Malay Grammar

Malay Grammar (1913)
oleh Richard Olaf Winstedt

[ 5 ]Page:Malay grammar (IA malaygrammar00winsrich).pdf/5 [ 6 ]Page:Malay grammar (IA malaygrammar00winsrich).pdf/6 [ 7 ]

PREFACE

Tins grammar was commenced to supply the want of a text-book for the second or higher examination in the Malay language, prescribed for officials.

In English there are no books in print dealing with the subject except Maxwell's Malay Manual, which is not strictly a grammar, and Shellabear's Practical Malay Grammar (printed in Singapore), which is quite elementary. This book will in no way supplant or interfere with those. Out of print are Crawfurd's Grammar, which among scholars hardly counts, and Marsden's, which so far as it goes is excellent, but it is a century behind modern research.

In Dutch there are several standard works, to which I owe a great debt, especially the grammars of Gerth van Wijk, Tendeloo, Spat, and van Ophuijsen; but Dutch is an insuperable obstacle for the casual student of Malay in the Peninsula. I too must ask forgiveness, if the refraction of an unfamiliar language has led me anywhere to distort the views of authorities I have quoted or criticized.

Arrangement is a difficult problem in Malay grammar. Before the chapter on Affixation it is desirable to deal with the simple forms of such parts of speech as will recur in that chapter as derivatives; and it is also important to deal with the radical form of the verb and then without a break to [ 8 ]proceed to derivative verbal forms. Now the important derivatives are substantival, verbal, and adjectival. Hence the arrangement in this book. While considering all the theories that have obtained on that terribly moot point the Malay verb simple and derivative, I have not hesitated to advance opinions of my own. At the risk of being egotistical, I may perhaps explain that I formed my views originally in the ward of a tropical hospital during an interminable illness, cut off from access to all books of reference except a few Malay classics, which I found it an amusement. to parse and analyse; and subsequent acquaintance with the results of modern comparative study has seemed to me on several important points to give those views support. Sometimes I found that I had unknowingly furbished up an old theory. The distinction I drew between the function of the simple and mě forms of the verb, though based on a very different premise, viz. that mě (and bě) derivatives are adjectival, would seem to underlie in a sense that unfortunately put ‘subjective passive’ theory, which in the bands of many grammarians went so far as to deny that the simple verb could ever be active!

I count it an important point in support of the theories I have ventured to advance, that all the examples quoted in this grammar are extracts from the Malay classics, notably the Sejarah Melayu and the Hikayat Hang Tuah.

I give a short bibliography of the principal works consulted. References to chapter and page it was decided, after some hesitation, to omit passim, since those who are export and interested enough to find their way about in Dutch grammars will have no difficulty in turning to the [ 9 ]passage required, while for others many and minute references are tiresome.

The arrangement or ground-plan of this work was suggested to me originally by Mr. R. J. Wilkinson, C.M.G., who has also read most of it in manuscript. I am greatly indebted to Mr. C. O. Blagden, member of the Council of the Royal Asiatic Society, for ungrudging and invaluable assistance extended over many months: if he will allow me to say so, his research work in Talaing is (directly) a loss to Malay scholarship. Above all, I must thank Dr. Fokker of Amsterdam, who read this book for my publishers with meticulous care, pointed out many inaccuracies, and suggested many improvements; I can only wish it were better worth the trouble he lavished upon it so generously.

Mr. Blagden and my brother Mr. E. O. Winstedt have seen the book through the press. [ 11 ]Page:Malay grammar (IA malaygrammar00winsrich).pdf/11 [ 12 ]Page:Malay grammar (IA malaygrammar00winsrich).pdf/12 [ 13 ]Page:Malay grammar (IA malaygrammar00winsrich).pdf/13 [ 14 ]Page:Malay grammar (IA malaygrammar00winsrich).pdf/14 [ 15 ]

CHAPTER I

ETYMOLOGY

§ 1. Malay is the tongue of the Malay Peninsula, which embraces the Straits Settlements of Singapore, Malacca, and Penang; the Federated Malay States, Perak, Selangor, Pahang, and Negri Sembilan: the states of Kedah, Kelantan, and Trengganu and Johore; and in the extreme north under Siamese protection Patani. It is also the tongue of the Riau Archipelago, of the East Coast of Sumatra, and of the West Coast of Borneo. It is as closely related to Menangkabau as Sundanese is akin to Javanese.

The language belongs to the Malayo-Polynesian or Oceanic or Austronesian family, as it has been termed variously, which covers an area from Formosa to New Zealand, from Madagascar to Easter Island, and includes the languages of the Philippines, the Malay Archipelago, Micronesia, Melanesia excluding Papua, and Polynesia. To the easternmost branch belong the languages of Samoa, Tahiti, and Tonga. To the western or Indonesian branch belong Malay; Malagasy; Tagalog Bisaya and Bontok in the Philippines; Batak and Menangkabau in Sumatra; Sundanese, Javanese, and Madurese Balinese; the Dayak dialects of Borneo; Macassar and Bugis in the Celebes, and many other less-known tongues.

This big Malayo-Polynesian family it has been attempted to connect with a family of Austro-Asiatic languages spread over the south-cast corner of Asia and embracing Munda of Central India; Khasi of Assam; Mon or Talaing and Khmer or Cambojan &c. of Indo-China; Nicobarese; and in [ 16 ]the Malay Peninsula Sakai and Semang. This connection was first definitely asserted by Professor Schmidt of Mödling, Austria, and is now generally accepted; it establishes an ultimate prehistoric relationship between Malay and the languages of the aborigines in the Peninsula.

In a mere introductory chapter to a grammar on one particular language, Malay, it is impossible to do more than summarize briefly the conclusions of philologists like Professors Kern, Brandstetter, Schmidt, Kuhn, Niemann, and other scholars, whose works may be found cited in the bibliography on pp. 8-10; referring especially to points concerning Malay types of grammatical structure; and suggesting problems and difficulties raised by a study of this particular language.

§ 2. Malay, which phonetically is well preserved, has become simplified morphologically. Under the modern system of affixation, which will be handled in the body of this grammar, comics a stratum, out of which it has developed, common to Indonesia. This stratum reveals:

(a) Prefixes m, b, p, k, t.
(b) Suffixes n and i.
(c) Infixes in, m and less widely spread l and r.

m appears as a prefix of the verb and of the adjective. Examples of the former are makan from a root kan eat, minum from inum drink, and from the roots idar and aleh the Javanese forms midar revolve, maleh move, that sometimes occur in Malay literature in place of mĕngidar and mĕngaleh. Instances of the latter are masin salty from asin salt, masam acid from asam a sour fruit, masing-masing several from asing apart.

As a verbal formative, m and the m compound forms generally denote the active, though there is uncertainty on the point. In modern Malay, even as a verbal formative m alone or with infixes would appear to have, in a sense, an [ 17 ]adjectival or participial force; adding an agent to an act, a verb to a subject (§ 38 note; §47). It is quite possible that substantival forms like měntua, měrlimau are abbreviations of orang mĕntua, buah mĕrlimau: in the Bodleian Sri Rama (early seventeenth century) rumah měrděrma occurs for almshouse.

b is an Indonesian prefix widely spread in the language group and is said to form intransitive verbs, e.g. běli buy from root ěli, bělah to split from ělah. Less widely found in the family, b occurs like m as an adjectival prefix.

In modern Malay it occurs mostly as bě(r), a prefix that may best be described as denoting reflexivity, reciprocity, addition, and possession. It would probably be safer to define the Indonesian b as having the same general functions rather than to define it as a formative of active or intransitive verbs.

p serves everywhere as a causal prefix. Examples of its verbal use in Malay are pantul (also antul) cause to rebound, pěngap (also ngap) to make a pant, to pant; and to form a substantive, pangkat promotion, rank from angkat raise, be raised. Professor Kern identifies it with the essential part of apa something.^[1] In modern Malay it figures as pě, pĕ + nasal, pěr + r (or l) (§§ 54-6).

k, which is probably the preposition ka signifying place whither and so state into which, is a prefix common to the Indonesian family and has a passive nuance, forming especially verbal adjectives and abstract nouns. It is also employed with numerals and e.g. in kini now as formative for adverbs. of time. In modern Malay (§ 57) it is used with numerals; forms many abstract nouns with the help of the suffix an; and survives in a few stereotyped words to which may be added kapit supporters of a bridegroom from apit press on either [ 18 ]side; alah, kalah he worsted; aleh, kaleh turn; antil, kantil, anting, kanting swaying; apong, kapong drift. A passage in the early seventeenth-century MS. of Sri Rama in the Bodleian Library, Oxford, contains the verbal use, rare in Malay, fakir miskin sĕmua-nya kĕanugĕrah ayapan raja beggars and the religious poor were all in receipt of food from the prince.

t a prefix corresponding to the modern Malay tě(r) (§ 52) is a formative of function clear and intelligible. It has even the same nuances in different languages of the Malayo-Polynesian family. ‘In Fiji’, says Professor Kern, ‘words thus formed differ from adjectives and passive verbs generally in this respect, that they imply a thing has become so of itself. But it appears to be used also, when they do not wish to mention or when they do not know the agent by whom the thing has come into the state expressed by this or that form of the verb.’

