e ndonesian lanae siaion and linisics prosecs and

H. SteinhauerThe Indonesian language situation and linguistics; Prospects and possibilities In: Bijdragen tot de Taal-, Land- en Volkenkunde, 150 Volumes of Bijdragen; A Backward Glimpseand a Forward Glimpse 150 (1994), no: 4, Leiden, 755-784

This PDF-file was downloaded from http://www.kitlv-journals.nl

Downloaded from Brill.com04/15/2022 08:46:42AMvia free access

HEENf STEINHAUER

The Indonesian language situationand linguistics

Prospects and possibilities

Nobody knows how many languages are spoken in the world today, but itis estimated that one-tenth of them are spoken in Indonesia. Wurm andHattori (1981, 1983) distinguish just under 500 languages within theborders of Indonesia, but the information on which their findings are basedis heterogeneous, and for some areas nearly nonexistent. The criteria theyuse to distinguish dialects from languages are not always explicit and insome respects problematic. Particularly for varieties of Malay that arespoken in pockets throughout eastern Indonesia and in discontinuouschains of dialects throughout Sumatra, mainland Malaysia and Kalimantan,it is difficult to ascertain whether they represent different languages, and ifso, to establish their boundaries. If different criteria than those of Wurmand Hattori are used for defining a language, the number of languages mayincrease dramatically.1 Whereas Wurm and Hattori (1983:map 40) classifyLamaholot as one language, spoken in eastern Flores and on the islandseast of Flores, Keraf (1978:299) concludes on lexicostatistical grounds thatthe 33 varieties of Lamaholot he investigated (he did not include theLamaholot pockets on the islands of Pantar and Alor; see also Stokhof1975:8-10, 43-4) should be classified as 17 different languages. Accordingto the atlas (Wurm and Hattori 1983:maps 43 and 44) there are 21 lan-guages spoken in the province of South Sulawesi, belonging to four differ-ent Austronesian 'supergroups' or groups of the same status. Grimes andGrimes (1987:15-8), however, also on the basis of lexicostatistical calcula-tions, distinguish 35 separate languages, divided into three 'stocks' withcomplicated subdivisions and one isolate.

Given this unsatisfactory state of the art, Indonesia has started a large-scale language survey project, which ten years from now should result inan up-to-date geographical picture of the Indonesian language situation,based on clearly defined and consistently applied criteria. However, such apicture is by definition out-of-date as soon as it is completed - certainly fora developing, multilingual, polyglossic society such as Indonesia. The lin-guistic situation in Indonesia is highly dynamic and subject to pressures,

1 This is illustrated by the 11th edition of Ethnologue (Grimes 1988), which lists669 living languages for Indonesia.

BKI 150-IV (1994)Downloaded from Brill.com04/15/2022 08:46:42AM

via free access

756 Hein Steinhauer

shifts and changes, which cannot be captured by traditional dialectologicalmethods alone.

In the following sections I shall first present a rough picture of theIndonesian linguistic scene, based on Wurm and Hattori (1983)2 for a staticview, and on the published results of the censuses of 1971, 1980 and 1990for a more dynamic view. Then I shall discuss some of the social factorsbehind the dynamic character of Indonesian languages, and conclude witha survey of future tasks and perspectives of linguistics as I see them.

Indonesia's linguistic sceneAs indicated above, the number of indigenous languages in Indonesiatoday can be conservatively estimated to be nearly 500. These languagesare divided into eight different unrelated language families. Some lan-guages in Irian Jaya are still unclassified and may very well each constitutea family of its own. Table 1 indicates the eight identified language families,each with the number of languages belonging to it, grouped by number ofspeakers. The figures are based on Wurm and Hattori (1983) and includesome 35 languages that are spoken only in eastern Malaysia and BruneiDarussalam, but probably not beyond the borders of Indonesia.

The language families in the table that are not Austronesian are com-monly known as non-Austronesian or Papuan. It should be stressed,though, that our present knowledge cannot justify viewing the non-Aus-tronesian languages as a single group of linguistically related languages.

Table 1. Languages in Indonesia, eastern Malaysia and Brunei Darussalam, grouped by languagefamily and number of speakers

languagefamily

AustronesianTrans New GuineaWest PapuaGeelvinkEast Bird's HeadSkoKwomtariSepik-Ramuunclassified

total

1 = 9 3

2 = 0 4

1

33

1

I

8

= <200= 200-1000

2

2

i

3

3

2032341

111

_L56

5

6

number of speakers

4 5 6

62 126 8248 61 1310 8

11 11

_ 2

121 199 104

= 1000-10,000= 10,000-100,000

7

28

28

7 _

8 =

8 languagesnumber of

14 337157

2163

2115.

14 533

100,000 - 1 million> one million

Table 1 includes three languages with no speakers. Wurm and Hattori(1981:map 2) mention only Mapia (once spoken on the island of that name,

2 For a discussion on the method and merits of this atlas see Steinhauer 1986.


The Indonesian language situation and linguistics 757

and known from Stokhof 1988) and Batumerah (once spoken in thevillage of that name, a village that has now been absorbed by AmbonTown; see Collins 1983:62) as being extinct. Another extinct language(not mentioned in the atlas) is the Tambora language, which was destroyedby the eruption of Mount Tambora in 1815. The language is known from ashort wordlist in Raffles (1817:appendix F, cxcviii-ix)3 and a subsequentaccount by Zollinger (1850).

To a certain extent it is possible to derive patterns of language shift fromthe results of the successive censuses of 1971, 1980 and 1990.

The census figures of 1971 distinguish respondents that know Indo-nesian from those that do not, these 'others' being divided into speakersof a local vernacular and those speaking a foreign language (such asChinese or Arabic, presumably). In the published data on the 1971 census(Biro Pusat Statistik 1974) the categories of those knowing Indonesianand 'others' are furthermore broken down by residence, age group, andsex.4 The figures in Biro Pusat Statistik 1974:60 show that knowledge ofIndonesian is more widespread among men than among women, that it ismore widespread in urban than in rural areas, and that there is somecorrelation with education. This is suggested by a sharp rise in the figuresindicating knowledge of Indonesian in age groups older than theyoungest group and a gradual fall in these figures for the higher agegroups. In Table 2 these observations are examined further, by comparingthe percentages of people knowing Indonesian with those able to readand write Latin script. The categories are further broken down by sex,residence, and age group.5

It is apparent from Table 2 that a significant number of people learned toread and write in a language other than Indonesian (a regional languagefor instance, or Dutch, in the case of some of the older generation). For allage groups illiteracy is greater among women than among men. The ideamust have been widespread that education for women was a waste of timeand money, if not contrary to Nature or God's Plan.

3 The words given by Raffles are obviously not Austronesian. So far, however, nonon-Austronesian language has been identified either to which the Tambora languagecould be related.4 Residence is either urban or rural. The age groups used here are 0-9, 10-19, 20-29, 30-39, 40-49, 50-59, 60-69 and 70+, with the understanding that a respondentborn on 23 August 1943 is still included in the age group 4 0 ^ 9 on 22 August 1993,but in the next group on the next day.5 The percentages of respondents who know Indonesian have been calculated on thebasis of the figures given in Biro Pusat Statistik 1974:60. The percentages ofrespondents literate in Latin script are taken from Biro Pusat Statistik 1974:61-5.Where the percentages for knowledge of Indonesian exceed those for literacy, theyhave been printed in bold.


758 Hein Steinhauer

Table 2. Knowledge of Indonesian and literacy in 1971, broken down by age group, residence andsex

Age groupand residence

0-9

10-19

20-29

30-39

40-49

50-59

60-69

70+

UR

UR

U

R

UR

UR

U

R

U

R

uR

number ofrespondents

men

3,089,40016,041,982

2,459,68510,481,291

1,634,9955,944,574

1.287.4296,350,334

899,8864,590,972

470,1412,548,731

226,6981,404,890

116,125774,454

women

2,991,76715,737,625

2,460,32710,103,727

1,725,2247,651,364

1.340,2326,904,677

844,2004,416,357

474,4012,586,596

268,9511,580,914

150,859783,983

% knowing

Indonesian

men

51.316.6

87.359.2

91.461.7

86.349.1

82.140.3

77.434.9

70.228.5

61.324.8

women

51.416.4

85.353.7

83.542.1

70.526.1

63.019.5

55.015.2

49.013.1

45.013.0

% literate inLatin script

men

_

-

92.680.9

94.279.3

86.264.4

78.851.6

71.141.1

58.528.6

43.618.6

women

88.673.5

80.955.1

57.630.2

43.818.0

30.19.8

20.45.5

13.83.9

U = urban R = rural

Table 2 also shows that knowledge of Indonesian can be acquired notonly through education, but through exposure and inter-ethnic contacts aswell. Again, this is less apparent among women, whose social mobility usedto be much more restricted than it is now. Obviously the exposure factor ismuch stronger in urban than in rural environments: inter-ethnic contacts,and domains and topics for which the use of Indonesian is moreappropriate than a regional language, are a common feature of urban life,while they are rare phenomena in traditional rural communities. Therelative importance of exposure is greater with the higher age groups,because they have had a longer period of exposure and have generallyhad less education. The effects of exposure get the upper hand earlier forwomen (especially rural women), because the education factor was still oflittle significance for them in 1971. But even for urban men, exposurebecomes a relatively strong factor at an early age.

The most important question pertaining to language in the 1971 censuswas 'Do you know Indonesian?', and 'If not, do you know a regional or aforeign language?' In the 1980 and 1990 censuses, the same informationwas collected, with more detailed questions about daily language use:'What is your daily language at home?'; 'If not Indonesian, do you knowIndonesian?'

