20

As a final (Sanskrit visarga) it is only heard before a pause, otherwise it disappears, except when preceded by an a-sound and immediately followed by", in which case it replaces the initial '^-

So: pum.Sih-\-''6' p&n becomes p M-m ah o'-p a fl (a man's house) but: gagaA -j- papkdsa becomes ga-ga-pap-ka-sa (glorious, prop, powerful and renowned) and : panasuJ^ + '^itM becomes p a-n a-s n-w i-t u (that nurse) thus: s am bah (a respectful salutation) + y a n (god) have become one word samba'yan.

Consequently in combination with a word or a suffix a final syllable closed by h is assimilated with one ending in a vowel ').

Spelling. Malay spelling is a very arbitrary one. We might compare it in that respect with ours in the seventeenth century, when one word was found in three different spellings on the same page. According to the original system almost every word is written separately. Consequently sandhirules do not exist or rather are never applied. That such a system can never represent the true pronunciation of the words arranged in a sentence is self-evident.

The words being written mostly as though each of them were sounded indepently from the word preceding or following, we might expect never to find a written sign for h at the beginning of a sentence or part of a sentence , and always to meet such a sign when heard in a single word. But the Arabic-Persian alphabet Malays make use of is not quite fit for the language and Malays cannot spell. So we find f and sb as initials, almost indiscriminately, (t) 1, (.) ^, (^) ^, (j) "J, (a) "x, and .g. or so as medials, and s or x as a final.

1) Compare J. CrawfuiJ. A grammar and diet, of the Mai. L. Orthography p. 3.

21

So: ''ap? (fire) is written ,yl

"dfam (hen) „ „ f^\.s> and j^j

''zkan (fish, meat) is written ^yol and ^yC^5>

ba'M (odour) is spelt jL and ^Lj and _^Lj

na'Mk (rise) „ „ libU and t>5CjlJ but pafe't (bitter) is nearly always found spelt like this : c>^lj , without the least good reason. sSmb^^ is spelt hj^^.

In all these cases the only fit letter for expressing h would be I without '^ ; for it agrees with French h written between a and another vowel (in bahut f. i.) , as well as . corresponds with our w written between u, o or oo and another vowel, (5 with our y written between a or e and another vowel (as in Tower or layer). The definition given by Arabian grammarians of each of these consonants authorizes this comparison.

The Arabian sn at the beginning of a wofd representing exactly the sound of our h in house, it ought never to be used for ^, but Malays very often do.

As to ^ and ^ they are but substituting letters for f, when it has a labial or palatal vowel. Malays very often omit the hamzah, but even then it would be wrong to transliterate Id hut (sea) 1 a w m t etc. (as f. i. Klinkert did in his „Zamenspraken" : baiooe, oe meaning u, i.e. oo in „mood"); for . and ^ are but substitutes for t (Caspari, Arabische Grammatik etc.). But Idut with a diaeresis would be as wrong as that, if the diaeresis be intended to denote a hiatus, for there is no hiatus at all in the word. Then Idut (without any separating sign between the vowels) is by far the best way of transhterating the word , unless we make use of a new letter for the uniting aspiration, as we do in this dissertation {h).

24

tends to be so, and even faults made by Arabs writing or speaking- Malay are sometimes most zealously imitated.

4. — k. — Malay k is identical with what in Dutch is written as such, not always with the sound represented by the same sign in Engl, or Germ. In the latter an initial kh in speaking very often corresponds to k or c in writing. In Spanish and Italian the c before a , o or u expresses quite the same consonant as that written k in this treatise.

It mostly occurs as an initial of a stem or syllable; as a final only after i and a in the last syllable of a stem. Examples: 'kuku (nail) daddk (bran) ta'pik (pull). In a Malay dissyllabic stem k as initial of one of the syllables is never met along with g as an initial of the other. After the syllable mS,, when used as a verbal prefix the initial k is dropped. I venture this hypothesis for such forms.

In an older period of the language, ma stood apart as a separate word (v. d. Tuuk), then annexed to the following word it took n between them; afterwards k was dropped. This evolution can still be observed in the Javanese (mS,katan used along with mankatan and 'a a tan in colloquial style). Some names of trees (e.g. mankitdw) and a few other isolated cases (manka'wan) would show then the transitional state of these forms. Perhaps they have been better conserved being less used in speaking than verbal forms.

Spelling. — The consonant k in genuine Malay words is usually rendered by d- Sporadically ^ is seen instead by mere mistake.

The deviating rendering of it in words from Sanskrit ori-

25

gin, where another consonant comes next, will be treated afterwards (Sanskr. element).

5. — n. — This consonant is usually rendered by two signs in English , Dutch or German spelling , i. e. ng. In most English and Dutch books treating of Malay this rendering has also been adopted for the Malay a. Indeed its pronunciation is wholly identical with what is written ng in English and Dutch, when these letters close a word. But if we wish to render only that consonant and nothing else, ng is not fit for it, for in the middle of an Enghsh word it represents the consonant + g (f. i. singer) or n + j (f- i. in ginger). Moreover , the single consonant ought to be represented by a single sign. So we chose for it the usual one in transliterating Sanskrit.

In Malay n is rather frequent. It occurs as an initial , medial and final.

As a medial it never closes a syllable, always belonging to the syllable which contains the next following k of g. So ti'ngal (remain, stay, live) has to be devided:

te'-ngal

pankalan (pier, landing place):

p a - n k a' - 1 a n

panwlw (chief, headman):

pa-fiM-lw

t^ngipi (a fish):

tai-ng{ -pi. n very seldom opens and closes a syllable at a time, when this contains also g or k,

b«-nun (puzzled etc.) and ta/)-ca-n^n-ca-na"n (astonished) are the most current examples.