Pacific Languages: An Introduction (OA)
Transcript of Pacific Languages: An Introduction (OA)
Pacific Languages
PacificLanguages
AN INTRODUCTION
John Lynch
copy 1998 University of Hawailsquoi PressAll rights reserved
Printed in the United States of America
98 99 00 01 02 03 5 4 3 2 1
Library of Congress Cataloging-in-Publication Data
Lynch JohnPacific languages an introduction [John Lynch]
p cmIncludes bibliographical references and indexISBN 0ndash8248ndash1898ndash9 (alk paper)1 Pacific AreamdashLanguages I Title
P381P3L96 1998499mdashdc21 97ndash24552
CIP
University of Hawailsquoi Press books are printed on acid-free paper and meet theguidelines for permanence and durability of the Council on Library Resources
Designed by Josie Herr
This book is licensed under the terms of Creative Commons Attribution-NonCommercial-NoDerivatives 40 International (CC BY-NC-ND 40) whichpermits readers to freely download and share the work in print or electronic for-mat for non-commercial purposes so long as credit is given to the authorDerivative works and commercial uses require permission from the publisherFor details see httpscreativecommonsorglicensesby-nc-nd40
The Creative Commons license described above does not apply to any materialthat is separately copyrighted Please refer to the credit lines and source nota-tions in the book to determine the copyright holders for images and other third-party material
The open-access ISBNs are 9780824842581 (PDF) and 9780824881832(EPUB)
The open-access editions of this publication were made possible by The Univer-sity of the South Pacific (USP) OER Course Conversion Grant This competitivefunding program is open to all USP faculty and staff wishing to integrate openmaterials into mainstream teaching and learning
To
AndoniaBrendan
andSteven
Contents
Illustrations x
Preface xiii
Acknowledgments xvii
Terms Used xix
Chapter 1 Linguistics Some Basic Concepts 111 The Structure of Language 112 Common Grammatical Categories and Functions 513 Reconstructing Linguistic History 9
Part 1 Geography and History 21
Chapter 2 The Languages of the Pacific 2321 How Many Languages 2322 Linguistic Demography 2723 Language Names 4024 A Brief History of Pacific Language Research 41
Chapter 3 The History of the Austronesian Languages 4531 The Austronesian Family 4532 The Oceanic Languages 46
VII
33 The Settlement of Oceania 5134 Reconstructing Culture 58
Chapter 4 The History of the Papuan and Australian Languages 6041 Interrelationships of Papuan Languages 6142 Interrelationships of Australian Languages 6843 Possible External Links 6944 Implications for Prehistory 69
Part 2 Structure 73
Chapter 5 Sound Systems 7551 Oceanic Languages 7552 Papuan Languages 8753 Australian Languages 9154 Orthographies 9455 Summary 99
Chapter 6 Oceanic Languages Grammatical Overview 10061 Pronouns 10062 Nouns and Noun Phrases 10563 Possessive Constructions 12264 Verbs and the Verb Complex 13065 Sentences 14866 Similarities and Differences 165
Chapter 7 Papuan Languages Grammatical Overview 16671 Pronouns 16672 Nouns and Noun Phrases 16973 Possessive Constructions 17174 Verbs and the Verb Complex 17275 Sentences 177
Chapter 8 Australian Languages Grammatical Overview 18581 Pronouns 18582 Nouns and Noun Phrases 18783 Possessive Constructions 19484 Verbs and the Verb Complex 19585 Sentences 198
VIII Contents
Part 3 The Social and Cultural Context 203
Chapter 9 Languages in Contact 20591 The Social Context of Language Contact 20592 The Linguistic Effects of Contact 20893 Three Case Studies 21394 Historical Implications 218
Chapter 10 Pidgins Creoles and Koines 220101 Pidginization Creolization and Koineization 220102 Melanesian Pidgin 221103 The Pidgins of the Motu Traders 232104 Fiji Hindi 235
Chapter 11 Language Society and Culture in the Pacific Context 237111 The Vocabulary of Pacific Languages 237112 Counting Systems 244114 Languages in Use 256115 Language Use in Pacific Nations 261116 Shift Survival Death Revival 268
Conclusion Ideas about Pacific Languages 272
Suggestions for Further Reading 279
Appendices 283
1 Data Sources 2852 Phonetic Symbols 2913 Sample Phoneme Systems 2954 Glossary of Technical Terms 299
Notes 313
References 321
Index 345
Contents IX
Illustrations
Tables
1 Pacific languages by region and country2 Polynesian Outliers3 Languages of Melanesia4 Papuan language families5 Anejom pronouns6 Baniata independent pronouns
Figures
1 Genetic relationship2 Subgroups of a family3 Subgroups of Proto East-Central Papuan4 Establishing migration patterns5 Time limits on comparative linguistics6 An Austronesian family tree7 The Polynesian subgroup8 Partial classification of Anejom terms for marine life9 A moiety system
10 Njamal kinship terms11 Anejom kinship terms
X
Maps
1 Reconstructing migration patterns2 The Pacific3 Languages of Polynesia4 Languages of Micronesia5 Languages of Fiji and Rotuma6 New Caledonia7 Vanuatu8 Solomon Islands9 Papua New Guinea
10 Irian Jaya11 Australia12 Austronesian languages13 Oceanic subgroups in Melanesia14 Austronesian migrations15 Papuan language families16 Possible wider groupings of Papuan families
Illustrations XI
Preface
This book was born out of frustration I have lost count of the number oftimes people have asked me to recommend to them a ldquogood general bookon the languages of the Pacificrdquo There are any number of good specialist ortechnical books on the Austronesian languages as a whole or on the Papuanlanguages or on Australian languages or on certain subgroups or individuallanguagesmdashbut virtually all of these are aimed at readers who have studieda considerable amount of linguistics
There are however many nonlinguists who want or need to know some-thing about the languages of this region Language is an important topic ofconversation an important political and social phenomenon in many if notall Pacific countries and territories Pacific peoples want to know more abouttheir languagesmdashwhat other languages they are related to where they camefrom how they compare with say English and French what the other lan-guages in the region are like People working in Pacific countries need somegeneral information on the languages of the country or the region to assistthem in their work and in their appreciation of the cultures and societiesof the Pacific Teachers sociologists community workers government offi-cers high school and university studentsmdashall are affected by language andmost would like to know more
Hence this book It has been a long time in the making but I hope thatit will serve a useful purpose I have tried to steer a middle course betweenbeing too simplistic and being too technical Obviously to provide detailedcoverage in any book of the sound systems and grammars of fourteen hun-dred languages their interrelationships and connections with languagesoutside the region their history and current status and the relationshipsbetween language culture and social organization is quite impossible
XIII
What I have tried to do is to give the general reader a feel for what theselanguages are like (with a minimum of references) and at the same timeoffer linguists something to get their teeth into (with references to sourcesthey can follow up)
The book has three major sections Part 1 describes the geographicaldistribution of Pacific languages and attempts to summarize what is knownof their history Part 2 is an overview of the phonological and grammaticalstructure of these languages This discussion is far from exhaustive Manyareas (eg complex sentences) and many thorny problems (eg the Polyne-sian ldquopassiverdquo) are omitted or glossed over But there is enough informationto give a general picture of what Pacific languages are like in what waysthey are similar and how they differ both from each other and from met-ropolitan languages like English or French Part 3 looks at the relationshipbetween Pacific societies and cultures and their languages from a number ofdifferent points of view In the Pacific as elsewhere language is very much asocial and cultural phenomenon
The careful reader will notice a bias toward Oceanic languages in part2 This results partly from my own professional background and partly fromthe fact that while there are good general surveys of Papuan and Australianlanguages (Foley 1986 and Dixon 1980) there is nothing comparable forOceanic languages
The orthography I use in citing language data is generally the standardorthography of the language For languages lacking such an orthographyI have used a standard set of phonetic symbols (see appendix 2) This hasoften meant modifying the orthography of the original sources Similarly Ihave consistently used the same name for the same language even whensome sources use different names
Phrase and sentence examples are presented as shown below
FijianE ā rai-ci irau na yalewa na cauravouhe PAST see- (TRANS) themtwo the woman the youth
lsquoThe young man saw the two womenrsquo
bull The first line in italics is the phrase or sentence in that languagewith hyphens marking morpheme breaks within a word underlin-ing is used to focus on the particular aspect of grammar beingdiscussed (in this example the transitive suffix -ci)
bull The second line is a word-by-word and morpheme-by-morphemetranslation Where a single morpheme expresses more than oneitem of meaning these are separated by a colon (thus irau lsquothe
XIV Preface
two of themrsquo is glossed as lsquothemtworsquo) Grammatical categories aregiven in capitals a few of these are abbreviated (thus TRANS = tran-sitive) and a list of all such abbreviations appears below I havetried not to be too technical with these grammatical terms andhave used for example ldquotherdquo or ldquowithrdquo rather than abbreviationslike ART (for article) and COM (for comitative) even if these aresometimes too general Readers interested in more technical as-pects of the grammars of any languages cited should consult thesources (appendix 1)
bull The last line in single quotation marks is the free English translation
I have tried to be consistent in my use of grammatical terms throughoutthe book even where this means using a different term from that in the orig-inal source So for example I consistently use ldquocontinuousrdquo even thoughsome writers may have used terms like ldquoprogressiverdquo or ldquodurativerdquo and Iuse ldquocompletiverdquo where others use ldquoperfectiverdquo I have used small capitalswhen a technical term is introduced for the first time in the text There is aglossary of such terms in appendix 4
I have generally not directly quoted sources of language data in the textsince this would unnecessarily clutter the text with references However alist of data sources for all languages from which data are cited can be foundin appendix 1 and the languagesrsquo locations are indicated on maps 3 through7 I have also provided suggestions for further reading at the end of thebook
Preface XV
Acknowledgments
This book originally appeared in 1993 as an in-house text for the Universityof the South Pacific course ldquoStructure of Pacific Languagesrdquo I am grateful tostudents for their feedback about this earlier version and to Donn BayardBarbara Haulsquoofa Andrew Horn Ross McKerras Macha Paris Mere PuleaJeff Siegel Matthew Spriggs Jan Tent Randy Thaman Howard Van TreaseVilimaina Vakaciwa and the Department of Geography of the University ofthe South Pacific who either commented on parts of the earlier version orprovided useful information
I take particular pleasure in expressing my gratitude to Niko BesnierTerry Crowley Ken Rehg Malcolm Ross and Nick Thieberger withoutwhose assistance this book would not be what it is I reserve of course allblame for errors and misinterpretations
I am also deeply indebted to a number of Pacific people who haveopened the doors of their languages to me I am especially grateful to JohnDavani Tom Hiua John Naupa Julie Piau Tulsquoa Taumoepeau-Tupou PhilipTepahae and Apenisa Tusulu to the people of Uje and Anelcauhat (Ane-ityum) and west Tanna in Vanuatu and to my tambus in Kond and Anigl inPapua New Guinea
My sons Brendan and Steven have lived with this book over the past fewyears and the backs of discarded drafts have been of great use in helping todevelop their artistic talents My wife Andonia has been a source of constantencouragement and I am eternally grateful for her love and support I ded-icate this book to them
XVII
Terms Used
The following abbreviations are used in the text
(a) marks a phrase or sentence as ungrammatical (b) marks aphoneme or word as having been reconstructed for someproto-language
oslash zero1 first person2 second person3 third personCS construct suffixDIFFSUBJ different subjectEXC exclusiveINC inclusiveNOM nominalizerO objectPL pluralPOSS possessiveS subjectSG singularSAMESUBJ same subjectTRANS transitiveV verb
XIX
CHAPTER
1
LinguisticsSome Basic Concepts
11 The Structure of LanguageLinguistics is the systematic study of language and descriptive lin-guistics is the branch of linguistics that deals with the analysis and descrip-tion of languages Each language is a system with various units and rulesfor the combination of these units into larger units These rules are not al-ways formulated in grammar books but they are there neverthelessmdashin thebrains of speakers of the language
One simple way of seeing the operation of these rules is through the mis-takes children make when they are learning a language When a four-year-old says The mans goed away the sentence is clearly incorrect English Itdoes however follow a pattern First the child has deduced that to make anoun plural in English you add s to it She has already produced large num-bers of plural nouns like dogs cats cups bananas and so on following thisrule Second she has also deduced that to put a verb into the past tenseyou add ed to it Again she has already produced many English verbs in thepast tense this waymdashlaughed cried kicked washed etc
In producing the sentence The mans goed away the child is not imitat-ing what adults say since no adult speaker of English would say that sen-tence Instead she is applying two of the many rules she has formulated onthe basis of observing how English is spoken
1 NOUN + s = PLURAL2 VERB + ed = PAST TENSE
The only problem is that the noun man happens to be an exception to rule(1) and the verb go an exception to rule (2) Looking at this ungrammatical
1
utterance gives us insight into how the childrsquos brain is functioning in termsof rules that combine units into larger units
What are these units I have been talking about If you asked a non-linguist that question the answer would probably be sounds words andsentences Unfortunately the situation is more complex than that
111 The Sounds of LanguageAt the ldquolowestrdquo level of language we have sounds which linguists enclose insquare brackets [ ] to distinguish them from letters Individual sounds like[t] [e] and [n] are meaningless in themselves Only combinations of soundsprovide meaningful utterances [t] + [e] + [n] = ten [n] + [e] + [t] = net
No language uses all the speech sounds human beings can make andthe sound systems of different languages are organized in different waysThe study of sounds is known as phonetics and the study of the way inwhich sounds are organized into a system in a language is called phonology(or sometimes phonemics) (A chart of all phonetic symbols used in thisbook appears in appendix 2)
Let us take as an example the sounds [p] (represented by the letters por pp) and [f] (represented by f or ff) These are quite different sounds butis the difference between them important In some languages for exampleEnglish it is as the pairs of words below show
pull fullpig figsupper suffercup cuff
The only difference in sound between the words in each pair is the differencebetween the sounds [p] and [f] but each word has a very different meaningIn English the sounds [p] and [f] belong to different phonemes that is theyare different significant units of sound in the language And linguists writephonemes in slant lines to distinguish them from both sounds and lettersThus English has the phonemes p and f
Compare the same two sounds in the Tok Pisin language of Papua NewGuinea
paia faia both mean lsquofirersquopasim fasim both mean lsquotiersquomipela mifela both mean lsquowersquolap laf both mean lsquolaughrsquo
2 CHAPTER 1
In this language the difference between [p] and [f] is not significant You canuse either sound without changing the meaning of a word In Tok Pisin [p]and [f] belong to the same phoneme usually written p The same sounds indifferent languages may therefore have quite different functions in the sys-tems in which they occur and quite different relationships with each other
Note that we are dealing with sounds and phonemes here not with theletters that are used to write them In the English words we looked at abovethe phoneme f is represented by the letter f in full as well as by the combi-nation ff in suffer The same phoneme f is also represented by ph in phoneby gh in enough and so on Our principal concern is with the sound systemsof Pacific languages though we will also look at their orthographies orwriting systems
112 The Composition of WordsPhonemes combine to form larger units Consider the following English ex-amples
actactedreactreacted
Each of these consists of a number of phonemes and each is also a word since ithas meaning by itself and in the written language appears with a space beforeandafterThesecondandthirdwordshowevercanalsobedivided into twomean-ingful parts act lsquocarry outrsquo + ed lsquopast tensersquo and re lsquobackrsquo + act The fourth wordconsistsof threemeaningfulparts re+act+ed
These smallest meaningful units are called morphemes Some single mor-phemes are words (act dog house desire for example) Other words (acted re-act reacted dogs housewife desirable for example) consist of multiple mor-phemes The study of morphemes and of the way morphemes combine to formwords isknownasmorphology a termalsousedtorefer to thepatternsbywhichmorphemescombine to formwords inaparticular language
The examples given above show one other feature of morphemes Whileact can stand on its own as a word (as a free morpheme) re and ed cannotMorphemes like re and ed are known as affixes and they must be attachedto another morpheme There are a number of different kinds of affixes themost common being prefixes which like re come before the root in a wordand suffixes which like ed come after the root The convention in linguis-tics is to write prefixes with a following hyphen (re-) and suffixes with apreceding hyphen (-ed) the hyphen indicating where the join takes place
Linguistics Some Basic Concepts 3
Another kind of affix occurs in some Pacific languages namely the infixwhich is placed within the root In Roviana (Solomon Islands) for exampleverbs are converted to nouns by inserting the infix -in- (note the hyphensboth before and after the infix) before the first vowel of the root
habu lsquoto fishrsquo hinabu lsquoa catch of fishrsquokera lsquoto singrsquo kinera lsquoa songrsquomoho lsquoto be sickrsquo minoho lsquosickness diseasersquotoa lsquoto be aliversquo tinoa lsquolifersquozama lsquoto talkrsquo zinama lsquolanguagersquo
When morphemes combine to form words the sounds at the boundariesof these morphemes often change For example I said above that the four-year-old had learned to form plurals by adding the suffix -s but this is notstrictly true The regular plural morpheme has two spellings and three orfour pronunciations in English The pronunciation of the letter s in pluralslike cats cups socks is indeed the phoneme s but the letter s of pluralslike dogs bugs homes is pronounced as the phoneme z not as s and thesame letter in plurals like inches buses dishes is pronounced IZ or əz de-pending on the dialect I also said that the child had learned to form the pasttense by adding -ed to verbs Again this is not strictly true The pronunci-ation of -ed is ɪdorəd in words like banded and slotted d in killed andconned and t in laughed and kissed
In these examples the sound at the end of the noun or verb determinesthe pronunciation of the plural or past-tense suffix The study of soundchanges that take place when morphemes combine to form words is knownas morphophonemics
113 Above the Word LevelWords combine to form phrases A phrase is a group of words that func-tions as a unit in a sentence Look at the following English sentence (where marks the boundary between phrases)
The young boys were killing the cats on the beach
Each of these phrases is a unit When each is moved to some other positionin the sentence it must be moved as a whole entity For example the passiveequivalent of the sentence above is
The cats were being killed by the young boys on the beach
and not something like
The young the cats were being killed by boys on the beach
4 CHAPTER 1
(The asterisk marks the sentence as ungrammatical) That is it is not justthe noun boys that moves in this change from active to passive but thewhole noun phrase the young boys
There are different types of phrases In this book I refer to noun phraseswhich are phrases that function like nouns and can be replaced by a single nounor a pronounmdashthe young boys and the cats in our sentence above are both nounphrases (and could be replaced for example by they and them) I also refer toprepositional phrases which are noun phrases introduced by a prepositionon the beach and by the young boys in the examples above are prepositionalphrases introduced by the prepositions on and by I use the term verb com-plex to refer to phrases that function like verbs were killing and were beingkilled in the sentences above are both verb complexes1
Phrases combine to form clauses A clause is a group of phrases containinga subject (the topic being talked about) and a predicate (what is being saidabout the topic) A sentence is a group of one or more clauses that can standalone If we return to our example of the cat-killing boys none of the following isa sentence since each requires other phrases to make it complete2
The young boysWere killing the catsThe young boys on the beach
English and many other languages usually require each predicate to con-tain a verb complex so that a sentence must have at least one verb Manylanguages of the Pacific however do not require this since in these languagesthere is no verb equivalent to English be (with its various forms is are etc) Sofor example English demands the verb be in equational sentences like That manis a doctor but many Pacific languages have no verb in equivalent sentences Inthe Lenakel language of Vanuatu for example the same sentence would be Wusaan tokta literally lsquoman that doctorrsquo with no verb
12 Common Grammatical Categories and Functions121 Subject and ObjectThe terms subject and object traditionally refer to the performer and re-ceiver of the action of the verb respectively In the sentence The boy ispetting the pig the performer of the action the boy is called the subjectand the receiver the pig is the object In many languages the verb changeswith a change of subject In the sentence The boys are petting the pig theplurality of the subject the boys causes the verb to change from singular (ispetting) to plural (are petting)
Linguistics Some Basic Concepts 5
This fact is important because the subject is not always the performerof the action Look at these sentences
The boy likes the pigThe boy was bitten by the pig
In these sentences the boy is still the subject because we can see the samekinds of changes in the verb when the boy becomes plural
The boys like the pigThe boys were bitten by the pig
In the second case however the boy is not performing the action The pig isperforming the action on the boy
In other languages the subject and the object behave in ways differentfrom the way in which English subjects and objects behave and we cannotgive a universal definition of these concepts But the subject often performsthe action and the object usually receives it
122 Transitivity and VoiceA sentence that contains no object is intransitive while one that does con-tain an object is transitive Examples
Intransitive Mele is eatingThe dogs are sleeping
Transitive Mele is eating a bananaThe dogs chased the children away
An active sentencemdasha sentence in the active voicemdashis one in which thesubject performs the action or where the object has the action performed onit A passive sentence is one in which the action is performed on the subjectFor example
Active Mele ate the bananaThe men cut down the tree
Passive The banana was eaten by MeleThe tree was cut down
123 Adjectives and VerbsMany Pacific languages do not distinguish between adjectives and verbs in thesame way English does The distinction in English is related to the existence ofthe verb be In English an adjectivemdashlike good for examplemdashcan either pre-
6 CHAPTER 1
cede the noun it describes or follow the verb be (or similar verbs like seem orappear) as in A good chief looks after his people and Our chief isseems good
In many Pacific languages however adjectives belong to a class ofstative verbs verbs that indicate a state rather than an action In Fijian forexample a verb is marked as stative by one of a number of markers (eg elsquothird person singular subjectrsquo) In the first sentence below the verb is kanalsquoeatrsquo and the word levu lsquobigrsquo follows the noun it modifies vuaka lsquopigrsquo
E kana na vuaka leva oqōlsquoThis big pig is eatingrsquo it eats the pig big this
In the next sentence the word levu lsquobigrsquo behaves like a verb that is just askana lsquoeatrsquo does in the sentence above
E levu na vuaka oqōlsquoThis pig is bigrsquo it big the pig this
A stative sentence is an intransitive sentence expressing a state rather thanan action Thus while Mele is eating expresses an action Mele is fat or Meleis a teacher express a state
124 Person Number and GenderIn English we are used to distinguishing first second and third personpronouns as well as subject object and possessive forms Both nouns andpronouns occur in singular and plural and in some cases they have mascu-line feminine or neuter gender The English subject object and possessivepronouns illustrate this
Singular PluralFirst person I me my we us ourSecond person you your you yourThird person
Masculine he him his they them theirFeminine she her they them theirNeuter it its they them their
Pacific languages differ in a number of ways from the English model
1 Most Pacific languages do not show gender in pronouns Raroton-gan (Cook Islands) ia or Fijian o koya or Anejom (Vanuatu) aen allmean lsquohersquo lsquoshersquo and lsquoitrsquo
2 A large number of Pacific languages distinguish two types of first per-son pronouns Inclusive first person pronouns refer to the speakerand the addressee(s) Exclusive first person pronouns refer to thespeaker and some other person(s) but not the addressee(s) In Bis-
Linguistics Some Basic Concepts 7
lama the national language of Vanuatu for example yumi is the firstperson inclusive pronoun (lsquoI + yoursquo) while mifala is the first personexclusive pronoun (lsquoI + hesheitthem [not you]rsquo)
3 Many Pacific languages distinguish more than two numbers the mostcommon (apart from singular and plural) being the dual numberwhich refers to two and only two the trial number referring to threeand only three and the paucal number used for a few (three to six orso) or to a small group that is part of a much larger one
The function of the plural changes depending on how many num-bers a language recognizes In a language with a singular a dual atrial or a paucal and a plural the role of the plural is much smallerthan it is in a language with only a singular and a plural In Fijian forexample we have o koya lsquohesheitrsquo (singular) o irau lsquothey tworsquo (dual)o ira-tou lsquothey (a few)rsquo (paucal) and o ira lsquothey (many)rsquo (plural)
4 Many Pacific languages have separate object and possessive formsof the pronoun as English does But in addition and unlike Englishmany also distinguish between an independent pronoun and a sub-ject pronoun The independent pronoun can be used as an answerto a question and may be used as a subject but when it is it isusually emphatic In Lenakel for example in is the third personsingular independent pronoun and r- is the corresponding subjectpronoun The sentences In r-am-apul and R-am-apul both mean lsquoHeshe is asleeprsquo But while the second one is a neutral statement thefirst emphasizes that it is he or she not someone else who is asleep
125 Possessives and ClassifiersIn languages like English there is usually only one kind of possessive con-struction No matter what the possessed noun refers to or what the posses-sorrsquos relationship is to that noun the same construction is used my handmy father my house my dog are all possessed in the same way by means ofthe possessive my
Now look at translations of those four phrases in Motu (spoken aroundPort Moresby in Papua New Guinea) in which the suffix -gu translates lsquomyrsquoThe nouns are ima tama ruma and sisia
ima-gu lsquomy handrsquotama-gu lsquomy fatherrsquoe-gu ruma lsquomy housersquoe-gu sisia lsquomy dogrsquo
Here we can see that there are two different constructions The words forlsquohandrsquo and lsquofatherrsquo attach -gu directly to the noun I call this type a direct
8 CHAPTER 1
possessive construction The words for lsquohousersquo and lsquodogrsquo do not attach -gudirectly to the noun but attach it instead to the morpheme e- and this word(e-gu) precedes the noun This I call an indirect possessive construction
In one way or another most Pacific languages distinguish two types ofpossessive constructions to which different linguists have given differentlabels and which have different semantics These two types could be classi-fied as follows
bull Close or subordinate or inalienable possession is often manifestedby direct constructions This involves the possession of something overwhich the possessor has no control and which cannot (normally) be ac-quired or disposed of It may be an integral part of the possessor (like ahand) or a relative (we cannot control who our father is)
bull Remote or dominant or alienable possession is frequently mani-fested by indirect constructions This involves the possession ofsomething over which the possessor has control It can be acquiredand disposed of given away or sold like a house or a dog
Some languages are more complex than this using a system of classifiersoften in both possession and counting to show what type of thing the noun isjust as in English we normally do not say ten cattle or four breads but ten headof cattle or four loaves of bread using head and loaf as kinds of classifiers Lookat the following examples from Ponapean (spoken in Pohnpei Micronesia)
kene-i-mahiediblething-my breadfruitlsquomy breadfruitrsquonime-i uhpwdrinkingthing-my coconutlsquomy drinking coconutrsquosehu pah-sopsugarcane four-stalklsquofour stalks of sugarcanersquo
Ponapean has more than twenty possessive classifiers (like kene- and nime-above) and approximately thirty numeral classifiers (like -sop above)
13 Reconstructing Linguistic History131 Genetic RelationshipAll languages change The process of change is gradual but it is also con-stant There are various kinds of evidence for this For example earlierwritten records show a version of the language different from the modern
Linguistics Some Basic Concepts 9
version though both are often still recognizable as the ldquosamerdquo languageThe two examples given below of the beginning of the Lordrsquos Prayer in theEnglish of about 1400 and in modern English illustrate this principle
Oure fadir that art in heuenes halowid be thi name thi kyngdom come tobe thi wille don in erthe es in heueneOur Father who is in heaven may your name be kept holy May your king-dom come into being May your will be followed on earth just as it is inheaven
Even if a language does not have written records going back a longtime the fact that people of different generations speak the same languageslightly differently shows that languages change We can even observechanges taking place in a language when we notice competing forms likethe two different pronunciations of a word like either in English (one with aninitial vowel sound like that of niece and the other with a vowel like that ofnice) or the past tense of the verb divemdashdived and dovemdashin many dialects ofAmerican English Perhaps the most obvious example of language changehowever is the continual introduction of new words into all languages (andless obvious but also quite frequent the gradual loss of words that for onereason or another have become obsolete)
Imagine now that we have a single speech community speaking a lan-guage we will call X This community splits into four separate groups AB C and D Because language change is inevitable and continuous after afew hundred years these four communities would speak different dialects ofthe same language3 But after a thousand years or more these four dialectswould have changed so much that they had become separate languages asshown in figure 1 The languages would share many similarities in vocabu-lary and grammar since language change is relatively slow But a speakerof language A would have considerable difficulty in holding a conversationwith a speaker of B C or D
Languages A B C and D in figure 1 are all genetically related toeach other because they all descend from language X which is their com-mon ancestor Languages A B C and D are often referred to as daughterlanguages of X and all four languages belong to the same language fam-ily Figure 1 which represents their relationship is their family tree
Where there are historical records of the ancestor language and of thewhole period of change it is easy to establish the relationship between thedaughter languages and to see how diversification took place But in thePacific as in many other parts of the world such records do not go back any-where near far enough for us to have concrete proof of diversification andrelationship How then do linguists establish such languagesrsquo relationship
Related languages share a number of similarities in vocabulary pronun-
10 CHAPTER 1
Figure 1 Genetic Relationship
ciation and grammar Linguists look for similarities between various lan-guages and if the similarities are numerous enough they assume that thelanguages involved are related despite the absence of documentary proofand derive from a hypothesized common ancestor which is referred to as aprotolanguage
But not all similarities between languages can be attributed to genetic re-lationship There are two other possible explanations One is that the similar-ities are purely accidental In Motu Fijian and many other Pacific languagesthe word for lsquoeyersquo is mata while in Modern Greek the word for lsquoeyersquo is matiThis however is a purely accidental resemblance as there are no other con-nections between Greek and Motu or Fijian If two or more languages shareonly a few similarities these are probably coincidental It is virtually impossi-ble however that languages could accidentally have hundreds of similarities
The second explanation for similarities between languages is copyingor borrowingmdashthat a language has adopted a word (or some other linguisticfeature) from some other language For example in many Pacific languagesthe word for lsquoradiorsquo is something like retio or ledio This word has beencopied from English but this does not mean that these languages are relatedeither to English or to each other
Copying is a very common phenomenon in all languages (see chapter 9)When new items of technology new social practices or new ideas are intro-duced into a society from outside often the words for them modified to fitlocal pronunciation will be brought in at the same time English is full ofwords copied from other languages Algebra boomerang coup demonstra-tor ghetto junta taboo thug and yen are just a few examples
Copying is more likely to take place in certain areas of the lexicon thanin others For example words like snow coconut ice cream church teamand television could be easily introduced into a language since they representthings or concepts that are by no means found in all cultures or environments
Linguistics Some Basic Concepts 11
But words like hand leg one two black white eat sleep are much less likelyto be taken from another language since all languages probably have theirown words for these concepts irrespective of the culture of their speakers orthe physical environment in which they live There would be no need for a lan-guage to supplement its vocabulary by borrowing them For similar reasonscertain aspects of grammar (the morphological structure of words for exam-ple) are less likely to be borrowed than others (like word order)
If similarities between two languages are only in areas where we mightexpect to see copying they do not constitute evidence of genetic relation-ship If however the similarities are in areas of vocabulary and grammarwhere borrowing is much less likely to take place we can reasonably con-clude that these are not due to chance or borrowing but to genetic in-heritance The words and structures were present in some form in an an-cestor language and have been retained usually in a modified form in thedaughter languages This then leads to the conclusion that the languagessharing these similarities are related belong to the same language familyand derive from the same protolanguage
132 Reconstructing a ProtolanguageIn addition to being able to show with reasonable confidence that a setof languages are related and derive from the same common ancestorhistorical-comparative linguists can reconstruct what many of the soundswords and grammatical structures in the protolanguage were probably like
An important principle in reconstruction especially in dealing with simi-larities in vocabulary is that of the regularity of sound correspondencesLook at the following examples from the Aroma Hula and Sinagoro lan-guages spoken on the coast east of Port Moresby in Papua New Guinea
Aroma Hula Sinagorolsquofatherrsquo ama ama tamalsquomilkrsquo laa laa latalsquosewrsquo uli uli tulilsquograndparentrsquo upu upu tubulsquosagorsquo lapia lapia labialsquopigeonrsquo pune pune punelsquoskinrsquo opi kopi kopilsquobirdrsquo manu manu manulsquomosquitorsquo nemo nemo nemo
There are a number of correspondences between identical phonemesAroma m corresponds to Hula m and Sinagoro m This correspondence is
12 CHAPTER 1
abbreviated as mmm We can also see all the vowels (aaa iii and soon) But there are also some correspondences between different phonemesFirst although we have the set ppp (as in pune pune pune lsquopigeonrsquo) wealso have another set ppb (as in lapia lapia labia lsquosagorsquo) Then we alsohave the set θθt (where θ represents the absence of a sound) as in uli ulituli lsquosewrsquo The important thing about both types of correspondence sets isthat they are regular They are not random but occur again and again inmany words Even in the short list above you can see a number of examplesof each
In the case of correspondence sets of the type mmm the original lan-guage almost certainly had m and the daughter languages have not alteredit The protolanguage then had a phoneme m where the asterisk denotesa reconstructed form
In the case of correspondence sets of the type ppp and ppb howeverone or more daughter languages has changed The logical assumption hereis that the set ppp reflects an original p while the set ppb represents anoriginal b which Aroma and Hula have changed to p The merger of pho-netically similar phonemes is a very common phenomenon and this is whatseems to have happened The distinction between the two phonemes p and bhas been lost in these two languages (in the same way as the distinction be-tween the voiced w in witch and the voiceless w in which is being lost in mostvarieties of English) Similarly the set θθt probably represents an earliert which has been lost in Aroma and Hula again loss of a phoneme is farmore common and natural than the addition of a phoneme
Using this principle of regularity of correspondence and also makinguse of what linguists know generally about language change it is possibleto reconstruct elements of a protolanguagemdashto make an educated guessabout what the phonemes words and grammar of the ancestor languagemight have been Given that Aroma nemo Hula nemo and Sinagoro nemoall mean lsquomosquitorsquo for example and that the correspondences nnn eeemmm and ooo are regular linguists would reconstruct the word nemolsquomosquitorsquo in the language ancestral to these three languages The full set ofprotoforms for the words given above would be
tama lsquofatherrsquolata lsquomilkrsquotuli lsquosewrsquotubu lsquograndparentrsquolabia lsquosagorsquopune lsquopigeonrsquo
Linguistics Some Basic Concepts 13
kopi lsquoskinrsquomanu lsquobirdrsquonemo lsquomosquitorsquo
133 Families and SubgroupsThe original split of a community may be followed by later splits Similarlythe original split of a protolanguage may be followed by subsequent splitsin intermediate ancestral languages sometimes called interstage lan-guages Look at the family tree in figure 2 which represents the followinghistorical sequence of events
First the original ancestral language X initially split into three daugh-ter languages P Q and R Some time later (1) language P suffered sufficientdivisions to result in the modern languages A and B (2) language Q splitinto Z and the modern language C (3) language Z itself underwent a furthersplit into the modern languages D and E and (4) language R split givingrise to the modern languages F G and H
All of these languages are related since they all derive from a commonancestor X There are differing however degrees of relationship in this fam-ily tree For example languages A and B are more closely related to eachother than either is to any other modern member of the family because theyshare a period of common development that the other languages do notmdashtheperiod when language P was separated from the others Similarly languagesF G and H are more closely related to each other than to any other modernmember of the family Languages C D and E can also be grouped together
Figure 2 Subgroups of a Family
14 CHAPTER 1
but within the group D and E are more closely related to each other than ei-ther is to language C
Linguists generally use the term subgroup to refer to two or more lan-guages within a family that are more closely related to each other than tothe rest of the family In figure 2 A and B form one subgroup and F G and Hanother C D and E make up a third subgroup within which exists a furtherlower-level subgroup (sometimes called a subsubgroup) D and E
When the history of a language family is known through written recordsthe subgrouping of languages within that family can also usually be estab-lished by examining those records But how do we determine subgroups ofa language family in an area like the Pacific where written records of lan-guages either do not exist at all or date only from recent times
One technique for doing this is known as lexicostatistics This involvesthe comparison of the basic vocabulary of the languages we are interestedin (using a standard one-hundred-or two-hundred-word list) and expressingthe degree of relationship between any two languages in the sample as apercentage which represents the cognates (similar vocabulary items pre-sumed to derive from the same original word in the protolanguage) sharedby each pair of languages A higher percentage corresponds to a closer rela-tionship and members of subgroups should show the highest percentages
Lexicostatistics has the advantage of allowing quick formulation and quan-tification of the internal relationships of a language family but it also has manyproblemsSomeof theseare theoretical ormethodological andneednotconcernus here One obvious problem however is that a list of even two hundred wordsrepresents only an extremely small part of a whole language and the figures ob-tained from comparing such lists may not accurately represent the relationshipbetween two languages Today most linguists do not rely heavily on lexicostatis-tics as a method for subgrouping languages although they might use it to get apreliminary indication of the possible subgrouping
The chief method linguists use to establish subgroups is examination ofshared innovations If you go back to the Aroma Hula and Sinagoro exam-ples in the last section you will see that two changes or innovations havetaken place (1) original t has been lost in both Aroma and Hula (but not inSinagoro) and (2) the distinction between original b and p has been re-tained in Sinagoro but it has been lost in both Aroma and Hula where thesetwo phonemes merge as the single phoneme p
Aroma and Hula share two innovations that Sinagoro does not whichwould suggest that the two languages are more closely related to each otherthan either is to Sinagoro The family tree in figure 3 shows how these threedescendants of Proto East-Central Papuan are related
Rather than suggesting that Aroma and Hula both quite independently
Linguistics Some Basic Concepts 15
Figure 3 Subgroups of Proto East-Central Papuan
made the changes t gt θ b gt p inferred from a comparison of cognates itseems logical to assume that the changes happened only once in the inter-stage language Proto Aroma-Hula In this way Aroma and Hula came toshare two innovations missing in Sinagoro which suggests that they belongtogether in a subgroup
There are various kinds of innovations which if shared by two or morelanguages exclusive of others in the family can be solid evidence for assign-ing those languages to the same subgroup Phonological innovations (likethe example above) and innovations in morphology are fairly strong evi-dence innovations in vocabulary and syntax (sentence structure) are lessstrong since changes take place in vocabulary much more easily and rapidlythan in phonology or morphology Quantity is also a factor Generally speak-ing if languages share more innovations (of the stronger kind) the hypothe-sis that they form a subgroup is more secure
134 Reconstructing Linguistic and Cultural HistoryWhat use can linguists and others can make of the conclusions reachedabout the relationships between languages the subgroups of a languagefamily and the reconstructed protolanguage
The branch of linguistics I have been discussing is known as com-parative linguistics or historical-comparative linguistics It involvescomparing languages in order to find out something about their historyThis branch of linguistics is one of the disciplines contributing to the studyof prehistory the time preceding the existence of written records (Othersuch disciplines include archaeology social anthropology the study of oralliterature and oral traditions and so on) So what can comparative linguis-tics tell us about prehistory
16 CHAPTER 1
First the fact that languages are related implies that they have a com-mon origin This often (though not always) implies that the people whospeak those languages have a common origin as well telling us somethingabout the origins of and historical connections between the peoples of aregion
Second information about subgroupings can give us an idea of thechronology of language divisions (and presumably also divisions in a com-munity) as well as providing indications about the directions in which peo-ple migrated As an example of this let us consider just the following Pacificlanguages Fijian Tongan Pukapuka (spoken in the Cook Islands) Tahitianand Rapanui (Easter Island) A simple family tree for just these five lan-guages would look like the one in figure 4
The most recent split in this family (which includes hundreds of otherlanguages) is that between Tahitian and Rapanui with the next most recentthat between Pukapuka and the ancestor of Tahitian and Rapanui Some-what earlier Tongan and ldquoProto Pukapuka-Tahitian-Rapanuirdquo divided andthe first split was between Fijian and all the other languages As you can seeby looking at map 1 the splits proceeded from west to east
On the basis of this subgrouping most linguists would assume (1) thatthe original homeland of this group of people was probably somewherearound the Fiji-Tonga area and (2) that the general direction of migrationof these peoples was probably from west to east as shown in map 1 Notethat I have used the terms ldquoassumerdquo ldquoprobablyrdquo and ldquosomewhererdquo Theseconclusions are merely the best educated guesses we can make from thedata We would still want to find supporting evidence from other disci-plinesmdasharchaeological dates oral traditions or the likemdashbefore adoptingthese conclusions firmly
