An Introduction to Indonesian Linguistics/Essay 3

by Renward Brandstetter

ESSAY III

THE INDONESIAN VERB :

A DELINEATION BASED UPON AN

ANALYSIS OF THE BEST TEXTS IN

TWENTY-FOUR LANGUAGES

(The original was published in 1912.)

SUMMARY

1-16. Section I: Method and Sources.

17-25. Section II : The Verbal Word-base.

26-42. Section III : The Formatives of the Verb. 26. Verbs with and without Formatives. 27. Formatives and Auxiliary Words of Form. 28-9. Phonetic Laws. 30. The Formative ṅ. 31-2. Variation. 33. Infixes. 34-5. Explanation of certain Formatives. 36. Three Principal Types of Formatives. 37. Meaning and Function of the Formatives. 42. Doubling of Words.

43-73. Section IV : The Three Genera, the Principal Characteristic of the Indonesian Verb. 43-5. General description. 46-55. Active. 56-62. Causative. 63-72. Passive. 73. Verbal Systems.

74-92. Section V: The Moods. 74. General Description. 75-85. Imperative. 86-7. Conjunctive. 88-91, Optative, Potential, etc. 92. Irrealis.

93-117. Section VI : The Tenses. 93. General Description. 94. Present Active. 95-104. Past Active. 105-14. Future Active. 115. Tenses of the Causative 116-7. Tenses of the Passive.

118-33. Section VII : The Persons. 118-25. Full and Short Forms of Pronouns. 126-32. Relation of Pronoun and Verb. 133. Other Forms of Verbal Function.

134-8. Section VIII : Verbal Phrases and Other Modes of Expression. 139-62. Section IX : The Verb in the Sentence. 139. The Connecting Links of the Sentence. 140. Prepositions. 141. The Copula. 142. The Status Constructus. 143. Emphasizing the Predicate. 144. Sentences without a Subject. 145. Sentences without a Verbal Predicate. 146-62. Connexion of the Predicate with other Parts of the Sentence.

SECTION I : METHOD AND SOURCES.

1. I have observed that comparative philologists, whether they happen to operate in the Indo-European, Indonesian, or any other branch of the subject, seem, for the most part, to diverge along two different lines. The one school delves deeply into the texts of the several literatures that bear upon the subject, the other is inclined to depend more on manuals and vocabularies. The second method, though it may not give perfect satisfaction, certainly has the advantage of greater facility and rapidity; but as the special character of the present IN ( = Indonesian) monograph compels me to follow in the footsteps of the first school, I will endeavour to justify my procedure in the eyes of those who pursue other methods.

2. There are, to be sure, a large number of IN grammars and vocabularies, and amongst them we meet with not a few that deserve to be styled “exemplary”. Still, numerous as these are, there are quite as many languages that are represented only by inadequate manuals, or none at all. — Moreover, the point of view of the grammarian dealing with a single language is not the same as that of the comparative philologist. The grammarian will fail to observe some things which are of interest to the comparative philologist, or even if he does observe them, he may perhaps omit to include them in his delineation. For example: we shall have to deal later on with a causative formative paka- which occurs e.g. in Bugis (in Southern Celebes), in which language we find e.g. the form pakatanre, “to heighten”, derived from the WB (= word-base) tanre, “high”. In the language of Nias (an island at the back of Sumatra), in conformity with its phonetic laws, as to which more will be said hereafter, this paka - has to become faqa -. Now the formative faqa - is not mentioned in the Nias grammar, but we find it in the texts. Illustration: in the story of "Samagowaulu in the South " is the sentence: " Do let me go, Father ! " = Let-go me only yet, father = faqamoi do mano sa, ama. Here we have the causative faqamoi, " to let go ", derived from the WB moi, "to go ".

3. The literatures of the IN nations are rich, particularly in popular productions which are most welcome material, not only for the folklorist, but also for the student of comparative philology. And fortunately such texts have been published in plenty, thanks to the zeal of Dutch scholars especially, who, in this department, headed by Kern, have done admirable work. In a good many cases we have got texts of languages for w^hich there are at present no manuals, and that happens just to apply to some languages which are particularly important to the comparative student, as for instance Kupangese. — It is only in the special department of Phihppine studies that the texts published up to now are deficient in number. It would, therefore, be a fruitful task for such scholars as Scheerer, Conant and Seidenadel to remedy this defect.

4. There is one scientific operation that is practicable only on the basis of a study of texts, viz. enumeration. I do not by any means regard enumeration as a species of child's play : it is, amongst other things, a matter of scientific importance to know how often a linguistic phenomenon occurs. — Specimen enumeration: as Section IX shows, IN, like other languages, have reflexive verbs; Malay has an expression parallel to the German "sich begeben " ("to betake oneself"), Bugis a parallel to the French " se repentir ". Now an analysis of the Malay popular historical romance Hang Tuah yields a dozen reflexive verbs in 112 printed pages. A German or French text of the like compass would exhibit more of them. The result of the enumeration, therefore, is: Malay has reflexive verbs as German and French have, but they are less numerous in Malay than in these two languages. Moreover, we shall find on more than one occasion that an enumeration actually decides a doubtful question. 5. Even those scholars who merely wish to study comparative phonology, must not think that this limitation of their aims dispenses them from the study of texts. I will emphasize that point by means of an example: one phenomenon of very frequent occurrence in IN is metathesis. Thus the Common IX* word pari, " ray " (a species of fish), appears in Tontemboan, of Northern Celebes, under the form pair. Such metatheses are inexphcable without the aid of a study of Kupangese texts. In Kupangese, a language spoken in Timor (an island lying near New Guinea), metathesis appears quite regularly in certain contexts. Common IN laku, " to go", and kali, " to dig ", are lako and kali in Kupangese; but the sentence " Then I went and dug a hole ", in the Story of the Fool (literally " then I went (to) dig hole ") is mo auk laok kail bolo. — Or, to give another example of the importance of texts even for the phonologist : in Minangkabau (in Sumatra) pronunciation and spelling diverge to a marked degree, and as a rule the spelling represents an older phase of pronunciation; hence the written language is of importance for comparative phonology. Now the grammar, with its practical tendency, only gives the spoken forms of words, not the written; and even the very carefully compiled vocabulary occasionally gives merely the spoken form. In such a case we can find the form which for us is the more important one only in texts printed in the native character. The word for " generation ", for instance, appears in the vocabulary only under the form sunduiq; but in the texts I find sundut, and that this find may be rehed on, i.e. that this spelling really embodies the older pronunciation of the word, is proved by the fact that Karo, another Sumatran language, actually says sundut. Here then we have a phonetic phenomenon which only a text could reveal, both grammar and vocabulary having failed us. Illustration: in the Minangkabau popular story entitled Manjau Ari, in the third line from the beginning, we find the words " from generation to generation ", which in the spoken language are sunduiq basunduiq, written s + u + n + d + u + t b + r + s + u + n + d + u + t.

*Common IN = occurring in all or at any rate most IN language; [See Essay II, especially §§ 2-4.] 6.Lastly, I have not infrequently met with discrepancies as between texts and manuals, and in such cases it was always the text that was in the right. The Makassar grammar says that the auxiliary word of form la is " generally " employed in order to indicate future time. So I said to myself: If that is the case, I shall meet with the particle indicating the future particularly frequently in the two prophecies contained in the romance Jayalangkara, for there the predicates refer to future events. But in the first prophecy, among a dozen predicates, the sign of the future only figures three times, and in the second one not at all. And the same proportion is exhibited, apart from the prophecies,, by the whole text of the romance. So the grammar ought to have formulated the rule thus: " Makassar possesses the capacity of expressing the future; but usually, when future time is intended, it simply uses the general form of the verb, which does not imply any time in particular."

