RSPAS Working Papers in Linguistics

Linguistics Dept Home Working Papers Home  

        Will Taba survive? A 'stressed' language from eastern Indonesia

John Bowden
Linguistics Department
Research School of Pacific and Asian Studies
Australian National University

John.Bowden@anu.edu.au

This paper discusses the potential for language death amongst Taba (East Makian) speakers from North Maluku, Indonesia. After sketching out some of the relevant historical background and examining recent changes in how the language is used, a variety of structural changes in the language, including the ongoing loss of speech levels and decay in the numeral clasifier system, as well as the adoption of a large number of closed class loan words from Malay are discussed. While none of these changes in the language, if taken on its own, would be of too much concern for the future health of Taba, the sheer number of changes currently taking place suggest that the long term prognosis for survival of the language is not good.

1. Introduction

This paper discusses the potential for language death amongst Taba (otherwise known as East Makian or Makian Dalam) speakers from the Halmahera region of North Maluku, Indonesia. Before discussing the Taba situation in detail, however, it is necessary to say a few general words about the notion of language death and what we know about the kinds of circumstances that characteristically lead to language death in communities around the world.

When languages 'die', it is not usually the case that they just stop being spoken overnight. Language death is usually a gradual process brought about within the context of certain kinds of social situations through which people gradually, usually over a course of a few generations, switch from speaking one language to speaking another language. Of course, the underlying reasons that people give up one language are never purely linguistic, but are primarily social. Sasse has provided a general overview of the kind of process that is typically seen in cases of language death (1992: 9-10). Sasse maintains that that three sets of conditions generally involved in the process of language death, and that these occur more or less sequentially. Only the last of his three factors is truly purely linguistic. Sasse lists these three factors as:

  • External setting
The external setting involves the extralinguistic factors which are present in situations of language death: 'cultural, sociological, ethnohistorical, economic, etc., processes, which create, in a certain speech community, a situation of pressure which forces the community to give up its language'

  • Speech behaviour
Such external settings, then, induce changes in people's speech behaviour. This is characterised as 'the regular use of variables, which in a given speech community, are bound with social parameters, e.g. the use of different languages in multilingual settings, he use of different styles of one language, domains of language and styles, attitudes towards variants of language, and so on'

  • Structural change
This kind of speech behavior in turn results in certain structural consequences, i.e. 'the purely structural, substantial linguistic set of phenomena, e.g. changes in the phonology, morphology, syntax and lexicon of the language threatened by extinction'

These three processes are seen as occurring in a fixed temporal sequence, and Sasse illustrates this sequence as illustrated in figure 1.

Stages of Language Death

2. The external setting: a brief history

Taba is spoken chiefly on the eastern side of Makian island, on parts of Moti and the Kayoa islands, as well as in a number of villages on adjacent parts of the South Halmahera coast, as illustrated in figure 2. A sizeable community of Makianese people also live in a transmigration area in Malifut, near Kao on the northern peninsular of Halmahera island. In addition, there is a significant community of Taba speakers living in the regional metropolitan centre of Ternate. There are reports of further speakers of the language on Bacan and Obi islands as well as in the hinterland of Jailolo on Halmahera. While I have been able to confirm that there exists at least one Taba speaking community on Bacan island, I have not been able to confirm any of the other reports about communities around Jailolo or on Obi personally.

Makian island itself is home to two different languages: Taba is spoken on the eastern side of the island while a Papuan language is spoken on the western half of the island. In English this language has been called West Makian and in Indonesian it is referred to as Makian Luar or 'Outer Makian'. Taba speakers refer to the language as Taba Lik 'Outer Taba', while the speakers of the language itself refer to it (and Makian island) as Moi.