Third comparative linguistics can tell us something about the cultureof the people who spoke the protolanguage and about the changes thathave taken place in that culture If a set of words can be reconstructed for
Figure 4 Establishing Migration Patterns
Linguistics Some Basic Concepts 17
Map 1 Reconstructing Migration Patterns
a protolanguage the items or concepts they refer to were probably alsopresent in the protoculture For example if we could reconstruct for a pro-tolanguage words for taro yam coconut and breadfruit then we couldpresume that these items were in the original culture of the people whospoke that language And if the daughter languages have quite unrelatedwords for peanut rice coffee and sweet potato then we could assume thatthese items were not in the original culture but represent later innovationsThe identification of copied words can also tell us quite a bit about anotheraspect of linguistic and social historymdashcultural contact between groups ofpeople speaking (related or unrelated) languages
135 Time DepthsFinally a word of warning The principles and techniques of comparative lin-guistics allow linguists to trace relationships between languages going backperhaps eight or ten thousand years and to make associated conclusionsregarding migrations cultures and so on If however the initial breakupof a language family took place longer ago than about ten thousand yearslinguists often cannot find sufficient evidence to prove that the languages in-volved are related The changes that have taken place in each language overthe millennia are usually so great that very few similarities can be distin-guished or reconstructed
The hypothetical family tree in figure 5 helps illustrate this point Thesimilarities currently existing between the modern languages P through Ywould probably lead comparative linguists to divide them into four unre-lated families
18 CHAPTER 1
Figure 5 Time Limits on Comparative Linguistics
1 the C family with members P and Q2 the D family with members R S and T3 the E family with members U and V and4 the F family with members W X and Y
The true historical picture is presented in the diagram which showshow all these languages are related deriving from a common ancestor XBecause of the length of time involved however the changes have been sogreat that most similarities between say languages P and Y have been lostwhich is why linguists would treat these languages as belonging to four dis-tinct families
The study of prehistory relies heavily on comparative linguistics formany different kinds of information But it is important also to realizethatmdashat least with the techniques currently at our disposalmdashcomparativelinguistics has limitations
Linguistics Some Basic Concepts 19
PART ONE
Geography and History
CHAPTER
2
The Languagesof the Pacific
When different people speak of the Pacific region they often mean differentthings In some senses people from such Pacific Rim countries as Japan andKorea Canada and the United States and Colombia and Peru are as mucha part of the region as are those from Papua New Guinea Fiji the MarshallIslands Tonga and so on In this book however I use the term ldquothe Pacificrdquoto refer to the island countries and territories of the Pacific Basin includingAustralia and New Zealand
This Pacific has traditionally been divided into four regions MelanesiaMicronesia Polynesia and Australia (see map 2) Australia is clearly sepa-rate from the remainder of the Pacific culturally ethnically and linguisti-cally The other three regions are just as clearly not separate from oneanother according to all of these criteria There is considerable ethnic cul-tural and linguistic diversity within each of these regions and the bound-aries usually drawn between them do not necessarily coincide with clearphysical cultural or linguistic differences These regions and the bound-aries drawn between them are largely artifacts of the western propensityeven weakness for classification as the continuing and quite futile debateover whether Fijians are Polynesians or Melanesians illustrates
Having said this however I will nevertheless continue to use the termsldquoMelanesiardquo ldquoMicronesiardquo and ldquoPolynesiardquo to refer to different geographi-cal areas within the Pacific basin without prejudice to the relationships ofthe languages or the cultures of people of each region
21 How Many LanguagesThis book deals mainly with the indigenous languages of the Pacific regionThere are many other languages that can be called ldquoPacific languagesrdquo for
23
Map
2Th
ePa
cific
24 CHAPTER 2
example languages that have developed fairly recently such as HawaiianCreole Fiji Hindi Hiri Motu (Papua New Guinea) Melanesian Pidgin(known as Tok Pisin in Papua New Guinea Pijin in Solomon Islands and Bis-lama in Vanuatu) Broken and Kriol (Australia) and others There are alsothe languages of the metropolitan powers particularly English and Frenchwhich are widely used throughout the region but also Bahasa Indonesia inIrian Jaya and Spanish in Easter Island And there are small but substantialnumbers of speakers of various Chinese languages of Vietnamese and ofother ldquointrusiverdquo languages in Pacific towns (These languages receive someattention in part 3)
When it comes to what we might call ldquotruerdquo Pacific languages we findthat this region is probably the most linguistically complex in the worldThere are or were almost fourteen hundred distinct languages spoken inthe Pacific or about one quarter of the worldrsquos languages And these four-teen hundred languages are spoken by not much more than 01 percent ofthe worldrsquos population1 Further so far as we can tell these languages donot all belong to a single language family There are a number of languagefamilies in the Pacific
Let us look first at the nature of the differences between languagesin this region Many people describe the languages of the Pacific as ldquodi-alectsrdquo partly because most are spoken by small populations and areunimportant in terms of world politics and partly because many are un-written But linguists use the terms ldquolanguagerdquo and ldquodialectrdquo with quitespecific meanings
Speakers of the same language living in geographically separateareas often speak differently though these differences are usually notgreat enough to prevent communication between them For examplemany Americans say sidewalk diaper and flashlight where English peo-ple would use footpath nappy and torch And while most English peoplepronounce words like half past and mast with the same vowel as thefirst vowel in father most Americans pronounce them with the samevowel as in hat But despite these obvious differences in vocabulary inpronunciation and in grammar as well the Americans and the Englishcan still communicate quite easily We would therefore say that they arespeaking different dialects of the same language But Americans or Eng-lish people must learn French to understand a French person as Englishand French are different languages
Mutual intelligibilitymdashwhether speakers from one group can or cannotcarry on a normal conversation with speakers of anothermdashis just one way oflooking at the distinction between language and dialect In many parts of thePacific it is difficult to test for mutual intelligibility because people not only
The Languages of the Pacific 25
speak the language of their own community but also acquire an understand-ing either active or passive of the languages of neighboring communitiesfrom a very early age People from two communities can quite often carry ona conversation in two different languages so testing for mutual intelligibil-ity is fraught with all sorts of problems In cases like these linguists have touse their own judgment about how many languages are involved
Perhaps more important than the issue of mutual intelligibility is the is-sue of social identity People believe that their language is the same asmdashor isdifferent frommdashanother grouprsquos language for a variety of social rather thanlinguistic reasons Here are two examples of this
1 On the basis of mutual intelligibility Hindi and Urdu would beclassified as dialects of the same language Hindi is the nationallanguage of India It is written in the Devanagari script and isclosely associated in peoplersquos minds with Hinduism Urdu is thenational language of Pakistan It is written in Arabic script andis closely associated with Islam For these nonlinguistic reasonsmost speakers would say that Hindi and Urdu are two different lan-guages
2 Many people refer to Fijian as if it were one language It is associ-ated with a group of people who are ethnically and culturally fairlyhomogeneous and there is just one written version which all lit-erate Fijians read and write But people in the eastern part of Fijicannot understand people from the western area when they speak(unless they have learned the western Fijian language)
There is a further problem with differentiating and counting languagesthat relates to the phenomenon known as a dialect chain A dialect chainis found in a series of communities in which each community has a differentdialect Close neighbors can quite easily understand each other but peoplehave greater difficulty in understanding or communicating with people fromcommunities farther along the chain Imagine that the following villages arespread along the coast of a large island
A B C D E F G H I J
People from say village C can easily communicate with their close neigh-bors (A and B to the west D and E to the east) they have some difficultycommunicating with people from F and G and they cannot communicatewell at all with people from H I and J On the other hand people fromvillage E can communicate easily with those from C D F and G havesome difficulty with those from B H and I but find people from A and Junintelligible People from A would be unable to communicate with those
26 CHAPTER 2
from J so it would seem from looking just at the two ends of the chainthat two different languages are involved But there is nowhere in themiddle of the chain where we can draw a language boundary since every-one can communicate with their immediate neighbors So are we dealingwith one language or two
In one sense this is really a problem only when one tries to count thenumber of languages to tidy up the situation with a neat classificationSome linguists would say that the villages I have described share one lan-guage made up of a complex dialect chain Others would say it is two withhowever no distinct boundary between the western language and the east-ern one Situations like this are found in Fiji in the Caroline Islands ofMicronesia and in a number of areas in Papua New Guinea This is one rea-son that different authorities give different numbers of languages for certainareas of the Pacific
Despite these complications when I say that there are about fourteenhundred languages spoken in the Pacific I do mean languages not dialectsSome of course are quite similar to each other as French is to Spanish andItalian or even as Hindi is to Urdu But there are also differences of thesame order of magnitude as those between English and Chinese And manyof these languages are spoken in a number of dialects as well
There are two other reasons why we cannot be exact about the numberof languages in the Pacific Some languages are moribundmdashthat is at lastreport they were spoken by just a small number of old peoplemdashand there-fore are almost extinct Many Australian languages fall into this categorybut there are some in Melanesia as well The other reason is that at leastin certain parts of the Pacific we have insufficient information The interiorof Irian Java is an especially good example though not the only one In suchcases we are forced to make educated guesses
Table 1 gives the number of languages spoken in each of the main re-gions of the Pacific and in each of the countries and territories within eachregion For the reasons discussed above the figures given are approximate
22 Linguistic Demography221 Polynesia and MicronesiaWith a few exceptions we can say that in Polynesia there is generally onelanguage per island or per island group Ignoring minor problems (ldquoAre theytwo languages or two dialectsrdquo) there are twenty-one languages spoken inwhat is referred to as the Polynesian Triangle (including the extinct Moriorilanguage)2 Map 3 shows the location of all these languages
The Languages of the Pacific 27
Table 1 Pacific Languages by Region and Country
Melanesia 1151+Irian Jaya 205+Papua New Guinea 750+Solomon Islands 63Vanuatu 105New Caledonia 28
Micronesia 16Belau 1Northern Marianas and Guam 2a
Marshall Islands 1Kiribati 1Nauru 1Federated States of Micronesia 11
Fiji and Polynesia 22Fiji including Rotuma 3Tonga 2Niue 1The Samoas 1Tuvalu 1Tokelau 1Wallis and Futuna 2Cook Islands 3Hawairsquoi 1French Polynesia 5Easter Island 1New Zealand 1
Australia 200b
Total 1389+aOne of these is a dialect of Carolinian other dialects of which are spoken inthe Federated States of MicronesiabMany of these have become extinct or are moribund
Speakers of many of these languages now live outside their home coun-tries There are significant communities of speakers of for example EastUvea (Wallisian) in New Caledonia and Vanuatu and of Tongan and Samoanin both New Zealand and the United States About as many Rarotongan
28 CHAPTER 2
The Languages of the Pacific 29
speakers and many more speakers of Niuean live in New Zealand as in theCook Islands and Niue respectively
Some Polynesian languages have large numbers of speakers Samoanprobably has about 250000 speakers Tongan Tahitian and New ZealandMāori each approximately 100000 Rarotongan with more than 30000speakers and Wallisian with 10000 are also large in Pacific terms In con-trast some of the languages of French Polynesia other than Tahitian arespoken by fewer than a thousand people
Micronesia is similar to Polynesia in havingmdashas a rulemdashonly one languageper island or island group although there are difficulties in deciding exactlyhow many languages there are Bender and Wang (1985 54ndash56) have a goodbrief discussion of this problem While many of the speech traditions of Microne-sia are clearly identifiable as discrete languages the Trukic group of speechcommunities extending from Chuuk (Truk) Lagoon to Tobi presents a majorproblem Different linguists have divided this complex continuum into threeseven and eleven distinct languages which makes the exercise of counting lan-guages difficult and probably futile I have taken Bender and Wangrsquos figure ofthree languages for this continuum and this gives the somewhat arbitrary figureof sixteen languages spoken in Micronesia Map 4 shows the location of theselanguages but also indicates the named varieties of the three Trukic languagesthat some linguists treat as distinct
Many speakers of Micronesian languages also live outside their home coun-tries particularly in Guam and the United States Fiji Nauru and Solomon Is-lands possess sizable Kiribati-speaking communities Kiribati and Chamorroeach with more than 50000 speakers have the greatest number of speakers inMicronesia Lagoon Trukese Ponapean and Marshallese all have about 20000speakers and most of the other languages (depending on how they are defined)number in the thousands A number of languages or dialects how-evermdashincluding Sonsorolese Satawalese Namonuito Ngatikese Kapingama-rangi and Nukuoromdashhave fewer than a thousand speakers
222 MelanesiaFor the purposes of this discussion Melanesia is taken as including the in-dependent states of Papua New Guinea Solomon Islands Vanuatu Fiji theIndonesian province of Irian Jaya and the French overseas territory of NewCaledonia Melanesia differs from Polynesia and Micronesia here it is therule rather than the exception for there to be many languages per island Inthis general survey of the linguistic situation in Melanesia maps 5 through10 locate all the languages of Melanesia mentioned in this book
30 CHAPTER 2
The Languages of the Pacific 31
Fiji and Rotuma
Rotuma is home to a distinct language spoken by around 10000 peoplebut linguists disagree about how many languages are spoken in the rest ofFiji Certainly there are many different varieties of ldquoFijianrdquo spoken by the300000 or so ethnic Fijians in Viti Levu Vanua Levu and the offshore is-lands The situation is further complicated by the fact that the dialect of the
Map 5 Languages of Fiji and Rotuma
32 CHAPTER 2
island of Bau southeast of Viti Levu has been adopted as the standard forthe written language for education and for certain public occasions so thatmany Fijians who speak another dialect also know that one ldquoFijianrdquo consistsof a chain of perhaps thirty or forty dialects Most linguists would probablydivide this chain into two languages Western Fijian (spoken in the westernhalf of Viti Levu) and Eastern Fijian (spoken in the rest of the country ex-cluding Rotuma)
New Caledonia and the Loyalty Islands
There are twenty-eight languages in the French territory of New Caledoniaall spoken by small populations The two languages with the largest numberof speakers are Drehu with about 7000 speakers and Paicicirc with just under5000 but five of the territoryrsquos twenty-eight languages have fewer than twohundred speakers (Map 6 shows only those languages that I mention in thisbook)
Map 6 New Caledonia (showing languages referred to in the text)
The Languages of the Pacific 33
Vanuatu
The Republic of Vanuatu is home to between one hundred and 110 lan-guages (Tryon 1976) As in New Caledonia all of these are spoken by verysmall populations Recent estimates (Tryon and Charpentier 1989) showthat Northeast Ambae on Ambae Island with 4300 speakers Lenakel andWhitesands on Tanna each with 4000 and Apma on Pentecost with 3800have the largest number of speakers3 Forty-one languages or almost halfthe languages of the country have two hundred speakers or fewer five ofthese forty-one have fewer than fifty speakers (Map 7 shows only those lan-guages mentioned in the text)
Solomon Islands
The most recent linguistic survey of Solomon Islands (Tryon and Hackman1983) lists sixty-three languages as being spoken in that country Those withthe largest populations are the North Malaita dialect chain with 13500 andKwararsquoae with 12500 both on Malaita No other language has more than10000 speakers Twelve languages have fewer than two hundred speakerssix of these twelve have fewer than fifty (Map 8 includes only those lan-guages discussed in the text)
Papua New Guinea
Papua New Guinea is probably the most linguistically diverse nation inthe modern world A population of around four million people speak wellover seven hundred distinct languages Wurm and Hattorirsquos (1981) lin-guistic atlas of the region estimates that there are 750 languages spokenin Papua New Guinea This may be a slightly conservative figure Otherestimates usually count more than these Some differences lie in the dis-tinctions made between dialect and language Map 9 shows only a fewof these languages
According to Wurm and Hattorirsquos figures in the 1970s nine of the lan-guages of Papua New Guinea were spoken by more than 40000 people Allof these except Tolai are spoken in the Highlands These languages are
Enga 165000 Huli 60000Kuman (Simbu) 140000 Kewa 48000Hagen 100000 Mendi 45000Kamano 85000 Wahgi 45000Tolai 65000
At the same time a staggering 114 languages in Papua New Guinea arelisted as being spoken by populations of fewer than two hundred people
34 CHAPTER 2
Map 7 Vanuatu (showing languages referred to in the text)
The Languages of the Pacific 35
36 CHAPTER 2
The Languages of the Pacific 37
Irian Jaya
The situation in Irian Jaya is less clear than anywhere else in the Melanesianregion since much less research has been done on these languages thanon those of any other part of the Pacific Wurm and Hattori (1981) believethat slightly more than two hundred languages are spoken in this Indonesianprovince only four of them by 40000 people or more These four are
Western Dani 100000Grand Valley Dani 75000Ekagi 65000Biak-Numfor 40000
In contrast Wurm and Hattori list forty languagesmdash20 percent of thosein the provincemdashas being spoken by two hundred or fewer people (Map10 names only the languages mentioned in this book)
Map 10 Irian Jaya (showing languages referred to in the text)
38 CHAPTER 2
223 Australia
Dixon (1980) says that at the time of first European settlement there wereprobably about two hundred different languages spoken in Australia We willnever know the exact figure since many languages had disappeared beforeany linguistic work had been carried out on them Of these two hundred theWestern Desert language had the largest number of speakers around 6000It was spoken over an area of about 13 million square kilometers4
The survival of Australian languages (and of the people who speak them)has been severely threatened in the last two centuries Whole tribes andtheir languages died out in many areas while other tribes assimilated tovarying degrees to the invading culture losing their languages in theprocess Of the current language situation in Australia Dixon says
Of the 200 languages spoken in Australia before the Europeaninvasion 50 are now extinct the last speakers having died someyears ago in most cases there are still some people who wouldclaim tribal membership but they know only a dozen or so isolatedwords of what was once a full and flourishing language Then there
Map 11 Australia (showing languages referred to in the text)
The Languages of the Pacific 39
are probably around 100 languages that are on the path towardsextinction Some are remembered by only a handful of old peo-ple and will cease to be spoken or remembered within a very fewyears others may be being learnt by children in a few families butthe total number of speakers is so smallmdasha few dozen or lessmdashthatthese languages seem bound gradually to drop out of use over thenext few generations Finally perhaps 50 languages are in a rel-atively healthy statemdashspoken as first language by a few hundred(or in one or two cases by a few thousand) people and preservingtheir full range of use in everyday affairs and in ceremony and rit-ual (Dixon 1980 18)
While the languages of the rest of the Pacific region are generally quiteviable the Australian languages which once spread right across the conti-nent are in rapid decline The number of speakers of each diminishes withthe shift toward English and the decimation of the population
23 Language NamesLanguage names in the Pacific can be problematic Some languages areknown by one and only one name It may be the peoplersquos own name for thelanguage (Nakanamanga in Vanuatu) the name of the people themselves(Motu in Papua New Guinea) an English version of a local name (Tongan)or a compound expression referring to some feature of the language (Pitjan-tjatjara the name of a dialect of the Western Desert language of Australiawhich means ldquohaving the word pitjantja lsquocomersquordquo)
In some areas however people do not have a name for their own lan-guage but refer to it as ldquothe languagerdquo ldquoour languagerdquo or ldquocorrectgoodlanguagerdquo The Tolai language of New Britain for example has been re-ferred to in the literature as Tuna from a tinata tuna lsquothe real languagersquoLanguages of this kind are often named in the literature after the locality inwhich they are spoken For example linguists call the languages spoken onthe islands of Paama and Mota in Vanuatu Paamese and Mota
People sometimes invent names for languages lacking a specificappellation Discussing the names of some languages of the Torricelli Phy-lum in the Sepik area of Papua New Guinea Laycock (1975b 774) says ldquoThelanguages are named when not after a village or area by the translationinto that language of no or there u none this practice has been widespreadin the Lumi area for some time and may antedate European contact and theprinciple has been extended in naming languages outside the Lumi areardquoThis practice explains why a number of languages in this area have very sim-ilar names (Olo Alu Galu Aru Aruop and so on)
40 CHAPTER 2
Even when people do have their own name for a language some othername is often given usually a geographical one On the island of Tannain Vanuatu for example the ldquorealrdquo names of the languages spoken in theLenakel and Whitesands areas are respectively Netvaar and Nɨrak Butthese languages are almost universally known as Lenakel and Whitesandsnot only to outsiders but also to their speakers
In many cases the same language goes by a number of different namesa name in the local language and a geographical name or a series of namesfor different dialects or different localities in the language area for exampleThe language spoken (in two dialects) on the islands of Rennell and Bellonain Solomon Islands is known variously as Rennell-Bellona Rennellese Bel-lonese Moiki Munggava and Munggiki The Nakanamanga language ofcentral Vanuatu is perhaps better known to linguists as Nguna or Ngunesewhich is the dialect that has received the most attention5
Again there are cases where names may refer only to different dialectsWest Guadalcanal (Solomon Islands) for example has a number of nameddialects some of which appear in the linguistic literature as if they were sep-arate languages (Gari or Ghari Kerebuto Nggae Sughu and Vaturanga)Early mission grammars or dictionaries often named the language afterthe location of the mission while the name in current use is different(Lamalanga [name assigned by missionaries] for Raga spoken in PentecostIsland in Vanuatu) Hyphenated language names (eg Mono-Alu in SolomonIslands) can indicate that there are (at least) two named dialects but no over-all local name for the language Spelling variations also occur The name ofthe Baniata language of Rendova in Solomon Islands has also been spelledBantildeata and Mbaniata while another Solomons language spoken in NewGeorgia has been variously spelled Bareke Bariki Mbareke and Mbariki
In this book I try to use the most generally accepted name for any lan-guage with consistency even if (1) the language has other names and (2)these other names are used in my sources
24 A Brief History of Pacific Language ResearchThe first information on Pacific languages came from European navigatorswho published lists of words and occasional sentences in various languages(and sometimes commented on the similarities between some of them) Mis-sionaries followed translating religious materials into various Pacific lan-guages but also producing grammars dictionaries and the like Some colo-nial government officials also made contributions
Professional linguists were rather late on the scene In general their in-terests have been threefold
The Languages of the Pacific 41
1 Comparative-historical attempting to establish relationships be-tween languages within the Pacific and between Pacific languagesand languages outside the region thus contributing to the study ofPacific prehistory
2 Descriptive analysis of the grammars of Pacific languagescompilation of dictionaries and so on
3 Theoretical testing or formulating general claims about the natureof language and of language change on the basis of data from Pa-cific languages
241 Fiji and PolynesiaOur knowledge of the Fijian and Polynesian languages is more completethan our knowledge of most other Pacific languages for a number of reasonsThere is usually only one language per country (or island) The languagesare not especially difficult phonologically and are quite closely related sothat a knowledge of one makes a good stepping-stone to learning anotherAnd in general the Polynesian languages and Fijian have been studied forfar longer than have those of the rest of the region
Missionary endeavors and the work of some colonial officials provideda firm foundation for the description of many of these languages with agood number of grammatical studies and dictionaries being written in thenineteenth and early twentieth centuries The first grammar and the firstdictionary of Fijian for example were published in 1850 (Hazlewood 1850a1850b) and there were also early studies of the languages of Tonga Samoaand various parts of eastern Polynesia including New Zealand In many ofthe countries of Polynesia governments have also taken a keen interest inthe preservation of traditional culture and language encouraging the useof Polynesian languages in schools and churches on radio and television inbooks and newspapers and elsewhere in the public domain So there aregood grammars andor dictionaries for most of the languages of Fiji andPolynesia and there are numerous publications in and on these languagesof both an academic and a general nature
242 MicronesiaGiven Micronesiarsquos checkered colonial history it is not surprising that littlewas known about most of its languages until after the Second World WarSome of the early information on Micronesian languages was written in Ger-man or Japanese
Bender (1984 viiindashx) gives a brief summary of the history of Microne-
42 CHAPTER 2
sian linguistics since 1945 Initial studies focused on applied linguistics toassist the American government in education and other areas But thesestudies often had a more academic side as well The decision in 1966 to sendPeace Corps volunteers to Micronesia meant that language courses had tobe written providing a fresh impetus for linguistic research These languagelessons often developed into full-scale grammars and dictionaries mainlyunder the auspices of the University of Hawairsquoi which continues to be themajor center for the study of Micronesian languages
As a result of the last fifty yearsrsquo research good grammars or dictio-naries exist for most Micronesian languages Orthographies have been de-veloped for virtually all the languages and many are or have been used asclassroom languages in Micronesian educational systems
243 MelanesiaIn Melanesia some languages have been well known to linguists for a longtime but a very large number remain almost completely unstudied Apartfrom a few wordlists published by early explorers it was once again the mis-sionaries who undertook the first serious study of any of the Melanesianlanguages For many of these languages missionary grammars and dictio-naries (in French German or Dutch as well as English) remain the onlypublications of a linguistic nature By the turn of the twentieth century therewere publications on a handful of these languages including the compara-tive studies of von der Gabelentz (1861ndash1873) Codrington (1885) and Ray(1926) which presented grammatical sketches of a number of languagesBut even into the 1920s very little indeed had been published about the lan-guages of Melanesia
During the twentieth century missionary linguistic work has continuedin anglophone Melanesia Scholars from various universities have also pub-lished grammatical and lexical studies of a number of Melanesian languageswhile the Summer Institute of Linguistics has engaged in a massive amountof research into languages of the New Guinea area especially Until recentlythe pioneering work of Leenhardt (1946) remained the major source of infor-mation for the languages of francophone Melanesia though recent work by anumber of French and other linguists has dramatically increased our knowl-edge of the languages of New Caledonia and the Loyalty Islands
244 AustraliaApart from a few missionaries and colonial officials very few of the earlywhite settlers paid much attention to Australian languages Given their atti-
The Languages of the Pacific 43
tudes toward Aboriginal people and Aboriginal society which ranged fromclassifying them as primitive attempting to assimilate them and treatingthem with ldquobenign neglectrdquo to downright extermination and genocide onewould not have expected much linguistic work to be done on these lan-guages in the first century of contact
In the earlier part of the twentieth century some linguistic study accom-panied anthropological studies In his survey of the languages of AustraliaDixon notes that in the fifty years between 1910 and 1960 there was onlyone linguist Arthur Capell active in the field In more recent years linguistsfrom a number of universities in Australia and elsewhere as well as thoseworking with the Summer Institute of Linguistics have produced a consid-erable body of descriptive and comparative work Much of this falls into thecategory of salvage linguistics recording a language before it becomes ex-tinct Many salvage attempts are just sketches containing gaps in lexiconand grammar that can never be filled
44 CHAPTER 2
CHAPTER
3
The History of theAustronesian Languages
Comparative-historical linguists have divided the fourteen hundred or solanguages of the Pacific into three broad groups About 450 are classified asbelonging to the Austronesian family a very large family of languages withanother six or seven hundred members spoken outside the Pacific BasinSeven hundred or so languages spoken on the island of New Guinea or onislands not far from it belong to a number of apparently unrelated familiesAll are grouped under the cover term Papuan The two hundred Australianlanguages belong to a third broad genetic grouping We know much moreabout both the present and the past of the Austronesian languages of the Pa-cific than we do about the Papuan or Australian languages For this reason Idiscuss the history of the Austronesian languages first
31 The Austronesian FamilyThe Austronesian language family is one of the two largest language fami-lies in the world in number of member languages (The other is the Benue-Congo family in Africa) The family as a whole has somewhere between athousand and twelve hundred languages spoken by almost three hundredmillion people1 Map 12 shows the distribution of Austronesian languagesOutside the Pacific Basin Austronesian languages are spoken in Taiwan inMalaysia and a few small pockets on the Asian mainland in Madagascar andin almost all of island Southeast Asia All the languages of the Philippinesand almost all the languages of Indonesia (excluding most of Irian Jaya) areAustronesian
About 450 Austronesian languages are spoken within the Pacific regionThese include all the languages of Polynesia Micronesia Fiji New Caledo-nia and Vanuatu as well as almost all the languages of Solomon IslandsOnly about one quarter of the languages of the New Guinea area belong
45
Map 12 Austronesian Languages
to this family however Speakers of these languages generally occupy NewGuinearsquos offshore islands and some coastal areas but very few inhabit in-land areas
While linguists are still not in full agreement as to the major subgroupsof Austronesian figure 6 shows one widely accepted view of the higher-or-der branches of this family Nearly all of the Austronesian languages dis-cussed in this book belong to the Oceanic subgroup The family tree sug-gests an Asian origin for speakers of Austronesian and the archaeologicalevidence tends to corroborate this
32 The Oceanic LanguagesTwo languages spoken in Micronesia Palauan and Chamorro belong to oneof the Western Malayo-Polynesian subgroups of Austronesian and the Aus-tronesian languages of the western part of Irian Jaya belong to the SouthHalmaherandashWest New Guinea subgroup All of the other Austronesian lan-guages in the Pacific belong to the Oceanic subgroup This subgroup wasoriginally established by the German linguist Dempwolff (1934ndash1938) Hereferred to it as Urmelanesisch lsquoProto Melanesianrsquo All Oceanic languagesshare a number of phonological grammatical and lexical innovations thatare absent from the other Austronesian languages
321 Internal Relationships of the Oceanic LanguagesScholars have been debating the internal relationships of Oceanic for sometime They agree that the initial branching of Oceanic was in the western
46 CHAPTER 3
part of the Pacific but the poor state of our knowledge of Melanesian lan-guages has made it difficult to determine just what that initial branchinglooked like Fijian and the Polynesian languages have been thoroughly stud-ied for more than a century and their interrelationships are fairly clearThey form however only one small subsubgroup of Oceanic and studyingthem has not helped a great deal in determining the overall structure of theOceanic subgroup
Figure 6 An Austonesian Family Tree
The History of the Austronesian Languages 47
Only in fairly recent years has a coherent picture of the Oceanic sub-group begun to emerge Currently the groups within this subgroup include
1 Yapese spoken on the island of Yap in Micronesia (Ross 1995) Thismay prove to form part of the Admiralty Islands group
2 The Admiralty Islands group namely the languages of Manus andneighboring islands to the north of the New Guinea mainland
3 The Saint Matthias Islands group two languages spoken on smallislands immediately to the north of New Ireland in Papua NewGuinea This also may prove to be part of the Admiralty Islands group
4 The Western Oceanic group a very large grouping consisting ofa The North New Guinea subgroup comprising all the Oceanic
languages of western and southern New Britain plus those spo-ken along the northern coast of Papua New Guinea from justsouth of the Markham Valley westward to the Irian Jaya border
b The Papuan Tip subgroup all the Oceanic languages of thePapuan mainland and the neighboring islands
c The Meso-Melanesian subgroup made up of the Oceanic lan-guages of northern and eastern New Britain New IrelandBougainville (and their offshore islands) and the Oceanic lan-guages of the western half of the Solomon Islands (excluding ahandful of Polynesian Outlier languagesmdashsee 322 below)
d The Sarmi-Jayapura subgroup made up of the Oceanic lan-guages of the northeast coast of Irian Jaya (Ross 1996) (Theseare included here because they may turn out to be part of theNorth New Guinea subgroup)
5 The Southeast Solomons group includes the Oceanic languagesof Guadalcanal Malaita and Makira plus Bughotu on Isabel Thisgroup may possibly also include the languages of Utupua andVanikoro in the Temotu Province of Solomon Islands though it ismore likely that these form one or even two separate subgroups
6 The Southern Oceanic group (Lynch 1997) consisting ofa The North-Central Vanuatu subgroup in which are the non-
Polynesian languages of north and central Vanuatu from theTorres Islands in the north to Efate in the central south
b The Southern Melanesian subgroup with the non-Polynesianlanguages of Southern Vanuatu (Erromango Tanna and Ane-ityum) New Caledonia and the Loyalty Islands
7 The Micronesian group all non-Polynesian Oceanic languages ingeographical Micronesia excluding Yapese note that the status ofNauruan within this group is still problematic
48 CHAPTER 3
The History of the Austronesian Languages 49
8 The Central Pacific group consisting of Rotuman the languagesof Fiji and all Polynesian languages including the Polynesian Out-liers discussed below2
Attempts have been made to try to link two or more of these groupingstogether into a higher-order group but they have so far been unsuccessfulGroups 5ndash8 above have recently been linked into a putative Central-EasternOceanic subgroup (Lynch Ross and Crowley 1998) whose validity is still be-ing investigated Because of this trying to present a family tree of Oceanicwould serve no real purpose at this stage of our research
322 Oceanic Subgroups and Geographical RegionsGiven the subgrouping of Oceanic just outlined it should be obvious that theboundaries dividing the three traditional geographical-cultural regions of thePacificmdashMelanesia Micronesia and Polynesiamdashdo not correspond to the lin-guistic facts About twenty languages are spoken in the geographical areaknown as Polynesia Outside Polynesia are fourteen other languages that arevery clearly genetically ldquoPolynesianrdquo These are referred to as PolynesianOutliers and most scholars assume that they are the result of migrations intoMelanesia and Micronesia from western Polynesia after its settlement by theancestors of the modern Polynesians Table 2 gives a list with locations of thefourteen Polynesian Outliers (See also maps 4 6ndash9) Figure 7 shows the in-terrelationships of the Polynesian languages and their immediate relatives inthe Central Pacific group The primary split in Polynesian occurred betweenthe Tongic subgroup (consisting of just Tongan and Niuean) and the NuclearPolynesian subgroup (consisting of all other Polynesian languages includingthe Outliers) The closest Outliersrsquo relatives within Polynesian appear to beSamoan Tokelauan Tuvaluan East Uvea East Futuna Niuafolsquoousbquo and Puka-puka Although all the languages of Polynesia are Polynesian in the geneticsense not all Polynesian languages are spoken in Polynesia
In Micronesia the situation is somewhat different The ldquoMicronesianrdquo sub-group consists of most but not all of the languages of geographical Microne-sia Not only are two Polynesian Outliers Nukuoro and Kapingamarangi spo-ken in Micronesia but Yapese appears to be a single member of a subgroupseparate from all other Oceanic languages To complicate matters further thenature of the relationship of Nauruan to the other Micronesian languages isunclear and Palauan and Chamorro are not even Oceanic languages at all buthave as their closest relatives languages in Indonesia and the Philippines
Nowhere however is the mismatch between so-called cultural areasand linguistic classification more glaring than in Melanesia Hundreds of
50 CHAPTER 3
Table 2 Polynesian Outliers
Country or territory Location LanguageFederated States of Mi-cronesia
Nukuoro Island Nukuoro
Kapingamarangi Island KapingamarangiPapua New Guinea Nukuria Island Nukuria
Mortlock Island TakuuTasman Island Nukumanu
Solomon Islands Ontong Java LuangiuaStewart Island SikaianaRennell Island BellonaIsland
Rennellese
Duff Island PileniTikopia Island AnutaIsland
Tikopia-Anuta
Vanuatu Emae Island EmaePort Vila harbor Ifira-MeleFutuna Island AniwaIsland
West Futuna
New Caledonia Ouveacutea Loyalty Islands Fagauvea (WestUvea)
Papuan languages are spoken in Melanesia as are a number of Oceanic lan-guages including a dozen or so Polynesian Outliers (see table 3)
But more important is the fact that although we can speak of a Polyne-sian subgroup and even of a Micronesian subgroup that have some corre-lation with geography there is no such thing as a Melanesian subgroup ofOceanic Of the eight major subgroups of Oceanic six are located wholly orpartly in Melanesia
33 The Settlement of OceaniaLinguists construct hypotheses about the interrelationships of languages toattempt to find out about the past These theories about past languages andlanguage splits generally lead to theories about the origins and migrationsof peoples In many cases one can compare linguistic and archaeological hy-potheses in an effort to put both on a firmer footing
331 Origins of Oceanic SpeakersThe Oceanic subgrouprsquos position on the Austronesian family tree (figure 6)indicates that the speakers of Proto Oceanic migrated from Southeast Asia
The History of the Austronesian Languages 51
Figure 7 The Polynesian Subgroup
52 CHAPTER 3
Table 3 Languages of Melanesia
Austronesian Papuan TotalNew Caledonia 28 mdash 28Vanuatu 105 mdash 105Solomon Islands 56 7 63Papua New Guinea 210 540 750Irian Jaya 45 160 205Totals 444 707 1151
to the Pacific region This thesis is almost universally accepted Some evi-dence suggests that the closest external relatives of the Austronesian lan-guages may be (1) the Thai-Kadai group of languages spoken mainly inThailand and Laos and (2) the languages of the neighboring Austroasiaticgroup spoken mainly in Cambodia and Vietnam Both of these groups alsohave members in southern China and in parts of Malaysia Archaeologistssuspect that dramatic improvements in agricultural practices accompaniedby significant population growth led to expansions of human populations onthe Southeast Asian mainland around 5000 BC (Bellwood 1995)
The Austronesians were one of these populations The linguistic familytree presented in figure 6 is compatible with the archaeological evidencepointing to an Austronesian homeland on the Asian mainland The first no-ticeable expansion was into Taiwan and then after some centuries fromTaiwan to the Philippines Later some Austronesian speakers migrated toMalaysia Indonesia and Madagascar
The closest relatives of Oceanic are its immediate western neighborsin the Cenderawasih Bay area and the Halmahera Islands in western IrianJaya The immediate ancestors of the Proto Oceanic speakers migrated fromeastern Indonesia through western Irian Jaya into the Bismarck Archipelago(Manus New Britain and New Ireland) and settled theremdashpossibly aroundthe Willaumez Peninsula in New Britainmdashfor some time Map 14 gives someidea of the various migrations
332 The Dispersal of Oceanic SpeakersOceanic speakers were not the first to arrive in the New Guinea area speak-ers of Papuan languages had been there for a long time The New Britainarea for example has been settled for more than thirty thousand yearsand parts of the mainland of New Guinea for much longer even than thatContact between the original Papuan-speaking settlers and the invading
The History of the Austronesian Languages 53
Map 14 Austronesian Migrations
Austronesians must have been varied in nature In some situations the twogroups probably engaged in open warfare In others the relationship wouldhave been uneasy but not particularly hostile Yet others no doubt involvedtotal integration and intermarriage
Some speakers of Proto Oceanic and its early descendants limited theirsettlements moving slowly to settle the Admiralties and New Britain Otherswent farther Yap may have been settled from the Admiralties for exampleas were New Ireland and the western Solomons Oceanic speakers alsocrossed the Vitiaz Strait to reach the New Guinea mainland with one groupprogressively settling the north coast from east to west and another movinginto the Milne Bay area and the south coast
Some Oceanic speakers seem to have been more adventurous still Ifindeed they originated in the New Britain area they have left no tracethere but seem to have moved first southeast into the Solomons then southinto northern and central Vanuatu and north into Micronesia probably theKiribatindashMarshall Islands area from which location they settled the rest ofMicronesia There were also movements further south into southern Vanu-
54 CHAPTER 3
atu the Loyalty Islands and New Caledonia and further east to Fiji fromwhere Polynesia was settled Map 14 outlines these movements
We should be careful however not to think of all of these migrationsas major colonizing expeditions Spriggs (1995) for example suggests thatthere were probably initial long-distance scouting parties followed by morethan one movement of people along fairly well defined routes Back-migra-tions of some people also took place The migrations may have been deliber-ate as such factors as population pressure food shortages or political tur-moil forced people to seek somewhere else to live They may also have beenaccidental at least initially as fishermen were blown off course and endedup on new islands Many settlements succeeded but a great number nodoubt failed because of disease attacks by speakers of Papuan languagesand all kinds of other reasons
Such factors complicate the neat splitting of communities suggested byfamily treesOntheonehanda languagecommunitymaynothaveactually splitbut rather slowly diversified as contact between its subgroups became less andless intense On the other hand different related languages could have influ-enced each other blurring any innovations that might have been developing inone or another of them Speed of settlement is another complicating factor Inthe islands east of the Bismarcks no one seems to have stayed in one place longenough for telltale linguistic innovations to appear Under these circumstancesdefinitive higher-order subgroups of Oceanic are hard to establish
If the settlement of the Pacific proceeded in the direction and at thespeed discussed above it begins to make sense that Micronesians and Poly-nesians although originating in Melanesia nevertheless physically resem-ble their Southeast Asian ancestors more than they do Melanesians SomeOceanic speakers moved through Melanesia quickly enough to retain Asiangenetic features and these people ldquobecamerdquo Polynesians and MicronesiansOthers remained in Melanesia where centuries of intermarriage with thephysically different Papuan speakers have led to quite different genetic de-velopments (Pawley and Ross 1995 60)
333 Dating Dispersals The ldquoLapita PeoplerdquoTrying to establish a chronological framework for these migrations purelyon linguistic grounds is presentlymdashand will probably remainmdashimpossible Afamily tree provides relative datings of language divisions telling us thatone such split occurred before or after another but it does not offer anyabsolute dating
