Background of the Indonesian language

In 1928 a congress of nationalist youth in Batavia pledged to ‘uphold the language of unity, the Indonesian language’, but such a language did not yet exist.Footnote ⁶ Instead, students were making this pledge in what was (up until the moment of their oath) a popular dialect of the language indigenous to the Strait of Malacca and used more broadly for trade, Dutch-run indigenous schools, and religious teaching. This dialect, called school Malay, had also been ‘the language most associated with Islam and its dissemination throughout the lands below the winds’, and its use throughout the archipelago ‘represented a continuation of the tradition of Islamic scholarship in the region’.Footnote ⁷ School Malay was very much tertiary in the Dutch colony, though, behind Dutch as the official language of administration and local languages as popular vernaculars.

This changed very quickly after 1942. When the Japanese conquered the archipelago, they required that Indonesian be used to replace Dutch in government administration. In 1945, Indonesians proclaimed an independent state, with Indonesian as the official language. Both of these changes prompted a furious pace of linguistic reform. The editorial of the inaugural issue in July 1948 of a new magazine called Pembina Bahasa Indonesia (Builder of the Indonesian Language) stated that ‘Among the changes that are clearly visible, that have been brought by the last six years, no one could have expected the progress of the Indonesian language.’Footnote ⁸ Commentators in the magazine believed that Indonesian could become ‘the best language in the world’ if it were guided well enough to become a thoroughly modern language.Footnote ⁹ Indonesian linguists saw their opportunity to define Indonesian for an expectant nation as ‘the broadest opportunity to use the results of modern linguistics and the experience of languages in other countries to make our language the best possible tool of society and tool of human thought in scientific fields, etc.’Footnote ¹⁰

This scramble to become modern and linguistically perfect, though, had unforeseen consequences for the country's strong Islamic community. Whereas Malay had been associated with its past as a trans-ethnic language for communication of theological ideas between religious leaders, after independence technocrats, scholars, and politicians intentionally developed Indonesian for the communication of policy and scientific ideas, and in so doing unintentionally alienated pious Muslims and restricted their ability to function in the new linguistic landscape.

Disappearance of Jawi script

At the turn of the twentieth century, one could write Malay in two ways: the Roman alphabet or the Arabic script. Malay written in Arabic script is called jawi, a derivative of the name used in Arabic to refer to Southeast Asia.Footnote ¹¹ Literacy in jawi was common among the cosmopolitan Muslim elite around the Strait of Malacca, both men and women, in the nineteenth century.Footnote ¹²

While based on Arabic script, jawi included several additional letters for the precise recording of distinct Indonesian consonants. In this way, jawi was the most complete system for writing Malay consonants, with every consonant sound represented by its own letter, thus meeting the gold standard of ‘one-letter-one-sound’ in the design of orthographic systems. In contrast to consonants, however, the jawi script for writing Malay did not transcribe short vowels, offering the option to vary the vocalisation of spoken words to fit a range of local dialects. The lack of vocalisation meant that jawi required greater personal initiative for reading the text, yet at the same time it provided the opportunity for greater mutual intelligibility between regions. For example, the standard Malay word kambing (goat) and the Minangkabau variant kambiang would be written the same way, and this was useful for mutual understanding. At the same time, words with the same consonant structure but differing vowels, like bidak (pawn), bidik (aim), and biduk (riverboat), were written identically, which could cause ambiguity or lack of clarity.Footnote ¹³ The famous colonial linguist William ‘Oriental’ Jones disliked this indeterminacy and (possibly subversive) multi-interpretability, however, and so labelled the jawi system of writing as ‘degenerate’, foreshadowing the attitudes held by post-independence Indonesian experts.Footnote ¹⁴

In Indonesia both religious and secular groups employed the Arabic script for writing Malay up through the early twentieth century, although the Roman script for Malay was also in circulation. In fact, through the 1920s Muslim Indonesians were probably more likely to know jawi than romanised Indonesian, especially Muslim Indonesians of a lower class. A prominent Muslim politician of the 1950s described his best friend from his youth, a tailor's son, as not attending school but still coming to the mosque to learn to read and recite the Qur'an.Footnote ¹⁵ Presumably, there was a substantial population across Indonesia of Muslims like this tailor's son who had learned to read the Qur'an (and thus jawi script) in recitation circles, but who had never attended a formal school where they could be exposed to the Roman alphabet.

