
1he second approach to clause conjoining
in (!!9) is juxtaposition. lulling the posts out ol the ground (vuti na tughuru) which
lollows temporally on taking down the timber, is not introduced with a conjunction.
kather, it immediately lollows the preceding clause.
1here is one example in the data, presented in (!40), ol an m- clause that does
not appear to be conjoined to another earlier clause.
(!40) ! nea te roropo i muri-a na lotu m-u dea
i nea tena roropo i muri-a na lotu m-u dea
lOc yesterday lOc morning lOc behind-O:!8o Ak1 worship cON}-18o go
i vae ni talu-a nigua na 4ible.
lOc house olN put-O:!8o l8:18o Ak1 lible
'Yesterday in the morning alter church l went to [my] house to leave my lible.'
lxample (!41) is the lirst sentence ol a text in which the speaker recounts what he did
the previous day. leing the lirst sentence, there is no preceding clause lor this to join
to. 1he subject relerence pronoun marked with m- is curious in this instance.
while as a native lnglish speaker it would be nice to have recourse to a separate
contrastive conjunction (i.e., 'but'), there is no essential dillerence between 'and' and
'but' in lengo. 8o in the lollowing examples, where an lnglish speaker might preler a
gloss ol 'but', lengo speakers continue to use the same word ma 'and'.
141
8imons (1986:28) presents live aspects encoded by 1o'abaita subject pronouns (1o'abaita is a lellow
member ol the 8outheast 8olomonic language lamily lrom north malaita). 1he one ol interest here is
that which she calls 'serial'. ln narrative discourse, the general lorm is used at the outset ol an event
and the serial lorm is used in all lollowing clauses which are linked to it by a continuity ol action and
participants." 1he sense ol continuity ol action and participant(s) expressed by a subject pronoun
with 'serial' aspect is quite intriguing with relerence to lengo m- 'cON}'.
192
(!41) a. + ngao-a na iti niu m-e viti na tua-gu.
18o want-O:!8o Ak1 up coconut cON}-!8o pain Ak1 leg-l8:18o
'l want to climb the coconut [tree] and my leg hurts.'
b. 0 utha m-e lalai.
!8o rain cON}-!8o stop.raining
'lt was raining and it stopped.'
clearly the clauses marked with m-e above are in contrast with the clauses preceding,
the subject ol (!41)a. would be up the tree il his leg didn't hurt. lt is the 'but' that
prevents him lrom going up. lut there is no way to mark adversative clauses in lengo
other than with the conjunction ma.
1he locative tena has a clause conjoining counterpart, mena 'but now'. lt is used
to join two clauses, the second ol which stands in contrast to the lirst. 1he locative
mena is another case ol m- 'conjunction' taking the place ol t- 'realis' at the beginning ol
a word.
142
whereas tena adds an emphatic 'now' to a clause,
14!
mena adds the 'now' with
an additional contrastive element: 'but now'. consider the examples in (!42).
(!42) a. t-ara be(bere-ngi-a dea i beti m-i tathi m-ara rongo-ni-a
kl-!ll kul-see-1k-O:!8o go lOc river cON}-lOc sea cON}-!ll hear-1k-O:!8o
po teigha ghani-ra na kau mena ia e dea ghani tinoni t-i
just Nlo bite-O:!ll Ak1 dog lOc ulm !8o go bite person kl-lOc
'they watched it [a crocodile] go in the river and the ocean and they heard it
ate nothing but dogs but now it has gone and bitten a person'
b. Ma na bona t-u gari vaolu m-e doku na thuli-gu ni bebeu g-u
cON} Ak1 time kl-18o boy young cON}-!8o good Ak1 body-l8:18o olN play llc1-18o
tangomana na bebeu, mena deni u ghua thairo kikiki ba k-e viti.
able Ak1 play lOc ulm:8o 18o cONu work little lL1 lkk-!8o pain
'And when l was a young boy my body [lelt] good lor play and l was able to
play, but now il l work a little bit it hurts.'
ln both examples mena 'lOc' is lollowed by a demonstrative which helps situate the
142
cl. m-i 'cON}-lOc' and t-i 'kl-lOc' in (!42)a.
14!
1he sentence lollowing (!42)b. in the source text begins tena deni u thudu le po 'now l just sit still'.
19!
temporal location at the deictic centre ol the text.
when the second ol two conjoined clauses is in irrealis mood (k-) and the subject
relerence pronoun ol that clause is a single syllable (i.e., u '18o', o '28o', e 'l8o', or a
'1lN.ll'), the conjunction ma joins with a subject relerence pronoun ol the same person
and number and precedes the irrealis marked subject relerence pronoun (e.g., m-e k-e
sobo 'cON}-!8o lkk-!8o lloat, and it lloats', m-o k-o sau-i 'cON}-!8o lkk-!8o wash-O:!ll, and you
wash them'). ll the subject relerence pronoun is one ol the multiple syllable lorms (i.e.,
ami, amu, ara), the conjunction ma is lound (e.g., ma k-ami dea 'cON} lkk-1lX.ll go, and we
go').
1he conjunction ma J m- is not lound belore t- or g- prelixed subject relerence
pronouns in the same way they are lound belore k-8kl. lnstead, where a sentence is
marked, lor example, realis (t-8kl) at the outset and a coordinate clause is added, the
subject relerence pronoun in the second clause is marked with m-. 1he prelix m- takes
on the aspect ol t-, lorming a composite 1Am J conjunction marker.
144
1he conjunction m- also joins to the locative i (compare g-i and t-i).
(!4!) *ra-ko leta t-i vini vi-leuth-i8 na leoni bona ara-ko
!ll-uL 'no save' kl-lOc actually klcl-light-O:!ll Ak1 all time !ll-uL
vi-doku-i po t-i i levu m-i levu.
klcl-good-O:!ll llm kl-lOc lOc side cON}-lOc side
'1hey never lought each other, they were always just good to each otherlrom
both sides.'
kather than *i levu i levu in (!4!), i levu m-i levu 'lOc side cON}-lOc side' is used. 1he second
locative in a series ol locatives is conjoined by m- (there is no evidence in the data ol
144
compare yet another description ol oela: 1he copula ma 'and' is compounded with the verbal
particle e in the lorm me, and is used thus to carry on the tense signilicance even when the tense ol
the preceding verb is past, the subject not being lurther expressed" (lvens 19!:109).
194
three or more locatives in series).
ln example (!44) there is an interesting use ol m- 'cON}' as indicating a sort ol
alternation relation.
(!44) 0 ngitha na ivolo ni pana5
!8o how.many Ak1 ten.garden.row olN pana
'low many ten.garden.rows ol pana [are there in your garden]7'
#ima teigha m-e ono na ivolo.
live Nlo cON}-!8o six Ak1 ten.garden.row
'live, or il not, six ten.garden.rows.'
1his presents a kind ol middle ground between conjunction and alternation. 1he
speaker, not knowing exactly how many garden rows ol pana (a root crop) he has,
lrames his answer in a teigha m-e construction: 'could be live, could be six, somewhere
around that many'. lt is curious that, given how this kind ol an answer leans more
toward alternation (see below) than conjunction, that the m- conjunction lorm is used.
. 1. ! Al ternate
Alternates are joined by pa J p- 'or'. with pa J p- the situation is much the same
as with ma J m-, in that it is sometimes presented as an allix on the subject relerence
pronoun in the 1Am slot. 1his raises the important question ol whether or not there is
1Am associated with pa J p-. At lirst glance, one would be inclined to say 'no'
conjunctions do not normally have 1Am. lut consider the lollowing: when one is
presented with alternatives, as is the case with pa J p- 'or', the alternatives remainat
the moment and lor the purposes ol the utteranceunrealised. lt is expected that one
ol the alternates will be (or already is) realised, but to the speaker the alternates remain
mere possibilities. 8een in that light, pa J p- exhibits shades ol irrealis mood, or at least
a (potential) shilt lrom irrealis to realis. lowever, until more convincing prool arises
19
that pa J p- does in lact indicate 1Am, it will be analysed as a conjunction.
1he examples in (!4) demonstrate the role ol pa J p- 'or' in holding together
two alternates.
(!4) a. @ari mane pa na gari ghaoka5
child boy Al1 Ak1 child girl
'ls it a boy or a girl7'
b. ko mo lubathia moa pe dea
2sg Nlo let.!sg Nlo or.!sg go
'uon't let it out (release it) or it will run away.' [8ymons, !..!]
c. ? oti-ra p-ara dea5
28o hold-O:!ll Al1-!ll go
'Are you holding them or did they go7'
d. 4a k-u oti-a p-u k-u lubathia.
lL1 lkk-18o hold-O:!8o Al1-18o lkk-18o release-O:!8o
'l will hold it or l will release it.'
1he conjunction pa J p- 'or' behaves exactly the same as ma J m- 'and, but' with
relerence to serial single-syllable subject relerence pronouns one ol which is marked
irrealis, as (!4)d. demonstrates.
14
ln addition to the irrealis marked subject relerence
pronoun (k-u) a subject relerence pronoun marked with the alternate conjunction p- is
lound (p-u). multi syllable subject relerence pronouns behave as does ara '!ll' in (!4)c.
Another means lor presenting alternatives, as has been shown in (!44) above, is
with the construction teigha ma = m- '[il] not, then'. when using this construction the
speaker holds signilicantly less commitment to the alternatives than with pa = p-. 1here
is an openness to options lurther up and down the scale ol alternatives that the speaker
has presented.
14
8ee the discussion on page !.