In the published results of the 1980 and 1990 censuses (Biro Pusat Sta-tistik 1983 and 1992) only eight major languages are distinguished besides



Indonesian, all other languages being grouped as 'other'.6 Both censusesreport a considerable rise in the avowed knowledge of Indonesian.7

It should be stressed that answers to the above questions are necessarilysubjective. Even for a trained linguist, criteria for establishing knowledgeof Indonesian are difficult to establish. Is Riau Malay Indonesian? OrAmbonese Malay? Besides, the answer 'a bit' is not allowed. There mayhave been a tendency to answer 'yes' as a sign of one's good citizenship,where 'a bit' or perhaps 'no' would have been more in accordance withthe facts. If the benevolent response towards the foreign tourist whomanages to stammer a one-word phrase in heavily accented Indonesian, O,bahasa Indonesianya lancar sekali! ('Oh, you speak fluent Indonesian!'),is to be taken as indicative of the average Indonesian's evaluation oflinguistic competence, then the results of the censuses regardingknowledge of Indonesian should be viewed with some scepticism.8

However, I assume that such factors did not change dramatically in thenineteen years between 1971 and 1990, so that a comparison of the resultsof the censuses does make sense.

Table 3 shows the census count of knowledge of Indonesian inabsolute numbers and as percentages of the rural and urban populationsbroken down by age group and sex.9

The picture which arises from this table is consistent with thetendencies already observed. For all age groups there is a gradual increasein knowledge of Indonesian between 1971 and 1990, ranging up topractically 100% for urban adolescents and young adults. Women arecatching up with men, except for the higher age groups. Also, the gapbetween rural and urban areas is narrowing. Not shown here, but apparentfrom the published data, is an increasing correlation between knowledgeof Indonesian and literacy in Latin script.

Let us now look at the census figures on language use at home,collected for the 1980 and 1990 censuses. In Table 4 (see the appendix)the absolute numbers of first-language speakers of Indonesian, Javanese,Sundanese, Madurese, Minangkabau, Buginese, Batak, Balinese, Banjarese

6 The raw census data reported some 200 possible answers to the question aboutone's daily language, other answers being qualified as 'other'.7 The data are not immediately comparable, as the breakdown into age groups in thelatter two censuses differs from the 1971 census: 0-4, 5-9, 10-14, 15-24, 25-49,50+. The 1990 census, moreover, does not try to qualify the youngest age grouplinguistically.s The question 'what is your daily language at home?' is also problematic: dailybilingualism is excluded as a possible choice.9 The data are from Biro Pusat Statistik 1974:60, 1983:76-7 and 1992:190-1. Toenable comparison with 1971 the age groups distinguished in the three censuses havebeen rearranged to achieve a common denominator, i.e. a breakdown into age groupsof 0-9, 10-49 and 50+; 5+ (everyone age 5 or above) has been added as the sumtotal for the comparable figures of 1980 and 1990. The percentages are those of thetotal age group specified for sex and residence.


760 Hein Steinhauer

Table 3. Absolute numbers and percentages of respondents knowing Indonesian in 1971, 1980 and

1990, broken down by age group, residence and sex

agegroup

0-4

5-9

10-14

15-24

25-49

5+

0-9

10-49

50+

mfmf

mfmf

mf

mf

mfmf

mfmf

mfmf

mfmf

mfmf

mfmf

m = male

1971

-

-

-

-

-

-

1,586,0641,537,162

51.3%51.4%

5,490,2245,014,797

87.4%78.7%

601,813460,855

72.6%51.5%

URBAN1980

979,069922,606

42.1%42.0%

1,495,5541,447,591

69.1%69.9%

1,925,9791,864,874

97.2%96.7%

3,665,9643,753,961

98.1%95.7%

4,474,4093,950,585

95.8%85.7%

12,861,63112,009,977

91.1%84.5%

2,474,6232,370,197

55.1%55.6%

10,066,3529,569,420

96.9%91.5%

1,299,725992,966

83.5%59.1%

f= female

1990

-

2,728,8412,614,674

81.3%85.2%

3,191,9843,089,446

99.9%99.9%

6,179,7976,558,836

100%99.9%

8,813,8108,266,011

98.2%94.1%

23,488,32222,728,429

95.2%91.2%

-

18,185,59117,914,293

98.8%96.7%

2,573,8902,199,462

88.4%68.5%

22

1411

1

1970

-

-

-

-

-

-

,662,016,573,553

16.6%16.4%

,835,663,314,664

54.2%38.9%

,482,628701,505

31.4%14.2%

R U R A L1980

725,356687,071

8.5%8.4%

3,205,5183,174,575

37.0%38.1%

6,240,9455,636,076

87.3%85.9%

8,517,4708,555,687

87.3%78.7%

12,109,9258,824,598

76.6%53.3%

33,438,48327,630,334

69.6%56.1%

3,930,8743,860,093

22.9%23.4%

26,868,34023,016,361

82.1%67.8%

3,364,6251,439,421

50.8%20.8%

1990

-

4,913,2194,743,485

57.3%58.5%

7,571,2797,092,541

96.7%96.8%

10,558,73710,664,903

96.9%94.2%

16,484,80114,605,662

88.7%75.5%

44,781,78340,107,559

83.0%73.3%

-

34,614,81732,363,106

92.8%85.1%

5,253,7473,000,960

65.2%34.8%

and 'other' regional languages are compared for 1980 and 1990, brokendown by age group, sex, and residence.10

10 The data are from Biro Pusat Statistik 1983:70-2 and 1992:174, 177, 180. Thepercentages are based on the total numbers of respondents specified for language andage group; the category (language) 'not stated' has been left out. For the 'Total'numbers and percentages, the small numbers of respondents whose age group wasunknown were also left out.



Table 4 corroborates the finding that Indonesian is closely connectedwith urban culture. For all age groups, not only the percentages but alsothe absolute numbers of its daily speakers in urban environments exceedthose of the rural population. For all regional languages the proportionsare the other way around.

Table 4 shows that there are striking differences among the regionallanguages. They show an overall increase of speakers in absolute numbers,with the exception of certain age groups, and with considerable differencein degree. The youngest age groups of rural speakers of Javanese, Madu-rese, Minangkabau, Balinese and Buginese show a decrease in speakers (initalics in Table 4). As the figures show, urbanization cannot be the onlyexplanation. Language shift appears to be only a marginal factor. Perhapsbirth control makes the difference. The marked regional differencescertainly deserve investigation. Especially striking is the differencebetween neighbouring communities, such as Sundanese and Javanese.Javanese shows an overall decline in percentage. It is the only languagerepresented in Table 4 for which the absolute number of rural speakers ofanother group besides the youngest age group declines, in this case theage group of adolescents and young adults. Sundanese, on the other hand,shows a steady percentual increase for practically all age groups (urban,rural and total). In Table 4 percentual increase is printed in bold.

A somewhat less spectacular difference shows up for two otherneighbouring languages, Minangkabau and Batak. Whereas the relativeimportance of Batak is increasing among the rural population, it is theurban percentages which are increasing for Minangkabau. The reason isapparently the phenomenon of merantau, migration to towns and cities(often outside the home area) by adolescents and adult Minangkabau men.Compare the numbers of urban Batak and Minangkabau speakers in 1990,broken down by age group and sex, in Table 5.11

Table 5. Urban Batak and Minangkabau speakers in 1990

age group

5-910-1415-2425-4950+

women

24,72724,53060,46477,85030,777

Batak

men

27,71526,87957,75679,87424,679

Minangkabau

women

74,71369,11097,708

120,0078,869

men

62,38059,673

123,179163,80751,695

Table 4 also shows that Balinese is reasonably persistent among theadolescent and higher age groups, but less so for the two youngest agegroups; the figures also show a high rate of urbanization not accompaniedby language loss.

11 The data are based on Biro Pusat Statistik 1992:172. No comparison has beenmade with 1980, because the published data for 1980 are not broken down by sex.


762 Hein Steinhauer

Compared with Madurese and Buginese, Batak shows great vitality,with its growing role in the rural mosaic of languages. Madurese andBuginese seem to be the least vital of the languages represented in Table 4.The reasons for this deserve investigation. For Madurese a plausibleexplanation is that the mixed language situation in East Java (Javaneseand Madurese; see Sodaqoh Zainudin et al. 1978:8 and Wurm and Hattori1983:map 39) causes a disproportionate shift to a common language.However, the increase of Indonesian in East Java is below the average (seeTable 6). When Madurese shift to another language, they shift to Javanese.The slight numerical increase of Javanese in East Java, as against a declinein Yogyakarta, Central Java, and West Java (according to the data in Table6), may be an indication of this, although differing transmigration ratesshould also be taken into account.

Table 4 clearly shows that it is Indonesian that is increasing everywhere,in absolute figures as well as percentages (with the only exception beingthe oldest age group in rural areas). This is only to be expected. What isamazing, however, is the position of Banjarese, with a total increase ofmore than 55%: 32.9% for the urban population, which is well below thenational urban population increase of 74.7%, presumably because of alower urbanization rate; but 66.6% for the rural population, for which thenational increase is only 9.8%. This strong position of Banjarese is causedby the rapid disappearance of the smaller regional languages in South andCentral Kalimantan, whose speakers switch to Banjarese as their firstlanguage (see below).