Even more competent in jawi were those Muslim youths who attended a religious primary or secondary school; they, too, would have had little exposure to the Roman alphabet.Footnote ¹⁶ These students and graduates could read the many journals pushing forward the religious debate, all written entirely in Arabic script.Footnote ¹⁷ This world of jawi literacy led one particularly vocal Indonesian Muslim activist, the Sumatran author and religious scholar Hamka, to argue in the 1950s that jawi was ubiquitous across the archipelago through the late nineteenth century, and only began to disappear after that because of the influence of the Dutch colonial power.Footnote ¹⁸ Yet the jawi script was used for instruction in the Indonesian language even in some Dutch-run schools, with the Roman alphabet reserved for Dutch and European languages. This system was in place in West Sumatra in the nineteenth century,Footnote ¹⁹ and remained in place in Banjarmasin until 1942.Footnote ²⁰

By the 1920s, romanised Indonesian was more common for the non-Muslim (or non-pious Muslim) community and in government affairs; however, the increasingly firm position of romanised Indonesian does not mean that jawi was displaced. Arabic script remained the dominant form among the Islamic community, even as Romanisation made inroads. The Muhammadiyah press in Yogyakarta was particularly active in printing religious materials in Indonesian and Javanese, for both of which it used the Arabic script,Footnote ²¹ but they also printed the organisation's 1926–27 almanac in romanised Indonesian.Footnote ²² Private libraries were still lending handwritten books in jawi through the 1920s.Footnote ²³ Although some minor works and periodicals may have circulated in Latin letters earlier, the first Islamic theological text to become popular in romanised Indonesian was Ahmad Hassan's al-Boerhan published in 1928, a mere seventeen years before Indonesian independence.Footnote ²⁴ Ahmad Hassan's romanised scholarship appealed to a new class of Islamic leaders: those trained by the Dutch at Western-language institutions who still wanted to explore their Muslim roots. These kinds of leaders, men like Agoes Salim, Kasman Singodimedjo, S.M. Kartosuwirjo, Abu Hanifah, and Mohammad Natsir, became very prominent in Islamic politics in the late 1930s and through the 1940s, and they were the first to bring a romanising impulse into Islamic political circles.Footnote ²⁵ This may have been the death knell of jawi as a system of writing. Soon after the Dutch-trained Islamic politicians gained prominence, reformist Islamic schools started teaching romanised Indonesian instead of jawi, and increasing numbers of scholars published Islamic works or taught in romanised script.Footnote ²⁶ Still, the colonial government retained jawi as a way to communicate with the religious community.

Thus, at the height of Dutch colonialism, the jawi script still had a distinct place: a separate (and, to Western minds, inferior) way to write, intended by the colonial power for relating to Indonesia's Muslim masses, but not for high scholarship. By the end of the Dutch period, even the community of Islamic leaders had begun a transition to romanised Indonesian for some scholarship.

Under Japanese colonialism, the Indonesian language took a more prominent role in public affairs, prompting serious linguistic choices. The Japanese mandated that all official business be conducted in Indonesian (Dutch was quickly and fully outlawed), and this forced large numbers of Dutch-educated Indonesians to study and use the Indonesian language for the first time. These Indonesians generally employed the Roman alphabet, solidifying its dominance.Footnote ²⁷ Still, jawi continued to hold its own as a separate but viable script. When the Japanese began to publish a biweekly magazine targeted at Indonesian Islamic scholars in March 1944, they chose to use the Arabic script for all three languages in which the magazine circulated (Indonesian, Sundanese, and Javanese).Footnote ²⁸ Thus, while jawi continued in both official and unofficial use under the Japanese, political Indonesians in the capital coalesced around a romanised standard for the language.

The complete disappearance of jawi as a viable alternative to romanised Indonesian occurred later but rather quickly, from roughly 1948 through 1956. Although jawi was never used by the central Indonesian government for administration, during the Revolution it remained more heavily used in certain regions, such as in Aceh.Footnote ²⁹ Furthermore, Dutch publishers must have thought that the Arabic script for Indonesian still had a strong future, as evidenced by the books they published in 1949 and 1950 to teach this script.Footnote ³⁰

The preference of the independent Indonesian government for romanised Indonesian effectively guaranteed the death of the jawi system of writing. Not only government laws and regulations, but also education in government schools, were entirely romanised. Religious schools became obliged to teach the Roman alphabet, while secular schools did not teach jawi, even at the highest level.Footnote ³¹

Signs of the disappearance of jawi as a script accepted in everyday life were particularly conspicuous on Indonesia's money. Coins had employed jawi since the earliest days of the Republic, again a sign that the lower classes were more likely to read the Arabic script than Roman letters; lower-class individuals were also more likely to handle coins than paper money. By contrast, paper money in independent Indonesia never incorporated jawi.Footnote ³² The images in Fig. 1 of the fifty-cent coins in 1952 and 1955 demonstrate visually the erasure of jawi. In 1952, the obverse of the fifty-cent coin shows a picture of a Javanese Muslim rebel prince, labelled with his name in romanised letters on the left (Dipanegara) and jawi () on the right. On the 1955 impression, the exact same image appears on the fifty-cent coin, but with the jawi text removed.

Figure 1.

Fifty-cent coins from 1952, on the top, and 1955, on the bottom. Note how the jawi name for Dipanegara appears to the right of the portrait on the earlier coin, but not the later one.

I am thankful to Dr Luke Treadwell and colleagues at the Ashmolean Museum, Oxford, for these images.

