Language research in Papua New Guinea: A case study of Awar.

Levy, Catherine

Abstract

Papua New Guinea features today as one of the nations with the highest linguistic diversity on Earth, with 820 languages for just over five million people. Each language represents a cultural identity and a vision of the world; all are of oral tradition, and quite a few of them are threatened with disappearance. A linguist's passion is to write down and describe these languages for cultural conservation and literacy purposes. In this article, after a brief overview of language groups in PNG in general, the author shares her experience of linguistic research in the Awar community in the northern part of Madang province. By using illustrations from Awar, she explores the importance of language in forming one's worldview as well as the goals, methods and tools of the linguist.

The diversity of languages in PNG

Papua New Guinea has three official national languages: English, Tok Pisin (PNG pidgin) and Motu(1). But traditionally, Papua New Guineans speak 820 languages belonging to two language groups: the Austronesians and the Papuans

The Austronesians are part of a language group originating from South China, spreading from the Philippines, Indonesia, and as far west as Madagascar. In New Guinea they are mostly encountered along the coasts and on the adjacent islands, then further east across the whole South Pacific, all the way to Easter Island, then north to Hawaii and south to New Zealand. They represent about 800 languages today, all relating to a common proto-, that is, ancestral, language.

The other group, the Papuans languages, are spoken on mainland New Guinea and stretch west to Halmahera, and as far east as the Solomons. These languages are not related and thus appear in the actual state of research as belonging to about 60 different families. That is why we prefer to call them 'Non-Austronesians', to emphasize that they do not all belong to one group.

With respect to these language groups, it is important to note two concepts: the difference between a language and a dialect on the one hand, and the number of speakers per language on the other hand.

In the case of Papua New Guinea, the number 820 refers to languages, not dialects. A dialect is a variant of a standard language. Two dialects of one language may be defined as 'mutually intelligible' variants of that language; two people speaking two dialects of the same language would understand each other. The difference between language and dialect may be established on the basis of a comparative vocabulary list: 90% common vocabulary or more between two vocabulary lists indicates that we are dealing with two dialects of the same language; 80 % or less common vocabulary indicates that we are dealing with two separate languages, that might be related, that might belong to the same family and spring from a common 'mother-language', but are nevertheless now distinct from each other. Between 80% and 90% common vocabulary, the assessment depends on the linguist. Linguistics is not an exact science.

Take for instance the language studied in this paper: the Awar language. It is part of the Lower Ramu family, composed of five languages spoken in Boroi, Bosmun, Kaian and Watam, all located in the northern part of the Madang Province. The Awar language is spoken by a total of 1100 speakers in three villages and features two dialectal forms. The version spoken in the Awar village is presented as the 'reference'; it gives its name to the language. The other villages are referred to as speaking variants of this 'main' language.

The 820 languages in PNG are, thus, languages, not dialects. They are grouped into families with ancestors as different as those of French and Chinese, for instance, and this sometimes only a few kilometres apart.

After the assurance that we are dealing here with languages, and not dialects, another feature particular to PNG languages is the low number of speakers per language. The total number of languages spoken in the world today is approximately 6000. Throughout the world, 90% of the world population speaks only 10% of these living languages. In this category we find widely spoken languages such as Mandarin, English, and Spanish.

For the 90% of remaining languages, the number of speakers per language is very low. The languages of PNG are among those with a small number of speakers. Take, for example, the Madang Province, which has about 120 to 150 languages for a population of around 400,000. These figures indicated that each language may be spoken by an average of less than 3000 people.

There are a number of commonly accepted reasons for this linguistic diversity. First there is the dynamic of the time span. PNG has been inhabited for 50,000 years. If one takes the very conservative hypothesis that the migrants came with but one original language, and the other conservative hypothesis that one language split into two languages every 1000 years, after 50,000 years one still arrives at figures largely above 820 languages.