This shift is seemingly at variance with the favourable figures for the'other' languages in Table 4. These figures, however, can hardly be takenat face value. In the first place, differences similar to those betweenMadurese and Banjarese may be encountered among the 'other'languages too. Second, it should be kept in mind that these 'other'languages represent nearly 500 different languages, many of them withfew speakers, whose shift to another language would not show up in thestatistics. Furthermore, most of the minor languages are spoken in areas ofIndonesia which are less developed than the rest of the country, andrelatively less affected by urbanization, education, mass media, increasedmobility, and easy access, so that a major shift to Indonesian is yet to come.Finally, some unusually large differences between 1980 and 1990 for the'other' category seem to be indicative of a classification problem, ratherthan of a real increase in 'other' languages. Such differences tend to becoupled with unusually high figures for the category tak terjawab,translated as 'not stated', but also meaning 'unanswerable'.

Table 6 gives daily language (Indonesian or 'other') per province, bothin absolute numbers and as a percentage of the total population of theprovince (including the 'not stated' cases). Javanese has been added,



Table 6. Indonesian and Javanese vis-a-vis the minor regional languages in 1980 and 1990, perprovince, in absolute numbers, and as a percentage of the total population of the province

Provincei

Aceh

North Sumatra

West Sumatra

Riau

Jambi

South Sumatra

Bengkulu

Lampung

Jakarta

West Java

Central Java

Yogyakarta

East Java

Bali

West Lesser Sunda

East Lesser Sunda

Moluccas

Irian Jaya

North Sulawesi

Central Sulawesi

South Sulawesi

Census

1980199019801990

1980199019801990

1980199019801990

1980199019801990

1980199019801990

1980199019801990

1980199019801990

1980199019801990

1980199019801990

1980199019801990

19801990

Southeast Sulawesi 19801990

Indonesiannumber

199,060373,355

2,517,1233,532,362

74,38780,200

255,525506,267

176,793228,763430,962344,737

59,90785,761

436,522680,956

5,956,8636,889,0812,769,6854,690,654

268,353373,10655,20982,826

654,786787,599

81,785134,686

82,734103,728374,603515,421

693,651587,547314,364513,019

582,714911,351288,215479,512

477,591861,862

99,970239,673

%

7.612.730.140.0

2.22.3

11.817.9

12.213.19.36.3

7.88.4

9.413.0

91.993.210.115.1

1.11.5

2.0

3.1

2.2

2.7

3.35.3

3.03.6

13.718.4

49.237.128.437.4

27.641.022.432.4

7.914.010.621.0

'Other'number

2,108,2692,317,3501,095,648

951,661

72,04959,921

1,390,2111,440,372

883,456921,912

3,302,7513,871,042

550,558651,079758,827797,050

81,020162,655147,922278,212

36,13519,1747,9783,682

60,90838,25832,09126,174

2,560,1652,672,5152,300,6882,265,739

694,007922,638622,365748,291

1,499,2491,264,562

683,419711,506

2,688,8932,565,259

374,619660,087

%

80.878.813.110.8

2.11.7

64.251.0

61.252.771.471.2

71.763.916.415.3

1.3

2.2

0.50.9

0.10.1

0.3

0.1

0.2

0.1

1.31.0

94.092.684.180.9

49.358.356.254.6

70.956.971.648.0

44.441.739.857.9

Javanesenumber

174,920176,845

2,905,8662,806,129

55,59694,478

183,936378,572

245,337363,231573,458

1,018,461

118,094181,928

2,886,2193,091,330

235,951188,170

3,652,0503,240,577

24,579,06924,413,734

2,682,8932,584,834

21,719,92721,947,986

27,51241,492

6,95910,0242,5562,548

15,95554,4844.299

74,377

20,39524,78737,36978,156

44,218108,39034,22878,679

%

6.7

6.021.016.3

1.62.78.5

13.4

17.020.812.418.7

15.417.962.459.2

3.6

2.513.310.4

96.996.097.696.2

74.574.6

1.11.6

0.30.30.1

0.1

1.13.4

0.4

5.4

1.01.1

2.95.3

0.71.83.66.9


764

Province

South Kalimantan

Census

19801990

Central Kalimantan 1980

East Kalimantan

West Kalimantan

1990

1980199019801990

Hem Steinhauer

Indonesiannumber

50,84265,83737,88060,366

335,357637,141230,422233,268

%

2.52.9

2.95.0

27.638.8

9.38.4

number

560,21659,813

639,711579,300

560,690378,011

1,941,5332,151,605

Other'%

27.22.6

67.048.1

46.223.078.177.4

Javanesenumber

44,218198,00237,369

124,376

122,840222,61441,424

125,787

%

4.78.6

4.910.3

10.113.5

1.74.5

which because of spontaneous and organized transmigration is becominga major language in several provinces outside Java.12

For Irian Jaya the 1980 census reports a high figure for the category takterjawab: 161,785 (vs 20,394 for 1990). The reason is that in 1980 theinterior of the province was not subjected to a first-hand survey. Probablyall of these unclassified 'respondents' belong to the 'other' category. Thisalso holds for the 327,472 unclassified respondents in Southeast Sulawesiin 1980 (vs only 4,772 in 1990). The discrepancy may be due to localcircumstances or to insufficient instruction of census fieldworkers in 1980.

As Table 6 indicates, there must have been classification problems inmost of the Sumatran provinces and in the Moluccas. In South Sumatra, forinstance, there is an increase in speakers of 'other' languages of more thanhalf a million. At the same time the figures show an unexpected decline ofIndonesian, while for 1980 there are 224,680 unclassified respondents (vs51,940 for 1990, which is also a relatively high number). It can be assumedthat the Malay-like character of most Sumatran regional languages causedclassificatory dilemmas (Indonesian, 'other' or tak terjawab), resulting in adifferent outcome for 1990.13 Indonesian shows a decline in the Moluccasas well, whereas 'other' languages increase. Apparently a considerablenumber of speakers of Ambon Malay or another regional language wereclassified as speakers of Indonesian in 1980, but as speakers of 'other'languages in 1990.14

West Java presents another classification problem, since the mixedlanguage of Cirebon may have been classified as Sundanese, Javanese,'other', or tak terjawab. It was probably different ways of census takingin 1990 which resulted in a sharp increase of 'other' speakers, while the

12 The data are from Biro Pusat Statistik 1983:75 and 1992:189. East Timor has notbeen included in the table, because it was left out of the 1980 census. The figures for1980 include the age group 0-4.13 Either because of a change in self-perception among the respondents, or becauseof different instructions to census fieldworkers.14 In 1980, the anti-Indonesian protests by radical Moluccans in the Netherlands in1975 were still fresh enough in people's minds for an avowed Ambonese identity tobe looked upon with suspicion.



number for the category tak terjawab also jumped (120,944 vs only 5,187in 1980).

It is now clear that the numerical increase of 'other' languages in Table4 is misleading, since in 12 of the 26 provinces in Table 6, 'other'languages decline. This is the case not only in provinces with noindigenous minor languages (East Java, Central Java, Yogyakarta, Bali,West Sumatra), but also in the Eastern Lesser Sunda Islands and in NorthSulawesi. In each of these provinces there is a variety of Malay which isintruding on the minor languages, namely Kupang Malay in the EasternLesser Sunda Islands and Manado Malay in North Sulawesi, apparentlyboth counted as Indonesian in the census. A decline also shows up inSouth Sulawesi and in North Sumatra (in both cases more than matched byan increase in Indonesian), and in South Kalimantan and neighbouringprovinces (to the benefit of Banjarese).

As Table 6 shows, Javanese is making inroads in practically everyprovince of Indonesia. The Eastern Lesser Sunda Islands are aconspicuous exception. In the long run, contact with other languages inthe transmigration areas may cause language shifts. Such a shift may havecaused the relative decline of Javanese in provinces where the Javanesepresence dates back several generations, namely Lampung and NorthSumatra. However, in the more recent transmigration destinations (SouthSumatra, Jambi, Riau, Bengkulu, Southeast Sulawesi, Central Sulawesi,Irian Jaya and Kalimantan) the number of speakers of Javanese is stillincreasing, presumably because of new arrivals, thus making up for thenumerical decline of the language in Java itself.

Transmigration has caused linguistically relatively homogeneous areasto become more heterogeneous, especially Sumatran target provinces suchas Lampung, South Sumatra, Jambi, Riau and North Sumatra. To supple-ment the picture of migration, see Table 7. For each major language exceptIndonesian and Javanese, this table gives the provinces where the numberof daily speakers of the language exceeded 4% of the total population ofthe province or was more than 100,000, in either 1980 or 1990.15

Table 7 shows that the most widespread language after Javanese isBuginese (see also Table 6), found in discontinuous areas and increasing insome of them. But again it is Banjarese that appears to show the mostgrowth. Table 8 shows the position of Banjarese vs Indonesian and'other' languages in South Kalimantan, the homeland of Banjarese, as wellas in the other provinces of Kalimantan.16

15 The figures have been calculated on the basis of the data from Biro Pusat Statistik1983:75 and 1992:189. The data for 1990 do not include the age group 0-4, whilethose for 1980 do.16 The data are based on Biro Pusat Statistik 1983:75 and 1992:189. Again it shouldbe stressed that the figures for 1980 include the age group 0-4, while those of 1990do not.