The erasure of jawi was not an explicit government policy, like forced romanisation in Turkish,Footnote ³³ but it was also a stronger mandate than comparable script reforms in China or Japan.Footnote ³⁴ Unlike the use or avoidance of certain characters, which allowed for these characters to still exist on the periphery of the mainstream writing system, the elimination of jawi from official use created a vacuum in which jawi, and by extension Arabic, could not influence the other processes of standardisation. This was also possible because jawi was not a local ethnic language and did not have the protection of certain mechanisms for the preservation of first languages.Footnote ³⁵

For a broad swath of local Muslim scholars and their followers, the disappearance of jawi as a legitimate script carried serious and very real consequences. This is a case of linguistic reform causing more than just shifts in meaning and symbolic power — the consequences included real shifts in power away from local Islamic leaders.

The basic mechanism of this change is quite straightforward: Muslims who were literate only in the jawi script became illiterate in the modern state. Certainly this was true for a large number of lower-class Indonesians, as suggested by the earlier example of the tailor's son, but for most of these working-class individuals literacy or illiteracy (in any script) did not significantly change their role in society. There was a much greater impact on Muslim leaders, many of whom were also rendered illiterate.Footnote ³⁶ These leaders can be broken down into three groups.

First, many Islamic leaders were (or became) competent in both jawi and romanised script, but they preferred jawi. The famous Muslim author and polemicist Hamka was fluent in both systems from a young age, but continued to use jawi for his private use and for correspondence with close friends as late as the 1970s.Footnote ³⁷ Other leaders had been unfamiliar with the Latin alphabet, but were able to learn it. Muhammad Zainuddin Abdul Madjid, the founder of a large Islamic organisation on Lombok, felt much more comfortable writing in jawi and conducted all his personal business in that script, but he took the time to become familiar with romanised Indonesian so he could participate in forums where it was used, such as the national Constitutional Assembly to which he was elected in 1955.Footnote ³⁸

Second, there was a class of Indonesian Muslims who remained illiterate in romanised Indonesian but engaged with that system of writing through proxies, usually students. This was the case for Sayyid Idrus al-Jufri, founder of East Indonesia's largest Islamic organisation, Alkhairaat, who could read and write only in Arabic script, but who used students as his scribes for romanised Indonesian.Footnote ³⁹ Anecdotal evidence suggests that this method was used by a large number of senior scholars across the archipelago.

Third, the most bitter choice for Islamic leaders was to continue to use the scripts with which they were familiar. They were thus rendered legally illiterate and excluded from mainstream and governmental discussion and debate, but they were able to maintain the written worlds in which they had functioned all their lives.

One man who took this third choice was Gurutta Haji Ahmad Bone, a well-respected leader in the Islamic community of South Sulawesi. During the Dutch colonial period, Ahmad Bone had founded a major regional organisation for Islamic scholars, making him arguably the foremost organiser in the traditionalist community on Sulawesi at the time.Footnote ⁴⁰ However, although Ahmad Bone was a venerated religious scholar and could read and write Arabic, school Malay (in jawi), and Buginese (in the local Indic lontara script), he could not write in Roman letters. For this reason, when in 1954 he wanted to weigh in on an issue of policy at the provincial level, he wrote a bilingual letter, duplicating all of his opinions in Indonesian (written in jawi) and Buginese (written in lontara) (Fig. 2).

Figure 2.

From these first few lines of Ahmad Bone's letter, one can see his use of jawi and lontara script. The fact that the same message is presented in both the Buginese and jawi versions implies that Ahmad Bone was searching for a way to convey his thoughts to the government in the absence of an ability to write in romanised Indonesian. Arsip Propinsi Sulawesi Selatan, Collection 11, Kantor Wilayah Departemen Agama Propinsi Sulawesi Selatan, Periode Tahun 1947–1990, #184.

I am grateful to Ahmad Taslim for his help in acquiring this image.

Because of the linguistic reforms undertaken with the standardisation of Indonesian, Ahmad Bone, who had been a leading Islamic figure during the Dutch period and held sway over thousands of traditionalist Muslims, was unable to participate fully in the policy discussion. In his province's debate on the subject of eliminating prostitution, the letter he wrote was disregarded by the governor because of Bone's inability to voice his opinion in Roman letters.Footnote ⁴¹

Just as it limited their participation in policy discussions, this bias in favour of romanised writing also would have disqualified ulama from governmental positions, such as district head, or even possibly the head of a local Ministry of Religion office. Historians of Indonesia have speculated on how the former bureaucratic and aristocratic elite from the Dutch period, who had been thrown out and replaced — often by religious leaders — during the Revolution, were able to recapture their positions in the 1950s.Footnote ⁴² Linguistic reforms must have been a contributing factor. The formalisation of certain linguistic reforms (in this case, the selection of romanised Indonesian as the official writing system) favoured those who had been trained in Dutch colonial schools and were fluent in Latin letters.