In the 1950s and 1960s linguists made an attempt to derive actual datesfrom lexicostatistical data3 (The term for this is glottochronology) Concrete
The History of the Austronesian Languages 55
dates for the breakup of Proto Indo-European and its various subgroups werefor example proposed Glottochronology however was strongly criticized bymany scholars not only because of some of the dates it generated but also be-cause of inherent weaknesses in its methodology and underlying assumptionsThe practice has been almost universally abandoned
But although there is no linguistic technique for determining absolutedates for divisions in protolanguages or for migrations linguists can try tomatch their relative dating sequences with archaeological evidence whichis on surer ground when it comes to absolute dating In the Oceanic regionthis cooperative enterprise has led to some interesting results
Archaeologists use the term Lapita to refer to a distinctive style of pot-tery (The name comes from a place in New Caledonia one of the first sitesexcavated with this pottery) The term Lapita culture refers to the cul-tural complex associated with this pottery style including the introductionof pigs dogs and chickens distinctive stone adzes and shell ornamentsthe development of larger villages and the intensification of agriculture(Spriggs 1995 116ndash118)
Lapita culture first appears in the archaeological record of the BismarckArchipelago about 1600 BC It seems to have reached as far as Vanuatu andNew Caledonia by about 1200 BC and Fiji and western Polynesia by about1000 BC In Vanuatu and islands farther south and east Lapita people werethe first settlers There is no evidence of any pre-Lapita people (Papuans orothers) in eastern Melanesia and Polynesia and this absence of competitionfor land would have made settlement much easier than it was farther northand west
This notion of a very rapid movement of people through island Melane-sia correlates very well with the linguistic subgrouping that I discussed inthe last section That is the fact that the Oceanic group seemed to havea number of first-order subgroups (a ldquoflatrdquo tree) rather than two or threesubgroups that themselves have only two or three subsubgroups suggestsfairly quick movement over a wide area Much slower settlement patternswould have allowed more time for distinctive innovations and would presenta more layered family tree with the eastern languages much lower down thetree than the western ones
Archaeologists tell us that the original Polynesians settled in the Samoa-Tonga area about 1000 BC remaining in that area for five hundred or evena thousand years At around the turn of the era some moved into easternPolynesia while others migrated to the western Pacific to establish the Out-liers By about AD 1000 all the major eastern Polynesian island groups hadbeen settled (Bellwood 1978 318)
In Micronesia there is evidence that the Mariana Islands may have been
56 CHAPTER 3
settled from Southeast Asia about 1000 BC The rest of Micronesia howeverappears to have been settled for only about two thousand yearsmdashprobablyby Lapita people from somewhere in Melanesia though neither linguisticsnor archaeology is able to tell us precisely where
Significantly less archaeological work has been done in the western andnorthern parts of the Pacific than in the eastern part so that linguists work-ing on the Polynesian languages who are dealing with a relatively shortperiod have reliable archaeological information with which to correlatetheir findings But those working in Melanesia and Micronesia have to dealwith a longer period of time much less archaeological information againstwhich to test their hypotheses and in some areas at least occupation bypre-Oceanic peoples
334 Rapid Diversification in MelanesiaMany linguists have commented on and tried to explain the much greaterdiversity exhibited by the Austronesian languages of Melanesia than by anyother part of the Austronesian family This is partly a function of time Aus-tronesian languages have had more time to change in Melanesia than inPolynesia or Micronesia and so appear less similar to each other But thereis more to the problemmdashand to Austronesian language historymdashthan thetime factor After all Austronesians have not been in Melanesia for count-less eons longer than they have been in Polynesia and Micronesia
Some of the Austronesian languages of Melanesia seem to havechanged more rapidly than others This is due in part to contact betweenAustronesian and Papuan languages Fairly clear evidence shows that somelanguages of the Oceanic subgroup have changed radically as a result ofcontact with Papuan languages Among the most radical are languages likeMagori and Maisin in Papua New Guinea where linguists have had difficultyin deciding whether or not they are Austronesian at all So the history ofthe Austronesian languages of Melanesiamdashespecially western Melanesiamdashiscomplicated by the fact that they not only neighbor Papuan languages buthave in many cases been in intimate contact with them
But certainly all of the major differences between Melanesian languagescannot be explained by Papuan contact Many of the more aberrant Oceaniclanguages in Melanesia like those of New Caledonia are far away from thenearest Papuan language Rapid change can be an internal matter as wellas an external one and many of the differences between languages in thisregion have come about without external influence The small scale of manyMelanesian societies can allow changes to spread more quickly than theymight in larger societies although smallness does not cause rapid change
The History of the Austronesian Languages 57
The notion of the emblematic function (Grace 1981) of language inMelanesia is an important one to mention here Linguistic differences canbe important as badges of membership in a particular social group and peo-ple often focus on these differences as markers of in-group or out-groupstatus (in the same way that young people in many societies use slang ex-pressions to mark their in-group status) In Melanesia especially differencesbetween neighboring languages may have been exaggeratedmdasheven man-ufacturedmdashin order to preserve this emblematic function Such a processleads to more rapid diversification than normally expected
34 Reconstructing CultureMuch of the effort of comparative-historical linguists has gone into the re-construction of the vocabulary of Proto Oceanic An examination of thisreconstructed vocabulary gives us insights into the culture of the speakersof the language in a number of ways
1 An examination of words that can be reconstructed for ProtoOceanic can help us make inferences about the culture of thespeakers of that language
2 Identification of widespread cultural items for which terms can notbe reconstructed for Proto Oceanic suggests that such items weremore recent introductions
3 An examination of reconstructed Proto Austronesian words not re-flected in Proto Oceanic can indicate which original Austronesiancultural items were lost or abandoned by Oceanic speakers as theymoved eastward into the Pacific
As an example of the last point we can reconstruct Proto Austronesianwords referring to different kinds of rice and millet and to rice and milletcultivation but no such reconstructions can be made for Proto OceanicPresumably these crops were abandoned by Oceanic speakers in their mi-gration from Southeast Asia to the Pacific
Terms we can reconstruct for Proto Oceanic embrace a wide culturalrange4 A few of the subject areas are
bull Canoes and fishing Terms for two types of outrigger canoes (large andsmall) outrigger float and boom matting sail paddle bailer launch-ing rollers rudder and anchor as well as terms for various parts of thecanoe and for steering and sailing There are also many reconstructedterms for a number of aspects of fishing technology and of coursenames of many different kinds of fish shellfish and crustaceans
58 CHAPTER 3
bull Pottery Various kinds of pots clay and techniques of clay pot manu-facture decorations and accessories like lids as well as terms fordifferent kinds of cooking (roasting boiling steaming stone or earthoven etc)
bull Food crops Several kinds of yam taro banana pandanus bread-fruit sago and sugar cane as well as terms associated with horti-cultural practices
bull Fruits and nuts A wide range of terms relating to the coconut hasbeen reconstructed including those for different stages of growthand parts of the fruit or tree The words for a number of fruit and nuttrees for betel nut and for plants like ginger and turmeric have alsobeen reconstructed
bull Animals and birds Proto Oceanic terms in this area include wordsfor wild and domesticated pig dog fowl rat bandicoot cassowarycuscus (possum) and numerous bird names
bull Social structure A fairly complete set of kinship terms has been re-constructed as have terms relating to chieftainship and the societalhierarchy
These and other reconstructed words paint the following picture of earlyOceanic culture The original speakers of Proto Oceanic were clearly a mar-itime people They used outrigger canoes fished with hooks and nets andgenerally exploited the resources of the maritime environment They grew anumber of crops including yam taro banana and sugar cane and gatheredfruits and nuts They had domesticated fowls pigs and dogs (and suffered therat) used the spear and the bow and arrow for hunting or warfare made claypots and built houses with shelves and platforms (as well as probably build-ing more temporary shelters in gardens or on the beach) They had a fairlyhierarchical society with chiefs and probably other social ranks as well Theybelieved in gods or spirits and probably practiced sorcery
But some words cannot be reconstructed for Proto Oceanic despite thefact that the items they name are found in most parts of the Pacific today Thesweet potato for example is grown and eaten across the Pacific yet thereis no Proto Oceanic reconstruction for it Apparently the sweet potato wasintroduced after the dispersal of Oceanic speakers Archaeological evidenceconfirms this Other items that also fall into this category include the paw-paw and the cassava (manioc) Our linguistic evidence particularly whenpaired with the archaeological testimony gives us a partial understanding ofPacific prehistory although much remains to be done
The History of the Austronesian Languages 59
CHAPTER
4
The History of the Papuanand Australian Languages
Almost a thousand languages spoken in the Pacific region do not belong tothe Austronesian family Of these more than seven hundred are spoken in ornear New Guinea and are known by the general term ldquoPapuanrdquo the remain-ing two hundred or so are or were spoken in Australia We know much lessabout the history of these languages than about the history of the Austrone-sian languages
The majority of Papuan languages are located in the interior of theisland of New Guinea This area experienced no European contact untilshortly before (and even in some cases some time after) World War II Sowhile many of the languages east of New Guinea had been written for ahundred years or more and had been studied in some detail most Papuanlanguages were unknown to the outside world until very recently
Missionary linguists (especially those working with the Summer Insti-tute of Linguistics) were largely responsible for dramatically increasing ourknowledge of Papuan languages in the decades after 1945 and the pictureis considerably clearer than it was in say the 1960s Nevertheless thereare still very many Papuan languages about which almost nothing is knownand the work of comparative linguistics has barely begun Where Australiais concerned the death of many languages before they had been properlyrecorded leaves us with gaps of a different Kind Much of the evidenceneeded to make historical inferences has disappeared and formulating andtesting historical hypotheses is hampered at every turn
As if these problems were not enough we are faced with a much longerperiod of human habitation in both Australia and New Guinea than in mostof the rest of the Austronesian-speaking world The longer a group of lan-guages have had to diversify the fewer will be their apparent similarities
60
In dealing with the history of both Papuan and Australian languages I canmake only general and tentative statements
41 Interrelationships of Papuan Languages411 Papuan Language FamiliesThe term Papuan refers to those languages of the Pacific region excluding Aus-tralia that are not members of the Austronesian language family It does notrefer to a single family of languages ldquoPapuan languages are not all geneticallyrelated They do not all trace their origins back to a single ancestral languageOn the basis of present knowledge they belong to at least sixty different lan-guage families all with their own common ancestral languagerdquo (Foley 1986 3emphasis mine) Some linguists prefer the label ldquoNon-Austronesianrdquo for theselanguages since it does not imply the genetic unity that a positive label likeldquoPapuanrdquo does ldquoNon-Austronesianrdquo however like any negative label has itsown problemsmdashRussian Chinese English and Swahili after all are also non-Austronesian languagesmdashso I use the term ldquoPapuanrdquo in this book
While Foley does not explicitly define the criteria he uses in decidingmembership or nonmembership of a family it is clear from his conclusionsthat relatively close relationship is involved As far as these sixty or so fami-lies are concerned their ldquowider relations [are] not yet conclusively demon-strated Undoubtedly with more careful and complete comparative workthis picture will become simpler a number of families will probably com-bine into larger families as Romance Germanic and Slavic combine into theIndo-European familyrdquo (Foley 1986 3)
In this initial discussion of Papuan language families I follow Foleyrsquosconservative view later I discuss proposed combinations of these familiesThe list of Papuan families in table 4 proceeds generally on a west-to-eastbasis with the number for each family corresponding to that on map 15 Thelocations given in table 4 refer to geographical regions in Irian Jaya and toprovince names in Papua New Guinea
The situation is however even more complicated than table 4 indicatesNot every Papuan language belongs in a (smaller or larger) family A numberof Papuan languages are currently classified as isolates The term isolaterefers to a one-member family a language that on the basis of current evi-dence appears to have no relatives
412 Possible Wider Groupings of Papuan FamiliesNaturally enough the existence of so many language families in such a rela-tively small geographical area has caused many linguists to look for wider
The History of the Papuan and Australian Languages 61
Table 4 Papuan Language Families
Family Locationa Number oflanguages
West of the New Guinea mainlan1 Timor-Alor-Pantar Timor area 182 Northern Halmahera Halmahera Islands 11
Mainland Irian Jaya onlyb
3 West Birdrsquos Head Birdrsquos Head 64 Central Birdrsquos Head Birdrsquos Head 45 Borai-Hattam Birdrsquos Head 26 South Birdrsquos Head Birdrsquos Head 107 East Birdrsquos Head Birdrsquos Head 38 MairasindashTanah
Merahwestern 4
9 West Bomberai western 310 East Cenderwasih
Baywestern 4
11 TorndashLake Plain northern 2012 Nimboran northeast 313 Kaure northeast 314 Pauwasi northeast 415 Sentani northeast 416 Dani-Kwerba central 1117 Wissel Lakes central 418 Mek (Goliath) eastern 919 Kayagar southeast 320 Yelmek-Maklew southeast 221 Kolopom Frederick Hendrik
Island3
Both sides of the Irian JayandashPapua New Guinea borderc
22 Sko north coast 823 Border northern 1224 Kwomtari northern 525 Senagi northern 226 Central-South New
Guineacentral 54
27 Marind southern 628 Trans-Fly south coast 25
62 CHAPTER 4
Mainland Papua New Guinea onlyd
29 Torricelli East amp West SepikMadang
48
30 Upper Sepik East Sepik 1631 Ram West Sepik 332 Tama East and West Sepik 533 Yellow River West Sepik 334 Middle Sepik East Sepik 1235 Sepik Hills East Sepik 1536 Leonhard Schulze East Sepik and West-
ern6
37 Nor-Pondo East Sepik 638 Yuat East Sepik 639 Mongol-Langam East Sepik 340 Waibuk Enga Madang 441 Arafundi East Sepik 242 Keram (Grass) East Sepik Madang 543 Ruboni East Sepik Madang 844 Goam Madang 1145 Annaberg Madang 346 Arai East Sepik 647 Amto-Musian West Sepik 248 Mugil-Isumrud-Pi-
homMadang 28
49 Josephstaal-Wanang Madang 1250 Brahman Madang 451 Mabuso Madang 2952 Rai Coast Madang 2953 East New Guinea
Highlandsall Highlandsprovinces
42
54 Finisterre-Huon Morobe 6555 Gogodala-Suki Western 356 Kutubuan Southern Highlands 557 Turama-Kikorian Gulf 458 Teberan-Pawaian Simbu Gulf 359 Inland Gulf Gulf 560 Eleman Gulf 761 Angan Gulf 1262 Binanderean Oro 1663 Central-Southeast
New GuineaCentral Milne Bay 36
The History of the Papuan and Australian Languages 63
East of the New Guinea mainland64 New Britain East New Britain
New Ireland8
65 South Bougainville Bougainville 466 North Bougainville Bougainville 467 Yele-Solomons Milne Bay Solomon
Islands5
68 ReefsndashSanta Cruz Solomon Islands 4aGeographical designations in Irian Jaya province names in Papua New GuineabThere are a number of isolates in addition to the languages listed herecThe isolate Yuri belongs in this groupdSeveral isolates occur in this group
relationships between them If the neighboring Austronesian languages canapparently be classified into a single large family then can we not at leastreduce the number of Papuan language families Scholars at the AustralianNational University particularly S A Wurm have attempted to establishlarger groupings of Papuan languages on the basis of what seem to beshared features Lack of adequate information about many languages hashampered this work While some of the proposals rest on solid data othersare much more impressionistic Map 16 shows the locations of proposedwider groupings
Wurm borrowed terms from the biological sciences to refer to someof these wider groupings of languages A stock is a group of languagefamilies that appear to be reasonably closely related to each other whilea phylum is a group of distantly related families or stocks In table4 I have often treated as families groups that other linguists referto as stocks The degree of relationship between Papuan languages ofthe same stock roughly parallels that between geographically dispersedmembers of the Austronesian family but the concept of a phylum isquite different as it implies only a very distant relationship The tech-niques and procedures of comparative linguistics have not yet been ableto prove the existence of such attenuated relationships
One proposed phylum is the West Papuan phylum consisting of theNorthern Halmahera West Birdrsquos Head Central Birdrsquos Head and Borai-Hat-tam families (families 2 through 5 in table 4) along with the Amberbakenisolate for a total of twenty-four languages all in the extreme west of IrianJaya These languages have in common a certain amount of lexical similarity
64 CHAPTER 4
The History of the Papuan and Australian Languages 65
66 CHAPTER 4
and some grammatical features (eg the marking of subject and object byverbal prefixes rather than suffixes [Wurm 1982 208])
Another suggested wider grouping is the Sepik-Ramu phylumconsisting of more than one hundred languages belonging to sixteen differ-ent families (numbered 30 through 45 in table 4) and spoken mainly in theEast Sepik West Sepik and Madang provinces of Papua New Guinea (A fewnearby isolates would also be members of this phylum) These languagesshare a number of distinct phonological features such as a very small num-ber of vowel phonemes and also have some common grammatical features(Wurm 1982 210)
The Torricelli family (29 in table 4) is treated by Wurm as the Torricelliphylum composed of perhaps five or six families Foley (1986 241ndash242)however treats this as a single family largely because these languagesshare a number of grammatical features not found elsewhere among Papuanlanguages (Subject prefixes and complex noun-class systems are two exam-ples)
Wurm has also grouped the Papuan languages spoken east of the NewGuinea mainland into the East Papuan phylum This consists of twenty-five languages belonging to the New Britain South Bougainville NorthBougainville Yele-Solomons and ReefsndashSanta Cruz families (64 through 68in table 4) There appears to be some lexical and grammatical evidence forthe existence of this group though it is not very strong and the situation iscomplicated by the heavy Austronesian-language influence on some of themembers of the phylum
The largest and possibly most controversial genetic grouping Wurmproposes is the TransndashNew Guinea phylum This hypothesis in its most ex-treme form proposes that almost all the rest of the Papuan languagesmdashwiththe exceptions of a few small families and some isolatesmdashbelong to a singlegenetic group of about five hundred languages stretching from Timor in thewest to Milne Bay in the east It would include all of the languages of thesouthern and central part of the mainland as well as some spoken in thenorth (1 6 8 9 11ndash21 23 25ndash27 and 48ndash63 in table 4) There are certainphonological and grammatical features shared by at least some members ofthis group but the existence of the phylum as a wholemdashat least at this stageof our researchmdashseems tenuous to say the least Some support for the hy-pothesis can be found in Pawley (1995)
A number of the families listed in table 4 cannot at present be assignedto any phylum even under the most liberal application of the comparativemethod These lone families are the East Birdrsquos Head Cenderawasih Bay(plus the isolate Yava) Sko Kwomtari Arai and Amto-Musian families(see map 16)
The History of the Papuan and Australian Languages 67
42 Interrelationships of Australian Languages421 Mainland AustraliaSome linguists have divided the languages of mainland Australia into twopseudogenetic groups The Pama-Nyungan group of languages occupiesabout three-quarters of the mainland Its name comes from the words mean-ing lsquomanrsquo at the northeastern and southwestern extremes of the group(Dixon 1980 221) These languages are very similar typologically in bothphonology and grammar The remaining languagesmdashreferred to by the neg-ative term NonndashPama-Nyunganmdashoccupy the northwest of the mainland(see map 11 in chapter 2) They are phonologically fairly similar to the Pama-Nyungan languages but grammatically quite different
In the 1960s a lexicostatistical classification of the Australian languagesdivided the languages into twenty-six ldquophylic familiesrdquo (those sharing 15percent basic vocabulary or less) Of these groups one was Pama-Nyunganand each of the remaining twenty-fivemdashNonndashPama-Nyunganmdashgroups washeld to be a ldquophylic familyrdquo on a par with Pama-Nyungan (Dixon 1980 263)
Dixon himself is highly distrustful of this classification A majority of lin-guists now agrees that all the mainland languages belong to a single Aus-tralian family The differences between Pama-Nyungan and NonndashPama-Nyungan languages are developmental rather than original ldquoIt seems clearhellip that nearly all the languages of Australia form one genetic family goingback to a single ancestral language proto Australianrdquo (Dixon 1980 228) Be-cause of the thousands of years of contact between Australian languageshowever shared innovations supporting subgrouping hypotheses are ex-tremely hard to find ldquoPresent knowledge of the relationships between Aus-tralian languages is not sufficient to justify any sort of fully articulated lsquofam-ily treersquo modelhellip It could perhaps be that the continual levelling due todiffusion of features of every sort has obscured those genetic splits that didtake place in the development of Australian languages so that it will not bepossible to reconstruct themrdquo (Dixon 1980 264ndash265)
422 TasmaniaGenocide in Tasmania has led to the loss of all Tasmanian languages AnAboriginal population of possibly five thousand people at the time of first Eu-ropean contact speaking somewhere between eight and twelve languageswas exterminated in less than eighty years The last full-blooded Tasmaniandied in 1888 although there are still about four thousand people of partialTasmanian Aboriginal descent living in Tasmania and elsewhere So littlewas recorded of these languages that it is almost impossible to say anything
68 CHAPTER 4
about them (Crowley 1993) As regards their history Dixon (1980 233) saysldquoAll we can conclude is thismdashthere is NO evidence that the Tasmanian lan-guages were NOT of the regular Australian type They have been separatedoff for so long and the available materials are so poor that the likelihood ofa genetic connection cannot be confirmed The genetic affiliation of Tasman-ian is and must remain unprovenrdquo
43 Possible External LinksSuggestions about the wider relationships of Papuan and Australian lan-guages have not been lacking but given the long periods of time involvedmost of these can remain no more than suggestions Greenbergrsquos (1971)Indo-Pacific Hypothesis attempts to link Papuan languages with those ofTasmania (but not mainland Australia) and of the Andaman Islands in theIndian Ocean Greenberg speculated that Australian languages are relatedto the Dravidian languages of South India Scholars and amateurs have alsolooked for relationships between Papuan or Australian languages and thoseof Africa and Asia None of these hypotheses seems to be based on any ev-idence more solid than typological similarities or a few possibly accidentallexical similarities
Foley (1986 271ndash275) however has recently presented a small but tan-talizing amount of evidence for the existence of a genetic link between Aus-tralian languages and the languages of the Eastern Highlands of Papua NewGuinea A small number of basic vocabulary items look as if they might be cog-nate But he says that this evidence ldquoin no way constitutes proof of a geneticrelationship between Australian and Eastern Highlands languageshellip Muchmore detailed and careful research needs to be done before a convincing proofis provided and given the time depth that may never be possible Rather theabove data represent a first attempt at marshalling some evidence for a ge-netic link between Australian and Papuan languagesrdquo (Foley 1986 275)
Up until around eight thousand years ago New Guinea and Australiawere one continent Only then did sea levels start rising after the last Ice Ageto form what is now the Torres Strait It is distinctly possible that Australiawas settled from the New Guinea area so the idea of a genetic link betweenthe two areas cannot be ruled out
44 Implications for Prehistory441 OriginsThe New Guinea mainland was probably occupied at least fifty thousandyears ago presumably by the ancestors of the speakers of (some) Papuan
The History of the Papuan and Australian Languages 69
languages Australia was also settled at about the same time Since compar-ative linguistics cannot reach back more than about eight or ten thousandyears most of that fifty-thousand-year period is lost to linguistics
It is possible that all the Papuan families are related descending from asingle ldquoProto Papuanrdquo ancestor that we cannot even dimly imagine A singlelanguage spoken somewhere in the New Guinea area around 50000 BCcould have given rise over time to all of the modern Papuan languages andthis language may have been the ancestral language from which all Aus-tralian languages ultimately derive
We also have no evidence to indicate the origins of the first speakers ofPapuan or Australian languages Linguistic connections with Asia or Africaare nothing more than highly speculative nor would we expect otherwiseIf the time elapsed is too great to show interrelationships among all Papuanlanguages it is certainly too great to show genetic relationships betweenthese languages and those to the north south or west
But perhaps this very lack of evidence for external relatives means thatthe Papuan languages do or did form a genetic unity and that the same couldbe said about the Australian languages If the diversity existing among mod-ern Papuan and Australian families is due to different origins and differentmigrations of people at various times from various locations one might expectto find some genetic connections between individual Papuan or Australianfamilies and Asian or African language families The fact that we do not whilenot strong enough to be called evidence does suggest that the Papuan lan-guages may have formed a single linguistic grouping in the very distant pastand that the same may be true of Australian Tasmanian included
In only a few instances has there been anything in the way of comprehen-sive reconstruction of the phonology grammar and vocabulary of any of thelarger Papuan families and the situation in Australia is pretty similar Littlecan be said about relations with other families origins and migration routesand earlier stages of Papuan or Australian culture and the little that has beensaid on these topics must be treated as highly speculative
442 DiversificationOne question that must be asked in any study of the Papuan and Australiansituations is if both New Guinea and Australia have been settled for aboutthe same length of time why do we find such incredible genetic diversityamong Papuan languages whereas Australian languages all seem to belongto just a single family
Physical geography and its effect on wide-ranging human movementis one contributing factor ldquoMost of New Guinea is difficult country indeed
70 CHAPTER 4
steep forest-covered mountains with precipitous drops swirling riversdense nearly impenetrable rainforests and endless tracts of swamplandThe terrain thus poses some genuine barriers to human social interactionand would certainly favour rather than inhibit linguistic diversityrdquo (Foley1986 9) Geographical barriers like these were often bolstered by socialbarriers Hostile relations were the rule between neighboring communitiesand the tendency of language to take on emblematic functions and to beconsidered as a mark of group identity throughout Melanesia is one sign ofcommunitiesrsquo desires to set themselves apart People often created linguis-tic differences or exaggerated differences that already existed in order topreserve their group membership
Kulick (1992 2ndash3) for example quotes the following observation madeby Ken McElhanon who worked among Selepet speakers in Papua NewGuinea ldquoThe people living in the Selepet-speaking village of Indu had gath-ered together for a meeting During this meeting a decision was reached tolsquobe differentrsquo from other Selepet speakers It was agreed that the villagers ofIndu would immediately stop using their usual word for lsquonorsquo bia which wasshared by all their fellow speakers of Selepet Instead they would begin say-ing buŋε which they did and have continued doing since that timerdquo Thereare many other similar examples In Buin (spoken in Bougainville) speakersof the Usai dialect have reversed all gender agreements masculine mark-ers in other dialects become feminine in Usai and feminine markers becomemasculine (Laycock 1982a) Similar phenomena can be observed in Oceaniclanguages Speakers of Ririo (Choiseul Solomon Islands) seem to have de-liberately transposed the last consonant and vowel of words to make themsound more different from their counterparts in the neighboring closely re-lated language of Babatana (Laycock 1982b 274ndash276)
Babatana Ririososole susuel lsquonakedrsquovumi vuim lsquobeardrsquopiru piur lsquowildrsquobose bues lsquomanrsquo
The Australian continent is in some ways less difficult geographicallythan New Guinea and physical barriers to long-distance communicationare generally much less extreme Though Australians belonged to distincttribal and linguistic groups there was much more social interaction be-tween those groups typically accompanied by transfer of vocabulary itemsfrom one group to another Rather that accentuating differences Austral-ians seem to have made an active effort to keep different languages frombecoming too different Dixon (1980 239) refers to ldquoa gradual but constant
The History of the Papuan and Australian Languages 71
shifting of tribal groupsrdquo through which people came into contact with dif-ferent languages He also mentions mergers of different tribal groups whosenumbers had been reduced by famine or disease Such factors conspired tokeep Australian languages more similar to each other than one might ex-pect especially in comparison to Papuan languages
Both Papuan and Australian languages have been in the Pacific region fora very long time indeed We know nothing of where they came from and littleof how their speakers moved around the region We do not know how far eastor west of the New Guinea mainland Papuan speakers might have originallysettled or much about their connections with Australian languages All we canreasonably deduce is that by the time speakers of Austronesian languagesarrived in this area about four to five thousand years ago or so speakers ofPapuan and Australian languages were very much entrenched
72 CHAPTER 4
PART TWO
Structure
CHAPTER
5
Sound Systems
The sound systems of languages in different parts of the Pacific vary enor-mously sometimes even when the languages themselves are closely relatedMajor similarities and differences exist between languages of the threebroad genetic groupsmdashAustronesian Papuan and Australian Below I dis-cuss the vowel systems consonant systems stress and tone and the way inwhich words are structured in each group touching briefly as well on thedevelopment of orthographies1
51 Oceanic Languages2
511 Vowel SystemsThe great majority of Oceanic languages have five vowel phonemes whichis also the commonest system found among the worldrsquos languages generallyA vowelrsquos position in the diagram corresponds to how it is described eg iis a high front vowel
i ue o
a
This system is universal in the languages of Polynesia and widespread inMelanesia though among Micronesian languages only Kiribati has fivephonemic vowels This same system has also been reconstructed for ProtoOceanic In many languages there is also a phonemic (significant) differencebetween short vowels and long vowels a long vowel being one that takesalmost twice as much time to articulate as a short vowel The examples
75
below show in various languages that vowel length alone is sufficient todistinguish two otherwise identical words A long vowel is marked with a fol-lowing colon a is long and a is short
Samoanmalo lsquoloinclothrsquo malo lsquohardrsquolulu lsquobarn owlrsquo lulu lsquoshakersquo
Nukuoronui lsquococonutrsquo nui lsquogreenrsquoahe lsquogo backrsquo ahe lsquowhenrsquo
Paamesemen lsquoitrsquos ripersquo men lsquohis tonguersquovati lsquohe stoppedrsquo vati lsquohersquoll bite ifrsquo
A handful of languages have fewer than five vowels One Micronesianlanguage Marshallese has been analyzed as having only four vowelphonemes These are written ię e and a but they have wide variations inpronunciation The vowel e for example is variously pronounced [ε] [ә] and[o] depending on the neighboring consonants Some languages in the Mo-robe Province of Papua New Guinea also have fewer than five vowels Marifor example has just i a u Adzera i a o u
Quite a few languages have more than five phonemic vowels Rotumanfor example has ten Almost all of the languages of Micronesia have morethan five vowels Kosraean has twelve Lagoon Trukese and Saipan Carolin-ian each have nine Yapese and Ulithian eight Mokilese seven and Nau-ruan Chamorro and Palauan six Some dialects of Ponapean have sevenvowel phonemes others six (See appendix 3 for the vowel inventories ofKosraean and Mokilese) Vowel length is also significant in Micronesian lan-guages as the example shows
Mokilesepaj lsquonestrsquo paj lsquohollow of canoersquoros lsquodarknessrsquo ros lsquoflowerrsquo
In Melanesia most languages with more than five vowels have just one ortwo extra ones In Tanna and Malakula (Vanuatu) as well as in New Ireland(Papua New Guinea) languages with six vowels generally add ә (the sound of ain English words like ago or sofa) Some languages in Melanesia have developedseven-vowel systems the basic five vowels plus either front rounded vowels (likethe vowels in French rue and heureux) or a contrast between two different e-sounds and two different o-sounds (e and ε o and ͻ)
The most complex Oceanic vowel systems however are almost cer-tainly those of New Caledonia (see appendix 3) Iaai in the Loyalty Islands
76 CHAPTER 5
for example has eleven short vowels all of which can also occur longwhile Xacircracirccugraveugrave on the mainland has ten oral and seven nasal vowelphonemes each of which can occur short or long yielding thirty-four vowelcontrasts
How do such complex vowel systems evolve out of an original five-vowelsystem The changes that took place in different Oceanic languages are verydifferent Here I give just two kinds of examples First phonemes often havemore than one pronunciation depending on their phonetic environmentImagine that the phoneme a was pronounced [aelig] (the sound representedby a in English cat) when the vowel in the next syllable was i but as [a] (likein father) elsewhere We would have pairs of words like
mati lsquosickrsquo [maeligti]mata lsquoeyersquo [mata]
The pronunciation of phonemic amdash[aelig] or [a]mdashis totally predictable Nowimagine that this language drops out all vowels at the end of words as hashappened in many Oceanic languages The following changes occur
[maeligti] gt [maeligt] lsquosickrsquo[mata] gt [mat] lsquoeyersquo
Now the contrast between [aelig] and [a] creates a minimal pair and aelig hasbeen added to the languagersquos phonemic (as opposed to phonetic) inventory
Rotuman illustrates a second kind of process Most Rotuman wordshave ldquolongrdquo and ldquoshortrdquo forms that are used in different grammatical con-texts3 In some cases the short form simply drops the final vowel of the longform (Note that the symbol ŋ represents the ng sound in English sing whileʔ represents the glottal stop)
RotumanLong form Short formhaŋa haŋ lsquofeedrsquoheleʔu heleʔ lsquoarriversquo
Metathesis howevermdashtwo phonemes exchanging placesmdashis more com-mon With some vowel combinations metathesis has no further phonologi-cal repercussions
RotumanLong form Short formhosa hoas lsquoflowerrsquotiko tiok lsquofleshrsquopepa peap lsquopaperrsquo
Sound Systems 77
But with other combinations the two vowels that came into contact havefused to produce a third different vowel (The vowel ouml is a bit like the vowelin French heureux while uuml is the vowel in French rue)
RotumanLong form Short formmose (gt moes gt) moumls lsquosleeprsquofuti (gt fuit gt) fuumlt lsquopullrsquo
Because of this Rotuman which originally probably had five vowels now hasten
512 Consonant SystemsPolynesian Languages
In general terms the Polynesian languages have the simplest consonantsystems of all the Oceanic languages Tongan has the largest inventoryof consonant phonemes of all of the Polynesian Triangle languages withtwelve A number of Polynesian languages for example Hawaiian have onlyeight consonants
Tongan HawaiianP t k ʔ P k ʔv wf s h hm n ŋ m n
l l
The consonant systems of the Polynesian Outlier languages are gener-ally slightly more complex (Krupa 1982) In some cases this is a result ofcontact with neighboring non-Polynesian languages First unlike any Poly-nesian Triangle language quite a few Outliers among them West FutunaIfira-Mele Emae and Takuu make a distinction between l and r Secondin addition to the normal Polynesian stop consonants p t k ʔ some Outliersshow a contrast with the aspirated stops ph th kh (eg Takuu and Kapinga-marangi) with the voiced stops b d g (Fagauvea) or with the prenasal-ized stops mb nd (eg Emae and Pileni) Third there is contrast betweenthe ordinary nasals m n ŋ and one or more of the voiceless nasals m n inKapingamarangi Fagauvea and Pileni
Micronesia
The consonant systems of the languages of Micronesia are quite differentfrom those of the Fijian and Polynesian languages Lagoon Trukese is fairly
78 CHAPTER 5
typical of the majority of these languages It has the following fourteen con-sonants (tʃ represents a sound something like ch in English church butwith the tongue turned back)
Lagoon Trukesepw P t tʃ kmw m n ŋ
f sr
w y
All consonants except w and y have both short and long forms
Lagoon Trukesesɨk lsquoappearrsquo sɨk lsquobleedrsquokamwe lsquoclamrsquo kamwet lsquosweetheartrsquotʃimw lsquoheadrsquo tʃin lsquospeedyrsquotakir lsquolaughrsquo takitʃ lsquotorch-fishingrsquo
Most other Micronesian languages have similar consonant systems (in-cluding the distinction between short and long consonants) although Kiri-bati has no phonemic fricatives A number have in addition to the trilled reither a flapped r or an l Some like Kosraean Nauruan and Yapese (seeappendix 3) have more complex systems of consonants
Melanesia
There is a considerable variety of consonant systems in Melanesia and al-though neighboring languages often have similar systems one cannot makebroad generalizations on a geographical basis It is fair to say however thatthe consonant systems of New Caledonia are considerably more complexthan those of the rest of this region
Some of the simpler consonant systems in this region are found in theNew Guinea area Below for example are the consonants of the Tigak lan-guage of New Ireland
Tigakp t kb g
svm n ŋ
rl
Sound Systems 79
Probably half of the Melanesian languages would fall into a category ofmedium complexity as far as any classification of consonant systems is con-cerned This complexity usually involves one or more of the following (1)contrast between oral and prenasalized stops (2) contrast between simpleand aspirated stops (3) contrast between voiced and voiceless fricativesand (4) contrast between simple and labialized or velarized consonantsStandard Fijian and the Tolsquoabalsquoita dialect of North Malaita (Solomon Is-lands) illustrate such phonological systems
Fijian Tolsquoabalsquoitap t k t k kw ʔmb nd ŋg mb nd ŋg ŋgw
f s f θ sv ethm n ŋ m n ŋ
l lr
w y w
An unusual phonological feature of some of the languages of north Malakulaand east Santo in Vanuatu are the apico-labial consonants p m whichare produced with the tip of the tongue touching the upper lip
Themostcomplexconsonantsystems inMelanesiaare thoseof the languagesof New Caledonia (see appendix 3 for two examples) The Drehu language of theLoyalty Islands has twenty-eight consonant phonemes including a contrast be-tween the alveolar stops t d and the retroflex stops ṭ ḍ (similar to that found inmany Indian languages)mdasha fairly rare contrast t d (similar to that found in manyIndian languages)mdasha fairly rare contrast in Oceanic languages Both Drehu andPijea languageofthenorthernmainlandthathasthirty-fiveconsonantphonemescontrast voicedandvoicelessnasal lateral andsemivowel phonemes
513 Prosodic FeaturesThe system of consonants and vowels in a language is often referred to asthe segmental phonology of the language since linguists break up (seg-ment) a stream of speech into discrete units Other features of speech thatdo not belong to individual segmentsmdashconsonants or vowelsmdashbut to sylla-bles or words are known as suprasegmental or prosodic features Stressand tone are two of the most important of these
Stress
The term stress refers to the relatively greater prominence given to onesyllable in a word through extra effort extra loudness a change in pitchor some combination of these factors The underlined syllables in the Eng-
80 CHAPTER 5
lish words temptation absolute absolutely and resist receive greater stressthan the other syllables in those words
In the majority of Oceanic languages the position of stress in a word ispredictable Let us take Samoan as an example The basic pattern in Samoanis one of penultimate stress Stress (marked here by an acute accent overthe vowel of the syllable) falls on the next-to-last syllable of the word
Samoantuacuteli lsquodismissrsquo taacutema lsquochildrsquotuliacuteŋa lsquodismissalrsquo tamaacuteʔi lsquoyoung of animalsrsquo
When a suffix is added to a word in Samoan the stress shifts to the right sothat it still falls on the penultimate syllable tuacuteli becomes tuliacuteŋa
When a Samoan word ends in a diphthong (like ae ai au for example)or in a long vowel stress falls on this final diphthong or long vowel
Samoanatamaacutei lsquocleverrsquo faifeʔaacuteu lsquopastorrsquotamaacute lsquofatherrsquo paʔuacute lsquofallrsquo
Most Oceanic languages seem to have a predictable pattern of penulti-mate stress but in some languages while stress is predictable the patternsare different One such language is Māori There are three rules involvedin the assignment of stress in Māori (a) The first long vowel in a word isstressed as in the examples in (a) below (b) if there are no long vowels thefirst vowel cluster is stressed as in (b) and (c) if there are no long vowels orvowel clusters as in (c) then the first vowel is stressed
Māori(a) manaacuteki lsquosupportrsquo paacutetutahi lsquoa villagersquo(b) tamaacuteiti lsquochildrsquo taacuteutau lsquobarkingrsquo(c) taacutemariki lsquochildrenrsquo hoacutero lsquofastrsquo
Languages with unpredictable stress patterns are relatively uncommonin the Pacific although they do exist In many languages of this type how-ever there seems to be one common stress pattern other patterns beingvery much in the minority The Big Nambas language of Malakula in Vanuatuis an example of this type of language In it the majority of words arestressed on the penultimate syllable
Big Nambasaacuteγau lsquogo awayrsquo patiraacuteni lsquoput it uprsquoipaacuteli lsquohersquoll burn itrsquo iputakmaacuteni lsquohersquoll spoil itrsquo
But sometimes stress falls on the final syllable Compare the two pairs be-low identical except for stress
Sound Systems 81
Big Nambasaacuteγau lsquogo awayrsquo aγaacuteu lsquochiefrsquos wifersquoipaacuteli lsquohersquoll burn itrsquo ipaliacute lsquohersquoll tie itrsquo
Further as is not the case in Samoan the stress remains in its original posi-tion even when suffixes are added
Big Nambasγaacutepat lsquochiefrsquo γaacutepatak lsquomy chiefrsquopraacutepar lsquosow (pig)rsquo praacuteparan lsquohis sowrsquo
Tone
Phonemic tone refers to contrasting pitch occurring at the word level Thesame string of consonants and vowels can mean different things if the pitchof the voice is high or low rising or falling While common in Asian andAfrican languagesmdashand in Papuan languages as wellmdashtone is fairly rare inthe rest of the Pacific Among Oceanic languages just a few in New Cale-donia (like Cegravemuhicirc) and a few more in the Morobe Province of Papua NewGuinea (like Yabecircm) have phonemic tone
Cegravemuhicirc has three tones high (marked here with an acute accent) mid(marked with a macron) and low (marked with a grave accent) as exempli-fied in the following words
Cegravemuhicirctiacute lsquodestroyrsquotī lsquogatherrsquotigrave lsquowritersquo
Yabecircm has two tones high and low
Yabecircmaacutewaacute lsquovaluablesrsquo agravewagrave lsquohisher mouthrsquowaacute lsquomangorsquo wagrave lsquocrocodilersquosaacute lsquoto hammerrsquo sagrave lsquoput on top ofrsquooacuteliacute lsquobodyrsquo ograveligrave lsquowagesrsquo
Because tone is a rare phenomenon in Oceanic we assume that the fewlanguages that have it have developed it some time after they split off frommost of their other relatives But how do languages develop tone systemsLet us look briefly at what seems to have happened in Yabecircm and closely re-lated languages (Bradshaw 1979 Ross 1993)
At one time there was probably a rule in Yabecircm that a syllable contain-ing a voiceless stop or fricative (like p t k s) would have high tone but onewith a voiced stop or fricative (b d g j) would have low tone4 For exam-ple kaacutepuacuteŋ lsquoI plantrsquo and kaacutetaacuteŋ lsquoI make a soundrsquo but gagravebugrave lsquoI insultrsquo and
82 CHAPTER 5
gagravedugraveʔ lsquoI bowrsquo Some consonants that conditioned high or low tone havesince changed their voicing (or even disappeared) but they have left theirtone ldquotracerdquo behind For example earlier s remained s in Yabecircm and be-cause it is and was voiceless it is associated with high tone
Yabecircmsipo gt seacutep lsquogo downrsquosaqit gt siacute lsquosewrsquo
On the other hand earlier j was voiced and it conditioned low tone on thefollowing syllable but later became voiceless s
Yabecircmjoacuteŋi gt soacuteŋ lsquostop up plugrsquojoRi gt sograve lsquotiersquolejan gt lέsέŋ lsquonitrsquo