766 Hein Steinhauer

Table 7. Provinces where a major language other than Indonesian and Javanese was spoken bymore than 4% of the population or more than 100,000 speakers, in 1980 or 1990

Language

Sundanese

Madurese

Batak

Minangkabau

Balinese

Buginese

Banjarese

Census

1980

1990

1980

1990

19801990

1980

1990

19801990

1980

1990

1980

1990

Province

West JavaLampungCentral JavaJakartaWest JavaLampungCentral Java

East Java

West KalimantanEast JavaWest Kalimantan

North SumatraNorth SumatraWest Sumatra

West SumatraRiauWest SumatraRiau

BaliBaliCentral Sulawesi

South SulawesiCentral SulawesiSoutheast SulawesiEast KalimantanSouth SulawesiEast KalimantanSoutheast SulawesiCentral SulawesiJambi

South KalimantanCentral KalimantanEast KalimantanSouth KalimantanCentral KalimantanEast KalimantanRiau

Number

20,858,324487,701479,071151,765

22,746,859514,749528,625

6,705,232143,724

6,535,702164,589

2,905,8662,806,129

119,437

3,107,857181,697

3,133,631185,760

2,315,4732,301,337

63,266

2,803,917152,46093,06177,874

2,557,690140,504126,406120,07870,911

1,295,806167,310112,920

1,874,015389,409229,761137,421

Percentage

76.010.6

1.92.3

73.19.9

2.1

23.05.8

22.25.9

34.831.8

3.4

91.28.4

89.46.6

93.891.1

4.3

46.311.99.9

6.441.6

8.611.18.14.1

62.817.59.3

81.732.314.04.9

It is obvious from Table 8 that Banjarese and not Indonesian is thelanguage to which the other regional languages are losing ground, aprocess which does not seem to leave much room for Indonesian. In theneighbouring provinces of Central Kalimantan and East Kalimantan,Banjarese is making strong inroads. In Central Kalimantan the otherregional languages are still relatively strong. In East Kalimantan, however,the pressure of Indonesian and Javanese (see Table 6) is also considerable,thanks to the oil industry, forestry, and transmigration. West Kalimantan isstill the most stable, even showing an absolute increase in the number of



Table 8. Banjarese compared to Indonesian and 'other' languages in the Kalimantan provinces, inabsolute numbers of speakers and as a percentage of the total population of the province

Province

South Kalimantan

Central Kalimantan

East Kalimantan

West Kalimantan

Census

19801990

19801990

19801990

19801990

Indonesian

50,84265,837

37,88060,366

335,357637,141

230,422233,268

Banjaresenumber

1,295,8061,874,015

167,310389,409

112,920229,761

9124,507

percentage

62.881.7

17.532.3

10.114.0

0.00.9

'Other'

560,21659,813

639,711579,380

560,690378,011

1,941,5332,151,605

speakers of 'other' regional languages.The census data on daily language use given in Tables 4 to 8 primarily

compare similar groups of respondents. Because of the irregularbreakdown into age groups these groups cannot be followed throughtime. The only age group in 1990 that corresponds exactly to one in the1980 census data is the 15-24 age group. In Table 9 the data on dailylanguage use for this age group are compared with the combined data forthe 5-9 and 10-14 age groups of the 1980 census.17

As Table 9 shows, the total population of these age groups decreased by9.3% between 1980 and 1990, whether by death or emigration. Under thecombined effects of death, emigration and ongoing urbanization (with a57.3% increase for these age groups in urban areas) the total ruralpopulation decreased by 27.2%. Two languages that clearly deviate fromthe national figures are Balinese and Minangkabau: a high degree ofurbanization among groups speaking these languages is not accompaniedby a shift to Indonesian (increases of 97.6% and 72.4%). Madurese, Batakand Buginese, on the other hand, participate less in urbanization (withincreases of 28.4%, 38.0%, and 39.1%, respectively, well below thenational average), or, if they do, they do not maintain their language aswell as do Balinese. Use of Minangkabau, though strong in urban areas, ismarkedly decreasing in the villages (by 39.1%). Batak and Buginese arealso losing ground in rural areas. Batak, Buginese, Madurese andMinangkabau all show an overall decline well below the national averageof 9.3% (28.8%, 26.0%, 20.3%, and 19.3%). Batak, Buginese andMinangkabau are major languages of the provinces of North Sumatra,South Sulawesi and West Sumatra, respectively (see Table 7). The censusesshow a clear shift towards Indonesian for North Sumatra and SouthSulawesi, but not for West Sumatra (see Table 6).

The explanation for the relative decline of Minangkabau in West

17 The data are from Biro Pusat Statistik 1983:70-2 and 1992:174, 177, 180.


768 Hein Steinhauer

Table 9. Daily language of the 5-14 age group in 1980 and the 15-24 age group in 1990

Language

Javanese

Sundanese

Madurese

Batak

Minangkabau

Balinese

Buginese

Banjarese

Other

Indonesian

Total

Census

1980 (5-14)1990 (15-24)increase abs.

in%

1980 (5-14)1990 (15-24)increase abs.

in%

1980 (5-14)1990 (15-24)increase abs.

in%

1980 (5-14)1990 (15-24)increase abs.

in%

1980 (5-14)1990 (15-24)increase abs.

in%

1980 (5-14)1990 (15-24)increase abs.

in%

1980 (5-14)1990 (15-24)increase abs.

in %

1980 (5-14)1990 (15-24)increase abs.

in%

1980 (5-14)1990 (15-24)increase abs.

in %

1980 (5-14)1990 (15-24)increase abs.

in %

1980 (5-14)1990 (15-24)increase abs.

in %

Urban

2,565,6773,903,571

+ 1,337,894+52.4%

1,129,6881,619,886+490,198

+43.4%

163,786210,247+46,461

+28.4%

85,638118,220+32,582

+38.0%

147,573254,478

+ 106,905+ 72.4%

78,398154,916+76,518

+ 97.6%

94,212131,081+36,869

+39.1%

101.593196,042+94,449

+ 93.0%

780;7871,185,702+404,915

+51.9%

2,981,6175,013,797

+2,032,180+ 68.2%

8,128,96912,787,940+4,658,971

+57.3%

Rural

12,796,2598,947,967

-3,848,292-30.1%

4,876,2143,619,467

-1,256,747-25.8%

1,435,3331,063,832-371,501

-25.9%

843,943543,587

-300,356-35.9%

805,904515,295

-290,609-36.1%

581,855433,229

-148,626-25.5%

841,998562,946

-279,952-33.2%

352,463433,987+82,524

+ 23.5%

6,269,8764,674,126

-1,595,750-25.5%

1,608,9071,361,365-247,542

-15.4%

30,411,75222,154,901-8,256,851

-27.2%

Urban+Rural

15,361,93612,851,538-2,510,398

-16.3%

6,005,9025,239,353-766,549

-12.8%

1,599,1991,274,079-325,040

-20.3%

929,581661,807

-267,774-28.8%

953,477769,773

-183,704-19.3%

660,253588,145-72,108

-10.9%

936,210693,127

-243,083-26.0%

453,056630,029

+ 176,973+ 3 9 . 1 %

7,050,6635,859,828

-1,190,835-16.9%

4,590,5246,375,162

+ 1,784,638+38.9%

38,540,72134,942,841-3,597,880

-9.3%

Bold type indicates figures that deviate significantly from the average figures (found in 'Total' atthe bottom of the table)



Sumatra, and no obvious shift to Indonesian, must be related to themerantau phenomenon discussed above. Apparently, sizeable numbers ofMinangkabau migrating to other provinces do not maintain their nativelanguage.

The general decline of Javanese is also remarkable. The percentages donot deviate very dramatically from the national figures, but the latter areheavily influenced by the figures for Javanese, which outnumbers all otherregional languages in Indonesia. The only two languages whose rural andoverall figures do not conform to the national pattern of decline are againIndonesian and Banjarese.

Endangered languages

I have already pointed out that the data for 'other' languages are not asfavourable (in terms of prospects for the survival of the languages) as theymay seem to be. In fact, since the time when Wurm and Hattori collectedthe data for their atlas, several minor languages have become endangered,if not extinct. I present here a list of such languages I have come across.18

The Moi in the Bird's Head peninsula of Irian Jaya are too close to thetown of Sorong not to be affected by urban culture and the variety ofMalay/Indonesian that goes with it (Raymond Menick, personalcommunication). Moreover, reports in Indonesian newspapers in 1993indicate that maintaining the traditional Moi way of life is made difficult bythe aggressive presence of logging companies in their homeland.

The Inanwatan (on the southern coast of the Bird's Head peninsula) arealso shifting to Malay. Children no longer use the language of theirgrandparents (Lourens de Vries, personal communication).

Malay/Indonesian has also become the main language of the Waropen.A new administrative centre and a road around the swamps along the eastcoast of Geelvink Bay, the original homeland of the Waropen, has resultedin the abandonment of their traditional villages and in considerable inter-ethnic contact. Waropen people below the age of 40 exclusively use localMalay; the only active speakers of Waropen are old people (over 65).Consequently, the place the language is best preserved is the smallcommunity of Waropen immigrants in Serui on the island of Yapen: for thesimple reason that life expectancy appears to be higher there than on themainland (Paul van Velzen, personal communication).

Silzer and Haikkinen (1984:99-103) indicate that in about half of theYapen villages where one of the languages Ambai, Ansus, Serui-Laut andWoriasi are spoken, 'the majority of the preschool children use Indonesianwhen playing'. In the other villages they use 'non-Indonesian [...] butmany adults speak some Indonesian' (Silzer and Haikkinen 1984:4). Theirinformation is based on a report from 1981. Today the prospects of several

18 It should be noted that in the vast majority of language descriptions, ecologicalinformation on the language is missing.