Two administrative moments from Indonesia help to demonstrate the distinct shift in government linguistic policy from the Revolution to the 1950s. During the Revolution, in the 1948 election law, indirect electors were chosen for every 250 people, with the requirement that the electors be literate. However, this literacy could be ‘in Latin, Arabic or some local alphabet’.Footnote ⁴³ In the 1950s, the selection of local executives came not through elections (and not with the recognition of several writing systems); instead, candidates were appointed to these positions by the national government, which favoured long-standing bureaucrats who would naturally be adept at romanised Indonesian.Footnote ⁴⁴ Thus, religious leaders lost access to local government positions because of their inability to read and write the new, romanised language of the state.

Religious leaders saw another concrete implication of jawi's gradual death: a widened gap between Indonesian and the Arabic language, leading to greater barriers for Indonesians to study Arabic or engage with classical Islamic texts. Had jawi become the dominant script for Indonesian, many Muslims believed that Arabic thought would have continued to exercise a great influence on the Indonesian public sphere. As it was, Indonesia became irrevocably aligned in modern philosophy with the Roman-alphabet-using Western powers. Hamka wrote in 1952, ‘the more time passes the more the new Indonesian drifts away from the soul of Islam’.Footnote ⁴⁵ One political result of the distance from Islamic thought was the decreasing appeal of the idea of an Islamic state. As Michael Laffan has noted, ‘The ultimate failure … to achieve an Islamic state in Indonesia was prefigured by their choice of script.’Footnote ⁴⁶

Finally, Indonesians' decreased interaction with texts in Arabic also exposed them to fewer sources for potential loanwords. This had consequences as they developed their language's vocabulary.

Vocabulary selection

To fuel Indonesian's expanded use, especially after its acknowledgement as the ‘language of unity’ in 1928, the language itself had to expand to accommodate more uses and more contexts. This expansion involved two major trends: nationalisation and modernisation.

The first trend sought to find indigenously acceptable ways to express certain concepts and was primarily concerned with finding alternatives to Dutch words that were frequently used for certain concepts. This process had begun in Islamic groups even under the Dutch. For example, a Muslim student organisation on Kalimantan rewrote their organisational constitution in the late 1930s to include a switch from Dutch loanwords to Indonesian words, including many which were Arabic-derived.Footnote ⁴⁷ This process quickened after independence. In West Sumatra, for example, the Persatuan Tarbiyah Islamiyah (Perti) replaced the Dutch-derived kongres for large organisational and party meetings with an Arabic root word, muktamar. The word kongres ‘smelled colonial’ whereas muktamar was ‘more Islamic and fitting with the environment of independence’.Footnote ⁴⁸ This continued even into the 1950s, when members of parliament would try to amend bills that sounded too Dutch.Footnote ⁴⁹

Beyond merely replacing a few Dutch words in common usage in Indonesia, however, it was necessary to coin new words for concepts where no words had yet been adopted into common usage. This was the push of modernisation.

The process began institutionally under the Japanese. The Japanese established the Indonesian Language Commission on 20 October 1942, shortly after their invasion, bringing together more than a dozen leading political figures, journalists, and literati to address the language's grammar, new words emerging from society, and technical terminology. Sutan Takdir Alisjahbana, a leading Minangkabau journalist, became the head of the Language Office, the centre of day-to-day activities for the commission.Footnote ⁵⁰ His efforts concluded with a two-volume dictionary published during the Revolution, but still the Indonesian language needed to add significantly more words.

To this end, the Ministry of Education and Culture of independent Indonesia established a Working Committee on the Indonesian Language in June 1947. Once again, Sutan Takdir Alisjahbana served as the head.Footnote ⁵¹ In July 1950, after the end of the national revolution and the recognition of Indonesia's sovereignty, the new government established a Terminology Commission, which produced a prodigious 15,000 terms in its first six months.Footnote ⁵² In 1952, the Minister reorganised language development efforts under a new centre at Jakarta's University of Indonesia: the Lembaga Bahasa dan Budaja (henceforth, the Language and Culture Institute).Footnote ⁵³ Headed by the Javanese Dr Prijana, in the first two years of its existence this Institute coined almost 60,000 new words for the Indonesian language in a process that one observer called ‘something out of Gilbert and Sullivan’.Footnote ⁵⁴ The Institute distributed sheets of the new words to newspapers, schools and government offices, where they were then copied, recopied, and passed around until the words came into popular usage. It also published the new words every two months as an addendum to the Institute's journal, Bahasa dan Budaja (Language and Culture).

New words and technical terminology, according to the Institute, should come from a list of source languages, ranked by priority:

1. 1. Bahasa Indonesia

2. 2. Well-known terms from regional languages (i.e., Javanese)

3. 3. Arabic

4. 4. Sanskrit

5. 5. Widely used international terms (often Latin).