196
. 2 8ubordi nate cl auses
8ubordinate clauses ol complement, relative, purpose, reason, and condition
types are lound in lengo. 8ubordinate clauses normally lollow main clauses, though
they can be moved ahead ol the main clause lor the purpose ol topicalisation.
. 2. 1 compl ement
complement clauses in lengo are introduced by the complementiser a thi = na
tha 'Ak1 cOml'.
146
1he presence ol the article indicates the nominal nature ol the
complement clause. 1he examples in (!46) illustrate.
14
(!46) a. k-amu mai rongo-ni-a [na tha ba k-a ghali-a]
lkk-2ll come hear-1k-O:!8o Ak1 cOml lL1 lkk-1lN.ll do-O:!8o
'you come listen to what we will do'
b. e ghilaghana-a [na tha t-e doku vani-ghita]
!8o know-O:!8o Ak1 cOml kl-!8o good uA1-O:1lN.ll
'he knows what is good lor us'
c. ba k-a ghe memele k-a ghilaghana-a [a thi ba k-e
lL1 lkk-1lN.ll continue choose lkk-1lN.ll know-O:!8o Ak1 cOml lL1 lkk-!8o
tangomana lavi-ghita tena niulu vaolu]
able take-O:1lN.ll lOc year new
'we will choose, we will know who will be able to lead us in the new year'
d. ba k-u tugu vani-ghamu [na tha ba k-e sara tena arere]
lL1 lkk-18o tell uA1-O:2ll Ak1 cOml lL1 lkk-!8o arrive lOc 8aturday
'l will tell you what [is going] to happen on 8aturday'
1he complement clause as nominalised Vl serves as an object argument to the head
verb.
ln each ol the examples in (!46) the object complement lollows the verb.
lxample (!4) demonstrates an entirely dillerent ordering.
146
Again, the dillerence between a thi and na tha is humanity: the lormer is used lor humans and the
latter lor all else.
14
1he complement clauses in these examples are indicated with square brackets.
19
(!4) [na tha ba k-amu ghali-a] ighamu tovothi na rurutu ni vanua
Ak1 cOml lL1 lkk-2ll do-O:!8o ll:2ll separate Ak1 group olN village
ba k-amu rongo-ni-a t-i ngeni de
lL1 lkk-2ll hear-1k-O:!8o kl-lOc today ulm
'what you dillerent village groups will do you will hear today'
ln (!4)a preamble to instructions lor a leastthe object complement ol rongonia
'hear it' is lound at the beginning ol the sentence.
. 2. 2 kel ati ve
kelative clauses are common in lengo. loth subject and object noun phrases are
modilied by relative clauses. 1he most common lorm that a lengo relative clause takes
is ol a postnominal externally headed clause and is realised as a clause in realis mood
lollowing the modilied head.
148
lt is not surprising to lind a relative clause, which is by
nature descriptive, to be set in realis mood.
149
1here are also a lew instances ol relative
clauses in which the pronoun thi, in these cases a relative pronoun marked 'human',
takes the place ol the relativised noun. 1here is no overt marking to dillerentiate
restrictive lrom non-restrictive relative clauses.
A noun phrase subject may be relativised with a clause in realis mood. lxample
(!48) shows the subject noun phrase na tinoni 'the people' lollowed by a relative clause
t-ara kuki 'they cooked'.
10
(!48) m-ara lavi mai-a [na vanga] [na tinoni [t-ara kuki ]]
cON}-!ll take come-O:!8o Ak1 lood Ak1 person kl-!ll cook
Vl O 8 kel
8
'the people who cooked [it] brought the lood'
As is evident, the relative clause lollows the head noun. lxample (!49), an intransitive
148
1his is consistent with layne's (199:!26) observation that, languages which are dominantly VO in
main clause constituent order always have postnominal relative clauses."
149
lt needs to be made clear that not every clause marked realis is relative.
10
ln the lollowing examples the various constituents are delimited with [square brackets].
198
sentence, also leatures a postnominal relative clause.
(!49) m-e thai mo [na dani [t-ara-ko ghunughunu-a]]
cON}-!8o arrive more Ak1 day kl-!ll-uL plan-O:!8o
Vl 8 kel
8
'and the day that they planned lor arrived'
ln both (!48) and (!49), the relativised subject is alter the verb. A topicalised
subject, that is, a subject at the beginning ol a sentence, may also be relativised as in
the examples in (!0).
(!0) [na ghai [t-e doku]] boro k-e ghoe thaghata [na vua-na]
Ak1 tree kl-!8o good impossible lkk-!8o produce bad Ak1 lruit-l8:!8o
8 kel
8
Vl O
'it's impossible lor a tree that's good to produce poorly its lruit'
uespite the relative clause's location belore the verb, it still lollows the noun it
relativises.
A subject noun phrase may also be relativised with the relative pronoun thi. 1his
type ol relative clause is apparently only used lor humans as subject, the data does not
contain any examples ol thi used with relativised human objects, nor any instances ol
the non-human pronoun tha used with relativised non-human subjects or objects.
lxample (!1) demonstrates this type ol relative clause.
(!1) a. [na tinoni ketha [a thi ga deni ]] ba k-e mai lau-a
Ak1 person dillerent Ak1 kll there ulm lL1 lkk-!8o come take-O:!8o
8 kel
8
Vl
[pilea na vanga de]
little.bit Ak1 lood ulm
O
'a dillerent person who is over there, he will come take a bit ol this lood'
b. [a thi [ighoe] t-o mono i buri ti dea] k-o gharasu mai
Ak1 kll ll:28o kl-28o stay lOc back go lkk-28o move come
kel
8
- 8 - kel
8
Vl
'you who are at the lar back move this way'
199
lxample (!1)b. presents an interesting case. 1he subject, the second person singular
emphatic pronoun ighoe 'you', is interposed between the relative pronoun and the
remainder ol the relative clause.
A noun phrase object is relativised with the same type ol construction: a clause
in realis mood. consider the examples in (!2).
(!2) a. m-u ghoni-a [na pou [t-ami uru-vi -a tena rodo]]
cON}-18o prepare-O:!8o Ak1 b.d.m. kl-1lX.ll dive-1k-O:!8o lOc night
Vl O kel
O
'l prepared the beche-de-mer that we dove lor during the night'
b. ami-ko kere-ra [na leoni igha [t-ami -ko l avi ]]
1lX.ll-uL burn-O:!ll Ak1 all lish kl-1lX.ll-uL caught
Vl O kel
O
'we cook the lishes that we caught'
c. tena deni u thudu le po ni bungu-ti-ra po [na gari vaolu
lOc ulm 18o sit still llm olN watch-1k-O:!ll llm Ak1 child new
Vl O
[t-ara ghal i -a na thai ro]]
kl-!ll do-O:!8o Ak1 work
kel
O
'now l just sit still to just watch the youths that do the work'
As is the case with a relativised subject noun phrase, the relative clause is lound
lollowing the relativised object.
. 2. ! lurpose
1here are two means lor marking a purpose clause in lengo. 1he lirst is with the
genitive ni.
11
lound belore a verb, ni is a genitive ol purpose.
12
Normally, the subject
relerence pronoun is dropped belore the verb ol purpose. 8o where one might expect to
lind a subject relerence pronoun in (!!) belore the verb ol purpose tughia 'hammer
11
According to crowley (2002a:29) oela also uses ni to indicate lLklJclAkJlO88 (purpose J character J
possession).
12
Other aspects and uses ol the genitive ni have been discussed beginning on page 4.
200
s.t.' (i.e., *m-ara-tu iti dea ni [m-ara-tu] tughi-a na kava), it is absent.
(!!) m-ara-tu iti dea ni tughi-a na kava vete oti na tughuru
cON}-!ll-lAL up go olN hammer-O:!8o Ak1 copper lLkl hold Ak1 post
'they climbed up to hammer the iron lor holding the posts'
likewise, in (!4) *m-u dea i vae ni [m-u] talu-a nigua na 4ible is not lound.
(!4) muri-a na lotu mu dea i vae ni talu-a nigua na 4ible
behind-O:!8o Ak1 worship cON}-18o go lOc house olN leave-O:!8o my Ak1 bible
'alter church l went to my house to leave my lible'
Again, the subject relerence pronoun is absent belore talua 'leave something', the verb
ol purpose. koss (2004:22) notes that in longgu, a 8outheast 8olomonic language
which borders lengo, desententialisation [ol adverbial and complement clauses] is
atypically widespread and brings about corelerential subject deletion." koss indicates
that it is not the norm lor canonic Oceanic languages to delete corelerential subjects in
purpose clauses, and cites longgu as somewhat ol an exception. lengo ought to be
added to the list ol exceptions as it also makes common practise ol corelerential subject
deletion, but it is not strictly obligatory (see (!6) below).
ln the case ol a complex verbin (!) a serial verb marked imperlective (bo dea
pogho 'going to bathe')ni stands at the beginning ol the construction.
(!) M-e thivo ni bo dea pogho m-e mai na vua.
cON}-!8o down olN lmll go bathe cON}-!8o come Ak1 crocodile
'As she was going down to bathe a crocodile came.'
while it is most common lor the subject relerence pronoun to be dropped
belore the purposed verb, it is possible to include it. when the subject relerence
pronoun is present, it is lound lollowing ni, as in (!6).
(!6) k-o duara niu ni k-o pulu-i
lkk-28o pour coconut olN lkk-28o roll.up-O:!ll
'you pour the coconut so that you can roll them'
201
A second means lor marking purpose is with ne 'so that'. 1he particle ne is lound
between the subject relerence pronoun and the verb. whereas with ni the subject
relerence pronoun may be deleted (as demonstrated above), the subject relerence
pronoun is never dropped in a ne purpose clause.
