Journal of the Straits Branch of the Royal Asiatic Society/Volume 9/On the Transliteration of Malay in the Roman Character
ON THE TRANSLITERATION OF MALAY IN THE ROMAN CHARACTER.
by W. E. MAXWELL.
SOME years ago, in compliance with the directions of the Secretary of State for the Colonies, a system was adopted by the Government of the Straits Settlements for the spelling of native names, in which a want of conformity was complained of. It is convenient and desirable that there should be some standard for the spelling of names which may appear in official correspondence, which may be printed in Blue-books, and quoted in Parliament. But a system may satisfactorily secure uniformity which may nevertheless be wanting on the score of scholarship, and, unless sound in the latter respect, it will not answer the purposes of the philologist or geographer.
The adoption of the Government system by the Council of the Straits Branch, Royal Asiatic Society, as that which members are invited to adopt,[1] lays it open to their criticism. It may be questioned if it is satisfactory from a scientific point of view, or in accordance with principles of true scholarship. Two distinct subjects—transliteration and pronunciation—are confused, and the report which deals with them does not sufficiently distinguish between instructions how to spell and instructions how to pronounce.
The subject is a difficult one. Marsden, Crawfurd and Logan have failed to find a satisfactory settlement of it, but I do not think that the last word on it has yet been said. The following remarks on the transliteration and pronunciation of Malay words are offered to the Society with the view of drawing the attention of the Council to the advisability of the adoption for literary and scientific purposes of some better systein of rendering Malay words in Roman lottors than that hitherto recommended.
GENERAL PRINCIPLES.
There are two objects to be kept in view in deciding upon a system by which to render Malay in Roman characters:—
1st. To obtain a faithful transliteration of the Malay character.
2nd. To clothe the words in such a form that they may be pronounced correctly by an English reader.
The first regards letters before sounds, the second regards sounds before letters.
Either of these objects may be attained separately, but to combine both without perplexing the reader is more difficult of accomplishment. If the reproduction in some form or other of native letters (for some of which the English alphabet has no equivalent) is too exclusively attended to, the result may sometimes be a word which is difficult of pronunciation to the uninitiated. Crawfurd claims the advantage of simplicity for his system, yet few persons probably would recognise in S’ex[2] the common Arabic word Sheikh. On the other hand, if the system be purely phonetic, the ear must be entirely depended on; sounds which nearly approach each other will be mistaken one for another, and persons professing to use the same system will very likely spell words differently.
Another important point must be borne in mind, Malay contains a large number of pure Sanskrit and Arabic words; it is necessary, therefore, to avoid any serious departure from the principles sanctioned by European scholarship of transliterating those languages. Any system of spelling Malay would be discredited which should present common Sanskrit and Arabic words in uncouth forms hardly recognisable to students of those languages.
It is submitted, therefore, that in a really sound system of Romanised Malay,—(1) the native spelling must be followed as far as possible; (2) educated native pronunciation must be followed in supplying vowels which are left unwritten in the native character; (3) native pronunciation may be disregarded where the written version is not inconsistent with the true pronunciation of a Sanskrit or Arabic word.
Examples:—
1. ماري Mari, come. (Here the four letters m, a, r and i exactly transliterate the four native letters).
2. تمڤڠ Tampang, a coin. Tampong, a patch. Tempung, a game. Tempang, lame. Tumpang, to lodge.
These five words are spelt in the same way in the native character, in which only the consonants, t m p n g, are written. Regard must, therefore, be had to pronunciation in assigning the proper vowels to them when rendered in Roman letters.
3. منتري Mantri, a minister. This word is pronounced by Malays M’ntri, as if there were no definite vowel between the m and n, but its Sanskrit origin shews clearly that a is the vowel which ought to be supplied.
ڤترا؛ڤتري Putra, a prince, Putri, a princess; in these words, too, the vowel-sound in the penultimate is indefinite, but the vowel a is properly supplied, both being common Sanskrit words; to write them petra and petri would be to disguise their origin.
VOWELS.
The difficulty of arriving at a satisfactory system of transliteration of Malay is caused partly by the insufficiency of the Arabic vowels to render the Malay vowel-sounds.
The vowels borrowed from the Arabic are four:—
ا Alif, ă, as the a in father. باڽق bāniak, many, much, very; لام lāma length of time.
و Wau, ō, ū, as the o in nose and the u in truth. تولق tōlak, to push; ݢون gūna, quality, use.
ي Ya, ē, ī, as the ê in fête and the double e in thee. بيد bēda, difference; بيني bini, wife.
ع Ain, ’a, ’e, ’i, ’u. This vowel conveys a deep and somewhat nasal sound which must be heard to be understood; examples: عمر ’umur, life, age; عقل ’akal, mind, intelligence; علم ’ilmu, science.