A sub-form ti has been detected. Kern sees it in tiba arrive, for which he suggests the same root as in rěbah fall, and in the Javanese tilem sleep, which he contrasts with malam night (cp. tilam mattress); tiarap, tiada are other instances. In Malay one may compare unjok offer with tunjok show outright; anggul pitch (of a boat) with tanggul bob right up.

n. The modern Malay suffix an (§ 58) conceals two old. suflix formatives, one substantival as in labuhan anchorage from labuh to lower (anchor or curtain, &c.); the other superlative or intensative, a use that may be traced in words like lautan ocean from laut sea, sayuran the vegetable world from sayur vegetable.

i, which still survives as an intensative suffix for verbs (§ 62) was once also like n a substantival suffix. tui master in Fiji corresponds to the Malay tuan; and the use may be seen in the Malay pělangi rainbow from pělang stripe; rambuti rough woollen cloth from rambut hair. Like prefix k, this [ 19 ]suffix i was originally a preposition. It signified place where and corresponds to the modern Malay di. ‘In Indonesian tongues article and preposition are often identical’ and i was also an Indonesian article.^[2]

-in- is said to be a passive formative and is found in the Dayak kinan eaten from kan eat. It perhaps survives in the Malay word binatang, but it has no grammatical significance in Malay. Cp. also sěnantan milk white (of game cocks) from santan coco-nut milk; chěnonut pope's nose in fowl from chonet projecting.

-m- occurs in several Indonesian languages in kuman from the root kan eat. Professor Kern derives even the Malay word těmpat place from the old Javanese těpět spacious, vast + this infix, which he calls a durative. Schmidt summarizes its use in Austronesian tongues generally as expressing ‘manner, internal movement, happening’. It is fairly common in Malay as a crystallized relic and conveys the notion of duration and repetition:—

kunchup closing (of a flower), kěmunchup sensitive plant; kuning yellow, kěmuning a yellow-wood tree; santan coco-nut milk; nyiur sěmantan a coco-nut producing milk. It is commonest with reduplicated forms:—chĕrlang-chěměrlang radiant; gilang-gemilang repeated glillering; guroh-gĕmuroh prolonged roll of thunder; turun-temurun continuous descent; tabur-těmabur all sprinkled; silir-sĕmilir waving to and fro; tali cord, tali-těmali cordage.

-r- and -l- are infixes, common enough in Malay in crystallized forms, but no longer living formatives. So far as can be judged from examples, they appear to denote duration, intensity, plurality, reciprocity, confusion and [ 20 ]kindred ideas. They are employed mostly with reduplicated forms. Their derivatives are nouns and adjectives as well as verbs:^[3]

-r-. jubong, jĕrubong a stretch of awning over cargo; chondong leaning, aslant, chondĕrong leanings (of the heart); kuping, kĕruping a scab; gondong, gĕrondong permanent swelling, goitre; jongkah jagged, jěrongkah jagged (of a mouthful of teeth); kědut, kĕrdut wrinkled; kusut, kĕrusut entangled: sabut fibre, sěrabut fibrous; kas (old Javanese), kĕras hard; kibas shake, kĕribas shake thoroughly; kělip, kěrlip twinkle; kuit, kĕruit wag; titek drop, tĕritek drop continually; sadong, sĕradong trip up; kĕmut (normal) throb of the pulse, gěrěmut (faster) throb of a boil; sendeng, sĕrendeng heel right over.

-l-. biku zigzag, běliku a river bend; kebak, kĕlebak gaping (of a deep cut); kěmbong inflated, kělěmbong a bubble, a blister; tingkah conduct, tělingkah line of conduct; sirat, sělirat mesh-work; kangkang a-straddle, kělangkang the perineum; kědut crumpled, kělědut much crumpled; gĕtar tremble, gělětar (also gěmětar and gěměntar) prolonged trembling; gigit bite, gěligit keep biting; gosok rub, gělosok keep rubbing; kupas peel, kělupas keep peeling; kětak, kětok rap, kělětak, kělětok keep rapping; sidek search, sĕlidek keep searching, search thoroughly; siseh, sĕliseh quarrel one with another.

These single or simple affixes or ‘affixes of the first degree’ may be traced in a few sets of words like adu contest, madu rival, padu weld; alang across, cross, malang thwarting, vexatious, palang cross (kayu palang cross-bar), kalang or galang roller, thwarts; ělok curve, pělok embrace, [ 21 ]kělok curve, arc, tělok bay, jělok deep-curved (of a boat), kělun (suffix u) spirals (of smoke).^[4]

§ 3. What has made it harder to solve the functions of prefixes even in the restricted Indonesian group is that a prefix and an infix frequently combine to form a compound prefix^[5] or ‘prefix of the second degree’ giving rise to a new and crystallized formative. The functions of the obsolete simple separate prefixes and infixes, it is then supposed, are obliterated.

In Malay we get

(a) m + nasal infix, § 44, especially II.
p + nasal infix, § 54.
(b) m + r that is to be seen as a relic in certain nouns

and adjectives, especially plant-names, and for example in the rare variants mĕrawan and mĕrapi where bĕrawan and bĕrapi are the ordinary modern forms.

b + r,^[6] § 49, which is said by Kern to be a fusion of adjectival měr and verbal b.
t + r, § 52.
p + r, §§ 54-56.

[ 22 ]This r is commonly written in all cases in literary Malay, but it is often omitted in colloquial Malay. Dr. Fokker's analysis gives a widespread but not universal practice. ‘A

genuine Malay stem ending in r and having another initial than a never takes a prefix closed by r.’ bakar forms těbakar; labur, pělabur; layar forms bělayar, pělayaran; chichir, běchichir-chichiran; tengkar, bětěngkar. Similarly, ‘Malay stems with medial r and having another initial than a.’ From kirim we get běkirim; from sĕrta, běsĕrta, pěsĕrtakan; from siram, běsiram; from dorong, tědorong; from pěranjat, těpěranjat; from kěrja, běkěrja, pěkěrjaan; from pěrměna, těpĕrměnaï. ‘Before a stem with initial s the prefix without r is preferable; thus běsisek is better than bĕrsisck.’ From ribu we get měribu, běribu not měrribu, bĕrribu; from rangga, mĕrangga; from ragi, měragi; from ragong, tĕragong-ragong; from rajok, pěrajok and so on.

In the Kedah dialect to quote an example from the Peninsula, this r is never sounded at all.

(c) A few crystallized survivals^[7] like kěndudok, sěndudok, sěkědudok a plant; diri, sěndiri, kěndiri self; sĕngkarut interlaced; sembělit costiveness; sěnjolong long-snouted gavial; sěmbuang offering; sěmbulu rough- [ 23 ]hewn; buku, těmbuku knot; gulong roll up, těnggulong a millipede which rolls up.

§ 4. Finally Malay has a few instances of ‘prefixation of the third degree’, where a prefix will be compounded with two infixes dalu, měndalu and bĕndalu, kěměndalu mistletoe; from root lap flash, rělap a flashing rope (used in fishing), gěrlap to flash, gěměrlapan flashing.

§ 5. Comparison of words even in Malay alone shows below the above stratum a very old system of suffixation, which has not yet been unravelled.

(1) gigi tooth, gĕrigi, gĕrigis serrated, gigil, gigir chatter (of teeth), ĕnggil-běrěnggil serrated (of hills), ringgit tooth-edged, milled. (2) iku zigzag, siku elbow, sigong rest on elbows, biku zigzag pattern, bengkok, chongkok, chělengkok zigzag, twisted. (3) kuku claw, kokol curred, kokot clare-shaped, kokong, kĕrukut very claw-shaped, kukur rasp, rasper. (4) gětu, gětil pinch, kětit a pinch, pinched off, kětip nip between the teeth. (5) kait hooking, kail fish with a hook, kais clutch, grab.

To look for verbal or adjectival or other meaning referable to the scheme of our own grammar were futile considering how hard it is to discover such import in the clearly defined Indonesian affixes. It seems probable that words like gigir, gigil, ěnggil, kokol, kukur exhibit what are the prefixes and infixes of Indonesian grammar cropping up as suffixes. But the question belongs to the province of comparative philology rather than to that of Malay grammar.

§ 6. Depending mainly, so far as it has been accepted, on identities in grammar, Schmidt's brilliant synthesis of Austronesian and Austro-Asiatic languages can appositely be recapitulated here. IIis arguments are as follows:

(a) The two families have an absolute identity of phonetic system. [ 24 ] (b) Both place the genitive after the noun.

(c) Both use inclusive and exclusive forms for the pronoun of the first person plural.

(d) There is absolute similarity in structure of words. The old view that Austronesian words were disyllabic had been dissolving for years. Professor Pijnappel a quarter of a century ago analysed words like kělětek, kělětak, kělětok (and one may add bělatek, jělatek sparrow and bělatok, jělatok woodpecker) through kětek, kětak, kětok, down to the onomatopoeic monosyllables tik, tak, tok imitating various notes in tapping. In English, Sir William Maxwell, following Logan, ventured the suggestion, that in tangan hand, tangkap seize, tongkat walking-stick, there existed a monosyllabic root (found in Sakai and Talaing) teng hand. Again scholars have shown that while Indonesian languages have many disyllables like langit sky, bulan moon, puteh white, ulu head, mata eye, they have running through all the group many monosyllabic roots kan eat, tut wind (Malay kentut break wind), num drink, pas loose (Malay lěpas), tong hang (as in gantong), lit (Malay kulit) rind, peel. And critical study of Austro-Asiatic languages on the other side has shown that they contain disyllabic as well as monosyllabic words.