7.The present monograph treats of the IN verb. It is based upon a detailed analysis of IN texts. I do not say that I owe nothing at all to grammars, but I do say that for me the analysis of texts has everywhere been the primary factor, and has formed the groundwork of my edifice. And in all cases, where I have consulted a grammar, I have verified its assertions by the help of texts. All the illustrations occurring in this monograph have been collected by myself out of the texts mentioned in § 11 (or the other texts referred to in § 14) ; should any of the illustrations also figure in some manual, it would be a coincidence that has escaped my notice.

8.A monograph should have something of an artistic form. Now there is no art in raking things together into a heap. Art involves selection and limitation, lucidity of structure, and intelligible exposition. Particularly selection and limitation; I shall therefore by no means submit to the reader every one of the observations I have made, but only such as appear to me to be especially characteristic and interesting. By " characteristic " I mean in relation to the IN .structure, and when I say " interesting ". I have in mind. above all, the interests of Indo-European and general linguistic study. I will explain that by an example. The IN languages of the Philippines have a copula, which links the subject with the predicate, and has the form ay, or i, or ya. So the passage near the beginning of the Tagalog version of "Wilhelm Tell": "The boy fell asleep ": reads aṅ bata i naidlip. There are, however, certain cases in which the copula is omitted, e.g. where Tell says to Johannes Parricida: "Stand up!" = Stand + up you = tumindig kayo. Now the appearance of the copula is a linguistic phenomenon which is characteristic from the point of view of IN and interesting for the Indo-European student, and therefore I shall speak of it in Section IX; but the limitations of its use are of much less importance, and accordingly I shall say nothing about them. If it should appear that any part of this monograph has been expressed too concisely, the defect can easily be remedied : I shall simply expand such portion into an additional monograph.

9. Amongst the numerous IN languages I have chosen the following for the basis of my delineation:

Philippines: 1. Bontok. — 2. Tagalog.

Northern Celebes: 3. Tontemboan.

Central Celebes: 4. Bareqe.

Southern Celebes: 5. Makassar. — 6. Bugis.

Borneo: 7. Dayak.* — 8. Basa Sangiang.

Java: 9. Old Javanese. — 10. Modern Javanese.

Islands towards New Guinea: 11. Kamberese. — 12. Kupangese.

— 13. Bottinese. — 14. Masaretese.

Sumatra: 15. Minangkabau. — 16. Toba. — 17. Karo. — 18. Gayo. — 19. Achinese.

Islands at the back of Sumatra: 20. Mentaway. — 21. Nias.

Malay Peninsula: 22. Malay.

Madagascar: 23. Hova. — 24. Old Malagasy, i.e. the more archaic dialect of Ferrand's texts, which is indeed related to Hova but nevertheless independent of it.

10. Justification of the choice of these twenty-four languages. — The reason why I selected just these languages

• [See Essay I, § 10, footnote.]

consists merely in this, that they appeared to me to be the most fruitful for my theme. It is only the inclusion of the Basa Sangiang that demands a more detailed explanation. The name Basa Sangiang means "language of the spirits", the genitive "of the spirits" being indicated merely by position, without any special formative. It is the liturgical language of the Dayaks. It differs from the Dayak proper in vocabulary and morphology. Thus in the last of the Songs of the Dead we find the sentence: "She has for cradle a spider's web" = Has + cradle spider's + web" = batuyaṅ lawa. Here by means of the formative ha- is derived from the WB tuyaṅ, "cradle", the verb batuyaṅ, "to have a cradle"; ordinary Dayak says hatuyan. However the Basa Sangiang may have originated, its formatives are strictly IN. Thus one of the formatives which it specially affects, viz. ṅa-, or ṅam-, occurs in another language of Borneo, namely in Tidung :

Tidung: ṅalikut, "to bind", from WB likut.

nampuki, "to abuse", from WB puki.

Basa Sangiang: ṅalayan, "to rest", from WB layan.

ṅampelek, "to interrupt", from WB pelek.

In a similar way all the special features of the verb in the Basa Sangiang can be shown to be genuine IN; therefore I am justified in including it among the twenty-four languages.

11. Now follows a list of the texts of the twenty-four languages selected as a basis for my investigations and delineations:

1. Bontok: mythical stories, accounts of battles, headhunters' ceremonies, working songs.
2. Tagalog: Guillermo Tell ni Schiller.
3. Tontemboan: mythical stories, descriptions of sacrifices, prayers at sacrifices, legends, ghost stories, tales.
4. Bareqe: stories about animals, funny stories, popular songs, riddles.
5. Makassar: the war epic Maqdi, the romance Jayalangkara, elegies, children's songs.
   6. Bugis: the edifying tale of King Injilai with the three moral tales of the executioners interwoven therewith, love

   songs, epigrams against cowardice, letters.
6. Dayak: popular stories.
7. Basa Sangiang: the songs at the festival of the dead.

10. Modern Javanese : the history of the kingdom of Kediri, the historical drama Prabu Dewa Sukma.
11. Kamberese: stories about animals, dancing songs, harvest songs, songs at house-building.
12. Kupangese: the Story of the Fool.
13. Rottinese: the play "Cock and Ape", wherein the characters are animals.
14. Masaretese: stories about animals, historical legends, forms of oaths.
15. Minangkabau: the popular tale Manjau Ari.
16. Toba: the Contest of Sangmaima for the spear that was an heirloom.
17. Karo: the Story of the Glutton.
18. Gayo: the Legend of the Blue Princess. Small vignettes illustrating social life.
19. Achinese: the Story of the Pelican.
20. Mentaway: love dialogues,polite dialogues, conversations about the priepthood, medicine, custom, and law.
21. Nias: popular tales, wedding songs, proverbs, the great heroic hymn of Lagemann.
22. Malay: the famihar epic Bidasari, the historical romance Hang Tuah, Abdullah's Journey to Kelantan, etc.
23. Hova: the ethical Testament of Umbiasa, the old funeral oration of Imerina, Rahidy's fables.
24. Old Malagasy: Muhammadan sermons and prayers.

12. Justification of the choice of these texts. — By far the greater number of the selected texts are of an original and popular kind; they are therefore precisely such as a student of language desires. The exceptions are the Tag., Old Jav. and Old Mlg.* texts, whose inclusion must accordingly be justified.

I. The Tag. translation of Tell is by Eizal. To everyone who knows the name of Rizal the idea will at once suggest itself that this translation must contain the purest and most genuine Tag.

II. The Old Jav. literature, or so much of it as has been published up to the present time, is in the highest sense a product of conscious art, dependent in a great measure on the ancient Indian literature. This relation of dependence is reflected by the word-store of Old Jav., which displays a large percentage of ancient Indian loan-words. The Rāmāyaṇa begins with a characterization of considerable length, wherein the "epitheta ornantia" are for the most part Sanskrit words, the native ones being pretty well confined to rĕṇön, "renowned", dumilah, "brilliant", and māsih, "kindhearted". — But this alien element has only affected the vocabulary; morphology and syntax, and therefore also the character of the verb, have hardly been modified in the slightest. The same is true of the foreign element in other IN languages, as the researches on that subject, e.g. those of Van Ronkel, have shown. — Accordingly we shall not only use the Old Javanese texts without scruples, but shall also find them to be the most fruitful of all texts for our purpose.