6. 6. Other languages (usually Dutch or English)Footnote ⁵⁵

In practice, though, it seems that the new committee reviewing words actively preferred Javanese, Sanskrit, and Dutch more than official policy dictated. Sutan Takdir Alisjahbana, the former head of several language commissions, complained about the attempt to monopolise vocabulary expansion in one office, when traditionally any Indonesian who had need of a new word could create whatever neologism she thought useful. Alisjahbana also bemoaned the high number of Javanese words selected by the commission.Footnote ⁵⁶

Under the leadership of this Dutch-educated, Javanese head of the Institute, Dr Prijana, the Indonesian language was oriented away from Arabic. Instead, the focus was not only Javanese but also Sanskrit, which became the source language when future administrations sought to coin powerful monikers for new ideas.Footnote ⁵⁷ Besides the influence from these two Asian languages, in 1952 Hamka also bemoaned the new Indonesian language as ‘developed from Dutch and filled with Dutch’.Footnote ⁵⁸ This was, of course, clear hyperbole — the vast majority of Indonesian terms and all of the language's structure derived from Malay — but it was an impression that grew out of new bureaucratic realities.

The Culture and Language Institute openly looked to Dutch influences for organisation of the Indonesian language. One of the articles the Director, Dr Prijana, cites as a key source of thinking is an article by the Dutch Orientalist language expert, C.C. Berg.Footnote ⁵⁹ Other authors in the official journal Bahasa dan Budaja suggested systematic adjustments for Dutch words to make them appropriately Indonesian (e.g., the suffix -et from the Dutch -eit, and -in from -ien), but made no commentary on how to adjust Arabic words.Footnote ⁶⁰ The rival journal Pembina Bahasa Indonesia explicitly rejected Arabic suffixes as proper Indonesian, showing that Arabic was not to be seen as an inspiration for developing the national language in this new, scientific environment.Footnote ⁶¹

By the late 1950s, language scholars were going so far as to ridicule and disdain the adoption of Arabic words into Indonesian. Umar Junus, a lecturer at a teacher training college in Malang, wrote that the adoption of Arabic (and Dutch) words into the Malay language during the colonial period demonstrated how that language had functioned as a tool of foreign oppression. That made it unthinkable for Indonesian (which Junus identifies as an independent-minded language of the national revolution and liberation) to rely on the importation of foreign, Arabic terms.Footnote ⁶² The popular language critic Armijn Pane argued similarly against the use of imported words, lumping together Dutch and Arabic as foreign influences. Although Dutch is singled out for special condemnation because it has such a pervasive influence, Pane also admits that Dutch retained its place as a language of prestige — which probably saved it from elimination as a source of new vocabulary. Arabic was not so well protected, and so it fell victim to linguistic nativism, which Pane also joined in, saying ‘Thinking in a way that requires foreign languages will not lead us to think in a lively way, and will not lead us to think in a way that creates.’Footnote ⁶³ Thus, the importation of Arabic words, or the creation of Indonesian words from Arabic roots, practically stopped after Indonesian independence.Footnote ⁶⁴

Although the Language and Culture Institute in Jakarta and the Western-educated intellectual elite eschewed the adoption of Arabic words, grassroots forces immediately after independence still looked to Arabic. Words were not always standardised immediately, especially major imports and Dutch loanwords with a natural Arabic alternative. Take, for example, the word used for Indonesia's proclamation of independence on 17 August 1945. This has now been canonised in popular imagination as the ‘Proklamasi’, a Dutch-derived neologism that is reserved almost exclusively for that one historical moment. And yet, during the Revolution not all Indonesians knew the event by the term Proklamasi. In 1946, an Islamic publication in Medan was still calling it Perm'aloeman (in modern spelling, Permakluman), an Arabic-derived word still common today in Indonesian (especially in Islamic circles), although superseded for referring to this 1945 event.Footnote ⁶⁵ Even when using the word Proklamasi, for the first decade or so of independence this term often had to be explained in context with another word — usually an Arabic loanword. This led Mohammad Yamin, Minister of Education and Culture at the time of the Indonesian Language Congress in Medan in 1954, to the peculiar turn of phrase ‘proklamasi jang dimaklumkan Soekarno-Hatta’ (‘the proclamation [Dutch loanword] that was proclaimed [Arabic loanword] by Soekarno-Hatta’).Footnote ⁶⁶ This use of the word maklum to explain the new proklamasi shows the strength of Arabic loanwords in many sectors of society, and furthermore demonstrates how the official preference for European borrowings was a result of the contingencies of history (a Dutch-trained Javanese man running the Language and Culture Institute, among other things).