The terrain of PNG is another factor. Valleys, mountains and swamps are common features of the New Guinea landscape. The resulting forced isolation has favoured the splitting of original languages into a number of related but distinct languages.

The role of language as an identity and cultural manifesto is another reason for such linguistic diversity. Language defines the speaker; it gives them an identity and the assurance of belonging inside a cultural group. This characteristic also insures the protection and survival of the language.

Finally, the environment also shapes the culture that, in turn, influences a language. A harsh environment might favour contact between different groups of people to insure their survival, as in Australia and some parts of Africa. In areas where climatic and environment conditions are more benevolent, as in PNG, the reliance on the neighbour is not as vital as the survival of the closed group on a defined piece of land. Thus, together with more antagonistic tendencies between groups, such areas have been observed to present a greater cultural and linguistic variety in PNG.

The significance of language: words to see the world

A language is a series of conventions set in place for the purpose of communication. As a convention, different people have decided that this bright ball in the sky that gives us light during daytime is called 'sun' in English, 'soleil' in French', 'Sonne' in German, and 'rao' in Awar. All speakers of one language agree on this, and thousands of other conventions.

The study of non-European languages for a European is particularly enlightening for it shows what view of the world is contained in a language and, in turn, how the acquisition of a language, especially a mother tongue or first language, influences the way the speaker views the world. Indeed, what the speaker sees! To illustrate this point, I will take a series of examples from the Awar language: the perception of colours, the expression of comparison, the local designations for social relations, conflicts and spiritual manifestations.

My first example concerns colours. Speakers of European languages may take it for granted that the whole world sees blood as red, the sky as blue, an orange as orange and leaves as green, just as, for example, the English, Italians, French and Germans see them.

Nevertheless, a large number of languages present only three words to differentiate colours: red, light and dark. Red seems to be universal, maybe because of the colour of blood. Light and dark may also be translated as white and black. Many languages do not have words for other colours.

Awar recognises yellow as well as red, white and black. It also recognises blue, and acknowledges that the sky and the sea are blue. But when it comes to green, when it wants to describe all the bush and gardens that surround the village, Awar uses an expression: 'blue like a tree leaf'. In other words, green is perceived as a variant of blue. Awar, a village literally bathing in what a European speaker would see as green, does not see green as green, but as blue.

The expression of comparison is another linguistic example that throws light on the way Awar people see their world. In Awar, you cannot say that 'one is better', 'one is bigger', or 'one is smaller'. At best, you will say 'one is big, one is small', 'one is good, one is not good'. It is perhaps a small thing, yet something very distant from the Western spirit of competition.

Other examples give insight into social relationships and conflict. There is a word, for instance, to designate an in-law within Awar speakers, and another word to designate in-laws within the Sepik-Ramu area, but there is no word to designate an in-law from anywhere else, because one did not traditionally marry outside one's language group.

The same expression 'mot vor', by way of another example, designates 'an old man' and 'a big man', 'an old woman' and 'a big woman' (referring to 'social weight'). An old man is automatically 'a big man', worthy of respect. Finally, there is no word in Awar to designate 'an orphan'. A child may lose her or his biological parents, but they will always have a father and a mother.

Language helps illustrate the local view on conflict as well. The Awar traditionally had rather antagonistic relations with a number of neighbours, not linguistically related, with whom they used to enter into periodic warfare. Despite this fact, the language does not have any translation for the European concepts of 'winning' or ' losing' a battle. You could lose your knife, but not a battle. You did not win a battle either.

Contact with the Western system, however, has prompted the Awars to coin a new word to express this concept. In Tok Pisin, the contact language, 'to win' and 'wind' are both translated by 'win'. Thus the Awars took their word for 'wind'--'mbombe'--and coined the expression 'mbombe mongre', 'to make wind' which has now become the Awar expression for 'to win'.