Useful introductions to the history of North Maluku can be found in van Fraassen (1981) and Andaya (1993) from which most of the following notes, unless otherwise cited, have been culled. Makian, as well as Ternate, Tidore, Moti and Bacan islands are the islands to which cloves (Eugenia aromatica) are indigenous. From as early as the first century AD, there were reports of cloves in Chinese, Indian and Roman sources. Trade routes direct from Java to the North Moluccan region were in existence from at least the fourteenth century, and it appears that Islam arrived in the region not long after. Before European contact, the people of these islands and the nearby region were to a large degree subject to the power of sultanates based in Ternate, Tidore, Jailolo and Bacan. Makianese oral tradition suggests that the sultanate of Bacan may originally have been Makianese. The sultanates acted as conduits for the trade and their rulers extracted tributes from the people living in the area. It appears that Malay has been spoken in the region from about the 15th century. Given the fairly widely dispersed nature of power centres up to this time, it is unlikely that any particular lingua franca had any real dominance in the area until later historical events intervened.

In 1512 (just after Malacca's fall to the Portuguese in 1511), the Portuguese explorer Francisco Serrčo first arrived in Ternate. Although the Portuguese did not immediately found any settlements in the area, Serrčo remained in Ternate until his death early in 1521, and a number of Portuguese ships passed through the region over this period. Meanwhile, the Spanish had begun to take an interest in the region, and in late 1521, after Ferdinand Magellan's death in the Philippines, two ships from his fleet managed to reach Tidore where their crews were received hospitably. After staying in Tidore for about six weeks (and loading their vessels with a cargo of cloves) they returned to Spain as heroes and sparked off great interest in the region back in Europe.

The Portuguese returned to Ternate in 1522 and built a fort where they maintained a permanent settlement until 1574. They also established a presence in Ambon and a few other places. The Portuguese saw their role in Maluku as twofold: first, they wished to control the trade of cloves from the region, and second, they wanted to propagate Christianity. Both of these aims brought them into conflict with the Ternatans. After Sultan Hairun of Ternate was murdered in 1570, the Ternatans struck back. Assisted by Tidore and Bacan, they forced the Portuguese back to Ambon in 1574.

Tidore, fearful of the growing power of Ternate, invited the Portuguese back a few years later. They tried establishing a fort on Tidore in 1578, but Ternate remained the centre of the clove trade. A succession of brief Dutch expeditions to Ternate around the turn of the seventeenth century resulted in an alliance between the two powers against the Portuguese who were finally expelled from Tidore (and Ambon) in 1605.

However, the Dutch alliance with Ternate eventually secured control of the region. The Spanish maintained a presence on Tidore and the southwestern part of Ternate for a time, but the Dutch took control of Moti, Bacan and also Makian islands and the western part of Halmahera. During this period, Makian island was brought firmly into the struggle between the colonial powers. The construction of three forts on Makian took place after the Dutch expelled the Spanish from the island during 1610 and 1611. By 1663 the Spanish had given up any attempt at maintaining a presence in the region and withdrew to Manila.

It was probably during this period that the use of both Malay and Terrnatan as lingua francas in the area began to emerge. However, once the Dutch had taken secure control, cloves started being grown in a number of other places, most notably Ambon, and the strategic significance of North Maluku declined. The new lingua francas were not yet in a position to put any of the local indigenous languages under any real threat yet.

The Dutch remained in control of what were the Dutch East Indies and became Indonesia until World War II, when the Japanese invaded. After Japan's eventual defeat, the Dutch attempted to reassert control over their former colony but the attempt was unsuccessful, and Indonesian independence was proclaimed on 17 August 1945. Makian has been a part of the Republic of Indonesia since that time. National independence and economic development have now brought a huge number of changes to Makian.

Most important of these changes as far as the language is concerned is a huge growth in schooling. All schooling is conducted in Indonesian, and this has contributed to the increasing 'Malayicisation' of Taba. Increased communications over this time, particularly radio, and to a lesser extent television, have also contributed to the increasing impact of Malay on the language.

Traditionally, society was organised into endogamous patrilineal clans or soan. Today, however, the traditional soan no longer plays a very important role in Taba society and the old rules of endogamy are no longer enforced. On Makian today, intervillage (and thus inter- soan) marriage is common. Marriage with people from other Islamic ethnic groups is also widespread.