III. The Old Mlg. texts display much the same character as the Old Jav.; they contain a fairly considerable quantity of Arabic loan-words, but for the rest their inclusion may be justified by the same argmnents as have been used concerning the Old Jav. texts.

13. Some IN dictionaries give such lengthy illustrative quotations in support of the words they explain, that they may be said to amount to complete, though short, texts. This may be seen, for example, in Aymonier and Cabaton's ^[1] Cham dictionary or in Hazeu's dictionary of Gayo. The "small vignettes illustrating social life", mentioned in my list of sources, consist of such lengthy quotations in Hazeu's dictionary.

14. The texts enumerated in § 11 vary in compass and contents, but in almost every instance these sufl&ced for the requirements of the present monograph. Where that was not the case, I have drawn upon additional texts. An example: in Hova we meet with an interesting imperative fuha, "wake up !", a form which constitutes an exception to the regular modes of formation. But this fuha occurs neither in Umbiasa's Testament, nor in the funeral oration, nor in the fables; but it does figure in the oracular formulas of Amurunkay and Vunizungu; in treating of the imperative I shall therefore have to quote from these formulas. — Just as I shall occasionally make use of other texts besides those mentioned in § 11, so too it will occasionally happen that other languages besides those enumerated in § 9 will be called upon to give evidence.

15. The majority of the texts mentioned in § 11 are accompanied by translations. The comparative philologist prefers such translations as, without being woodenly literal, do not depart widely from the wording of the original. I consider as a model in this respect the style and manner in which Kern, the two Adriani's, and Blagden do translations. Of the two versions of the Sangmaima, the more elegant one by Pleyte is more convenient for the student of literature, the more literal one by Schreiber more suitable for the student of language. Van der Toorn's translation of the Manjau Ari is in places too free for the requirements of the linguistic student. For instance, in the description of the character of the Bandaharo, there is a sentence: "He used to slay and pay no wergild, he wounded and paid no fine" = Killed, not paid + wergild, wounded, not paid + fine = mambunuah indaq mambanun, mancancaṇ indaq mamampeh. This he renders by "He disposed freely over the life and death of his subjects". — In my translations, which constituted the beginning of my IN studies and a preparation for my work in comparative philology, I have taken several difterent lines. My Hang Tuah and Paupau Rikadong may serve both for the student of literature and the comparative philologist ; the Jayalangkara is a decided abridgment of a somewhat diffuse original, so the student of language would do well not to tackle it; on the other hand, the translation of the Injilai has been specially designed to meet the requirements of the linguistic student, and even the beginner. Parts that were more than usually troublesome to read have been transliterated in the footnotes and every passage that offers any sort of difficulty is literally translated and explained. I did this because in my opinion an accurate knowledge of Bug. is indispensable to IN comparative philologists. — The requirements of students are met even morefully by Snouck Hurgronje's translation of the Blue Princess or Jonker's Kupangese translation : the former translates word for word, the latter gives a double version, an interlinear one and a free one.

16. My monograph has been preceded by five works, all of great value, deahng with some parts of its theme. Kern has written several essays on the Old Jav. verb; Jonker has described the ways in which the IN verb indicates person; Brandes has dealt with the infix -in- ; Van Ophuijsen has discussed certain phenomena of the Mai. Verb; Adriani has scattered in his manuals many acute observations. — It has been my endeavour to find something new to say even on these subjects.

---

17. In relation to the objects of our monograpli we can divide IN words into two classes: word-bases and words derived from word-bases by means of formatives. The latter may be called "derived" words, for short. The WB, which is mostly a disyllable, occasionally a monosyllable or a trisyllable, is the shortest formation actually existing in Kving speech; the fact that it is possible to analyse the WB further theoretically does not concern us in this monograph. In the Nias verse from the heroic hymn of Lagemann: "He went and clasped the shaft of the spear" = W.cl.sh.sp. = moi muraqu dotoa hulayo, the word moi, "to go", is a WB, muraqu, "to clasp", is a derived word, formed from the WB raqu. So here there are both kinds, a WB and a derived word, used in living speech, in the sentence.

18. WB's that denote action, or it may be suffering, or a state, we style verbal WB's. In the fourth canto of the Malay epic Bidasari there is a verse: "Day by day he sat there sorrowing" = Every every d. sat sor. = tiyap tiyap hari duduq bĕrcinta. Here duduq is a verbal WB, it means "to sit"; and everywhere, wherever it occurs, it means "to sit", not "seat ", for that is kadudukan.

19. There are not a few verbal AVB's which, either quite unchanged or modified only in strict conformity with phonetic law, run through so many IN languages that we have to call them Common IN. Such a word is takut, otherwise takot, etc., "to fear, to be afraid". — Illustrations: In the Tagalog Tell where Friesshardt says: "And we are not afraid of the waters of the Alps" = And not are + afraid of the w. of the A. = at di natatakot sa maṅa Hog nan Alpes. In the Bugis History of the Founding of Luwuq we find: "His servants Avere afraid"

= Feared s. his = metauqna ata na. In the Malay work Hang Tuah, Hang Jĕbat says: "I am not afraid to die" = Not I fear to die = tiyada aku takut akan mati. In the Hova Fable of the Donkey we find: " There was no one who did not fear him" = Not was, not feared him = tsi nisi tsi natahutra azi.

Note. — In support of the assertion that takut is Common IN I have only given illustrations from four areas of distribution, viz. the Phihppines, Celebes, the Malay Peninsula, and Madagascar. Strictly speaking, I ought to give illustrations from all the twenty-four languages, or at any rate from a majority of them, in order to convince the reader that takut really is Common IN. But that procedure would involve such an accumulation of ballast as to deprive my monograph of the character which, as stated in § 8, I wish it to have. Accordingly both here and in the following I confine myself to mentioning merely three or four of the illustrations I have collected; but in doing so I always strictly observe the precaution of quoting from languages which in each case are most remote from one another both from the point of view of relationship and also geographically, as in the present instance of Tag., Bug., Mai. and Hova.