(e) Both families exhibit a remarkable identity in their systems of affixation simple and compound. So they have prefixes k, p, m; infixes m, n, r, l; suffixes n and i. And so far as they can be defined, the functions of these affixes in both families are similar.^[8] [ 25 ] § 7. Foreign loan-words.^[9]

The oldest foreign loan-words in Indonesian languages are Sanskrit, which are found pre-eminently in Javanese, and [ 26 ]then in Malay, but also as far afield as the Celebes, the Philippines, and Madagascar. The borrowing includes not only words for religious, moral, and intellectual ideas but some astronomical, mathematical and botanical terms, a court vocabulary, and a large number of everyday words. A few examples only can be given here:

agama religion; alpa negligence; anggota limbs; angkara violence; angkasa heavens; angsa goose; aniaya oppression; antara between; anugerah gift; atau or; bahagia blessing; bahaya danger; bahasa language; bakti meritorious service; bangsa race; bĕnda thing; bĕntara herald; beta servant; biasa accustomed; bijaksana wise; binasa ruin; buta evil spirit; budi intellect; bumi earth; chahaya lustre, glow; chakĕrawala the revolving vault of heaven; chĕrana bowl; chětěria a kshatriya, warrior; chinta love, regret; chintamani a certain kind of snake; chuka vinegar; daksina south; dělima pomegranate; děnda a fine; děrma alms; dina poor; dosa sin; duka grief; gaya conduct, walk; gĕmpita uproar; gĕnta bill; gĕta divan; harga price; harta property; hasta cubit; jaga to watch; jampi magic; jělma incarnation; jĕntěra wheel; jiwa life; juta million; karna because; kĕranda three-plank coffin; kěrja work; kěsumba red; kětika time; kosa goad; kuasa might; kurnia gift; maha great; makota crown; manek bead; mangga mango; mangsa carrion; mantĕri vizier; manusia man; mara danger; masa time; mělati jasmine; merpati pigeon; mulia illustrious; mutiara pearl; nadi the pulse; nama name; naraka hell; pala nutmeg; papa poor; pěnjara prison; pěrkara affair; pěrkasa brave; pěrtama first; pěrwira warrior; puja prayer; putĕra princeling; putĕri princess; raja prince; rajawali eagle, hawk; rupa appearance; saksi witness; sakti supernatural power; sěgěra quickly; seksa punishment; sěmpurna perfect; sěmua all; sěndi muscle; sěnjata [ 27 ]weapon; singa lion; surga heaven; tenggala plough; upaya resources.

Arabic loan-words deal especially with the sphere of religion and law; but as Marsden maintained not so many are so common in conversation that they can be considered Malay by adoption. Among the commonest are:

alam world; arif wise; akal ingenuity; adat custom; dunia world; fikir think; kadar power, ability; kubur a grave; kuat strong; shak doubt; sěbab cause; sujud kneel in prayer.

A few Persian words occur, mainly in literature, and perhaps they came through a Hindustani channel:

astana palace; bandar seaport; běděbah unlucky; biadab discourtesy; biapĕri merchant; darya sea; diwan court of justice; gandum corn; jadah bastard; jam clock; juadah cakes; kawin marry; kenduri feast; lashkar soldiery; mohor di-mark; nafiri trumpet; nakhoda master of a ship; nesan grave-stone; pahlawan champion; pĕri fairy; sakar sugar; sakhlat broadcloth; sěrban turban; těrmasa show.

Hindustani are jori buggy; lagam bit; ras reins; sardi glanders; tan stable.

Tamil supplies a small number of established loan-words:

kapal steamer; katil bed; ketumbar coriander; kolam pond; maligai tower; měmpělai bridegroom; mutu carat; tandil overseer; tirai curtain.

Chinese loan-words are rare rare and apply to Chinese things:

kongsi a (secret) society; loki Chinese courtesan; loteng upstairs floor; lu you; pekong joss; tanglong lantern.

Of Western languages Portuguese has left a great number of words describing articles of European culture:

bangku bench; bělědu velvet; bola ball; bomba pump; garfu fork; kěbaya gown; lelong auction; meja table; [ 28 ]měrinyu superintendent; padĕri priest; pěluru bullet; pěniti pin; pěrada gold-leaf; pita ribbon; renda lace; rial dollar; sěpatu shoe; tĕratu torture; těrungko prison; tuala towel.

The commonest Dutch loan-words are:

engsel hinge; duit cent; pělěkat placard; sěkopong spade suit in cards; sěturup syrup; ransum rations; tong tub; senapang rifle; bom carriage shaft and landing-place; sopi gin; gelas glass; botol bottle; lampu lamp. [ 29 ]

CHAPTER II

PHONETICS

§ 8. In studying the Malay language which has for the vehicle of its expression an alien alphabet, it is necessary to remember that a language is built not of letters but of sounds. It were too elaborate here to deal with sounds so clusive that they have no graphic symbol. But Malay words are composed of the following sounds represented by letters.

§ 9. Consonant.

(a) Guttural class.

A glottal check or abrupt closing of a final vowel made by stopping the breath. It is represented by ء or ق, romanized ’ or k, as in ڨوکوء poko’ or ڨوکق pokok (and in a few cases rendered confusingly by ك e.g. بايك baik, تيليك tilek and so on). Whatever the symbol in Arabic or Roman script, there need be no difficulty, if it is remembered that final k, no matter how represented, in Peninsular Malay never indicates any other sound than this of the glottal check.

h like the h in Ah Amy, Ah Isabel, Ah uncle; a semivowel rather than an aspirate. It is sounded distinctly only between two similar vowels a … a, o … o as in rahang, bohong, leher. [The Arabic ھ occurs like alif at the beginning of Malay words as a graphic prop for a vowel, which in Arabic cannot stand as part of the syllable but must be accompanied by a consonant; e.g. for u and i except when they are used as semivowels w and y—ايسڤ or هيسڤ isap, اولو or هولو ulu: and to indicate the presence of the indeterminate vowel ĕ, [ 30 ]for which there is no Arabic symbol, an initial alif or ھ are both used, sometimes indifferently:— امس ěmas, هلي hělai, امبوس or همبوس ĕmbus. Its function as a semivowel may be seen in the spellings a tuhan a variant of توهن tuan, توها tuha, ڤاهت pahit, where there is no aspirate sound at all.]

g as geese, gaunt, good, e.g. gigi, gagah, gusi; never as in germ.

k, usually represented by کـ and occasionally by ݢـ, is identical with k in kiss, Kaffir, Koran.

ng as in fling, long, never as in tingle, sponging.

r not the English cerebral (or lingual) but the Scotch guttural r, distinctly but not too emphatically enunciated. It differs in different parts of the Peninsula. In the south it is lingual, only more trilled than in English: in the north it is guttural.

(b) Palatal class.

y as in mayer, ratepayer; it exists unexpressed between words like he ambles, she-ass.

ny is the equivalent of the Spanish ñ or the individual consonant sound represented by n in new, nude, by ni in pinion, onion, by gn in vignette.

(c) Dental class.

ch nearly as in chat, chisel, channel, but really an affricate, where the tongue stops and then glides, while in English it is fricative, produced by a mere glide of the tongue.

j nearly as in Jenny, jump, but like Malay ch a dental, and not a palatal as in English.^[10]

s a superdental as in sister or as in hiss rather than his.

d a superdental as in plunder, binding, landing.

t a hard superdental identical with the initial t in topple, tort, Tom. [ 31 ] n ‘may be compared with the superdental represented in English by the same sign when written before a d though it is a little more distinctly pronounced, i.c. more with the top of the tongue’.

l not quite identical with the English sound. What the English write with the same sign seems very often to be a superpalatal (lingual or cerebral), i.e. a consonant formed at the higher part of the palate with the tongue turned backward, the top pressed with its full breadth against the roof of the palate. The Malay l is produced at the lower part of the palate like what the Dutch write l, when pronounced by well-bred people.' Elsewhere, Dr. Fokker compares it with l in hill-top not in hill.

(d) Labial class.

w (not expressed in the system of romanizing employed in this grammar) as in coward, power, sower; the semi-vowel unexpressed between words like rue it or in a word like dual.

b as in English boot, bean, tub.

p as in English: pig, pup, pant.

m as in English.

§ 10. Vowels.

Broadly the vowels in Malay are:

a as in langit, api, kayu, ratus.
e as in bela.
i as in ingin.
o as in gopoh.
u as in kayu.
ě as in ĕnam.

and these are the only vowel sounds which will be distinguished in the body of this grammar, though a work on phonetics would make further distinctions and mark them by means of accents. The problem of definition is increased by the great [ 32 ]differences in Malay dialects, of which for the Peninsula at least no exhaustive study has been made. Dr. Fokker, who wrote primarily of pronunciation in West Borneo and who moreover, I believe, has since modified his views, distinguished in each of the first five vowels three variations of sound;

(a) Long and clear. a as in Ah: ragut, jadi, chabut, sabut, batu. e like ea in swear: beta, merah. i as ee in week: siram, kirim. o as in bore: oleh, orang, tolak. u like oo in moor: kurang, surut.

(b) Short and clear. a nearly equivalent to u in cup: ikat, isap, kĕrap. e as in padre or nearly like i in kitchen: oleh, ekur, pĕrentah. i as in German Sinn or almost like ee in been: pipi, manis, tapis. o as in pillow: taroh, jatoh, bongkar, pohon. u as in put: pulang, siku.

(c) Short and toneless surd. a as in canto or nearly like o in some: lĕnyap, kurap, lapis, banding, ganti. e as in German Brell or shorter than ai in said: oleng, chĕbek, ejek. i as in kitten, sing: sĕring, bĕtis. o as in German Gott or nearly as o in lot: bohong, komeng, korek. u in German Butter, dumm or the French pronunciation of géranium, album: těrus, takut, kapur, subur.

Some of these examples at any rate do not fit with the pronunciation of the Peninsula. It is easy to exaggerate the long clear a; and jâdi, pâdi, pâdam, châbut, are not heard in the Peninsula. Again, the distinction between a (b) and a (c) would seem to be merely that in unaccented syllables the vowel is less definite that in accented: and this fact leads to the blending of the sounds represented by i and e, by o and u, § 20 (c); e.g. gosol or gesil, ekor or ekur, a blending never found in accented syllables. Moreover, elaborate as it is, Fokker's analysis certainly fails to exhaust all the sounds in Malay the first o in bodoh is closer than o in bore but longer than o in pillow; and merah has an e more open than that in beta. [ 33 ]§ 11. Diphthongs.

ai as in German Kaiser or like i in light, but sometimes shortened in unaccented syllables till it approximates to ai in maid.

au as ow in cow, but sometimes shortened in unaccented syllables till it approximates to ow in know.