Makian island is an active volcano with a long history of devastating eruptions. Verbeek (1908: 14) reports that there were eruptions in 1646, 1760, 1861-64, and 1890. The most recent eruption in 1988 resulted in the temporary evacuation of all the island's inhabitants, and this evacuation has meant that all contemporary Taba speakers, no matter where they live today, have all experienced substantial portions of their lives living in locations which are predominantly non-Taba speaking. Over the years, eruptions have led to widespread migration both from and to Makian island. Although the propensity for eruptions has continued to provide a motivation for Makian's inhabitants to settle elsewhere a countervailing motivation to return to the island, or to occupy land left vacant by those who have fled, is provided by the island's rich volcanic soils. Lucardie (1980) discusses Makianese migratory traditions, and, in a paper published not long before the most recent eruption, Lucardie (1983) examines the Indonesian government's transmigration scheme which was designed to encourage the Makianese to settle in Malifut on Halmahera island. This scheme was motivated by geological reports which predicted that a further major eruption was imminent.

Before the advent of government sponsored transmigration this century, large communities of Makianese people had already moved elsewhere. They set up villages on nearby islands where they maintained their native language. This occurred, for example, in the Bacan community referred to above, where Makianese people emigrated after the 1890 eruption.

Makian island itself has been less affected by modernisation than other parts of Indonesia, even other parts of its less developed eastern provinces such as Maluku. Inducements to the population designed to encourage their transmigration to Malifut have included the withdrawal of all government services from Makian island, including government health clinics, schools, and other government offices such as police and military posts. While the local population has found the resources to fund local schools (and pay their teachers), money is scarce on Makian, and the economy is to a fair degree a subsistence one.

3. Speech behaviour: recent changes

Any sensible discussion of recent changes in speech behaviour of the Makianese people must take into account where they are currently living. Although I do not have good information about Taba speakers in all of the areas they have settled over recent years, I can talk to some extent about the situation in Malifut, Ternate, and on Makian island itself.

In Malifut, the Makianese are giving up their language. The transmigration area is not just home to Taba speakers, but also to speakers of a number of other languages. Most significantly, there is a sizeable comunity of speakers of Makian Luar, or Moi, the language of the western side of Makian island. These people came to Malifut as a result of the same transmigration programme that brought Taba speakers to the area. There is also a sizeable community of Pagu speakers in the area who lived here before the advent of trasmigration. In addition, there are quite a number of migrants from a number of diffeent parts of Indonesia, most notably South Sulawesi and Java. In this environment, then, it is hardly surprising that Taba speakers are failing to pass on their language fully to their children. Although the younger Makianese who are being brought up here understand the language reasonably well, they are themselves imperfect speakers at best. While oder Makianese often still do use their langauge amongst themselves, Taba is clearly dying in Malifut.

In Ternate, the situation is similar to that just outlined above for Malifut, except perhaps that the language is in an even more marked state of decline. Ternate is a fairly large regional centre with a population drawn from an even wider catchment than that of Malifut. In Ternate, intermarriage with people from different ethnic groups is probably becoming the norm rather than the exception. Although there are suburbs of Ternate such as Kampung Pisang which have predominantly Makianese communities, those children who do learn anything of the language while living here are generally even more imperfect speakers than those from Malifut.

While Taba speakers are generally proud of their ethnic identity as Makianese, the conception of ethnicity they hold does not appear to be dependent on the language they use to any significant degree. Taba speakers refer to the whole of Makian island as Taba, whether the eastern side where their language is spoken, or the western side where the speakers of Makian Luar live. The label Taba is somewhat ambiguous as to whether or not it refers to the east or to the whole island. Speakers of Makian Luar are generally perceived as manusia Taba 'Taba people', although they can be further subclassified as manusia Taba lik if anyone feels a strong need to disambiguate. It is thus easy to maintain one's identity as manusia Taba, particularly in Malifut where such a large community of Taba people reside, without an overwhelming need to maintain the language.(1) In both Malifut and Ternate, then, the language is clearly on its last legs.

On Makian island itself, the situation is more complicated, and most of the rest of this paper will be devoted to discussing the situation on Makian. Children on Makian are still brought up speaking the language at the moment, but more and more, they are being taught Malay alongside Taba from an early age. When adults address very young children, in fact, they tend to use more Malay than Taba, especially when issuing orders. Baby talk, as I have heard it being used, tends to be Malay rather than Taba. Although as they get older, people start speaking to children in Taba rather than Malay, it is clear that at the moment, children on Makian are being brought up to speak both languages at the same time. How long this is likely to continue into the future is a moot point.