Theological leaders, on the other hand, including some in smaller Islamic political parties, still saw Arabic as an important language not only for religious devotion but also for state purposes in a future Indonesian Islamic state. These leaders found Arabic-derived words, and even directly-imported Arabic words, unavoidable in pursuing their Islamic political goals. The congress of the small Islamic party Perti, held in Jakarta in August 1955, issued its draft constitution for an Islamic State of Indonesia, which they gave an Arabic (not Indonesian) name: ‘al-Daulah al-Jumhuriyah al-Islamiyah al-Indonesia’. Furthermore, Perti had several Islamic concepts (such as ‘regulation or general guidelines’ [nushuh au mabadi' al-‘ammah] and ‘coordinating board’ [ahl al-hilli w al-‘iqdi]) that it felt uncomfortable including in its constitution just in Indonesian, so they were incorporated in Arabic as parenthetical explanations.Footnote ⁶⁷

The choice of Islamic leaders to continue to employ, or even coin, Arabic loanwords in Indonesian impacted their place in Indonesia's new society. Because Arabic loanwords were foreign and rejected by leading linguistic commentators across the board, their use made Muslim leaders appear foreign, possibly even subversive, or more likely as backward bumpkins, in the eyes of the new bureaucratic elite. More importantly, it rendered their arguments and justifications unintelligible to the majority of young Indonesians, the generation pushing forward and shaping the new state. These ideas did not have the benefit of coming from a prestige language, as Dutch loanwords (and later English ones) did, and so they must have rung very hollow and unappealing for the generation of young Indonesians being taught to forge the country's identity with European influences and Indonesian spirit. The real consequence, then, was the rejection of ideas from Islamic leaders: certainly their idea about an Islamic state, which failed to win the expected majority of Constitutional Assembly seats in 1955, but also presumably lesser policy and governance proposals. This is not to suggest that linguistic hurdles were the only, or even the main, reason that voters rejected the Islamic platform, merely that they would have contributed to the alienation between Islamic party leaders and the Indonesian public.

Spelling reforms

Just as much as the addition of new words to the Indonesian language demonstrated how language bureaucrats wanted the Indonesian language to develop (projecting the future), the regulation of spelling demonstrated the way in which they wanted to deal with Indonesian's linguistic heritage (managing the past). Which linguistic markers to preserve became a hotly-debated topic in the late 1940s and 1950s. Several small yet systematic spelling reforms took place that demonstrated the adoption of Western sensibilities over Arabic ones and, equally important, reduced the Islamic potential of the new national language.

In early 1953, the Language and Culture Institute, the same body tasked with creating new vocabulary for modern Indonesian, sent out letters to luminaries in the areas of language and literature to ask for their opinions on spelling in the national language. The Minister of Education and Culture in 1947 had already issued a directive with a few quick guidelines to modernise spelling, but by 1953 the environment was ripe to consider further revisions. After soliciting feedback from experts across the archipelago, the Institute reprinted their experts' replies in the October 1953 edition of Bahasa dan Budaja, and the consensus that emerged shows a distinctly European sensibility about language.

Among the several points at issue in the spelling of modern Indonesian, many concerned how to adopt Arabic derivatives into the Indonesian language. In some cases, Arabic loanwords had already morphed to make themselves sufficiently easy for Indonesians to pronounce (e.g., the Indonesian word waktu, time, from the Arabic waqt , had permanently added a final vowel that appears only on the nominal case in the Arabic, helping Indonesians to vocalise the /t/). These Indonesianised words needed no further adaptation. However, a few Arabic letters posed particular troubles for Indonesian: the glottal stop hamza (ء), the voiced pharyngeal fricative ‘ain (ع), and the voiceless velar fricative kha (خ). These are not the only Arabic letters without a direct romanised equivalent.Footnote ⁶⁸ However, these three letters posed the greatest challenge because they were sufficiently common and could not be easily reassigned to another letter. This suggests that these three letters had been vocalised by Indonesians, especially those used to reading these letters in jawi, thus making it harder to push them out of popular usage. Whereas most language standardisations around the world pushed for the goal of ‘one-letter-one-sound’, Indonesia's reforms threw this model on its head to prioritise conformity with modern (read: European) languages, and thus blocked the possibilities of Arabic words in the new national language.

The glottal stop is a sound also native to Southeast Asia, and found in several common Malay words of Austronesian origin, such as the words for ‘no’ and ‘father’ (today spelled tidak and bapak, respectively, with the final ‘k’ symbolising a glottal stop). In Malay words from Austronesian, the glottal stop always appears at the end of a word, whereas in Arabic it can appear in other positions. In romanised transcriptions, this sound was commonly signified by an apostrophe (thus, tida', bapa') through the 1940s, but the trend in the early 1950s was to signify this sound instead with the letter ‘k’.Footnote ⁶⁹ This solution avoided the use of an apostrophe to replace a phoneme. Because the apostrophe did not appear in this way in Dutch or English, Indonesian linguistic scholars saw it as undesirable to make it a phoneme in Indonesian. Even the Ministry of Education and Culture weighed in against the use of apostrophes in words, and supported the implementation of ‘k’ as a replacement.Footnote ⁷⁰ However, this created ambiguity and violated the ‘one-letter-one-sound’ principle as well, something that language scholars at the time claimed they wanted to avoid.Footnote ⁷¹ In certain Indonesian words, a final ‘k’ is, in fact, pronounced as /k/ and not a glottal stop, as in masak (to cook), or at the end of syllables such as in saksi (witness) and bukti (evidence). Thus, by using a final ‘k’ to symbolise hamza, the new Indonesian spelling fell short of the ‘one-letter-one-sound’ ideal and lost some clarity, all in the attempt to eliminate Arabic influence and become more European in form.