770 Hein Steinhauer

of the indigenous languages of Yapen (among them the non-AustronesianYawa language) seem to be even bleaker than those of Waropen (Paul vanVelzen, personal communication).

In Wurm and Hattori (1983:map 45) the Kayeli language of Bum Islandis estimated to have 1,000 speakers. Grimes (1991:31) argues that theremust have been at least three dialects of this language: Leliale, Kayeli andLumaete. When he conducted his research on the Buru language in the1980s people still remembered the existence of the Lumaete dialect, but thedialect had recently become extinct. The last speaker of the Leliali dialectdied in March 1989. In that year 'the four remaining speakers of [the third]dialect [...] were all over 60 years old [and] did not use [the dialect] amongthemselves' (Grimes 1991:32).

An earthquake and tidal wave on 30 September 1898 meant thebeginning of the end for the Paulohi language in Seram. According to localtradition, the church in which most of the speakers of the language weregathered simply slid into the ocean. Survivors enabled Stresemann todescribe the language (Stresemann 1918), but they were unable topreserve their identity in the face of Wemale and Amahai speakers movingdown from the interior. In the mid 1970s only a few very old people wereable to speak the language (James T. Collins, personal communication).Collins (1983:38, 99) lists Naka'ela, Hulung, Loun, Paulohi, Kamarian-Rumakai, Nusalaut, Amahai and Eti-Kaibobo as languages near extinction.These languages are marked 'dying' in Wurm and Hattori (1983:map 45).For Nusalaut a note is added that the language 'is today only spoken bythe oldest generation in the village of Titawae on the Nusa Laut Island'. Inthe mid 1970s each of the villages Naka'ela, Hulung and Loun on SeramIsland appeared to have only three speakers left for the language of thesame name (Collins 1983:38). The original Loun community was decimatedby the Spanish influenza in 1919. The survivors fled into the woods andsomehow never regrouped afterwards. The last three remaining 'speakers'Collins could find in the mid 1970s had only a defective knowledge oftheir language (James T. Collins, personal communication).

In North Sulawesi Malay has been used since the nineteenth century asthe language of the Protestant church and schools. It has greatly enhancedthe prestige of the local variety of Malay, Manado Malay, to the detrimentof other local languages. The decline of these latter languages is a processwhich was lamented as early as the 1920s. Today the use of ManadoMalay in daily oral communication in North Sulawesi is growing rapidly,spreading from city to town all over the province. According to AkunDanie (1991:31) it is beginning to be used even in parts of theneighbouring province of Central Sulawesi. One of the languagesthreatened with extinction is Bantik, which is still maintained by olderpeople in five villages which have all recently been absorbed by the city ofManado. In a village that was still Bantik speaking some 25 years ago, butwhich had been absorbed by the city earlier, all that remains of the Bantik



community is two speakers and a graveyard (George Bawole, personalcommunication).

Table 8 makes clear the special position of Banjarese vis-a-vis the otherregional languages in South, Central and East Kalimantan. The weakposition of the 'other' languages is evidenced by various sources.Darmansyah (1985:1), for instance, states that 'most Maanyan speakersspeak Banjarese [...] but not vice versa'. Usop et al. (1992:1) observe thatthe Lamandau language of Central Kalimantan is still used, but that inofficial, nontraditional meetings it is being replaced by Indonesian andBanjarese; the authors sketch an ongoing process of culture mixing,already visible in the capital of the district, which will eventually result inthe complete disappearance of the language.

In his dialectological study of Banjarese in South Kalimantan, Kawi(1991) finds an unusually high number of doublets and triplets claimed tobe synonyms. There is little doubt that this points to a variety of substrata.Kawi (1991:11-2) asserts that nearly all speakers of Dayak languages(which constitute the 'other' languages of Kalimantan aside from varietiesof Malay) in the provinces of South, Central and East Kalimantan are ableto communicate in Banjarese, which is why Banjarese is often usedalongside Indonesian in inter-ethnic communication. Soetoso et al.(1989:1) describe how the 4,000 people who constituted the Paku tribe in1987 had practically stopped using the Paku language as a result of theirfrequent contacts with speakers of Maanyan, Lawangan and Banjarese.Even at home, the majority of Paku people use Maanyan or Banjarese.Children no longer learn the Paku language.

Nanang et al. (1988:1) observe that the Bayan Dayak are always readyto accommodate visitors, even in their linguistic behaviour. Consequentlythey often speak Banjarese, Bakumpai or Indonesian. Furthermore, theauthors make the important observation that there are also Bayan Dayakcommunities that have openly given up the Bayan language because oftheir conversion to Islam.

Durasid (1990:41-2) remarks that the speakers of Malayic languages inKalimantan have been at an advantage for a very long time - during thesultanates of Pontianak, Kutai and Banjarmasin, under Dutch colonial rule,and in the Indonesian Republic. Politically, economically, and culturally,the coastal Malayic people were more advanced than the inhabitants ofthe interior. For speakers of the Barito languages the Malayic languages(Iban, Kutai Malay, and Banjarese) have been languages of prestige. Fordaily communication some speakers of a Barito language therefore preferto use a Malayic language (Durasid 1990:42).

Factors causing shifts and changes

Language death is a popular topic in linguistics today. Like rape andincest it has always been there, but only now are we aware of it. What isdifferent today is that languages becoming extinct are not offset by the


772 Hein Steinhauer

rise of new ones; these days dialects no longer develop into languages.Also new is the worldwide character and speed of language death.

The general scenario of gradual language death is always the same:external pressure, subsequent changes in language behaviour, and finallyinterruption of language transmission to a new generation, resulting insemi-speakers who have only a defective command of the language (seeSasse 1992a, 1992b). The details differ from case to case; what deservessystematic study is the relationship between the types of external pressureand the changes in language behaviour they induce, as well as theresulting patterns of decay.

Most external factors capable of initiating gradual language death havebeen mentioned in passing above (see also Steinhauer 1993). Some ofthem are typically Indonesian, while others are more general. The particularcombination is always community specific.

War, partial genocide, and acts of God may put a language communityat such a disadvantage vis-a-vis its neighbours that it succumbslinguistically in the long run. This is what happened to the Paulohi andLoun language communities in the Moluccas. Complete genocide or actsof God with the same effect (as in the case of the Tambora language) causeimmediate, rather than gradual, language death, which from a linguisticpoint of view is a deplorable loss, but not interesting as a process.

War by peaceful means, such as eviction of people from their lands,erodes the prosperity of the people as a community, and inevitably affectstheir linguistic behaviour. This is what seems to be happening inKalimantan, where logging concessions and transmigration projects havegreatly reduced the traditional territory of the Dayak tribes. On 24 January1994 the newspaper Kompas reported on a Dayak village which wasincluded in a logging concession, so that the people were forbidden to cutdown the trees they had planted in their own gardens and were forced touse their front doors for making coffins.

In general, contact with outsiders has linguistic consequences. The moreso, if these outsiders are economically or culturally superior, or areperceived as such. This explains part of the success of Banjarese. Tradersand government officials who visit the interior of South and CentralKalimantan are mostly Banjarese. Moreover, they are Muslims. Dayakswho give up their tribal religion and convert to Islam appear toconsciously abandon their own language and to shift to Banjarese as asign of total conversion (as in the Bayan Dayak case mentioned above).

Increased mobility as an aspect of modernization has its effects on thelanguage behaviour of individuals. This is the background of the statisticaldisappearance of many Minangkabau who move out of West Sumatra.Whole communities are affected by being resettled closer to other groups(as in the case of the Waropen). They are also affected when they have toshare their traditional lands with other people (loggers, transmigrants,tourists).



On the national level, a very important factor that diminishes theprestige of regional languages is the prevalence of diglossia, wherebyIndonesian functions as the national, supra-ethnic, official language,whereas the regional languages are used for unofficial intra-ethniccommunication and local cultural events. Mass education and masscommunication, along with the omnipresence of government institutions,representatives, and regulations, as well as of religious institutions (giventhe obligation for an Indonesian citizen to adhere to an officiallyrecognized (world) religion), have created a multitude of domains in whichIndonesian is the only appropriate means of communication. Domains inwhich the regional languages are or were appropriate have been takenover by Indonesian, or have become of secondary importance. 'Progresscannot be stopped.' Modernization implies cultural genocide.

But this should not be viewed as entirely negative. In a world in whichthe ideals of ethnic cleansing have a greater rallying power thantraditional-style political programmes, the threat of ethnic strife is never faroff. In Indonesia, the Indonesian language is one of the major instrumentsof national and supra-ethnic unity, and a major counterforce againstthreats of ethnic divisiveness. A very important factor contributing to thesuccess of Indonesian is the fact that it has never been the language of aspecific dominant group. Neither can it be stigmatized as the language of aculturally or economically identifiable section of the population. Moreover,as Indonesian is the language of most mass media and institutionalizedinstruction, knowledge of it is a precondition for personal development.Consequently, widespread knowledge of Indonesian is a precondition fornational development. It is official policy, therefore, that Indonesian beknown by every Indonesian.

This state of affairs is the reason why parents of the same linguisticbackground decide not to transmit their own language to their children. Inorder 'not to put them at an educational disadvantage' they useIndonesian (or, rather, their own variety of Indonesian) instead. This is afrequent phenomenon, especially in urban communities, where ties withtraditional culture are weak and career perspectives dependent ondiplomas.