§ 12. In the Peninsula, there are great differences between Riau-Johor and Kedah pronunciation and even between the pronunciation of each separate state. And these differences require exhaustive treatment, before the pronunciation of the Peninsular Malay can be profitably discussed. Moreover such discussion would be beyond the scope of a grammar. The Peninsula is fortunate in that ‘the language of Malacca (Riau-Johor) played a prominent part in fixing the κοινή διάλεκτος of the golden age’ and that its pronunciation is roughly in accord with classical spelling.

§ 13. Foreign sounds.

Malays have attempted in Arabic and other foreign words to imitate alien sounds, and they have incorporated Arabic letters representing rare sounds in their alphabet, though they have seldom conserved the right and original pronunciation. These are as follows:

Tha ث th as in thin, but commonly corrupted by Malays into an s sound: thalatha pronounced as sĕlasa, ithnain as isnain.

Ḥâ ح a strong aspirate, but not distinguished by Malays from the soft ھ, except that unlike the soft aspirate, it is never dropped out in spelling, e.g. huruf never uruf, hukum never ukum.

Khâ خ a hard guttural like ch in German or in the Scotch word loch, commonly corrupted by Malays to k.

Dzal ذ pronounced by Malays as dz, z, or j. [ 34 ]Za ز = English z, e.g. zaman. But it is often corrupted by Malays to j, e.g. pauh janggi for pauh zanggi.

Shin ش = sh, but often corrupted to s.

Ṣâd ص a very strongly articulated s popularly pronounced as an ordinary s. But educated Malays pronounce this and the four following letters very gutturally. This habit gives a throaty sound to the accompanying vowel, while leaving the consonant practically unaffected.

Dlad ض a sort of aspirated d pronounced something like the combination dth. Educated Malays pronounce it as a soft th: retha, kathi, but the ignorant as l: rela. Sometimes it is pronounced as d: fěduli. Its common romanized form dl as in redla, hadlir combines the Arabic d and the Malay l.

Ṭâ ط a strongly articulated palatal t, but Malays pro- nounce it as ordinary t.

Tlâ ظ a strongly articulated palatal z, but like dlâd pronounced by Malays as l or dh: lohor, dhohor.

'ain ع a strong guttural commonly pronounced by Malays like an alif: adat. In the middle of a word its presence is often marked by doubling the vowel or by a pause: maana, ma’na, but not always, mana also being heard.

Ghain غ pronounced by Malays like r with a burr: e.g. ghaib as raib, mashghul as mashrul.

Fa ف = f, often pronounced by Malays as p, e.g. paham, arip.

Ḳâf ق a deep faucal k, but commonly pronounced by Malays as ordinary k. The letter is wrongly used (but not sounded as a ق) in Sanskrit words, e. g. لقس laksa.

§ 14. Accent.

The need to discuss accent tends to give an exaggerated idea of its importance in Malay. It is necessary to bear in mind that there is no strong accent on any syllable in a [ 35 ]Malay word that words like pěrkataan, pěrbuatan, aluran, kěděngaran, di-katakan, for example, are pronounced practically with the same stress on every syllable.

Ordinarily in the Malay word, the accent falls on the penultimate except that

(1) when the penultimate is ĕ in an open syllable and rarely in a closed, then the accent falls on the last syllable, ěnám, těngáh;

(2) when a derivative is built up by prefixes from a monosyllabic root, the accent sometimes remains on that root, namely, on the last syllable;

(3) in the vocative, the stress is sometimes thrown on the last syllable.

The case of words built up of suffixes presents a special problem. The Arab system used the huruf saksi و ا and ي to indicate quantity, and in its application to Malay treated accent as the equivalent of quantity, putting the huruf saksi in accented open syllables. The early spelling has left this mark of accent in the penultimate of derivatives ڤرکتأن pěrkatáan from کات káta, چمبرون chěmburúan from چمبور chěmbúru, کتهوي kětahúï from تاه tahu, and even جديکن jadíkan from جاد jádi, کتاڽ katá-nya from کات káta, کداڽ kudá-nya from کود kúda, and so on-except in the case of the particles lah, tah, kah. The Indonesian rule is that the accent falls on the penultimate whether of simple or of derivative words. The general opinion of Dutch scholars, from Werndly down to Tendeloo and Fokker, has insisted on the same rule holding good of Malay. Two notable exceptions, however, are von de Wall and Gerth van Wijk. Only the scientific study of dialect throughout the Peninsula and Archipelago could explain fully how experts have come to differ. In the Peninsula I confess I had supposed in common with Europeans who have lived there a quarter [ 36 ]of a century that the Malay had generally gone back on the old Indonesian rule. But special observation for the purposes of this work has led me to revise my opinion, and to think that while practically there is hardly any accent at all in the words in question, still the Malay does say pěrkatáan, ingátan, kudá-nya, namá-nya, and jadíkan—though the suffix kan has not this shifting influence when the stem ends in a consonant, and tímbangkan, támbatkan will be correct. [ 37 ]

CHAPTER III

SPELLING

§ 15. The following are the letters of the Malayo-Arabic alphabet with their Roman equivalents. The writing, of course, runs from right to left. The form of the letters differs according to their position and their connection with other characters; some letters never connect with others to the left of them, and therefore are found sometimes in isolation; both of these points are shown in the table.

Letter.	Unconnected.	Conventional Roman Equivalent.	Form in connection.			Examples.
Letter.	Unconnected.	Conventional Roman Equivalent.	With following letter.	With preceding letter.	With both.	With following letter.	With preceding letter.	With both.
Alif	ا	…	…	ـا	…	…	دفا	…
Ba	ب	b	بـ	ـب	ـبـ	بلت	ترتيب	ببل
Ta	ت	t	تـ	ـت	ـتـ	تليڠا	بولت	تتق
Tha	ث	th	ثـ	ـث	ـثـ	ثلج	حديث	عثمن
Jim	ج	j	جـ	ـج	ـجـ	جاري	ثلج	تنجڠ
Cha	چ	ch	چـ	ـچ	ـچـ	چاري	ڤنچ	ڤنچڠ
Ha	ح	h	حـ	ـح	ـحـ	حال	صح	ضححى
Kha	خ	kh	خـ	ـخ	ـخـ	خبر	شيخ	تخت [ 38 ]
Dal	د	d	…	ـد	…	احد	…	…
Dzal	ذ	dz	…	ـذ	…	عذر	…	…
Ra	ر	r	…	ـر	…	لنتر	…	…
Za	ز	z	…	ـز	…	فلزم	…	…
Sin	س	s	سـ	ـس	ـسـ	سيكت	تبس	بسر
Shin	ش	sh	شـ	ـش	ـشـ	شهدان	بخشيش	شمشير
Sad	ص	ṣ	صـ	ـص	ـصـ	صحبت	قميص	فصل
Dlad	ض	dl	ضـ	ـض	ـضـ	ضرورة	فرأيض	فضولي
Ta	ط	ṭ	طـ	ـط	ـطـ	طبيب	غلط	شيطان
Tla	ظ	tl	ظـ	ـظ	ـظـ	ظهر	محفظ	عظمت
Ain	ع	‘	عـ	ـع	ـعـ	عدل	طمع	فعل
Ghain	غ	gh	غـ	ـغ	ـغـ	غأيب	بالغ	صغير
Nga	ڠ	ng	ڠـ	ـڠ	ـڠـ	ڠري	ڨگڠ	بلڠكس
Fa	ف	f	فـ	ـف	ـفـ	فصل	شريف	عفريت
Pa	ڤ	p	ڤـ	ـڤ	ـڤـ	ڤوسيڠ	سليڤ	تمڤت
Ḳaf	ق	k, q	قـ	ـق	ـقـ	قدرت	فلق	فقير
Kaf	ک ك	k	کـ	ـک ـك	ـکـ	كنا	نأيك	فكين [ 39 ]
Ga	گ	g	گـ	ـگ	ـگـ	ݢلير	رڠگ	ڤگڠ
Lam	ل	l	لـ	ـل	ـلـ	لنتق	تبل	بلم
Mim	م	m	مـ	ـم	ـمـ	مالس	ليم	لمبق
Nun	ن	n	نـ	ـن	ـنـ	نيلي	ايكن	نند
Wau	و	u, w	…	ـو	…	…	ڤيلو	…
Ha	ھ	h	ھـ	ـە	ـھـ ـہـ	هارس	لمبه	بهو
Ya	ي	y	يـ	ـي	ـيـ	يا	سقسي	تيڨو
Nya	ڽ	ny	ڽـ	ـڽ	ـڽـ	ڽات	بتلڽ	ڨڽو

Of these letters چ ch is borrowed not from the Arabic but from the Persian, as also is the form گ g for which ك simply is often written: ڤ p is not Arabic. ڠ ṅ or ng, پ (or ڽ) ñ or ny appear to have been constructed from غ and ن respectively. The fourteen letters already dealt with (§ 13) represent Arabic sounds and occur only in Arabic loan-words, except that, as stated in the aforesaid section, guttural q ق is employed as a symbol in Malay words. ة (ta bersimpul, as Malays term it) is used often by Malays for final t, e.g. اية itu, سورة surat, that is, in accordance with Persian and Hindustani usage, but wrongly, of course, according to Arabic usage, since in that language it is a variant of ه and called ha-ta, being pronounced like la only when followed by a vowel. ف fa is commonly used by Malays for ڤ and pronounced [ 40 ]as p. In Malayo-Javanese works ڎ represents a palatal d found in Sanskrit and in Javanese, though not elsewhere in Malayo-Polynesian languages; it is exotic and not employed now. لا lam alif, a combination of two letters, is sometimes regarded as a separate letter of the alphabet. So too is hamzah ء.