(!) a. + ngao-a na dea laka inau k-u ne lavi-a nigua na ghau.
18o want-O:!8o Ak1 go also ll:18o lkk-18o so.that get-O:!8o my Ak1 knile
'l want to go also so that l can get my knile.'
b. 6oropo rukana k-o pelu vani-u ruka na vugho
morning day.alter.tomorrow lkk-28o purchase uA1-O:18o two Ak1 net
k-u ne vu(vugho ko-gha-da igha.
lkk-18o so.that kul-net uL-oral.cll-l8:1lN.ll lish
'Next tomorrow morning you buy lor me two nets so that l can net some lish
lor us to eat.'
A linal means lor marking purpose is with vete. whereas ni and ne mark an
agent's purposive action, vete has more ol a sense ol something's reason lor existence.
consider (!8).
(!8) m-ara-tu iti dea ni tughi-a na kava vete oti na tughuru.
cON}-!ll-lAL up go olN hammer-O:!8o Ak1 iron lLkl hold Ak1 post
'they climbed up to hammer the iron lor holding the posts.'
lere the purpose ol the 'iron'
1!
to hold the posts in placeis in view.
. 2. 4 keason
keason clauses lollow the main clause they substantiate. A reason clause is
introduced with pukua 'because' and are always lound in realis mood (t-8kl).
(!9) a. ?h, t-e ghani-a tinoni ia pukua t-e gora.
lN1} kl-!8o bite-O:!8o person ulm because kl-!8o hungry
'Oh, it bit that person because it was hungry.'
1!
lron (lijin kava 'copper') is a generic term lor metal among lengo speakers.
202
b. t-ara ghali-a na thairo pukua t-e viti na aro-gu
kl-!ll do-O:!8o Ak1 work because kl-!8o pain Ak1 shoulder-l8:18o
'they are doing the work because my shoulder hurts'
A reason is negated within the subordinate clause, as in (!60).
(!60) a *. e thaghataleo vata po pukua t-e teigha na logho-a
Ak1 A !8o bad.inside continue llm because kl-!8o Nlo Ak1 have-O:!8o
siki dae-na
any child-l8:!8o
'A. just continued to worry because he didn't have any children'
1his amounts to an X because not Y" type ol clause.
. 2. condi ti onal
1he lengo conditional, as was discussed in relation to the verb phrase on page
114 and lollowing, is marked with ghua 'il'. lere it remains to discuss some ol the
pragmatic aspects ol the use ol the conditional ghua. 1he conditional is lound with
three varieties ol 1Am marking on the accompanying subject relerence pronoun:
unmarked, realis (t-), and irrealis (k-). lach ol these bears signilicantly on the meaning
ol the conditional.
An unmarked subject relerence pronoun used in conjunction with ghua may
indicate a 'zero conditional', that is, a conditional whose probability ol happening is as
good as certain. consider the examples in (!61).
(!61) a. ma na bona t-u gari vaolu m-e doku na thuli-gu ni bebeu g-u
cON} Ak1 time kl-18o boy young cON}-!8o good Ak1 body-l8:18o olN play llc1-18o
tangomana na bebeu, m-e na deni u ghua thairo kikiki ba k-e viti
able Ak1 play cON}-!8o Ak1 ulm:8o 18o cONu work little lL1 lkk-!8o pain
'and when l was a young boy my body [lelt] good lor play and l was able to
play, but now il l work a little bit it hurts'
20!
b. e dea ghua i namo na vonu m-e tumuri po na igha
!8o go il lOc deep Ak1 turtle cON}-!8o lollow just Ak1 lish
'il the turtle went deep the lish just lollowed'
A 'lirst conditional', that is, a conditional whose level ol probability is likely, can
be introduced with either an unmarked or realis subject relerence pronoun, as in (!62).
(!62) a. Ma k-ami-tu mai soni iti ga i T. t-ami -tu ghua mavulo.
cON} lkk-1lX.ll-lAL come throw up there lOc 1 kl-1lX.ll-lAL il willing
'And we will throw [the sticks] up at 1. il we are willing.'
b. e ghua logho bobona vata i nganiba na nulavi m-ami-tu
!8o il have time continue lOc later Ak1 alternoon cON}-1lX.ll-lAL
dea lebo-i laka t-i
go lloat-O:!ll also kl-lOc
'il there is still time in the alternoon [and we do not know il there will be or
not], we will also lloat them [down the river]'
c. o ghua ngao-a na ghali-a siki 'e.f.' t-i i leo-a t-i
28o il want-O:!8o Ak1 make-O:!8o any e.l. kl-lOc lOc inside-O:!8o kl-lOc
nimoa na thara k-o ghali mai-a
l8:28o Ak1 least-lood lkk-28o make come-O:!8o
'il you want any 'extra llavour' in your least-lood [and l do not know il you
do or not], you make it and bring it'
1he irrealis marked subject relerence pronoun together with ghua indicate a
'second conditional'one whose probability is unlikely to be realised in the luture.
(!6!) m-e k-e ghua teigha tu-mana-miu nga geia po na tha t-ami
cON}-!8o lkk-!8o il Nlo lAL-enough-l8:2ll there ll:!8o llm Ak1 kll kl-1lX.ll
tangomana ighami
able ll:1lX.ll
'and il it's not enough lor you, it's all that we're able [to do] . . . '
1he expectation is that what 'we' have done &ill be enough. As such, there is no
contingency plan (il there &as a backup plan it would indicate an expectation that it
&ouldn't be enough).
204
1he conditional magi 'suppose' introduces an assumption that is presented as a
hypothesis lor argument's sake. 1here is an element ol 'condition' in magi statements,
but it is only to carry the argument lorward, the 'possibility' is as good as done.
consider (!64), which is lrom a conversation lollowing a crocodile attack.
(!64) %ana-thi mate-a magi veghe na agotha ni mate-a tavu-a na theutu
shoot-1k die-O:!8o il like Ak1 hunt olN die-O:!8o lind-O:!8o Ak1 path
i vi ga t-o tangomana na mate-a.
lOc where there kl-28o able Ak1 die-O:!8o
'll [you want to] shoot it deadlike hunt it to deathlind a way that you are
able to kill it.'
1he assumption is that everyone does want the crocodile shot dead. 8ince that is the
case, it remains to lind a way to make it happen.
clause conjoining is a signilicant part ol lengo grammar and discourse. 1he
briel treatment that coordinate (simultaneous, sequential, and alternate) and
subordinate (relative, purpose, reason, and conditional) complex clauses received here
could be lruitlully explored in greater depth. lor the present, l leave it to another
researcher to rise to that challenge.
20
8 kllOk1lu 8lllcl
1here are a number ol ways to indicate reported speech in lengo. 1he lirst is
with a quote margin comprised ol an optional 1Am marker ( or g-Jm-), a singular lorm
ol the subject relerence pronoun, and -na 'say': (1Am-)8kl-na. 1here are no plural lorms.
(!6) prelix 8kl 'say'
{
E
g D
mD
} {
u
o
e
}
-na
ly lar the most common lorm ol the quote margin is m-8kl-na. 1he next most common
lorm is unmarked subject relerence pronoun - na. 1here is a single example ol g-8kl-na
in the data (see (!68) below). 1here is no evidence, however, ol t-8kl-na 'kl-8kl-say' nor k-
8kl-na 'lkk-8kl-say'.
1he quote is in the lrame: (quote margin) quoted speech" (quote margin). 1hat
is, the quote margin can be lound belore, alter, or surrounding the quoted speech. ln
(!66) the quote margin is belore the quote.
(!66) oko m-ena T., B@ari, k-amu mai thanga-u.C
linish cON}-!8o:say 1 boys lkk-2ll come help-O:18o
'1hen 1. said, loys, come help me."'
ln (!6) the quote margin is alter the quote.
(!6) B)-a-tu oli ba t-i,C una m-u tapa.
lkk-1lN.ll-lAL return lL1 kl-lOc 18o:say cON}-18o run
'we need to go back now," l said and l ran.'
ln (!68) the quote margin elements are lound both belore and alter the quote. 1his
latter construction is, however, rare.
(!68) *ra-ko g-ena na ghaoka dini, . . . ," g-i igeira ena.
!ll-uL llc1-!8o:say Ak1 woman ulm:ll llc1-lOc ll:!ll !8o:say
'1hose two women said, . . . ," they said.'
206
Note in (!68) that there is a lack ol agreement in number between ara-ko '!ll-uL', dini
'ulm:ll', igeira 'll:!ll' and g-ena 'llc1-!8o:say'. 1he pronouns ara-ko, dini and igeira are
plural while the quote margin g-ena is singular. oiven that this quote marker is limited
to singular lorms ol the subject relerence pronoun, agreement in number is not always
possible. 1he lack ol agreement between non-singular subjects and singular quote
margins gives rise to clauses with two lorms ol the subject relerence pronoun.
ln order to circumvent this 'lack ol agreement', there is another quote margin at
the lengo speaker's disposal. lt is the dative bosa vani 'say uA1' - object person pronoun.
(!69) ami -tu dea bosa vani -ra balu gari laka BMai . . . C
1lX.ll-lAL go speak uA1-O:!ll some boys more come
'we lew went and spoke to some more boys, come . . . "'
lere the person and number ol the speaker(s) is not constrained as it is above with the
8kl-na construction.
Quoted speech can also be introduced with verbs ol speech such as ghu 'shout'
and thuge 'ask'.