Even greater confusion arose with attempts to replace the letter ‘ain. Several common Indonesian words derived from Arabic include this letter, such as rakjat (people/masses, from ), Djumat (Friday, from ), sjair (poetry, from ), taat (obedient, from ), and adil (just / fair, from ). Here again, the ‘ain had commonly been symbolised in Dutch-era romanisation by an apostrophe (thus ra'ajat or ra'jat, Djum'at or Djuma'at, sja'ir, ta'at, and ‘adil). Because the apostrophe could be interpreted as either hamza or ‘ain by those who did not know the word, confusion was also possible here, but because many Indonesians pronounced the ‘ain as a glottal stop, this was only a minor problem. Unlike in the case of hamza, however, the linguistic experts consulted by the Language and Culture Institute could not come to a consensus as to how to transcribe ‘ain in a new romanised system of spelling. Some, including R. Satjadibrata, felt that ‘ain could not be eliminated, and had to continue to be represented with an apostrophe, so that the pronunciation and meaning of certain words would not be lost; he notes the example of sjair or sja'ir (poetry), which cannot be pronounced with an /ai/ diphthong as one would assume without an apostrophe.Footnote ⁷² Others, such as ta'at and ‘adil, saw the elimination of the apostrophe but no new letters; in the case of Djumat there continued to be debate on whether the word should have one ‘a’ or two.Footnote ⁷³ By far the most confusing were words in which the ‘ain preceding a consonant or in a series of consonants became written as a ‘k’. The letter ‘k’ here could also symbolise the sound /k/ or hamza, so that the many Indonesian people learning the national language for the first time from its transcription understandably could not identify which sound was intended. The long-term result is that the regional pronunciation of several words originally pronounced with an ‘ain is now a hard /k/ sound (including rakyat, yakni, makna, and nikmat), especially on the Outer Islands.Footnote ⁷⁴ The pronunciations with a hard /k/ seemingly came into existence during the 1950s, when people across the archipelago were learning the language from scratch based entirely on its written form. These pronunciations certainly were not apparent in the various spellings recorded in turn-of-the-century jawi dictionaries.Footnote ⁷⁵ This spelling reform did more than abandon the ‘one-letter-one-sound’ ideal. It went further by eliminating both the written representation and later even the spoken replication of a sound previously present in many dialects of Malay, and so the reform fundamentally changed the pronunciation of many words.

The fate of the Arabic letter kha, traditionally transcribed ‘ch’ in Indonesian, was also thoroughly confused. The Ambonese high school teacher A.W.J. Tupanno noted that the 1947 Soewandi directive did not list ‘ch’ as one of the compound letters of Indonesian, leaving the public confused as to how to tackle this sound. People in Ambon, he reported, variously pronounced Indonesian words like chusus (especially, from ) with an initial /h/ sound, /tʃ/ as in ‘church’, or sometimes with a hard /k/, instead of the intended velar fricative. He suggested the elimination of the consonant kha from Indonesian altogether.Footnote ⁷⁶ In the end, Arabic kha was sometimes represented with a ‘k’ (as in kabar [news, from ] and Kamis [Thursday, from ]); this led to increased confusion with the other two phonemes from Arabic also transcribed ‘k’. Alternatively, kha was sometimes kept as ‘kh’ but pronounced usually as /h/ (as in khas [unique, from ] and akhir [final, from ]). Leading up to the 1954 Indonesian Language Congress in Medan, some editorials came out calling for ‘kh’ to be simplified down to ‘k’ or ‘h’, with the rationale that ‘foreign letters or words that are used in the Indonesian language, their spelling should be brought into line with the official spelling of the Indonesian language’.Footnote ⁷⁷ In certain areas, like Sumatra, where Arabic influence was greater, the pronunciation kha remained, thus forcing the new writing system to include it as an additional two-letter phoneme, and yet it represented a non-European sound that the Dutch-educated scholars felt uncomfortable accommodating.

All three phonemes examined could collapse into the letter ‘k’, on top of the letter's traditional burden of the voiceless velar plosive (the /k/ sound in Dutch or English). This imploded the ‘one-letter-one-sound’ standard and thus rendered the language standardisation process self-contradictory: in their efforts to make the language a ‘modern’ and ‘scientific’ one by eliminating non-European sounds and symbols, Indonesian linguists violated the most modern and scientific rule of orthographic reform. Unlike the colonial Philippines, where the letter ‘k’ became a symbol of opposition to European colonial heritage and control,Footnote ⁷⁸ in independent Indonesia the letter ‘k’ could be taken as a symbol of ongoing European colonial heritage: postcolonial scholarship bowed to ill-fitting European standards.