20. We have seen above that both simple WB verbs and derived verbs can do duty as the predicate of a sentence. In most IN languages the derived verbs bulk much more largely than the others, and yet we always find alongside of them a minority of simple WB verbs. This state of things must be regarded as Common IN. Only in the languages spoken near New Guinea, e.g. in Masaretese, do I observe the reverse, viz. that the simple verbs predominate in the texts. — This statement shall be emphasized by an enumeration. In the part of the Tag. Tell, where Baumgarten relates Wolfenschiess' suggestions, all the verbs are derivative; in the passage "The castellan lies in my house" the idea of lying is replaced by that of being (i.e. being present in) and is not expressed verbally, for reasons which will be dealt with here after. In the Tontemboan Storv of the Defeat of the Antelope by the Water Snail there are hardly any but derived verbs in 23 lines of print. On the other hand, in the Masaretese Story of the Ghost with Seven Cords we find in 39 lines of print only 5 derived verbs: eptea < formative ep + WB tea, "to set", ephatak, "to sacrifice", danewen < da + newen, "to live", damata, "to die , and epmata, "to kill",

21. Having learnt that verbal WB's, without any formatives, are capable of being used in the sentence, if we now proceed to enquire whether any particular categories of verbs are used in that way, we get the following result from an analysis of the texts: the WB's most commonly thus used as predicates are those which have a passive or neutral sense; more rarely the dative ones, i.e. such as link themselves with the object by means of a preposition; and very seldom accusative* ones, which take an object without a preposition. — Illustrations: Day., from the Story of Sangumang: "He wishes to be addressed" = blaku tiṅak. Kamberese, from the Riddle about Maize: "He stands up" = na hadaṅ. Old Jav., from the Ādiparwa: "To be versed in the spiritual life" = wĕruh ri ambĕk. Mal., from the Hang Tuah: "To have breeding" = Know speech = tahu bahasa. — Here then we have the passive WB tiṅak, "to be spoken to", the neutral hadaṅ, "to stand", the dative wĕruh, "to be versed in", and the accusative tahu, " to know ", used in sentences.

That the passive WB's are really passive in their nature, is proved by the circumstance that they can be accompanied by an agent linked with them by means of the same preposition as in the case of a derived passive, e.g. in Mal. by means of oleh, "by". — Illustration from the Sĕjarah Mĕlayu: "This king was defeated by King Alexander" = Was + defeated k. this by K. A. = (maka) alah (lah) raja itu oleh raja Iskander.

Note. — The IN languages are rich in particles. Such particles often merely serve the purpose of beginning the sentence, like the above maka, or laying stress on some part of it, like

• [I.e. what are commonly called transitive verbs, the others (so far as they are active) being intransitive.] the above lah, or marking an antithesis, and then they are untranslatable. For the greater convenience of the reader I put them between parentheses: let the reader simply pass them by.
  

22. In practically all the IN languages the verb "to be" — not our copula "to be" but "to be" in the sense of "to exist, to occur somewhere" — is devoid of a formative; this phenomenon must be styled Common IN. —Illustrations: Magindanao, from the collection of dialogues in Juanmarti's work: "There is someone there" = Is someone = aden sakatau. Mad., from the story Paman Manceng: "There was once a man" = Was one man male = bada settoṅ oreṅ lakeq. Sěraway, from the story Riṅgan Sëdayu: "There was once a king" = Formerly w. a k. = bëmidaw adaiv suqatu rajaw.
23. Enumerations bearing on §§ 21 and 22.
I. In the Old Jav. story of the evil serpent Takshaka in the Adiparwa of the Mahābhārata we find many neutral, some passive, some dative, and one accusative verbal WB, viz. tuṅgaṅ, "to sit on". — Illustration: "It sat on the hill" = tuṅgaṅ parwata.
II. In the part of the Mal. Hang Tuah which relates the early history of the hero we meet with the same state of things as in Old Jav. Amidst many neutral WB's there are a few passive and dative ones, and a single accusative one, tahu, "to know". — Illustration: "To have breeding" = tahu bahasa.
III. In the Day. story Asang Baratih the proportion is again similar, only there we come across more accusative WB's, namely such as denote motion, which will be dealt with hereafter.
IV. In the collection of Hova fables by Rahidy the neutral and passive WB's are in the majority, the dative ones are represented by a single case: "to say to (a person)" = hui; accusative ones are wanting. But in contrast with the Old Jav. (see I. above) the relation between the neutral and passive WB's is that the latter, e.g., hita, "to be seen", balance the former.
24. Now when it is desired to use derived verbs and not simply verbal WB's, the language is able to fashion them out of the most diversematerials, from any part of speech and even from "complexes" or agglomerations of words.
I. Derived verb formed from a WB which in itself is already of a verbal nature. In Old Jav., at the beginning of the episode of the death of Abhimanyu in the Bharata-Yuddha: "The son of Dharma was dismayed" = san dharmmasuta atěgěg, we have a derived verb atěgěg, formed from the WB těgěg, "to be dismayed", which in itself is also verbal.
II. Derivatives from substantival WB's. Example: Old Bug. pajuṅ, "royal parasol", maqpajuṅ, "to wield the royal parasol".—Illustration from the History of the Founding of Luwuq: "They went to him who wielded the royal parasol" = Went to r. + p. + wielding the = menreq ri maqpajun e.
III. Derivatives from other parts of speech. From the interrogative pronoun apa Old Jav. forms the verb aṅapa, "to do what, to desire what ?"; from the word en, "yes", in Tontemboan comes men < ma + en, "to say 'yes'", the past tense of which is nimen; from the interjection of clearing the throat, ehem, Day. forms the verb nanehem, "to clear the throat, to say 'h'm'".—Illustrations: Old Jav., from the Mausalaparwa: "What ailed the Brahmans that they cursed?" = W. + a. the B. c. ? = aṅapa (ta) sira brāhmana śumāpa. Tontb., from the Story of the Python: "The youngest, she said 'yes'" = si caakaran isia (ka) nimen. Day., from the Story about saying H'm: "He cleared his throat 'h'm'" = iä ňaṅehem eheehem.
IV. Derivatives from conglomerates or "complexes". In Mak., "to ask" is palaq and "help" is tuluṅ, but "to ask for help " is papalaqtuluṅi. In this case a verb has been formed by means of the two formatives pa- and -i out of the conglomerate palaq tuluṅ. — Illustration, from the Jayalangkara: "None other can we ask to help save only the serpent" = Not other can we ask + for + help except serpent the only = taena maraeṅ maka kiq papalaqtuluṅi pasaṅalinna naga ya ji.
25. Concluding observation of Section II: Delimitation of the verb as against the substantive and the adjective.
I. Verb and substantive. There are verbal WB's and there are substantival ones. The Mal. WB duduq means "to sit", and not "seat", the Bug. WB api signifies "fire", and not "to burn"; in Mal. "seat" is expressed by kadudukan, in Bug. "to burn" by tunu. It is true that the vocabularies speak of certain WB's as being both verbs and substantives; but in very many cases the texts tell us that that is merely apparently the case. In all Mal. dictionaries we find the statement that tidor means "sleep" and "to sleep"; but tidor, "sleep", when used in a context, requires different pronouns from tidor, "to sleep". "He sleeps" is tidor iya, "his sleep" on the other hand is tidor ña. So verb and substantive are at least distinguished by the construction. Again another case: Toba pintu means both "door" and "shut"; but pintu, "door", has the accent in the usual Common IN fashion on the i, while pintu, "shut", has it on the u. — It often happens that the same formatives are used to form both verbs and substantives; Bug. -ěṅ serves both purposes. But here too the language knows how to guard itself from confusion. Bug. WB's often have two or three forms, differing from one another in the final, a subject discussed by me in a former monograph.* Thus gauq, the WB for the idea "to make", also occurs in the variants gauk and gaur. And the variant gauk is used by the language to form the substantive: gaukěṅ, "thing", while from the variant gaur it creates the verb: gaurěṅ, "to make".
II. Verb and adjective. In the department of the adjective we also find the phenomenon that formatives are used
* ["Sprachvergleichendes Charakterbild eines indonesischen Idiomes", especially §§ 46 seqq.] which do duty for the verb as well. Thus in Hova ma- forms both verbs and adjectives, but still it hardly ever happens that this ma- is employed with one and the same WB to form a verb and an adjective.From the WB tewina, " thick ", the language does, it is true, make the adjective matewina, but for the verb it employs another mode of formation: manatewina, " to make thick ".