§ 16. Besides the alphabet, Malays have borrowed from the Arabs certain diacritical signs, most of which have become almost obsolete in their writing. These are:

(1) The vowel points, in Arabic called harakat, in Malay baris or sěnjata, which represent short vowels and when followed by ي ,ا and و become long vowels.

Fathah or baris di-atas ــَـ = short a or if followed by alif long a.

Kasrah or baris di-bawah ــِـ = short e or i or if followed by ya, long e or i.

Dlammah or baris di-hadapan ــُـ = short o or u, or if followed by wau long o or u.

These vowel points have fallen into disuse, being replaced against Arabic usage by huruf saksi (§ 18 (d)).

(2) Hamzah ء is found in Arabic at the commencement of word or syllable with alif as a prop; alif by itself having no sound except that after a consonant it serves to prolong the vowel fathah: this use of hamzah is not practised in Malay. For its Malay use sec § 18 (e).

(3) The jazm ــْـ which shows that the consonant over which it is placed closes the syllable and does not begin a fresh one: for example, placed over kh in bakhshish it signifies that the word is pronounced bakh-shish and not bakhěshish.

(4) The tashdid ــّـ : see § 18 (e).

§ 17. From the evidence of the earliest Malay manuscripts extant, it is clear that there was a fixed standard for the [ 41 ]spelling of Malay in Arabic characters, at the very beginning of the seventeenth century, a standard obtaining in many different places in the Malayan Archipelago. The introduction of this foreign alphabet was a direct consequence of the conversion of Malays to Islam. The earliest and most important missionary centre was N. Sumatra, which strictly was not Malay in speech, though Malay was used for commerce, literature, and religion; most of the old Malay MSS. were written there and von de Wall alludes to eja acheh Achinese spelling as the original style. A system of spelling there adopted naturally would spread with the spread of Islam to the rest of Sumatra, the coasts of Borneo, the Moluccas, to Malay settlements in Java and at Malacca. But it is possible that something more than repetition and imitation went to account for the uniformity of system. ‘If the Arabs had attempted to make an adaptation of their own system of spelling to suit the peculiarities of the Malay language, the result would have been that in different parts of the Archipelago there would have been different modifications of the Arabic spelling, and a variety of Malay spellings would have been unavoidable. The uniformity in the spelling of the earliest MSS. would lead us therefore to expect that the system of orthography according to which the Arabs originally began to write the Malay language and which they taught subsequently to the Malays, was the same as they themselves used in writing their own language.’ Certainly in the main they did attempt to apply Arabic principles.

The notes of the early seventeenth-century system^[11] were:

(a) The use of vowel points: at any rate they were used on unusual words at their first occurrence in a work, so that بَدَنُل Badanul, a proper name, is fully vowelled on its first mention in the Bodleian Sri Rama but not subsequently. [ 42 ](b) The tashdid is inserted to indicate that the letter over which it is placed, whether vowel or consonant, is sounded twice, e.g, u as uw, i as iy, s as ss, ng as ngng, and so on.

بوّت buwat, دوّ duwa, ديّم diyam, ايّ iya, سيّڠ siyang, سودّه suddah, بسّر bĕssar, تڠّه těngngah.

But the use of the tashdid to double the consonant, i.e. when it follows a short vowel (as in our words better, fellow, galant or gallant) is not generally observed in Malay MSS. and is not Arabic, nor does it represent any real phonetic doubling of the consonant. (One is tempted to compare with it the double kk in such words as فرتنجقكن pĕrtunjokkan, which are often said to be in imitation of Javanese double-letter forms and to exhibit a survival probably due to the existence of two forms of k; but Javanese influence hardly existed at all in Acheen, and it is more probable that the first k ق never represented anything but the mere glottal check (§ 9).)

(c) The omission of final و ا and ي.

كاي kayu, اك aku, ترلال terlalu, ڤنتpintu, مريب měribu, برتم bĕrtěmu, اڤ api, جاد jadi, مات mati, لاك laki, ڤرگ pergi, بوم bumi, هات hati, سكل sa-kali, سوك suka, دو dua, توه tuha.

Exceptions:

(1) A final vowel is inserted when required for the explanation of a foreign word. In the Bodleian Sri Rama the Sanskrit puri is written sometimes ڤور but oftener ڤوري and pĕrkasa is written فركشا.

(2) It is inserted when ا represents a uniform long a, and ي, و the diphthong sounds au, ai.

کرا kěra, بلا běla, دڤا děpa, کيلو kilau, ريسو risau, توڤي tupai, هلي hělai.

(d) The omission of any symbol to represent the indeterminate short vowel ĕ unless the tashdid as used above (b) can be taken as an illegitimate symbol. [ 43 ](e) The omission of و ,ا and ي is as medial vowels in closed syllables.

برت bĕrat, بلم bĕlum, سبت sĕbut, اورڠ orang, کمبڠ kambing, رڠگت ringgit, بمبڠ bimbang, تمبه tumboh.

Exceptions:

(1) When the spelling of Arabic words is retained (though Arabic pronunciation may not be followed), e.g. اسلام islam, فقير fakir, کتاب kitab— in Arabic of course one would get kitabu, fakiru, &c., and the syllables would not be closed.

(2) The monosyllables دان dan and ڤون pun.

(f) و ,ا and ي and are inserted in an open syllable upon which the accent falls—in Malay usually the penultimate.

تون túan, روس rúas, کڤيتڠ kĕpíting, تليڠا tĕlínga, سکدودق sĕkĕdúdok, لراڠن larángan, تمباڠن tambángan, رجراج raja-rája, اڤبيل apa-bíla, هلبالڠ hulubálang.

Exceptions:

A few words like مك máka, فد páda, در dári, سده súdah, سگل ségala, in which the vowel may have been omitted because it was short and had little stress on it.

[In certain derivative words formed by the addition of suffixes, و ,ا and ي are shifted to the penultimate of the derived word: ايڠت ingat, ايڠاتن ingatan, چمبور chěmburu, چمبرون chěmburuan, تاه tahu, کتهوي kětahuï, جاد jádi, جديکن jadikan, کود kuda, کداڽ kuda-nya, کات kata, کتاڽ kata-nya. Many have thought that the accent does not now, in the Peninsula at any rate, fall on that syllable, but still remains on the penultimate of the root word. Consequently it has been suggested that this shifting arose from vicious analogy with Javanese forms and phonetics, which seems improbable considering the small influence Java had in the north of Sumatra: or that it might have arisen from vicious analogy with the spelling of Malay roots, but that again seems very improbable. [ 44 ]As I have stated on p. 32, it is really due to the prevalence even now of the old Indonesian system of accentuation (whereby the accent falls on the penultimate of simple and derivative words alike). Of course that accentuation may have been more pronounced at the time when the Arabic system of spelling Malay was introduced than it is now.]

(g) Reduplication of words was cominonly signified by the Arabic cipher ٢ 2 angka dua after a word, a symbol not so used in its native tongue: راج٢ raja-raja, بركات٢ běrkata-kata.

§ 18. The modern spelling of Malay in Arabic characters differs from the script of three centuries ago in several ways.

(a) It never uses vowel points except over words quoted from the Arabic or over foreign and ambiguous words.

(b) It has dropped the tashdid.

(c) It employs و ,ا and ي as finals, even when these final vowels are neither long vowels nor diphthongs. کايو kayu, اکو aku, ترلالو těrlalu, ڤنتو pintu, مريبو měribu, اڤي api, جادي jadi, ماتي mati, لاکي laki, دوا dua, توا tua.

Exceptions:

(1) A few stereotyped common words preserve the old style: ايت itu, اين ini, سوات suatu, سڤرت sěpĕrti, اد ada, اڤ apa, اي ia, دي dia, کيت kita, جوك juga, ڤول pula, ڤد pada, سرت sěrta, مان mana, منسي manusia most finals in a.

(2) The change in the use of a is not established, the old practice of omission being followed or neglected arbitrarily according to the will of the writer in the spelling of most words. [ 45 ](d) و ,ا and ي is are used more and more as medial vowels in closed syllables—except again in the case of some common stereotyped words like منتا minta, مليا mulia, ايسق esok, راتس ratus. ‘The use of these letters و ,ا and ي and the huruf saksi, as they are called, is opposed to Arabic alphabetical theory. The syllable should consist of two letters; the introduction of a third,’ started even in our earliest MSS. ‘must have been a bold innovation … The innovation may have been to some extent countenanced by the disuse of case-endings in Arabic words such as kitab, islam, &c., but it was certainly developed, if not introduced by European influence. It must be remembered that until recently the printing of Malay books has been entirely in the hands of Europeans, especially in those of missionaries, and that the influence so exercised must have been great. The power of public instruction under European direction has also to be reckoned with … It may be predicted that if nothing is done to check existing tendencies the use of the huruf saksi as English vowels will extend to all words. This solution of the present system would not be an unsatisfactory one. It would certainly make Malay spelling consistent and easy.’ At present, the extended use of the huruf saksi is commoner among police and clerks trained on European lines than among pundits proud of a little Arabic learning. The latter propound three theories, which pretend to be practical rather than scientific but fail even of their limited object:

(1) و and ي and should be inserted in closed syllables, except that when the vowel sounds of the two syllables are alike, the و or the ي should be inserted only in one of the two syllables: بيڠوڠ bingong, بوتير butir, کوتيڤ kutip, کمڤوڠ kampong, تنجوڠ tanjong, بيمبڠ bimbang: but تڠگي tinggi, دنديڠ dinding, بورڠ burong. Apply this theory to the case of tulang bones and tolong help or tunggal solitary and tunggul tree-stump; either the spelling of each pair [ 46 ]must be identical, viz., تولڠ and توڠگل or else the well-established and convenient use of the huruf saksi to represent the accented vowel, i.e. the penultimate, must be abandoned. The theory is artificial, arbitrary and impracticable.