514 Word StructureSome Oceanic languages allow only open syllables meaning that eachsyllable may begin with a consonant but may not end with one These lan-guages do not permit consonant clustersmdashtwo or more consonants com-ing together without an intervening vowel Using C for consonant V forvowel and ( ) to indicate that whatever is enclosed is optional the generalstructure of words in languages of this type is built on the pattern (C)V(C)Vhellip where vowels (and in some languages consonants) may be short or long
Languages that allow only open syllables occur in some parts of PapuaNew Guinea and Vanuatu the southeastern Solomons most of Fiji and Poly-nesia Examples
Mekeoakaikia lsquogreatrsquooisofai lsquooff you gorsquoekapaisau lsquohe made mersquo
Arositaroha lsquonewsrsquoamamu lsquoyour fatherrsquohaʔaheuheu lsquochange formrsquo
Fijianveitau lsquofriendsrsquovakasalataka lsquoadvisersquombatambata lsquocoldrsquo
Sound Systems 83
Hawaiianpauloa lsquoeverythingrsquohoahanau lsquocousinrsquokukonukonu lsquoexcessiversquo
Probably the majority of Austronesian languages however allow bothopen and closed syllables (syllables ending in a consonant) In some casesonly a few consonants (most frequently nasals) can close a syllable In suchcases there are few consonant clusters and they mainly occur across mor-pheme boundaries Here are some Banoni examples (note that ts repre-sents a single phoneme in Banoni not a consonant cluster)
Banonimatam lsquoyour eyersquoβatamumam lsquomake us eatrsquoteŋtapatsi lsquobroken off and scatteredrsquo
In other cases however consonant clusters are frequent and can occur insyllable-initial position as well as across syllable boundaries
Adzeratatariʔ lsquofowlrsquoromgam lsquoyourselfrsquotafa-ŋga-ŋʔ lsquoour ancestorsrsquo
Maringefnakno lsquofamousrsquokntildeaokntildearoo lsquobe stringyrsquosnaplu lsquoslip outrsquo
Big Nambasprapar lsquosow (female pig)rsquovənmaran lsquoold womanrsquokətəγsrasr lsquoyoursquove sweptrsquo
Most Oceanic languages have a large amount of reduplication aprocess wherein all or part of a word is repeated Look at the following ex-amples from Hawaiian
Hawaiianʔaki lsquoto take a nip and let gorsquoʔakiʔaki lsquoto nibble (as a fish)rsquoʔaʔaki lsquoto nip repeatedlyrsquo
84 CHAPTER 5
The basic verb is ʔaki The verb ʔakiʔaki shows complete reduplicationwith the whole verb root being repeated while the verb ʔaʔaki is an exam-ple of partial reduplication in which only part of the verb (in this casethe first syllable) is repeated Reduplication commonly has a number offunctions in the languages in which it is productive Take a look at these ex-amples
1 Repetition or continuous action
Māoripaki lsquopatrsquo pakipaki lsquoclaprsquokimo lsquowinkrsquo kimokimo lsquoblink wink repeatedlyrsquo
2 Intensity
Tahitianhiʔo lsquolook atrsquo hiʔohiʔo lsquostare atrsquoparau lsquoconversersquo parauparau lsquotalk a lotrsquo
3 Similarity or diminution The reduplicated word refers to some-thing similar to but often smaller or more moderate than itsunreduplicated counterpart
Tonganviku lsquowet all overrsquo vikuviku lsquodamprsquohavili lsquostrong windrsquo havilivili lsquogentle wind breezersquo
4 Change in part of speech eg making a noun into an adjective
KosraeanpΛk lsquosandrsquo pΛkpΛk lsquosandyrsquopweŋ lsquonewsrsquo pweŋpweŋ lsquofamousrsquo
5 Change from transitive to intransitive (In the transitive verbs be-low the suffix -i marks the third person singular object)
TigakTransitive Intransitive
nol-i lsquothink aboutrsquo nonol lsquobe thinkingrsquovis-i lsquohit himrsquo visvis lsquofightrsquokalum-i lsquosee itrsquo kalkalum lsquolook appearrsquo
6 Indication of plurality usually of the subject of a verb but some-times of some other aspect of the action (In the examples belowthe reduplicated form is used if the subject of the verb is plural)
Sound Systems 85
SamoanSingular Pluralʔai ʔaʔai lsquoeatrsquotu tutu lsquostandrsquoŋalue ŋalulue lsquoworkrsquo
Nearly all the examples so far have been from Polynesian and Microne-sian languages Here is a set of examples from a Melanesian language theNguna Island dialect of Nakanamanga (Vanuatu) The function of each ex-ample of reduplication is given in the right-hand column
Nakanamanga (Nguna dialect)kati lsquobitersquo katikati lsquonibblersquo diminutiontaki lsquothrowrsquo tataki lsquocontinually
throwrsquorepetition
namalo lsquopiecersquo namalomalo lsquopiecesrsquo pluralityvano lsquogorsquo vanovano lsquotravel aroundrsquo randomnesstare lsquowhitersquo tareare lsquovery whitersquo intensification
When reduplication is partial it may be prefixed suffixed or infixed oc-curring before after or in the middle of the root A rare example of infixedreduplication given above is Samoan ŋalue lsquowork (singular)rsquo ŋaluluelsquowork (plural)rsquo Below are four examples from Manam The first two showpartial prefixed reduplication and the last two partial suffixed reduplication
Manamsalaga lsquobe longrsquo sasalaga lsquolong (plural)rsquoeno lsquosleeprsquo eneno lsquoalways sleeprsquosapara lsquobranchrsquo saparapara lsquohaving branchesrsquoʔulan lsquodesirersquo ʔulanlaŋ lsquodesirablersquo
The last Manam example shows that there are often morphophonemicchanges involved with reduplication so that the reduplicated part of the wordis not always phonologically identical to the unreduplicated part In Tonganvowels undergo changes in many reduplicated words Some of these changesinvolve differences in length others differences in vowel quality
Tonganpoʔuli lsquobe darkrsquo popoʔuli lsquobe somewhat darkrsquomafi lsquopowerfulrsquo mafimafi lsquoalmightyrsquoteliŋa lsquoearrsquo taliŋeliŋa lsquofungusrsquomuʔa lsquofrontrsquo muʔomuʔa lsquogo in frontrsquo
In Ponapean when certain categories of consonants come togetheracross a morpheme boundary as a result of reduplication the first is re-
86 CHAPTER 5
placed by a nasal as in (a) below In other cases a vowel is introduced tobreak up the consonant cluster as in (b)
Ponapean(a) pap lsquoswimrsquo pampap
kak lsquoablersquo kaŋkaksas lsquostaggerrsquo sansastit lsquobuild a wallrsquo tintit
(b) tsep lsquobeginrsquo tsepitsepkatsore lsquosubtractrsquo katsikatsorekatek lsquobe kindrsquo katakatekmasukun lsquobe blindrsquo masamasukum
52 Papuan Languages521 Vowel SystemsThe majority of Papuan languages have the standard five-vowel systemfound among the Austronesian languages as well
i ue o
a
Although this is the most common system some Papuan languages in-cluding many of those in the Sepik area of Papua New Guinea have fewerthan five phonemic vowels while others have more Compare Iatmulrsquos threevowels to Vanimorsquos eight
Iatmul Vanimoɨ i uə e ə o
ɛ ͻa a
Foley (1986 54) says that no Papuan language with more than eight phone-mic vowels has been attested
A number of Papuan languages for example Pawaian contrast oral andnasalized vowels (The examples below are all low tone)
Pawaiansugrave lsquogingerrsquo yegrave lsquoancestorrsquosṹ lsquoroadrsquo yẽ lsquotype of nutrsquo
Distinctions of vowel length do occur in Papuan languages though this fea-ture is much rarer than it is in Oceanic languages
Sound Systems 87
512 Consonant SystemsConsonant inventories in many Papuan languages are relatively small (asample of Papuan consonant inventories is given in appendix 3) No lan-guage in the world has a smaller consonant inventory than Rotokas (spokenon Bougainville) which has only six consonant phonemes5
Rotokasp t kv r g
There are however Papuan languages with more complex consonantsystems A number of languages distinguish prenasalized and simple stopswhile some languages (like Kacircte for example) have coarticulated labial-ve-lar stops In addition to the labial stops p and b made by closing the lipsand the velar stops k and g made by putting the tongue up in the back ofthe mouth there are the coarticulated stops kp and gb produced by simul-taneously closing the lips and raising the tongue at the back of the mouth
Languages of the Highlands of Papua New Guinea are well known foramong other things their range of laterals (or l-like sounds) Kobon forexample has three laterals an alveolar lateral l rather like English l aretroflex lateral ḷ with the tip of the tongue turned back to the roof of themouth and a palatal lateral λ a bit like the ly in the English word hal-yard Melpa also has three laterals dental l (made with the tongue betweenthe teeth) velar ɫ (with the tongue raised at the back of the mouth) andflapped (where the tongue flaps against the tooth ridge) Both Kobon andMelpa also have an r phoneme that contrasts with all of these laterals
Perhaps the most complex Papuan phonological system however isfound in Yele (or Yeletnye) the language of Rossel Island In addition to aset of simple phonemes Yele also has labialized palatalized prenasalizedand postnasalized consonants plus in some cases coarticulated consonantsas well So in addition to simple p there is labialized pw palatalized pyprenasalized mp postnasalized pm and coarticulated tp and kp Andsimilar statements could be made about many other Yele consonants
523 Prosodic FeaturesMany descriptions of Papuan languages do not mention stress perhapsbecause it is often associated with tone and it is difficult to find general pat-terns In some languages stress appears to be predictable though there isa range of patterns Waskia for example tends to stress the last syllable ofa word whereas Kewa prefers the first
88 CHAPTER 5
Waskiakadiacute lsquomanrsquonauacuter lsquococonutrsquobagesaacuten lsquoit staysrsquonamerukoacute lsquohe must gorsquo
Kewa6
poacutena lsquocutrsquoruacutemaa lsquoportion outrsquoroacutegoma lsquoclayrsquo
In other Papuan languages though stress is not predictable as the fol-lowing examples from Koita illustrate
Koitaoacutemo lsquoheadrsquo omoacute lsquoadzersquoγuacutedi lsquodigging stickrsquo γudiacute lsquolimersquoγuacutema lsquopathrsquo γumaacute lsquoaxersquo
Quite a number of Papuan languages have phonemic tone Tone lan-guages are mainly found in the central Highlands and in parts of the Morobeand Sepik provinces of Papua New Guinea but they do occur in other partsof the Papuan region as well Most Papuan tone languages contrast only highand low tones
Pawaiansuacute lsquotoothrsquo sugrave lsquogingerrsquoyeacute lsquonewrsquo yegrave lsquoancestorrsquo
Foreagravesigraveyuacutewegrave lsquoI stand uprsquo agravesigraveyugravewegrave lsquoI peel itrsquonagraveyagraveneacute lsquomy hairrsquo nagraveyaacuteneacute lsquomy kidneyrsquo
Foley (1986 63) says that in many such languages tone is closely associatedwith the stress system with high tone correlating with accented syllablesand that these are not strictly speaking tone languages7
In some languagesmdashespecially in the Eastern Highlands of Papua NewGuineamdashtonal systems are more complex These seem to be true tonal sys-tems The following words in Awa which has four phonemic tones illustratethis8
Awapǎ lsquofishrsquo rising tonenacirc lsquotarorsquo falling tonenaacute lsquobreastrsquo high tonenagrave lsquohousersquo low tone
Sound Systems 89
524 Word StructureSome Papuan languages have only open syllables A number of these lan-guages allow combinations of vowels sometimes quite a few vowels appear-ing in sequence without any intervening consonant
Toaripipasisa lsquoladderrsquoeaso lsquofish spearrsquomaeamariti lsquoshamersquoeae lsquoerroneouslyrsquo
Mountain Koiarineinuvueabe lsquotheir mothersrsquoneiniai lsquoproperlyrsquosaiamo lsquoslowrsquoialelua lsquoconsequentlyrsquo
Some Papuan languages that generally have open syllables (see the firsttwo words in the example below) allow syllables to be closed with a nasal
Buinitaka lsquofreshwater shrimprsquotopituumoru lsquofish-killerrsquokuikuiŋ lsquodriftwoodrsquorempo lsquobattle axersquo
Probably the majority of Papuan languages allow fairly widespread conso-nant clustering Words may end in a range of consonants
Wahgiamŋa lsquoyawnrsquooksnal lsquoavoidrsquomolmŋe lsquothey werersquoendzmo lsquowastersquokopsnde lsquocut openrsquokandzIp lsquothey sawrsquo
Kamasauberyi lsquobeanrsquotorbiŋ lsquomouth harprsquofraŋgi lsquotomorrowrsquosuŋgrum lsquotype of grassrsquosurog lsquocaterpillarrsquowand lsquospeechrsquo
90 CHAPTER 5
Reduplication is a much less common feature of Papuan than of Oceaniclanguages
53 Australian LanguagesIn comparison with Oceanic and Papuan languages Australian languagesare probably of moderate phonological complexity None of them hasphonemic tone for example and in most stress is predictable occurringon the first syllable of the word Many have quite small vowel inventoriesthough a few Australian languages rival those of New Caledonia in theirlarge number of vowels Consonant inventories are neither small norlarge
531 Vowel SystemsMost Australian languages have just three vowel phonemes though many ofthese also distinguish vowel length for a total of six vowel contrasts
i u i ua a
Exceptions are generally of two kinds First there are a few languagesin Central Australia that have only two vowel phonemes Kaitij for examplehas just ɨ and a (though each of these has a number of different pronuncia-tions in different phonetic contexts) Second some languages in the northand northwest have a four-or five-vowel system for example Alawa andKunjen
Alawa Kunjeni u i ue e o
a a
But a few languages especially those in the Cape York area havedeveloped complex vowel systems from what was probably an ancestralthree-vowel system One such system that of Anguthimri appears in ap-pendix 3
532 Consonant SystemsIn discussing the consonants of Australian languages it is helpful to use twotechnical terms Apical refers to sounds made with the tip of the tongueand laminal describes sounds made with the blade of the tongue Many
Sound Systems 91
Australian languages distinguish apical and laminal stops and nasals andmany have two sets of apicals and two sets of laminals Apicals include theapico-alveolar (tongue tip on the tooth ridge) consonants d t n and theapico-postalveolar or retroflex (tip on the roof of the mouth) consonantsḍṭṇ Laminals occur as laminodentals (tongue blade on the teeth) namelydtn and laminopalatals (blade on the roof of the mouth) dy ty ntilde
Australian languages generally have bilabial (b p m) and velar (g kŋ) stops and nasals as well Along the east coast languages usually haveonly one lateral but elsewhere they have two or more Most Australian lan-guages have two rhotics or r-sounds One is usually a retroflex semivowelṛ (rather like English r) and the other a flapped or trilled r
Consonant inventories for four languages illustrate some general pat-terns Wargamay is an example of an east-coast language with no contrastbetween apicals or between laminals and with one lateral Kunjen is an east-ern language with a laminal contrast but no apical contrast and with onelateral Wajarri a western language exhibits apical contrast but no laminalcontrast and has more than one lateral Pitta-Pitta is a central Australian lan-guage that contrasts both apicals and laminals and has more than one lateral
Wargamay
b d d gm n n ŋ
lṛ
w y
Kunjen
p t t ty kb d d dy g
f eth γm n n ntilde ŋ
lṛ
w y
Wajarri
p t t ṭ km n n ṇ ŋ
l l ḷr ṛ
w y
92 CHAPTER 5
Pitta-Pitta
p t t ṭ ty km n n ṇ ntilde ŋ
l l ḷ λr ṛ
w y
Two other patterns emerge from an examination of the four consonantsystems given above First contrast between voiceless and voiced stops iebetween p t k and b d g is not common though it does occur in a minority oflanguages Second fricative phonemes are rare Of the languages above onlyKunjen has fricative phonemes (f eth γ) (But in some languages stops like bare pronounced as fricatives say [f] or [v] in some phonetic contexts)
533 Word StructureAustralian languages show remarkable similarity in the way in which conso-nant and vowel phonemes combine to form words As in other Pacific lan-guages words of one syllable are extremely rare Most words contain twosyllables some more than two Words seldom begin with a vowel and se-quences of vowels are also rare Two-consonant clusters are common in themiddle of words but not initially or finally Words may end in either a conso-nant or a vowel The typical pattern is CVC(C)V(C) and words of more thantwo syllables simply build on this pattern
There are commonly restrictions on where consonants occur Typicallylaterals and rhotics do not occur in word-initial position and stops do notoccur finally Rules also govern the formation of two-consonant clusters inmedial position Here are some examples from Bandjalang showing the dis-tribution of laterals rhotics and stops as well as a limited range of medialtwo-consonant clusters (rb ntildeb ŋb mb)
Bandjalangdyadyam lsquochildrsquo babaŋ lsquograndmotherrsquoburbi lsquokoalarsquo ŋuntildeba lsquosnakersquoguluŋbay lsquoflursquo yalantilde lsquotonguersquodyimbaŋ lsquosheeprsquo balaya lsquodiersquo
There are exceptions to these constraints Anguthimri mentionedabove as an atypical Australian language for its vowel system is excep-tional in other ways as well It contrasts voiceless and voiced prenasal-ized stops and possesses five fricative phonemes It also has a phonemicglottal stop (see appendix 3) Besides these phenomena Anguthimri has
Sound Systems 93
many monosyllabic words and allows word-initial vowels and consonantclusters It does not however allow word-final consonants (except wand y) Some examples
Anguthimripweke lsquogroperrsquo paeligŋa lsquoelbowrsquokyabara lsquoalligatorrsquo iγiti lsquobrownrsquoubu lsquored gumrsquo baw lsquotoothrsquodwa lsquoeyersquo drya lsquowingrsquo
Reduplication is often used in Australian languages to form the plural ofnouns and adjectives
Dyirbalbari lsquoaxersquo baribari lsquoaxesrsquobulgan lsquobig onersquo bulganbulgan lsquobig onesrsquo
It sometimes has such other functions as intensity (Kalkatungu)diminution (Diyari) or unreality (Western Desert)
Kalkatungujagabi lsquolistenrsquo jagabijagabi lsquolisten intentlyrsquobuyud lsquohotrsquo buyudbuyud lsquoVery hotrsquo
Diyarikintala lsquodogrsquo kintalakintala lsquopuppyrsquo
Western Desertwati lsquomanrsquo watiwati lsquochild playing at
being an adultrsquo
54 OrthographiesNo Pacific languages were written before European contact9 and even to-day not all Pacific languages are written This usually means that no mis-sionaries or linguists have done sufficient work on these languages to designan orthography Languages in this category are found almost exclusively inMelanesia and Australia
Many languages in Melanesia and Australia are used for a much nar-rower range of written purposes than are other Pacific languages The mainwriters are probably linguists One reason for this has to do with the rela-tively small numbers of speakers of these languages and the fact that theygenerally write in a more widely understood language (English French orMelanesian Pidgin for example)
94 CHAPTER 5
541 General IssuesThe Latin alphabet (in which English and most European languages are writ-ten) is universally applied to the writing of Pacific languages Orthographiesfor most of the written languages of the Pacific were developed by Chris-tian missionaries during the nineteenth and twentieth centuries althoughlinguists have also made their contributions
In developing orthographies for Pacific languages missionaries andlinguists faced a number of problems that reliance on the spelling systemof say English or French could not always resolve The first of these ofcourse is that the spelling systems of these two European languages arethemselves not always consistentmdashor at least not transparently so In Eng-lish for example we now spell Fiji as Fiji but earlier explorers wrote it asFejee or Feejee the French write it as Fidji The ldquocorrectrdquo Fijian spellinghowever is Viti
There are also distinctive phonological features in Pacific languages thatlanguages like English or French do not have For these there is no ldquonaturalrdquoorthographic representation Two examples common to many parts of the re-gion are (1) the contrast between short and long vowels and (2) the glottalstop phoneme Different solutions were often found for these kinds of prob-lems in different areas For vowel length the macron (as in ā ē) has been usedin many Polynesian languages although double vowels (aa ee) are used inothers The glottal stop has most often been indicated in Polynesia by a quota-tion mark (as in Hawailsquoi) though in some parts of Melanesia letters like c or qwhich are not otherwise needed in the spelling system have been used
The problem with additional marks like apostrophes and macrons isthat because they are not perceived as ldquonormalrdquo letters they are veryoften left out by people when they are writing the language10 For examplealthough Hawaiian has both the glottal stop and the distinction betweenlong and short vowels many people do not indicate either of these distinc-tions when they write Hawaiian Thus the words pau lsquofinishedrsquo and paʔulsquolavalava sarongrsquo are often both written as pau although a more accuratewriting system (and the one recently officially re-endorsed) would write theword for lsquofinishedrsquo as pau and the word for lsquolavalavarsquo as pālsquoū
The problems have not only been technical however There are generalprinciples on the basis of which a good orthography can be developed butthere is often a certain amount of choice even after the application of thesescientific principles For example it makes equal scientific sense to write aas ā as aa or in a number of other ways (like ah in parts of Micronesia)Orthographic design in many parts of the Pacific has often revolved aroundthese areas of choice and reflects the fact that speakers of a languagemdashand
Sound Systems 95
outsidersmdashhave very strong feelings about how a language ought to be writ-ten regardless of any scientific approach to the situation
Factionalism of various kinds shows itself in spelling controversies allover the Pacific There has been a long debate in Kiribati over whether towrite brsquo and mrsquo or bw and mw for the phonemes bw and mw The NauruanLanguage Board is currently preparing a Nauruan dictionary in two differ-ent orthographies pending a final decision on spelling One of these systemsderives from the Protestant Bible translation while the other was developedby Catholics and endorsed by an earlier official body In the spelling of TokPisin in Papua New Guinea before the Second World War there were the fol-lowing competitive orthographic decisions11
g ŋLutherans g ŋCatholics g ngMethodists q g
There have also been other nonlinguistic factors at work English andFrench as international languages have considerable prestige in the Pa-cific Although linguists have their own phonetic symbols for sounds manyof these are not standard letters in the English or French writing systemsmdashβeth θ ʔ ə ŋ for example Attempts to use letters like these to represent soundsin Pacific languages are often met with resistance by speakers of those lan-guages who donrsquot want their languages to look ldquofunnyrdquo in comparison withEnglish or French
Other problems are also related to the orthographies of the prestige lan-guages In general a scientific approach to orthographic design requiresthat wherever possible each phoneme should be represented by a singleletter12 Following this principle the early missionaries used the single let-ter g to represent the phoneme ŋ (the sound written ng in English singer)in a number of Polynesian languages Pago Pago the capital of AmericanSamoa for example is pronounced paŋopaŋo This principle was extendedby Methodist and related missionaries to some other parts of Polynesia toFiji and to certain areas in Melanesia
But though this decision may follow scientific rationality there is aconflict with the spelling system of English where the letter g has a verydifferent value In Tongan for example original g was later changed tong since it was felt that Tongans learning English would be confusedby the two different values of the letter g in these two languages Manylanguages in Melanesia and Micronesia use ng for this sound but thishas led to problems of a different sort On the one hand English ngrepresents both the sound ŋ as in singer and the sounds ŋg as in fin-
96 CHAPTER 5
ger and outsiders often mispronounce words written in Pacific languageswith this letter combination (Tonga frequently being pronounced by Eng-lish speakers as if it were Tongga for example) On the other hand ifng is used for ŋ then designers of writing systems are often forced touse the somewhat unsightly three-letter combination ngg to representŋg There have then been a number of problems in the development ofspelling systems in the Pacific by no means all of them having to do withthe nature of the languages
542 Polynesia and FijiBecause of their relatively simple phonological structures the developmentof writing systems for the Polynesian languages has been a fairly straightfor-ward matter There have been different approaches to the velar nasalphoneme ŋ written g or ng and to long vowels written with macrons orwith double vowels Sometimes even in the same language some writershave used macrons and some double letters while others have ignoredvowel length altogether Māori Maaori and Maori have all had some cur-rency in New Zealand for example though the first seems now to be thepreferred spelling
The designers of the Fijian writing system fairly consistently applied theone-phoneme-one-letter principle although not without controversy13 In Fi-jian the prenasalized stops mb nd ŋg have been written with the singleletters b d and q rather than mb nd and ngg According to the same prin-ciple ŋ is written as g and eth as c (rather than the ng and th of English)Where vowel length is written the macron is used but many writers of Fi-jian ignore this feature
543 Melanesia and MicronesiaIn some parts of Melanesia the early missionaries made similar kinds of de-cisions as those made for Fijian and Polynesian languages In a number oflanguages in Vanuatu especially g is used for ŋ and in some c is used forγ Additional single symbols were created to try to adhere to this principlep and m being used to represent pw and mw Many of these languagesalong with those of the Solomons have only five vowels which caused noproblems Vowel length (where it was recognized) however was generallyrepresented by doubling vowels
Further west in the New Guinea area the Methodist traditions fromFiji and Polynesia had less influence and orthography designers have gen-erally kept fairy closely to English spelling at least as far as consonants
Sound Systems 97
are concerned In these languages for example the prenasalized stops mbnd ŋg tend to be written b d and g in word-initial position (where the pre-nasalization is fairly weak) and mb nd and ngg in other positions Thevelar nasal [ŋ] is usually written ng although in some areas where theLutheran church is strong the letter ŋ is used The occurrence of morethan one lateral in Highlands languages has required the use of two lettersto represent a single phoneme like tl dl gl and so on in addition to sim-ple l while gh is frequently used for the velar fricative γ In dealing withlanguages which have more than five phonemic vowels both digraphs(two-letter combinations) and diacritics (additional marks like accents)have been used Thus where there is a contrast between i I and e (asin English seat sit set) for example these vowels are written i icirc e or iecirc e or ii i e
The complex nature of the consonant and vowel systems of most NewCaledonian languages has forced linguists to use both diacritics and combi-nations of letters The vowels of Xacircracirccugraveugrave for example are a acirc auml e eacute egrave ecirc euml i icirc oocirc ouml u ugrave ucirc uuml and the long vowels are written by doubling these letters Writingthe consonant phonemes of Pije involves single letters (p m h w) digraphs(pw ph hm hw) and even trigraphsmdashcombinations of three letters repre-senting a single phonememdashlike phw hmw hny hng
In Micronesia digraphs are usually used to help represent complexvowel and consonant systems A number of Micronesian languages use oafor ͻ when this contrasts with o (written o) and h is often used to marklong vowels thus i represents i while ih represents i14 Digraphs and tri-graphs are also widely used in writing consonant phonemes Carolinian forexample distinguishes bw gh mw pw rh sch and tch from b g m p r sand t Long consonants are usually represented by doubling the consonant(as in ll for long l) In the case of digraphs only the first letter is doubled(mmw represents long mw)
544 AustraliaIn general the small number of vowel phonemes in Australian lan-
guages has not posed many problems for designers of orthographies Longvowels have sometimes been written as double vowels sometimes with a fol-lowing h thus a is written aa in some languages but ah in others
Decisions made about writing consonants vary but a common patternis to write retroflex sounds with a preceding r dentals with a following hand palatals with a following y palatal stops are sometimes written j InGooniyandi for example the stop and nasal phonemes given on the left be-low are written with the letters on the right
98 CHAPTER 5
Gooniyandi
Phonemes Letters
b d d ḍ dy g b th d rd j gm n n ṇ ntilde ŋ m nh n rn ny ng
Similarly multiple laterals are generally written lh l rl and ly (or lj) whilethe two rhotics are generally written r and rr
I have adopted these spelling conventions here and transliterated sym-bols in this way from sources that use phonetic symbols Note however thatthere is pressure to spell Australian languages following English conven-tions For example the Bandjalang (bantildedyalaŋ) people now choose to writetheir language name Bundjalung to avoid its possible mispronunciation asbaeligntildedyəlaeligŋ by English speakers
55 SummaryPacific languages show a great diversity of phonological systems Vocalicallythey range from Australian languages with just three short vowels to NewCaledonian languages with seventeen short vowels Consonant inventoriescan be very small and simple or extremely large and complex Some lan-guages have phonemic tone others do not Some allow a great deal ofconsonant clustering others allow none
Various social issues surround and affect the development of orthogra-phies for these languages In the remainder of this book I use the standardwriting system in italics for each language from which I give examples Inthe case of languages without a generally accepted writing system I use amodified set of phonetic symbols also in italics
Sound Systems 99
CHAPTER
6
Oceanic LanguagesGrammatical Overview
61 PronounsI use the term ldquopronounrdquo fairly loosely Oceanic languages generally haveonly one set of free pronouns but they also have one or more sets of pronom-inal forms that are more or less bound to nouns verbs or other morphemesWhile only the free forms might qualify as pronouns under a strict definitionI discuss the other forms here as well
611 PersonAlmost all Oceanic languages make a distinction between inclusive firstperson referring to the speaker and the addressee or addressees (ldquoI +yourdquo) and exclusive first person referring to the speaker and some otherindividual or individuals (ldquoI + hesheittheyrdquo) For example
Motu Mono-Alu Nakanamanga PuluwatSingular
I lau maha kinau ngaangyou oi maito niigo yeenhesheit ia elsquoa nae yiiy
PluralweINC ita maita nigita kiirweEXC ai maani kinami yaacuteaacutememyou umui maang nimu yaacuteaacutemithey idia relanalsquoi naara yiir
100
Exceptions to this general statement are found in a few languages that seemto have lost the inclusiveexclusive distinction These include the five lan-guages of the Siau family in the West Sepik Province of Papua New Guinea(Sera Sissano Ali Tumleo and Ulau-Suain) Kiribati and possibly also oneor two varieties of Fijian
Sissano KiribatiSingular
I ya ngngaiyou e ngkoehesheit i ngaia
Pluralwe eit ngairayou om ngkamiithey ri ngaiia
Very few Oceanic languages mark gender in pronouns In all the exam-ples above the third person singular refers to male or female animates aswell as to inanimates Maringe (Isabel Solomon Islands) is one of the fewOceanic languages that does have a gender distinction though it differsfrom the English one Female speakers use only one set of third personforms but male speakers use two setsmdashone referring to males and the otherin all other cases
Maringe
Male speaker Female speakerhe mana nalsquoasheit nalsquoa nalsquoathey (males) mare relsquoethey (non-males) relsquoe relsquoe
Some languages in Melanesia have no third-person pronouns at allMari (Morobe Province Papua New Guinea) is one such It uses demonstra-tives (roughly translated ldquothis onerdquo ldquothose onesrdquo) instead of pronouns likeldquohesheitrdquo or ldquotheyrdquo
612 NumberA three-way distinction between singular dual and plural number is per-haps the commonest pattern in Oceanic languages the dual number refer-ring to two and only two This pattern is found in Polynesian languages and
Oceanic Languages Grammatical Overview 101
Rotuman as well as in many languages in Melanesia and Micronesia For ex-ample
Yapese Nakanai Alsquojieuml SamoanSingular
I gaeg eau gegravenya alsquouyou guur eme gegravei lsquooehesheit qiir eia ce ia
Dualwe twoINC gadow etalua goumlrru tālsquouawe twoEXC gamow emilua goumlvu mālsquouayou two gimeew emulua goumlu lsquoouluathey two yow egirua curu lālsquoua
PluralweINC gadaed etatou gegraveveacute tātouweEXC gamaed emiteu gegraverreacute mātouyou gimeed emutou geumlveuml lsquooutouthey yaed egiteu ceacutereacute lātou
There are two common departures from this pattern A number of lan-guages in Melanesia and Micronesia show only a two-way distinction be-tween singular and plural The examples given in 611 above from MotuMono-Alu and Nakanamanga (in Melanesia) and Puluwat and Kiribati (inMicronesia) illustrate this
The other variation is quite common in Melanesia (including Fiji)though not elsewhere in the Pacific It involves a four-way distinction be-tween singular dual trial or paucal and plural Some of these languageshave a trial number which refers to three and only three
Tolai AnejomSingular hesheit ia aenDual they two dir aarauTrial they three dital aattajPlural they (gtthree) diat aara
Others have a paucal number which refers to a few (perhaps three to six orso) or to a small group in comparison with a larger group1
Paamese Nadrau FijianSingular hesheit kaie i kwayaDual they two kailue i kirau
102 CHAPTER 6
Paucal they (a few) kaitelu i kiratouPlural they (many) kaile i kira
613 FunctionsThe pronouns cited so far are known as independent pronouns They maystand alone as the answer to a question and may also act as subject of averb (though they often have an emphatic function in this usage) There arehowever other pronominal forms in many Oceanic languages although theymay not always be able to stand alone
Most Oceanic languages for example have a separate set of subjectmarkers which are formally different from the independent pronounsThese subject markers mark the person and number of the subject and usu-ally occur within the verb complex In some languages they are preverbalparticles in others prefixes to the verb In many of these languages the in-dependent pronoun is used in subject position only for emphasis Contrastthe following sentences in Lenakel
LenakelI-es-ol-aan
I-not-do-notlsquoI didnrsquot do itrsquo
Io i-es-ol-aanI I-not-do-notlsquoI didnrsquot do itrsquo lsquoIt wasnrsquot I who did itrsquo
In both sentences the person and the number of the subject are markedwithin the verb by the prefix i- lsquoIrsquo The first sentence with no independentpronoun is a neutral statement In the second however emphasis is placedon the subject lsquoIrsquo through the use of the independent pronoun io
Below are some examplesmdashin just singular and plural num-bersmdashillustrating the formal difference between independent pronouns andsubject markers The Nehan and Fijian subject markers are free preverbalparticles while the Trukese ones are verbal prefixes2
Nehan Trukese FijianIND SUBJ IND SUBJ IND SUBJ
SingularI ingo ku ngaang wuacute- o yau auyou inga ko een ke- o iko ohesheit git ke iiy e- o koya e
Oceanic Languages Grammatical Overview 103
PluralweINC ingeg ki kiich si- o keda daweEXC ingam king aacuteaacutem eacutewuacute- o keimami keimamiyou ingam kung aacuteaacutemi wo- o kemunī nīthey gisit ka iir re- o ira ra
Rather fewer Oceanic languages have formally distinct object mark-ers many using the independent pronoun in this role Above for examplewe saw the use of the Lenakel independent pronoun io lsquoIrsquo as an emphaticsubject This same form is also used in object position
LenakelR-ɨs-aamh-aan io
he-not-see-not melsquoHe didnrsquot see mersquo
Languages with distinct object pronouns are found in Melanesia and Mi-cronesia In some of these languages (like Anejom in the example below)these are free forms while in others (like Kiribati) they are suffixed to theverb3
Anejom KiribatiIND OBJ IND OBJ
SingularI antildeak ntildeak ngngai -aiyou aek yic ngkoe -ikohesheit aen yin ngaia -ia
PluralweINC akaja caja ngaira -iiraweEXC ajama camayou ajowa cowa ngkamii -ingkamithey aara ra ngaiia -iia -i
For more about the functions of both subject and object markers see section64
Virtually all Oceanic languages also have a set of possessive affixes(normally suffixes) marking the person and number of the possessor Thesediffer from independent pronouns and subject markers (though they are fre-quently identical or similar to object markers) The grammar of possessionin Oceanic languages is quite complex (refer to section 63 below) For exam-ple the Fijian possessive suffix -qu lsquomyrsquo is attached directly to certain typesof possessed nouns (like tama lsquofatherrsquo in the example below) but when usedwith nouns of other types it is attached to a possessive marker or classifier(as with vale lsquohousersquo)
104 CHAPTER 6
Fijianna tama-qou but na no-qau valethe father-my the POSS-my houselsquomy fatherrsquo lsquomy housersquo
These affixes are almost always suffixes But in a few languages they occuras prefixes in some grammatical contexts
Wayan Fijiano mna-m but m-uluthe mother-your your-headlsquoyour motherrsquo lsquoyour headrsquo
A comparison between the singular and plural independent object andpossessive pronouns in Anejom and Kiribati is given below
Anejom KiribatiIND OBJ POSS IND OBJ POSS
SingularI antildeak ntildeak -k ngngai -ai -uyou aek yic -m ngkoe -iko -mhesheit aen yin -n ngaia -ia -n(a)
PluralweINC akaja caja -ja ngaira -iira -raweEXC ajama cama -mayou ajowa cowa -mia ngkamii -ingkami -miithey aara ra -ra ngaiia -iia -i -ia
It follows from all of this that while some Oceanic languages have apronoun system as simple as that of English many have pronoun systemsof considerable complexity Table 5 lists the full set of independent objectand possessive pronouns in Anejom along with the three sets of subjectmarkers used in the aorist past and inceptive tenses to illustrate thiscomplexity
62 NOUNS AND NOUN PHRASESThe notion of parts of speech as we understand it in English does not nec-essarily apply to Oceanic languages While some Oceanic languages clearlydistinguish nouns from other parts of speech in some formal or functionalway many others do not The Fijian word tagane for example can functionas a noun meaning lsquomanrsquo as a verb meaning lsquoto be malersquo and as an adjectivemeaning lsquomalersquo
Oceanic Languages Grammatical Overview 105
Table 5 Anejom Pronouns
1 INC 1 EXC 2 3IndependentSingular mdash antildeak aek aak aen aanDual akajau ajamrau ajourau aarauTrial akataj ajamtaj ajoutaj aattajPlural akaja ajama ajowa aara
ObjectSingular mdash ntildeak yic -c yin -nDual cajau camrau courau rauTrial cataj camtaj coutaj ettajPlural caja cama cowa ra
PossessiveSingular mdash -k -m -nDual -jau -mrau -mirau -rauTrial -taj -mtaj -mitaj -ttajPlural -ja -ma -mia -ra
Subject (aorist)Singular mdash ek na etDual tau ekrau erau erauTrial taj ettaj ettaj ettajPlural ta ekra eka era
Subject (past)Singular mdash kis as isDual tus eris arus erusTrial tijis eris atijis etijisPlural eris ekris akis eris
Subject (inceptive)Singular mdash ki an intildeiyiDual tu ekru aru eruTrial tiji etiji atiji etijiPlural ti ekri aki eri
106 CHAPTER 6
FijianE lako mai na tagane oyā (tagane = noun)he come here the man thatlsquoThat man is comingrsquo
E tagane na vuaka oqō (tagane = verb)he male the pig thislsquoThis pig is malersquo
E mate na vuaka tagane (tagane = adjective)he die the pig malelsquoThe boar diedrsquo
In this and subsequent sections when I use the word ldquonounrdquo I am referringto words functioning as nouns in a particular context For our purposesthen tagane is a noun in the first Fijian sentence above though not in theother two
621 Form of the NounNouns in Oceanic languages are generally invariable in form That is a noundoes not change form to mark singular and plural nor generally do nounstake prefixes and suffixes (apart from possessive affixes discussed later) Fi-jian vuaka and Hawaiian pualsquoa for example both mean lsquopigrsquo or lsquopigsrsquo
In languages of this type plurality is expressed either by a separatemorpheme in the noun phrase (see 625) or by a subject or object marker inthe verb complex Often a combination of strategies is used as in the Vin-mavis example below in which the noun itself (matoro lsquoold manrsquo) remainsinvariable
VinmavisMatoro i-fwelemoldman he-comelsquoThe old man camersquo
Matoro ar at-fwelemoldman PL they-comelsquoThe old men camersquo
There are however some exceptions to the generalization that nounsare invariable in form First in some languages of Polynesia and Melanesiathere is a small set of nouns referring to human beings that form the pluralby a change in the position of stress or by partial reduplication as in Motuor by lengthening a vowel as in Māori and Hawaiian
Oceanic Languages Grammatical Overview 107
Singular PluralMotutau lsquomanrsquo tatauhahiacutene lsquowomanrsquo haacutehinemero lsquoboyrsquo memerokekeacuteni lsquogirlrsquo keacutekeni
Māoritangata lsquomanrsquo tāngatatupuna lsquoancestorrsquo tūpunatuahine lsquosisterrsquo tuāhine
Hawaiianluahine lsquoold womanrsquo luāhinekahuna lsquopriestrsquo kāhunakupuna lsquograndparentrsquo kūpuna
In Kiribati vowel lengthening also occurs in nouns but it indicates genericreference rather than plurality
Kiribatite tina lsquo(the) motherrsquo tiina lsquomothers in generalrsquote ika lsquo(the) fishrsquo iika lsquofish in generalrsquote ben lsquo(the) coconutrsquo been lsquococonuts in generalrsquote bong lsquo(the) dayrsquo boong lsquodays in generalrsquote biti lsquo(the) knifersquo biiti lsquoknives in generalrsquo
In Rotuman the long form of a noun (see 511 above) marks a noun asdefinite Indefinite nouns occur in the short form4
RotumanFamori lsquoeapeople saylsquoThe people sayrsquo
Famoumlr lsquoeapeople saylsquo(Some) people sayrsquo
There are also some languages geographically and genetically fairlywidespread that mark plurality of nouns by a prefix or a suffix Among theseare the non-Oceanic languages Palauan and Chamorro
Singular PluralPalauanchad lsquopersonrsquo rę-chad
108 CHAPTER 6
kangkodang lsquotouristrsquo rę-kangkodangsęchel-ik lsquomy friendrsquo rę-sęchel-ik
Chamorroestudiante lsquostudentrsquo man-estudiantepalelsquo lsquopriestrsquo mam-alelsquosaina lsquoparentrsquo mantilde-aina
Some languages in Vanuatu have fairly complex pluralization strategiesIn Sye for example there is a general plural prefix ovn- (This varies slightlyaccording to the following consonant) Kinship terms without possessive suf-fixes (like namou lsquomotherrsquo) may take this prefix and also the suffix -me theymust take one of these Kinship terms with possessive suffixes (like asu-glsquomy husbandrsquo and ma-n lsquoher brotherrsquo) must take the suffix -me and may takethe prefix r(o)- Thus
Sye
Singular Pluralkuri lsquodogrsquo ovn-kurinakeh lsquoaxersquo ov-nakehneteme lsquopersonrsquo ovo-temenamou lsquomotherrsquo ov-namou namou-me ov-namou-measu-g lsquomy husbandrsquo asu-g-me r-asu-g-mema-n lsquoher brotherrsquo ma-n-me ro-ma-n-me
In Anejom nouns beginning with n or in drop this in the plural Nounsreferring to humans must take a plural prefix elpu- those referring to thehigher animates may take this prefix other nouns take no plural prefix
Anejom
Singular Pluralnatamantilde lsquomanrsquo elpu-atamantildenatimi lsquopersonrsquo elpu-atiminepcev lsquosharkrsquo elpu-epcev epcevincai lsquotreersquo caiinhat lsquostonersquo hat
Some Oceanic languages make no formal distinction between nounsand say verbs or adjectives Those that do make this distinction (and alsosome that do not) have one or more nominalizersmdashmorphemes that con-vert verbs or adjectives into nouns Some examples are presented below
Lenakelaklha lsquostealrsquo i-aklha lsquothiefrsquo
Oceanic Languages Grammatical Overview 109
n-aklha-aan lsquotheft robberyrsquok-aklha lsquohouse-breaking toolrsquo
Mokilesekaraja lsquoexplainrsquo karaja-poa lsquoexamplersquowia lsquomakersquo wia-poa lsquoconstructionrsquowoaroai lsquoto lastrsquo woaroai-n lsquodurationrsquo
Māorikimi lsquoto seekrsquo kimi-hanga lsquoa searchrsquonoho lsquositrsquo noho-anga lsquoseatrsquoinu lsquoto drinkrsquo inu-manga lsquoa drinkrsquo
622 ArticlesArticles are morphemes marking the class or reference of a noun InEnglish the article the marks a noun as definite while aan marks it assingular indefinite in French un and le mark singular masculine nouns(indefinite and definite respectively) while une and la mark singularfeminine nouns
Generally speaking the languages of the New Guinea mainland and theislands of Papua and those spoken in Vanuatu have no articles5 Examples
ManamTamoata roa tolsquoa i-tilsquoin-iman hisspouse hisolderbrother he-show-herlsquoThe man showed his wife to his older brotherrsquo
KilivilaE seki Kilagola yena guyauhe give Kilagola fish chieflsquoThe chief gives Kilagola the fishrsquo
SyeNatmonuc y-omonki nacavechief heDISTANTPAST-drink kavalsquoThe chief drank (the) kavarsquo
Most of the remaining Oceanic languagesmdashthose of the islands to thenortheast of New Guinea the Solomon Islands New Caledonia MicronesiaFiji and Polynesiamdashdo have articles although there are some exceptions
Fijian languages generally have two articles In Standard Fijian o is theproper article and is used before pronouns proper nouns (names of specificpeople or places) and some kinship terms Na is the common article and is
110 CHAPTER 6
used before other nouns that are definite in some sense6 Indefinite nouns (likeyaqona in the second example below) are not marked by articles Examples
FijianE gunu-va na yaqona o Seruhe drink-TRANS the kava thePERSONAL SerulsquoSeru is drinking the kavarsquo
E gunu yaqona o Seruhe drink kava thePERSONAL SerulsquoSeru is drinking kavarsquo
Polynesian languages have a slightly larger number of articles InHawaiian for example the articles are
Hawaiianka ke7 definite article singular lsquothersquonā definite article plural lsquothersquohe indefinite article lsquoarsquoa personal article
Mokilese and Yapese provide illustrations of different kinds of Microne-sian article systems In Mokilese a noun may occur with no article (ordemonstrative) The reference is usually generic
MokileseMahnsang kin wia ahr paj in pohn suhkoabird HABITUAL make their nest in top treelsquoBirds build their nests in treetopsrsquo
A koah kak wiahda warQUESTION you can build canoelsquoCan you build canoesrsquo
When the reference is specific but indefinitemdashthe addressee does not knowwhich individual is being referred tomdashMokilese nouns take as a suffix the ap-propriate numeral classifier (see 625 below) in the singular and -pwi in theplural
MokileseNgoah kapang lih-men oI see woman-CLASSIFIER therelsquoI saw a woman therersquoNgoah kapang lih-pwi oI see woman-aPL therelsquoI saw women therersquo or lsquoI saw some women therersquo
Oceanic Languages Grammatical Overview 111
When the reference is both specific and definite the suffix -o (sometimes-u)is used
MokileseNgoah kapang lih-oI see woman-thelsquoI saw the womanrsquo
Yapese is similar to Fijian or the Polynesian languages in that it hasthree articles all of which come before the noun fa definite ba indefinitesingular and ii which is used optionally before personal names
Yapesefa rea kaarroothe SG carlsquothe carrsquo
ba kaarrooa carlsquoa carrsquo
ii Tamag (or just Tamag)thePERSONAL TamaglsquoTamag (a manrsquos name)rsquo
Most languages of the Bismarck Archipelago and the Solomon Islandsalso have a small number of articles that precede the noun Gender distinc-tion is not uncommon Tolsquoabalsquoita for example has a common article ngaand two personal articles tha (used with masculine names) and ni (usedwith feminine names)