Other aspects of language use on Makian are also clearly changing. Formal speeches are not generally made in Taba any more. This is something which has largely come about since the eruption. The current kepala desa 'village head' in Ngofakiaha village, for instance, always uses Indonesian when making public addresses, whether the audience is a Taba speaking one or not, although his predecessor before the eruption used to use Taba. At weddings and other significant commuinity feasts, etc., while Taba is used during informal conversation, Malay is invariably used for the formal parts of the occasion. Presumably this behaviour has emerged at least partially as a response to the fact that more and more people are marrying others from different ethnic groups, Malay is still used when both marriage partners and most of their respective families are Makianese. During my residence on Makian, I only saw Taba being used on an official occasion once. This occurred when the Sultan of Ternate made a brief visit to Makian island before returning to Ternate during the afternoon and leaving behind his deputy, the Jogugu Moluku Kie Raha, a Makianese man who now resides in Ternate, as the sultan's representative for further fesitivities in the evening. The Jogugu made his speech to the assembled crowd in Taba. I assume he did this for affective reasons, in order to show his identification with the Makianese, although his social standing was way beyond that of most of his village dwelling audience.

All of the population of Makian aged older than about ten years have now spent significant periods of their lives in residence somewhere other than Makian. After the last eruption, the whole population of the island was evacuated and resettlement did not start to take place until about a year after the event. It is perhaps not surprising, then, that significant changes in people's speech behaviour have taken place.

4. The language itself: structural change

One of the most obvious way of seeing linguistic change in progress is to examine the speech of older and younger cohorts of speakers and to plot the differences between the ways in which they speak. The speech of older people can be seen as representative of earlier stages of a language while that of younger people might be seen as representative of the way in which the language is headed. All languages change, however, whether they are 'healthy' or not. In Taba, as in all languages, there is evidence that linguistic change is afoot: there are many characteristics of younger Taba speakers' speech that differ from that of their elders. Many of these changes are clearly the result of influence from the regional lingua franca of North Moluccan Malay which, as we have seen, all contemporay speakers of Taba are also fluent in.

Change due to contact with other languages is one of the most common types of change found in any language, whether it is dying or not, and we need to be able to find ways of distinguishing between change in healthy languages and change in dying languages. While much of the change that occurs in healthy languages bears a close resemblance to change that occurs in dying languages, change in dying languages is generally much mnore drastic and faster occuriing than in healthy languages (see, eg. Dorian, 1981 and Schmidt, 1985).

In this section of the paper, I intend to review just a few aspects of language change in Taba, dealing with a selection of different changes that are presently under way, each of which I believe can be explained in different ways, not all of them diagnostic of language death by any means. We will discuss a avriety of borrowed grammatical words, the use of numeral classifiers,and a decline in the use of speech levels.

4.1. Borrowed 'grammatical words'

It has long been argued that languages tend to be resistant about borrowing grammatical function words, e.g. prepositions or conjunctions, etc. from other languages (see, e.g. Weinreich 1953, republished as 1974: 34). The existence of many such borrowed forms would appear on the face of it to be evidence of a language in trouble. In contemporary Taba usage, there are in fact quite a large number of such function words in common use and we will examine a few of them here: the existential verb ada 'to exist', the conjunctions dadi 'so' (< Malay jadi 'to become') and karna 'because', and the preposition untuk 'for'. All of the examples included below have been taken from natural speech and forms which have been borrowed from Malay are all presented in bold type.

    • ada 'to exist'
    This form is very commonly found in use by younger Taba speakers and even quite often found amongst older speakers of the language. Its use in Taba is exemplified in (1) below, which is taken from a conversation in which the protagonists are discussing some plastic fishing baits that I had brought with me from Australia as a present.

    (1) Yang model ine...   suntung... ine      ada
        yang model i-ne     suntung    i-ne     ada
        REL  model DEM-PROX squid      DEM-PROX exist
        'This type here, squid , this one exists.'
    