SECTION III: THE FORMATIVES OF THE VERB.

26. We have already learnt that the verbal WB may do duty as a predicate either with or without the addition of other linguistic elements. In the Nias proverb: “You need not close your hand when you have no tobacco in it” = Not shut h. hen not t. = boi goxoi daṅa, na lo bago , the word goxo is an unmodified WB and acts as a predicate. In the passage from the Masaretese oath formula: “He will die in eight days' time” " = In days eight die he = la beto etruwa damata di, the WB mata has taken on another linguistic element, the formative da-, in order to play the part of a predicate.

27. Now of such linguistic elements there are two kinds. Either they are syllables which unite with the WB to make a new formation, which is a unit and is governed by a single accent. Or they are independent words, which, though they attach themselves to the WB, do not coalesce with it. The first are called formatives, the latter auxiliary words of form. In the phrase from a Bug, love-song : “Indifference changes into passion” I. the becomes p. = lěbba e mañcaji seṅěrrěṅ, the syllable mañ- in mañcaji is a formative. But in another Bug. love-song: "He has deceived " = Has he d. = pura na bělle, the word pura, which really means " done " but here indicates the past, is an auxiliary word of form. — In the course of the present Section we shall only concern ourselves with the formatives; the auxihary words of form will be discussed hereafter.

28. As phonetic law constitutes the basis of comparative philology, we must now, after the introductory observations of the two preceding paragraphs, concern ourselves with the phonetic conditions of the IN verbal formatives. But not all the twenty-four languages call for special notice under this head.

I. Bontok. Common IN pĕpĕt, that is to say the rapidly pronounced, indifierent vowel ĕ, becomes e in Bont.; hence the Common IN formative -ĕn appears in Bont. as -en.
II. Tagalog. Common IN ĕ becomes i in Tag., hence the same formative becomes -in < -ĕn.—Some of the Common IN r’s appear in Tag. as g, by the RGH-law; hence the Tag. formative mag- < mar-. Thus beside Toba mar-somba there is a Tag. mag-simba, “to worship”.
III. Bugis. A Common IN consonant, other than a nasal, immediately preceding another consonant becomes q in Bug.; thus beside Toba martaru there is a Bug. form maqtaro < mar + taro, “to put”. By analogy this maq- can also be put before a vowel, as in maqeñeq, “to shine”, from the WB eñeq.—All Common IN final nasals are unified into ṅ in Bug., which therefore has -ĕṅ and -aṅ for Common IN -ĕn and -an.
IV. Makassar. Mak. shares the laws which under III. above have been ascribed to its near relative Bugis, and adds to them the following: Common IN ĕ is represented in Makassar by a, hence Bug. -ĕṅ and -aṅ both appear as -aṅ in Mak.
V. Toba. Common IN ĕ appears in Toba as o, hence the Toba -on < -ĕn. That is the reason why in I. above we found Toba marsomba corresponding with Tag. magsimba, both being from an original marsĕmbah.
VI. Hova. Common IN initial b becomes w in Hova, hence the formative wa- < ba-.—Unaccentuated Common IN ĕ is represented in Hova by i; a Common IN final nasal of every kind by -na; hence the repeatedly mentioned formative -ĕn appears in Hova as -ina. — Common IN k becomes a spirant in Hova, hence it has hu-, as compared with the Karo ku-.
29. The phonetic laws of the above-named languages have all been dealt with by me in previous monographs, so I need only state them here, without giving any evidence in support of them. It is otherwise with Nias. No one has as yet said anything about the phonetic laws of Nias, and therefore I must be more discursive on that matter.
I. Common IN ĕ becomes o in Nias.* Hence:

Common IN ĕnĕm, “ six ”	:	Nias ono
Common IN tĕlĕn, “ to swallow ”	:	Nias tolo
Common IN kĕna, “ hit ”	:	Nias gona.

II. Common IN final consonant disappears in Nias.† Hence:

Common IN ĕnĕm, “ six ”	:	Nias ono
Common IN tĕlĕn, “ to swallow ”	:	Nias tolo
Common IN takut, “ to fear ”	:	Nias taqu.

III. Common IN initial k appears in Nias as g.‡ Hence:

Common IN kaka, “ elder brother ”	:	Nias gaqa
Common IN kima, “ shell-fish ”	:	Nias gima
Common IN kĕna, “ hit ”	:	Nias gona.

IV. Common IN k in the interior of a word turns into q.§ Hence:

Common IN takut, “ to fear ”	:	Nias taqu
Common IN buku, “ knot, joint ”	:	Nias buqu
Common IN kaka, “ elder brother ”	:	Nias gaqa.

V. Common IN initial p becomes a spirant.‖ Hence:

Common IN pitu, “ seven ”	:	Nias fitu
Common IN puri, “ behind ”	:	Nias furi
Common IN panah, “ shooting weapon ”	:	Nias fana.

VI. In conformity with these phonetic laws the formatives show the following changes:

Common IN -ĕn	:	Nias -o
Common IN ka-	:	Nias ga-
Common IN paka-	:	Nias faqa-.

* [See also Essay IV, § 5, IV.]	† [See Essay IV, § 205.]
‡ [But see Essay IV, § 349, II.]	§ [See also Essay IV, § 143.]
‖ [See also Essay IV, § 112, II.]

30. Particular attention should be paid to the phonetic processes which take place in the various IN languages in connexion with the prefixing of the formative ṅ:

I. There is a Common IN formative ṅ-, about which we shall have a good deal more to say. Thus from the WB atta Bugis forms the verb ṅatta, “ to be ready ”, from the WB golek Modern Javanese makes the verb ṅgolek, “to seek”.
II. Now IN very rarely tolerates two consonants together at the beginning of a word. Consequently the addition of the formative ṅ- to WB's beginning with a consonant has led to a variety of compromises. The most important are the following, which may be styled Common IN:

ṅ + k > ṅ

ṅ + t > n

ṅ + p > m.

Examples from Modern Javanese:

ṅ + WB kirim > ṅirim, “to send”

ṅ + WB tumbas > numbas, “to buy”

ṅ + WB pakah > makah, “to ramify”.