Tolsquoabalsquoitanga lsquoai lakoo kithe wood this PLlsquothe firewoodrsquo
ai tha Gereawife theMASCULINE GerealsquoGerearsquos wifersquo
maka ni lsquoOinafather theFEMININE lsquoOinalsquoOinarsquos fatherrsquo
The most complex article systems are those of New Caledonia In theselanguages articles precede the noun and markmdashamong other featuresmdashdef-
112 CHAPTER 6
initeness number and gender Drehu has the following articles and article-like particles
Drehula definite near speaker visiblelai definite near addressee visiblelo definite not present or visibleketre indefinite singularxaa indefinite non-singularisa lsquoeachrsquoitre o paucalnoumljei pluralhaa collective
Cegravemuhicirc has an even more complex system Its articles distinguish gen-dermdashfeminine and nonfeminine which both treat the noun as a person orindividual as well as neuter which treats the noun as a thing or idea)numbermdashsingular dual and plural and referencemdashdefinite indefinite andneutral (Neutral reference marks the noun as a noun without specifyingwhether it is definite or indefinite)
CegravemuhicircNeutral Definite Indefinite
Singular Nonfeminine pā pācɛ pāliFeminine ɛ ɛcɛ ɛgiNeuter ā ācɛ āli
Dual Nonfeminine lūpwɔ lūpwɔcɛ lūpwɔliFeminine lū lū cɛ lū li
Plural Nonfeminine lēpwɔ lēpwɔcɛ lēpwɔliFeminine lē lēcɛ lēliNeuter ni cɛ li ili
623 DemonstrativesDemonstratives are words that locate the noun in space andor time gen-erally with reference to the speaker and the addressee though sometimeswith reference to some other focus English has a simple two-way distinction(between thisthese and thatthose) and this system is found in a fewOceanic languages in Melanesia
Manam Maringengae gne lsquothisrsquongaedi gre lsquothesersquo
Oceanic Languages Grammatical Overview 113
ngara gno lsquothatrsquongaradi gro lsquothosersquo
Almost universal in Oceanic languages however is a three-way direc-tionallocationaltemporal distinction in demonstratives corresponding tothe three grammatical persons The three categories are often referred toas proximatemdashnear the speaker and corresponding to the first person (thespeaker) intermediatemdashnear the addressee and corresponding to the sec-ond person (the person spoken to) and distantmdashaway from both speakerand addressee and corresponding to the third person (some other person orthing)
Some languages simply mark this distinction without specifying number
Motu Fijian Māoriina (o)qō nei PROXIMATE lsquothisrsquoena (o)qori na INTERMEDIATE lsquothisthatrsquouna (o)yā ra DISTANT lsquothat (yonder)rsquo
Others however not only make the three-way contrast but also indicate sin-gular and plural
Nakanai Kiribati Rotumanaleie aei telsquoisi PROXIMATE lsquothisrsquoini aikai lsquoi lsquothesersquo
alele anne talsquoa INTERMEDIATE lsquothisthatrsquoene akanne lsquoo lsquothesethosersquo
aleio arei taelige DISTANT lsquothat (yonder)rsquounu akekei lsquoie lsquothose (yonder)rsquo
There are further complications in some languages Tolsquoabalsquoita for ex-ample has not only a regular distant demonstrative labaa but two othersspecifying vertical orientation loo lsquothat yonder and higher uprsquo and fuulsquothat yonder and lower downrsquo Anejom possesses not only the three-way dis-tinction noted above but also has a set of anaphoric demonstratives whichmark a noun as having been previously referred to Example
Anejomniom iyiikihouse thatlsquothat house (the one I was talking about before)rsquo
In addition Anejom distinguishes number in demonstratives and so has thefollowing
114 CHAPTER 6
AnejomSingular Dual Plural
Proximate inintildeki erantildeki ijintildekiIntermediate enaanai mdash ijeknaaDistant enaikou erantildekou ijeknaikouAnaphoric iyiiki eraaki ijiiki (recent)
ijekentilde (distant)
624 AdjectivesI mentioned earlier that there is often difficulty in rigidly assigning a wordto a specific part of speech in Oceanic languages This is especially apparentin the distinction or lack of it between verbs and adjectives
Words that translate into English as adjectives generally have two func-tional possibilities in most Oceanic languages First they may occur withina noun phrase almost always following the noun which they modify
Fijianna waqa levu
the canoe biglsquothe big canoersquo
Samoanlsquoo le teine putaFOCUS the girl fatlsquothe fat girlrsquo
Second and more frequently adjectives function as stative verbs That isthey function in the same way as other intransitive verbs (being marked forsubject tense and so on) but they express a state rather than an actionwith the subject being the experiencer of that state
FijianE levu na waqait big the canoelsquoThe canoe is bigrsquo
SamoanUa puta le teineSTATIVE fat the girllsquoThe girl is fatrsquo
Many languages in Melanesia however do have a category of adjectivesthat differs from the category of stative verbs although both of these cate-
Oceanic Languages Grammatical Overview 115
gories include words that would translate as adjectives Lenakel for exam-ple has a set of stative verbs similar to those illustrated for Fijian andSamoan vɨt lsquogoodrsquo and esuaas lsquosmallrsquo may function as adjectives followingthe noun in a noun phrase
LenakelR-ɨm-aamh nimwa v ɨ t kerhe-PAST-see house good onelsquoHe saw a good housersquo
Kova esuaas ka r-ɨ s-apul-aanchild small that he-not-sleep-notlsquoThe small child is not asleeprsquo
They may also occur as the head of a verb complex taking prefixes markingsubject and tense aspect just like any nonstative verb (compare the behaviorof vɨt and esuaas in the examples below with that of aamh lsquoseersquo and apullsquosleeprsquo above)
LenakelNimwa taha-n r-ɨm-vɨt akɨnhouse POSS-his it-PAST-good verylsquoHis house wasused to be very nicersquo
Kova ka r-ɨs-esuaas-aanchild that he-not-small-notlsquoThat child is not smallrsquo
There is however a set of words that can only be adjectives like vi lsquonewrsquoand ituga lsquoforeignrsquo These also follow the noun in a noun phrase
LenakelR-ɨm-ol nimwa vihe-PAST=make house newlsquoHe built a new housersquo
Nɨkava ituga r-ɨs-vɨt-aankava foreign it-not-good-notlsquoAlcohol (lit foreign kava) is not goodrsquo
Words in this category never function as stative verbs and utterances likethe following ones are unacceptable
LenakelNimwa r-(ɨm)-vihouse it-(PAST)-new
116 CHAPTER 6
Nɨkava r-ɨs-ituga-aankava it-not-foreign-not
625 Numerals and QuantifiersTwo classes of words or morphemes relate to counting Numerals are exactnumbers in a counting system one two three four and so on Oceaniclanguages exhibit a range of numeral systems the commonest are simpledecimal (base 10) or quinary (base 5) systems but there are variations onthese systems and other systems are also represented (see chapter 11 fora detailed discussion) Quantifiers are morphemes that mark grammaticalnumber (singular dual plural) or express less mathematically exact quanti-ties like ldquosomerdquo ldquomanyrdquo ldquofewrdquo ldquoallrdquo and so forth
In many Oceanic languages numerals and quantifiers function as sta-tive verbs The following Fijian examples illustrate this
FijianE moce na gonehe sleep the childlsquoThe child sleptis sleepingrsquo
E dua na gonehe one the childlsquo(There is) one childrsquo
Two features of such systems are (1) that a noun modified by a numeraloccurs in what is effectively a relative clause in the sentence and (2)that numerals above one usually take singular rather than plural subjectmarkers
FijianErau moce e rua na gonetheytwo sleep he two the childlsquoThe two children sleptare sleepingrsquo
Such systems are common in Polynesia and are also found in some lan-guages in Melanesia
TahitianlsquoUa holsquoi mai na talsquoata lsquoe toruPAST return here thePL person itis threelsquoThree people came back herersquo
Oceanic Languages Grammatical Overview 117
AnejomA noupan is ithii is amen a natimi is esejat time itPAST one itPAST live SUBJECT person PAST threelsquoOnce upon a time there were three peoplersquo
In most other Oceanic languagesmdashwhich tend to be those that distin-guish adjectives from stative verbs (like Lenakel in 624 immediatelyabove)mdashnumerals and quantifiers function much like adjectives That isthey occur within the noun phrase not as stative verbs For example
Manam Lenakelaine rua peravɨn (mil) kiuwoman two woman (DUAL) twolsquotwo womenrsquo lsquotwo womenrsquo
lsquoaleti lsquoolsquoolsquoo neram ituga asuul (miin)whiteman many people foreign many (PL)lsquomany Europeansrsquo lsquomany foreignersrsquo
This is perhaps the commonest pattern among the languages of Melanesiaand it is also found in a few Micronesian languages
There are in many of these languages however vestiges of an earliersystem in which the numerals were once stative verbs Compare the Vin-mavis and Lenakel numerals for two through five with the Proto Oceanicforms from which they originate8
Proto Oceanic Vinmavis Lenakellsquotworsquo rua iru kiulsquothreersquo tolu itl kɨsillsquofourrsquo vati ifah kuvɨrlsquofiversquo lima ilim katilum
Here we can see that the roots of the numerals have something in front ofthem i in Vinmavis k (+ vowel) in Lenakel In Vinmavis i- is a third personsingular non-future verbal prefix and in Lenakel k (+ vowel) is a third per-son non-singular verbal prefix Many languages of this type may once havetreated numerals as stative verbs but over time the verbal prefix has be-come attached to the numeral and the numeral has lost its verbal nature
The third kind of system involves what are known as numeral clas-sifiers Some Micronesian languages have an elaborate system of theseclassifiers and they are perhaps the best known representatives of this typealthough such classifiers also occur in the Admiralty Islands languages AsRehg says of Ponapean ldquoEvery concrete noun in Ponapean belongs to one ormore classes When we use a numeral with a noun an appropriate numeral
118 CHAPTER 6
classifier must be employed More simply stated the choice of the numeralsystem one uses is dependent upon what one is countingrdquo (Rehg 1981 125)
Here are three such numeral systems in Ponapean with the forms ofthe numerals one through nine The words in the second column are usedwith the word mwutin lsquoheap or pile ofrsquo Those in the third column are usedto count stalks of things And those in the last column are used for countingslices or chips of something
Ponapeanlsquoheaps ofrsquo lsquostalks ofrsquo lsquoslices ofrsquo
1 emwut osop edip2 riemwut riasop riadip3 silimwut silisop silidip4 pahmwut pahsop pahdip5 limmwut limisop limadip6 wenemwut wensop wenedip7 isimwut isisop isidip8 walimwut welisop welidip9 duwamwut duwasop duwadip
Examples
Ponapeanmwutin dihpw pahmwutpileof grass four-CLASSIFIERlsquofour piles of grassrsquo
sehu pah-sopsugarcane four-CLASSIFIERlsquofour stalks of sugarcanersquo
dipen mei pah-dipsliceof breadfruit four-CLASSIFIERlsquofour slices of breadfruitrsquo
As you can see from these examples the numeral is made up of a mor-pheme representing the number itself (sili- lsquothreersquo pah- lsquofourrsquo etc) anda suffix which is the classifier Ponapean has twenty-nine such classifierswhich include the following (the first three being those exemplified above)
-mwut used to count heaps or piles-sop used to count stalks-dip used to count slices chips or shavings of something-pak used to count times
Oceanic Languages Grammatical Overview 119
-pit used to count strips or strands of something-mwodol used to count small round objects-pali used to count body extremities-pwoat used to count long objects-men used to count animate beings
It has as well a general classifier -u which can be used with a range of nounsThese classifiers may also be used without any numeral in which case they
functionas indefinitearticles (comparesection622 in relation toMokilese)
Ponapeanpwihk men tuhke pwoatpig CLASSIFIER tree CLASSIFIERlsquoa treersquo lsquoa pigrsquo
Other languages in Micronesia with elaborate systems of numeralclassifiers include Kiribati (with sixty-six classifiers) Ulithian (forty-three)Trukese Nauruan and Yapese In contrast Mokilese has only four classi-fiers and Marshallese only vestiges of a classifier system Kosraean mayhave never had a classifier system at all (See Bender and Wang [1985 79]for a brief discussion of this)
The languages of the Kilivila family in the Trobriand Islands in PapuaNew Guinea (Kilivila Muyuw and Budibud) have sets of classifiers like thoseof Ponapean but they are used with other items in noun phrases as well (seesection 626) Other Oceanic languages have numeral classifiers but thesesystems are more limited than the Micronesian ones Some Polynesian lan-guages fall into this category Tongan for example requires the classifiertoko when numerals refer to persons or animals
Tonganha kau faifekau lsquoe toko fitua PL minister itis CLASSIFIER sevenlsquoseven ministersrsquo
Other quantifiers behave similarly Compare the following Tongan phrasewith the one above
Tonganha kau faifekau lsquoe toko fihaa PL minister itis CLASSIFIER howmanylsquohow many ministersrsquo
626 Noun Phrase StructureAs a general rule articles (where they occur) precede the noun in a nounphrase in Oceanic languages while adjectives and demonstratives follow
120 CHAPTER 6
the noun The position of numerals and quantifiers is more variable In somelanguages these precede the noun in others they follow it In the examplesbelow the head noun is underlined to illustrate these patterns
Labugwa kege agravenigrave hanocirc anamocirc maipi lene
canoe small one house big five thislsquoa small canoersquo lsquothese five big housesrsquo
Banonina tavana kota numa ghoom bangana bubuPL person all house new big redlsquoall peoplersquo lsquothe big new red housersquo
Tolsquoabalsquoitaroo wela loo ki nga fau balsquoitatwo child this PL the stone biglsquothose two childrenrsquo lsquoathe big stonersquo
Port Sandwichnavuumls xavoi minac ngail pwici isa-n raibow real other PL all POSS-his onlylsquoall his other real bows onlyrsquo
Ponapeanpwutak reirei sili-men-oboy tall three-CLASSIFIER-thatlsquothose three tall boysrsquo
Kiribatiteni-ua te boki akannethree-CLASSIFIER the book thoselsquothose three booksrsquo
Fijianna wai batabatā na vinivō damudamu oqōthe water cold the dress red thislsquo(the) cold waterrsquo lsquothis red dressrsquo
Tahitiante mau pōtilsquoi purotu tē-ra talsquoata lsquoinothe PL girl beautiful the-that man badlsquothe beautiful girlsrsquo lsquothat bad manrsquo
The Kilivila language has a system of classifiers similar to but muchricher than the Bantu languages of Africa with close to two hundred dif-
Oceanic Languages Grammatical Overview 121
ferent classifiers altogether Not only numerals but also demonstrativesand adjectives have to be attached to a classifier The examples belowillustrate the use of the classifiers to lsquomale humansrsquo and bwa lsquotrees orwooden thingsrsquo
Kilivilatau m-to-na to-kabitam
man this-CLASSIFIER-this CLASSIFIER-intelligentlsquothis intelligent manrsquo
ma-bwa-si-na bwa-tolu kaithis-CLASSIFIER-PL-this CLASSIFIER-three treelsquothese three treesrsquo
63 Possessive ConstructionsIn virtually all Oceanic languages the grammar of possession is more com-plex than it is in English (as mentioned briefly in chapter 2) In this discus-sion of possession we look first at those languages that most closely reflectthe reconstructed Proto Oceanic system then at major departures from thissystem
631 Possessive Constructions Similar to Proto OceanicThe languages most closely reflecting the original Proto Oceanic system ofpossession are found in parts of Island Melanesia especially the more east-erly parts of this region (including Fiji) These languages indicate whetherpossession is direct or indirect and then discriminate between several dif-ferent types of indirect possession
In direct possession the possessive pronoun is attached directly to thepossessed noun These constructions generally encode a semantic relation-ship between the possessor and the possessed noun that has been referredto as close or subordinate or inalienable They most commonly imply thatthe possessor has little if any control over the fact of possession and are typ-ically used with normally irremovable and integral parts of the body and ofthings and with all or some kinship terms For example
Paamesenati-n mete-nchild-hisher eye-hisheritslsquohisher childrsquo lsquohisherits eyersquo
122 CHAPTER 6
Fijianna tina-qu na ulu-quthe mother-my the head-mylsquomy motherrsquo lsquomy headrsquo
In indirect possession on the other hand the possessive pronoun isnot attached to the possessed noun but rather to a separate morphemethat I refer to as a possessive marker These constructions generally en-code a relationship between possessor and possessed that can be calledremote or dominant or alienable They most commonly imply that the pos-sessor has control either over the possession itself or at least over the fact ofpossession Such constructions are typically used with items of disposableproperty nominalized verbs of which the possessor is the underlying sub-ject and nouns that the possessor owns or controls in some way or another
The languages with which I am dealing in this section have a smallnumber of subclasses of indirect possession each with its own possessivemarker Paamese and Fijian for example have the following markers9
Paameseaa- food passivemo- drink or for domestic useso- social relationship determined by law or customono- general active
Fijianke- food passiveme- drinkno- general active
Some examples
Paameseauh aa-k ipu aa-myam POSSFOOD-my loss POSSPASSIVE-yourlsquomy yam (to eat)rsquo lsquoyour lossdisadvantagersquo
oai mo-m aisin mo-nwater POSSDRINK-your clothes POSSDOMESTIC-hislsquoyour water (to drink)rsquo lsquohisher clothesrsquo
meteimal so-m telai ono-mvillage POSSCUSTOM-your axe POSSGENERAL-yourlsquoyour villagersquo lsquoyour axersquo
Oceanic Languages Grammatical Overview 123
Fijianna ke-mu madrai na ke-na itukutukuthe POSSFOOD-your bread the POSSPASSIVE-his reportlsquoyour bread (to eat)rsquo lsquohis report (the one made
about him)rsquo
na me-qu bia na no-na valethe POSSDRINK-my beer the POSSGENERAL-his houselsquomy beer (to drink)rsquo lsquohis housersquo
When the possessor is a noun what is known as a construct suffix(abbreviated here CS) is often added to the possessed noun in a directconstruction and to the possessive marker in an indirect construction InPaamese for example the construct suffix is -n
Paamesemete-n huli kailueye-CS dog DUALlsquothe two dogsrsquo eyesrsquo
vakili one-n iseicanoe POSSGENERAL-CS wholsquowhose canoersquo
The distinction between direct and indirect possession and between thevarious types of indirect possession depends partly on the semantics of thepossessed noun and partly on the nature of the relationship between the pos-sessor and the possessed It follows therefore that at least some nouns mayparticipate in more than one kind of possessive construction depending onthe nature of that relationship For example
Fijianna yaca-qu na no-qu yacathe name-my the POSSGENERAL-my namelsquomy namersquo lsquomy namesakersquo
na ke-na niu na me-na niuthe POSSFOOD-his coconut the POSSDRINK-his coconutlsquohis coconut (meat to eat)rsquo lsquohis coconut (water to drink)rsquo
na ke-mu itaba na no-mu itabathe POSSPASSIVE-your photo the POSSGENERAL-your photolsquoyour photo (the one taken ofyou)rsquo
lsquoyour photo (the one you tookor have)rsquo
124 CHAPTER 6
632 Simplification of Indirect PossessionOne common departure from this original system the loss of contrast betweensome or all of the half-dozen or so indirect possessive subtypes is widespreadin western Melanesia and occurs also in parts of Micronesia (for example inYapese and Kiribati) Many of the languages of New Guinea and Solomon Is-lands distinguish direct and indirect constructions but have only two indirectpossessive markers One of these refers to food and drink (and often to itemsinvolved in producing or cooking food) and the other to all other alienablepossessions In Manam the markers are lsquoana- (food and drink) and ne- (other)
Manam
mata-ng tama-gueye-your father-mylsquoyour eyersquo lsquomy fatherrsquo
bang lsquoana-gu suru lsquoana-θtaro POSSFOOD-my soup POSSFOOD-hislsquomy tarorsquo lsquohis souprsquo
uma lsquoana-ng lsquoaula lsquoana-gugarden POSSFOOD-your fishhook POSSFOOD-mylsquoyour gardenrsquo lsquomy fishhookrsquo
lsquousi ne-gu mata ne-dalavalava POSSGENERAL-my custom POSSGENERAL-ourINClsquomy lavalavarsquo lsquoour customrsquo
Other languagesmdashdistributed somewhat randomly throughout thisareamdashsimply contrast direct and indirect constructions with no subclassifi-cation of indirect possession
Syenoru-g etme-nhand-my father-hislsquomy handrsquo lsquohisher fatherrsquo
nimo horu-g nup horo-mhouse POSS-my yam POSS-yourlsquomy housersquo lsquoyour yamrsquo
Kiribatitina-na kuni-umother-his skin-mylsquohis motherrsquo lsquomy skinrsquo
Oceanic Languages Grammatical Overview 125
a-na boki a-u benPOSS-his book POSS-my coconutlsquohis bookrsquo lsquomy coconutrsquo
633 Development of Classifier SystemsBy contrast some Oceanic languages have developed a complex systemof classifiers (similar to those discussed in relation to the numerals) tomark categories of indirect possession10 Many Micronesian languagesfall into this category as do a few in Melanesia (like Iaai in the LoyaltyIslands)
Ponapean like almost all Oceanic languages distinguishes direct andindirect constructions
Ponapeanmoange-i nime-i uhpwhead-my CLASSIFIER-my coconutlsquomy headrsquo lsquomy drinking coconutrsquo
Indirectly possessed nouns belong to a number of different classes in Pona-pean the members of each class usually having some semantic feature thatdistinguishes them from other nouns Possessive classifiers mark the nounas belonging to a particular class and possessive suffixes and the constructsuffix are attached to these classifiers There are more than twenty posses-sive classifiers in Ponapean some of which are
PonapeanClassifier Used with nouns referring tokene- edible thingsnime- drinkable thingssapwe- landwere- vehicleskie- things to sleep onipe- things used as coveringspelie- peers counterparts opponentsmware- garlands names titlesnah- small or precious things and people or things over
which the possessor has a dominant relationship
There is also a general classifier ah- which is used with nouns that do notfall into any other class Examples
Ponapeanah-i seht lsquomy shirtrsquoah-i pwutak lsquomy boyfriendrsquo
126 CHAPTER 6
ah-i mahi lsquomy breadfruit treersquoah-i rong lsquomy newsrsquo
Many nouns may occur with more than one classifier with slight seman-tic changes So the noun pwihk lsquopigrsquo may be possessed with the dominantclassifier nah- the general classifier ah- and the edible classifier kene- eachwith different meanings
Ponapeannah-i pwihk lsquomy (live) pigrsquoah-i pwihk lsquomy (butchered) pigrsquokene-i pwihk lsquomy pork my pig (as food)rsquo
The system in Iaai is similar to that of Micronesian languages like Pon-apean Kinship terms body parts and certain other nouns closely related tothe possessor are directly possessed
Iaaihinyouml-k ba-nmother-my head-hislsquomy motherrsquo lsquohisher headrsquo
hwakeci-m i-fuuc-incustom-your NOM-speak-hislsquoyour customrsquo lsquohisher way of speakingrsquo
But there is also quite a large number of markers used in indirect con-structions
Iaaia- foodbele- drinkhanii- something caught (eg through hunting or fishing)houmlne- a contributionhwa- a noiseiie- a piece of something to chewii- landdee- a roadhnacircacirc- something done to oneanyi- general (none of the above)
Examples include
Iaaianyi-k thaan a-n koumlnyingCLASSIFIER-my chief CLASSIFIER-his tarolsquomy chiefrsquo lsquohis taro (to eat)rsquo
Oceanic Languages Grammatical Overview 127
bele-n trii hanii-ny wacircacircCLASSIFIER-his tea CLASSIFIER-his fishlsquoher tea (to drink)rsquo lsquohis fish (which he caught)rsquo
There are also specific possessive markers in Iaai These are derived fromnouns and are used to indicate possession of those same (or related) nouns
Iaaiumwouml-k uma lsquomy housersquonuu-k nu lsquomy coconut treersquohuu-k hu lsquomy boatrsquowaii-k wai lsquomy reefrsquo
634 Loss of Direct ConstructionsA few languages in the New Guinea area and Rotuman and all the Polyne-sian languages have for the most part lost the distinction between directand indirect constructions Only indirect constructions are used11 In Labufor example there is only one set of possessive pronouns and it is used withall nouns
Labuyecirc na ana yecirc na hanocircyou your mother you your houselsquoyour motherrsquo lsquoyour housersquo
Rotuman has two indirect possessive-markers lsquoe(n) used with possessednouns that refer to food drink a personrsquos turn at doing something and withsome nouns to do with contests and challenges and lsquoo(n) used with all othernouns including kinship terms and nouns referring to parts of things
Rotumanlsquoe-n lsquoalsquoana lsquoalsquoan lsquoe le FauholiPOSSFOOD-his taro taro POSSFOOD thePERSONAL Fauholilsquohis tarorsquo lsquoFauholirsquos tarorsquo
lsquoo-n lelelsquoa lsquoo-n lsquoalaPOSSGENERAL-his children POSSGENERAL-his teethlsquohis childrenrsquo lsquohis teethrsquo
lsquoo-n lsquoeap lsquoeap lsquoo le FauholiPOSSGENERAL-his mat mat POSSGENERAL thePERSONAL
Fauholilsquohis matrsquo lsquoFauholirsquos matrsquo
128 CHAPTER 6
Almost all Polynesian languages (except Niuean and Takuu) have re-tained the dichotomy between inalienable (or subordinate) and alienable(or dominant) possession but this is expressed by two different indirectconstructions Generally speaking inalienable or subordinate possessionis expressed by a possessive morpheme based on the vowel o whilealienable or dominant possession is expressed with the vowel a For ex-ample
Samoanlsquoo lo-lsquou tama lsquoo le ulu o TavitaFOCUS POSS-my father FOCUS the head POSS DAVIDlsquomy fatherrsquo lsquoDavidrsquos headrsquo
lsquoo la-lsquou talsquoavale lsquoo le naifi a TavitaFOCUS POSS-my car FOCUS the knife POSS Davidlsquomy carrsquo lsquoDavidrsquos knifersquo
Nukuoro12
to -no potu te potu o SoanPOSS-his wife the wife POSS Johnlsquohis wifersquo lsquoJohnrsquos wifersquo
ta-na naivi te naivi a SoanPOSS-his knife the knife POSS Johnlsquohis knifersquo lsquoJohnrsquos knifersquo
As in most other languages we have looked at there are many examplesof the same noun being possessed in both constructions with a concomitantsemantic difference
Nukuoroto-ku ngavesi ta-ku ngavesilsquoPOSS-my box lsquoPOSS-my boxlsquomy coffinrsquo lsquomy storage boxrsquo
to-no potopoto ta-na potopotoPOSS-his short POSS-his shortlsquohis shortness (perma-nent condition)rsquo
lsquohis shortness (temporary condition aswhen hunched over)rsquo
te kkai o Vave te kkai a Vavethe story POSS Vave the story POSS VavelsquoVaversquos story (told abouthim)rsquo
lsquoVaversquos story (that he tells)rsquo
Oceanic Languages Grammatical Overview 129
Niuean has lost even this distinction using only a- forms in all cases
Niueanhaa-ku ihu haa-ku falePOSS-my nose POSS-my houselsquomy nosersquo lsquomy housersquo
64 Verbs and the Verb ComplexI use the term verb complex to refer to a phrase consisting of a verb whichmay be preceded and followed by particles of various kinds13 In someOceanic languages the verb itself is fairly simple in structure but numer-ous particles may occur in a verb complex In others a verb may take quitea number of prefixes and suffixes and the verb complex is usually simpler instructure
641 General Structure of the VerbIt is common in many Oceanic languages for the verb to consist simply of theverb root (underlined in the next set of examples) with no prefixes or suf-fixes This is particularly true of the Micronesian and Polynesian languagesbut is also common in languages of Melanesia
NehanA mahoh ene pak-e rikin wahthe old this should-heNONPAST lie restlsquoThis old man should lie down and restrsquo
TolsquoabalsquoitaNau ku bilsquoi fulaI I justnow arrivelsquoI arrived just nowrsquo
AlsquojieumlGouml yeacute vi koumlyoumlI will CONTINUOUS playlsquoI am going to go on playingrsquo
PonapeanSoulik kin pirida kuloak isuhSoulik HABITUAL getup clock sevenlsquoSoulik gets up at seven orsquoclockrsquo
130 CHAPTER 6
FijianE lailai na valeit small the houselsquoThe house is smallrsquo
RotumanlsquoEap ta la hoalsquomat the FUTURE takelsquoThe mat will be takenrsquo
MāoriI kai te rangatiraPAST eat the chieflsquoThe chief atersquo
Languages of this type do however have a fairly small set of verbalprefixes and suffixes The most frequently used prefixes mark causativity(see the Tolsquoabalsquoita example below) and reciprocality (Alsquojieuml) while suffixescommonly mark the person and number of the object (Tolsquoabalsquoita Alsquojieuml)transitivity (Fijian) or the passive (Māori)
TolsquoabalsquoitaNia lsquoe falsquoa-faalu-a rabolsquoahe he CAUSATIVE-clean-it bowllsquoHe cleaned the bowlrsquo
AlsquojieumlCuru vi-yalsquo-rutheytwo RECIPROCAL-hit-themtwo
lsquoThey hit each otherrsquo
FijianE rai-ci irahe see-TRANS themlsquoHe saw themrsquo
MāoriKa pūhi-a te poaka e waiINCEPTIVE shoot-PASSIVE the pig by wholsquoBy whom was the pig shotrsquo
Grammatical features that are marked by particles in languages likethese are marked by prefixes in another set of languages found mainly inMelanesia In these languages the verb root almost never occurs aloneWhen it does it marks the (singular) imperative
Oceanic Languages Grammatical Overview 131
LenakelAmnuumwdrinklsquoDrink (it)rsquo
In such languages however verbs typically take prefixes and suffixes mark-ing subject tense-aspect and a range of other grammatical features In theexamples below the verb root is underlined
ManamlsquoU-lele-lsquoamayou-lookfor-usEXC
lsquoYou looked for usrsquo
Malsquoasi-lo i-ngara-ngaraocean-in he-CONTINUOUS-swimlsquoHe is swimming in the oceanrsquo
Natu i-laba-doichild he-big-COMPLETIVE
lsquoThe child has grown uprsquo
Robulsquoa i-ro-rolsquoalsquo-i-ramo-larubbish it-HABITUAL-throw-them-randomly-persistentlylsquoHe keeps throwing rubbish all over the placersquo
LenakelR-ɨm-kɨn mun akɨnhe-PAST-eat again verylsquoHe ate a lot againrsquo
K-n-ai-ami ru apus am nɨkomthey-COMPLETIVE-PL-urinate try extinguished just firelsquoThey just tried to put the fire out by urinating on itrsquo
K-ɨm-am-ai-akar-atu-pn kam ilar miinthey-PAST-CONTINUOUS-PL-talk-RECIPROCAL-there
to they PL
lsquoThey (pl) were talking to one anotherrsquo
K-ɨm-uɨni-uas to nahutothey-PAST-DUAL-say-together to crowdlsquoThey were both talking at once to the crowdrsquo
132 CHAPTER 6
642 Tense Aspect and MoodTense refers to the time of the action or state referred to by the verb Forexample in English one makes a basic tense distinction between past pre-sent and future Aspect refers to the way in which the action is carried outor is seen to be carried out English distinguishes completive habitual con-tinuous and punctiliar aspects in each tense14 In many languages there aresome markers of tense some of aspect and some that mark a combination oftense and aspect Descriptions of these languages often refer to the tense-aspect system
Some languages have tense systems similar to or simpler than that ofEnglish For example Fijian has only two markers of tense ā lsquopastrsquo and na lsquofu-turersquo which are particles coming before the verb The past tense marker isoptional once the time has been established as in the second example below
FijianE ā lako mai o Jonehe PAST go here thePERSONAL JohnlsquoJohn camersquo
E nanoa e (ā) lako mai o Joneon yesterday he (PAST) go here thePERSONAL JohnlsquoJohn came yesterdayrsquo
E na lako mai o Jonehe FUTURE go here thePERSONAL JohnlsquoJohn will comersquo
Rotuman has only one tense-marker la (sometimes taeligla) which marks thefuture The non-future is unmarked
RotumanIa lsquoea ia la leumhe say he FUTURE comelsquoHe says he will comersquo or lsquoHe said he would comersquo
Tͻn ta sun-lsquoiawater the hot-STATIVElsquoThe water is (now) hotrsquo
Fā ta leume-aman the COME-COMPLETIVElsquoThe man has already comersquo
Other languages have more complex tense systems than that of EnglishLenakel for example distinguishes four non-future tenses
Oceanic Languages Grammatical Overview 133
Lenakeln-ak-ol lsquoyou do itrsquon-ɨm-ol lsquoyou did itrsquon-n-ol lsquoyou have done itrsquon-ep-ol lsquoyou did it (after you did something else)rsquo
A future prefix t- can be used in combination with two of the tense prefixesabove to produce two different future tenses
Lenakelt-n-ak-ol lsquoyou will do it soonrsquot-n-ep-ol lsquoyou will do it some time laterrsquo
In yet other languages tense is not really marked at all Let us considerwhat Rehg (1981 268) has to say about Ponapean
Ponapean may be described as a tenseless language This is not to saythat in Ponapean it is impossible to express notions of timehellip What ismeant by saying that Ponapean is tenseless is that it expresses consid-erations of time in a way different from English Rather than using atense system to signal time relations Ponapean employs what we willcall an aspect system The basic difference between these two sys-tems is this in a tense system when an event occurred is importantin an aspect system the time contour of the event is crucial
This idea of a time contour can be clarified by looking at four aspects markedin Ponapean
1 Habitual aspect is marked by the preverbal particle kin this im-plies that the action is or was a customary or habitual one which isor was done regularly
2 Continuous aspect (Rehg calls this ldquodurativerdquo) marked by redupli-cating the verb signals that the action or state of the verb iscarried out or takes place over some length of time
3 Completive aspect marked by the suffix -ehr indicates that theaction has reached or is on the way to reaching some kind of con-clusion or completion
4 Irrealis aspect marked by the preverbal particle pahn impliesthat the action is not complete or realized (often it corresponds toa future tense in other languages)
Some examples
PonapeanSoulik kin kang raisSoulik HABITUAL eat ricelsquoSoulik eats ricersquo
134 CHAPTER 6
Soulik kang-kang raisSoulik CONTINUOUS-eat ricelsquoSoulik is eating ricersquo
Soulik kang-ehr raisSoulik eat-COMPLETIVE ricelsquoSoulik has eaten ricersquo
Soulik pahn kang raisSoulik IRREALIS eat ricelsquoSoulik will eat ricersquo
A verb may also occur without any of these aspect markers as in
PonapeanSoulik kang raisSoulik eat ricelsquoSoulik is eating ricersquo lsquoSoulik ate ricersquo etc
This simply indicates that Soulik was involved in eating rice No time is spec-ified although this can of course be included if it is necessary
PonapeanSoulik kang rais nan sounpar samwalahroSoulik eat rice on year lastlsquoSoulik ate rice last yearrsquo
Soulik kang rais metSoulik eat rice nowlsquoSoulik is eating rice nowrsquo
Ponapean illustrates the use of an aspect rather than a tense system AsI mentioned above however many Oceanic languages have particles or af-fixes that mark both tense and aspect Here is the list of Māori tense-aspectparticles
Māorika inceptive Beginning of a new actioni past Action in the pastkua completive Action (fairly recently) completedkia desiderative Desirability of an actionme prescriptive Action should take placee non-past Present or future (when used with ana
following the verb indicates incompleteor continuous action)
Oceanic Languages Grammatical Overview 135
kei warning lsquoDonrsquotrsquo or lsquolestrsquoina conditional lsquoIfrsquo or lsquowhenrsquo
Examples
MāoriKa takoto te tamaiti ka moeINCEPTIVE lie the child INCEPTIVE sleeplsquoThe child lay down and sleptrsquo
Kua mate ta-ku hoaCOMPLETIVE die POSS-my friendlsquoMy friend has diedrsquo
Me hoki te tamaiti ra ki te kāingaPRESCRIPTIVE return the child that to the homelsquoThat child should go homersquo
E haere ana te wahine ki te moanaNONPAST go CONTINUOUS the woman to the sealsquoThe woman is going to the searsquo
Kia āta kōrero tātou kei rongo mai a-ku hoaDESIDE-RATIVE
careful talk we INC LEST hear here POSSPL-my friend
lsquoWe should talk quietly lest my friends hearrsquo
A final set of examples from Nakanamanga illustrates a different fea-ture the concept of mood and shows a pattern of root-initial consonantalternation that is found in a few areas within Oceanic (particularly cen-tral Vanuatu and the Morobe Province of Papua New Guinea) Mood(sometimes referred to as modality) does not relate so much to time asto actuality An actual state or event is said to be in the realis moodwhile a nonreal or non-actual state or event is in the irrealis moodRealis often refers to something that happened is happening or will def-initely happen whereas irrealis refers to something that only might takeplace Oceanic (and other) languages differ in the treatment of negativesIn some languages the negative is in the realis mode (because it actuallydid not happen) in others it is in the irrealis mode because the actionwas not real
Nakanamanga has a set of preverbal particles marking tense-aspectSome of these are used in realis mood others in irrealis mood It is alsoone of the many central Vanuatu languages in which there is alternation be-
136 CHAPTER 6
tween some initial consonants of verbs Verbs with initial v w k and r retainthese consonants in irrealis mood but change them to p p g and t re-spectively after any preverbal particle (The verb root is underlined in theexamples)
NakanamangaIrrealis mood Realis mooda ga vano e panoI INTENTIONAL go he golsquoIrsquom goingrsquo lsquohe goesrsquo
e pe rogo e poo togohe CONDITIONAL hear he COMPLETIVE hearlsquoif he hearsrsquo lsquohe has heardrsquo
643 SubjectMost Oceanic languages mark the person and the number of the subjectsomewhere in the verb complexmdasheither as a prefix to the verb or as a pre-verbal particle15 In some cases a single morpheme marks both person andnumber
PaameseNa-mū mon alokI-makeit puddinglsquoI made the puddingrsquo
Ro-mūmon alokweINC-makeit puddinglsquoWe (inclusive) made the puddingrsquo
Ma-mūmon alokweEXC-makeit puddinglsquoWe (exclusive) made the puddingrsquo
KiribatiE ata-aihe know-melsquoHe knows mersquo
A ata-aithey know-melsquoThey know mersquo
Oceanic Languages Grammatical Overview 137
In other languages person and number are marked by separate mor-phemes
LenakelN-ak-am-kɨn menuk uayou-PRESENT-CONTINUOUS-eat chicken orlsquoAre you (singular) eating chickenrsquo
N-ak-am-ia-kɨn menuk uayou-PRESENT-CONTINUOUS-DUAL-eat chicken orlsquoAre you two eating chickenrsquo
N-ak-am-ar-kɨn menuk uayou-PRESENT-CONTINUOUS-PL-eat chicken orlsquoAre you (plural) eating chickenrsquo
In a number of languages in Melanesia the marking of tense-aspect ormood is combined with the marking of the subjectrsquos person and number ina single morpheme Manam for example has two sets of subject prefixes toverbs one used in realis mood and the other in irrealis mood
ManamSingular Plural
Realis Irrealis Realis Irrealis1 u- m- 1 INC ta- ta-2 lsquou- go- 1 EXC lsquoi ga-3 i- nga- 2 lsquoa- lsquoama-
3 di- da-
For example
ManamEu i-matedog itREALIS-dielsquoThe dog diedrsquo
Eu nga-mate lsquoanadog itIRREALIS-die likelylsquoThe dogrsquos going to diersquo
In languages like these the subject marker occurs whether the subjectis a full noun phrase or a pronoun and whether that subject is expressed inthe sentence or not By contrast languages in western Polynesia use prever-bal subject-marking pronouns only when the subject is a pronoun
138 CHAPTER 6
TonganNalsquoe lsquoalu lsquoa e tangata (noun phrase subject)PAST go SUBJECT the manlsquoThe man wentrsquo
Nalsquoa ne lsquoalu (pronoun subject)PAST he golsquoHe wentrsquo
SamoanUa sau le alilsquoi (noun phrase subject)COMPLETIVE come the chieflsquoThe chief has comersquo
Ua lsquouo sau (pronoun subject)COMPLETIVE I comelsquoI have comersquo
Languages in eastern Polynesia have lost this preverbal subject-markingsystem altogether
TahitianlsquoUa tāpū te vahine lsquoi te vahie (noun
phrase subject)PAST cut the woman OBJECT the woodlsquoThe woman cut the woodrsquo
lsquoUa tāpū vau lsquoi te vahie (pronoun subject)PAST cut I OBJECT the woodlsquoI cut the woodrsquo
644 Object and TransitivityMost Oceanic languages have suffixes that mark a verb as transitive mdashthatis as having an object
NakanamangaA ga munuI INTENTIONAL drinklsquoIrsquoll drinkrsquo
A ga munu-gi noai nagaI INTENTIONAL drink-TRANS water thatlsquoIrsquoll drink that waterrsquo
Oceanic Languages Grammatical Overview 139
FijianE buluhe burylsquoHesheit is buriedrsquo
E bulu-t-a na benuhe bury-TRANS-it the rubbishlsquoHeshe buried the rubbishrsquo
There are a number of features of the marking of transitive and objectThe first is the form of the transitive suffix In many languages this is simply-i
AnejomAdapo-i upni yin aakcover-TRANS good him youlsquoCover him up wellrsquo
In other languages however the transitive suffix is -Ci where C is a the-matic consonant This consonant (1) is not present when the root occurs byitself (2) is present when the suffix is added and (3) is different with differ-ent verbs Look at the following Fijian intransitive and transitive verbs (thetransitive is in the form used before a pronoun or proper noun)
FijianIntransitive Transitivebulu bulu-ti lsquoburyrsquorai rai-ci lsquoseersquotuku tuku-ni lsquotellrsquokaci kaci-vi lsquocallrsquoviri viri-ki lsquothrow atrsquokila kila-i lsquoknowrsquo
As you can see from the examples the form of the suffix (which is sometimessimply -i) is unpredictable One simply has to learn that bulu for exampletakes -ti but rai takes -ci16
The second feature is that many Oceanic languages in fact have twotransitive suffixes the first deriving from Proto Oceanic -i and the secondfrom -aki or -akini This second suffix is sometimes called the applicativeIt often refers to the instrument with which the action is carried out the rea-son for performing the action or some other more indirect transitive notionIn the Fijian examples below I have used the form of the suffix that incor-porates a third person singular object -a In Fijian -Ci-a becomes -Ca and-Caki-a becomes -Caka In some cases the thematic consonant is the same inboth suffixes
140 CHAPTER 6
FijianTransitive Applicativecici-va lsquorun for itrsquo cici-vaka lsquorun with itrsquocabe-ta lsquoascend itrsquo cabe-taka lsquoascend with itrsquooso-va lsquobark at itrsquo oso-vaka lsquobark because of itrsquouso-ra lsquopoke itrsquo uso-raka lsquopoke with itrsquo
In other cases the thematic consonants are different
FijianTransitive Applicativekaki-a lsquoscrape itrsquo kaki-taka lsquoscrape with itrsquoyaqa-va lsquocrawl to itrsquo yaqa-taka lsquocrawl with itrsquomasu-ta lsquopray to itrsquo masu-laka lsquopray for itrsquotala-a lsquosend himrsquo taka-vaka lsquosend itrsquo
Mention of the Fijian third person singular object suffix -a brings us toa third feature the specific marking of object Marking the objectrsquos personand number within the verb complex is less common than marking subjector transitivity For example although in Lenakel separate prefixes mark boththe person and the number of the subject and although some verbs take atransitive suffix the object is not marked in the verb complex at all Evenpronominal objects occur as free forms17
LenakelR-ɨm-eiua-in mun iikhe-PAST-lie-TRANS again youlsquoHe lied to you againrsquo
A large number of languages however do mark the person and the num-ber of the object within the verb complex either with a suffix to the verb (asin Manam and Kiribati) or as a postposed verbal particle (as in Fijian)
ManamBang u-naghu-serelsquo-itaro IREALIS-pierce-split-itlsquoI split the taro by piercing itrsquo
KiribatiE ata-a tama-uhe know-him father-mylsquoHe knows my fatherrsquo
Oceanic Languages Grammatical Overview 141
FijianE ā rai-ci irau na yalewahe PAST see-TRANS themtwo the womanlsquoHe saw the two womenrsquo
Generally if a language has transitive and object suffixes both occur suf-fixed to the verb in that order18
UlithianYule-mi-ya cale leedrink-TRANS-it water thislsquoDrink this waterrsquo
Xa-si-ya doxo cale laacarry-TRANS-it here water thatlsquoBring that water herersquo
In other languages the object suffix occurs when the object is a pronounbut not when it is a noun or noun phrase
NakanamangaA ga munu-gi-aI INTENTIONAL drink-TRANS-it (pronoun object)lsquoIrsquoll drink itrsquo
A ga munu-gi noai nagaI INTENTIONAL drink-TRANS water that (noun object)lsquoIrsquoll drink that waterrsquo
645 The PassiveOnly a small number of Oceanic languages contrast active and passive voiceA couple of Micronesian languages mark the passive by means of a suffix tothe verb (eg Kosraean -yuhk) The example below contrasts an active sen-tence with the corresponding passive one
KosraeanTuhlihk sacn tuhlakihn pinsuhl nuhtih-k ahchild that snatch pencil CLASSIFIER-my thelsquoThat child snatched my pencilrsquo
Pinsuhl nuhtih-k ah tuhlakihn-yuhk (sin tuhlihk sacn)pencil CLASSIFIER-my the snatch-PASSIVE (by child that)lsquoMy pencil was snatched (by that child)rsquo
142 CHAPTER 6
Most languages of eastern Polynesia have a passive This is usually markedby the suffix -Cia where C once again represents a thematic consonant19
Examples
HawaiianUa lsquoai ka mākalsquoi i ka poiCOMPLETIVE eat the policeman OBJECT the poilsquoThe policeman ate the poirsquo
Ua lsquoai-lsquoia ka poi (e ka mākalsquoi)COMPLETIVE eat-PASSIVE the poi (by the policeman)lsquoThe poi was eaten (by the policeman)rsquo
MāoriI inu te tangata i te waiPAST drink the man OBJECT the waterlsquoThe man drank the waterrsquo
I inu-mia te wai (e te tangata)PAST drink-PASSIVE the water (by the man)lsquoThe water was drunk (by the man)rsquo
TahitianlsquoUa hohoni te uri lsquoi te tamaitiPAST bite the dog OBJECT the boylsquoThe dog bit the boyrsquo
Ua hohoni-hia te tamaiti (lsquoet e uri)PAST bite-PASSIVE the boy (by the dog)lsquoThe boy was bitten (by the dog)rsquo
In examples of the passive given so far I have put the agent in paren-theses In these languages a passive sentence may occur with or without anagent
TahitianlsquoUa hohoni-hia te tamaiti lsquoe te uriPAST bite-PASSIVE the boy by the doglsquoThe boy was bitten by the dogrsquo (agent specified)
lsquoUa hohoni-hia te tamaitiPAST bite-PASSIVE the boylsquoThe boy was bittenrsquo (no agent specified)
Oceanic Languages Grammatical Overview 143
Very few languages in Melanesia have a passive Those that do are spo-ken in the western Solomons In these languages only the passive withoutagent is permitted Indeed in Roviana at least the passive is used only whenthe agent is generic or is not recoverable from the context
RovianaSeke-a sa tie sa sikihit-it the man the doglsquoThe man hit the dogrsquo
Ta-seke sa sikiPASSIVE-hit the doglsquoThe dog was hitrsquo
646 The Causative and the ReciprocalA very widespread causative prefix in Oceanic languages whose form de-rives from Proto Oceanic paka- expresses the notion that the subject makesor causes the action of the verb to happen The causative can convert a sta-tive or an intransitive verb into a transitive one
FijianE davo-r-ahe lie-TRANS-itlsquoHe lay on itrsquo
E vaka-davo-r-ahe CAUSATIVE-lie-TRANS-itlsquoHe made himherit lie downrsquo
Further examples of this function are
ManamDang i-alsquoa-gita-iwater he-CAUSATIVE-hot-itlsquoHe heated the waterrsquo
RovianaLopu va-mate tie si raunot CAUSATIVE-die person SUBJECT IlsquoI didnrsquot kill anybodyrsquo
144 CHAPTER 6
MokileseLih-o ka-loau-i mwingeh-uwoman-the CAUSATIVE-becooked-TRANS food-thelsquoThe woman made sure the food was cookedrsquo
West FutunaNe-i faka-sara aia ta vetokaPAST-he CAUSATIVE-beopen he the doorlsquoHe opened the doorrsquo
The causative prefix often has a number of other functions in these lan-guages One common one is to form ordinal or multiplicative numerals fromcardinal numerals which are stative verbs20
Kiribatiteniua lsquothreersquo ka-teniua lsquothirdrsquonimaua lsquofiversquo ka-nimaua lsquofifthrsquo
Samoanlua lsquotworsquo falsquoa-lua lsquotwicersquotolu lsquothreersquo falsquoa-tolu lsquothree timesrsquo
There is also a widespread reciprocal prefix deriving from ProtoOceanic paRi- that marks both reciprocality (the subjects perform the ac-tion on each other) and often also mutual common united or concertedaction The following pair of examples illustrates reciprocality
FijianE loma-ni koyahe love-TRANS shelsquoHe loves herrsquo
Erau vei-loma-nitheytwo RECIPROCAL-love-TRANSlsquoThey (two) love each otherrsquo
The next examples show concerted action
FijianEra butu-k-athey tread-TRANS-itlsquoThey trod on itrsquo
Era vei-butu-yak-athey CONCERTED-tread-TRANS-itlsquoThey trampled it all overrsquo
Oceanic Languages Grammatical Overview 145
Samoan has taken this one step further There the reciprocal prefix fe- inaddition to normal reciprocal functions has also come to mark some verbsas having plural subjects (perhaps deriving from the idea of united or con-certed action)