Language gives one a glimpse into the spiritual world of the Awars as well. Spirits are part of everyday life in Awar. Rivers, reefs and caves, to name only a few, have the names of the spirits--indeed they are the spirits--that alternatively protect or endanger those who come close to them. The sacred bamboo flute, used for secret ceremonies, has the name of the spirit, lets out the voice of the spirit, indeed is the physical representation of the spirit.

Finally, there is no Awar word to say 'thank you'. In a system of sharing and reciprocating, 'thank you' would be redundant; the one who gives is simply acting according to an expected social pattern. 'Thank you' may be necessary in a Western system to acknowledge the giver's choice, as the Western system is not one of automatic sharing and giving. But such is not the case in Awar. The closest approximation in Awar to 'thank you' is 'yandi', that is, 'it is good'.

All such examples, reflecting idiosyncrasies of 'conventions for communication', gives one some insight into the way the world is perceived by its speakers, and the way the language selects what the speaker will see. Clearly, if the language does not suggest that you compare, for instance, or see bigger or smaller things, or recognize the colour green, you probably will not perceive the world in these ways. If the language does not suggest that you win or lose a battle, your relationship to your opponent will be different than if your language allows you to do so. And if the river is the spirit, you will tend to respect it more than you might a mere body of water.

In short, as illustrated with these examples from Awar, each of the 820 languages in Papua New Guinea simultaneously reflects, creates and partakes in a specific cultural environment. The people, with their languages and cultures, have not only managed to survive and adapt for the last 50,000 years, they have also been successful with their languages. The speakers have all the words and concepts necessary to feel completely at home in their physical and social surroundings and contribute to the self-confidence and self-reliance that has shaped traditional New Guinea culture.

Today, however, change is taking place at a rate that endangers New Guinea cultures and languages. A language disappears, quite simply, when there is nobody to speak it anymore. Speakers may choose to adopt a language that will be more directly - even superficially - beneficial, or that is culturally dominant. To speak English or Tok Pisin, for example, in PNG today, is closely associated with success in a new, promising, material culture.

On another level, a foreign language might be not chosen, but rather forced upon a community, as, for example, a group of landowners who feels helpless when facing legal documents in English that are presented to them to sign. The landowners may have a working knowledge of English, or they may have none. The official jargon - and social pressure - is such that they find themselves in a very dependent situation. Little provision seems to be made on a legal level to make sure that the landowners understand what they sign, what they commit themselves, their families and their ground to. This is particularly striking when one considers how foreign the idea of giving up their ground is to the Melanesian understanding of the world.

A borrowed language as a common language, be it English, Tok Pisin or Motu, will then act as a 'lowest common denominator', suitable for all, for basic exchange, but leaving out identity and originality, because it cannot express the riches and the sophistication of the independent culture in its own environment. Worse yet, it will express another view of the world, of the people, altogether, abandoning the old ways to turn to new ideas.

Most Papua New Guineans are raised in a multilingual environment: the local language with their parents, Tok Pisin in the village or town, and English at school. Their multilingualism must certainly be an object of envy - not all Westerners can boast as good a level of a third language as New Guineans can with their spoken English. Still it remains that evolving within three languages is a complex affair as it implies dealing with three sets of vocabularies and concepts, referring to three different social contexts.

Furthermore, it is widely accepted today that literacy programs are most fruitful when the students, whether children or adults, are first taught in their mother language, when they learn about the concepts of reading and writing in the language they are most familiar with. Only later should they move on to another language--English in the case of Papua New Guinea.

Literacy in a local language also helps prevent the creation of a gap between the school and the home, allowing the parents to remain involved in their children's education, and allowing the child to move smoothly between both environments.

In Papua New Guinea today, where traditional languages and cultures are weakened and threatened by 'foreign' influences, the official language becomes a social divider. Feelings of self-confidence and self-reliance are replaced by helplessness and frustration, often turned towards other members of the groups, such as the women! And who would not feel helpless, and be angry, when their interests and those of their clan are at stake and yet one cannot converse with the other party, and one's traditional language is not considered, or perhaps is looked down upon as not suitable?