(!0) a. 0 ghu-vi -a dae-a, B*i1 0 ghani-a na vua
!8o shout-1k-O:!8o child-l8:!8o lN1} !8o bite-O:!8o Ak1 crocodile
a tina-mu,C ena.
Ak1 mother-l8:28o !8o:say
'8he shouted to her son, ley! 1he crocodile is biting your mother," she said.'
b. m-u dea thuge-a a J., BJ., . . . 5C 0na, B . . . C
cON}-18o go ask-O:!8o Ak1 } } !8o:say . . . "
'l asked }., }., . . . 7" le said, . . . "'
oiven, though, that ena '!8o:say' is stated at the end ol (!0)a. but absent lrom b. (ena in
(!0)b. is the quote margin preceding }.'s response), it seems that thuge 'ask' is a closer
equivalent to the speech margin subject relerence pronoun - na than ghu 'shout'. 1he
20
quote margin subject relerence pronoun - na is used lor statements and thuge 'ask' lor
questions.
Questions posed and answers given can be presented back to back, with quote
margins belore the question and alter the answer to dillerentiate the speakers.
(!1) m-ara-ko thuge-u, B . . . 5C B . . . ,C una m-u tapa
cON}-!ll-uL ask-O:18o 18o:say cON}-18o run
they two asked me, . . . 7" . . . ," l said and l ran
1he same holds true lor quoted indicative statements lrom dillerent interlocutors: they
can also be presented back to back with quote margins belore the lirst and alter the
second quote to dillerentiate the speakers.
(!2) M-ena po t-i, B . . . C B . . . ,C ena t-i na vonu.
cON}-!8o:say llm kl-lOc !8o:say kl-lOc Ak1 turtle
lt just said, . . ." . . . ," the turtle said.
lengo quote margins stand out lor the lack ol plural lorms. while speakers are
not averse to using a singular quote margin lor plural participants, they do have
recourse to the dative serial verb construction bosa vani - object person pronoun to
ensure agreement ol number.
208
9 cONclL8lON
1his grammar presents some ol the more readily evident leatures ol the
8outheast 8olomonic language lengo. l have presented an overview ol the phonology
along with more detailed discussions ol word, phrase, clause, and sentence level syntax.
oiven the paucity ol grammar descriptions available lor 8olomonic languages, it is my
hope that this grammar will prove uselul lor typological or comparative work in the
region. Ol course, there are some aspects ol lengo that have not been covered to the
depth they deserve in this description and there remains signilicant room lor lurther
exploration and discovery about the syntax ol lengo.
At the level ol phonology, lengo is interesting lor the predictable process ol
vowel harmony in the case ol third person singular possessive -a joining to words
ending in JuJ. 1he two JuJs raise to an JeJ as does any consecutive JuJ to the lelt (e.g.,
[lima] 'hand' - [-a] 'l8:!8o' [lime] 'her hand', [atha] 'name' - [-a] 'l8:!8o' [ethe] 'his
name'). Any vowel to the lelt other than JuJ disrupts the harmonisation process (e.g.,
[kabula] 'inner thigh' - [-a] 'l8:!8o' [kabule] 'his inner thigh').
Nouns are divided into classes based on possession typedirect (semantically
inalienable) and indirect (semantically alienable)and also based on which article
common or properthey take. 1he articles are not marked lor deliniteness, the same
article is used lor delinite and indelinite relerence. Verbs can be derived lrom nouns by
a process ol stem-initial cV reduplication.
lengo has a rich paradigm ol pronouns. 1hey include emphatic, subject
relerence, object, and possessor. 1here are lour persons (1 inclusive J exclusive, 2, and
!) and lour numbers (singular, plural, dual, and paucal). ln addition to these there are
209
rellexive and interrogative J relative pronouns. 1he latter have animate and inanimate
lorms.
Verbs are divided into two typesA-type and L-typebased on the role ol the
subject in corresponding transitive and intransitive lorms. with A-type (dynamic)
verbs, the subject ol the intransitive corresponds to the agent ol the verb's transitive
lorm. with L-type (stative) verbs, the subject ol the intransitive becomes the undergoer
ol the transitive. lntriguingly, in the same manner as above, nouns can be derived lrom
verbs by the same process ol stem-initial cV reduplication.
1he verb phrase is interesting lor the subject indexing subject relerence
pronoun which is inllected lor 1Am. 1his verb phrase constituent takes prelixes lor
perlective aspect (g-), and realis (t-), irrealis (k-), and apprehensive (b-) moods.
lmperative mood and luture tenses (possible and impossible) are expressed within the
irrealis mood. An unmarked subject relerence pronoun expresses indicative mood and,
with the auxiliary bo, imperlective mood.
lengo makes extensive use ol serial verb constructions. 1he various types
include directional, sequential, causative, manner, and ambient. 1here are also
prepositional verbs that express comitative, dative, and instrumental meanings. linally,
modal verbs are used to indicate abilitive, desiderative, negative, prohibitive, and non-
volitive moods.
8ince much ol my attention was given to word, phrase, and clause level
structures, prolitable study could be made in the area ol complex clauses. ln particular,
the roles ol the clause conjoining particles ma J m- 'and' and pa J p- 'or' and their
relationship to the 1Am allixes is not entirely clear. when m- or p- join to the subject
210
relerence pronoun they displace any 1Am-marking prelix that might otherwise be there.
complement and relative clauses, though discussed brielly, could also be explored to
greater depth.
ln conclusion, l suspect there are any number ol motivations lor a work like this.
lenneth like, the American linguist, developed the analogy ol the world's languages as
an orchestraeach 'instrument' adding a unique timbre to the collection ol
instruments and the whole orchestra being somehow less should one instrument be
absent.
14
Alter spending lour years in a lengo community, my motivation lor writing a
grammar ol their language is to let the lengo voice" be heard. lt is a beautilul voice, as
each language is. my dream is that some day a lengo speaker will set this warbling
rendition aside and let the world hear lengo as it was meant to be heard.
14
http:JJwww.wyclille.caJwordaliveJarchiveJ2004winterJindex.html
211
AlllNulX A: 8outheast 8ol omoni c l anguage lamil y
lynch et al. (2002:110) and the lthnologue (oordon 200:6!) have slightly
dillerent presentations ol the 8outheast 8olomonic language lamily tree.
1

lynch, koss, crowley lthnologue
8outheast 8olomonic lamily 8outheast 8olomonic (26)
lugotuJoelaJouadalcanal lamily oela-ouadalcanal ()
lugotu lughotu (1)
oelaJouadalcanal lughotu
oelic: oela and lengo oela (2)
west and 8outh ouadalcanal
16
oela
lonnguJmalaitaJmakira lengo
lonngu ouadalcanal (4)
malaitaJmakira lirao
8a'a, Llawa and Lki ohari
makira malango
malaita 1alise
'Are'are, Oroha, marau malaita-8an cristobal (19)
central and northern malaita malaita (14)
lonngu (1)
lonngu
Northern (9)
8outhern (4)
'Are'are
8an cristobal ()
1he dillerences between the trees are evident. where both language lamily trees agree,
however, is that lengo and oela are siblings", whereas lengo's immediate geographical
neighbours are cousins" (with longgu, and to a greater extent 'Are'are, having
somewhat distant lamilial relationships despite their geographical proximity). 1he map
in lllustration 4: lengo language neighbours shows lengo in geographic relation to
1
1he reader should note that only the languages proximate to ouadalcanallonngu and 'Are'areare
included in the malaita-8an cristobal branch ol the lthnologue tree. 1here are seventeen other
languages lrom that branch that are not listed here. 1he tree lrom lynch et al. is presented as it
appears in their work. lt is also worth mentioning that makira" and 8an cristobal" are dillerent
names lor the same island.
16
ln the discussion accompanying their language tree, lynch et al. mention the west ouadalcanal-
1alise-malango-lirao group ol languages. l assume their west ouadalcanal is the lthnologue's ohari.
212
surrounding languages.
21!
!llustration F: #engo language neighbours
ollA
oLAuAlcANAl
8AN1A l8Alll
mAlAl1A
1.
2.
!.
4.
.
6.
. 8.
9.
languages:
1. lengo
2. oela
!. lughotu
4. ohari
. malango
6. lirao
. 1alise
8. longgu
9. 'Are'are
AlllNulX l: Orthography
1here are three orthographies in use among lengo writers. leople tend to
lollow the 'conventions' ol their respective church denominations based on what the
various churches have introduced and adopted over the past century.
1
lven so, the
orthographies are not necessarily used consistently, even by the same author. lt is the
prenasalised JqJ that lies at the middle ol the issue. Or the JpJ. Or . . . Actually, it
depends where one begins. ligure 9.1: Orthography options presents some ol the
decisions that have been made by the various churches to come up with letters or
groups ol letters to represent the sounds in the language:
llA Anglican catholic Lnited
p 'bar-over-n'
18
'ng' 'g' or 'n'
p
q 'g' 'ngg' 'q'
'gh' 'g' 'gh'
Figure P.": ?rthograph< options
llA Ji
p
qeiuJ 'll:!8o', lor example, could be written as: igeia, inggeia, or iOeia. Obviously, il
the letter 'g' is used to represent one ol these three sounds ([], [
p
q], or [p]), some other
letter(s) have to be used lor the other two. oiven that prenasalised [
m
b] and [
n
d] are
written by all lengo writers as 'b' and 'd', l have chosen to do the same with [
p
q],
writing it as 'g'.