Samoan
Singular Pluralalsquoa fe-alsquoa lsquokickrsquoinu fe-inu lsquodrinkrsquofefe fe-fefe lsquobe afraidrsquotagi fe-tagi-si lsquocryrsquooso fe-oso-fi lsquojumprsquo
647 The Structure of the Verb ComplexIn some Oceanic languages the verb root may take a fairly large numberof verbal affixes but the verb complex usually contains relatively few parti-cles Languages of this type are mainly found in Melanesia though not allMelanesian languages fit this pattern There is no clear correlation betweenthe morphological complexity of the verb and the geographical location orgenetic affiliation of the language
In other languagesmdashespecially those of Polynesia Micronesia Fiji andsome parts of Melanesiamdashthe verb is simpler morphologically The verbcomplex usually contains a number of particles marking tense aspect andvarious other adverbial features
This difference can best be illustrated by looking at a couple of verbcomplexes in two languages Fijian which uses a range of preverbal andpostverbal particles and Lenakel which relies heavily on affixes The Fijianexamples below are from Schuumltz (1985) while the Lenakel sentences aretranslations of these The verb root is underlined in each example
FijianE sā qai tau-r-a maishe ASPECT then bring-TRANS-it herelsquoThen she brought it herersquo
Eratou sā lako vata sara yanitheyfew ASPECT go together intensive therelsquoThey (few) went off there togetherrsquo
E ā wili-k-a talehe PAST read-TRANS-it againlsquoHe read it againrsquo
146 CHAPTER 6
LenakelR-ep-os-i-pashe-then-take-TRANS-herelsquoThen she brought it herersquo
K-ɨm-hal-vɨn-uasthey-PAST-TRIAL-gothere-togetherlsquoThey (three) went off there togetherrsquo
R-ɨm-avhi-in munhe-PAST-read-TRANS againlsquoHe read it againrsquo
One further complication in Oceanic languages as in members ofmany other language families around the world is that a single verb com-plex may consist of more than one verb through a process known as verbserialization Usually the same participants (like subject and if a verb istransitive object) are involved with each verb in the series The followingexamples illustrate simple intransitive serialization Each serialized verb isunderlined
RovianaTotoso ene nuguru la ghami hellipwhen walk enter go weEXClsquoWhen we walked in helliprsquo
Turu saghe pule mae si raustand rise return come SUBJECT IlsquoI stood back uprsquo
In a transitive sentence transitivity or an objectrsquos features are usuallymarked only once In the following example note that the transitive suffixappears not on the first verb in the series (the transitive verb seke lsquohitrsquo) buton the last verb (the intransitive verb mate lsquodiersquo)
RovianaLopu seke mate-i rau pa lima-gunot hit die-TRANSthem I with hand-mylsquoI didnrsquot kill them with my handsrsquo
Paamese is a language in which negation is marked by a discontinuousaffix (see 654 below) that is a verb in the negative must take both the pre-fix ro- and the suffix -tei
Oceanic Languages Grammatical Overview 147
PaameseNi-ro-kan-tei ouhIFUTURE-not-eat-not yamlsquoI will not eat the yamrsquo
In serial constructions in Paamese the first verb in the series takes the pre-fix ro- while the last verb takes the suffix -tei
PaameseNi-ro-kan vīs-tei ouhIFUTURE-not-eat try-not yamlsquoI will not try to eat the yamrsquo
65 SentencesEnglish normally requires sentences to contain (at least) one verb butOceanic languages do not Below I follow Krupa (1982) in distinguishing be-tween verbal sentences and nominal sentences
651 Nominal SentencesNominal sentences have no verb They consist of a subject and a predicate(sometimes referred to as a topic and a comment about that topic) but thepredicate is usually a noun phrase specifying a person thing place andso on In languages in which the subject normally precedes the verb in averbal sentence (see below) the subjecttopic comes before the predicatecomment in nominal sentences
Subject Predicate
TolaiIau mamatiI fromherelsquoI am from herersquo
MotuIa na tau bada-nahe FOCUS man big-SGlsquoHe is a bigelderly manrsquo
TolsquoabalsquoitaThata-mu ni teiname-your thePERSONAL wholsquoWhat is your namersquo
148 CHAPTER 6
MokilesePediro kahdilik-menPediro Catholic-CLASSIFIERlsquoPedro is a Catholicrsquo
RotumanIa gagaja-the chief-alsquoHe is a chiefrsquo
In languages in which the verb normally precedes the subject (see below)the predicate in a nominal sentence comes before the subject
Predicate Subject
YapeseChitamngii-g Tamagfather-my TamaglsquoTamag is my fatherrsquo
RovianaVineki zingazingarana si asagirl lightskinned SUBJECT shelsquoShe is a light-skinned girlrsquo
FijianNa ke-na i-liuliu na kānalathe POSS-its NOM-lead the colonellsquoThe colonel is its leaderrsquo
TonganKo e faiako auFOCUS a teacher IlsquoI am a teacherrsquo
MāoriHe kātiro ātāhua a Māramaa girl beautiful thePERSONAL MāramalsquoMārama is a beautiful girlrsquo
Kiribati in which the verb comes first in a verbal sentence apparently al-lows either subject + predicate or predicate + subject with little if anydifference in meaning
Oceanic Languages Grammatical Overview 149
KiribatiTe berititenti ngaia (subject + predicate)the president helsquoHe is the presidentrsquo
Ngaia te berititenti (predicate + subject)he the presidentlsquoHe is the presidentrsquo
The translations of all these sentences contain some form of the verbldquoto berdquo which is used in equational sentences (ldquoHe is the presidentrdquo) insome kinds of locational sentences (ldquoI am from hererdquo) and so on ManyOceanic languages have no such verb expressing equational and locationalsentences as nominal sentences
652 Accusative and Ergative LanguagesIn discussing the structure of verbal sentences in Pacific languages (Oceanicand other) we need to introduce a distinction between accusative struc-tures and ergative structures21 English for example is a wholly accusativelanguage The subjects of transitive and intransitive verbs are marked in thesame way but the object of a transitive verb is marked differently
For example
1 She is sleeping2 She saw the man3 The man saw her
Sentences (1) and (2) are intransitive and transitive respectively Both haveshe as subject In sentence (3) the form of the object is her not she
The majority of Oceanic languages are accusative languages In the fol-lowing examples the subject is underlined
AnejomEt amjeg a natamantilde iyiihe sleep SUBJECT man thatlsquoThat man is sleepingrsquo
Et ecta-i natamantilde iyii a kurihe see-TRANS man that SUBJECT doglsquoThe dog saw that manrsquo
Et ecta-i kuri a natamantilde iyiihe see-TRANS dog SUBJECT man thatlsquoThat man saw the dogrsquo
150 CHAPTER 6
Southwest TannaKɨmlu i-ɨmn-la-gɨnwetwoEXC we-PAST DUAL-afraidlsquoWe (two) were afraidrsquo
K ɨ mlu i-ɨ mn-la-hai pukahwetwoEXC we-PAST DUAL-stab piglsquoWe (two) stabbed the pigrsquo
Pa l- ɨ mn-hai amluwho he-PAST-stab ustwoEXClsquoWho stabbed us (two)rsquo
In Anejom the subject of intransitive and transitive sentences is markedwith a preceding a while the object is unmarked In Southwest Tanna thesubject pronoun in both intransitive and transitive sentences is the same inform (kɨmlu lsquowe two EXCrsquo) but it is amlu as the object of a transitive verb Inboth languages the subject is marked by its position in the sentencemdashat theend in Anejom at the beginning in Southwest Tanna
Some Oceanic languages however have ergative structures In thesestructures the subject of a transitive verb called the agent is marked inone way (by the ergative case) while the subject of an intransitive verb (thesubject) and the object of a transitive verb (the object) are marked differ-ently by the absolutive case
Look at the following examples from Samoan In thefirstmdashintransitivemdashsentence the subject is underlined while in the secondand thirdmdashtransitivemdashsentences the agent is underlined
SamoanSa malsquoi le fafineSTATIVE sick the womanlsquoThe woman is sickrsquo
Na manalsquoo-mia le fafine e le tamaPAST want-TRANS the woman ERGATIVE the childlsquoThe child wanted the womanrsquo
E salu-ina e le fafine le falePRESENT sweep-TRANS ERGATIVE the woman the houselsquoThe woman sweeps the housersquo
In Samoan the absolutive case is unmarked le fafine lsquothe womanrsquo is subjectof the intransitive verb in the first sentence and object of the transitive verb
Oceanic Languages Grammatical Overview 151
in the second In the second and third sentences however le tama lsquothe childrsquoand le fafine are subjects of the transitive verb and are marked as such bythe ergative marker e
Note a similar pattern in Motu Morea is unmarked in the first two sen-tences where it is subject of the intransitive verb and object of the transitiveverb respectively But when it occurs as subject of the transitive verb as inthe third example it is marked by the following ergative marker ese
MotuMorea e-mahutaMorea he-sleeplsquoMorea is sleepingrsquo
Boroma ese Morea e-ala-iapig ERGATIVE Morea he-kill-itlsquoThe pig killed Morearsquo
Morea ese boroma e-ala-iaMorea ERGATIVE pig he-kill-itlsquoMorea killed the pigrsquo
653 Basic Structure of Verbal SentencesDifferent Oceanic languages have different basic phrase orders The orderof subject object and verb within the simple verbal sentence varies fromlanguage to language
SV(O) Languages
In the majority of Oceanic languages the subject (whether it is a pro-noun or a noun phrase) precedes the verb in both intransitive and transitiveclauses In transitive clauses the object follows the verb This order is foundin most languages of island Melanesia (including many of the PolynesianOutliers) as well as in nearly all languages of Micronesia For example
Subject Verb Object
NakanaiE pusi tetala eia parakukuruthe cat his it blacklsquoHisher cat is blackrsquo
E Baba kue-a la paiathe Baba hit-it the doglsquoBaba hit the dogrsquo
152 CHAPTER 6
LabuAse emewho comePASTlsquoWho camersquo
Ecircmaha mocirc-socirchocirc hanocircweEXC weEXC-build houselsquoWe built the housersquo
TolsquoabalsquoitaNau kwa-si matalsquoiI I-not sicklsquoI am not sickrsquo
Kini lsquoe ngali-a rediowoman she take-it radiolsquoThe woman took the radiorsquo
PaameseMail he-toMail heDISTANT-baldlsquoMail is going baldrsquo
Letau kail a-mūmo-n alokwoman PL they-make-TRANS puddinglsquoThe women madeare making the puddingrsquo
LenakelNakankɨp r-ɨm-am-apulNakankɨp he-PAST-CONTINUOUS-sleepldquoNakankɨp was sleepingrdquo
Pehe r-n-os nau kawho he-COMPLETIVE-take knife thatlsquoWho has taken that knifersquo
PonapeanLamp-o pahn pwupwidilamp-that FUTURE falllsquoThat lamp will fall downrsquo
Oceanic Languages Grammatical Overview 153
Kidi-e ngalis Soulikdog-this bite SouliklsquoThis dog bit Soulikrsquo
S(O)V Languages
While the Oceanic languages of much of the mainland of Papua NewGuinea particularly the southern part also prefer subject + verb order inintransitive sentences in transitive sentences the preferred order is subject+ object + verb22
subject Object Verb
MotuMorea e-mahutaMorea he-sleeplsquoMorea is sleepingrsquo
Morea ese boroma e-ala-iaMorea ERGATIVE pig he-kill-itlsquoMorea killed the pigrsquo
MaisinPita-ka i-maa-matuPeter-TOPIC he-CONTINUOUS-sleeplsquoPeter is asleeprsquo
Tamaate-seng sikoo-ka ti-fune-siMEN-ERGATIVE pig-TOPIC they-cut-itlsquoThe men cut up the pigrsquo
Verb-Initial Languages
Languages whose rules demand that the verb complex come first in thesentence are found in various parts of the Oceanic area23 Anejom in Vanu-atu many New Caledonian languages a few languages in Micronesia andmost Polynesian languages (especially those of the Polynesian Triangle) areverb-initial languages In some of these languages the normal order is verb+ object + subject
Verb Object Subject
AnejomEk hag antildeakI eat IlsquoI am eatingrsquo
154 CHAPTER 6
Is ecet Deto a ToseiPAST see Deto SUBJECT ToseilsquoTosei saw Detorsquo
IaaiA me walak wanakathe CONTINUOUS play childlsquoThe child is playingrsquo
A me kot wanakat thaanhe CONTINUOUS hit child chieflsquoThe chief is smacking the childrsquo
KiribatiE a mataku Itaiahe CONTINUOUS watch ItaialsquoItaia is watchingrsquo
E tenaa Itaia te kiriiit bite Itaia the doglsquoThe dog bit Itaiarsquo
In others verb + subject + object is the norm
Verb Subject Object
YapeseBea mool TamagPRESENT sleep TamaglsquoTamag is sleepingrsquo
Kea guy Tamag Tinaghe seeher Tamag TinaglsquoTamag saw Tinagrsquo
MāoriI kai te rangatiraPAST eat the chieflsquoThe chief atersquo
I inu te tangata i te rongoaPAST drink the man OBJECT the medicinelsquoThe man drank the medicinersquo
Oceanic Languages Grammatical Overview 155
TahitianlsquoUa tāmālsquoa te vahinePAST eat the womanlsquoThe woman has eatenrsquo
lsquoUa tāpū te vahine lsquoi te vahiePAST cut the woman OBJECT the woodlsquoThe woman cut the woodrsquo
Flexibility of Phrase Order
To some extent all Oceanic languages like most other languages in theworld allow some flexibility in basic phrase order In English for exampleemphasis or contrast can be laid on the object by moving it to sentence-ini-tial position Compare I just canrsquot stand that fellow with That fellow I justcanrsquot stand
In Oceanic languages it is generally possible to focus attention on anynoun phrase by moving it to the beginning of the sentence In some lan-guages there is a pause (marked in the examples below by a comma) ora special focusing morpheme after this phrase The first set of examples isfrom languages that are normally verb-initial The focus is on the subject
Subject Verb Object
AlsquojieumlMeumllsquou wegrave na kaniyam FOCUS it growlsquoAs for the yam itrsquos growing wellrsquo
IaaiWanakat a me walakchild 3SG CONTINUOUS playlsquoAs for the child heshe is playingrsquo
MālsquooriKo Wahieroa kua moe i a KuraFOCUS Wahieroa COMPLETIVE marry OBJECT thePERSONAL KuralsquoWahieroa [not someone else] has married Kurarsquo
The next couple of examples show focus on the object
156 CHAPTER 6
Object Verb Subject
AnejomNev-atimi iyii na ecta-i aekwhich-man that you see-TRANS youlsquoWhich man was it that you sawrsquo
Object Subject Verb
FijianE dua na qito levu keimami ākī-tak-ait one the game big weEXCPL PAST do-TRANS-itlsquoItrsquos a big game we playedrsquo
The examples below are from normally subject-initial (either SVO or SOV)languages with attention focused on the object
Object Subject Verb
NakanaiLa paia taume eau kama hilo-athe dog your I not see-itlsquoAs for your dog I havenrsquot seen itrsquo
TolsquoabalsquoitaNiu nelsquoe ki na ku ngali-a maicoconut this PL FOCUS I carry-it herelsquoIt was these coconuts that I broughtrsquo
MotuBoroma Morea ese e-ala-iapig Morea ERGATIVE he-kill-itlsquoThe pig Morea killed itrsquo
LenakelNimwa aan nɨmataag-asuul r-ɨm-atakɨnhouse that wind-big it-PAST-destroylsquoThat house was destroyed by the cyclonersquo
The following extract from a Banoni story (Lincoln 1976 229) showshow discourse features influence word order in these languages The nounphrase we are interested in is natsu-ri lsquotheir childrsquo
Oceanic Languages Grammatical Overview 157
BanoniVi natsu-ri ke vakekariana me-riathen child-their COMPLETIVE play with-themlsquoTheir child was playing with
na dzoko na kanisi Vi ka teviri na-riathe child the some then COMPLETIVE eat POSS-theirsome youngsters But they ate
borogho ke kota ke tai-ma natsu-ripig COMPLETIVE all COMPLETIVE come-here child-theirlsquoall the pork (before) their child camersquo
The story is about a man and his wife cooking pork When their child(natsu-ri) is introduced into the story it is obviously in focus It comes beforethe verb of which it is the subject (ke vakekariana lsquohe was playingrsquo) Oncethe child has been introduced however there is no necessity to focus on thechild again In the last clause in the above example natsu-ri follows the verbof which it is the subject (ke tai-ma lsquohe camersquo)
In some Oceanic languages however this variability in phrase order isa requirement of grammar Tolai for example has SV(O) in most sentencetypes but V(O)S in stative sentences Compare examples 1 and 2 with exam-ples 3 and 4
Tolai
Subject Verb Object1 A pap i pot
the dog it comelsquoThe dog camersquo
2 Iau gire ra papI see the doglsquoI saw the dogrsquo
Verb Object Subject3 I ga buka ra evu rat
it FARPAST full the two basketlsquoTwo baskets were filledrsquo
4 I ga tup dir a vinarubuit FARPAST tire themtwo the fightlsquoThe fight tired themrsquo
158 CHAPTER 6
Rotuman also has SV(O) as its normal order but this can change to VS incertain kinds of intransitive sentences (eg imperatives) Compare the firsttwo examples below with the last one
Rotuman
Subject Verb ObjectFā ta joni-enman the runaway-heSTATIVElsquoThe man ran awayrsquo
Iris tauɔki-a fuag tathey repair-TRANS breachinwall thelsquoThey are repairing the breach (in the wall)rsquo
Verb SubjectLeum lsquoaeligecome youlsquo(You) comersquo
654 NegationThere is some variety in the ways in which negation is marked in Oceaniclanguages The most widespread pattern is for negation to be marked by apreverbal negative particle
ManamTamoata tago nga-te-aman not heIRREALIS-see-melsquoThe man will not see mersquo
BanoniMa to tai no Ken ma to taiIRREALIS not go you Ken IRREALIS not golsquoIf you donrsquot go Ken wonrsquot go eitherrsquo
NakanamangaA ko taa munuI INCOMPLETE not drinklsquoI havenrsquot drunk yetrsquo
Oceanic Languages Grammatical Overview 159
AlsquojieumlCeacutereacute daa teuml kalsquouthey not still biglsquoThey are not big yetrsquo
KiribatiE aki kiba te moait not fly the chickenlsquoThe chicken didnrsquot flyrsquo
NukuoroIa e te hanohe PRESENT not golsquoHe is not goingrsquo
TonganNalsquoe lsquoikai lsquoalu lsquoa SialePAST not go SUBJECT SialelsquoSiale didnrsquot gorsquo
In a considerable number of Oceanic languages negation is marked bya discontinuous morpheme Two separate particles must both occur butthey are separated by some other elements (compare French Je suis maladelsquoI am sickrsquo with Je ne suis pas malade lsquoI am not sickrsquo) Generally one of theseparticles occurs before the verb and the other after it
RagaRan hav gita-u tehetheyCOMPLETIVE not see-me notlsquoThey didnrsquot see mersquo
RotumanTaunaeliglsquo ta kat sok rameeting the notNON-FUTURE happen notlsquoThe meeting did not take placersquo
Taunaeliglsquo ta kal sok rameeting the notFUTURE happen notlsquoThe meeting will not take placersquo
West FutunaA tata ni se kauna ma avau ki ta skulthe parent PAST not send not me to the schoollsquoMy parents didnrsquot send me to schoolrsquo
160 CHAPTER 6
Special mention must be made of the Lewo language of Epi Island Vanuatuwhich is probably unique in the world in requiring (in some grammaticalcontexts) a triple marking of negation
LewoPe ne-pisu-li re Santo polinot I-see-try not Santo notlsquoIrsquove never seen Santorsquo
Sa-na puruvi lala pe ka-la kinan-ena re si poliPOSS-his brother PL not POSSFOOD-their eat-NOM not again notlsquoHis brothers didnrsquot have any more foodrsquo
In languages with complex verbal morphology the negative is oftenmarked by a verbal affix rather than by a particle In a number of cases (likePaamese and Lenakel below) this affix is a discontinuous morpheme incor-porating a prefix and a suffix to the verb
MotuB-asi-na-ita-iaFUTURE-not-I-see-itlsquoI wonrsquot see itrsquo
PaameseInau na-ro-mesai-teiI I-not-sick-notlsquoI am not sickrsquo
LenakelWusuaas ka r-ɨs-ho-aan peravɨn taha-mboy the he-not-hit-not woman POSS-yourlsquoThe boy didnrsquot hit your wifersquo
Other Oceanic languages mark negation with a negative word thatcomes at the beginning of the clause or sentence but is not part of the verbcomplex Discontinuous marking also occurs in some of these languages(Rapanui in the examples below)
TahitianlsquoAita te talsquoata lsquoi hohoni-hia lsquoe te lsquourinot the man COMPLETIVE bite-PASSIVE by the doglsquoThe man was not bitten by the dogrsquo
Oceanic Languages Grammatical Overview 161
MāoriKāhore ngā wāhine e kōrero ananot thePL woman NONPAST talk CONTINUOUSlsquoThe women are not talkingrsquo
RapanuiIna matou kai malsquoa i te vānaga Magarevanot weEXC not know OBJECT the language MangarevalsquoWe ourselves donrsquot know the Mangareva languagersquo
In some Oceanic languages the negative is marked by a negative verbThe first Southwest Tanna sentence below is in the affirmative and the verb(asim lsquoto gardenrsquo) takes subject and tense prefixes
Southwest TannaMagau l-ɨmn-asim niɨvMagau he-PAST-garden yesterdaylsquoMagau worked in the garden yesterdayrsquo
In the negative equivalent of this sentence person and tense marking oc-curs on the negative verb apwah and the verb asim is nominalized
Southwest TannaMagau l-ɨmn-apwah n-asim-ien niɨvMagau he-PAST-not NOM-garden-NOM yesterdaylsquoMagau did not work in the garden yesterdayrsquo
Fijian behaves similarly with the negative verb sega
FijianE sega na kākanait not the foodlsquoThere is no foodrsquo
Au sega ni kilā na vosaI not that knowTRANSit the languagelsquoI donrsquot know the languagersquo
655 Prepositional and Postpositional PhrasesA preposition comes before a noun phrase and specifies that phrasersquosrelationship to the verb or to other phrases in the sentence Typicallyprepositions mark relationships like location time instrument cause
162 CHAPTER 6
and so on A prepositional phrase therefore is a noun phrase intro-duced by a preposition
Most Oceanic languages have a small closed set of prepositions Tolsquoaba-lsquoita and Samoan are typical
Tolsquoabalsquoitalsquoi location directionni purpose instrumentmala lsquolike asrsquolsquoana instrument goal comparison
Samoani location direction toward instrument causema comitative lsquowithrsquomā mō beneficiary lsquoforrsquo (the ao distinction paralleling
that ofpossessives)
mai ablative lsquofromrsquo
In the examples below the prepositional phrases are underlined
TolsquoabalsquoitaThaina-marelsquoa lsquoe nii lsquoi lumamother-ourtwoEXC she be in houselsquoOur mother is in the housersquo
Kasi-a lsquooko lsquoena lsquoana nini lsquoenacut it rope that with knife thatlsquoCut the rope with the knifersquo
SamoanUa sau le tama ma se lsquoau-falsquoiSTATIVE come the boy with a bunch-bananalsquoThe boy is coming with a bunch of bananasrsquo
lsquoO Malia oleā moe i le pōFOCUS Maria FUTURE sleep in the nightlsquoMaria will sleep in the eveningrsquo
As if to compensate for the fairly small number of basic prepositions mostof these languages make considerable use of compound prepositions Acompound preposition (underlined in the examples below) is composed of ageneral preposition plus a noun (often a body part) for greater specificity
Oceanic Languages Grammatical Overview 163
TolsquoabalsquoitaKa takalo-a gano fuu lsquoi maa-na bilsquou fuuhethen scatter-it soil that in face-its house thatlsquoThen he scattered the soil in front of the housersquo
lsquoOno lsquoi ninima-kusitdown at side-mylsquoSit down beside mersquo
Ni lsquoOina lsquoe nii lsquoi laa lumatheFEMININE lsquoOina she belocated in inside houselsquoOina is inside the housersquo
Although the languages of Polynesia and Micronesia and the majority ofthe languages of Melanesia use prepositions many of the languages of theNew Guinea mainland and the nearby offshore islands use postpositions tomark the same kinds of grammatical functions As the name implies a post-position comes after the noun phrase to which it refers rather than beforeit This kind of phrase is known as a postpositional phrase There is a verystrong correlation among the worldrsquos languages between SOV basic sen-tence order and postpositions Within Oceanic as well the languages thathave postpositions are usually also those in which the object comes beforethe verb
Below are the postpositions of Manam and Sinagoro The last twoManam forms are suffixes the Sinagoro forms are clitics suffixed to thelast word in the noun phrase whatever its grammatical category
Manamzaiza comitative lsquowithrsquolsquoana causeane oti ono instrumentbolsquoana lsquolike asrsquo-lo location-o lsquoonrsquo
Sinagoroai location lsquoin atrsquona instrument ablativeγoti accompanimentγana direction toward
Below are some examples in these two languages of sentences containingpostpositional phrases
164 CHAPTER 6
ManamRoa-gu uma-lo i-malipi-lipiwife-my garden-in she-work-CONTINUOUSlsquoMy wife is working in the gardenrsquo
Tanepwa zaiza lsquoi-purachief with weEXC-comelsquoWe came with the chiefrsquo
SinagoroAu γe-γu koko-na a-kwari-a-toI POSS-my axe-with I-hit-it-PASTlsquoI hit it with my axersquo
Kila na kwayalu baraki-na-γana γio piu-a-toKila ERGATIVE dog old-SG-toward spear throw-it-PAST
lsquoKila threw a spear toward the old dogrsquo
As with their preposition-using relatives many of these languages havecompound postpositions like the following in Sinagoro
Sinagoronuma gabule-na-aihouse underneath-its-atlsquounder the housersquo
numa muli-na-aihouse back-its-atlsquobehind the housersquo
mimiga potiati-aihole gonethrough-atldquothrough the holerdquo
66 Similarities and DifferencesOceanic languages exhibit a number of areas of similarity but also many ar-eas of difference Given the period of time in which many of these languageshave been developing separately from their relatives the quite large degreeof similarity is perhaps more surprising than the differences
Oceanic Languages Grammatical Overview 165
CHAPTER
7
Papuan Languages Gram-matical Overview
The seven hundred or so Papuan languages of the Pacific belong to a num-ber of distinct and apparently unrelated families For this reason alone it ismuch more difficult to make grammatical generalizations about them thanabout the Oceanic languages treated in chapter 6 I attempt here to give avery general feel for the diversity of Papuan languages focusing specificallyon differences between them and Oceanic languages The interested readeris referred to Foleyrsquos excellent survey of these languages (Foley 1986)
71 PronounsPronoun systems vary widely among Papuan languages but in general theyare not so complex as Oceanic systems Many Papuan languages distinguishonly singular and plural (sometimes only in some persons like Kuman in theexamples below) Some languages in Irian Jaya do not even do this Theysimply distinguish person though they usually have a special plural mor-pheme preceding or following (like king- in Manem)
Manem Kuman KoitaSingularI ga na dayou sa ene ahesheit angk ye au
Pluralwe king-ga no noyou king-sa ene yathey king-angk ye yau
166
Some Papuan languages however have a dual as well as a plural number inpronouns
Wiru AlamblakSingular
I no nanyou ne ninhesheit one reumlr
Dualwe two tota neumlnyou two kita nifɨnthey two kita reumlf
Pluralwe toto neumlmyou kiwi nikeumlmthey kiwi reumlm
The Wiru examples show another not uncommon feature of Papuan pro-nouns conflation of non-singular second and third persons
A number of Papuan languages distinguish gender in pronouns mostcommonly in the third person singular but occasionally in other persons aswell Note the following singular pronouns in Abelam
AbelamI wnǝyou masculine mǝ nǝyou feminine ntildeǝ nǝhe dǝshe lǝ
Very few Papuan pronominal systems distinguish inclusive and exclu-sive first person Nimboran in Irian Jaya is one language that makes thisdistinction though it does not distinguish singular and plural
Nimboranio I we inclusivengo I we exclusiveko you (singular and plural)no he she it they
The Papuan languages of Solomon Islands also have the inclusiveexclu-sive distinction All of them distinguish gender in the third person and somelanguages do so in other persons as well They also mark dual and in some
Papuan Languages Grammatical Overview 167
cases trial number The most complex of these pronoun systems is that of Ba-niata (see table 6)
Table 6 Baniata Independent Pronouns
Unspecifieda Masculine Feminine NeuterSingular
1 eei2 noe3 zo vo na ntildeo
Dual1INC be bebe1EXC eere eerebe2 bere berebe3 sere robe rede
Trial1INC meno menu1EXC eebeno eebenu2 mebeno mebenu3 nomo numo nafi
Plural1INC memo1EXC eebo2 mebo3 mo mo no
aGender is not distinguished in these persons and numbers
Many Papuan languages mark person and number of subject (and lessoften object) by verbal affixes usually suffixes but sometimes prefixes In-terestingly a number of languages make more distinctions in these affixesthan they do in free pronouns Kuman is one such language
Kuman
Independentpronouns
Subject suffixes
SingularI na -iyou ene -nhesheit ye -uw
168 CHAPTER 7
Dualwe two mdash -buglyou two mdash -bitthey two mdash -bit
Pluralwe no -munyou ene -iwthey ye -iw
Kuman (1) contrasts dual and plural in the subject pronoun suffixes (2) likeWiru it conflates non-singular second and third persons in the subject suf-fixes Neither of these features appears in the independent pronouns
Bilua on the other hand has subject prefixes One small class of verbsmarks the object by prefixes but most verbs take object suffixes Here are thesingular forms of these pronouns (along with the independent pronouns)
Bilua
Subject ObjectIndependent Prefixes Prefixes Suffixes
I anga a- l- -lyou ngo ngo- ng- -nghe vo o- v- -vshe ko ko- k- -k
72 Nouns and Noun Phrases721 Noun Class SystemsMany Papuan languages especially those in the central north of the main-land of New Guinea have elaborate noun class systems While a languagelike French for example grammatically distinguishes two genders (mascu-line and feminine) and a language like German three (masculine feminineand neuter) Abulsquo the language I use to exemplify this system in Papuan lan-guages (Nekitel 1986) has nineteen different noun classes These classesare based on a combination of semantic and phonological factors So whileclass 1 contains nouns referring to males and class 2 nouns referring to fe-males class 5 (which contains such diverse nouns as the words for ldquosongrdquoldquolegrdquo ldquosagordquo ldquovinerdquo and ldquotoothrdquo) is distinguished by the fact that the singu-lar form ends in h while the plural ends in lih
What is of interest in these systems is that other words in a clause that re-late to a nounmdashthe verb of which it is subject adjectives demonstratives andso onmdashare all marked morphologically to indicate that they refer to a noun of a
Papuan Languages Grammatical Overview 169
particular class The word order in the examples below is noun + demonstra-tive + adjective + verb The class marker is underlined in each case1
AbulsquoNoun Dem Adj VerbAleman ana afuni n-ahelsquo lsquoThis good man wentrsquoAlemam ama afumi m-ahelsquo lsquoThese good men wentrsquoNumatalsquo aulsquoa afulsquoi kw-ahelsquo lsquoThis good woman wentrsquoNumatawa awa afuweri w-ahelsquo lsquoThese good women wentrsquoAul ala afuli l-ahelsquo lsquoThis good eel wentrsquoAkuh akuha afukuhi h-ahelsquo lsquoThese good eels wentrsquoBahiataf afa afufi f-ahelsquo lsquoThis good river fish wentrsquoIhiaburuh aha afuhi h-ahelsquo lsquoThis good butterfly wentrsquo
722 Articles and DemonstrativesArticles are virtually nonexistent in Papuan languages As far as demonstra-tives are concerned some Papuan languages show the three-way distinctioncommon to Austronesian languages
Koitao PROXIMATE lsquothisrsquoe INTERMEDIATE lsquothis thatrsquovire DISTANT lsquothatrsquo
Other Papuan languages are more like English with a two-way contrast indemonstratives between proximate (near the speaker) and distant (not nearthe speaker) Barai though closely related to Koita is one such languagebut the demonstrative situation is complicated by the fact that other aspectsof the location of the noun referred to are also incorporated into the system
BaraiProximate Distantig- ij- generalmdash gar- gur- to the sidemdash gam- down at an anglemdash gaf- up at an anglemdash gum- straight downmdash guf- straight up
In languages with strongly developed noun class systems thedemonstratives usually incorporate a marker of the class membership of thenoun referred to The earlier examples from Abulsquo illustrate this
170 CHAPTER 7
723 Noun Phrase StructureAlthough there are exceptions in general the noun phrase in Papuan lan-guages has the head noun first and all modifying and descriptive wordsfollowing Below are a few examples with the head noun underlined in eachcase
Koitaata ahu inuhati vireman old all thatlsquoall those old menrsquo
Dagagutut otu ame uiwastory little that lastlsquothat last little storyrsquo
Abulsquoba-kuh a-kuha bia-kuh afu-kuhistick-CLASS this-CLASS two-CLASS good-CLASSlsquothese two good sticksrsquo
Engaakaacuteli eacutepeacute kituacutemende duacutepaman good four thoselsquothose four good menrsquo
73 Possessive ConstructionsPossessive constructions are less complex in Papuan languages than inOceanic languages Many Papuan languages simply mark a noun as beingpossessed with none of the various subtypes found in Oceanic languagesIn Koita for example the noun possessed is preceded by the independentpronoun and takes the suffix -Ce where the thematic consonant varies de-pending on the noun to which it is suffixed
Koitadi hete-re di ava-γeI chin-POSS I mouth-POSSlsquomy chinrsquo lsquomy mouthrsquo
di vaiγa-de di muni-veI spear-POSS I stone-POSSlsquomy spearrsquo lsquomy stonersquo
Papuan Languages Grammatical Overview 171
Other Papuan languages show a distinction between alienable and in-alienable nouns rather like that of the simplest systems in Oceanic In Dagafor example kinship nouns take possessive suffixes
Dagane goani-naI youngersibling-mylsquomy younger siblingrsquo
nu mama-nuwe father-ourlsquoour fatherrsquo
Other nouns do not take these possessive suffixes but are followed insteadby an independent pronoun plus a possessive marker
Dagane anu-t ne-gaI thing-NOM I-POSSlsquomy thoughtsrsquo
nu dugup nu-gawe clan we-POSSlsquoour clanrsquo
74 Verbs and the Verb Complex741 Person and Number Tense and AspectThe majority of Papuan languages mark person number and sometimesnoun class of the subject as well as tense-aspect and related categories bysuffixes to the verb stem In many cases this leads to complex strings of suf-fixes with concomitantly complex morphophonemics This complexity canalso mean that a clause or indeed a whole sentence may consist only of averb In the following examples from widely separated languages the verbroot is underlined
MagiOni-la-es-ago-IMPERFECTIVE-PRESENT-helsquoHe is goingrsquo
Oni-bi-ava-igo-CONDITINOAL-youtwo-IMPERATIVElsquoYou two gorsquo
172 CHAPTER 7
Oni-sa-lsquoa-i-deigo-FUTURE-I-IMPERATIVE-shorttimelsquoI will go now for a short timersquo
WahgiNa-pi-s-a-mbiɫ-monot-hear-CLASSIFIER-FUTURE-two-QUESTIONlsquoWill you two not hearrsquo
No-n-a-mb-uaeat-CLASSIFIER-FUTURE-they-QUESTIONlsquoCan they eatrsquo
Na-no-tang-e-r-indnot-eat-HABITUAL-COMPLETIVE-CLASSIFIER-IlsquoI do not always eatrsquo
Abelamwʌ-kʌ-wtǝ-kwʌtalk-FUTURE-I-NONPASTlsquoI will talkrsquo
gǝra-kʌ-ntildeǝnǝ-gwʌcry-FUTURE-you FEMININE-NONPASTlsquoYou (fem) will cryrsquo
kʌ-kʌ-wtǝ-kwʌ-yeat-FUTURE-I-NONPAST-notFUTURElsquoI will not eatrsquo
There are Papuan languages however in which at least some of thegrammatical information is carried by prefixes rather than suffixes In thefollowing examples from Yimas the verb root is again underlined
Yimasyan na-ka-kumprak-asa-ttree OBJECTCLASS-I-broken-CAUSATIVE-COMPLETIVElsquoI broke the treersquo
ka-n-wa-nlikely-he-go-PRESENTlsquoHersquos likely to gorsquo
antɨ-ka-wa-ntutmight-I-go-FARPASTlsquoI would have gonersquo
Papuan Languages Grammatical Overview 173
An almost bewildering variety of tense-aspect situations may be markedin the verbs of Papuan languages Let us take Korafe as an example Korafeverbs take one of a number of tense-aspect suffixes
Korafe-e present-are future-ete immediate past (something that happened today)-imuta very near past (something that happened yesterday)-a recent past (something that happened before yesterday
but not very long ago)-ise far past-erae habitual
These tense-aspect markers are followed by suffixes marking the subjectrsquosperson and number and then by a further set of suffixes markingmoodmdashindicative (statements) interrogative (yes-or-no questions) ques-tion (information questions) hortative subjunctive and imperative Theverb root is underlined in the examples
KorafeY-are-s-ago-FUTURE-you-STATEMENTlsquoYou will gorsquo
Re-da y-are-s-iwhat-to go-FUTURE-you-QUESTIONlsquoWhere will you gorsquo
Y-a-s-ago-RECENTPAST-you-STATEMENTlsquoYou went (recently)rsquo
Re-da y-a-s-iwhat-to go-RECENTPAST-you-QUESTIONlsquoWhere did you go (recently)rsquo
A further feature of the verb in Papuan languages is that complexmorphophonemic processes are involved It is often difficult to break downwhat follows a verb into its component suffixes Here are a few examplesfrom Kuman
KumanUnderlying form Surface formpit-i-ka-a gt prikahear-I-REALIS-STATEMENTlsquoI hearrsquo
174 CHAPTER 7
kumbt-uw-ka-a gt kumbrukwatwist-it-REALIS-STATEMENTlsquoIt twistsrsquo
ne-kit-mbugl-ka-a gt nekulkaeat-not-wetwo-REALIS-STATEMENTlsquoWe two didnrsquot eat (it)rsquo
kan-nagl-mba-t-a gt kanaglmbrasee-FUTURE-hopefully-EMPHATIC-STATEMENTlsquo(I) will hope to see itrsquo
742 Adjunct and Serial ConstructionsAdjunct and serial constructions a feature of many Papuan languages canbe introduced by examples from Kuman Look first at the following sen-tences (the relevant morphemes are underlined)
KumanBugla kinde suŋgwapig bad ithitlsquoThe pig is sickrsquo
Ambai giglaŋge duŋgwagirl song shesaylsquoThe girl is singingrsquo
These two sentences are examples of what are called adjunct con-structions in which the verb of the sentence is preceded by a morpheme ofsome other word class usually a noun or an adjective which is known as anadjunct Some other examples in Kuman are
Kumanka di gaugl ereword say laughter dolsquosayrsquo lsquolaughrsquo
kai ere nigl paitears do water lielsquocryrsquo lsquowash (self)rsquo
Serial constructions are similar but not identical to adjunct construc-tions as illustrated in the following examples
Papuan Languages Grammatical Overview 175
KumanYe mbo mbat narukwahe sugarcane cut hegivelsquoHe cut sugarcane for mersquo
Ye komboglo ake suŋgwashe stone hold shehitlsquoShe hit it with a stonersquo
In serial constructions the final verb is preceded by one or more otherverbs Some more examples include
Kumandi te di presay give say perceivelsquotellrsquo lsquoaskrsquosi bogl si goglhit cut hit dielsquosewrsquo lsquokillrsquoere kan pre pol sido see perceive undo hitlsquotryrsquo lsquounderstandrsquo
In all of these cases what other languages often view as a single state orevent and express by a single verb is broken up into components For exam-ple the sentence Ye komboglo ake suŋgwa is idiomatically translated lsquoShehit it with a stonersquo but is more literally lsquoShe held a stone and hit itrsquo In themore literal translation the two components of holding the stone and hittingsomething with it are separated
While many Papuan languages like Kuman make quite frequent use of ad-junct and serial constructions ldquothe closely related Kalam and Kobon are themost remarkable in applying this idea in the most thoroughgoing fashionhellipKalam immediately strikes one as a language in which the speakers are exces-sively specific in their description of eventsrdquo (Foley 1986 113) An example likethe following one gives an idea of just how specific these languages can be
KalamYad am mon pk d ap ay-p-ynI go wood hit hold come put-COMPLETIVE-IlsquoI fetched firewoodrsquo
While the sentence ldquotranslatesrdquo as ldquoI fetched firewoodrsquo the act of fetching isbroken down into its components in Kalam What the Kalam speaker is sayingis something like lsquoI went and chopped wood and got it and came and put itrsquo
176 CHAPTER 7
In languages with these kinds of constructions the number of actualverbs is often much smaller than in other languages ldquoKalam has under 100verb stems and of these only about twenty-five are commonly usedrdquo (Foley1986 115) The Kalam sentence above shows five verbs in a serial construc-tion Other serial constructions in Kalam include the following (The hyphenafter the last element indicates where subject and tense suffixes occur)
Kalamnb nŋ- ag tk-consume perceive sound severlsquotastersquo lsquointerruptrsquo
pwŋy md ay- d am yok-poke stay put take go displacelsquofix (by insertion)rsquo lsquoget rid ofrsquo
Kalam also uses adjunct constructions to a great degree
Kalamwdn nŋ- tmwd nŋ-eye perceive ear perceivelsquoseersquo lsquohearrsquo
kwnk g- joŋb tmey g-saliva do mouth bad dolsquospitrsquo lsquowhinersquo
ywg ntilde- mnm ag ntilde-lid give speech sound givelsquoput a lid onrsquo lsquoconfidersquo
75 Sentences751 Simple SentencesAny generalization about word order in Papuan languages would state thatthey tend to be verb-final languages The order of the core constituents is SVin intransitive clauses and SOV in transitive clauses2
Subject Object VerbBarai
Bu bajae fiad-iathey body pain-theylsquoThey (ie their bodies) are in painrsquo
Papuan Languages Grammatical Overview 177
Fu mave kana-ehe pig hit-PASTlsquoHe hit the pigrsquo
WahgiNa wo-tang-n-alI come-HABITUAL-CLASS-IwilllsquoI will always comersquo
Na mokine no-tang-indI food eat-HABITUAL-IhavelsquoI always ate foodrsquo
AnggorSonggo borǝ me-fe-ofowlegg broke-change-itSlsquoThe wildfowl egg brokersquo
Nindou ai songgo borǝma-r-ea-ndǝman he fowlegg broke-TRANS-itS-itOlsquoThe man broke the wildfowl eggrsquo
Grand Valley DaniAp nik-k-eman eat-REALIS-helsquoThe man atersquo
Ap palu na-sikh-eman python eat-FARPAST-helsquoThe man ate the pythonrsquo
While in many languages this is the usual order in others word order isnot significant for indicating functions like subject or object Many Papuanlanguages ldquomay be regarded as free word-order languages Although theverb is usually positioned clause-finally this rule is rigid only in some lan-guages In a great many Papuan languages peripheral nominals such aslocatives or temporals commonly occur after the verbhellip The general im-pression of clause structure in Papuan languages in comparison to Englishis its overall loosenessrdquo (Foley 1986 168)
Foley illustrates this statement using Yimas The following sentence fol-lows ldquostandardrdquo Papuan SOV order
178 CHAPTER 7
YimasSubject Object VerbPay-um nar-mang na-mpu-tayman-CLASSPL woman-CLASSSG her-theyMASCULINE-seelsquoThe men saw the womanrsquo
Each noun is marked as belonging to a particular noun class and the verbtakes prefixes corresponding to the noun class of the object and the subjectin that order In the verb namputay in the sentence above na- marks a third-person singular object of the human female noun class and mpu- marks athird-person plural subject of the human male noun class The verb nam-putay on its own means lsquoThey (male human) saw her (human)rsquo Conse-quently it is clear which noun is subject and which is object without relyingon word order The following Yimas sentences also mean lsquoThe men saw thewomanrsquo
YimasNarmang payum na-mpu-tay (object-subject-verb)Payum na-mpu-tay narmang (subject-verb-object)Narmang na-mpu-tay payum (object-verb-subject)
Nominal sentences are far less common in Papuan languages than inOceanic languages as many Papuan languages have existential verbs oftenmore than one Kuman for example has three yoŋgwa is used when thesubject is inanimate paŋgwa with animate and inanimate subjects that arein a specific place and molkwa with animate or inanimate subjects whoseexistence is being declared
KumanDi ta yoŋgwaaxe a itbelsquoThere is an axersquo
Usi gagl mina paŋgwacigarette bag in itbeinthatplacelsquoThere are cigarettes in the bagrsquo
Togoi ta molkwasnake a itexistlsquoThere is a snakersquo
Many of the languages of the Highlands are similar to Kuman Huli forexample has three existential verbs Sinasina four and Enga seven (Piau1981)
Papuan Languages Grammatical Overview 179
An extreme case is Anggor in the Sepik with eighteen verbs roughlytranslating ldquoberdquo What is important is the shape of the object its locationand its posture (Litteral 1981 128) So before choosing the appropriateverb one needs to know if the subject is masculine or feminine elongated orbunched up inside something else or not in a horizontal or vertical planehanging on something stuck to something and so on Some of the Anggorexistential verbs are
Anggoramar- be sitting on or insideanǝngg- be standing onenggor- be lying on a low planeanangg- be lying on a high planeapeningg- be attached flat toapaiyar- be attached and curling aroundapuiyar- used only of liquidsahetar- be hanging from a protrusion
This is not to say that there are absolutely no verbless sentences inPapuan languages They do occur in both subject-predicate and predicate-subject orders
Subject Predicate
KoitaAta bera yaga-uhu-geraman a house-in-thelsquoA man is in the housersquo
KumanYuŋgu-n awe
house-your wherelsquoWhere is your housersquo
Predicate Subject
DagaNe tata-na geI oldersibling-my youlsquoYou are my older siblingrsquo
752 Peripheral CasesPeripheral casesmdashrelations other than subject and objectmdashare generallymarked by postpositions or suffixes in Papuan languages More concretespatial notions tend to be expressed by postpositions
180 CHAPTER 7
KumanKagl-e mina yoŋgwafoot-her on itbelsquoIt is on her footrsquo
Komboglo pagl siŋgastone with IhitlsquoI hit it with a stonersquo
KewaAda ru-para paacute-luahouse inside-to go-IFUTURElsquoI will go inside the housersquo
Ada rolo-para paacute-luahouse underneath-to go-IFUTURElsquoI will go under the housersquo
But ldquothe more basic case relations are expressed directlyrdquo usually by suf-fixes (Foley 1986 93)
KumanMokona gagl-e krikagreens bag-in IpacklsquoI put the greens in the bagrsquo
Ye nigl-e molkwahe water-at hebetherelsquoHe is at the riverrsquo
KewaAda-para paacute-luahouse-to go-IFUTURElsquoI will go homersquo
Niacute-na meacuteaacuteaacute-riaI-for get-hePASTlsquoHe got it for mersquo
Many Papuan languages have a very wide range of morphemes markingperipheral case relations Koita which uses clitics to mark these relationsis a good example
Koita-γe lsquoto (rivers)rsquo-va lsquoto (things)rsquo-γasina lsquoto (persons)rsquo-he lsquoatrsquo
Papuan Languages Grammatical Overview 181
-da -na lsquoon torsquo-γore lsquowith (accompaniment) singularrsquo-ruta lsquowith (accompaniment) pluralrsquo-γahara lsquoforrsquo-ni lsquoforrsquo-γa -ma lsquowith (instrument)rsquo-ka partitive-Ce possessive (includes thematic consonant)
Some examples
Koitavani be-hetime some-atlsquosometimesrsquo
di dehiye-heI back-atlsquobehind mersquo
a-γoreyou-withlsquowith yoursquo
idi umuka-vatree root-tolsquonear the treersquo
753 Complex SentencesI discuss one syntactic feature of complex sentences in Papuan languagesswitch reference briefly here This feature is typical of most groups ofPapuan languages (and is also found for example in a number ofAmerindian language groups) but it is rare in the Pacific
The following examples from Enga illustrate what I am going to talkabout First here are some basic verbs3
EngaBaaacute p-eacute-aacutehe go-PAST-helsquoHe wentrsquo
Baa-meacute kalaacutei p-i-aacutehe-ERGATIVE work do-PAST-helsquoHe workedrsquo
182 CHAPTER 7
Nambaacute p-eacute-oacuteI go-PAST-IlsquoI wentrsquo
In each case the verb takes a suffix marking the tense and another markingthe subjectrsquos person and number
When two or more clauses are put together to form a complex sentencethe last verb in the clause (final verb) retains this subject-tense markingbut the other verbs (medial verbs) do not Rather they incorporate a suffixindicating whether the subject of the verb is the same as or different fromthe subject of the following verb Look now at the following Enga examples
EngaBaa-meacute paacute-o kalaacutei p-i-aacutehe-ERGATIVE go-SAMESUBJ work do-PAST-helsquoHe went and worked (at the same time)rsquo
Nambaacute p-e-oacute-pa baa-meacute kalaacutei p-i-aacuteI go-PAST-I-DIFFSUBJ he-ERGATIVE work do-PAST-helsquoI went and he workedrsquo
In both of these sentences the final verb lsquodorsquo has the suffix marking subjectand tense but the verb preceding it does not In the first example the verbpaacute lsquogorsquo takes the suffix -o which indicates that the subject of this verb is thesame as the subject of the next one and the actions happened at the sametime In the second case the verb lsquogorsquo (now with the form p) takes both tenseand subject markers and the suffix -pa which indicates that the subject ofthe next verb is going to be different from the subject of this verb This iswhat is meant by switch-reference
Languages with switch-reference systems are generally a little morecomplex than I have shown For example in the first sentence we find thesuffix -o which marked the verb as having the same subject as the next oneand indicated that the actions of the two verbs were roughly simultaneousIf the second action occurred after the first however we would have to usethe suffix -(a)la rather than -o Here are some suffixes found on Enga medialverbs