    One of the first things that is remarkable about the example given above is the very large number of Malay loans found throughout what is otherwise a stretch of Taba discourse. The only 'pure' Taba forms found here are the demonstratives in fact. The relativiser yang, the nouns model and suntung, as well as the existential verb ada are all borrowed from Malay. While on the face of it, the existence of such utterances in 'Taba' speech is disturbing, none of the Malay forms seen above should be viewed with alarm on their own. Taba has no relativiser of its own and the use of Malay yang here simply serves to focus on model ine 'this type' more closely than would otherwise be the case (in much the same way as it would be used in North Moluccan Malay). The words model and suntung are both common nouns, and as such, can be expected to be highly 'borrowable' in situations of language contact. In this case, although there is a perfectly acceptable Taba alternative for suntung 'squid'; in saisuak, the Malay form is probably used here because the speaker is referring to the kinds of fishing baits that are available in Ternate shops, where he would use Malay rather than Taba if he were seeking to buy some himself. Finally, although ada is one of the function words which might normally cause alarm, there is no existential verb in 'pure Taba', and in many cases it is just simpler to use the Malay word rather than come up with the circumlocutions that would be required if the speaker was going to stick with pure Taba such as 'they sell these there'.

    • dadi 'so' and karna 'because'
    These two conjunctions / discourse connectors were clearly borrowed into Taba at different times. The Taba conjunction dadi 'so, thus' is obvioulsy a rather early borrowing from Malay jadi 'become, thus' since the Taba form reflects Malay /j/ as /d/. The Taba phonological inventory has 15 inherited consonant phonemes and four further phonemes which have ben borrowed. Amongst the borrowed phonemes in contemporary Taba is /j/ which can be seen in words such as jeregen 'jerry can' borrowed ultimately from English but via Malay. Dadi was clearly borrowed very early from Malay since the non-native /j/ was nativised as Taba /d/ at a time when /j/ was still not a productive phoneme in the language as it has since become. Further evidence for the rather fully nativised Taba status of dadi can be seen from the fact that it also occurs as a verb meaning 'to become' in which case its Actor is cross-referenced by the same person and number marking proclitics as are other verbs. In example (2), which illustrates the use of the verb dadi 'to become' we can again see just how common Malay loans (again given in bold) are in normal Taba speech.

    (2) Ni       suka   ndadi      guru
        ni       suka   n=dadi     guru
        3sg.POSS desire 3sg=become guru
        'He wants to bcome a teacher.'
    

    Although dadi is a conjunction / discourse connector borowed from Malay, it can hardly be seen as evidence of incipient language death because it has been used in the language for a very long time without any major decline in the use of the language.

    The conjunction karna 'because' appears to be a much later loan than dadi. While dadi is commonly heard being used by people of all ages, karna is much more commonly used by younger, rather than older speakers. Its use is illustrated in (3).

    (3)  Lomosi    layok        karna    lkiu              kwat
         lomo=si   l=ha-yok     karna    l=kiu             kwat
         other=PL  3pl=ACT-cry  because  3pl=be.frightened EMPH
         'Others were crying because they were very scared.'
         

    Taba clause sequences are noteworthy for the fact that there isa very strong preference for iconic ordering. When two clauses encode temporally sequential events, the event which occurred first must be reported first, followed by the event which came second. There is no Taba conjunction analogous to 'after' in English or sesudah in Indonesian wich allow non-iconic ordering of clauses as in 'I went to the shop after I had eaten lunch'. In Taba, one would have to say the equivalent of 'I ate my lunch and then I went to the shop.' Purely indigenous Taba likewise provides no conjunction that would allow us to express a cause after the clause expressing the result of that causing event. Without using karna in example (3) above, the speaker would have been forced to say somethinmg along the lines of 'others were very scared so they cried'.

    Again, although karna appears on the surface to be a suspicious borrowing since it is a grammatical function word, I would argue that it should not really be seen as evidence for the decline of the language since it is a form which has had no ready equivalent in indigenous Taba. Speakers clearly find it very useful to be able to order their clauses in a non-iconic way and it is hardly surprising that they have adopted this Malay form to allow them to achieve this end.