In the few cases where Arabic sounds could not be ignored, Indonesians still tried to domesticate them into European normativity. A key example is the word ‘Qur'an’, which employs both a deep qaf sound in Arabic as well as a glottal stop at the start of the second syllable. Being a crucial word for religious purposes, even the most avid moderniser was loath to eliminate its Arabic flavour entirely, but many still tried to tame it into Western norms. Thus, the public intellectual Ghazali Hasan writing in the Medan newspaper Tangkas in October 1954 spelled it ‘Qur-an’, replacing the awkward apostrophe (which did not appear in Dutch words) with a hyphen, which was much more common for Dutch or English words.Footnote ⁷⁹

The confusing and inconsistent approach to the Arabic phonemes analysed above — phonemes that were frequently pronounced in some regions — shows the tendency of Dutch-educated, mostly Javanese, language experts to erase Arabic influences. The experts found these sounds, which occurred neither in spoken Javanese nor in the Dutch alphabet, inappropriate for the modern, national language, and so they removed representations of those sounds in the transcription of the Indonesian language.

The importance of these small spelling reforms on Islamic-minded individuals was not immediately apparent in the 1950s, in part because many of those who would have been concerned about them were not literate in romanised Indonesian. Some impacts became more apparent over the following two decades, though. In seminars leading to the creation of the ‘perfected’ spelling system (ejaan yang telah disempurnakan) in 1972, the Islamic author Hamka protested mightily at the final death of the apostrophe (for the ‘ain sound in particular). Among other problems, he realised that without separate letters or symbols for various Arabic sounds, certain dissimilar words in Arabic would be rendered homophones in Indonesian, such as nikmat, which could represent either comfort (the common meaning, from Arabic ni'mah ) or indignation (from Arabic niqmah ).Footnote ⁸⁰ This pointed to two problems for the Islamic community moving forward. First, this obviously made certain importations from Arabic impossible. Second, and more broadly, scholars worried that this deficiency in the spelling system hampered Indonesian transcriptions from aiding in the process of learning Arabic pronunciation, with implications for non-experts learning to recite the Qur'an, say their prayers, or engage in religious contracts such as marriage. Islamic leaders called for, at the very least, the retention of the apostrophe for ‘ain, but to no avail; it was finally and officially removed from Indonesian in 1972.Footnote ⁸¹

There were some more immediate consequences of the spelling standardisation, but they were more subtle. One was that making jawi or Arabic-inspired spelling non-normative made Islamic scholars look foolish in their writing of Indonesian. An example from 1952 will demonstrate this issue clearly. In that year, traditionalist Muslims split from the umbrella Islamic political party Masjumi, which was led by Dutch-educated politicians. The leader of the traditionalists in this action, Kyai Wahab Hasbullah, wrote a letter to the leaders of Masjumi threatening to initiate the exodus, and this letter came under scrutiny. The historian Deliar Noer, when evaluating the exchange, noted in dismissive language that ‘It appears that this letter was written by Kyai Wahab himself. This is seen in the style of language and spelling of the letter.’Footnote ⁸² Because Kyai Wahab employed jawi for his own use almost exclusively, one can be reasonably certain that the tell-tale signs seen by Noer are markers of jawi usage. Taking the single sentence provided by Noer, one can examine these trends.

Noer quoted Kyai Wahab as saying in Indonesian, ‘saja minta ma-af beribu-ribu ma-af, saja akan coba2 berdjuwang untuk mencapae tuntutan tersebut tiada dengan melalui Masjumi lagi,’ meaning, ‘I apologise a thousand times, I will try to struggle to reach these aforementioned points without doing it through Masjumi anymore.’Footnote ⁸³ The sentence contained four key variations from standard Indonesian: ‘ma-af’ (usually maaf), ‘coba2’ (usually coba-coba), ‘berdjuwang’ (usually berdjuang), and ‘mencapae’ (usually mencapai). The first was given a marker in the middle to symbolise a glottal stop. The second employed a numerical marker for repetition, which originated from jawi, although it later became common in casual writing.Footnote ⁸⁴ The third used a spelling that appears to be a more natural transcription of the jawi spelling for this word (), in which the Arabic letter waw (و) could be interpreted to function either as a long vowel /u/ sound or as a consonant /w/ sound flanked by short vowels. Finally, the word mencapae reflected a similar unfamiliarity with standard vowels for the transcribing of jawi vowel combinations. Although consonants were consistent, vowels in Arabic script were left basically to the pronunciation of the user, and in this case Kyai Wahab used a vowel transcription that better reflected the Javanese accent. All of these linguistic choices were understandable when recognising their jawi roots. However, to someone without knowledge of jawi (or someone who denigrated jawi as a valid linguistic choice), Kyai Wahab looked like an unintelligent man who could not remember proper spellings. Thus, the standardisation of spelling, done as it was to Western norms, made Islamic leaders appear foolish in the short term, in addition to limiting their vocabulary choices and hampering Arabic education in the long-term.

As with the other linguistic reforms in the standardisation process of the 1950s, it is almost certain that the technocrats in charge of the reform process did not intend to sideline the Islamic community through this change. Nevertheless, that was one unintended consequence of the standardisation of spelling.