Enga-o same subject simultaneous action-(a)la same subject sequential action-pa different subject simultaneous or sequential action-nya same subject next verb expresses purpose or desire-niacute-mi same subject next verb expresses intense desire
Papuan Languages Grammatical Overview 183
Some other Papuan languages have even more complex switch-referencesystems
Languages with switch-reference systems generally have no or fewconjunctions The information that is carried by conjunctions in most Aus-tronesian languagesmdashand in languages like Englishmdashis carried by theswitch-reference verbal suffixes
184 CHAPTER 7
CHAPTER
8
Australian LanguagesGrammatical Overview
In attempting to make generalizations about the structure of the two hun-dred or so languages of the Australian continent we have to remember thatmany of them have disappeared virtually without trace while many othersbecame extinct after only a small amount of linguistic workmdashand that lit-tle usually the effort of linguistically untrained peoplemdashwas done on themTo some extent any general statement about Australian languages is anextrapolation from the languages for which we have reasonable amounts ofdata and an educated reanalysis of those languages recorded by amateursin the last century1
81 PronounsAlmost all Australian languages distinguish at least three numbers in pro-nounsmdashsingular dual and pluralmdashthough a few have a trial or a paucal aswell About half the languages of Australia have an inclusiveexclusive dis-tinction like nearly all Oceanic languages while the rest (like most Papuanlanguages) do not There appear to be no geographical correlates of thesedifferent systems They are scattered fairly randomly across the continent
Below are examples of the two most common types of pronoun systems2
Wargamay WajarriSingular
I ngayba ngajayou nginba nyintahesheit nyunga palu
185
Dualwe two ngali we two INC ngali
we two EXC ngalijayou two nyubula nyupalithey two bula pula
Pluralwe ngana we INC nganyu
we EXC nganjuyou nyurra nyurrathey jana jana
In many Australian languages the third person ldquopronounsrdquo are not reallypronouns at all especially in the singular but rather demonstratives with ameaning something like ldquothis onerdquo or ldquothat onerdquo as opposed to ldquohesheitrdquo
Apart from languages with two or four numbers there are some othervariations in these general patterns Pitta-Pitta for example distinguishesbetween masculine and feminine in the third person singular In addition allthird person pronouns have to take a locational suffix so the full range ofthird person pronouns is
Pitta-PittaSingular Dual Plural
lsquohe lsquoshersquo lsquothey tworsquo lsquotheyrsquoNear nhuwayi nhanpayi pulayi thanayiGeneral nhuwaka nhanpaka pulaka thanakaFar nhuwaarri nhanpaarri pulaarri thanaarri
Lardil is one of a number of languages in which non-singular pronounstake different forms depending on the relationship between the people in-volved One set is used for people of the same generation or two generationsapart the other for people who are one or three generations apart Here arethe dual pronouns
LardilSame generation or two
generations apartOne or three
generations apartwe two INC ngakurri ngakuniwe two EXC nyarri nyaankiyou two kirri nyiinkithey two pirri rniinki
Pronouns generally vary in form according to case that is their functionin the sentence These case suffixes are usually the same as those for nouns
186 CHAPTER 8
As well as the free or independent pronouns discussed above many Aus-tralian languages also have a set of bound pronouns which must be attachedto some other constituent in the sentence Bound pronouns typically marksubject or object and they are often attached to verbs
Western Desertpu-ngku-rna-ntahit-FUTURE-I-youOBJECTlsquoI will hit yoursquo
pu-ngku-rni-nhit-FUTURE-me-youSUBJECTlsquoYou will hit mersquo
In some languages bound pronouns are attached to an auxiliarymdasha specialword in the sentence whose main function is to carry these suffixesmdashratherthan to the verb In the Walmajarri example below the verb is yi- lsquogiversquo butthe bound pronouns are attached to the auxiliary ma-
WalmajarriYi-nya ma-rna-ny-pilangu-lu kakajigive-PAST AUXILIARY-weEXCPL-toyoutwo-DUALO-PLS goannalsquoWe gave the goanna to you tworsquo
82 Nouns and Noun PhrasesIn Australian languages nouns are sometimes reduplicated to mark plural-ity or other features The major feature of interest in the morphology ofnouns in Australian languages however is the marking of case
821 Case MarkingIn most Australian languages a noun phrase must take a suffix indicating itsfunction in the sentence Pronouns also take these case-marking suffixes
Yidiny illustrates the kinds of case-marking systems common in Aus-tralian languages3 It marks a number of cases as listed below (Differentforms of the same case marker occur after different noun-final phonemesSee the discussion of Wargamay on pp 189ndash190 below for an example ofthis)
Yidinyabsolutive oslashergative -nggu -du -bu -ju
Australian Languages Grammatical Overview 187
locative allative instrumental -la -da -ba -jaablative causal -mu -mdative -ndapurposive -gucausal -mu -maversive -jida -yidapossessive -ni
The absolutive case marks the subject of an intransitive verb and the ob-ject of a transitive verb The ergative case marks the agent (the subject of atransitive verb)
YidinyWagaal-du mujam wawa-lwife-ERGATIVE motherABSOLUTIVE lookat-PRESENTlsquo(My) wife is looking at Motherrsquo
The locative allative and ablative cases have to do with direction and loca-tion Locative refers to the location allative marks direction toward andablative marks direction from
YidinyMujam gali-ng digarra-mumotherABSOLUTIVE go-PRESENT beach-ABLATIVElsquoMother is going from the beachrsquo
Here are some examples of some of the other case suffixes
YidinyYingu gurnga mangga-ng waguja-ndathisABSOLUTIVE kookaburra
ABSOLUTIVElaugh-PRESENT man-DATIVE
lsquoThis kookaburra is laughing at the manrsquo
Mujam dubuurrji wuna-ng minya-mmotherABSOLUTIVE fullup lie-PRESENT meat-CAUSALlsquoMother is lying down satiated with meatrsquo
Yingu waguuja garba-ng bama-yidathisABSOLUTIVE manABSOLUTIVE hide-PRESENT people-AVERSIVElsquoThis man is hiding for fear of the peoplersquo
The possessive case suffix -ni marks the possessor A noun with this suf-fix also takes the case suffix of the possessed noun (since it functions like anadjective describing that noun)
188 CHAPTER 8
YidinyWagal-ni-nggu gudaga-nggu mujam baja-lwife-POSS-ERGATIVE dog-ERGATIVE mother
ABSOLUTIVEbite-PRESENT
lsquo(My) wifersquos dog is biting Motherrsquo
Case markers in Australian languages play the same kind of role asprepositions or postpositions in other languages of the Pacific They indicatevarious kinds of grammatical relations between a noun phrase and the verbor between the noun phrase and another noun phrase Because of this Aus-tralian languages have no prepositions or postpositions4
Nouns in Australian languages may also take other suffixes referred toas derivational suffixes Dyirbal illustrates some typical kinds of nominalsuffixes
Dyirbal-jarran plural-garra one of a pair-manggan one of a group-mumbay lsquoallrsquo-barra lsquobelonging to a placersquo-bila comitative lsquowithrsquo-ngarru similative lsquolike asrsquo
For example
Dyirbalgambil-barratablelands-belongingtolsquotablelands peoplersquo
mijiji-garrawhitewoman-oneofpairlsquoa white woman and someone elsersquo
Morphophonemic changes are common when suffixes are added tonouns and verbs Two patterns are particularly widespread First in manylanguages the initial consonant of some suffixes changes according to the fi-nal phoneme of the root to which it is suffixed In Wargamay for examplethe ergative suffix has (at least) five forms
Wargamay-nggu after a vowel bari-nggu lsquostonersquo-ndu after l maal-ndu lsquomanrsquo-dyu after ny munyininy-dyu lsquoblack antrsquo
Australian Languages Grammatical Overview 189
-du after n rr gururr-du lsquobrolgarsquo-bu after m walam-bu lsquotickrsquo
Second in some languages there is vowel copying The vowel of a suf-fix mimics the final vowel of the root to which the suffix is attached Forexample Anguthimri has an ergative suffix on nouns of the form -gV withthe vowel repeating the final vowel of the root
AnguthimriRoot Ergativekyabara lsquocrocodilersquo kyabara-gaszliguumlyi lsquoashesrsquo szliguumlyi-giku lsquostickrsquo ku-gu
822 Noun ClassesQuite a number of Australian languages especially in the northern part ofthe continent have a gender or noun class system Nouns belong to one ofa number of classes determined partly at least on a semantic basis Theclass membership of a noun may be marked on the noun by an affix (a pre-fix in some languages a suffix in others) or a particle and it is often alsomarked on adjectives and other modifiers referring to the noun In prefix-ing languages the noun class membership of subjects or objects may also bemarked in the verb
Tiwi has a noun class system rather like that of French Nouns are eithermasculine or feminine Inherently masculine or feminine nouns may not beovertly marked as such but other nouns often take a suffix (What is inher-ently masculine or feminine is of course culturally defined Crocodiles forexample are thought of as masculine [some] crabs as feminine)
TiwiMasculine Feminine
tini lsquomale personrsquo tinga lsquofemale personrsquomatani lsquomale friendrsquo matanga lsquofemale friendrsquokirijini lsquoboyrsquo kirijinga lsquogirlrsquo
In Tiwi adjectives demonstratives and possessives have to agree with theclass membership of the noun they refer to
TiwiMasculine Feminine
arikula-ni yirrikipayi arikula-nga kiripukabig-MASCULINE crocodile big-FEMININE crablsquoa big crocodilersquo lsquoa big crabrsquo
190 CHAPTER 8
ngi-nanki kirijini angi-nanki pilimungaMASCULINE-this boy FEMININE-this roadlsquothis boyrsquo lsquothis roadrsquo
ngini-wutawa alawura angi-wutawa pulagumokaMASCULINE-they boss FEMININE-they femaledoglsquotheir bossrsquo lsquotheir bitchrsquo
Yanyuwa has sixteen noun classes each one marked by a prefix that oc-curs also with adjectives and numerals (The class marker is underlined ineach example below)
Yanyuwa
rra-muwarda rra-walkurra rra-jakardaFEMININE-canoe FEMININE-big FEMININE-manylsquomany big dugout canoesrsquo
na-lungundu na-walkurra na-jakardaARBOREAL-shelter ARBOREAL-big ARBOREAL-manylsquomany big bark sheltersrsquo
ma-murala ma-walkurra ma-jakardaFOOD-wildcucumber FOOD-big FOOD-manylsquomany big wild cucumbersrsquo
narnu-yabi narnu-arrkulaABSTRACT-good ABSTRACT-onelsquoone good thingrsquo
nya-yabi nya-arrkulaMASCULINE-good MASCULINE-onelsquoone good manboyrsquo
The last two examples show how noun class prefixes can occur even withoutan accompanying noun The class marker makes the referent clear5
823 Modifiers to NounsDemonstratives
Australian languages have no articles but they do have a number of demon-stratives referring to spatial relations The complexity of the demonstrativesystem varies from language to language Here are the demonstratives inthree Australian languages
Australian Languages Grammatical Overview 191
Gumbaynggiryaam lsquothis these herersquoyarang lsquothat those therersquo
Yaygiradyi ngadyi lsquothis these herersquoila yila lsquoherersquodyaadyi lsquothere (not too far away)rsquoalaara yalaara lsquothere (a long way off)rsquo
Djapudhuwai lsquothis these herersquodhuwali lsquothat those there (nearby)rsquongunha lsquothat those there (a long way off)rsquongunhi anaphoric lsquothe one we are talking aboutrsquo
Adjectives
Adjectives behave like nouns in many ways They take the same case suffixesand very often occur as the head of a noun phrase (In these examples fromGumbaynggir the ergative suffix takes the forms -du and -dyu)
GumbaynggirNiiga-du barway-dyu buwaa-ng giibar dyunuyman-ERGATIVE big-ERGATIVE hit-PAST childO smallOlsquoThe big man hit the small childrsquo
Barway-dyu buwaa-ng dyunuybig-ERGATIVE hit-PAST smallOlsquoThe big one hit the small onersquo
824 Noun Phrase StructureThe tendency in Australian languages is for possessives and demonstrativesto precede the head noun in the noun phrase while adjectives follow it Thefollowing examples illustrate this tendency (The head noun is underlined ineach case)
Pitta-Pittanganya-ri murra wimaI-POSS stick biglsquomy big stickrsquo
Guugu Yimidhirrnambal warrga-alstone big-withlsquowith a big stonersquo
192 CHAPTER 8
Yukultartathinta pirwanta ngawu pirtiyathat their dog badlsquothat nasty dog of theirsrsquo
This is however only a tendency As with other areas of grammar inAustralian languages word order in the noun phrase is usually relativelyfree In the Gumbaynggir sentence (repeated from the previous section) theadjectives follow the noun
GumbaynggirNiiga-du barway-dyu buwaa-ng giibar dyunuyman-ERGATIVE big-ERGATIVE hit-PAST childO smallOlsquoThe big man hit the small childrsquo
But either or both could precede the noun in a noun phrase yielding thesepossibilities
GumbaynggirBarway-dyu niiga-du buwaa-ng giibar dyunuybig-ERGATIVE man-ERGATIVE hit-PAST childO smallOlsquoThe big man hit the small childrsquo
Barway-dyu niiga-du buwaa-ng dyunuy giibarbig-ERGATIVE man-ERGATIVE hit-PAST smallO childOlsquoThe big man hit the small childrsquo
This freedom of order sometimes extends beyond the phrase ldquoNot onlycan words occur in any order in a phrase and phrases in any order in asentence [but] in addition words from different phrases may be freely scat-tered through a sentencerdquo (Dixon 1980 442) Look first at the followingsentence
WargamayYibi-yibi ngulmburu-
ngguwurrbi-bajun-du
buudi-lganiy malan-gu
child-PL woman-ERGATIVE
big-very-ERGATIVE
take-CONTINUOUS river-to
lsquoThe very big woman is taking the children to the creekrsquo
In this sentence the adjective wurrbi-bajun-du lsquovery bigrsquo immediately fol-lows the noun ngulmburu-nggu lsquowomanrsquo and it also clearly refers to itbecause both adjective and noun are marked with the ergative suffix (mor-
Australian Languages Grammatical Overview 193
phophonemically-nggu and -du) Because of this other orders are possibleThe same sentence could be said
WargamayNgulmburu-nggu buudi-lganiy malan-gu yibi-yibi wurrbi-bajun-duwoman-ERGATIVE
take-CONTINUOUS
river-to child-PL big-very-ERGATIVE
lsquoThe very big woman is taking the children to the creekrsquo
Here the noun subject ngulmburu-nggu lsquowomanrsquo is separated from its mod-ifying adjective wurrbi-bajun-du lsquovery bigrsquo by the verb the allative phraseand the object
83 Possessive ConstructionsOne case suffix added to nouns and pronouns in many Australian languagesis a possessive suffix
Djapudjamarrkurlilsquo Milyin-guchildren Milyin-POSSlsquoMilyinrsquos childrenrsquo
ngarra-ku-ny dhuway-lsquomirringu-nyI-POSS-EMPHATIC lsquomy husbandrsquo husband-kinship-EMPHATIC
Many Australian languages also distinguish between alienable and inalien-able possessive constructions Alienable possession is marked by thepossessive suffix as in the examples above and is used with all possessednouns except parts of wholes The part-to-whole relationship uses an in-alienable construction in which there is no specific marking Possessed andpossessor nouns are just put one after the other in that order
DjapuDharpu-ngal ngarra-n dhandurrung-dhu gatapanga-ypierce-COMPLETIVE
I-OBJECT horn-ERGATIVE buffalo-ERGATIVE
lsquoThe buffalorsquos horn has pierced mersquo
Rluku ngarra gara-thi-nfoot I spear-INCHOATIVE-COMPLETIVElsquoMy foot has been spearedrsquo
194 CHAPTER 8
84 Verbs and the Verb ComplexIn the verb system a major grammatical difference exists between Pama-Nyungan languages and those of the rest of Australia The Pama-Nyunganlanguages are nonprefixing They use suffixes exclusively to mark verbalcategories like tense aspect and the like Many of the languages of ArnhemLand and the Kimberleys use both prefixes and suffixes
841 Verbs in Pama-Nyungan LanguagesThe general structure of the verb in nonprefixing languages is root + (de-rivational suffixes) + inflection There may be one or more derivationalsuffixes following a root and there will definitely be an inflectional suffix
Derivational Suffixes
Some derivational suffixes convert a transitive verb into an intransitive oneor an intransitive verb into a transitive one Others mark continuous habit-ual and other aspects of the verb as well as expressing meanings for whichother languages often use adverbs
Below are some examples of a number of derivational suffixes in twolanguages Pitta-Pitta and Wargamay In each case only the verb root (plusderivational suffix) is given the final hyphen means that a tense-aspect in-flection needs to be added
Pitta-Pittakathi- lsquoclimbrsquo kathi-la- lsquoput uprsquomari- lsquogetrsquo mari-la- lsquoget forrsquomirrinta- lsquoscratchrsquo mirrinta-mali- lsquoscratch selfrsquongunytyi- lsquogiversquo ngunytyi-mali- lsquoexchangersquothatyi- lsquoeatrsquo thatyi-li- lsquowant to eatrsquothatyi- lsquoeatrsquo thatyi-linga- lsquogoing to eatrsquortinpa- lsquorunrsquo rtinpa-ma lsquorun aroundrsquothatyi- lsquoeatrsquo thatyi-yarnrta- lsquoeat while walking alongrsquo
Wargamaybaadi- lsquocryrsquo baadi-ma- lsquocry forrsquodyinba- lsquospearrsquo dyinba-ma- lsquospear withrsquomayngga- lsquotellrsquo mayngga-ba- lsquotell each otherrsquodyuwara- lsquostandrsquo dyuwara-bali- lsquobe standingrsquobimbiri- lsquorunrsquo bimbiri-yandi- lsquorun awayrsquo
Australian Languages Grammatical Overview 195
Inflectional Suffixes
Inflectional suffixes to verbs mark tense (or tense-aspect) They often mark averb as being imperative or as occurring in a subordinate clause Most Aus-tralian languages like Latin have more than one conjugation or conjuga-tional class of verbs Verbs in the same conjugation take the same suffixesbut verbs in another conjugation take a different set of suffixes and thereis no semantic explanation for why a particular verb belongs to a particularconjugational class
To illustrate both the idea of conjugations and the kinds of grammaticalfunctions they mark here are some data from the Atampaya dialect ofUradhi which has four conjugations (labeled I II III and IV)
UradhiI II III IV
past -γal -kal -n -ntilde -npresent -ma -al -ntildea -oslashfuture -maŋka -awa -ntildeaŋka -ŋkaimperative -ethi -ti -ri -yi -γu
For the verb roots wa- lsquoburnrsquo rima- lsquotwirlrsquo lapu- lsquoblowrsquo and ruŋka- lsquocryrsquowhich belong to conjugations I II III and IV respectively the verb forms ineach tense are
UradhiI II III IV
lsquoburnrsquo lsquotwirlrsquo lsquoblowrsquo lsquocryrsquopast wa-γal rima-n lapu-ntilde ruŋka-npresent wa-ma rima-al lapu-ntildea ruŋkafuture wa-maŋka rima-awa lapu-ntildeaŋka ruŋka-ŋkaimperative wa-ethi rima-ri lapu-yi ruŋka-γu
842 Verbs in Prefixing LanguagesVerbs in the prefixing languages of Australia have a quite different and usu-ally more complex structure I take the Wunambal language as an examplehere
In Wunambal there are two classes of verbs which I call I and II Thereappears to be no good semantic explanation as to why any particular verbbelongs to one class rather than the other In class I verbs features of thesubject are marked by a prefix but features of the object are marked by asuffix In class II verbs prefixes mark both subject and object
196 CHAPTER 8
Class I verbs begin with one of the prefixes marking person of the sub-ject (and if third person noun class membership)6
Wunambalng- first persong- second personb- w- m- a- n- nj- third person different noun classes
The number of the subject is variously marked Non-singular is marked by aprefix but specifically dual and trial subjects take an additional suffix Forexample
Wunambalgu-r-wanban gu-r-wanban-miyayou-NONSG-fall you-NONSG-fall-DUALlsquoyou (pl) fallrsquo lsquoyou two fallrsquo
Some tense-aspects and moods are marked by prefixes others by suffixes
Wunambalgu-nu-ma gu-ma-yayou-not-come you-come-FUTURElsquoyou didnrsquot comersquo lsquoyou will comersquo
For class I verbs the objectrsquos person and number are indicated by a suffix
Wunambalba-nbun-bun-wuru ba-nbun-bun-nguhe-spear-PRESENT-them he-spear-PRESENT-itlsquohe spears themrsquo lsquohe spears itrsquo
Class II verbs have much the same tense-aspect marking system as classI verbs The difference lies in the fact that the objectrsquos person and numberare marked by prefixes (underlined in the examples below) which precedethe subject prefixes
Wunambalgu-nga-nbun gu-r-nga-nbunyouOBJECT-I-hit youOBJECT-PL-I-hitlsquoI hit yoursquo lsquoI hit you (pl)rsquo
bu-r-nga-nbun bu-r-nga-ru-nbunhim-PL-I-hit him-PL-I-PL-hitlsquoI hit themrsquo lsquowe hit themrsquo
Both classes of verbs also have a set of derivational suffixes along thelines of those found in nonprefixing languages
Australian Languages Grammatical Overview 197
843 The Verb ComplexSo much semantic information is contained in the verb itself especiallybecause of the system of derivational suffixes that the verb complex in Aus-tralian languages often consists of no more than the fully inflected verb(Numerous examples of this have been given in preceding sections) But averb complex may include adverbs or locative demonstratives as in
GumbaynggirBirmading yilaamingrunPAST herePASTlsquo(She) ran over herersquo
Mudang giduudaming yaraang yilaaunablePAST onsandPAST there nearspeakerlsquo(He) was unable to go on the sand herersquo
85 Sentences851 Nominal SentencesEquational stative and locational sentences generally have no verb in Aus-tralian languages The most common order is subject + predicate
Subject Predicate
WajarriPakarli maparnpaman sorcererlsquoThe man is a sorcererrsquo
Warla parntiegg goodlsquoThe egg is goodrsquo
Kuwiyari marta-ngkagoanna rock-onlsquoThe goanna is on the rockrsquo
AnguthimriAngu rtalawatiI redlsquoI am redrsquo
Ma ngu-tyanaman clothes-withoutlsquoThe man is nakedrsquo
198 CHAPTER 8
852 Verbal SentencesIn discussing the structure of verbal sentences in Oceanic languages I intro-duced the distinction between accusative and ergative structures Most Aus-tralian languages have ergative structures so that the subject of a transitiveverb (the agent) is marked differently from the subject of an intransitiveverb The subject of a transitive verb is in the ergative case that of an intran-sitive verb in the absolutive case The object of a transitive verb is also in theabsolutive case
Look at the following examples from Wargamay
WargamaySubjectAgent Object VerbMaal gagaymanABSOLUTIVE golsquoThe man is goingrsquo
Maal-ndu ganal ngundayman-ERGATIVE frogABSOLUTIVE seelsquoThe man is looking at the frogrsquo
Ganal-ndu maal ngundayfrog-ERGATIVE manABSOLUTIVE seelsquoThe frog is looking at the manrsquo
In Wargamay the absolutive case is unmarked so the noun maal lsquomanrsquoappears as maal when it is the subject of an intransitive verb (as in the firstsentence) and also when it is the object of a transitive verb (as in the thirdone) The ergative case is marked by one of a number of suffixes (cf 821above) one of which is -ndu When maal lsquomanrsquo or ganal lsquofrogrsquo is the agent(the subject of a transitive verb) as in the last two sentences it must takethe ergative suffix
In many Australian languages however pronouns behave differentlyfrom nouns in marking subjects and objects Look now at the following set ofsentences
WargamaySubjectAgent Object VerbNgali gagaywetwo golsquoWe two are goingrsquo
Ngali ganal ngundaywetwo frogABSOLUTIVE seelsquoWe two are looking at the frogrsquo
Australian Languages Grammatical Overview 199
Ganal-ndu ngali-nya ngundayfrog-ERGATIVE wetwo-OBJECT seelsquoThe frog is looking at us tworsquo
These sentences show that although the noun ganal lsquofrogrsquo behaves erga-tively the pronoun ngali lsquowe tworsquo behaves accusatively It has the same form(ngali) when it is subject either of an intransitive or a transitive verb but adifferent form (ngali-nya) when it is the object of a transitive verb In thisrespect it behaves exactly like its equivalent weus in English Australianlanguages like Wargamay that treat nouns and pronouns differently are re-ferred to as split-ergative languages7
Because Australian languages clearly mark the case or function of nounphrases in a sentence by affixes to the noun phrase (as in Wargamay) byaffixes to the verb or in both of these ways it is obvious from looking at anoun phrase what its function in a sentence is Because of this ldquothe orderof words and phrases can in most Australian languages be extraordinarilyfree it has little or no grammatical significance A preferred order can usu-ally be perceivedhellipBut there can be unlimited deviation from this preferredorder dictated partly by discourse considerations (lsquotopicrsquo and the like) andpartly by the whim of the speakerrdquo (Dixon 1980 441)
Where there is a preferred word order it is usually subject + verb in in-transitive sentences and agent + object + verb in transitive sentences asin the Wargamay examples above Object + agent + verb however is just asfrequent Both versions of this Wargamay sentence are acceptable
WargamayAgent Object VerbGanal-ndu ngali-nya ngundayfrog-ERGATIVE wetwo-OBJECT seelsquoThe frog is looking at us tworsquo
Object Agent VerbNgali-nya ganal-ndu ngundaywetwoOBJECT frog-ERGATIVE seelsquoThe frog is looking at us tworsquo
This is possible because (1) it is clear from the suffix -ndu that ganal-ndulsquofrogrsquo is the agent and (2) it is also clear from the suffix -nya that ngali-nyalsquous tworsquo is the object
Other phrases are also relatively free as far as their order is concernedsometimes occurring before the verb and sometimes after it It is rarehowever for the verb to occur in sentence-initial position In the followingexamples the verb complex is underlined
200 CHAPTER 8
BandjalangMali-yu ngagam-bu yalany-dyu giyay bunybeh-lathe-ERGATIVE dog-ERGATIVE tongue-with salt lick-PRESENTlsquoThe dog is licking salt with its tonguersquo
YidinyWaguuja-nggu wagal bunja-ng banggaal-daman-ERGATIVE wife hit-PRESENT axe-withlsquoThe man hit his wife with an axersquo
WajarriYamaji-lu kuka marlu ngura-ki kangkarni-manyaman-ERGATIVE meat kangaroo camp-to bring-PRESENTlsquoA man is bringing kangaroo meat to the camprsquo
Australian languages do not have a passive construction but they do havesomething similar In accusative languages the original object in an activesentence becomes the subject of the passive sentence and the original activesubject is either moved to a peripheral phrase (The man chopped down thetree gt The tree was chopped down by the man) or deleted altogether (Thetree was chopped down) Some Australian languages have what is called anantipassive Look first at the following normal ergative Dyirbal sentence
DyirbalObject Agent VerbBala yugu banggul yara-nggu gunba-nit tree heERGATIVE manERGATIVE cut-PASTlsquoThe man was cutting the treersquo
In the antipassive the agent (banggul yara-nggu lsquothe manrsquo) becomes thesubject of what is now an intransitive verb and the object (bala yugu lsquothetreersquo) becomes a peripheral phrasemdashin this case a dative phrase The verbis also marked differently Here is the antipassive form of the sentenceabove
DyirbalSubject Verb DativeBayi yara gunbal-nga-nyu bagu yugu-guhe man cut-ANTIPASSIVE-PAST itDATIVE tree-DATIVElsquoThe man was cutting the treersquo
Note that the translations of the two sentences are the same Dixon saysthat in Dyirbal at least ldquoa regular transitive sentence and its antipassivecorrespondent hellip have the same basic meaning and differ only in emphasis
Australian Languages Grammatical Overview 201
rather like an active and its corresponding passive in Englishrdquo (Dixon 1980449) Like the peripheral agent in a passive sentence the peripheral dativephrase in an antipassive sentence can be deleted
DyirbalSubject VerbBayi yara gunbal-nga-nyuhe man cut-ANTIPASSIVE-PASTlsquoThe man was cuttingrsquo
A small group of languages mainly in Arnhem Land the Pilbara regionand the Kimberleys in the northwest of the continent offer exceptions tothese generalizations They are accusative not ergative in structure thepreferred order is frequently SVO (though OVS is also common) and somehave a passive Lardil illustrates these languagesrsquo structure
LardilSubject Verb ObjectOtherPirngen rikurwoman crylsquoThe woman is cryingrsquo
Pirngen rnethakun rtang-anwoman hit man-OBJECTlsquoThe woman hit the manrsquo
Rtangka rnethakun pirngen-inman hit woman-OBJECTlsquoThe man hit the womanrsquo
Rtangka rneyikun pirngen-inman hitPASSIVE woman-OBJECTlsquoThe man was hit by the womanrsquo
But these languages are in the minority and most Australian languages ad-here to the ergative model
202 CHAPTER 8
PART THREE
The Socialand Cultural Context
CHAPTER
9
Languages in Contact
Languages are normally not spoken in totally isolated communities Peoplespeaking one language usually come into contact either occasionally or ona more regular basis with speakers of one or more other languages and thesmaller the society that speaks a particular language the greater is the like-lihood of their being in contact with outsiders This social contact very oftenhas both major and minor linguistic effects
91 The Social Context of Language Contact911 Peaceful Contact between Settled SocietiesOne common kind of social contact between different language communi-ties in the Pacific is that between relatively equal and settled societiesIn many parts of the region for example marriage regulations require aman to marry a woman from outside his own clan and community Thispractice of exogamy often means that husband and wife speak differentlanguages Where a number of men in the same village choose wives fromthe same outside community a foreign-language enclave will form at leasttemporarily in the village Since women have the primary responsibility forlooking after younger children those children will often grow up hearingtwo languages spoken in the home The women may not much influencethe way the men speak but they do influence how their children speakThese children often end up incorporating some aspects of their mothersrsquolanguage into their own
A second kind of peaceful contact involves regular trade There are nu-merous cases in the Pacific where for example people of a coastal village
205
trade with inland villagersmdashthe former supplying fish and other marine pro-duce the latter vegetables and other non-maritime commodities Such asituation occurred in Central Papua where the coastal Motu traded with theinland Koita and Koiari people Many words for maritime concepts in Koitaare originally Motu words while the Motu have taken into their languageKoita words for non-maritime things
912 Peaceful Contact Involving TravelNot all trade takes place between sedentary peoples it may also involvesome or all of the parties traveling considerable distances In the westernPacific for example anthropologists have documented such large-scaletrading complexes as the hiri of the Gulf of Papua the kula of the islands ofthe Milne Bay area and the moka of the highlands of Papua New Guinea Agreat deal of long-distance trade in all kinds of commodities seems to haveoccurred in nearly every part of Australia and the Pacific
Such trading expeditions of course bring people speaking different lan-guages together at least for short periods and also often bring new thingsand ideas into at least one of the societies involved In such cases borrow-ing or copying often takes place That is the society into which somethingnew is introduced often takes the word for that thing from the language ofthose introducing it This is how English acquired such words as alcoholcurry tomato pasta tapioca sago and hundreds of others
Migrationmdasheither temporary or permanentmdashalso brings people speak-ing different languages together Temporary migration at least in the Pa-cific generally means that people leave their home area to work in townsor on plantations or ships for a period of time and then return home oftenbringing with them new things and ideasmdashalong with the words for them insome other language Permanent migration involves long-term settlementin a new area often because of overcrowding or sociopolitical problems athome or because of natural disasters like volcanic eruptions A smallishcommunity speaking one language may live in the middle of a larger com-munity speaking a different language and the potential exists for each lan-guage to influence the other
The whole of the Pacific region was settled from its western extremesand Australia was probably settled from the north Some of these migrantswould have been the first people to settle a particular area while otherswould have come into contact with descendants of the original settlersIn more recent times not only have rural people moved into urban areasbut whole communities have been relocated Mission stations all over theregion government settlements in Australia and the resettlement of the Ba-
206 CHAPTER 9
nabans on Rabi Island in Fiji represent three such cases There have alsobeen significant movements of populations from Micronesia and Polynesiainto the United States and New Zealand (see chapter 2) All of these situa-tions bring languages into contact with various degrees of closeness
913 Conquest Colonization and ConversionPolitico-military takeovers by one society of another represent a less peace-ful kind of social and linguistic contact Once again the Pacific aboundsin examples of which the Tongan domination of the Lau group in Fiji andthe warlike Orokaiva and Mailu in Papua New Guinea who enslaved con-quered peoples are three As with the Norman conquest of England athousand years ago these takeovers produced dramatic changes in lan-guage as the conquered peoples were forced to learn their conquerorsrsquolanguage to survive
European and Asian colonization of Australia and the Pacific rep-resents a more recent but thoroughgoing example of politico-militaryconquest The Spanish Dutch Germans French and Japanese as well asEnglish speakers from a number of nations have all made incursions intothe region over the last four centuries French and English currently beingthe dominant metropolitan languages in the Pacific These outsiders intro-duced new forms of government and education brought in a vast numberof new technological items and social customs and were responsible forthe establishment of plantations and urban centers English and Frenchhave been the major languages of government education and inter- andintraregional communication and are looked on in many parts of the Pa-cific as the prestige languages Because of both the attitudes toward thesetwo European languages and the new concepts introduced by EuropeansEnglish and French have had a considerable influence on most Pacific lan-guages
The founding of missions preceded colonization in some cases and fol-lowed it in others This process could be viewed as a conquest of a differentkindmdashthe displacement of traditional religious systems in favor of westernChristian beliefs and religious practices as missionaries aimed for a con-quest of the souls and minds of aboriginal Australians and Pacific IslandersThe establishment of churches and schools as well as the more or less suc-cessful abolition of some traditional customs resulted in the introduction ofnew words for new conceptsmdashin some cases even new ways of speaking asformal prayers and hymns were developed
In multilingual Melanesia particularly missions were responsible forsetting up certain vernaculars as church languages for example Tolai
Languages in Contact 207
Gedaged Yabecircm Kacircte Dobuan Suau and Wedau in Papua New GuineaRoviana in Solomon Islands and Mota in Vanuatu Faced with a multiplicityof languages in a relatively small area missionaries often chose one lan-guage as the language of the mission requiring speakers of neighboringand usually related languages to use the chosen language in religious con-texts This practice has helped create a complex situation in which bothEuropean and Pacific church languages influence other languages in theregion
92 The Linguistic Effects of Contact921 Lexical ChangeVirtually all languages borrow or copy words from other languages Englishis an excellent example as it has taken in thousands of words from verydiverse sources In the Pacific the influence of both local and intrusive lan-guages on other Pacific languages has led to the incorporation of new wordsinto those languages
Speakers of the non-Polynesian languages of Southern Vanuatu whohave probably been in the area for well over two thousand years came intocontact with speakers of the Polynesian language West Futuna about a thou-sand years ago These immigrant Polynesians introduced their neighbors tokava-drinking and refined their maritime skills especially those involvingdeep-sea fishing The words below are West Futuna loans into Kwamera aTanna language suitably adapted to Kwamerarsquos phonological and grammat-ical structure (Lynch 1994 1996)
Kwamera Loans from West Futuna
Kava terminologyKwamera West Futunanɨkava lsquokavarsquo kavatapuga lsquochiefrsquos kavarsquo tapugatamafa lsquoritual spitting of kavarsquo taumafanafunu lsquofood eaten after drinking kavarsquo fonoakona lsquodrunkrsquo konataporoka lsquokind of canoe-shaped
kava bowlrsquota poruku lsquokind of canoersquo
nafaacuteu lsquokava strainerrsquo fao lsquococonut branchused as a kavastrainerrsquo
208 CHAPTER 9
Maritime terminologyKwamera West Futunatira lsquomastrsquo shiranɨkiatu lsquooutrigger boomrsquo kiatokwan-metau lsquofishhookrsquo metaotakwarau lsquoprevailing windrsquo tokorautafra lsquowhalersquo taforatataua lsquobarracudarsquo tataotagarua lsquosea snakersquo tagaroa
Trukese provides a good example of the influence of succeeding colonialpowers Much of western and central Micronesia was under Spanish controlfrom the late seventeenth century until the Spanish-American War in 1898when Guam was ceded to the United States and the rest of Spainrsquos posses-sions went to Germany which had already colonized the Marshall Islands tothe east Japan succeeded Germany at the outbreak of World War I and theUnited States succeeded Japan at the end of World War II The influence ofeach of these colonial languages can be seen in borrowed words in Trukese(Goodenough and Sugita 1980)
Trukeseantiyos lsquofishing gogglesrsquo lt Spanish anteojoskoopwure lsquocorrugated ironrsquo lt Spanish cobre lsquocopperrsquopaatere lsquopriestrsquo lt Spanish padre
kiiwuacutefer lsquosuitcasersquo lt German Kofferkkumi lsquorubberrsquo lt German Gummimaak lsquomoneyrsquo lt German Mark (monetary unit)
kooyeng lsquoplaygroundrsquo lt Japanese kōen lsquoparkrsquoosiroy lsquobaby powderrsquo lt Japanese oshiroiramuacutene lsquomarblesrsquo lt Japanese ramune
miniyon lsquomillionrsquo lt English millionpinakpwoot lsquoblackboardrsquo lt English blackboardsekit lsquojacketrsquo lt English jacket
Samoan is a good example of missionary influence on a language Manynew words came into it from the biblical languages Greek Latin and He-brew
Samoanperitome lsquocircumcisersquo lt Greek peritomeagelu lsquoangelrsquo lt Greek angelos
Languages in Contact 209
tilsquoāpolo lsquodevilrsquo lt Greek diabolossātauro lsquocrossrsquo lt Greek stauroslsquoaila lsquodeer gazellersquo lt Hebrew lsquoayyallsquooreva lsquovulturersquo lt Hebrew lsquoorebh lsquoravenrsquolsquourosa lsquobearrsquo lt Latin ursus
922 Semantic ChangeContact may also bring about changes in the meanings of existing wordsin a language This may involve expanding the meaning of a word to referto something newly introduced For example Fijian dakai originally meantlsquobow (for shooting)rsquo and Lenakel kopwiel means lsquostonersquo but both have takenon the additional meaning lsquogun riflersquo In Ponapean sakau originally referredonly to kava but now it refers to any intoxicating beverage Nting meant lsquototattoorsquo in Ponapean but now also means lsquoto writersquo
Semantic change may also involve narrowing the meaning of a wordLenakel niko originally meant both lsquocanoersquo and lsquomoietyrsquo (since it was be-lieved that the first members of the two moieties arrived on the island in twodifferent canoes) But the form kenu (from English via Bislama) is now thecommon word for lsquocanoersquo and for most Lenakel speakers niko now meansonly lsquomoietyrsquo
923 Phonological ChangeWhen a language takes in words from another language it often adaptsthem to its own phonology The English words restaurant miracle princeroyal and court for example all derive from French but they are notpronounced as the French pronounce them They have been adapted toEnglish phonological patterns Sometimes however the copying of wordsfrom one language into another may bring about a change in the phonolog-ical system of the borrowing language either through the introduction of atotally new sound or through the reorganization of the existing sounds ina language
Dyirbal for example is typical of many Australian languages in thatthe phoneme l is not permitted at the beginning of words However theintroduction of words like lada lsquoladderrsquo and laymun lsquolemonrsquo from English hasbrought about a change in the phonological structure of Dyirbal which nowpermits word-initial l
Motu originally had no contrast between the sounds [t] and [s] [s] oc-curred before [i] and [e] while [t] occurred before other vowels Englishwords copied into Motu originally fit this pattern1
210 CHAPTER 9
Motu[sesi] lsquoshirtrsquo[makesi] lsquomarketrsquo[sini] lsquotinrsquo[tupu] lsquosouprsquo[topu] lsquosoaprsquo
Due to the persistent influence of English however younger generationsnow pronounce these words as follows
Motu[seti] lsquoshirtrsquo[maketi] lsquomarketrsquo[tini] lsquotinrsquo[supu] lsquosouprsquo[sopu] lsquosoaprsquo
What has happened here is that the distribution of [s] and [t] has changedand there is now contrast between them
In addition to changing the distribution of existing sounds in a languagecontact may also lead to the introduction of a new sound Samoan for exam-ple originally had an l but no r As the result of contact with other languageshowever a number of words with r have been introduced
SamoanlsquoAperila lsquoAprilrsquo lt Englishlsquoareto lsquobreadrsquo lt Greek artoslsquoario lsquosilverrsquo lt Tahitian ariofaresaio lsquophariseersquo lt Greek farisaiosmisionare lsquomissionaryrsquo lt Englishlsquooreva lsquovulturersquo lt Hebrew lsquoorebh lsquoravenrsquorosa lsquorosersquo lt Englishteropika lsquotropicsrsquo lt Englishlsquourosa lsquobearrsquo lt Latin ursus
924 Grammatical ChangeFinally contact between languages may also bring about changes in gram-matical structure Polynesian Triangle languages are normally verb-initial(see chapter 6)
Languages in Contact 211
TahitianVerb Subject ObjectlsquoUa tāpū te vahine lsquoi te vahiePAST cut the woman OBJECT the woodlsquoThe woman cut the woodrsquo
Polynesian Outlier languages however are much more flexible allowingboth VSO and SVO orders with SVO probably being more common
NukuoroVerb Subject ObjectNe kake ia te nuiPAST climb he the coconutlsquoHe climbed the coconutrsquo
Subject Verb ObjectIa ne kake te nuihe PAST climb the coconutlsquoHe climbed the coconutrsquo
This has almost certainly come about at least in part from contact withneighboring non-Polynesian languages which are almost exclusively SVO
A similar change seems to have occurred on the mainland of NewGuinea The original Oceanic languages spoken there almost certainly hadverb + object order (whether SVO or VOS is a matter of some discussionbut is irrelevant here) Then they came into contact with Papuan speakersfor whom SOV was the basic order and this contact led to a change in theOceanic languagesrsquo word order from SVO (or VOS) to SOV (Some examplesof languages with this order were given in chapter 6)
Let us look at one more example this time from the Papuan language Yi-mas
The formation of a negative verb from a positive one in Yimas is acomplicated affair involving alteration of the form and the positionof certain verbal affixes Many younger speakers do not know thismethod of negation but negate a verb by merely placing a particle inabefore it This is clearly a borrowing from Tok Pisin i no [= PREDICATEMARKER + NEGATIVE] but these speakers were totally unaware of its ori-gin regarding it as a native Yimas word until I pointed out its similarityto the Tok Pisin negative (Foley 1986 40)
212 CHAPTER 9
93 Three Case StudiesDirect and indirect inheritance of vocabulary in Rotuman borrowing as aresult of word taboo in Australia and contact between Austronesian andPapuan languages in northwest New Britain are three cases illustrating theeffects of language contact
931 RotumanThe Rotuman vowel system has undergone some interesting developments(chapter 5) But Rotuman is also a language where contact has led to a com-plex situation for the historical-comparative linguist
Rotuman words exhibit two sets of correspondences with proto-formshellip I propose to speak of directly [set I] and indirectly inheritedwords [set II] rather than inherited and loan words in order to em-phasize that all of the words with etymologies were once part of alanguage ancestral to Rotuman in the comparativistrsquos sense Some ofthem however re-entered Rotuman from a collateral related languageafter undergoing changes other than those which affected forms whichhad remained continuously in the Rotuman line (Biggs 1965389ndash390)
I am concerned here exclusively with the development of some of the conso-nants in Rotuman
What appears to have happened in Rotuman is this The original settlerswould have brought with them a version of Proto Central Pacific (PCP) Overtime some of the consonants changed their pronunciation with the resultthat the following regular developments can be identified (Biggsrsquo directly in-herited or Set I correspondences)2
PCP v pb t d r dr l k g ʔRotuman (Set I) h p f t r t l ʔ k oslash
Some examples of this set of sound correspondences are given below
Proto Central Pacific Rotuman (SetI)vitu lsquosevenrsquo hifukuli lsquoskinrsquo ʔuliʔatu lsquoline rowrsquo afuviri lsquoplaitrsquo hiri
Subsequent to the original settlement of Rotuma there seem to havebeen at least two later ldquoinvasionsrdquo by people speaking languages different
Languages in Contact 213
from but related to pre-Rotuman These invasions resulted in fairly large-scale borrowings of vocabulary Biggs (1965 411) sums up the situation asfollows ldquoIt is clear that Rotuman has borrowed extensively from a relatedlanguage or languageshellip Rotuman traditions are definite in associating atleast two occupations of their island with the Samoa-Tonga area particu-larly the islands of Savaiʔi [in Samoa] and Niuafoʔou [in Tonga]rdquo The lan-guages of the invaders had made somewhat different changes to the ProtoCentral Papuan consonant inventory Of Biggsrsquo corpus of Rotuman wordswith known etymologies 38 percent belong to Set I (as above) but 29 per-cent belong rather to the set of indirectly inherited correspondences (Set II)which are given below along with Set I for comparison3
PCP v pb t d r dr l k g ʔRotuman (Set I) h p f t r t l ʔ k oslashRotuman (Set II) f p t t roslash r r k k ʔ
Here are some examples of words containing Set II correspondencesin each case the expected but non-occurring Set I form is given as well(marked with a double asterisk)
Proto Central Pacific Rotuman (Set II)viti lsquospring uprsquo fiti (expected Set I hifi)tuki lsquopoundrsquo tuki (expected Set I fuʔi)kolo lsquodesirersquo koro (expected Set I ʔolo)robe lsquooverhangrsquo ope (expected Set I rope)
In some cases the same word has come into the language twice first di-rectly (Set I) and later indirectly (Set II) though with slight differences inmeaning For example
PCP RotumanSet I Set II
kuli lsquoskinrsquo ʔuli lsquoskinrsquo kiria lsquoleprosyrsquovidi lsquojump springrsquo hiti lsquostart with surprisersquo fiti lsquojumprsquotoka lsquocome ashorersquo foʔa lsquocome ashorersquo toka lsquosettle downrsquo
932 Word Taboo in AustraliaAustralian languagesmdashas well as many others in the Pacificmdashare character-ized by a system of word taboo This can take a number of forms One verycommon one is that ldquoa personrsquos name cannot be spoken for some time af-ter his death What is more any normal vocabulary itemmdashnoun adjectiveverb etcmdashthat is similar in form to the banned name must also be tabooedrdquo(Dixon 1980 28) Imagine if this were to apply in English When someone
214 CHAPTER 9
called Bill dies we could not use the word bill (meaning either lsquoaccountrsquo orlsquobeak of a birdrsquo) nor probably could we use phonologically similar words likebuild billet billy and perhaps pill We would have to find new wordshellip atleast for a time
In Australia the tabooed word is sometimes replaced by a synonym ornear synonym from within the language In our imaginary example abovebill could then be replaced by (1) check or account and (2) beak ldquoBut moreoften a new word will be borrowed from the language of a neighboring triberdquo(Dixon 1980 28) Examples