On the basis of the figures in Table 3, it can be predicted that if theincrease of knowledge of Indonesian is not checked - and there is noreason to assume that it will be - then within two generations allIndonesians will know Indonesian. It should be stressed again, though,that 'knowledge of a language' is an elastic notion. And with regard toIndonesian it may indeed be questioned whether all Indonesians willmaster the one and only 'good and correct Indonesian' (bahasaIndonesia yang baik dan benar). The Pusat Pembinaan danPengembangan Bahasa (National Center for Language Development andCultivation), which defines, monitors, and propagates this type ofIndonesian, still has a long way to go.


774 Hein Steinhauer

The possible role of linguistics

With regard to Indonesia the following questions should be answered bylinguists: what is the linguistic situation now, how did it come about, andwhat will it turn into?

In broad outlines the situation can now be described as a mosaic ofabout 500 different languages, most of them constituting chains of dialects(or continua), many of them with sociolectic variation, and with variantsfor special purposes (such as rituals, oral and written literary performances),and some of them with an ancient written tradition. Genetically theybelong to unrelated families; languages of the larger families (Austronesianand the Trans New Guinea Phylum) belong to widely differing subfamilies,with considerable typological variation.

Each language, or rather each variety of language (each dialect, eachsociolect) has its own lexicon and structure, and should be described assuch. Needless to say, such a description should not be made out of theblue, but against the background of what is known about languagestructure in general, while phenomena that are more language specific mayalways be described with reference to comparable phenomena in otherbetter-known languages, especially if they are genetically related orgeographically close.

Each language is used in its own way. One language community mayhighly value understatements and oblique remarks, while another relies onoverstatement to get an intention across. In one community a loud voice isappreciated as inspiring confidence; in another, anything else thansubduedness is thought to be inappropriate. Inter-ethnic communication (ajob interview, for instance) may be seriously hampered by unawareness ofdifferences in language use. Grammar is one thing; the proper way toconduct a conversation is quite another. Each language requires informa-tion on both grammar and usage?19

How the Indonesian linguistic situation will evolve depends on the sumtotal of language ecologies. For each language community it should bedetermined whether it is isolated or in close contact with other languages.It should be determined whether the nature of the community is rural orurbanized, traditional or modern, open towards other communities orclosed off, allowing free communication between its members or havingrestraints and restrictions. Furthermore, it should be determined what theposition of the language is vis-a-vis Indonesian, any local lingue franche,and neighbouring languages. Do all sections of the language communityuse the language in the same contexts? What are people's attitudestowards their language and other languages with which they come intocontact? Is the language transmitted to new generations? Answers to

19 An excellent illustration of this principle and its practical elaboration is Bird andShopen 1979.



these and similar questions will give some indication of what the future ofthe language will be. Other decisive factors are education and languagepolicy, as well as international political, economic, and environmentaldevelopments.

The present mosaic of languages arose by historical developmentswhich can only be reconstructed, to a large extent by purely linguisticmeans. Needless to say, such a reconstruction is a legitimate aim in itself.But it may also reveal unexpected relationships which otherwise wouldnot have been noticed, such as K.A. Adelaar's finding that the Tamaniclanguages in the interior of West Kalimantan (Kapuas Hulu District) areclosely related to South Sulawesi languages, especially Buginese andSouth Toraja (Adelaar 1994). Insight into the history of development of aless transparently structured language may also be extremely helpful forunderstanding its structure. A case in point is Noorduyn's fascinatingreconstruction of the history of proto-Gorontalo by comparing Gorontalowith Bolaang-Mongondow, thus very much reducing the highly opaquecharacter of the morphology of the former (Noorduyn 1982).

A prerequisite for language reconstruction is of course a fair knowledgeof the languages now in use. Because the knowledge of the vast majorityof Austronesian languages in Indonesia is still very imperfect, comparativehistorical reconstructions have been largely confined to sound changesand lexical reconstructions. But the conviction is gaining ground that 'theincreased refinement of the reconstructions of proto-Austronesian nowdemands a close scrutiny of its subgroups so as to proceed with greaterconfidence and effect' (Collins 1983:3). Reconstruction of languagehistory comprises not only the reconstruction of the phonemes of theearliest proto-language, but also of all changes (as much as possible in theirchronological order) which subsequently transformed that proto-languageinto the daughter languages. Where sufficient data are available - and onthe subgroup level this is more often the case - the search for soundchanges and proto-phonemes should be supplemented with comparativehistorical work in the fields of morphology and syntax. In other words, areconstructed proto-language should resemble real languages. Not onlyshould it have a grammar, but its phonemes should constitute aconceivable system, instead of - as is often the case in current practice -being a mere list of labels, each for a different string of soundcorrespondences in the languages under consideration.

Finally, concerted action should be taken on Anceaux's admonitionthat 'men de Papoea-talen bij de Austronesische studien [...] niet [kan]missen' ('one cannot do without the Papuan languages in Austronesianstudies'; Anceaux 1971:18), if only because New Guinea is the keystone ofOceanic linguistics (Anceaux 1953). Needless to add, it is the keystone ofnon-Austronesian linguistics too.

No linguistic change or relation can be described without knowingwhat it is that is changing or related. In other words, the keystone for all


776 Hein Steinhauer

other linguistic research is synchronic description. This conclusion,although it may seem rather obvious to non-linguists, will probably not beshared by those generative linguists who have adopted the view that it isthe main task of linguistics to explain why human infants are able to learnthe language of their environment, whether this is a polysyntheticlanguage such as Kwakw'ala (Anderson 1985:24-34) or an isolating tonelanguage such as Vietnamese. For a variety of non-linguistic reasons thisview has been fashionable since the early 1960s. The fashionableexplanation has been that there must be one innate universal grammar ofwhich all observed human languages are variations or realizations. Whatdistinguishes languages can therefore only be secondary (which is whythe universal grammar theory was hailed by Christian fundamentalists: itseemed to confirm the historicity of the biblical story of the confusion oftongues at the Tower of Babel). Being secondary, these distinctions wereconsidered less interesting, to the extent that observable linguistic factswere discarded as irrelevant.20 For adherents of this view, study of theIndonesian linguistic scene would be pointless. For other linguists, thefollowing observations hardly need repeating.

The main function of human language has always been that it enablesits users to get ideas across. Any human language, as long as it is still alive,has that capacity, because it is a conventionalized system of signs. Thisimplies that the form and meaning of these signs are invariant. Otherwise,language would be chaos. However, this principle of invariance does notexclude change and variation; in fact, without invariance there would notbe any change and variation.

In actual communication this invariant meaning of signs is instrumental(and not more than that) in conveying the speaker's intent, and again inthe interpretation thereof by the hearer. Speaker's intent, the referent ofthe utterance (in the real or imaginary world), and hearer's interpretationshould therefore be sharply distinguished from meaning.

Language, being defined as a system of signs, cannot be describedformally without reference to meaning. Any formal element which can bedistinguished in a language (whether phonemic, morphemic, syntactic orintonational) must have a semantic function. Any set of formally opposedelements must differ in semantic function. The importance of semanticanalysis cannot be stressed enough: too many studies on individualIndonesian languages (including Indonesian) overemphasize form andinventories of forms, discussing their meanings only in very broad terms,such as 'reciprocal', 'denominalizing', or even simply 'different from themeaning of the form X'. Even the second edition of the standard grammar

20 Illustrative in this connection is what the founder of this approach, NoamChomsky, says about linguistic facts: 'Even the absolute true grammar would, if itexisted, be refuted by masses of counterexamples [...] because we do not know what isthe right kind of evidence' (Chomsky 1978:10).



of Indonesian (Alwi et al. 1993) is not free of such descriptions.As implied in the previous paragraph, intonation is an essential aspect of

language: no utterance exists without it, and 'speaking with an accent' isalways at least partly a matter of using foreign intonation patterns.Paradigmatically, intonational oppositions tend to distinguish question,statement, member of an enumeration, and possibly other sentence types.Syntactically, intonation functions alongside word order as a means ofarranging words and word groups with respect to each other, its semanticcorrelate being a specific structuring of information. This syntacticfunction of intonation is rarely discussed in Indonesian linguistics.21 In thisconnection it should be stressed again that interpretation is not the same asmeaning: what structures the information of sentence X is not its context(a popular view in discourse analysis), but formal elements of X itself, suchas its word order and intonation.

The necessary complement to the study of the grammar of a language isthe description of its lexicon. Much work is still to be done in the field oflexicology, notably for Indonesian. Being the language of nationaldevelopment and of formal education on all levels, as well as the supra-ethnic access to regional Indonesian cultures, Indonesian covers thewidest range of domains. Ever since Indonesian became the officiallanguage (at the time of the Japanese occupation), its lexicon has beendeveloping in accordance with its new function. Vocabularies anddictionaries are the means for broadcasting these developments andmaking them common property. However, a prerequisite for success in thisrespect is that the definitions of the lexical items be as explicit as possible.In the latest edition of the monolingual standard dictionary (Tim Penyusun1993), many lexical entries are semantically described by a synonym only.Not only does this practice obscure the shades of meaning which in realitydistinguish these 'synonyms', it also creates vicious circles: strings of'synonyms' which begin and end with the same entry. Another handicapfor the lexicography of Indonesian is that the structure of the Indonesianlexicon has never been made explicit, that is, the systematic relationshipsbetween the various derivations of individual lexical roots or classes oflexical roots have never been subjected to comprehensive study. As aconsequence, entries which belong to the same derivational category arenot always explained semantically in the same way. These flaws inlexicographical practice are largely the consequence of the rapid extensionof the Indonesian lexicon; only when the language stabilizes will itslexicographers find time for lexicological study and contemplation.Needless to say, lexicographers of the regional languages should avoid theabove pitfalls from the start.