19
ln the interests ol avoiding diacritics, l decided against the bar-over J
bar-under-n, opting rather lor the digraph 'ng'. linally, given that two ol the three
1
1his is a complex issue and beyond the scope ol the present paper, but it should be noted that none ol
these orthographies were introduced specilically lor lengo. 1he Lnited church, lor example, has its
greatest sphere ol inlluence in western lrovince ol 8olomon lslands where koviana was adopted as
the church language". lowever, there are a lew Lnited church villages in the lengo language area
and the koviana J Lnited church orthography has been adopted by and adapted to lengo in those
villages.
18
1his is not reproduceable on a computer without resorting to the use ol a specialised lont. lor
example, uoulos 8ll can produce 'bar-over-n' (n) with the 'combining diacritic mark' L-0!04. lt is
somewhat easier to reproduce on a typewriter. 8ometimes technology constrains . . .
19
8imons (19:11) writes that, 1his solution deserves special consideration lor the luture il a system
ol standardized alphabets lor ouadalcanal languages is to be adopted."
214
orthographies use the digraph 'gh' lor the velar lricative, and that the 'g' is already
being used (lor pre-nasalised 'g', above), it seems a better choice than coming up with
something new such as, lor example, employing an otherwise 'unused' consonant such
as c, l, j, q, w, x, y, or z. 1he orthography used in this thesis is given in (!!).
(!!) llA u
m
b
n
d e
p
q i k l m n p p r s t Jz u Jv
lengo a b d e g gh i k l m n ng p r s t th u v
21
AlllNulX c: 8ampl e texts
Our lamily goes to l.
!ghami tamadae ami bo dea i ,. %uivuni ba k-ami oni(oni nimami
ll:1lX.ll lather.child 1lX.ll lmll go lOc l start lL1 lkk-1lX.ll kul-prepare l8:1lX.ll
we lamily are going to l. At the outset we will prepare our
na agutu. 9a tha t-ami ngao-a na lavi dea-i, ba k-ami lavi dea-i.
Ak1 thing Ak1 kll kl-1lX.ll want-O:!8o Ak1 take go-O:!ll lL1 lkk-1lX.ll take go-O:!ll
cargo. what we want to take we will take.
9a tha t-e mono t-ami teigha na ngao-i ba k-e mono oli i vae.
Ak1 kll kl-!8o stay kl-1lX.ll Nlo Ak1 want-O:!ll lL1 lkk-!8o stay return lOc house
what stays, what we don't want, will stay back at the house.
oko ba k-ami lavi dea-i i tathi vathai, ba k-ami dea talu-i ga.
linish lL1 lkk-1lX.ll take go-O:!ll lOc sea shore lL1 lkk-1lX.ll go put-O:!ll there
1hen we will take them to the sea shore, we will go put them there.
)-ami sakai ni sake thivo-a na thinaghe m-e k-e sobo i tena tathi,
lkk-1lX.ll one olN lilt down-O:!8o Ak1 canoe cON}-!8o lkk-!8o lloat lOc lOc sea
At the outset we lilt down the canoe and it lloats on the sea,
k-ami sakai ni lusa-ngi na tha t-ami ngao-a na lavi dea-i. oko k-ami
lkk-1lX.ll one olN load-1k Ak1 cOml kl-1lX.ll want-O:!8o Ak1 take go-O:!ll linish lkk-1lX.ll
at the outset we load what we want to take go. 1hen we
sakai ni togha. 9a bona k-ami togha ba k-ami kolivuti. )olivuti soko k-ami
one olN board Ak1 time lkk-1lX.ll board lL1 lkk-1lX.ll pray pray linish lkk-1lX.ll
board [the canoe]. when we are aboard we will pray. Alter praying we
sakai ni dea. )-ami dea ma k-ami dea ma k-ami dea sara i l. 4a
one olN go lkk-1lX.ll go cON} lkk-1lX.ll go cON} lkk-1lX.ll go arrive lOc l lL1
go. we go and we go and we go [until we] arrive at l.
k-ami-tu longa ma k-ami sake longa-i na leoni nimami na
lkk-1lX.ll-lAL landward cON} lkk-1lX.ll lilt landward-O:!ll Ak1 every l8:1lX.ll Ak1
we lew will [go] ashore and we lilt landward all our
agu(agutu. ! muri-a k-ami sakai ni sake longa-a na thinaghe.
kul-thing lOc behind-O:!8o lkk-1lX.ll one olN lilt landward-O:!8o Ak1 canoe
cargo. Alter that we lilt landward the canoe.
216
oko k-ami tavu theutu-mami ni dea i vae-mami t-e mono i tetete.
linish lkk-1lX.ll lind way-l8:1lX.ll olN go lOc house-l8:1lX.ll kl-!8o stay lOc hill.
1hen we lind our way to go to our house that is on the hill.
1he woman and the crocodile
9a tugu-a na ghaoka t-e ghani-a na vua. ! bongi ara-ko bo
Ak1 story-O:!8o Ak1 woman kl-!8o bite-O:!8o Ak1 crocodile lOc last.night !ll-uL lmll
1he story about a crocodile that bit a woman. last night they two were
dea pogho ara-ko ruka na ghaoka. *ra-ko dea pogho m-e tighi thivo a sakai na
go bathe !ll-uL two Ak1 woman !ll-uL go bathe cON}-!8o lirst down Ak1 one Ak1
going to bathe, they two two women. 1hey two went to bathe and one woman went
down lirst,
ghaoka, m-e tumuri-a a sakai. 0 thivo m-e theo-a na beti m-e
woman cON}-!8o lollow-O:!8o Ak1 one !8o down cON}-!8o scoop-O:!8o Ak1 water cON}-!8o
and one lollowed. 8he went down and she scooped some water and
e vuli iti ma na ruke e theo-a na beti m-e vuli dea i beti.
!8o pour up cON} Ak1 second !8o scoop-O:!8o Ak1 water cON}-!8o pour go lOc water
she poured it up[stream] and the second [time] she scooped some water and she
poured it across the river.
M-e thivo ni bo dea pogho m-e mai na vua. Mai na vua
cON}-!8o down olN lmll go bathe cON}-!8o come Ak1 crocodile come Ak1 crocodile
And she was going down to bathe and a crocodile came. A crocodile came
m-e mai gara-a. 0 gara-a m-e lavi dea-a. M-e ribe sakai
cON}-!8o come pull-O:!8o !8o pull-O:!8o cON}-!8o take go-O:!8o cON}-!8o surprise one
and it came [and] pulled her. lt pulled her and it took her. And one ol the women was
surprised
na ghaoka m-e ghu(ghu, e ghu-vi-a dae-a, J*i1 0 ghani-a na
Ak1 woman cON}-!8o kul-shout !8o shout-1k-O:!8o child-l8:!8o lN1} !8o bite-O:!8o Ak1
and she shouted, she shouted to her [the 'lirst' woman's] child, ley! A
vua a tina-mu,C e-na. Ma na vua e gara-a t-i na ghaoka
crocodile Ak1 mother-l8:28o !8o-say cON} Ak1 crocodile !8o pull-O:!8o kl-lOc Ak1 woman
crocodile is biting your mother," she said. And the crocodile pulled the woman
m-e gara dea m-e oti nga(ngasi na ghaoka tena buburu. @a(gara na vua
cON}-!8o pull go cON}-!8o hold kul-tight Ak1 woman lOc grass kul-pull Ak1 crocodile
and it pulled and she, the woman, held tightly to the grass. 1he crocodile pulled
21
m-e rongo-a dae-a t-e ghu-vi dea a tau-a ga m-e tapa
cON}-!8o hear-o:!sg child-l8:!8o kl-!8o shout-1k go Ak1 spouse-O:!8o there cON}-!8o run
and he, her child, heard her, the one who shouted to her spouse there and he ran
me mai. 0 mai me dea otia na lime tine me
m-e mai e mai m-e dea oti-a na lima-a tina-a m-e
cON}-!8o come !8o come cON}-!8o go hold-O:!8o Ak1 hand-l8:!8o mother-l8:!8o cON}-!8o
and he came.
160
le came and he went held his mother's hand and he
gara iti-a. @a(gara na vua8 ga(gara na da(dae8 m-e ga(gara ngata na
pull up-O:!8o kul-pull Ak1 crocodile kul-pull Ak1 kul-child cON}-!8o kul-pull strong Ak1
pulled her up. 1he crocodile pulled, the child pulled, and the child pulled
dadae me gara tine me gagara laka na vua. Aedea me
da-dae m-e gara tina-a m-e ga-gara laka na vua de-dea m-e
kul-child cON}-!8o pull mother-l8:!8o cON}-!8o kul-pull also Ak1 crocodile kul-go cON}-!8o
strong and the mother pulled and the crocodile also pulled. [As time] went on the child
was
ngata na da(dae m-e gara-a lithi-soni-a na vua. @-e
strong Ak1 kul-child cON}-!8o pull-O:!8o out-away.lrom-O:!8o Ak1 crocodile llkl-!8o
strong and he pulled her out away lrom the crocodile.
lubathia na vua. Me gara iti maia mo a tine
lubu-thi-a na vua m-e gara iti mai-a mo a tina-a
leave-1k-O:!8o Ak1 crocodile cON}-!8o pull up come-O:!8o more Ak1 mother-l8:!8o
1he crocodile lelt oll [lrom attacking] her. And he pulled his mother up toward him
more
m-e ngara koto. 0 ghani-a na tua-na m-e ghani voda laghi-ni-a
cON}-!8o injure lN18 !8o bite-O:!8o Ak1 leg-l8:!8o cON}-!8o bite hole through-1k-O:!8o
and she was badly injured. lt bit her leg and it bit a hole through
na lime. oko po mara lavi itia mara mai tavua na deresa 4.
na lima-a soko po m-ara lavi iti-a m-ara mai tavu-a na deresa l
Ak1 hand-l8:!8o linish llm cON}-!ll take up-O:!8o cON}-!ll come lind-O:!8o Ak1 nurse l
her hand. when it was all over they took her up and they came to lind the nurse l.