    • untuk 'for'
    The final 'grammatical word' I wish to examine here is the beneficiary marking untuk 'for'. Like karna 'because', untuk 'for' is not widely used by the oldest speakers of the language but it is common amongst younger speakers. It is illustrated in (4).

    (4)  Untuk  yak...  masure  hasole  lao    ne
         for    me      good    all     baits  PROX
         'For me, all these baits are good.'
    

    As with the case of karna 'because' discussed above, although we might at first suspect that the adoption of a preposition from Malay could be rather alarming in that it may be diagnostic of the drastic changes that so often take place in a language on the verge of extinction, closer examination of the place of untuk within the Taba system suggests that the situation may not be as bad as it first appears. As was the case with karna, Taba has no indigenous form that allows the introduction of a purely beneficial argument. Using purely indigenous Taba linguistic resources, the exact nature of the benefit bestowed on someone must be spelled out in some detail in a subordinate resultative clause. To translate something such as 'I sang a song for her', one would have to use something like the equivalent of 'I sang a song so she could listen to it'; to translate 'I cooked a meal for her' one would have to say 'I cooked a meal so that she could eat it'. In some cases, it is possible to license an argument that can be construed as a beneficiary by means of a possessive construction, as in (5).

    (5)  Mina npe      Mado ni       woya  do
         Mina n=pe     Mado ni       woya  do
         Mina 3sg=make Mado 3sg.POSS water REAL
         'Mina's made Mado's drink.' / 'Mina's made a drink for Mado.'
         

    In a number of Oceanic languages such as Saliba (Margetts, forthcoming) possessive morphology as in example (5) has been extended in meaning so that it may mark a purely beneficial NP. In Taba, however, possessive morphology can only be used if the possessor NP actually does end up with something in his/her possession. If Taba speakers had not all known Malay and had access to the Malay preposition untuk it is highly likely that any method for introducing beneficiaries into Taba grammar would have involved the extension of possessive morphology to fill this function as has been the case in Saliba and other languages. However, given the fact that Taba speakers do all know untuk and its functions, it is not really surprising that younger speakers have found it useful to adopt the form into their native language and save themselves the trouble of having to spell out the benefits of actions in such detail as they wouild otherwise have to. So again, while they have indeed adopted another word from a closed class by using untuk, this should not really be taken as evidence of incipient language death any more than the adoption of forms like ada 'to exist', karna 'because' or the now rather ancient adoption of dadi 'so' should be.

    4.2. The loss of speech levels by younger speakers:

    Taba has a system of speech levels, somewhat reminiscent of the better known systems found in languages such as Javanese and Balinese, albeit rather less elaborated than in these languages. The speech levels are named by Taba speakers using labels which have been borrowed from Malay: alus 'refined', biasa 'normal' and kasar 'coarse'. Alus forms should be used when talking to people older than oneself or when talking to people who are owed special respect for some other reason. Biasa forms should generally be used when speaking to those of the same age or younger while kasar forms are only countenanced when speakers intend to be rude to their addressees as when perhaps an adult chides an errant child. Some illustrative examples are given in (6) below.

    (6) alusbiasakasar 
     jouole-'yes'
     maledihamolam-'hungry'
     matutinhacarita-'to chat'
     paliharatoanam-'to bury'
     pasiartaggil-'to walk'
     mul ni asalmotbotang'to die'
     babasbabasbor'to bite'
     haloinahonhajaweda'to eat'
       hatuto
       matuto
     kudakudaburat'black'
       pies
     -hagahhanata'to steal'
     -kabotkaletaj'dirty'
    One of the most remarkable things about the alus forms shown in the listing above was that while it was certainly easy to collect the alus forms listed above there were many people who knew them and who were eager to make sure that I knew them as well it was a very rare occurrence when I actually heard young Taba speakers use them when addressing their elders. It seems fairly clear that while a number of people (mostly older) still have an extensive knowledge of the alus forms in Taba, the younger Makianese are not learning them any more and that they are falling out of use.