III. Now it often happens, as we shall explain in detail later on, that several verbal formatives together become attached to the verbal WB. Thus it is a Common IN phenomenon for the formative ṅ to combine with the formative ma-, as ma + ṅ; somewhat rarer are the combinations a- + ṅ, ar- + ṅ, and mar- + ṅ. The agglutination of these additional formatives entails no change in the function or meaning of the ṅ-. Alongside of the above-mentioned ṅgolek Modern Jav. also says aṅgolek, and both mean the same thing.—Here, in subsection III, the ṅ occurs under other conditions than in II.; it is in the interior of the word. Accordingly the phonetic phenomena that manifest themselves here are of a different nature, they are mostly simple assimilations; thus in Dayak ma + ṅ + tarik > mantarik, “to throw”.
IV. So we find, quite naturally, a different treatment of the case when the formative occurs in the middle of a word irom that which it receives when the formative is initial. Nevertheless a number of compromises have been made, consisting substantially in this, that the phenomena applicable to the case of the formative as initial have imposed themselves on the cases where it is medial. Thus in Day., ṅ + talusoṅ, “torch” > nalusoṅ, “to make torches”; ma + ṅ + talusoṅ, should, in conformity with III., become manalusoṅ, but in fact it is manalusoṅ. — Compromises of this kind are to be found in so many IN languages that we are compelled to style them Common IN.
V. As these compromises occur even in the interior of words, we can understand how it happens that there are sometimes alternative forms, an indication that the compromise is not yet a perfect one. Thus from the WB baläh, “to requite”, Day. forms both mambaläh and mamaläh.
31. In the last few paragraphs we have been discussing such phonetic phenomena affecting the formatives as we can grasp and comprehend from the point of view of phonetic law. But we also meet with a minority of phonetic phenomena which do not admit of that possibility at present. Beside the Common IN formative ta-, which forms a passive, the Sund. displays a ti-. Now a Common IN a is represented in Sund., without exception, by a; so the Sund. ti- is not a regular phonetic equivalent of the Common IN ta-. How then shall we explain the relation of ti- to ta- ? Shall we simply declare that they have nothing to do with each other ? That will hardly do, for after all the consonant is the same, and the meaning is identical. So we cherish the hope that the progress of research will throw light on this point, and we call phenomena like this ta : ti by the provisional name of “variation”.
32. There are, however, cases of variation that can be tackled more effectually; we will mention a few of them here.
I. Day. has a formative me-, which fashions verbs from onomatopœic words. Thus from the interjections kap!, bus!, rok!, are formed the verbs mekap, mebus, merok; but riṅ, “tinkle”!, produces miriṅ, “to tinkle”; so here apparently, we have a variation me : mi. If however we look into the matter more closely, it becomes plain to us that the interjections with the vowel i take the formative mi-, as we also find mi-tip, mi-sir. Here, therefore, the variation is a product of assimilation.
II. Alongside of the Common IN, or at any rate widely distributed, formatives ma-, maṅ- < ma + ṅ, bar-, tar-, Mal. has mě-, měn-, běr-, těr-. Now Mal. has the tendency to weaken into e the vowel (whatever it may be in Common IN) that precedes the accentuated syllable; examples: Common IN banuwa, “country” > Mal. běnuwa, Common IN kuliliṅ, “around” > kěliliṅ; and the two loan-words pěriksa, “enquiry”, and sěrdadu, “soldier”, also illustrate the same process. Now the above-mentioned formatives also invariably precede the accentuated syllable, and therefore they too have undergone this weakening, and so mě-, etc., are secondary forms of the more original ma-, etc.
An exception to this principle is the formative lĕ- which fashions verbs out of onomatopœic interjections, e.g., lětak, “to tap” < lě + tak. Here the equivalent of the ě in other languages is not a, but e.g. in Day. e, as in legop < le + gop, “to tap”, and in Toba o, as in loṅiṅ < lo + ṅiṅ, “to make a shrill sound”. But where the vowels correspond to one another in that way, the ě, as Mal. has it, represents the original condition, in conformity with the pěpět-law.
33. To conclude our considerations on the phonetic characteristics of the formatives, we will make some remarks on infixes. One of the Common IN verbal infixes is -um-, e.g. in Old Jav. lumaku, “to go”, from the WB laku. In place of this infix -um- we find in some other languages, e.g. Mentaway and Nias, a prefix mu-. Thus from the WB hede Nias forms the verb muhede, “to speak”. But the texts show that the mu + he in muhede can be replaced by hu + me, according to individual taste and fancy. In the Wedding Song we find the sentence: “Thus spake the old chieftain” — hulo muhede lafauluo; but in the Story of the Captain: “Why don't you speak?” = Why not s. you ? = hanawa lo
humede o. Here then a perfectly arbitrary metathesis is permitted; and as it seems to us more natural to assume that prefixes are the more primitive type of formative, we infer from this instance that infixes originated from prefixes through metathesis.
34. We have now dealt with the phonetic aspects of the formatives and turn to the further consideration of these verb-forming syllables. At the beginning of this Section we drew a distinction between formatives and auxiliary words of form. We will now supplement what we said on that subject by adding that the limits of the two concepts are often somewhat vague, inasmuch as the word of form can turn into a formative. By means of the word buah, “hit”, Day. fashions a passive formula, which is used when it is necessary to speak of pain, disadvantage, and the like; example: buah rugi, “to be damaged”, really “to be hit (by) damage”. Here buah is still felt to be an independent word, and therefore it is not joined to the substantive, in this instance rugi. But in Hova, where in conformity with phonetic law buah appears as wua, its use is no longer confined to words that denote damage, disadvantage, or the like; one can also say. for example, wuasuratra, “written”; so the root-meaning has faded, and wua is now felt to be a formative, and is written together with the WB, in this instance suratra. — Illustration from the Testament of Umbiasa: "Written in this book here" = W. in this b. this = wuasuratra amin iti taratasi iti.
35. From what has been said in the preceding paragraph it also follows that it is sometimes possible to explain the origin of the IN verbal formatives. I will give several such cases here:
I. Many formatives were originally prepositions. Thus a whole series of prepositions meaning “to, towards”, in Latin “ad, versus”, are used to form the future. In Mkb. ka, in Mal. akan (an extension of ka), in Hova hu < ku (a variant of ka), in Mak. la, in Bont. ad. In the Bont. sentence from the Story of the Stars: “Then their mother flies up to the sky” = Then flies the m. their to sky =
keceṅ tumayaw nan ina ca ad caya, the word ad is a preposition; in the sentence from the Battle of Kaloqokan: “ They will appear ” = Will a. they = ad-umali ca, it is the sign of the future. In Mak. la is no longer a preposition, but only the sign of the future, while in the languages of the islands that lie over against New Guinea it is still a preposition. That is shown by the following Kamberese sentence from the Dirge of the Crocodile for his Dead Friend: “ Let us go to the deep water ” = We go to w. d. = ta laku la wai mamanjoluṅ. In Sawunese we meet with a use of this la which represents a striking transition from the preposition to the sign of the future: la is only used when the idea of direction “ thither ” is combined with the idea of futurity. Thus in the text Bale ri ane there is a sentence: “ It was an order that one should buy rice” = Order, that “ la ” buy rice = li ta la wĕli lailudu. That means: “ The order was given to go and buy rice ”. Had the meaning been “ The order was given to come here and buy rice ”, la would not have been used.

II. Other formatives were originally articles. Thus the Nias active participle consists of the indicative verb and the agglutinated article si. In the Story of Buruti: “ Didn't you see any man passing just now ? ” = Not was + seen by + you just + now man passing ? = lo niila u mege niha sanoro, the form sanoro consists of the verb anoro and the article s < si.

Similarly the formative ṅ, discussed above, was originally none other than the widely distributed article ṅ. Originally, therefore, the Modern Jav. ula ṅuntal, “ the snake swallows ”, was ula ṅ untal, “ the snake (is) the swallowing (creature) ”. It is true that the fundamental meaning has generally faded away, but there are plenty of cases in which it can still be perceived. The sentence out of the Bol. dialogues in Beech: “ He abused me ” = saq ṅam'puki da-aku, can also, without doing violence to the meaning, be taken as He (is the one) that abused me ” = saq ṅ ampuki da-aku. — This phenomenon, of the formative ṅ being really an article, has another IN parallel, which will occupy us in Section VII. Just as they say “The snake is the swallowing creature”, so they also say, as we shall there see: “The snake, it is sleeping”. Here, too, the fundamental meaning has generally faded away, and the whole thing means no more than: “The snake is sleeping”.