Papua New Guinea is a land in transition, and there is no question as to whether or not it should be part of globalisation. It will be part of it, and the mastery of English, the language of globalisation, is a requirement. But with quick, brutal transition looms the danger of becoming a 'me-too' society, of adopting the 'external signs' of the new culture, rather than integrating it.

The other way to partake in the global change would be with the self-confidence and the self-respect of those who have roots and precise words to see their world. And one of the ways to do this is to support literacy and cultural conservation in the local language.

My own history with Awar

I came to PNG in 1986, as part of an ancillary team for a small research station set up by the University of Brussels, Belgium, in Hansa Bay, Madang Province. My connection to Papua New Guinea and its people may be called a love story from the start. I was instantly fascinated by the wealth of culture and information made available to me, by the readiness of the people to share their culture, and especially by what I saw particularly of the Awar culture, as that was the neighbouring village.

During the first years, I learned to speak Tok Pisin, and explored the environment. Having always had an interest in languages, I vaguely tried to learn Awar as a hobby. I could not help but notice that the children were not answering in the Awar language when spoken to by their parents, but preferred to use Tok Pisin. One day early in 1992, I met a white man on an almost deserted beach. Bill Foley, a reference in South Pacific linguistics, head of the Department of Linguistics at the University of Sydney, had just arrived from Watam, at the mouth of the Sepik, and was planning on spending a few days to do fieldwork in Awar for comparative purposes. We struck up a conversation and a couple of hours later I was determined to walk in his footsteps: that man was doing exactly what I had always wanted to do. He supported my - crazy wish to start working on the Awar language, but suggested, as I was going back to Belgium the following month, that I contact the university there and take some courses as formal training.

Back in Belgium, I learned that there was no such thing as a department of South Pacific linguistics in my home country, and that the nearest equivalent to it was the department of African linguistics, also dealing with languages of oral tradition. By a second stroke of luck, I got in touch with the head of that department, Madame Claire Gregoire. She did not laugh at my story of falling in love with a language spoken by 1000 people 17,000 km away. She smiled and told me that it was exactly her story: she had gone to Africa as a French teacher and, to make a long story short, became a specialist in African linguistics.

And so it was that I began to attend part-time classes in the department for two years, taking a number of courses that had to do with techniques to describe a language of oral tradition. In 1994 and 1995, I came back to PNG and did my first fieldwork. I would spend 6 months in Hansa Bay, then return to Belgium and work on my description, all the while working full time. Then in 1996, I moved permanently to PNG and started lecturing at Divine Word University shortly thereafter. While plodding along with linguistics, I very quickly became aware of how much I did not know about PNG culture and, in that respect, it was another stroke of luck to have the opportunity to lecture in the PNG Studies department at DWU.

My primary objective in Awar had been to put the language into writing, and use the results for literacy purposes. As time went by, however, and my research developed, I realized that putting a language into writing involves safeguarding the cultural memory of the community as well. There are a number of steps to this process and it is to them that I now turn.

The work of a linguist: phonology, morphology and syntax

The three successive parts of the description of a language are the phonology, the study of the sounds; the morphology, the study of the words; and the syntax, looking at the word order.

Phonology, the first step of linguistic description, has to do with the sound system that is specific to each language. Compare, for instance, the English sound 'th', as in 'that', or the German sound 'ch', as in 'ich', which are absent from Tok Pisin (PNG Pidgin. The linguist will look for the sounds that are relevant, those that convey a meaning through the use of 'minimal pairs'. For example, in English the words 'fair' and 'lair' are minimal pairs, composed of exactly the same sounds, except for the initial sound. From this specific minimal pair, we can establish that the sounds 'f' and 'l' in English are phonemes, that is, sounds that are relevant because they convey a difference in meaning. An example of a minimal pair in Tok Pisin would be: 'laus' (in English, 'louse') and 'haus', (in English 'house'), where 'l' and 'h' are phonemes, because they allow one to understand the difference between 'lous' and 'haus'.