    It appears that the reasons for the decline of speech levels are directly related to the fact that Taba is no longer used for formal addresses on ceremonial occasions. Since the alus forms are generally used only by younger people as they address their elders, most young Taba speakers are no longer being exposed to the alus speech level in such a way that they are able to master its use any more: older people do not use this level when addressing the younger people around them, and younger Taba speakers only seem to use the most common alus forms any more. From what I have seen of life on Makian, it appears that in the past, people gradually gained a command of alus speech by listening to formal speeches in which older speakers who had developed a command of the genre were forced to use it by reason of the fact that there might be older people in the audience. When I heard the Jogugu speak at the festivities occasioned by the arrival of the sultan of Ternate on Makian his speech contained an efflorescence of alus language such as I had never witnessed on any other occasion. Since such events happen much less frequently than they did in the past, it is no wonder that younger Taba speakers are not learning the genre nearly as well as their parents did.

    4.3. The loss of distinctions in numeral classification

    Taba has a rather elaborate system of numeral classifiers. When one wishes to count something, different forms of numbers are required depending on the nature of the thing it is being counted. There is a set of numeral roots which require extra morphemes that 'classify' the things being counted in different ways. These 'classifiers' may take the form of prefixes, proclitics, or independent words that occur before the numeral root itself. A listing of the numeral classifiers I know about is given in (7). (Although I tried to determine all of the classifiers that are used in the language it is quite likely that some less frequent forms have not yet been encountered.)

    (7) Taba numeral classifiers
     p-default
     ha=intervals of measurement, ordinal numbers ('first', 'second', etc.)
     i-single animate
     mat=more than one human
     sis=from 2-9 animals
     beit=mutiples of 10 animals
     mot=small square flat thin things
     gohathings assembled together
     watosmall oblong shaped things
     holapiece of wood / stick
     luklikrolled up things
     aitrees
     awastalks
     ising'hand' (of bananas, etc.)
     kopgrains
     bokaskewer
     coatbundle
     lofarmspan
     tonatten armspans
     odojoints (of bamboo)
     opadays before
     opodays ahead

    Example (8) shows how classifier-numeral collocations are formed with trhe default classifier p-, and the animal classifiers i- and si-.

    (8) a. pso                         plu                         ptol
           p-so                        p-lu                        p-tol
           CLASS-one                   CLASS-two                   CLASS-three
           one (piece of fruit, etc.)  two (pieces of fruit, etc.) three (pieces of fruit, etc.)
           
        b. iso                          silhu                       sithol
           i=so                         sis=lu                      sis=tol
           CLASS=one                    CLASS=two                   CLASS=three
           'one (animal)'               'two (animals)'             'three (animals)'
           
        c. ai          so               ai           lu             ai           tol
           CLASS(tree) one              CLASS (tree) two            CLASS (tree) three
           'one (tree)'                 'two (trees)'               'three (trees)'
           

    As I found when trying to learn alus forms in Taba, when it came to classifiers, I found that older people knew more of these forms than did younger people. Since people knew that I was trying to learn Taba asli 'authentic Taba' (ironically, asli 'authentic' is another Malay loan) I was sometimes coached by older Makianese people on what were the appropriate classifiers to be used when counting certain kinds of things, even when younger speakers were no longer using the forms. The system of Taba measurements lof 'armspan', etc. is giving way to modern metrical measurements; the traditional method for specifying ordinal numbers using the classifier ha= is being replaced by the Malay ordinals pertama 'first', kedua 'second', etc.; the less common classifiers such as hola 'long thin things' and mot 'small square flat thin things' have simply been replaced by the default classifier p-. On one occasion, for instance, when I was buying some envelopes in a local store and asked for amplop mot oenam 'envelopes - small flat thin things - six' I was corrected by the yound woman working in the store and told to say amplop p-oenam 'envelopes - default - six'. An elderly Makianese man who witnessed the exchange in turn chided the young woman for not being able to speak Taba 'properly' any more and berated her for having lesser skills at the language than the balanda-si 'foreigner'.