M-ara vou-a m-ara lavi dea-a i kliniki m-ara
cON}-!ll carry.on.a.pole.between.two.people cON}-!ll take go-O:!8o lOc clinic cON}-!ll
1hey carried her on a stretcher and they took her away to the clinic but they were
160
1he child ol the 'lirst' woman (the one bitten by the crocodile) was the spouse ol the 'second' woman
(the one who shouted).
218
teigha na tangomana na ghoni-a i kliniki m-ara lavi dea-a i 9aba 9ain
Nlo Ak1 able Ak1 make-O:!8o lOc clinic cON}-!ll take go-O:!8o lOc number nine
not able to make her [well] at the clinic and they took her away to Number Nine
161
te roropo i ngeni.
tena roropo i ngeni
lOc morning lOc today
this morning.
my shoulder hurts
Aoku. ! ngeni deni ami thairo ni vae ni kolivuti. 9a bona t-u thairo u
good lOc today ulm 1lX.ll work olN house olN prayer Ak1 time kl-18o work 18o
'Ol. 1oday we are working on behall ol the church. when l was working l
ngao-a po na tugu tauni-a na aro-gu t-e viti. ! ngeni u
want-O:!8o llm Ak1 tell around-O:!8o Ak1 shoulder-l8:18o kl-!8o hurt lOc today 18o
just wanted to tell about my shoulder that hurts. 1oday l
lavi-a na maul hammer m-u rosi-a na vae ki(kiki m-u
take-O:!8o Ak1 maul hammer cON}-18o tear.down-O:!8o Ak1 house kul-little.bit cON}-18o
took the maul hammer and l broke the house a little bit and l was
ne ribe na aro-gu mauli t-e rasa rasa m-e viti. 9a bona t-e
lmll surprised Ak1 shoulder-l8:18o lelt kl-!8o tear tear cON}-!8o hurt Ak1 time kl-!8o
really surprised that my lelt shoulder toretore and hurt. when it
viti m-u thudu varongo pukua i thau na bona t-u gari vaolu m-u bebeu
hurt cON}-18o sit still because lOc past Ak1 time kl-18o child young cON}-18o play
hurt l sat still because in the past, when l was a young boy, l played
tena rugb< na bona t-u mono i college: teacher's college. 9a bona t-u bebeu
lOc rugby Ak1 time kl-18o stay lOc college teacher's college Ak1 time kl-18o play
rugby when l was at college: teacher's college. when l played
rugb< m-e pisa8 pisa na aro-gu m-u ne vasangi viti koto
rugby cON}-!8o dislocate dislocate Ak1 shoulder-l8:18o cON}-18o lmll leel hurt very
rugby it dislocated, my shoulder dislocated and l really leel pain a lot
tena de. Ma na bona t-u gari vaolu m-e doku na thuli-gu ni bebeu g-u
lOc ulm and Ak1 time kl-18o child young cON}-!8o good Ak1 body-l8:18o olN play llc1-18o
now. And when l was a young boy my body was good lor play, l was
161
Number Nine is the common name ol the 8olomon lsland National kelerral lospital.
219
tangomana na bebeu. Mena deni u ghua thairo ki(kiki ba k-e viti.
able Ak1 play lOc ulm 18o cONu work kul-little.bit lL1 lkk-!8o hurt
able to play. lut now il l work a little bit it will hurt.
Tena deni u thudu le po ni bungu-ti-ra poa na gari vaolu t-ara ghali-a
lOc ulm 18o sit still llm olN watch-1k-O:!ll llm Ak1 child young kl-!ll do-O:!8o
Now l just sit still to just watch the young boys who are doing
na thairo pukua t-e viti na aro-gu. @eia po. Aoku.
Ak1 work because kl-!8o hurt Ak1 shoulder-l8:18o ll:!8o llm good
the work because my shoulder hurts. 1hat's all. Ol.
Yesterday
! nea te roropo muria na lotu mu dea i vae ni talua
i nea tena roropo muria na lotu m-u dea i vae ni talu-a
lOc yesterday lOc morning behind Ak1 worship cON}-18o go lOc house olN put-O:!8o
Yesterday morning, alter church, l went to [my] house to drop oll
nigua na bible. oko ma na tha t-u ghali-a u dea m-u vanga roropo.
l8:18o Ak1 bible linish cON} Ak1 kll kl-18o do-O:!8o 18o go cON}-18o eat morning
my lible. 1hen what l did [was] l went and l ate [the] morning [meal].
%anga roropo soko m-u dea i maketi, dea i maketi m-u dea pelu
eat morning linish cON}-18o go lOc market go lOc market cON}-18o go pay
[l] linished eating [the] morning [meal] and l went to market, [l] went to market and l
went [and] bought
kaleda te /. oko mu dea i vae mu dea tughu, mu ghonia
kaleda tena l soko m-u dea i vae m-u dea tughu m-u ghoni-a
calendar lOc l linish cON}-18o go lOc house cON}-18o go change cON}-18o make-O:!8o
a calendar lrom l. linishing [that] l went to [my] house to change and l prepared
na pou t-ami uru-vi-a tena rodo. oko m-ami lavi dea-i m-ami
Ak1 b.d.m. kl-1lX.ll dive-1k-O:!8o lOc night linish cON}-1lX.ll take go-O:!ll cON}-1lX.ll
the beche-de-mer we dove lor during [the] night. [when we] linished we took them and
we
kisi kutuni. Tavurake m-ami kuki-i m-ami pungu-i. oko mena T.,
cut belly then cON}-1lX.ll cook-O:!ll cON}-1lX.ll dry-O:!ll linish cON}-!8o:say 1
gutted them. 1hen we cooked them and we dried them. [when we] linished 1. said,
B@ari, k-amu mai thanga-u na turu-vaghini-a na vae-gu.C Aoku
kids lkk-2ll come help-O:18o Ak1 stand.up-cAL8-O:!8o Ak1 house-l8:18o good
ouys, come help me stand up my house." Ol,
220
m-ami-tu dea: inau, 0., 4.%., ami-tu dea8 m-ami-tu dea bosa vani-ra balu
cON}-1lX.ll-lAL go ll:18o l lV 1lX.ll-lAL go cON}-1lX.ll-lAL go tell uA1-O:!ll some
we lew went: l, l., l.V., we lew went, and we also said to some
gari laka, BMai k-a ba turu-vaghini-a na vae-a T.C 4a ngide. 9a bona
kid also come lkk-1lN.ll ulO stand.up-cAL8-O:!8o Ak1 house-l8:!8o 1 lL1 start Ak1 time
other kids, come, let's stand up 1.'s house." [1hat's how it] started. At that time
veghe ia ara rota ara kuki ghole, ara kuki raes, ara kuki viuvi ma na ti i
like ulm !ll busy !ll cook greens !ll cook rice !ll cook potato cON} Ak1 tea lOc
they were busy they were cooking greens, rice, potato, and tea at
vae-a T. ma vitu )., #., ma M., ma ira. Tavurake m-ami-tu dea
house-l8:!8o 1 cON} kin l l cON} m cON} ll:!ll then cON}-1lX.ll-lAL go
1.'s house the kin l., l., and m, and others. 1hen we lew went to
turu-vaghini-a na vae. 9a tha t-ami ghali-a ami-ke lavi leoa na tughuru,
stand.up-cAL8-O:!8o Ak1 house Ak1 kll kl-1lX.ll do-O:!8o 1lX.ll-7 take every Ak1 post
stand up the house. what we did [was] we took every post,
ami talu kolu-i, m-ami bere-ngi i vi na tughuru t-e tangomana
1lX.ll put together-O:!ll cON}-1lX.ll see-1k lOc where Ak1 post kl-!8o possible
we put them together, and we saw where [each] post could
turu-vaghini-a. oko m-e mai 9. laka m-ami maki mo na vae.
stand.up-cAL8-O:!8o linish cON}-!8o come N also cON}-1lX.ll mark more Ak1 house
stand up. Alter that N. came too and we [did] more marking [ol] the house.
Tighi agutu t-ami ghali-a ami lavi-a na maki ma vaghatha na vae
lirst thing kl-1lX.ll do-O:!8o 1lX.ll take-O:!8o Ak1 mark cON} measure Ak1 house
[1he] lirst thing we did [was] we took the measuring [tool] and measured the house
i vi k-e mono ga. M-ami vaghatha. %aghatha soko-a m-ami
lOc where lkk-!8o stay there cON}-1lX.ll measure measure linish-O:!8o cON}-1lX.ll
where it would be. And we measured. [laving] linished measuring
maki mo m-e poso-a m-ami lavi-a na spedi ma inau na tighi
mark more cON}-!8o straight-O:!8o cON}-1lX.ll take-O:!8o Ak1 spade cON} ll:18o Ak1 lirst
we marked more and it [was] straight and we took a spade and l [was] the lirst
mane t-u kabi. )abi-a sakai na gilu tavurake m-ara turu-vaghini-a
man kl-18o dig dig-O:!8o one Ak1 hole then cON}-!ll stand.up-cAL8-O:!8o
person to dig. [l] dug one hole and then they stood up
221
na tughuru m-e tu iti. oko-i e ruke na tuguru m-e kabi a 4.
Ak1 post cON}-!8o stand up linish-O:!ll !8o second Ak1 post cON}-!8o dig Ak1 l
a post and it stood up. [laving] linished those [things was] the second post and l. dug.