21 A welcome exception is Suparno 1993.


778 Hein Steinhauer

Conclusion

With its extreme linguistic diversity and unique dynamics, Indonesia is aMecca for linguists and linguistics. For the same reason (the field being toobroad), it is in reality rather a backwater, where individual linguists canroam for years without meeting a kindred soul. For the majority of linguistsIndonesia is terra incognita, and its linguistic treasures are unknown. Forthose who do know the treasures (Indonesian linguists included), there isalways the matter of money and other obligations. Indonesia does have agrowing core of trained linguists,22 but they are still outnumbered by thelanguages they have to deal with, not to mention by the research tasksawaiting them. Moreover, their opportunity to do research is largelyconfined to short-term projects on a low budget. And above all, their firstobligation is teaching, whether it is Indonesian or a foreign language.

Given the enormous amount of work to be done, the large number oflanguages to cope with, and the urgency of the tasks, the internationallinguistic community can hardly afford to stand aside and wait forsomething to happen. Concerted action should be taken. Time is runningout. What is needed is a 'Visit Indonesia Decade' campaign for linguists,with active support from academic institutions and organizations both inIndonesia and abroad, resulting in a special linguistic library or at least inan encyclopaedia of Indonesian languages.

May the coming 150 volumes of Bijdragen be at the forefront of thesedevelopments, dedicated as the journal always has been to the under-standing of 'taal' (language), rather than to universal grammatical 'kunde'(sophistication).

22 Thanks to structured academic training at a number of Indonesian universities, andto joint programmes such as the Indonesian Linguistics Development Project (ILDEPI, 1977-1985, ILDEP II, 1988-1992).

REFERENCES

Adelaar, K.A., 1994, 'The classification of the Tamanic languages', in: Tom Duttonand Darrell T. Tryon (eds), Language contact and change in the Austronesianworld, pp. 1-42, Berlin/New York: Mouton de Gruyter. [Trends in Linguistics;Studies and Monographs 77.]

Akun Danie, Julianus, 1991, Kajian geografi dialek di Minahasa Timur haul, Jakarta:Balai Pustaka. [Seri ILDEP.]

Alwi, Hasan, et al., 1993, Tata bahasa baku Bahasa Indonesia, Edisi kedua, Jakarta:Departemen Pendidikan dan Kebudayaan Republik Indonesia.

Anceaux, J.C., 1953, 'New Guinea: keystone of Oceanic linguistics', Bijdragen tot deTaal-, Land- en Volkenkunde 109:289-99.

-, 1971, Indonesia en Oceanie; Een taalkundige terreinverkenning, 's-Gravenhage:Mouton.



Anderson, Stephen R., 1985, 'Typological distinctions in word formation', in:Timothy Shopen (ed.), Language typology and syntactic description, Vol. 3:Grammatical categories and the lexicon, pp.3-56, Cambridge: Cambridge Uni-versity Press.

Bird, Charles and Timothy Shopen, 1979, 'Maninka', in: Timothy Shopen (ed.),Languages and their speakers, pp. 59-111, Cambridge, Mass.: Winthrop.

Biro Pusat Statistik, 1974, Sensus Penduduk 1971; 1971 Population Census, Jakarta:Biro Pusat Statistik.

-, 1983, Penduduk Indonesia; Population of Indonesia, Hasil Sensus Penduduk1980; Results of the 1980 Population Census, Seri/Series: S Nomor/Number 2,Jakarta: Biro Pusat Statistik.

-, 1992, Penduduk Indonesia; Population of Indonesia, Hasil Sensus Penduduk1990, Results of the 1990 Population Census, Seri/Series: S 2, Jakarta: Biro PusatStatistik.

Brenzinger, Matthias (ed.), 1992, Factual and theoretical explorations with specialreference to East Africa, Berlin/New York: Mouton de Gruyter.

Collins, James T., 1983, The historical relationships of the languages of CentralMaluku, Indonesia, Canberra: Australian National University, Department ofLinguistics, Research School of Pacific Studies. [Pacific Linguistics D-47.]

Darmansyah H. Gudai, 1985, A grammar of Maanyan; A language of CentralKalimantan. [PhD thesis, Australian National University, Canberra.]

Durasid, Durdje, 1990, Rekonstruksi Protobahasa Barito. [PhD thesis, UniversitasIndonesia, Jakarta.]

Grimes, Barbara D. (ed.), 1988, Ethnologue; Languages of the world, Eleventhedition, Dallas: Summer Institute of Linguistics.

Grimes, Charles E., 1991, The Bum language of eastern Indonesia. [PhD thesis,Australian National University, Canberra.]

Grimes, Charles E. and Barbara D. Grimes, 1987, Languages of South Sulawesi,Canberra: Australian National University, Department of Linguistics, ResearchSchool of Pacific Studies. [Pacific Linguistics D-78.]

Kawi, Djantera, 1991, Bahasa Banjar; Dialek dan subdialeknya. [PhD thesis, Uni-versitas Indonesia, Jakarta.]

Keraf, Gregorius, 1978, Morfologi dialek Lamalera, Ende: Arnoldus. [PhD thesis,Universitas Indonesia, Jakarta.]

Nanang, Yohanes Kalamper and Moses Usman, 1988, Morfologi dan sintaksis bahasaBayan, Jakarta: Departemen Pendidikan dan Kebudayaan.

Noorduyn, J., 1982, 'Sound changes in the Gorontalo language', in: Amran Halim,Lois Carrington and S.A. Wurm (eds), Papers from the Third International Con-ference on Austronesian Luinguistics, Vol. 2. Tracking the Travellers, pp. 241-61,Canberra: Australian National University, Department of Linguistics, ResearchSchool of Pacific Studies. [Pacific Linguistics C-75.]

Raffles, Th.S., 1978, The history of Java, with an introduction by John Bastin, Oxford:Oxford University Press. [First edition 1817.]

Sasse, Hans-Jiirgen, 1992a, 'Theory of language death', in: Brenzinger (ed.) 1992,pp. 7-30.

-, 1992b, 'Language decay and contact-induced change: Similarities and differ-ences', in: Brenzinger (ed.) 1992, pp. 59-80.

Silzer, Peter J. and Helja Haikkinen, 1984, 'Index of Irian Jaya Languages', Irian:Bulletin of Irian Jaya 12:i-v, 1-125.

Sodaqoh Zainudin, A. Kusuma Soegianto and Barijati, 1978, Bahasa Madura,Jakarta: Pusat Pembinaan dan Pengembangan Bahasa, Departemen Pendidikandan Kebudayaan.

Soetoso, Dewi Mulyani, et al., 1989, Fonologi Bahasa Paku, Jakarta: DepartemenPendidikan dan Kebudayaan.

Steinhauer, H., 1986, 'Austronesian geographical prospects', Bijdragen tot de Taal-,Land- en Volkenkunde 142:296-313.

-, 1993, 'The Indonesian linguistic scene; 500 languages now, 50 in the next cen-tury?', in: Sudaporn Luksaneeyanawin (ed.), Pan-Asiatic linguistics; Proceedingsof the Third International Symposium on Language and Linguistics, pp. 1462-77,Bangkok: Chulalongkorn University Printing House.


780 Hein Steinhauer

Stokhof, W.A.L., 1975, Preliminary notes on the Alor and Pantar languages, Can-berra: Australian National University, Department of Linguistics, Research Schoolof Pacific Studies. [Pacific Linguistics B-43.]

-, (ed.), 1988, Holle lists: vocabularies in languages of Indonesia, Vol. 11. Celebes,Alor, Ambon, Irian Jaya, Madura and Lombok, Canberra: Australian NationalUniversity, Department of Linguistics, Research School of Pacific Studies. [PacificLinguistics D-81.]

Stresemann, E., 1918, Die Paulohisprache; Ein Beitrag zur Kenntnis der amboi-nischen Sprachengruppe, 's-Gravenhage: Nijhoff.

Suparno, 1993, Konstruksi tema rema dalam Bahasa Indonesia lisan tidak resmimasyarakat Kotamadya Malang, Jakarta: Pusat Pembinaan dan PengembanganBahasa. [Also PhD thesis, Universitas Indonesia, 1991.]

Tim Penyusun Kamus Pusat Pembinaan dan Pengembangan Bahasa, 1993, KamusBesar Bahasa Indonesia. Second impression, Jakarta: Balai Pustaka.

Usop, KMA M., et al., 1992, Fonologi Bahasa Lamandau, Jakarta: Pusat Pembinaandan Pengembangan Bahasa.

Wurm, S.A., and Shiro Hattori (eds), 1981, Language atlas of the Pacific area, Part I,New Guinea area, Oceania, Australia, Canberra: Australian National University,Department of Linguistics, Research School of Pacific Studies. [Pacific LinguisticsC-66.]

-, 1983, Language atlas of the Pacific area, Part II, Japan area, Philippines andFormosa, mainland and insular South-East Asia, Canberra: Australian NationalUniversity, Department of Linguistics, Research School of Pacific Studies. [PacificLinguistics C-67.]

Zollinger, H., 1850, 'Verslag van eene Reis naar Bima en Soembawa, en naar enigeplaatsen op Celebes, Saleijer en Flores, gedurende de maanden Mei tot December1847', Verhandelingen Bataviaasch Genootschap 23-4:5-224.