    5. Conclusion

    The potential for the extinction of Taba is not evenly distributed amogst all the Taba speaking communities of North Maluku. It seems fairly clear that within a couple of generations, there will be hardly anyone speaking the language in places like Malifut or Ternate. The situation on Makian island itself is somewhat less clear. Assuming that there are no further earthquakes in the immediate future that will disrupt the community more than it has already been disrupted, however, the prognosis for the survival of the language many generations into the future is not all that great.

    If any one change in the grammars of younger Taba speakers is to be picked out as an example that might illustrate the potential for imminent language death in Taba, it would have to be the collapse of the classifier system which appears to be under way at present. The loss of speech level distinctions which is also taking place, while cause for concern, is not quite so alarming. Other cases of the death of particular formal genres of what are otherwise quite healthy eastern Indonesian languages have been documented (see Kuiper's 1998 documentation of the loss of 'angry speech' genres from Wejewa, a language spoken in western Sumba, for example). None of the particular cases of worrying loans that have been examined in this paper, taken on their own, would appear to be cause for too much concern, either. Good reasons have been found for the borrowing of forms such as untuk 'for' and karna 'because': reasons that could easily explain the borrowing of these terms in healthy languages. What is alarming about the current situation in Taba, though, is the sheer magnitude of potentially dangerous loans and structural shifts that are going on at present. The changes documented in this paper are only a few of those currently going on; many more could have been added to the list. While we have seen examples of early grammatical loan words such as dadi 'thus', there are only a few scattered examples of these to be found: the current onslaught is of a scale much higher than it has ever been in the past.

    A recent UNESCO publication (Wurm, 1996, ed.) discussing language death around the world documents hundreds of languages which it labels as either 'moribund' (with only a few elderly people remembering the language any more and not actually speaking it with anyone) or 'endangered' (still having communities who speak the language on a daily basis, but with these speakers being confined to people of middle age or older, and with no children being brought upo to speak the language any more). It seems pretty clear that a third category of 'stressed languages' (languages showing signs that they might move into the endangered category within a few generations) could easily be added to this list. The number of languages around the world that might fall into such a category would probably run into the thousands.

    On Makian island, there are certainly signs of stress. It does not yet appear that children will cease to learn the language altogether within the next 10 or 20 years, but there are clear signs that the language they will be learning will be drastically different from what their parents or their grandparents spoke. The outlook for beyond 20 or so years from now is potentially grim.


    References

    Andaya, Leonard Y. 1993.
    The world of Maluku: Eastern Indonesia in the early modern period. Honolulu. University of Hawaii Press.
    Dorian, Nancy, 1981.
    Language death. The life cycle of a Scottish Gaelic dialect. Philadelphia: University of Pennsylvania Press.
    Fraassen, Ch. F. 1981.
    'Historical introduction'. In Katrien Polman, ed. The North Moluccas: an annotated bibliography. KITLV Bibliographic Series 11. The Hague: Martinus Nijhoff., ed., 1-38.
    Kuipers, Joel C., 1998.
    Language, Identity, and Marginality in Indonesia: The changing nature of ritual speech on the island of Sumba. Studies in the Social and Cultural Foundations of Language 18. Cambridge: Cambridge University Press.
    Sasse, Hans-Jürgen, 1992.
    'Theory of language death.' In, Matthias Brenzinger, ed. Language death: Factual and theoretical explorations with special reference to East Africa Berlin & New York: Mouton de Gruyter, 7-30.
    Schmidt, Annette, 1985.
    Young people's Dyirbal: an example of language death from Australia. Cambridge: Cambridge University Press.
    Weinreich, Uriel, 1953. (republished 1974).
    Languages in Contact: Findings and Problems. The Hague: Mouton.
    Wurm, Stephen A., ed. 1996.
    Atlas of the world's languages in danger of disappearing. Paris: Unesco Publishing ; Canberra : Pacific Linguistics.


    Endnotes

    1 It is interesting to note that Makianese people who live in Jakarta profess that their children are being brought up to speak Taba as a first language alongside Indonesian. Presumably, the great distance of these people from their homeland and from sizeable communities of other Makianese means that they see maintaining the language as a more important prerequisite for maintaining their ethnicity than do the people living alongside many other Makianese. Of course, it is hardly conceivable that these children who appear to be learning Taba in Jakarta will go on to teach their own children the language.