These are always long. A short vowel is not written. In Arabic indeed it may bo denoted by what are called vowel-points placed above and below the consonants, but vowel-points have been generally adopted in Malay, and the short vowels are left to be supplied by the reader like vowels in our ordinary short-hand.
To show how completely the use and the accentuation of the vowels in Arabic differ from Malay, to which language nevertheless the Arabic alphabet (with some additions) has been applied, it is only necessary to examine a passage of Arabic transliterated in the Roman character, e.g., an extract from the Kur'an or from any other book, or to hear it correctly read.
The majority of the words, it will be found, end in open vowels, and in pronunciation the long vowels are strongly accentuated. A short e is of rare occurrence.
Take a sentence of equal length in Malay; it will be remarked that most of the words end in consonants, the exceptions being generally words of Sanskrit or other foreign origin, in many words the nominally short vowels, namely those not written, will have equal value in pronunciation with those which are written, and a sound which corresponds closely with the short e in the English words belong, bereft is abundant.
In writing Malay, therefore, the Arabic alphabet has to express sounds very different from those of the language to which it belongs.
The short e in Malay is often "a distinct and peculiar sound, which has a separate character to represent it in the Javanese alphabet,"[3] but for which there is no particular sign in the PersoArabic alphabet used by the Malays.
This sound can only be expressed in Arabic writing by the vowel-point called fathah (Malay, baris di-utas); it is a dash placed over the consonant to which the vowel belongs. The particles ber-, ter- would be written بَر، تَر.
(The fathah, however, denotes a short a as well as a short e as kapada كَڤَدَ).
In the words sembah, salutation, homage, bendang, a rice-field, senduk, a spoon, the first syllables are not pronounced like the English words gem, men. An indefinite sound is given to the syllables mentioned, as if it were attempted to pronounce the two consonants without an intervening vowel, s’mbah, b‘ndang, s‘nduk.
Some English scholars seeking a satisfactory mode of rendering Malay in Roman letters have attempted to do what the Malays have not thought it necessary to do for themselves, namely to denote this peculiar vowel-sound by a particular sign. Crawfurd professed to distinguish it by d; Keasberry wrote ŭ; there is perhaps good reason for this in works intended for the use of students beginning the study of the language, vocabularies, grammars and the like. But the authors of the Government spelling-system, who selected ĕ to express the sound in question, might have spared themselves this additional vowel-symbol.
As we have seen above, this sound can only be expressed in writing by Malays by the fathah, short a or short e. Why not be satisfied with a or e to express it in English? This is quite sufficient for purposes of transliteration, and scientific men do not want to burden their text with accents to denote sounds not expressed in the native text. We do not distinguish by a different sign each of the numerous ways of pronouncing e in the English or French language.
Once quit the safe ground of transliteration and trust to that uncertain guide—the ear—and all chance of uniformity is at an end. Let us see how the systems mentioned above have worked in practice. Take, for instance, the short syllable sa, which is frequently found as the first syllable of Malay words. The authorities who have been quoted are not agreed when to give the syllable the force of the vowel a and when to introduce their signs for the peculiar vowel-sound which they want to represent.
Keasberry writes samoa and sakarang, but săblah, sădikit and săbab.
Crawfurd writes sabenar, săbăb, sadikit and sâdikit, sakarang and sâkarang, sambilan and sâmbilan; one word is spelt in four different ways, sâpârti, sapârti, sapurti and sâpurti; he introduces the vowel in a curious manner in the Sanskrit words srigala, which he spells sârigala, and sloka, which he spells sâloka. The short vowels in the Sanskrit word sâbda and the Arabic word săbtŭ are represented in different ways.
The Spelling Committee of the Straits Settlements write Selangor, Sarâwak and sembilan, though it is not clear why sa is allowed to stand in Sarawak while Salangor is held to be wrong. The adoption of the syllable se in sembilan (nine) is still more singular, for the vowel is clearly a, sambilan being derived from sa-ambil-an, "one taken away (from ten)." In most instances this initial syllable is derived from the Sanskrit sa or sam (with) and it cannot be right to render it by se or sĕ, which do not more nearly approach the Malay pronunciation than sa.
Many other instances might be given. I have seen in Government publications the name of the Malay State "Patani," spelt "Pĕtani." Yet it can hardly be said that there is good reason for departing from the established mode of spelling this word (which has been spelt "Patani" from the days of James I.), when it is remembered that the Malay historical work called Sajarah Malayu says that the state was called after a fisherman who had a son called Tani and was therefore called Pa-Tani (Tani's father). However absurd this derivation may be, its occurrence in a purely native work is at all events conclusive as to the pronunciation of the first syllable.