The linguist, then, goes through a vast number of minimum pairs, to identify all the relevant sounds in the system. The phonemes are transcribed - written down using the International Phonetic Alphabet (IPA), a table that allocates a sign for every sound made by the human vocal system. Moreover, this table indicates, vertically, where the sound is expressed in the mouth, whether from the lips, labial (for example, 'p' or 'b'), or from the back of the throat (glottal). Horizontally, on the other hand, it indicates whether the passage of air is free fricative (like 'v', 's'), or obstructed - plosives (like 'p' or 'b'). Finally, the sound is identified as being voiced, using vocal chords (like 'b' or 'r'); unvoiced, not using vocal chords (like 'p' or 'k'); or nasal, when going through the nose, (like 'm' or 'n'). Thus, for instance, the phone 'p' will be described as unvoiced bilabial plosive; the phone 'n' will be described as voiced dental nasal.

Once the phonological system of the language has been established - the system of all relevant sounds, vowels and consonants - the alphabet may be finalized. The signs of the IPA are sometimes quite difficult to memorize, and do not all appear on the typewriter keyboard. The next step, then, is to determine conventions, to agree on the signs that will represent every phoneme in that specific language; one sign will represent one phoneme. In English, for instance, the sound 'tch', as in 'chicken' is represented by 'ch'. In German, that same convention 'ch' will represent the velar sound as in 'ich', which is, in turn, represented by 'j' in Spanish in such words as 'mujer'.

The step of choosing the signs for the alphabet is very important for the community who must be empowered with this decision. In the end, the survival of the language will be ensured by the use that the community makes of the alphabet and of the written language. In Awar, for instance, there have been lengthy hesitations and discussions about the writing of the glottal; ultimately, the letter 'h' was adopted. Another discussion took place for the velar fricative, the sound that is spelled 'ch' in German and 'j' in Spanish. A spelling 'ch' appeared disturbing because of its proximity to the English 'ch'. Finally 'x' was chosen, as it was otherwise not used in the language.

Once the work of phonology is complete, the following step for the linguist is to examine how words are formed in the language. This part of the analysis is called morphology. The morpheme is the smallest significant unit in the language that conveys a meaning. When a morpheme is attached to the noun or the verb, as in the examples below, it is called an affix: a prefix when it is attached before the word, a suffix when it is attached after the word.

Take for example the English word 'dogs' which can be analysed as 'dog' - 's'; the letter 's' is the morpheme of plural, the suffix indicating that we are dealing with more than one dog. In Awar, 'dogs' is 'kyaorih' where 'kyao' is 'dog' and 'rih' is the morpheme of plural. Or, take for another example the word 'washed' which can be analysed as 'wash'-'ed'; 'ed' is the morpheme that marks that the action is taking place in the past. In Awar, she 'washed' is translated by 'rwe-t' where 't' is the suffix of the past.

In the chapter on morphology, the linguist will also identify the different types of words present in the language. In English, for instance, one would encounter nouns, verbs, adjectives, adverbs, prepositions etc. The description of morphology would allow one to see how these words are formed as simple and compound nouns, singular, plural or even dual - a special mark for two elements; verbs and conjugation; and how one forms the present, past or future tense. Awar, for instance, features a dual for the nouns and the adjectives, and features postpositions, rather than prepositions as in English.

Subsequent to the study of phonology and morphology, the third step of linguistic description is the syntax, the description of the word order. Observed here is the order of the 'core constituents' of the sentence: the subject S, the object O, and the verb V. These three elements can permutate into six different sequences. All six possibilities are encountered in world languages, although a majority of the world languages features the SVO and SOV order.