Appendix

Table 4. First-language speakers of the regional languages of Indonesia, in absolute numbers andas percentages

Agegroup

5-9 UR

U+R%U

%R%U+R

10-14UR

U+R%U%R

%U+R

15-24UR

U+R%U%R

%U+R

25-49UR

U+R%U%R

%U+R

50+ UR

U+R

%U%R

%U+R

Total UR

U+R%U%R

%U+R

Javanese1980

1,320,0336,897,2248,217,257

31.2%41.0%39.0%

1,245,6445,899,0357,144,679

31.9%43.4%40.9%

2,443,2988,985,172

11,428,47032.0%44.0%40.7%

3,065,23413,938,13317,003,367

33.1%43.5%41.2%

1,337,9836,311,9257,649,908

41.4%46.9%45.8%

9,412,19242,031,48951,443,681

33.3%43.6%41.3%

1990

1,929,4906,218,9318,148,421

29.9%37.8%35.6%

1,949,9455,903,7497,853,694

30.9%39.1%36.7%

3,903,5718.947,967

12,851,53830.5%40.4%36.8%

5,594,35515,553,69121,148,046

31.6%41.1%38.1%

2,540,6087,723,521

10,264,12941.7%46.5%45.2%

15,917,96944,347,85960,265,828

32.2%41.9%38.8%

Sundanese1980

588,6502,734,9353,323,585

13.9%16.3%15.8%

541,0382,141,2792,682,317

13.9%15.8%15.3%

970,7263,074,7854,045,511

12.7%15.0%14.4%

1,179,7215,061,2256,240,946

12.7%15.89b-15.1%

444,6492,034,5062,479,155

13.8%15.1%14.9%

3,724,78415,046,73018,771,514

13.2%15.6%15.1%

1990

890,7372,760,8833,651,620

13.8%16.8%15.9%

880,5852,501,2493,381,834

14.0%16.6%15.8%

1,619,8863,619,4675,239,353

12.7%16.3%15.0%

2,169,1506,214,9888,384,138

12.2%16.4%15.1%

836,5892,661,7423,498,331

13.8%16.0%15.4%

6,396,94717,758,32924,155,276

13.0%16.8%15.6%

Madurese1980

89,267862,088951,355

2.1%5.1%4.5%

.74,519573,245647,764

1.9%4.2%3.7%

145,375971,748

1,117,1231.9%4.8%4.0%

259,8692,131,0152,390,884

2.8%6.6%5.8%

98,040828,497926,537

3.0%6.2%5.6%

667,0705,366,5936,033,663

2.4%5.6%4.9%

1990

104,888781,149886.037

1.6%4.7%3.9%

104,081654,353758,434

1.6%4.3%3.5%

210,2471,063,8321,274,079

1.6%4.8%3.6%

368,9632,335,5262,704,489

2.1%6.2%4.9%

156,6641,012,5651,169,229

2.6%6.1%5.1%

944,8435,847,4256,792,268

1.9%5.5%4.4%


782 Hein Steinhauer

Agegroup

5-9 UR

U+R%U

%R%U+R

10-14UR

U+R%U

%R%U+R

15-24UR

U+R%U

%R%U+R

25-49UR

U+R%U%R

%U+R

50+ UR

U+R%U%R

%U+R

Total UR

U+R%U

%R%U+R

1980

42,957451,976494,933

1.0%2.7%2.4%

42,681391,967434,648

2.0%2.9%2.5%

85,804501,164586,968

1.1%2.5%2.1%

90,993648,899739,892

1.0%2.0%1.8%

28,797305,470334,267

0.9%2.3%2.0%

291,2322,299,4762,590,708

1.0%2.4%2.1%

Batak1990

52,442482,268534,710

0.8%2.9%2.3%

51,409457,698509,107

0.8%3.0%2.4%

118,220543,587661,807

0.9%2.5%1.9%

157,724825,850983,574

0.9%2.2%1.8%

55,456375,271430,727

0.9%2.3%1.9%

435,2512,684,6743,119,925

0.9%2.5%2.0%

Minangkabau1980

76,581442,518519,099

1.8%2.6%2.5%

70,992363,386424,378

1.9%2.7%2.5%

161,453522,815684,268

2.1%2.6%2.4%

158,265734,247892,512

1.7%2.3%2.2%

57,414414,234471,648

1.8%3.1%2.8%

524,7052,477,2003,001,905

1.9%2.6%2.4%

1990

119,463417,851537,314

1.8%2.5%2.3%

117,443403,947421,390

1.9%2.7%2.4%

254,478515,295769,773

2.0%2.3%2.2%

321,573823,498

1,145,0711.8%2.2%2.1%

113,801440,250554,051

1.9%2.6%2.4%

926,7582.600,8413,527,599

1.9%2.5%2.3%

Balinese1980

39,135324,010363,145

0.9%1.9%1.7%

39,263257,845297,108

1.0%1.9%1.7%

74,393384,263458,656

1.0%1.9%1.6%

85,806620,633706,439

0.9%1.9%1.7%

36,912301,344338,256

1.1%2.2%2.0%

275,5091,888,0952,162,604

1.0%2.0%

1.8%

1990

55,608249,137304,745

0.9%1.5%1.3%

63,458259,871323,329

1.0%1.7%1.5%

154,916433,229588,145

1.2%2.0%1.7%

199,753745,203944,956

1.1%2.0%1.7%

73,832354,234428,066

1.2%2.1%1.9%

547,5672,041,6742,589,241

1.1%1.9%

1.7%



Agegroup

Buginees1980 1990 1980

Banjarese1990

5-9 UR

U+R

%U%R

%U+R

10-14UR

U+R%U

%R%U+R

15-24U

RU+R%U

%R

%U+R

25-49UR

U+R%U%R

%U+R

50+ UR

U+R%U%R

_ %U+R

TotalUR

U+R%U%R

%U+R

49,204481,651530,855

1.2%2.9%2.5%

45,008360,347405,355

1.2%2.7%2.3%

87,164485,417572,581

1.1%2.4%2.0%

113,679857,092970,771

1.2%2.7%2.3%

40,846335,147375,993

1.3%2.5%2.3%

335,9012,519,6542,855,555

1.2%2.6%2.3%

57,639398,853456,492

0.9%2.4%2.0%

62,629383,687446,316

1.0%2.6%2.1%

131,081562,046693,127

1.0%2.5%2.0%

189,113970,350

1,159,4631.1%2.6%2.1%

73,949399,216473,165

1.2%2.4%2.1%

514,4112,714,1523,228,563

1.0%2.6%2.1%

56,034202,525258,559

1.3%1.2%1.2%

45,559148,938194,497

1.2%1.1%1.1%

84,146232,639316,785

1.1%1.1%1.1%

103,480371,671475,151

1.1%1.2%1.2%

34,056140,356174,412

1.1%1.0%1.0%

323,2751,096,1291,419,404

1.1%1.1%1.1%

99,314320,850420,146

1.5%2.0%1.8%

103,454297,351400,805

1.6%2.0%1.9%

196,042433,987630,029

1.5%2.0%1.8%

271,405704,639976,044

1.5%1.9%1.8%

80,486247,733328,219

1.3%1.5%1.4%

750,7012,004,5602,755,261

1.5%1.9%1.8%


784 Hein Steinhauer

Othergroup 1980 1990 1980

Indonesian1990

5-9 UR

U+R

%U%R

%U+R

10-14UR

U+R

%U%R

%U+R

15-24UR

U+R

%U%R

%U+R

25-49UR

U+R%U

%R%U+R

50+ UR

U+R%U

%R%U+R

TotalUR

U+R%U%R

%U+R

409,8473,539,1503,948,997

9.7%21.0%18.8%

370,9402,730,7263,101,666

9.5%20.1%17.7%

707,1314,059,8714,767,004

9.3%19.9%17.0%

865,0986,160,2727,025,370

9.3%19.2%17.0%

315,1342,382,5542,697,688

9.8%17.7%16 2%

2,668,15218,872,57321,540,725

9.4%19.6%17.3%

645,2483,807,3384,452,586

10.0%23.1%19.4%

633,0413,380,6594,013,700

10.0%22.4%18.7%

1,185,7024,674,1265,859,828

9 .3%21.1%16.8%

1,594,6497,639,6999,234,348

9.0%20.2%16.6%

579,3632,930,1403,509,503

9.5%17.6%15 4%

4,638,00322,431,96227,069,965

9.4%21.2%17.4%

1.554.549891.460

2,446,00936.8%5.3%

11.6%

1.427.068717.447

2,144,51536.6%5.3%

12.3%

2.884.0251.219.2704,103,295

37.7%6.0%

14.6%

3.341.5501.532.8974,874,447

36.1%4.8%

11.8%

837.350406.018

1,243,36825.9%

3.0%74%

10,044,5424,767,092

14,811,63435.5%4.9%

11.9%

2.506.3971.014.7083,521,105

38.7%6.2%

15.4%

2.346.120870.111

3,216,23137.2%

5.8%15.0%

5.813.7971.361.3656,375,162

39 .2%6 . 1 %

18.2%

6.847.6452.013.5968,861,241

38.7%5 .3%

16.0%

1.589.207478.628

2,067,83526.1%

2.9%9.1%

18,303,1665,738,408

24,041,57437 .1%

5.4%15.5%

U: urban population, R: rural population, U+R: urban plus rural population, %U: percentage ofthe urban population of that age group, %R: percentage of the rural population of that age group,%U+R: percentage of the total population of that age groupBold type indicates an increase in percentage of speakers, Italic type indicates a decrease innumber of speakers


e ndonesian lanae siaion and linisics prosecs and

Documents