SYSTEM PROPOSED.
Vowels.
The only use of the accents which will be inserted is to denote that the vowel is expressed in the Malay text. No sign will be used to denote the accentuation of any particular syllable; transliteration, not pronunciation, is the first object to be kept in view. For goneral purposes, the accents may be omitted at option. It cannot matter whether مات, the eye, is rendered mâta or mata. Thus:—
â corresponds with ا written in Malay, as ڤاڤن pâpan.
a and e correspond with fathah where the Towel is omitted, as ڤنجڠ panjang, برچري ber-cherei.
î and ê correspond with ي written in Malay, as بين bînî, كيچيق kêchek.
i and e correspond with kesrah where the vowel is omitted, as دندڠ dinding, ضاهر zahir, ڤاتق pâtek.
û and ô correspond with و written in Malay, as بوت bûta, بوهڠ bôhong.
u and o correspond with dammah where the vowel is omitted, as تنتت tuntut, ڤندق pondok.
The Greek rongh breathing before a vowel denotes the presence of ع ain in the native writing, as عقل ‘akal, عمر ‘umur معلوم ma‘alum.
Dipthongs.
ai corresponds with ا and ي when followed by a consonant. as باءيق baik ناءيك naik.
au corresponds with و, as ڤولو pulau.
ei corresponds with ي, as سوڠي sungei.
Y and W.
Y should be written for ي when it precedes or is preceded by a long vowel, as سايڠ sâyang: لاير layar; بايڠ bayang; مويڠ moyang; بويڠ buyong. Exception, should never be rendered by iy for this gives two letters to one Malay character where one letter is sufficient; سياڠ siang, not siyang; سيوڠ siong not siyong.
W should be written for و when it precedes or is preceded by a long vowel, as باوا bawa: كاون kawan: لاوق lawak.
Exception:–
و should never be rendered by uw, for this gives two letters to one Malay character and one sufficiently expresses the sound: بوات buat, not buwai; كوالا kuala, not kuwala; توان tuan, not tuwan.
Liquids.
The combination of two consonants the latter of which is a liquid, which is so common in Aryan languages, is not to be found in indigenous Malay words. Where it apparently occurs its presence is caused by the elision of the vowel in one of the Polynesian prefixes ber, ter, ka, sa, and pe.
There are, of course, plenty of Sanskrit words in Malay in which the junction of two consonants, one being a liquid, occurs, such as sutra, indra, sri, mantri, but I believe that no instance of two consonants sounded together can be pointed out in Malay which cannot be accounted for either by foreign derivation or elision of the vowel of a particle.
Malay is an agglutinative language, and many of its dissyllabic radicals have been developed from monosyllables by the prefix of particles. Their origin has been forgotten and by the gradual growth of the language they may be now lengthened into words of three, four and five syllables by the addition of prefixes and affixes, each change giving fresh development to the simple idea embodied in the radical.
To analyse the origin of indigenous Malay words and to get some idea of their derivation, and of the connection between many which present distinct forns and get obvious similarity, it is necessary to identify the agglutinative particles and to distinguish them from the root. Where the syllables are distinct this is easy; in the words mekik, to cry out, to hoot; pekik, to squeal or scream as a woman; berkik, the snipe, literally, the squeaker,–the common root kik, and the agglutinative particles, me, pe, and ber, are easily distinguished.
But where the first letter of the root or radical is a liquid, there is a tendency in pronunciation to blend with it the first letter of the particle. Nevertheless, it is quite clear that in spelling such words as pelandok, the mouse-deer; pelantak, a ramrod; peluru, a ballet, the full value of the particle should be shewn, and that plantak, plandok and pluru are incorrect and unscholarly.
Pe is the sign of a verbal noun. I do not know of any Malay verb landok, but that the name of the mouse-deer is derived from a word having something to do with rapidity of motion is sufficiently shewn by the meanings of other words having the same root:–
- Lanchit and lonchat, to jump, spring.
- Lanchar, quick, direct, fluent.
- Lanchur, to flow, spurt out.
- Lanjut, long, stretching forward.
- Lantak, to strike home, transfix.
- Lanting, to fling.
- Langsong, to proceed direct, &c.
On the same principle, it is not incorrect to show, by the insertion of the vowel before the liquid, the existence of the forgotten particle in the first syllable of such words as, bri (be-ri), give; blanja (bel-anja), expend; blanga (bel-anga), a cooking pot; trang (te-rang), cleared; trima (te-rima), receive; trus (te-rus), through.
[4] Consonants.