English, for example, features the SVO order:

 I eat the apple
 S V O

Awar, like many Papuan languages, has an SOV order:

 Mo kan vasat
 You tree saw
 S O V

Further to the order of the core complements, syntax also examines the place occupied by other complements such as place, time or manner, as well as the way they are marked, for the speaker to recognize them. For instance, in English, the complement of direction or situation will be marked by a preposition such as 'in': I go to the village; or the dog jumped under the house. In Awar, direction is marked with a noun used as suffix: 'Mo tonik sangre', 'You house-inside go Present' or 'You go inside the house'.

The examples above are obviously not representative of the complexity of linguistic analysis, as one may judge by picking up any grammar of the English language. Every language features a level of subtlety and sophistication of expression that allows linguists to write hundreds of pages about them.

'By-products': the lexicon and the story book

Thus linguistic analysis produces a body of texts gathered from the study of the language of native speakers. One 'by-product' of this analysis then is a lexicon, that is, a list of all the words encountered during the analysis. The other will be a collection of stories such as traditional legends, or stories telling about skills such as sailing boat building and house building, pottery, garden clearing and others.

In the case of the Awar language, the lexicon counts over 1300 entries in Awar, English and Tok Pisin, including some grammatical notes, as well as the first book ever written in the Awar language, over 100 pages of traditional and recent stories.

All this material may be used for literacy purposes, for children and adults. But it also represents part of the community's collective memory. A striking example of this communal memory was the reaction of younger people that occurred during a proof-reading session. They were puzzled by vocabulary relating to boat building and realized only later, after referring to the elders who had dictated the story, that it was them, the youngest generation, who were not familiar with these terms that had left everyday conversation when the tradition was lost. The sailing boats were not used anymore, so the words relating to boat building were not used anymore either.

Languages change constantly. New words are coined, called neologisms. They may be borrowed, as English has borrowed the word 'hamac' from a South American language, and the word 'chair' from French. Awar also has borrowed words, for example, the concept 'mas' (in English 'must') from Tok Pisin to express the feeling of obligation. Languages can also enlarge the meaning of an existing word to express a new concept. In Awar, for instance, the word 'ngom', or 'corner', is now also used to refer to corned beef, packaged as it is in a square tin of 'all corners'.

Change and influence can also be visible in the word order. Take the example of complex nouns in English and Tok Pisin. English has a genitive-noun sequence for a complex noun: airport comes from air-port, and post office from post-office. Tok Pisin, a language with an Austronesian basis, on the other hand, features the noun-genitive order for complex nouns: For example, 'ples balus' is 'place plane', the place of the plane or airport; 'haus pos' is 'house post', the house of the post or post office; and 'tokples' is 'talk place' or the language of the village. However, a current advertisement in Tok Pisin for a famous fizzy drink displays the sign: 'Ples Coke!' not 'Coke Ples' as one might have expected. One must then ask, is this a linguistic or cultural influence?

Linguistic tools and methods

The methods for gathering linguistic information are simple. I was told from the beginning that being a linguist requires no material, apart from a good ear, a clear speaker, paper, pencil and an eraser.

There are a few provisos to this simple formula however. The speaker must be a native speaker and have no pronunciation faults. As for the tools, after a while, one is most thankful for the support provided by a good tape-recorder, a computer and email access. Finally, phonological analysis may require help from a sound laboratory.

The methods used to record the language are questionnaires and storytelling. Questionnaires are especially useful in the early stages, when one may need help to segment the enunciation given by the informant. Take for instance the utterance: 'mogurapnunihsangrene' which means, 'tomorrow you will go to the village'. How does one identify the different parts of speech, however, the various words that compose this sentence?

In the early stages of research, questionnaires are employed, using several commuting elements. For example, the linguist will ask the informant to translate such phrases as:

- The child sees the woman.

- The woman sees the child.

- The woman sees children.

- The woman saw children.

- The woman will see children.

And hundreds and thousands more.