III. We have already met with a verbal WB, buah, which has become a formative. Later on we shall come across an adjective or adverb, pura, that has had the same fate. The causative formative pa- is identical with the causal conjunction pa. And so on.

36. The IN verb possesses formatives to express, above all, the three genera, active, causative and passive; and that condition of things must be styled Common IN. In the Bis. Riddle about the Ship, in Starr's Collection of Riddles: “ It runs with its back ” = Runs goes + on + its + back = nagalakat nagahayaṅ, naga- is an active formative. In the Tarakan Story of the Tailed Man: “You made me go” = You made + go me = dudu palakaw daka, pa- is a causative formative. In the Talautese Cursing of the Fowl: “It shall be borne in mind” = papaghiana, -ana is a passive formative.

37. It happens not infrequently that verbal formatives have different meanings in different languages. We will mention some of these cases:

I. The prefix ma- sometimes forms causative, sometimes accusative,* sometimes neutral, and sometimes passive, verbs. Examples:

Day.: ma-haban, “to make sick”

Day.: ma-haga, “to guard”

Day.: ma-lelak, “to bloom”

Bont.: ma-oto, “to be cooked”.

The WB's are haban, “ sick ”, lelak, “ flower ”, etc. — We can comprehend these shiftings of meaning, if we assume that the neutral signification, as in malelak, “ to bloom ”, was the original one. One can very well understand a transition from the neutral meaning to the active and causative on the one hand , a nd to the passive on the other. And there are languages in which ma- is exchisively neutral. Similarly we often find uses that fluctuate between the one type and the other, as in the sentence from the Mentaway Fishermen's Stories: "The fat fish is now hemmed in " = Fat + fish is + hemmed + in = mokmok maipit.

II. The formative ka, originally, as shown above, a preposition, does duty in some languages as the sign of the future active, in others as the sign of the passive without any implication of tense. Thus the Minangkabau katiṅga, "to be about to stay", is future; the Bug. kacalla, "to be accursed", is passive. Here, too, we can give a rational account of the double evolution of the meaning. We shall see later on that in the IN sentence verbs of motion can often be omitted altogether, so that one may simply say: "I into the forest ", "I out of the town". Now if we think of sentences like " I (am going) to (= ka) dinner", we can well understand that a future in ka- might be evolved out of them;^[2] but it is equally comprehensible that a passive in ka- might grow out of sentences like " I (get) into (= ka) the curse ". — Illustrations: Minangkabau, from the Manjau Ari: "You will stay, I shall go" = anku ka-tiṅga, den ka-pai. Sundanese, from the Story of Nyai Sumur Bandung: "The story is told, how Rangga Wayang reached the centre of the town" = It is told, R. W. reaching to c. t. = ka-carios Raṅga Wayaṅ sumpiṅ ka hulu dayöh.

III. Common IN has a formative -ĕn for forming the imperative passive. Alongside of this there is a formative -ĕṅ that has an accusative or causative force ; in accordance with the phonetic laws set forth above it appears in Bug. as -ĕn, in Nias as -o, e.g. in Nias balio, "to transform", beside bali, "to turn back, to return." — Now here, in the case of these two -ĕn's, the imperative passive and the causative active one, I can think of no connecting link, nor have I found any-
where anything that looked like an intermediate stage between them.

38. The IN verbal formatives have as a rule three functions: a single IN verbal form represents, first, the infinitive, secondly, the participle, and, thirdly, the finite verb, of Indo-European languages. This state of things is to be styled Common IN, although we have, as above in Nias, found exceptions to it, and shall meet with others hereafter. Thus the Old Jav. atukar < a + WB tukar, according to the context, must sometimes be rendered by “ to brawl ”, sometimes by “ brawling ”, or by “ (I) brawl, (you) brawl ”, etc. The same thing applies to the Hova milefa, “ to flee ”, the Bug. maqrola, “ to prosecute ”, etc., etc. — Illustrations:

I. Old Jav., Jonker's Book of Laws, from the sections about brawling: “ Struck by the kĕris of the brawlers ” = S. through the k. of the brawling (persons) = kasuduk deni ṅ kĕris i ṅ atukar. “ If he begins to brawl ” = yen ambakalana atukar.

II. Hova, Book of the Laws of Ranawaluna, from the sections about fugitive slaves: “ A fleeing slave, if he steals ” = A sl. f., if st. = ni andewu milefa, raha maṅgalatra. “ The slave of a soldier, if he flees ” = ni andewu n ni miaramila, raha milefa.

III. Bug., Book of Laws of Amanna Gappa, from the sections about judicial procedure: “ The prosecutor speaks first ” = Person prosecuting the first speaks = to maqrola e riyolo maqtuqtu.

39. The addition of formatives to a WB not infrequently entails modifications, either slight or more pronounced, in the meaning of the word. In some languages, e.g. Day., this is less marked than in others, e.g. Bug. or Mak. Example.:

I. Dayak.

WB sala	“ wrong ”
basala	“ to be in the wrong, to do wrong ”
mañala	“ to accuse, to act wrongly ”

II. Makassar.

WB sala	"wrong, fault"
maqsala	"to be different"
maqñala	"to be guilty"
pisala	"to miss (in shooting)"
pisalai	"to frustrate (a plan)"
pasala	"to pay a penalty (in money)"
pasalaṅi	"to impose a fine"
pañalaṅi	"to infringe (a regulation)"
kasalai	"to be undutiful (towards one's parents)".

40. The number of formatives that can be attached to the WB at one time varies in the different IN languages, but hardly ever exceeds four. From the point of view of its capacity to form these combinations, the most interesting language is perhaps the Bug. In the Tiruray sentence from the collection of dialogues of an anonymous author: "I am hungry" = melayaf u, the verb has only the one formative, me-. In the Bug. sentence from the Injilai: "She was recognized everywhere" = W. + r + e.s. = riasiisěṅi (n) i, we have, to begin with, the WB isěṅ, "to know", and then four formatives: a, which simply turns the WB into a verb; i, which makes it transitive; si, which expresses the "everywhere" ; and ri, the passive formative.

41. The number of formatives possessed by the several IN languages varies greatly. The richest in this respect are the languages of the North, of Formosa, the Philippines, the intervening islands, and Northern Celebes; Sangirese, for instance, has about a hundred. The poorest in formatives are the languages of the East, Bim., for example, having only two, viz. ma- and ka-. The remaining regions occupy an intermediate position in this respect.

42. Lastly, the formative methods of fashioning verbs also include the method of reduplication, which is a particularly common IN feature in other parts of speech besides the verb, occurs in the most various shapes, and mostly indicates plurality, intensification, and the like. The following are selected cases of specifically verbal reduplication; none of them can be called Common IN.

I. Mentaway has a threeford or fourfold repetition, wherein the final consonant is omitted except at the very end, when the WB appears for the last time. Illustration from the Fishermen's Stories: "He goes, wanders continually, comes to his mother" = Goes he, w. c, c. to m. his = konat ña, toro-toro-toro-torot, šägä ka ina iña.