The following are the consonants used in writing Malay with the equivalents by which I propose to represent them in Roman letters:–
| ب | bâ | b |
| ت | tâ | t |
| ث | sâ | s [5] in Arabic th, pronounced as in thin. |
| ج | jim | j |
| چ | châ | ch |
| ح | hâ | ḥ |
| خ | khâ | kh[6] |
| د | dâl | d |
| ذ | dhâl | dh pronounced in Arabic like th in this. |
| ر | râ | r |
| ز | zay | z |
| س | sin, sim | s |
| ش | shin, shim | sh |
| ص | sâd | ṣ[7] |
| ض | ḍâd | ḍ[8] (in pronouncing this letter the tongue touches the back of the upper front teeth). |
| ط | ṭâ | ṭ |
| ظ | ẓa, ẓôi | ẓ[9] |
| غ | ghrain | ghr[10] |
| ڠ | ngâ | ng |
| ف | fâ | f |
| ڤ | pâ | p |
| ق | ḳâf | ḳ[11] |
| ک | kaf | k |
| ݢ | gâ | g hard |
| ل | lam | l |
| م | mim | m |
| ن | nun | n |
| و | wau | w |
| ه | hâ | h |
| ي | yâ | y |
| ڽ | nia | ni, ny, nia, nya |
Some of the foregoing letters represent sounds which do not belong to the native Malay language, but which are found only in words taken from Arabic. Uneducated Malays make little attempt to pronounce them, but every boy who learns to read the Kur'an has to do so and the present tendency of the language is to borrow more and more from the Arabs.
f is almost always turned by Malays into a p: e.g. pikir for fikir.
k and ḳ are generally pronounced alike by Malays and kh is not always distinguished from them.
ص س ث are all pronounced alike, as s, by the Malays.
In the same way little or no distinction is made in pronuncia- tion between t and ṭ. The letters denoted by ḍ and ẓ are generally mispronounced by Malays, who sometimes render them by l sometimes, as do Muhammadans in Persia and India, by z.
Spelling of Arabic Words.
Certain rules remain to be noticed which should be observed in transliterating Arabic words in Malay literature.
Al (el-) is assimilated before the solar letters, which are:—
- ل ظ ط ض ص ش س ز ر ذ د ث ت and ن.
The other letters are called the lunar letters and do not assimilate the ل, namely:-
- ء و م ک ق ف غ ع خ ح ج ب ا, and ي.
Examples: -r-rah-māni-r-rahim, the merciful, the compassionate; māliki yaumi-d-dīn, the Lord of the Day of Judgment; aleyhi-s-sa-lam, on him be peace. Proper names: Abdurrahman Dia-uddin.
The force of the orthographical sign called teshdid may be rendered by doubling the consonants over which it is placed as tammat, finished; jannat, Paradise (lit. "the garden"), Muhammad, Mohamed; Sayyid, a descendant of the Prophet.
- ↑ "Malay and English Spelling," Journal of the Straits Branch of the Royal Asiatic Society, Vol. I., p. 45.
- ↑ Crawfurd's Dictionary
- ↑ Crawfurd, Malay Grammar, p. 4.
- ↑
One advantage of inserting the vowel is that the separation of the particle from the root renders apparent etymological features which might otherwise be unsuspected. Thus, in the examples given above, the same root may perhaps be detached in the Malay words for "give" and "receive."
So the common derivation of belanga and other words having to do with heat or burning becomes apparent :-
- Bel-anga, a cooking pot.
- Hangat, hot.
- Hangas, burnt, scorched.
- Hangit, smell of something burning.
The meaning of ran or rang appears to be "to cut;" it occurs in such words as, rantas, to cut a passage through jungle, ranchang, to whittle to a point, etc.; terang, or trang, is "cleared," "cut away," and therefore "clear," "plain;" pa-rang, is "the cutter," the chopper or jungle-knife used in agriculture.
Us, the root of terus or trus, seems to convey the idea of admission or penetration:–
- Terus, through.
- Chelus, admissible.
- Lulus, admissible, permissible.
- Tumbus, pierced, perforated.
- Halus, fine, slender.
- Karus, thin, &c.
- ↑ Only two words are in common use in Malay which commence with this letter, namely the names of the second and third days of the week.
- ↑ خ is a strong guttural. It resembles the sound of ch, in the Scotch word loch.
- ↑ ص is strongly articulated palatal s, somewhat like ss in hiss.
- ↑ ض the true sound of this letter must be learnt by the ear. It is like a strong d.
- ↑ ظ the power of this letter is that of z, pronounced with a hollow sound from the throat.
- ↑ غ is a hard guttural g. It somewhat resembles the sound of the Northumbrian r.
- ↑ is a guttural k. This and the five preceding notes are taken from Faris-El-Shidiae's Arabic Grammar.