)abi a 4. m-e teigha na sara soko seghi-a m-e dea 4.%. m-e laua
dig Ak1 l cON}-!8o Nlo Ak1 arrive linish quick-O:!8o cON}-!8o go lV cON}-!8o grab
l. dug but he didn't arrive quickly [at] completion and l.V. went and grabbed
tulithia na spedi lime me kabi sokoa. Mami talu itia
tulithi-a na spedi lima-a me kabi soko-a m-ami talu iti-a
take.out-O:!8o Ak1 spade hand-l8:!8o cON}-!8o dig linish-O:!8o cON}-1lX.ll put up-O:!8o
the spade out ol his hand and he linished digging. we put up
sakai na talili i pala ghalagha. oko mena laka pala boko m-e vati
one Ak1 hall lOc side up.coast linish lOc also side down.coast cON}-!8o lour
one hall on the up-coast side. [laving] linished we stood up lour
na tughuru t-ami turu-vaghini-a. oko ami turu-vaghini-a laka
Ak1 post kl-1lX.ll stand.up-cAL8-O:!8o linish 1lX.ll stand.up-cAL8-O:!8o also
posts on the down-coast side also. [laving] linished we also stood up
i levugha e ruka, mi minim e ono, na tughuru. 9a bona veghe e para koto
lOc between !8o two l mean !8o six Ak1 post Ak1 time like !8o hot very
two posts in between, l mean six posts. At that time the sun was very
n' atho i nea te danikama mara thudu unga. Ma 9. po ma 4.%.
na atho i nea tena danikama m-ara thudu unga ma N po ma lV
Ak1 sun lOc yesterday lOc alternoon cON}-!ll sit shade cON} N llm cON} lV
hot yesterday in the alternoon and they sat [in the] shade. And just N. and l.V.
t-ara-ko iti dea ma 0. m-ara-tu iti dea ni tughi-a na kava vete oti
kl-!ll-uL up go cON} l cON}-!ll-lAL up go olN hammer-O:!8o Ak1 copper lLkl hold
they two went upand l.they lew went up and hammered the metal to hold
na tughuru. ?, na ghai-be kolu-a na tughuru. ! na soko m-u ghoni-a
Ak1 post oh Ak1 crossbeam with-O!8o Ak1 post lOc Ak1 linish cON}-18o make-O:!8o
the posts. Oh, the cross-beams with the posts. At the end l made
t-i na tughuru ni tu i levugha ni enoana ma na ghia-be tina.
kl-lOc Ak1 post olN stand lOc between olN jackpost cON} Ak1 crossbeam mother
the posts to stand in between the jackpost and the main cross-beam.
222
M-u ghoni soko-i m-ami-tu dea talu iti-i i-tu.ghami 4.%. ma 9. Aea
cON}-18o make linish-O:!ll cON}-1lX.ll-lAL go put up-O:!ll lAL-ll:1lX.ll lV cON} N go
l linished making them and we lew went [and] put them up, we lew l.V. and N. [and l].
m-ami-tu talu iti soko-i m-ami-tu thua-i soko-i mena soko
cON}-1lX.ll-lAL put up linish-O:!ll cON}-1lX.ll-lAL brace-O:!ll linish-O:!ll lOc linish
[As things] went we lew linished putting them up and we lew linished bracing them
and [it] [was] done
ga, m-ena T., B? k-amu sakai ni oli vae ba k-a-ke vanga.C M-ami
there cON}-!8o:say 1 oh lkk-2ll one olN return house lL1 lkk-1lN.ll-7 eat cON}-1lX.ll
there, and 1. said, Oh, belore you all return [to] [your] house we will all eat." we
thivo m-ara redim mo-a na vanga ). ma ira. Tavurake m-ara lavi
down cON}-!ll prepare more-O:!8o Ak1 lood l cON} ll:!ll then cON}-!ll take
[sat] down and they prepared the lood, l. and others. 1hen they took
thivo-i na vanga m-ami thudu t-i lighi ni vae-a M.*. tena kolova
down-O:!ll Ak1 lood cON}-1lX.ll sit kl-lOc beside olN house-l8:!8o mA lOc clover
down the lood and we sat beside m.A.'s house on the clover
m-ami vanga8 vanga m-ami inu ti. #avi soko-i na vanga lavi dea
cON}-1lX.ll eat eat cON}-1lX.ll drink tea take linish-O:!ll Ak1 lood take go
and we ate, [we] ate and we drank tea. [laving] linished taking (i.e., eating) the lood
[they] took it away
m-ara ghe kuki-a na ti m-ami-tu ghe inu ti. !nu ti soko
cON}-!ll continue cook-O:!8o Ak1 tea cON}-1lX.ll-lAL continue drink tea drink tea linish
and they continued to make tea and we lew continued to drink tea. [laving] linished
drinking tea
m-ami-tu thudu ni saluva m-ara dami tinoni dami, piala
cON}-1lX.ll-lAL sit olN talk cON}-!ll chew.betelnut people chew.betelnut smoke
we lew sat and talked, and those who chew betelnut chewed betelnut [and] those
who smoke
tinoni piala. oko m-e dea nulavi m-ami tovothi dea i nea.
people smoke linish cON}-!8o go alternoon cON}-1lX.ll separate go lOc yesterday
smoked. [when that was] done it [was] going [on] alternoon and we went [our]
separate [ways] yesterday.
M-u dea pogho i tathi. /ogho soko m-u dea pogho mami. oko m-u
cON}-18o go bathe lOc sea bathe linish cON}-18o go bathe lresh.water linish cON}-18o
And l went lor a swim in the ocean. [laving] swum l went[to have a] lresh-water bath.
[laving] linished l
22!
dea tughu m-u belo nulavi i nea. @eia po-a na tugu t-u
go change cON}-18o ring.bell alternoon lOc yesterday ll:!8o llm-O:!8o Ak1 story kl-18o
went to change and l rang the bell yesterday alternoon.
tangomana tugua tena bona deni. 0 tena tuaghai. @eia po. Tagiu.
tangomana tugu-a tena bona deni e teigha na tuaghai geia po tagiu
possible story-O:!8o lOc time ulm !8o Nlo Ak1 long ll:!8o llm thank.you
1hat just the story l'm able to tell at this time. lt's not long. 1hat's it. 1hank you.
1he turtle and the lish
162
Momono ti sakai na dani na vonu ti ma na igha. !koira di a sakai
mo-mono t-i sakai na dani na vonu t-i ma na igha iko-ira dini a sakai
kul-stay kl-lOc one Ak1 day Ak1 turtle kl-lOc and Ak1 lish uL-ll:!ll ulm Ak1 one
'1here was one day a turtle and a lish. 1hese two were
na udu la(laka. *ra-ko leta t-i vini vi-leu-thi8 na leoni na bona
Ak1 lriend kul-also !ll-uL not.know kl-lOc never klcl-light-O:!ll Ak1 all Ak1 time
also lriends. 1hey two never lought, all the time
ara-ko vi-doku-i po t-i levu m-i levu. Mo(mono mo a sakai na dani
!ll-uL klcl-good-O:!ll llm kl-lOc side cON}-lOc side kul-stay more Ak1 one Ak1 day
they two were just good to each other lrom side to side. 1here was a day,
ga, m-ara-ko ghunughunu-a t-i na visaghiri. Aoku m-e thai-a mo
there cON}-!ll-uL plan-O:!8o kl-lOc Ak1 contest good cON}-!8o arrive-O:!8o more
they two planned a contest. Ol, it arrived
na dani t-ara-ko ghunughunu-a m-ara-ko vuivuni-a na visaghiri. Aoku na igha t-i
Ak1 day kl-!ll-uL plan-O:!8o cON}-!ll-uL start-O:!8o Ak1 contest good Ak1 lish kl-lOc
the day they planned lor and they two started the contest. Ol, the lish
t-e vuivuni-a. ! vi ga t-e dea ga na igha e tumuri po na vonu.
kl-!8o start-O:!8o lOc where there kl-!8o go there Ak1 lish !8o lollow llm Ak1 turtle
started it. where the lish went the turtle lollowed.
0 dea i rughu ni vatu na igha m-e dea po na vonu. 0 dea ghua i namo
!8o go lOc beneath olN stone Ak1 lish cON}-!8o go llm Ak1 turtle !8o go cONu lOc deep
1he lish went beneath a stone and the turtle just went. ll the lish went extremely deep
tupi na igha m-e tumuri po na vonu. 0 na po t-i m-e dea kola po
extreme Ak1 lish cON}-!8o lollow llm Ak1 turtle !8o Ak1 llm kl-lOc cON}-!8o go tired llm
the turtle just lollowed. lt was just like this
162
1his story was presented to me in written lorm.
224
t-i na igha ga. M-ena po t-i, B? u boro laghini-gho i ghe
kl-lOc Ak1 lish there cON}-!8o.say llm kl-lOc oh 18o impossibe through-O:28o lOc contine
and the lish just got tired. lt just said, Oh, l can't outdo
ighoe ti tena de.C B? doku,C ena ti na vonu. Me atheathe kikiki
ighoe t-i tena deni o doku ena t-i na vonu m-e atheathe ki-kiki
ll:28o kl-lOc lOc ulm oh good !8o.say kl-lOc Ak1 turtle cON}-!8o rest kul-little.bit
you at this time." Oh good," said the turtle. 1he lish rested a little bit
po na igha ga m-ara-ko g-e sakai ni vuni-a po t-i na visaghiri.
llm Ak1 lish there cON}-!ll-uL llc1-!8o one olN start-O:!8o llm kl-lOc Ak1 contest
there belore they two started again the contest.