1 ldquoIn 1975 a man named Djaumlyila died at Yirrkala and as a result thecommon verb djaumll- lsquoto want to be desirous of was proscribed andreplaced by duktuk- probably a verb from another Yolŋu dialectthat did have this set of meaningsrdquo (Dixon 1980 28)
2 ldquoIn 1977 a Djapu man named Djewiny died and the loanword dhelsquotearsquo was at once tabooed at Yirrkala another loanword gopi lsquocof-feersquo had its meaning extended also to cover lsquotearsquo (little coffee is infact drunk at Yirrkala if disambiguation is necessary it can be re-ferred to as gopi yuwalk lsquoreal coffeersquo)rdquo (Dixon 1980 122)
933 Northwest New Britain
In a series of studies Thurston (1982 1987 1992) has documented the ef-fects of language contact among a number of languages spoken in the north-west of New Britain The area Thurston discusses is currently occupied by anumber of Oceanic languages (important to this discussion are the coastallanguages Kabana Amara Kove and Lusi) and the Papuan language AnecircmldquoThe Anecircm are now completely surrounded by speakers of Austronesian lan-guageshellip Evidence suggests that Anecircm is the sole surviving member of anon-Austronesian language family that once extended over much of whatis now West New Britain Province West of the Willaumez Peninsula all ofthese languages except Anecircm have been replaced by Austronesian lan-guages which retain features of a non-Austronesian substratumrdquo (Thurston1992 125)
Contact in this area between speakers of different languages related andunrelated has been going on for a long time with quite far-reaching effects
Generations of marriage and trade across linguistic boundaries thelongstanding tradition of regional multilingualism and the spread oflanguages by way of language shift have all conspired to produceregional similarity in phonology syntax semantics social structureeconomy cosmology and valueshellip Aside from lexical form the speak-
Languages in Contact 215
ers of Austronesian languages in northwestern New Britain sharemuch more with the Anecircm than they do with speakers of distant Aus-tronesian languages that are lexically more similar (Thurston 1992125)
There has been a large amount of lexical copying in both directions be-tween the Oceanic languages and Anecircm but of more interest are changesin grammar as a result of this prolonged contact Some of the grammaticalfeatures of Lusi (and some of the other Oceanic languages in the area) thatseem to have been introduced from Anecircm or its extinct relatives are de-scribed below
1 The reciprocal is marked by a suffix to the verb rather than by a pre-fix as is widespread in Oceanic languages (see 646 above)
LusiTi-rau-nga-rithey-hit-RECIPROCAL-themlsquoThey fought each otherrsquo
AnecircmI-pəl-akthey-hit-RECIPROCALlsquoThey fought each otherrsquo
2 Tenseaspect negation and similar categories are marked at the endof the verb phrase rather than by prefixes or preverbal particles asis common in Oceanic languages (see 642 above)
LusiI-rau γaea maohe-hit pig notlsquoHe didnrsquot kill a pigrsquo
I-la pa Rabaul γasilihe-go to Rabaul COMPLETIVElsquoHe has (already) gone to Rabaulrsquo
AnecircmU-b-ɨ aba mantuhe-kill-it Pig notlsquoHe didnrsquot kill a pigrsquo
U-k axɨ Rabaul bizanghe-go to Rabaul COMPLETIVElsquoHe has (already) gone to Rabaulrsquo
216 CHAPTER 9
3 Lusi has two postpositions aea purposive and iai locative as wellas a handful of prepositions Oceanic languages with postpositionstend to be restricted to the New Guinea mainland Although Anecircmdoes not have postpositions Thurston suggests that Lusi probablyacquired its postpositions as a result of contact with one of the nowextinct Papuan languages of the area
The contact has not been one way however An inclusiveexclusive dis-tinction in the first person is almost universal in Oceanic languages but isexceedingly rare among Papuan languages (see 71 above) Anecircm shows thisdistinction in possessive suffixes (though not in other pronominal forms)and it also has the inalienableedibleneutral contrast in possessive con-structions typical of western Oceanic languages but nonexistent in Papuanlanguages (Thurston 1987 91) The long-term intimate contact between lan-guages in this area has clearly produced major changes in the structure ofthese languages
934 ldquoMixedrdquo LanguagesThe Anecircm-Lusi situation just described gives rise to the following questionHow much can Language A be influenced by Language B and still remainLanguage A Or in different words can a language be truly ldquomixedrdquo notderiving from just one ancestor but in a sense from two There are numer-ous theoretical and philosophical questions involved here and they havegenerated considerable debatemdashnot to mention heat and acrimonymdashin thediscussion of certain languages in the Pacific especially in Melanesia4
I do not wish to go into these philosophical and theoretical questionshere There are however a number of cases where the influence of one ormore languages on another has led different reputable linguists to classifylanguages differently To take some extreme examples the following havebeen classified as Papuan by some linguists and as Austronesian by othersMaisin in the Oro Province of Papua New Guinea Magori and its neighbor-ing moribund relatives on the south coast of Papua the languages of SantaCruz and the Reef Islands in Solomon Islands and the languages of Ane-ityum New Caledonia and the Loyalty Islands
In all except the Reefs-Santa Cruz situation the general view today is thatthe languages involved are originally Austronesian Maisin and Magori havebeen very heavily influenced by Papuan languages whereas Aneityum andthe New Caledonian languages are probably so aberrant in their phonologi-cal and grammatical histories that they happen not to look very AustronesianThe Reefs-Santa Cruz languages however were probably originally Papuanlanguages that have been very heavily influenced by Austronesian languages
Languages in Contact 217
These decisions have been reached by ignoring vocabulary for the most partand looking instead at the core of the languagesrsquo grammatical systems
Even languages like these are not truly mixed in the sense of having twoco-equal ancestors They are however cases where the influence of anotherlanguage has been so strong as to make genetic affiliation very difficult todetermine
94 Historical ImplicationsIn chapter 1 I discussed the way in which historical inferences can be drawnfrom an examination of the relationships between languages Borrowing ofvocabulary phonology and grammar does not constitute genetic relation-ship The fact that the Fijian words sitoa and sitaba have been copied fromEnglish ldquostorerdquo and ldquostamprdquo does not mean that Fijian is related to English
But although the relationship between Fijian and English is not a ge-netic one there is still a historical connection between them An examina-tion of English words copied into Fijian for example provides us with in-formation of a cultural-historical nature They indicate what kinds of thingswere introduced to Fijian society and culture by English colonials mission-aries and settlers and what kinds of changes took place in Fijian societyand culture as a result of external influence
Let us go back to the example of Kwamera loans from West Futuna toillustrate this in a bit more detail (Recall that this copying took place longbefore Europeans came to the area) The following words relating to kavaand kava-drinking were borrowed by Kwamera
Kwameranɨkava lsquokavarsquotapuga lsquochiefrsquos kavarsquotamafa lsquoritual spitting of kavarsquonafunu lsquofood eaten after drinking kavarsquoakona lsquodrunkrsquotaporoka lsquokind of canoe-shaped kava bowlrsquonafaacuteu lsquokava strainerrsquo
Clearly this indicates a significant change in Kwamera culture This list ofwords represents not just the random borrowing of a few items but the tak-ing over of a whole cultural complexmdashthe preparation and drinking of kavawith its attendant rituals and behaviors (Lynch 1996)
Nor is this the end of the story of cultural contact between the Polyne-sian and non-Polynesian societies of southern Vanuatu The non-Polynesian
218 CHAPTER 9
languages have also borrowed heavily from Futuna maritime vocabularyespecially where long-distance voyaging or deep-sea fishing is concerned(Lynch 1994) Futuna on the other hand has borrowed a number of wordsfor varieties of yam taro and breadfruit from their more horticulturallyinclined non-Polynesian neighbors And perhaps most interesting of all Fu-tuna speakers appear to have been responsible for introducing a moietysystem to neighboring Tanna This system fell into desuetude on Futuna butwas reintroduced hellip by the Tannese (Lynch and Fakamuria 1994) An exam-ination of borrowed items in a language can give us significant informationabout the nature of contact-induced cultural change The influence of non-Pacific languages on those of the Pacific has been considerable over the pastcouple of centuries or so but Pacific Islanders have been moving around theregion for thousands of years and contact between languages has been partof the linguistic scene in the Pacific for the whole period
Languages in Contact 219
CHAPTER
10
Pidgins Creoles and Koines
Contact may have quite drastic effects on a language But it may also lead tothe creation of totally new languages which in some senses at least qualifyas ldquomixedrdquo languages Three of these new languages are in terms of num-ber of speakers among the largest languages spoken in the Pacific today(although not all speakers of any of these languages speak them as theirmother tongue) I use the term Melanesian Pidgin as a cover term for thethree languagesdialects known as Tok Pisin in Papua New Guinea Pijin inSolomon Islands and Bislama in Vanuatu spoken in all by perhaps three mil-lion people1 Hiri Motu is spoken in Papua New Guinea by about a quarter ofa million people And Fiji Hindi one of the two major languages of Fiji hasmore than 300000 speakers This chapter looks at these three languagesand at similar languages in various parts of the Pacific
101 Pidginization Creolization and KoineizationHow do languages like Melanesian Pidgin develop What is it about certainkinds of contact situations that gives rise to new languages
The term pidgin or pidgin language refers to a language that developsin a multilingual contact situation where the contact between the differentgroups is prolonged but relatively restricted Trade relationships planta-tions and shipsrsquo companies are typical breeding grounds for such lan-guages and in situations like these the process of pidginization begins totake place All speakers of a pidgin language use it as a second language tocommunicate with speakers of other languages when there is no other com-mon language In comparison with the first languages of its speakers a pid-
220
gin is usually simplified in pronunciation grammar and vocabulary In manycases especially in colonial situations the vocabulary of the pidgin is drawnmainly from the politically dominant (ie colonial) language whereas thegrammar is often based on the language(s) of the colonized people
Urbanization and marriage between people from different linguisticbackgrounds can turn a pidgin into peoplersquos first language especially whenthose people are the children of such mixed marriages growing up in townsIn these cases the pronunciation grammar and vocabulary of the originalpidgin language tends to expand rapidly and considerably The languagebecomes more complex because it is being used for all the communicativepurposes of a ldquonormalrdquo language This process of expansion is referred toas creolization A creole or a creole language is a language that has de-veloped from a pidgin but which is now the first language of many of itsspeakers
A different kind of mixingmdashwhat is known as dialect mixingmdashproducesa different kind of language When people speaking different geographicaldialects of a language are relocated and thrown together in a new commu-nity what is known as a koine often develops through a process known askoineization Each dialect contributes some elements and the resultantkoine is a blend of the original dialects While Melanesian Pidgin and HiriMotu are the result of the processes of pidginization and creolization FijiHindi is a koine2
102 Melanesian PidginMelanesian Pidgin and various Australian creoles are referred to asldquoEnglish-basedrdquo or ldquoEnglish-lexifierrdquo creoles This means simply that thebulk of their vocabulary is derived from English though some vocabularyand much of the grammar may have different origins (This does not meanthat these languages are ldquobroken Englishrdquo or ldquobaby-talkrdquo languages afterall although a very significant proportion of the vocabulary of Englishcomes from Romance languages like Latin and French we donrsquot considerEnglish to be ldquobroken Romancerdquo)
1021 Historical BackgroundThe late eighteenth and early nineteenth centuries saw the first prolongedand continuous contact between people living in the Pacific and outsiders Inthe Pacific Islands European explorers and missionaries were followed bywhalers sandalwooders pearlers becircche-de-mer3 fishermen and tradersall of whom had regular if sporadic contact with at least some people in
Pidgins Creoles and Koines 221
some Pacific islands In Australia and New Zealand of course contact wasmore intense in many areas as a result of European settlement This contactintensified during the nineteenth century as labor recruiters began recruit-ing Pacific Islanders to work on plantations in various parts of the regionespecially Samoa Fiji and Queensland In Queensland there was also somecontact between Pacific Islanders and aboriginal Australians who them-selves were often moved from their tribal homelands into situations wherethey lived and worked with speakers of other languages
In all of these situations numerous fairly unstable pidgins developedIn Melanesia and parts of Australia these unstable pidgins developed intorelatively stable languages as people who had learned different varietiesin different parts of the Pacific came into contact The contact betweenEuropeans mainly English speakers Pacific Islanders (almost exclusivelyspeakers of Oceanic languages) and aboriginal Australians was responsiblefor the very significant English input into the vocabulary of these creolesBut it was not just this contact that was significant in the development ofMelanesian Pidgin The contact between Pacific Islanders from different lin-guistic backgrounds was important from the beginning became even moreso later on and was probably responsible for the Austronesian contributionto the grammar of Melanesian Pidgin
By the latter part of the nineteenth century English-based pidgins werespoken in various forms and with various levels of sophistication in almostthe whole of the Pacific Basin from New Guinea to Pitcairn Island and fromthe Marshalls and Hawailsquoi to New Caledonia and New Zealand In most ofthese places however the pidgins died out4 In some places like New Cale-donia and the British colony of Papua (the southern half of what is now PapuaNew Guinea) this was as a result of government policy The governmentswere strongly opposed to a ldquobastardrdquo form of English being used thoughpossibly for different reasons (the British in Papua because they saw it as aldquobastardrdquo language the French probably because they saw it as a form ofEnglish) In other places like most of the countries of Polynesia the pidginsimply became unnecessary as people from other parts of the Pacific stoppedbeing recruited to work on plantations in these countries and as educationallevels improved In Samoa for example the cessation of labor recruiting andthe establishment of schools meant that pidgin English was no longer neededSamoan was the language of communication between Samoans while firstGerman and then English were used for communicating with foreigners
The situation in Melanesia and Australia was very different First thecountries are geographically larger and linguistically more diverse thanthose of Polynesia and Micronesia and it was more difficult for governmentsto exercise strong control over language use Second although recruitment
222 CHAPTER 10
of Melanesian laborers to overseas plantations stopped soon after 1900 thissimply meant that laborers began moving around their own country work-ing on newly established plantations frequently outside their own languagecommunities Any plantation might have a labor force drawn from a largenumber of different language groups Third as a result of this internal mo-bility men often married women who spoke a different language and thepidgin would have been the only language used in the home Finally theestablishment of urban centers attracted people speaking a multiplicity oflanguages from far and wide
Social conditions in Melanesia and in parts of Australia therefore wereripe not just for the preservation and retention of the pidgin but also forits development into a creole Children grew up speaking it as their firstlanguage adults who had not returned to their traditional homes for manyyears found that they were using the pidgincreole more and more and theirown language less and less As the twentieth century progressed Melane-sian Pidgin became the language of the people in what were to become theindependent states of Papua New Guinea Solomon Islands and VanuatuAustralian creoles and varieties of English spoken by Aboriginal people ac-quired similar importance
1022 Different HistoriesThe Melanesian Pidgin spoken by Papua New Guineans Solomon Islandersand ni-Vanuatu is recognizably the same languagemdashwith recognizable dif-ferences between how it is spoken in each of these three countries Thefollowing examples show both the similarities and the differences betweenthese three varieties
Tok PisinDispela pikinini i sindaun i stap na kaikai kiau wantaim kek
PijinDesfala pikinini i sidaon an kaekae eg weitim kek
BislamaPikinini ya i stap staon mo kakae eg wetem gatolsquoThis child is sitting down and eating eggs and cakersquo
Where do these kinds of differences come from To answer this question weneed to look more closely at the historical development of this language
Men from Vanuatu were first recruited to work on plantations inQueensland and Fiji in the 1860s and a little later men from the Solomon Is-
Pidgins Creoles and Koines 223
lands were recruited for the same work so there was considerable contactbetween ni-Vanuatu and Solomon Islanders at this time Only a few peoplefrom this part of the Pacific however were recruited to work in Samoa andthen only for a short time
Men from the German colony of New Guinea however did not go toQueensland or to Fiji which were British colonies Rather starting in the1880s they went to work on the plantations in Samoa then a Germancolony For a few years they were in contact with ni-Vanuatu and SolomonIslanders from whom they would have learned the basics of Melanesian Pid-gin but for the next few decades the New Guinea version of MelanesianPidgin known today as Tok Pisin developed in isolation both from othervarieties of the language and from English The German and Samoan lan-guages contributed some words to early Tok Pisin although many of thesehave disappeared The major contributing languages (other than English)have been Tolai (cf kiau lsquoeggrsquo in the example above) and other Austronesianlanguages of New Britain and New Ireland since Rabaul (where Tolai is spo-ken) was the headquarters of German New Guinea and the place wheremost of the laborers were recruited from or returned to
Pijin and Bislama did not undergo any of these influences However be-cause the French jointly ruled the New Hebrides (now Vanuatu) with theBritish for most of the twentieth century Bislama has incorporated a num-ber of words of French origin (like gato lsquocakersquo in the example above) It hasalso taken in quite a few words from local languages Neither French (forobvious reasons) nor local languages (for less obvious reasons) have madeany significant contribution to Pijin in Solomon Islands however The differ-ent colonial histories of each country along with different labor-recruitmentpatterns meant that there were significant differences in the contact situa-tions while each version of the language was developing
1023 The Structure of Melanesian PidginIn recent years the influence of English on Melanesian Pidgin has becomeeven more dominant than in the past not only in terms of vocabulary but tosome extent also in pronunciation and grammar as well At the same timethere is considerably more contact today between Melanesians from differ-ent countries and interdialectal influence is also beginning to be seen Oneof the features of a language undergoing creolization is that different peo-ple speak it with different degrees of fluency For some people it is theirfirst language For others it is very much a second language and the waythey speak it is often influenced by their first language Those who havebeen educated in English often incorporate words and other linguistic fea-
224 CHAPTER 10
tures from English into their Pidgin while less educated speakers do thismuch less frequently And although all languages are changing languageslike Melanesian Pidgin are changing much faster than others
Hence it is often difficult to say exactly what is or is not ldquoinrdquo a languagelike Melanesian Pidgin I try to describe the variety spoken by fluent but nothighly educated speakers but comment from time to time on common vari-ations from these patterns
Sound System
Melanesian Pidgin has the same five-vowel system as is found in the major-ity of the languages of the Pacific
i ue o
a
Educated speakers however sometimes incorporate English vowels intotheir speech An educated Papua New Guinean might say baeligŋ or baeligŋkfor lsquobankrsquo whereas someone less educated will say beŋ
The basic consonant system is also similar to that found in many of theOceanic languages of Melanesia
P t kb d gf s hv ʤm n ŋ
lr
w y
Notable omissions as a result of the pidginization process are the commonMelanesian fricatives x and γ which donrsquot occur in English and the Eng-lish fricatives θ eth z ʃ ʒ which are rare in Melanesian languages Both openand closed syllables may occur and consonant clusters are common
Variation in the pronunciation of consonants is of two kinds Pidginspeakers who also speak English often introduce phonemic distinctions fromthat language that are not made by less educated speakers So an educatedspeaker might say ʃu lsquoshoersquo and tʃetʃ or tʃətʃ lsquochurchrsquo whereas an unedu-cated speaker would be more likely to say su and sios
Pidgins Creoles and Koines 225
The other kind of variation is probably related to first-language interfer-ence Many speakers ldquoconfuserdquo similar sounds probably because thesesounds are not phonemically distinct in their own languages Among thepairs of sounds commonly confused by some speakers of Melanesian Pidginare p and b t and d k and g p and f t and s h and absenceof a consonant b and v s and ʤ f and v l and r v and wand n and ŋ For example some speakers of Tok Pisin say pis lsquofishrsquo andothers fis tasol lsquoonlyrsquo is often heard as tatol and while some speakerssayhaumas lsquohow muchrsquo others would say aumas or aumat
The orthography is fairly straightforward with ng being used to repre-sent ŋ (and j for ʤ in Pijin and Bislama) For most speakers voiced stopsdo not occur word-finally but etymological spellings are used in Pijin andBislama pik lsquopigrsquo and gut lsquogoodrsquo are written pig and gud in Pijin and Bis-lama but pik and gut in Tok Pisin The diphthongs ai oi and au arewritten ai oi and au in Tok Pisin but ae oe and ao in Pijin and Bislama sothe words for lsquorightrsquo lsquoboyrsquo and lsquohousersquo are rait boi haus in Tok Pisin butraet boe haos in Bislama
Sentence Structure
Melanesian Pidgin is a subject-predicate language and has both verbal andverbless sentences In verbal sentences the phrase order is SV in intransi-tive sentences and SVO in transitive sentences
Tok PisinSubject Verb ObjectWanpela man i kam 0one man PREDICATE come 0lsquoA man cameis comingrsquo
Maria i kilim pik bilong miMaria PREDICATE kill TRANS pig POSS melsquoMaria killedis killing my pigrsquo
Although there is no passive attention can be focused on the object of atransitive clause by moving it to the front of the sentence where it can be fol-lowed by the particle ia and a pause This often translates a passive Englishsentence Here is the object-focused version of the second sentence above
Tok PisinObject Subject VerbPik bilong mi ia Maria i kilimpig POSS me FOCUS Maria PREDICATE
killTRANS
226 CHAPTER 10
lsquoAs for my pig Maria killedis killing itrsquolsquoMy pig was killedis being killed by Mariarsquo
Verbless sentences follow a pattern similar to verbal sentences withsubject preceding predicate
PijinSubject PredicateHem i man blong mihe PREDICATE man POSS melsquoHe is my husbandrsquo
Pronouns
Pronouns in Melanesian Pidgin follow the Austronesian pattern They distin-guish at least three numbers and also show the inclusiveexclusive distinc-tion in the first person non-singular While the dual is common the trial isconsiderably rarer Here are the pronouns of Bislama
BislamaSingular
I miyou yuhesheit hem
Dualwe two INC yumituwe two EXC mitufalayou two yutufalathey two tufala
Trialwe three INC yumitrifalawe three EXC mitrifalayou three yutrifalathey three trifala
Pluralwe INC yumiwe EXC mifalayou yufalathey olgeta
There is virtually no morphophonemic variation in the pronouns5 Thesame form is used as an independent pronoun as subject or object or aftera preposition
Pidgins Creoles and Koines 227
BislamaMi hang-em ol klos blong mi long laenI hang-TRANS PL clothes POSS me on lineldquoI hung my clothes on the linerdquo
Hem i givim gato ya long yufala fromhe PREDICATE giveTRANS cake this to youPL becausehem i laekem yufala tumashe PREDICATE likeTRANS youPL verylsquoHeshe gave the cake to you (pl) because heshe likes you a lotrsquo
Nouns Noun Phrases and Prepositions
Nouns are almost universally invariable in form There are no articles andonly a small number of demonstratives Tok Pisin has dispela and Pijin des-fala lsquothisrsquo (sometimes lsquothatrsquo) both of which precede the noun Bislama yalsquothis thatrsquo follows the noun The following sentences all mean much thesame thing
Tok PisinDispela man i laik-im dispela merithis man PREDICATE like-TRANS this woman
PijinDesfala man i laek-em desfala womanthis man PREDICATE like-TRANS this woman
BislamaMan ya i laekem woman yaMan this PREDICATE likeTRANS woman thislsquoThis man likes thisthat womanrsquo
Adjectives numerals and other quantifiers normally precede the nounin a noun phrase6 although there are some modifiers that follow the headIn Tok Pisin all monosyllabic adjectives and numerals as well as some thathave two or more syllables must occur with the suffix -pela in this contextIn Pijin and Bislama the corresponding suffix -fala is less frequently usedoften occurring only when the adjective is emphasized Here are some ex-amples of noun phrases The head noun is underlined
Tok Pisintupela liklik meritwo small girllsquotwo little girlsrsquo
228 CHAPTER 10
ol gutpela pikininiPL good childlsquothe good childrenrsquo
dispela tripela bikpela popo tasolthis three big pawpaw onlylsquojust these three big pawpawsrsquo
Pijintufala boe nomoatwo boy onlylsquoonly two boysrsquo
wanfala pua womanone poor womanlsquoa poor womanrsquo
Bislamatu big haos yatwo big house thislsquothese two big housesrsquo
tu big-fala haos yatwo big-EMPHATIC house thislsquothese two particularly big housesrsquo
wan smol blu trak nomoone small blue car onlylsquojust a small blue carrsquo
There is only a small number of prepositions (but not so small a numberas some writers would have us believe) The following are the commonestprepositions in Bislama
Bislamalong location direction source instrument timeblong possession purpose beneficiaryolsem lsquolike aswetem accompaniment instrumentfrom cause
Examples
BislamaMi kam long Vila from wan kosI come to Vila becauseof one courselsquoI came to Vila for a coursersquo
Pidgins Creoles and Koines 229
Papa blong yu i stap wok wetem huiafather POSS you PREDICATE CONTINUOUS work with wholsquoWho is your father working withrsquo
Fis olsem hemia yu mas kat-em hem longfish like thisone you must cut-TRANS it withsap-fala naefsharp-EMPHATIC knifelsquoFor a fish like this one you have to cut it with a sharp knifersquo
The other two dialects are slightly different From does not occur inTok Pisin which uses the compound form bilong wanem lsquofor whatrsquo to markcause instead Pijin has the same prepositions as Bislama plus fo which isused to indicate purpose or tendency
PijinMifala laek fo goweEXC want for golsquoWe want to gorsquo
Hem i man fo dringhe PREDICATE man for drinklsquoHe is a drunkardrsquo
There are no special possessive pronouns in Melanesian Pidgin Thepossessive preposition (Tok Pisin bilong Pijin and Bislama blong) may be fol-lowed by either a noun or a pronoun possessor
Pijinnem blong yu belo blong siosname POSS you bell POSS churchlsquoyour namersquo lsquothe church bellrsquo
Verbs and the Verb Complex
Verbs are morphologically quite simple in Melanesian Pidgin The only com-mon affix is the transitive suffix
Tok PisinEm i rit i staphe PREDICATE read PREDICATE belsquoHeshe is readingrsquo
Em i rit-im dispela buk i staphe PREDICATE read-TRANS this book PREDICATE belsquoHeshe is reading this bookrsquo
230 CHAPTER 10
With certain verbs while the transitive form takes the suffix the intran-sitive form is often reduplicated
Tok PisinMama i was-im ol pikininimother PREDICATE wash-TRANS PL childlsquoMom washed the childrenrsquo
Ol pikinini i was-wasPL child PREDICATE INTRANSITIVE-washlsquoThe children washedswamrsquo
The verb complex does however contain a number of particles markingtense-aspect and other functions Verbs (and nonverbal predicates) take apreverbal particle i which marks what follows as a predicate This use of ican be seen in almost every example above7 In recent years however theuse of this predicate marker has become more and more optional especiallyin Tok Pisin and Pijin Thus the two Tok Pisin sentences above are just as of-ten heard as Mama wasim ol pikinini and Ol pikinini waswas
The verb is very often unmarked for tense and lack of marking canindicate either present or past Other tenses and aspects are marked byparticles some preverbal others postverbal
Tok PisinPREVERBAL POSTVERBALbai future pinis completedbin incomplete past i stap continuousken optative potentialinap abilitylaik intentionsave habitual
Three of these particles are or derive from verbs laik from laikim lsquolikewantrsquo save which as a verb means lsquoknow know how torsquo and i stap which asa verb means lsquoto be (in a place)rsquo Some examples of these tense-aspect par-ticles (plus the negative preverbal particle no) follow
Tok PisinEm i no save kaikai mithe PREDICATE not HABITUAL eat meatlsquoHeshe doesnrsquot eat meatrsquo
Yu bai wok-im pinisyou FUTURE do-TRANS COMPLETIVElsquoYou will have done itrsquo
Pidgins Creoles and Koines 231
Pita i inap karim ol kago bilong yuPeter PREDICATE able carryTRANS PL cargo POSS youlsquoPeter can carry your thingsrsquo
Mamok i no bin kamMamok PREDICATE not PAST comelsquoMamok didnrsquot comersquo
Reduplication of the verb for other purposes than to indicate intransitiv-ity is relatively common especially in Bislama There reduplication can havethe following functions reciprocal action random action repeated actionplurality intensity and the distributive
BislamaLeg blong hem i solapleg POSS he PREDICATE swelllsquoHisher leg is swollenrsquo
Leg blong hem i sol-solapleg POSS he PREDICATE INTENSITY-swelllsquoHisher leg is really swollenrsquo
Ol lif oli foldaonPL leaf PLPREDICATE falllsquoThe leaves fell downrsquo
Ol lif oli fol-foldaon long harikenPL leaf PL PREDICATE RANDOM-fall in cyclonelsquoThe leaves fell all over the place in the cyclonersquo
103 The Pidgins of the Motu TradersThe Motu people who live around Port Moresby speak an Oceanic lan-guage The western Motu particularly ldquoat the time of European contact(and for an unknown number of years before) hellip were involved in a complexnetwork of trading relationships with linguistically related and unrelatedgroups east west and inland of their present position The most spectac-ular and important part of this trade hellip was the hiri or annual tradingvoyage to the Gulf of Papua some 300 kilometres away to the westrdquo (Dut-ton 1985 20)
In the course of the hiri expedition two separate (and apparently un-named) pidgins developed One was based mainly on the Koriki language ofthe western Gulf of Papua the other on the Eleman languages of the east-ern part of the Gulf Dutton (1985) calls these the Hiri Trading Language
232 CHAPTER 10
(Koriki variety) and the Hiri Trading Language (Eleman variety) respec-tively
That however is by no means the end of the story The Motu also used apidginized version of their own language (Dutton calls this Simplified Motu)with other foreignersmdashoriginally probably in trade with their Oceanic-speaking neighbors and later with newcomers to the area After Europeancontact in the late nineteenth century they also used a variety of MelanesianPidgin with early colonial officials and other outsiders
The two Hiri Trading Languages were restricted to use on the hiri andwhen that trading expedition finally ceased toward the middle of the twen-tieth century the languages also died a natural death The English-basedpidgin died a less natural death it was proscribed by the British govern-ment which adopted instead the pidginized version of Motu as the languageof contact
The first British police force in Papua consisted of Fijians Solomon Is-landers and Kiwais from the Daru area of western Papua By the time thepolice force was being established there were a number of other foreignersof various origins settling in the Port Moresby area Simplified Motu soonbecame the lingua franca of this motley collection of people It was spreadoutside Port Moresby mainly by the police on their patrols along the coastand into the interior but also by released prisoners who were given posi-tions of authority as village constables The language acquired the namePolice Motu but in the 1970s as the connotations of the word ldquopolicerdquo weredeemed pejorative the name Hiri Motu was chosenmdashin the mistaken beliefthat Police Motu was a continuation of the language(s) spoken on the hiri
The differences between the Hiri Trading Languages and Hiri (or Po-lice) Motu can be seen in the following sentences (from Dutton 198533ndash34)
Hiri Trading Language (Koriki Variety)Enane pu miai anea Na okuaigo sago get come me givelsquoGo and bring some sagorsquo lsquoGive it to mersquo
Hiri Trading Language (Eleman Variety)Abuari pai avaia abusi Ara porohalaiago sago get come me givelsquoGo and bring some sagorsquo lsquoGive it to mersquo
Hiri (Police) MotuOi lao rabia oi mailaia Lau oi heniayou go sago you bring me you givelsquoGo and bring some sagorsquo lsquoGive it to mersquo
Pidgins Creoles and Koines 233
Two features give an idea of the simplified nature of Hiri Motu incomparison with Motu itself First Motu has the normal Oceanic contrastbetween direct and indirect possessive constructions and in indirect pos-session contrasts food (marked with a-) and other possessions (markedwith e-)
Motu(lau) tama-gu(I) father-mylsquomy fatherrsquo
(lau) a-gu aniani(I) POSSFOOD-my foodlsquomy foodrsquo
(lau) e-gu ruma(I) POSSGENERAL-my houselsquomy housersquo
Hiri Motu simply uses the general possessive form for all nouns lauegutamana8 lsquomy fatherrsquo lauegu aniani lsquomy foodrsquo lauegu ruma lsquomy housersquo
Second Motu has independent pronouns as well as subject prefixes andobject suffixes to verbs Hiri Motu uses free pronouns in all of these environ-ments
Motu Hiri MotuIndependent Subject Object All environments
SingularI lau na- -gu lauyou oi o- -mu oihesheit ia e- -(i)a ia
Pluralwe INC ita ta- -da itawe EXC ai a- -mai aiyou umui o- -mui umuithey idia e- -dia idia
Hiri Motu was once widely spoken in Papua (although virtually not atall in New Guinea the northern half of Papua New Guinea) and it is oneof the three official languages of Papua New Guinea (alongside English andTok Pisin) In recent years however the number of its speakers and conse-quently its status have tended to decline partly as a result of inroads intoPapua by Tok Pisin and partly because people who have been educated tendto use English in preference to Hiri Motu
234 CHAPTER 10
104 Fiji HindiBeginning in the late nineteenth century more than sixty thousand in-
dentured laborers were recruited from India to work on plantations in FijiInitially these recruits came mainly from northern India where languagesof the Indo-European family are spoken Many laborers spoke various di-alects of Hindi but many also spoke what was probably a pidgin knownas Bazaar Hindustani By the early part of the twentieth century howeveralmost half the laborers were being recruited from South India These work-ers spoke the quite unrelated languages of the Dravidian family The plan-tation environment brought into contact Hindi speakers from different di-alects (numerically the largest group of Indians) speakers of Hindi andother Indian languages (related and unrelated) speakers of Indian lan-guages Fijian and English and finally Indians and some of the twenty-seven thousand Pacific Islanders who were also recruited to work on Fijianplantations
A number of languages developed or were used on the Fiji plantationsan early variety of Melanesian Pidgin a pidginized variety of Fijian and apidginized variety of Hindi among them The first of these has died out inFiji but the other two are still used to some extent between people of differ-ent ethnic groups who have no other common language
In addition however another language also developed among Indians inFiji Through koineization Fiji Hindi or Fiji Bāt (= lsquolanguagersquo) evolved espe-cially among ethnic Indians born in Fiji Fiji Hindi incorporates elements froma number of Hindi dialects Some examples of the mixture of dialects involvedin the development of Fiji Hindi can be seen in the pronoun system
Fiji Hindi Hindi dialect sourceI ham Bhojpuri hamyou (intimate) tum Awadhi Braj tumyou (formal) āp Awadhi āp(u)hesheit (near) ī Bhojpuri Awadhi īhesheit (remote) ū Awadhi Bhojpuri ū
The formation of plural pronouns by the addition of log lsquopeoplersquo to the singu-lar as in ham log lsquowersquo is characteristic of Magahi
There were other contributors to Fiji Hindi as well Bazaar Hindustanithe Pidgin Hindi spoken on the plantations English (as one might expect)and also Fijian Some examples from Fijian follow
Fiji Hindi Fijian sourcedakāu lsquoreefrsquo cakaukūmāla lsquosweet potatorsquo kumalanangonā lsquokavarsquo na yaqona
Pidgins Creoles and Koines 235
tabāle lsquowifersquos brotherrsquo tavale lsquobrother-in-lawrsquotāmabūā lsquowhalersquos toothrsquo tabua
The result of this koineization process is a new form of Hindi different fromany spoken in India
A final complicating factor in the Hindi situation in Fiji has been thefact that although Fiji Hindi is the first language of virtually all Fiji In-dians who speak it in informal contexts it is not the language of formalsituations Standard (Indian) Hindi is used in schools on radio in printand in other formal contexts A situation of diglossia has developed inwhich people use one variety (Standard Hindi) in public meetings for reli-gious occasions and in other formal situations and the other variety (FijiHindi) in informal situations
236 CHAPTER 10
CHAPTER
11
Language Society and Cul-ture in the Pacific Context
111 The Vocabulary of Pacific LanguagesWesterners often evaluate people and their societies on the basis of theirtechnology People without advanced technology are considered primitivenot just technologically but intellectually as well
Linguists studying Australian and Pacific languages are often asked howmany words there are in those languages Underlying such a question is theassumption that such ldquoprimitiverdquo people must speak simple languages ldquoByand large the white population of present-day Australia has little knowledgeof the structure or nature of Aboriginal Australian languages Moreoverthey have serious misconceptions about them If you strike up a conversa-tion with even well-educated white Australians you may hear that hellip lsquo[Abo-riginal languages] have only a few score wordsmdashnames for common ob-jectsrsquordquo (Dixon 1980 4)
Nothing could as we have seen be further from the truth The gram-mars of Pacific languages are by no means simple or primitive How doPacific languages stand in terms of lexicon
1111 How Many WordsWhen linguists are asked how many words there are in a particular languagethe idea seems to be that the more words a language has the more sophisti-cated it is By implication Pacific languages probably have many fewer wordsthan English does and so are less sophisticated or more primitive
Even trying to compare dictionaries of two languages for this purpose isfraught with difficulties
237
How do we measure the number of words in a language First what isa word For instance should the compound lsquofirehosersquo be treated as asingle word different from lsquofirersquo and lsquohosersquo Languages differ widely asto what is considered a word Second are we talking about all wordsever used by any speakers of that language Or about all words usedcurrently Or about all the words used by an individual speaker andpresumably stored somehow in that speakerrsquos mind Or about all thewords ever recorded of the language These questions show how hardit is to compare languages with respect to the number of words inthem (Simpson 1993 123)
The other aspect of this problem relates to what forms the basis of ourcomparison The simplest way to compare languages in this way is to countthe entries in a dictionary (ignoring for the moment all the other problemsSimpson has pointed out) But some languages (like English) have a muchlonger and more intensive history of dictionary compilation than do others(like those of the Pacific) So even if this were a valid way of making com-parisons it would not be a particularly productive one
Let us pursue this a little further Crystal (1987 108) notes that the 1987edition of the Random House Dictionary of the English Language for ex-ample contains some 260000 headwords (ldquothe bold-face items that occurat the beginning of each entryrdquo) No dictionary of a Pacific language comesanywhere near that figure The monolingual Fijian dictionary currently inpreparation has considerably fewer than twenty thousand entries (Paul Ger-aghty personal communication) and this will probably be one of the largestPacific dictionaries when it is published Does this mean that Pacific lan-guage vocabularies are considerably more limited than those of Europeanlanguages
In a sheer numerical sense of course it does But we need to look a littlemore deeply into this question Let us do a quick experiment Below are thirtyconsecutive words beginning with the letter q as listed on pages 1415ndash1416of the 1981 edition of the Macquarie Dictionary How many of these do youthink are in common use How many could you give the meaning of1
quincuncial quinoline quinquepartitequincunx quinone quinqueremequindecagon quinonoid quinquevalentquindecennial quinoxaline quinsyquinella quinquagenerian quintquinic acid Quinquagesima quintainquinidine quinque- quintalquinine quinquefoliate quintan
238 CHAPTER 11
quinoid quinquennial quintequinol quinquennium quintessence
Dictionaries of languages like English include a vast number of highly tech-nical obsolete or obscure words (like many of those listed above) nearly allof which are not known to the ordinary speaker of the language
The average speaker of any language probably knows and uses some-where between five and ten thousand words in everyday life and mayvaguely recognize a few thousand more2 People in a particular professionor people who have a particular hobby or interest will have another set ofvocabulary related to that profession or interest but other speakers of thelanguage may probably not know those wordsmdashor at least not know how touse them accurately Most of the 260000 words in the Random House dictio-nary for example are probably unknown to almost all speakers of English
When we take all of this into account Pacific languages are not so differ-ent from English as might at first be assumed The average speaker of aPacific language also probably knows and uses between five and ten thou-sand words People who specialize in fishing weaving or other professionsand crafts will of course know additional technical terms not familiar toother people The difference with European languages lies in the fact thatmost Australian and Pacific societies are relatively small-scale ones Therange of specialization of professions crafts and hobbies is much smallerand hence the size of the technical or specialized vocabulary in those lan-guages tends to be much smaller as well
1112 Specialization Classification and AbstractionA second common misconception about Australian and Pacific languagesconcerns the degrees of abstraction they are capable of I quote Dixon(1980 5) again ldquoSome missionaries and amateur linguists who attempted tostudy an Australian language have contributed to the misconceptions [aboutthem] They have put it about that although there may be a superfluity ofterms for particular objects the languages are totally lacking in genericterms such as lsquofishrsquo or lsquofowlrsquo this is of course taken as a lack of mentalsophistication The fact is that Australian languages do have quite as manygeneric terms as European languagesrdquo
Specific Terms
Things of cultural importance are usually obvious from a languagersquos lexiconThe motor vehicle for example is of vital importance to western society andin English we have a large number of words referring to kinds of motor ve-hicles (car truck lorry van bus) to brands of vehicles (Toyota Cadillac
Language Society and Culture in the Pacific Context 239
Mercedes-Benz Rolls-Royce) to models (Toyota Corolla Toyota CressidaToyota Camry) as well as many hundreds of words referring to componentsof motor vehicles In many Pacific languages in contrast there is often justone word meaning lsquomotor vehicle (of any kind)rsquo
But in English we have just the one word yam to describe a particularroot-crop3 and we have had to borrow the word taro to describe anotherroot-crop Because these are not important food crops in English-speakingsociety one word for each is probably adequate In the Anejom language ofVanuatu where these are important sources of food there are accordingto my count so far names for forty different varieties of yam and words forover sixty named varieties of taro
A languagersquos vocabulary reflects what is important to its speakersWhere fine degrees of specialization are necessary or desirable they willshow up in the lexicon Stages of growth are one common area of specializa-tion in these languages The coconut of course is perhaps the prototypicalPacific plant and one should not be surprised to find lexical specialization inthis area In Kwamera for example the generic term for coconut is napueithe fruit itself goes through the following named stages of development
Kwameraiapwas small coconut coconut fruit budkwanapuirahaacutekw larger coconut fruit budkwatigɨs small coconut (about four inches in diameter)kapkapeki (intermediate stage between kwatig ɨs and tafa)tafa young coconut before meat has begun to formnafweruk nut with soft meat and effervescent waterkahimaregi nut with hard well-developed meatnapuei mhia ripe nut with developed flesh which falls from
the treekwarumahaacutekw fallen nut which has begun to sproutnuvera sprouted nut
In addition to these terms there are terms for different parts of the nut ofthe tree and of the fronds as well as for different varieties of coconuts anddifferent coconut products
The existence of areas of lexical specialization like this is not surprisingWe would expect coastal people to have numerous words for different kindsof fish and fishing equipment horticulturalists to have specialist terms forplants and their parts and growth stages warriors to have detailed vocabu-lary relating to weapons and so on But there are also a few less expectedareas of lexical specialization like the following set of names for differentkinds of noises in Yidiny
240 CHAPTER 11