By permutating one element at a time, the linguist can begin to identify individual elements. With the examples above, for instance, one would be able to identify the word order (examples 1 and 2), singular and the plural (examples 2 and 3), and the present, past and future tense (examples 3, 4 and 5).

Returning to the initially indecipherable utterance above, eventually, after much permutation, the linguist will discover that the sentence should read:

 Mo gurap nunih sangrene
 You tomorrow village go ... Future

As quickly as possible, however, the linguist will try to switch from questionnaires to storytelling techniques. These stories are recorded, thus giving the informant the freedom necessary for good storytelling. In other words, by allowing the speaker to tell the story at his or her own rhythm, without interruption, with complete spontaneity, one obtains material of the best possible quality for analysis. During storytelling, the researcher remains as neutral as possible, in order to keep any outside influence to a minimum. Besides this neutrality, there are a number of factors involving cultural sensitivity that play a part in the quality of the storytelling data gathered as well. A male informant, for example, may not be able to talk about the rituals surrounding pregnancy and childbirth, or may have equal problems telling a female researcher about male initiation. When it comes to ancient techniques, for another example, the vocabulary may be a problem if it has long been unused or the description of the technique itself may be vague, for not having been tried for over 60 years.

After recording - and having checked the quality of the recording - the researcher begins the next part of the work: the transcription of the text. With the help of an informant fluent in the language and in a contact language - Tok Pisin in my case - the researcher will set out to painstakingly write down the text, word for word, as dictated by the assistant. This part of the work is very tedious in the very early stages of the analysis, when the researcher seems to constantly encounter strange words and strange concepts requiring pause and explanation. In fact, at this juncture the language system can still 'go in any direction'; the researcher still does not know what the language conventions are. In the case of Awar, for instance, it took me ages to identify the meaning of movement verbs, all translated into Tok Pisin as 'I go'. After years of headaches, it finally dawned on me that Awar uses a different verb for 'to go' according to the direction one takes from one's place of origin, that is, the village. Moreover, as an added complication, the free storytelling language is often quite different from the sample gathered from questionnaires, which may lead to even more confusion and doubts.

In fact, this transition phase from questionnaire to storytelling and transcription is the part of the fieldwork that causes the most bouts of headaches and discouragement. My reference for this frustration is a quote from Margaret Mead, where she explains how she was repeating 'I will never make it ... I will never make it ...' until the day when the sentence popped into her mind - in the local language! Once over that hump, the researcher will feel positively exhilarated to be able to listen to a text and understand it almost fluently. Such exhilaration, of course, comes only after months and, more probably, years of effort!

Once a corpus of texts is constituted (several hundred pages of manuscript), the researcher may proceed with the analysis itself, reading and re-reading the texts until the meaning of every expression and the place of every word has become clear and predictable.

An ongoing process

My work started in 1992 with linguistic courses at the University of Brussels. Next came the fieldwork and analysis. I did not defend my thesis until December 2002. Since 2003, I have facilitated workshops in Awar to finalize the alphabet, proofread the first batch of texts and the lexicon, as well as gather more texts and work on the elementary school curriculum. And there is still work in progress. There is always more to write about a language; new English grammars are still being published today. In the case of Awar, we are presently working on more texts and an expansion of the lexicon. We also recorded a traditional celebration in January that will become the first Awar book illustrated with the pictures taken at the time of the sing-sing.

Future plans involve more channelling of the community effort, as a result of the momentum created by the distribution of the existing documents to every Awar household. We will then welcome - and record - such interventions as: 'This story is not told properly, the correct version is as follows', or 'That word is not in the dictionary yet; we must add it.'

Apart from the literacy work, another project involves investigating what appears to be traces of Austronesian influence on the Awar language such as loanwords and perhaps influence on a syntactic level. There appears to be evidence of Austronesian cultural influence as well. The Awars, originally from the Ramu and Sepik area, seem to have adopted Austronesian sailing boat building techniques from the Manam people, on the adjacent island. This comparative work is a project that will stretch over several years, as it implies gathering material from the languages of the Awar family - Bosmun, Boroi, Kaian, Watam - as well as analysing a Manam grammar and wordlist, and possibly even looking into anthropological studies.