(2) Another theory would revert for guidance in the employment of the huruf saksi to the same system that determined the use of the vowel points in the seventeenth century; namely, for the determination of vowel sound in rare words. It has historical basis, but what are rare words? For the pundit, simple terms of husbandry; for the peasant, the Sanskrit and Arabic loan-words of literature.

(3) A third theory would employ huruf saksi to distinguish words like tolong and tulang which otherwise would be spelt alike. This might be a serviceable empirical device, but strict scientific uniformity would entail encyclopaedic knowledge of every word in the language.

(e) Modern spelling has adopted hamzah ء, which is rare in early MSS., to indicate:

(1) As in Arabic a break between two vowels, of which the first is a and the second begins with the spiritus lenis, or in other words when an a sound passes on to another vowel only by means of a gentle aspirate—it never marks a break between vowel and semivowel. For instance کايّل = kai-yil, کايل = kail a diphthong; with hamzah کأيل = ka'il; and so too with لأين lain and لأوت laut, though such words are commonly written without hamzah. سئيکور sa-ekur, سئورڠ sa-orang, سئيکت sa-ikat, کألتن kěalatan, ککيأن kěkayaan, کأنم kěěnam, يأيت ia itu, نمأي namaï—it must be noted that when a word beginning with alif is preceded by sa the alif of that word is omitted.

(2) As an alternative to ق, indicating the glottal check, تيرؤ tirok, داتؤ datok, انچئ inchek. [ 47 ]§ 19. The following combinations of words are found:

(a) sa is joined to the following word سريبو sa-ribu, سسيکت sa-sikat.

(b) The prepositions di and ka are similarly joined کرومه ka-rumah, دتڠه di-těngah.

(c) And so too, the unemphatic short forms of the personal pronouns ك ku and کو kau, کليهت ku-lihat, كليه ka, كوباوا kau-bawa.

(d) م mu, ڽ nya (and ku and kau when they belong to a preceding word) are joined to the word preceding them; also daku, dikau and dia after akan and děngan. اولهم oleh-mu, باوهڽ bawah-nya, ديريكو diri-ku, اكندي akan dia, دڠنديکو děngan dikau.

(e) The particles lah, kah, tah, pun are joined to the preceding word, الڠکه alang-kah, اڤته apa-tah.

(f) يڠ yang may be joined with the preceding or following word, or with both. اورڠيڠ orang yang, يڠتردافت yang terdapat, اورڠيڠددافتي orang yang di-dapati.

(g) The component parts of compound words are joined افبيل apabila, هلبالڠ hulubalang.

The impossibility of formulating a logical system of spelling shows the inadaptability of Arabic characters for the expression of Malay phonetics. But the above general principles, such as they are, are better guides than merely empirical rules which are broken through and through by exceptions.

§ 20. Romanized Malay.

Two main systems of rendering Malay in Roman characters have been practised:

(1) The old Dutch method of reproducing the Arabic letters in Roman equivalents; the indeterminate vowel represented generally by e (and according to Dutch spelling, ancient and modern, u represented by oe, y by i, j by dj, ch by tj, kh by ch). [ 48 ] (2) The newer phonetic system, which disregards the fact that the language has been written in Arabic characters and seeks to reproduce the pronunciation of the educated Malay; a system which has been followed with varying degrees of failure and success by amateur native-born Chinese, by early voyagers, by the missionary press, by writers with an eye to popular needs and by scholars aiming at scientific phonetics.

The old Dutch method has been abandoned even by the Dutch in favour of the later system, which alone demands attention.

On the general principle of the modern system, there is nothing to add to what Dr. Snouck Hurgronje has written. ‘A purely phonetic system, to serve scientific purposes, inust be bound to one dialect and give a perfectly true image of its phonetic peculiarities. The semi-phonetic system that will fulfil practical requirements ought to give the average pronunciation of educated genuine Malays. So the texts of the most famous literary productions of the golden age of Malay civilization (sixteenth century A. D.), so far as they have not suffered from corruption, give us an image of the average pronunciation of Malay at that time. But this image is very imperfect, especially for the vowels, owing to the peculiar use of the Arabic alphabet; and besides that, the pronunciation of Malay has suffered many changes during the centuries which have elapsed since the gradual decay of Malay kingdoms. As political and literary centres lost their significance, dialects got increasing influence upon Malay style and orthography; the spreading of Malay over parts of the Archipelago widely distant from each other and without frequent intercourse, was very disadvantageous to the conservation of the unity of the written language andmade many dialects develop almost to separate languages. Our own period with its rapid means of communication makes the need of a certain uniformity in written Malay more. deeply felt than before. We cannot content ourselves with [ 49 ]attempts to restore the orthographical principles of three centuries ago, as (a) we have to make use of the Roman alphabet which is more suitable for expressing Malay vocalism and more in accordance with actual requirements of conversation than the Arabic; (b) the pronunciation of that time is very imperfectly known to us; (c) it has everywhere changed a good deal. So a mean will have to be sought between two extremes. The language of Malacca (Riau-Johor) having played a prominent part in fixing the κοινὴ διάλεκτος of the golden age, and so having left considerable traces in what has remained of the old uniformity of civilized speech, may be given more authority than most of its sister dialects, but these last nevertheless have to be taken very seriously into account. The local diversity of vocalism must keep us from marking too fine distinctions of sound and from fixing in script nuances which in a considerable part of Malaya are not observed.’

To put this general principle into practice:

(a) How is the Malay indeterminate short vowel to be represented? Some old writers used a: bassar great, marampas seize; Crawfurd affected â—bârchârai divorced; Marsden and older Dutch scholars e, besar. The missionary press in Singapore wavered between omitting it—bsar—and inserting ’ or ′—b’sar, b′sar. Swettenham, van Wijk and modern continental scholars generally have preferred ĕ—běsar. Now a and e represent its sound neither to the Malay nor to the foreigner; and e will be confused with e in words like beta and besan. Use no symbol and an agglomeration of consonants—kbsaran—results. Use ’ or ′ and ĕngkau becomes ’ngkau and kěĕnam becomes k’’nam! Obviously ĕ best represents its sound and is no harder to write than a dotted i or a French accent or the German diaeresis.

(b) How are ڠ ng as in دڠن dengan, اورڠ orang, and ڽ ny as in ڽاموق nyamok, باڽق banyak to be written? The scientific [ 50 ]Orientalist, especially the student of Sanskrit and Arabic, will of course shudder at anything but a scientific symbol; he will demand, for instance, ṅ or n̄ for ڠ and ñ for ڽ. But the student of Malay is in rather a different position. The great bulk of books on Malay, whether in the Straits Settlements or in the Netherlands Indies, have been printed by local presses, which have had no access to recondite symbols. Native readers of Romanized Malay abound; they have become used to the more slipshod way and cannot take kindly to change. And the student of Malay has to consider not only the wants of the native reader but the practice of Dutch scholars of high repute who outnumber the Englishman by twenty to one. The system adopted by the Dutch Government for the large literature of the Netherlands Indies and by Dutch scholars writing in Holland is not to be lightly regarded by other students. It is desirable that there shall be uniformity as far as possible; and it is futile for the solitary writer on Malay to pit the international scientific system against the weight of past prejudice and the current usage of two governments—for after careful consideration the Government of the Federated Malay States^[12] decided to follow the Dutch line of expediency against scientific perfection. Considering that Malay is a living language of great vitality, area, and adaptability, read and written in Romanized form by children in village schools, by Straits-born Chinese, Tamils and Eurasians, by immigrants Asiatic and European; a language moreover with thousands of living Malays to guide one in pronunciation; considering this, the use of symbols is certainly impracticable and perhaps not indispensable to scholarship. ng and ny never represent divided sounds in Romanized Malay but always the letters ڠ and ڽ, so that their use need lead to [ 51 ]no difficulty. The Dutch Government, the English Government, and most Dutch^[13] and English scholars have accepted this convention and it will be employed in this work.

(c) The same argument of expediency has swayed the Government with regard to letters representing exotic Arabic sounds. ه and ح are both represented by h, not by h and ḥ; ت and ط by t, not by t and ṭ; س and ص s, not by s and ṣ; ض ذ and ظ not by dz, dl and tl but all by dz; ز by z; ع and the hamzah both by an apostrophe; غ by gh; ث by th. Persian and Arabic words of established vogue in the Malay language have lost their native pronunciation and are spoken in Malay fashion, so that the Arabic double letters no longer represent differences in sound. Unusual literary and religious words anyhow require knowledge of Arabic on the part of Malay and European, if they are to be pronounced at all: and such readers will recognize them readily even in simplified Roman dress—especially as they will otherwise observe the foreign spelling; for example, khattu’l-istiwa. The spelling of exotic words is of very minor importance. The symbol for Malay final ق representing a glottal check is a more difficult matter. The q of scientific scholarship represents the Arabic ق in qudĕrat, and distinguishes it from ك k, but it does not represent the Malay final as in بورق burok. Spat uses the other Malay convention (§ 18 (e)) the hamzah and prints buro^ء which has a strange hybrid appearance. The Dutch and English Government spelling uses k alike for ك and ق. On the whole, this is as logical as any other method, seeing that the glottal check is represented in the Malay convention now by ق now by ك and now by hamzah—ڤاتيك, ڤاتيق, or ڤاتئي. k used for final need cause no difficulty, if it is remembered that k at the end of a Malay word always stands for the glottal check: ق elsewhere in [ 52 ]a word occurs only in a few Sanskrit and Arabic words which will be easily recognized.