! vi ga t-e dea ga na vonu e tumuri po na igha. 0 dea ghua
lOc where there kl-!8o go there Ak1 turtle !8o lollow llm Ak1 lish !8o go cONu
wherever the turtle went the lish just lollowed. ll the turtle
i namo na vonu m-e tumuri po na igha. T-e dea i vi na vonu m-e
lOc deep Ak1 turtle cON}-!8o lollow llm Ak1 lish kl-!8o go lOc where Ak1 turtle cON}-!8o
went deep the lish just lollowed. 1he turtle went wherever and
tumuri po na igha. *ra-ko naghonamane m-ara-ko k-e dea thai-a a sakai
lollow llm Ak1 lish !ll-uL surprising cON}-!ll-uL lkk-!8o go arrive-O:!8o Ak1 one
the lish just lollowed. 1hey twoand this is surprisingthey two went
na kokomu ki(kiki, m-e tu na vonu m-e tagu iti tena maone tena
Ak1 island kul-little cON}-!8o stand Ak1 turtle cON}-!8o crawl up lOc sand lOc
arrived at a little island, and the turtle stood and crawled up on the sand on
kokomu ia. Ma na tha laka, e kou laka t-i na liguni-a ghini-a igha
island ulm and Ak1 kll also !8o unwilling also kl-lOc Ak1 pass-O:!8o lN81-O:!8o lish
the island. And what's more, being unwilling to be outdone by it [the turtle] this lish
deni m-e ghe dea laka po t-i tena maone. oko m-e ghe tu
ulm cON}-!8o continue go also llm kl-lOc lOc sand linish cON}-!8o continue stand
continued to also just go on the sand. 1hen the turtle continued and stood
na vonu m-e ghe tagu oli beti. 0 na po t-i ga m-e
Ak1 turtle cON}-!8o continue crawl return sea !8o Ak1 llm kl-lOc there cON}-!8o
and continued and crawled back to the sea. lt was just like this: it was
boro po t-i na ghe oli dea i beti na igha m-e soko mate po
impossible llm kl-lOc Ak1 continue return go lOc sea Ak1 lish cON}-!8o linish die llm
just impossible lor the lish to continue to go back to the sea and it just died.
22
t-i. M-e na po ga m-e soko laka po t-i na vatari deni.
kl-lOc cON}-!8o Ak1 llm there cON}-!8o linish also llm kl-lOc Ak1 example ulm
And that there is just the end also ol this parable.'
226
8OLkcl8
lender, Andrea, and 8ieghard leller. 200!. lolynesian tapu in the 'deontic square': a
cognitive concept, its linguistic expression and cultural context." ln /roceedings
of the T&ent<-Fifth *nnual Conference of the Cognitive cience ociet<, eds. kichard
Alterman and uavid lirsh, 1!!-1!8. mahwah, N}: lawrence lrlbaum. Viewed on
world wide web at: http:JJwww4.psychologie.uni-
lreiburg.deJsignaturesJbellerJpublicationsJsb0!cogsci1apu.pdl, 12 lebruary
2008.
llevins, }uliette. 200!. A Note on keduplication in lugotu and cheke lolo." ?ceanic
#inguistics 42.2:499-0.
comrie, lernard. 196. *spect. cambridge: cambridge Lniversity lress.
coombe, lruce 1aylor. 2002. @ela @rammar. Lnpublished ms.
crowley, 1erry. 2002a. oela orammar.C ln The ?ceanic #anguages, }ohn lynch, malcolm
koss and 1erry crowley, 2-!. kichmond, 8urrey: curzon lress.
______. 2002b. erial %erbs in ?ceanic: * Aescriptive T<polog<. Oxlord: Oxlord Lniversity
lress.
crystal, uavid. 200!. * Aictionar< of #inguistics and /honetics 3Fifth 0dition'. malden, mA:
llackwell lublishing.
uixon, k.m.w. 1988. * @rammar of 4oumaa FiQian. chicago: Lniversity ol chicago lress.
uonohue, mark. 1999. lrregularity and pronominal markedness: where lavouritism sets
in." ?ceanic #inguistics !8.2:409-420.
uryer, matthew 8. 200a. Noun lhrase 8tructure." ln #anguage T<polog< and <ntactic
Aescription, %olume !!: Comple: Constructions 3econd 0dition', ed. 1imothy 8hopen,
11-20. cambridge: cambridge Lniversity lress.
______. 200b. word Order." ln #anguage T<polog< and <ntactic Aescription, %olume !:
Clause tructure 3econd 0dition', ed. 1imothy 8hopen, 61-1!1. cambridge:
cambridge Lniversity lress.
oivn, 1almy. 1984. <nta:: a functional-t<pological introduction 3%olume ! and !!'.
lhiladelphia: }ohn lenjamins lublishing co.
oordon, kaymond o., }r. (ed.). 200. 0thnologue: #anguages of the Morld 3Fifteenth 0dition'.
uallas, 1X: 8ll lnternational.
22
laspelmath, martin. 2006. Terminolog< of Case. lor malchukov, A. & A. 8pencer (eds.)
lorthcoming. ,andbook of Case. Oxlord Lniversity lress. Viewed on world wide
web at: http:JJemail.eva.mpl.deJ-haspelmtJcase1erminology.pdl, 1 may 200.
______. 200 (to appear). /re-established categories don't e:istNconseOuences for language
description and t<polog<. Viewed on world wide web at:
http:JJemail.eva.mpg.deJ-haspelmtJlreestablished.pdl, 1 August 200.
laspelmath, martin and 1homas mller-lardey. 2001. %alence change. Viewed on world
wide web at: http:JJemail.eva.mpg.deJ-haspelmtJ200val.pdl, 20 August 200.
lockett, charles l. 198. * Course in Modern #inguistics. New York: 1he macmillan
company.
lvens, w. o. 19!!. A grammar ol the language ol lugotu, Ysabel lsland, 8olomon
lslands." 4ulletin of the chool of ?riental and *frican tudies .1:141-1.
______. 19!. A orammar ol the language ol llorida, lritish 8olomon lslands." 4ulletin
of the chool of ?riental tudies (Lniversity ol london) .4:10-1110. Viewed on
world wide web at:
http:JJwww.paradisec.org.auJlieldnotesJimage_viewer.htm7lollA101,!,1,8, 16
}anuary 2008.
leesing, koger m. 1988. Melanesian /idgin and the ?ceanic ubstrate. 8tanlord, cA:
8tanlord Lniversity lress.
lichtenberk, lrantisek. 199. Apprehensional epistemics." ln Modalit< in @rammar and
Aiscourse, eds. }oan lybee and 8uzanne lleischman, 29!-!2. Amsterdam: }ohn
lenjamins lublishing company.
lynch, }ohn. 19!. Verbal Aspects ol lossession in melanesian languages." ?ceanic
#inguistics 12:69-102.
lynch, }ohn, malcolm koss and 1erry crowley. 2002. The ?ceanic #anguages. kichmond,
8urrey: curzon lress.
maddieson, lan. 1984. /atterns of sounds. cambridge: cambridge Lniversity lress.
Nordlinger, kachel and louisa 8adler. 2004. 1ense beyond the Verb: lncoding clausal
tenseJaspectJmood on nominal dependents." 9atural #anguage R #inguistic
Theor< 22:9-641.
lalmer, lill. 200. /assive /ossession in ?ceanic. laper presented at conlerence on Oceanic
linguistics , Noumea, 2-6 }uly 200.
228
lawley, Andrew. 19!. 8ome lroblems in lroto-Oceanic orammar." ?ceanic #inguistics
12:10!-88.
layne, 1homas l. 199. Aescribing Morphos<nta:: a guide for field linguists. cambridge:
cambridge Lniversity lress.
______. 2006. 0:ploring #anguage tructure: a student's guide. cambridge: cambridge
Lniversity lress.
koss, malcolm u. 1988. /roto ?ceanic and the *ustronesian #anguages of Mestern Melanesia
(8eries c, No. 98. Australian National Lniversity). canberra: lacilic linguistics.
______. 1998. lroto Oceanic phonology and morphology." ln The le:icon of /roto
?ceanic: The culture and environment of ancestral ?ceanic societ<. ": Material culture,
eds. malcolm koss, Andrew lawley and meredith Osmond, 1-!. canberra:
lacilic linguistics.
______. 2004. 1he morphosyntactic 1ypology ol Oceanic languages." #anguage and
#inguistics .2:491-41.
8apir, ldward. 1921. #anguage. New York: larcourt, lrace, and co.
8imons, oary l. 19. The @haimuta #anguage of !nterior @uadalcanal. working lapers lor
the language Variation and limits to communication lroject, Number 8. cornell
Lniversity and 8ummer lnstitute ol linguistics.
8imons, linda. 1986. 1he lronouns ol 1o'abaita." ln /ronominal <stems, (continuum, ),
ed. Lrsula wiesemann, 21-!. 1bingen: ounter Narr Verlag.
8ymons, lruce. 198. * ketch of *spects of #engo @rammar. Lnpublished ms.
williams-van llinken, catharina, }ohn lajek and kachel Nordlinger. 2001. 8erial verbs
in 1etun-uili: a preliminary account." ln tudies in the #anguages and Cultures of
0ast Timor, Volume 4, eds. oeollrey lull and lance lccles, 48-60. 8ydney:
Academy ol last 1imor 8tudies. Viewed on world wide web at:
http:JJwww.linguistics.unimelb.edu.auJaboutJstall_docsJrachelJsv.pdl, 11
}anuary 2008.
229