One important limitation to the progress of the project was the apparent initial lack of concern of the Awar community. Although the people were extremely friendly and welcoming, and helpful when it came to sitting down and telling a story, I failed to encounter genuine interest until in 2003 the first sample of the lexicon and of the book was put into their hands. Then all of a sudden it seemed to dawn on them the realisation of the worth of the material I had been collecting, as if only this tangible proof could rekindle a self-confidence that weakened as the Awar traditions were losing strength.

An enormous number of languages across the world have histories similar to the one of the Awar language: they are spoken by a minority, they represent traditions that appear obsolete in our modern world, and they are threatened with disappearance, in the same way that species of plants and animals are threatened when their environment is undergoing momentous change. Cultural diversity is on the same agenda as bio-diversity, and linguists all over the world work with communities, language per precious language, for cultural conservation.

References

Anceaux, J. C., 1965, The Nimboran Language, Phonology and Morphology, 'S-Gravenhage, Martinus Nijhoff.

Bussmann, Hadumod, 1996 The Routledge Dictionary of Language and Linguistics, Reference, New York.

Foley, William, 1986, The Papuan Languages of New Guinea, Cambridge University Press, Cambridge.

---. 1997, Anthropological Linguistics, Blackwell, Oxford

---. 2000, 'The languages of the New Guinea region', Annual Review of Anthropology 29, pp. 357-404.

Hoeltkoer, Georg, 1947, Die Maritime Ortung bei einigen Staemmen in NordNeuGuinea. Encyclopedia Helvetica, pp. 192-205.

Katamba, Francis, 1989, Introduction to Phonology. Longman, London.

---. 1994, Morphology, McMillan, Houndmills.

Longacre, Robert, 1972, Hierarchy and universality of discourse constituents in New Guinea languages, Georgetown University Press, Washington.

Mead, Margaret, 1935, Sex and Temperament in Three Primitive Societies, Morrow, New York.

Nettle, Daniel, 1999, Linguistic Diversity, Oxford University Press, Oxford.

Parrish, David, 1988, Grammatical Description of Bore, Pioneer Bible Translators, Madang (unpublished).

Pawley, Andrew, 1987, 'Encoding events in Kalam and English: different logics for reporting experience,' in R. Tomlin (ed.), Coherence and Grounding in Discourse, Benjamins, Amsterdam, pp. 329-360.

Pawley, Andrew. 1993. 'A language which defies description by ordinary means,' in W. Foley (ed), The Role of Theory in Language Description, Mouton de Gruyter, Berlin, pp. 87-129.

Reesink, Ger P.,1987, Structures and their Functions in Usan, John Benjamins Publishing Company.

Wurm, Stephen, 1982, The Papuan Languages of Oceania, Narr, Tubingen.

Z'graggen, John, 1971, Classificatory and typological studies in languages of the Madang Distirict, Pacific Linguistics C19.

---. 1972. Comparative Wordlist of the Ramu Language Group, New Guinea, Anthropos Institut, Alexishafen, Madang Province, PNG.

(1) PNG Pidgin (Tok Pisin) and Motu are two languages traditionally used as Lingua Franca in the country. The first is part of a family of pidgins spread throughout the South Pacific (Vanuatu, Solomons, North Queensland); the second is part of the Austronesian language family and first developed in the area of Port Moresby, the capital of PNG.

Until recently, Catherine Levy lectured in the Faculty of Arts, PNG Studies Department, at Divine Word University. She has a doctorate in linguistics from the University of Brussels, Belgium, and her thesis on the Awar language is presently being considered for publication by Pacific Linguistics in Canberra. Her research interests include community and grass-roots based projects especially regarding literacy and cultural conservation. Email: levyed@scubamadang.com.pg