(d) The representing of ي and و by i or iy and u or uw is debatable. Mr. Wilkinson^[14] has put the scholar's point of view succinctly: ‘The popular spelling of tuwan as tuan and of iya as ia … suggests that the second syllable … commences with the spiritus lenis when really it begins with a w or a y. It gives an incorrect idea of the value of alif by confusing the hamzated alif with the alif of prolongation. It is also faulty for etymological study; for instance, watu is the Javanese form of batu, and the bearing of this fact on the derivation of suwatu is not suggested by the spelling suatu which implies that atu and not watu is the second portion of the word. These points may seem trivial; but the confusion imported into Malay spelling is in great measure due to the inadequate comprehension of the alphabet fostered by the belief that the spelling tuan is not a mere approximation to the sound of a certain word but is an exact transliteration of certain Arabic letters.’ The popular spelling has been followed by Mr. Wilkinson in his smaller dictionary, and by the government committee. And it can be defended from the point of view of pure phonetics. No system of spelling derivatives can be formed solely with an eye to the requirements of etymology. Phonetically, a w slips in automatically when one says tuan, unless one deliberately makes the effort to say tu’an: and the spelling tuwan is calculated to suggest a more distinct w sound than Malays actually enunciate.

(e) The most difficult problem of all is the choice of the Roman vowel, when the Malay sound lies between o and u, between e and i. The F.M.S. Government, after consideration of Peninsular dialects and of the Dutch system, has [ 53 ]accepted the following convention for the selection of the vowel in final syllables:

(1) ong not ung; ch not uh; ok not uk; um not om; un not on; ul not ol; up not op; us not os; ur not or; ut not ot; u not o.

(2) eh not ih; ek not ik; ing not eng; im not em; in not en; il not el; ip not ep; is not es; ir not er; it not et; i not o.

(3) But if the penultimate vowel is e or o, the final should contain e or o in preference (o i or u—pohon not pohun, kotor not kotur, gesel not gesil, dongeng not donging, choket not chokit. ‘The fact is that in the Riau-Johor dialect the two vowels in question approach the sound of o and e, while in the Kedah dialect they approach that of u and i. Consequently there is a certain affinity between o and e and between u and i; were the spelling otherwise, it would suggest that the Riau pronunciation was accepted for one half the word and the Kedah pronunciation for the rest.’

§ 21. So much for general principles. The recommendation of the Government committee on the special case of certain classes of word will also be observed in this grammar. That committee decided as follows:

(a) Foreign words ‘magistrate’ ‘court’ khattu’l-istiwa shall pace the qualification in § 20 (e) supra be spelt as in their original language.

(b) Trisyllabic words where the first two syllables are separable by an h, بهارو ,سهاج ,سهيا, and so on, shall always be written with h between two letters a—sahaya, sahaja, baharu.

(c) In ‘the case of words like kěmudian, děmikian, and arakian, which are often pronounced kěmědian, děměkian, and arěkian, we are of opinion that there is sufficient variety [ 54 ]in local practice to justify the adoption of that form which lends itself best to the explanation of the composition of a word, especially as such a decision will tend to bring British and Dutch methods of Romanizing more closely into harmony. In the case of derivatives of combinations such as the word měnyabělah from sa-bělah, we consider there is no sufficient reason for treating ordinary derivatives of sa as distinct words with distinct spellings of their own, and we recommend that the a of sa be retained’ (i.e. měnyabělah and not měnyěbělah). At the same time, a distinction may be fairly drawn between the words keluar to go out and ka-luar outwards, though etymologically they are the same; and we recommend that where a derivative form has become specialized in use, it be written as a separate word.

(d) ‘The use of hyphens should be restricted as much as possible. Derivatives (such as měnyakiti from sakit) should not be divided up. But prepositional and other particles may be separated by hyphens from the word to which they are accretions: thus sa-kali pun, ka-rumah, di-dalam-nya.’ [ 55 ]

CHAPTER IV

THE MALAY WORD

§ 22. The Malay word may be:

I. Simple,

api fire, běsar big, chěkek strangle, tiga three, kurang less.

II. Derivative, i. e. built up by

(a) affixation (chapter vi):

běrapi fiery, těrběsar very big, měnchěkek strangling, kětiga third, těrkurang much less.

(b) reduplication (§ 63):

api-api mangrove, běsar-běsar fairly great, chěchěkek a ‘yanking’ noose, tiga-tiga three together, kurang-kurang at the lowest.

III. Compounded (§ 65): where compounded the words acquire a conventional meaning that would not belong to them taken separately. kayu-api firewood, orang běsar chief, chěkek kědadak violent strangling, vomiting, tiga-ratus three hundred, kurang akal stupid, rumah tangga wife.

Of course, not every simple word will undergo affixation, reduplication, and compounding. Some that take affixation are not reduplicated; some that are reduplicated are never compounded; some always remain simple.

§ 23. Often the Malay word cannot be assigned definitely to any one of our parts of speech. No hard and fast line exists between the radical used as substantive and the radical [ 56 ]used as adjective,^[15] for example: rumah běsar a large house; běsar rumah the size of a house; orang banyak many folk, banyak orang the number of folk; orang pandai a clever person, pandai běsi a blacksmith. sědikit few, běbĕrapa several, sakalian, sěmua, sěgala all stand sometimes before, sometimes after the substantive; in the first case they must be parsed as substantives, in the second as adjectives. Classical usage may tend to give a word currency rather as substantive than adjective, or as adjective than substantive, but often it cannot extinguish its essential versatility. In the conversational prose of Munshi Abdillah we constantly find such examples as pĕrgi main ka-panas go playing in the heat; mĕnurut adat dan bodoh orang following the customs and folly of men, where classical usage would prefer panas and bodoh to be parsed as adjectives, but where the chipped popular phrase of the writer has caught the historical as well as the living genius of the language.

Similarly substantive and verb are not always rigidly distinguished:

OXFORD: HORACE HART M.A.

PRINTER TO THE UNIVERSITY

↑ Mr. Blagden tells me that in Talaing pa = to do, to make. In view of Schmidt's synthesis, this seems worth recording.
↑ Like the other Indonesian articles a and ra, which are held to survive as prefixes and suffixes in such words as the Malay anu, bunga from O. J. bung, ratu (Malay dato) from ra + tu master, so too i has been traced in Malay sigi from O. J. sig, tubi from tub, rugi from rug.
↑ Dr. Fokker remarks that all infixes seem to be nothing but shifted prefixes, found in words where the infixing conveys greater facility of pronunciation. So lě and rě interchanging with li and ri, are common prefixes lě-ng-kiang, re-ng-kiang rice-holder; ri-mau (the roarer) tiger, -ěm- will be identical with mě and -in- or -ing- with i + ng, i + n.
↑ In addition to the ‘simple’ affixes given above, there are a few others which occur in one and another of the Indonesian languages, but which in Malay, at any rate, need not trouble the grammarian. Two may be mentioned. In Fiji there is d-, Javanese j-, Malay j- or ch-, examples of which Kern detected in jěbul (Malay chabul) rape from bul a hole and jebur (Malay chěbur) plunge into water from the onomatopoeic bur. Relics of such a formative are clear in such doublets as tengok, jengok see; pijak, jijak tread; kangkang, jangkang a-straddle; abu, jabu (cp. děbu, lěbu) dust; lotong, jĕlotong monkey; alit, palit, chalit smear; bělah split, chělah crevice; těgang, chěkang laut; kebek, sebek, chebek awry. s which in a few Bugis words betokens reciprocity and reflexive action may be seen in the Malay kepit, sepit pinch (one thing by another); kilau, silau flash (flash after flash).
↑ Real ‘compound prefixes’ are those given in § 43 (II).
↑ From ajar are formed bělajar, pēlajaran; l taking the place of r. [With regard to the theory that the -ng and -r terminations of such prefixes as mĕng- and bĕr-, &c., are mere phonetic links, it is to be noted that (1) it is not supported by the results of the comparative study of the Indonesian languages; (2) it is in conflict with the data of the Kota Kapur inscription (probably of the seventh century A.D.), which contains such forms as mangujāri to speak with; marjjahāti to do harm to; parsumpahan curse, and the like. (See Part 67 of the Bijdragen tot de Taal-, Land- en Volkenkunde van Nederlandsch-Indië.) It is more probable that ng and r were originally separate formatives.—C.O.B.]
↑ Dr. Fokker calls my attention to the fact that Kern considered any prefix to the name of animal, plant, thing, &c., an indefinite article; he suggests, rightly to my mind, that such a prefix may more probably be defined as a definite article, and instances siakap, kakap name of a fish with protruding eyes, with which one may compare kakap a spy, a mata-mata.
↑ In this context may be cited some interesting points discovered by Mr. R. J. Wilkinson in his study of central Sakai, a Peninsular language representative of the Austro-Asiatic family, which in its vocabulary possesses old Indonesian words not known in Peninsular Malay to-day, and only to be paralleled in the vocabularies of languages in distant islands of the Archipelago. ‘The introduction of an infix (n, ’u, ěn, or ön) in central Sakai makes the word substantival; jīs dayligh, jěnīs a day, twelve hours; pāp fire-warmed, pěnāp the thing warmed; köh striking, kěnöh club, striker; ehok prod, stab, chěnok predder, spike. A prefix per turns the root into a verb or a passive root into an active root dat die, pěrdat kill; nong journey, pěrnong to go; löt extinguished, pěrlöt to put out (a fire); bet sleep, pěrbet close the eye. These two forms can be combined to form a verbal noun; dat die, pěrěndat murder; pěrěnglöt extinguisher; goï be married, pĕrgöï wed, pěrěnggöï marriage. In certain cases the final letter of a Sakai word changes to n, ng, or m. Sometimes this follows a law of euphony owing to the coincidence of two consonants: chip bird, chīmklāk hawk; klāk hawk, klāng-blok roe; chěrök long, chěröng-sok long-haired vampire. But there are cases where the alteration cannot be so explained; mai person, un mam one person, dök house, nu d’ngnön a house; rōk dart, nar r’ngnon two darts. The conjugation of verbs shows göi to be married, ’nggöi (I) am married, en ’nggöi I am married; běrsōp to feed, ’mběrsōp (I) am feeding. This system has notable points of resemblance with the Indonesian. The euphonic nasal reminds one of such Malay forms as sělang-sěli, bengkang-bengkok, golang-golek. There is one other peculiarity of (Southern) Sakai word formation that is paralleled in Indonesian idiom, for the building of polite and honorific doublets to common words. ‘Given a word in Javanese’, writes Mr. Blagden in The Pagan Races of the Malay Peninsula, ‘with an open penultimate syllable (and a final syllable preferably open and generally ending in a), to turn it into a Krama or high form, close both syllables with a nasal (or the final one more rarely with a liquid), modify the initial consonant of the final syllable to suit the nasal which now closes the penultimate and change the vowel of the final syllable (as a rule ě, or it may be a, i, or u).’ Mr. Blagden quotes from the Javanese kira, kintěn accounts; segara (Skt.), sěganten ocean; sore, sontěn evening; kalapa, karambil (cp. Malay gělambir) coconut. And he points out how it occurs in other Malayan languages without ceremonial or specialized meaning: dara virgin (Malay), dantěn virgin, of buffalo or hen (Sundanese); jalu male (Sundanese), jantan (Malay); alu and antan, Malay variants for pestle; pěmali and pantang, Malay variants for taboo. Cp. piama, piantan due season, esp. for rice-planting. And again, how it occurs in the aboriginal dialects of the Peninsula āsu’, anjing, nyang dog; puteh, pěntol white; serigala, sĕgala’, sĕranggil jackal; without definite evidence of ceremonial use, except that inost big animals have honorific synonyms.
↑ Taken from Mr. R. J. Wilkinson's Malay-English Dictionary.
↑ See Dr. Fokker's edition of Beech's Tidong Dialects (Clarendon Press, 1908).
↑ Cp. Shellabear's ‘Evolution of Malay Spelling’, J.R.A.S., Straits Branch, xxxvi. 75-135.
↑ See Romanised Malay Spelling, F.M.S. Govt. Press, Kuala Lumpur, 1904.
↑ Dr. Tendeloo accepts ng for ڠ but uses ñ for ڽ.
↑ Wilkinson's Malay-English Dictionary, vol. ii. p. 714.
↑ ‘Probably adjectives neither derived nor foreign are at bottom words denoting a subject, that is a quality, and so coming in time to denote the possession of that quality.’—TENDELOO.

[1] Mr. Blagden tells me that in Talaing pa = to do, to make. In view of Schmidt's synthesis, this seems worth recording.

[2] Like the other Indonesian articles a and ra, which are held to survive as prefixes and suffixes in such words as the Malay anu, bunga from O. J. bung, ratu (Malay dato) from ra + tu master, so too i has been traced in Malay sigi from O. J. sig, tubi from tub, rugi from rug.

[3] Dr. Fokker remarks that all infixes seem to be nothing but shifted prefixes, found in words where the infixing conveys greater facility of pronunciation. So lě and rě interchanging with li and ri, are common prefixes lě-ng-kiang, re-ng-kiang rice-holder; ri-mau (the roarer) tiger, -ěm- will be identical with mě and -in- or -ing- with i + ng, i + n.

[4] In addition to the ‘simple’ affixes given above, there are a few others which occur in one and another of the Indonesian languages, but which in Malay, at any rate, need not trouble the grammarian. Two may be mentioned. In Fiji there is d-, Javanese j-, Malay j- or ch-, examples of which Kern detected in jěbul (Malay chabul) rape from bul a hole and jebur (Malay chěbur) plunge into water from the onomatopoeic bur. Relics of such a formative are clear in such doublets as tengok, jengok see; pijak, jijak tread; kangkang, jangkang a-straddle; abu, jabu (cp. děbu, lěbu) dust; lotong, jĕlotong monkey; alit, palit, chalit smear; bělah split, chělah crevice; těgang, chěkang laut; kebek, sebek, chebek awry. s which in a few Bugis words betokens reciprocity and reflexive action may be seen in the Malay kepit, sepit pinch (one thing by another); kilau, silau flash (flash after flash).

[5] Real ‘compound prefixes’ are those given in § 43 (II).

[6] From ajar are formed bělajar, pēlajaran; l taking the place of r. [With regard to the theory that the -ng and -r terminations of such prefixes as mĕng- and bĕr-, &c., are mere phonetic links, it is to be noted that (1) it is not supported by the results of the comparative study of the Indonesian languages; (2) it is in conflict with the data of the Kota Kapur inscription (probably of the seventh century A.D.), which contains such forms as mangujāri to speak with; marjjahāti to do harm to; parsumpahan curse, and the like. (See Part 67 of the Bijdragen tot de Taal-, Land- en Volkenkunde van Nederlandsch-Indië.) It is more probable that ng and r were originally separate formatives.—C.O.B.]

[7] Dr. Fokker calls my attention to the fact that Kern considered any prefix to the name of animal, plant, thing, &c., an indefinite article; he suggests, rightly to my mind, that such a prefix may more probably be defined as a definite article, and instances siakap, kakap name of a fish with protruding eyes, with which one may compare kakap a spy, a mata-mata.

[8] In this context may be cited some interesting points discovered by Mr. R. J. Wilkinson in his study of central Sakai, a Peninsular language representative of the Austro-Asiatic family, which in its vocabulary possesses old Indonesian words not known in Peninsular Malay to-day, and only to be paralleled in the vocabularies of languages in distant islands of the Archipelago. ‘The introduction of an infix (n, ’u, ěn, or ön) in central Sakai makes the word substantival; jīs dayligh, jěnīs a day, twelve hours; pāp fire-warmed, pěnāp the thing warmed; köh striking, kěnöh club, striker; ehok prod, stab, chěnok predder, spike. A prefix per turns the root into a verb or a passive root into an active root dat die, pěrdat kill; nong journey, pěrnong to go; löt extinguished, pěrlöt to put out (a fire); bet sleep, pěrbet close the eye. These two forms can be combined to form a verbal noun; dat die, pěrěndat murder; pěrěnglöt extinguisher; goï be married, pĕrgöï wed, pěrěnggöï marriage. In certain cases the final letter of a Sakai word changes to n, ng, or m. Sometimes this follows a law of euphony owing to the coincidence of two consonants: chip bird, chīmklāk hawk; klāk hawk, klāng-blok roe; chěrök long, chěröng-sok long-haired vampire. But there are cases where the alteration cannot be so explained; mai person, un mam one person, dök house, nu d’ngnön a house; rōk dart, nar r’ngnon two darts. The conjugation of verbs shows göi to be married, ’nggöi (I) am married, en ’nggöi I am married; běrsōp to feed, ’mběrsōp (I) am feeding. This system has notable points of resemblance with the Indonesian. The euphonic nasal reminds one of such Malay forms as sělang-sěli, bengkang-bengkok, golang-golek. There is one other peculiarity of (Southern) Sakai word formation that is paralleled in Indonesian idiom, for the building of polite and honorific doublets to common words. ‘Given a word in Javanese’, writes Mr. Blagden in The Pagan Races of the Malay Peninsula, ‘with an open penultimate syllable (and a final syllable preferably open and generally ending in a), to turn it into a Krama or high form, close both syllables with a nasal (or the final one more rarely with a liquid), modify the initial consonant of the final syllable to suit the nasal which now closes the penultimate and change the vowel of the final syllable (as a rule ě, or it may be a, i, or u).’ Mr. Blagden quotes from the Javanese kira, kintěn accounts; segara (Skt.), sěganten ocean; sore, sontěn evening; kalapa, karambil (cp. Malay gělambir) coconut. And he points out how it occurs in other Malayan languages without ceremonial or specialized meaning: dara virgin (Malay), dantěn virgin, of buffalo or hen (Sundanese); jalu male (Sundanese), jantan (Malay); alu and antan, Malay variants for pestle; pěmali and pantang, Malay variants for taboo. Cp. piama, piantan due season, esp. for rice-planting. And again, how it occurs in the aboriginal dialects of the Peninsula āsu’, anjing, nyang dog; puteh, pěntol white; serigala, sĕgala’, sĕranggil jackal; without definite evidence of ceremonial use, except that inost big animals have honorific synonyms.

[9] Taken from Mr. R. J. Wilkinson's Malay-English Dictionary.

[10] See Dr. Fokker's edition of Beech's Tidong Dialects (Clarendon Press, 1908).

[11] Cp. Shellabear's ‘Evolution of Malay Spelling’, J.R.A.S., Straits Branch, xxxvi. 75-135.

[12] See Romanised Malay Spelling, F.M.S. Govt. Press, Kuala Lumpur, 1904.

[13] Dr. Tendeloo accepts ng for ڠ but uses ñ for ڽ.

[14] Wilkinson's Malay-English Dictionary, vol. ii. p. 714.

[15] ‘Probably adjectives neither derived nor foreign are at bottom words denoting a subject, that is a quality, and so coming in time to denote the possession of that quality.’—TENDELOO.

[1]

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]