Note. —Initial š in Mentaway, as in šägä, sounds pretty much like sch in the German "schön".*

II. Bug. has a threefold repetition, with the formative ka- or si- interpolated in between. — Illustrations, from the Injilai: "He went hither and thither" = na lao na ka-lao-lao. "He wept continually" = tĕrri si-tĕrri-tĕrri na.

Note. —The first na, the one before lao, means "he"; the other two are particles of emphasis.

III. Mak. has a twofold repetition, with interpolation of the word of form saṅga or saṅge. — Illustration, from the Epic Maqdi: "Then were urgently summoned the four pillars of the state" = nikiyoq-m-i-saṅga-kiyoq toqdoq appak a.

Note A. — The m < mo is the particle of emphasis, and i means "they", appak, "four", a, "the".

Note B. — This Mak. type of reduplication is not common. An analysis of the whole of the Maqdi only yields one case, viz. the one quoted above. In the whole of the Epic Datu Museng there are two cases: kiyoq-saṅge-kiyoq and kape-saṅge-kape, "to beckon repeatedly".

IV. In several of the Philippine languages, which have a real system of moods and tenses, reduplication plays a great part, whereof we shall have to speak hereafter.

SECTION IV: THE THREE GENERA, THE
PRINCIPAL CHARACTERISTIC OF THE
INDONESIAN VERB.

43. The great majority of IN languages possess the capacity of forming he three genera of the verb: the active, the causative, and the passive. This phenomenon, therefore, is Common IN. It is the chief characteristic of the IN verb.

44. We have learnt that the languages of the East are very poor morphologically. So it is striking that even there we still find languages that possess the three genera, as, for example, Kamberese. Illustrations:

I. Active, formative ma-. From the Song at House-building: " Marapu, who created men " = na Marapu na mawulu tan.

II. Causative, formative pa-. From the Harvest Song: " Let this arrive at the top! "' = L. + a. this to top = patoama ña la pinu.

III. Passive, formative ka-. From the Song against the Son-in-law: " She runs around, as though maddened " = na laku biṅu katoaba.

45. We have seen above that there are WB's that can do duty in the sentence as active or passive verbs, without the help of any formative. But the usual rule is that the language requires formatives. I have never come across causatives without a formative.

46. The Active. The Common IN formatives for forming the active are: ma- ṅ-, or in place thereof maṅ- ᐸ ma + ṅ um-, or -um-. The formative um is a prefix with words having an initial vowel, an infix where they have an initial consonant.
47. The active formative ma-. Proof that it is Common IN:

Formosa, Form, dialect:	matagga, "to bleed".
Philippines, Bont.:	masuyep, "to sleep".
Celebes, Bungku:	mahaki, "to be sick".
Borneo, Basa Sangiang:	mahampan, "to have a border".
Near Java, Bal.:	mahumah, to dwell.
Islands near New Guinea, Kamberese :	malala, "to cook".
Sumatra, Lampong:	mabarsog, "to speak through the nose".
Islands at the back of Sumatra, Mentaway:	maloto, "to be afraid"
Madagascar, Sakalava :	mataotra, "to fear".

48. The active formative ṅ-, or its substitute maṅ-. Proof that it is Common IN:

Philippines, Bont.:	managni, "to dance".
Celebes, Bug.:	ṅanro, "to ask".
Togian Islands, Bajo:	ṅinum, "to drink".
Borneo, Basa Sangiang:	ṅujan, "to rain".
Java, Modern Jav.:	ṅutus, "to send".
Islands near New Guinea, Sumbawarese:	ṅaji, "to teach".
Sumatra, Karo:	ṅapit, "to pinch".
Islands at the back of Sumatra, Mentaway:	maṅaray, "to climb".
Madagascar, Old Mlg.:	nilu, "to shine".

Note A. — The WB's of these verbs are: sagni, kanro, inum, ujan, utus, aji, apit, karay, and ilu. The honetic processes here displayed, e.g. by ṅanro, have been discussed above. Old Mlg. nilu is for ṅilu, in strict conformity with phonetic law.
Note B. — Hova and the Mlg. of Ferrand's texts usually have the longer form of the prefix: man- < maṅ-, nilu is one of the few examples known to me of the shorter form; it occurs at the beginning of the sermon Tonih Zaṅahary. 49. The active formative um- or -um-. Proof that it is Common IN:

Formosa, Form, dialect:	comma, "to speak".
Phihppines, Bont.:	uminum, "to drink".
Celebes, Tontb.: kuman, "to eat".
Borneo, Day.:	kuman, "to eat".
Java, Old Jav.:	kuměmit, "to watch".
Sumatra, Toba:	sumuruṅ, "to improve".
Islands at the back of Sumatra, Simalurese:	lumaṅoy, "to swim".
Malay Peninsula, Mal.:	gumilaṅ, "to shine".
Madagascar, Hova:	humana, "to eat".

Note A. — The WB's are ka, inum, kan, kěmit, etc.
Note B. — The number of cases in which this formation occurs in Day. and Hova is small.
Note C. — The Form, comma has been left unaltered in its clumsy spelling; it corresponds with the Tontb. kuma, WB ka.
50. Illustrations of the three active formatives. Old Jav., from the Kuñjarakaṇa: "Others ran away" = waneh malayū; from the Śakuntalā: "He then saw a woman" = S. t. he w. = anon ta sira strī; from the Kamahāyānikan : "To penetrate into the holy mystery" = tumama ri saṅ hyaṅ paramarahasya. Modern Jav., from Meijer Ranneft's Collection of Eiddles: "A snake swallows a mountain" = ula ṅuntal gunuṅ.
Note. — anon is a + ṅ -- WB ton.
51. Specific signification of the three active formatives. In several languages ma- is intransitive, it- or man- transitive;

but in other languages the active formatives apparently only serve to form the active, without any other shade of meaning. The formative -um- usually plays the part of an aorist, inchoative, or future, and that state of things may perhaps be styled Common IN. — Illustrations of this force of the formative -um-. Bont., from the Head-hunters' Ceremonies: "They start for the settlement " = Start they to t. s. = sumaa ca is nan fohfüy. Tontemboan, from the Story of the Demon.

that haunts women at childbirth : "She came quickly, in order to clutch them" = sia mělaqu-laqus {omai) tumaṅkaq isera. Old Jav. from Mpu Tanakung's Prosody: "Startled by the birds bathing (they — the gaudily-coloured fish—) flashed upwards" = S. by the birds b., flashed-up = kagyat deni ṅ paksi madyus kumĕlab.

52. Alongside of the three principal active formatives, ma-, -ṅ, -um-, there are secondary formatives. I call them so because they are less widely distributed. I would mention the following as being the most interesting of them:

I. The formative r-, which can also unite with the formative a- to form ar-, and with ma- to form mar-. Here, too, as between r- and mar- we have the same relation as in the case of ṅ- and maṅ-. This r-, just like n-, was originally an article, in Old Jav. it is an unemphatic pronoun of the third person. The shorter form r- is rare, it is found in Karo, e.g., rělbuh, "to call" < r + WB ělbuh. The longer form is spread over Sumatra, Java, Celebes, Borneo, and the Philippines. In Bug. and Mak. the formative is maq- or aq-, in Tag. mag-, in conformity with the phonetic laws already explained. — Illustrations of the active formatives r- and mar-. Karo, from the Story of the Glutton: "There is somebody calling from down below " = Calls from below hither = rĕlbuh i těruh nari. Bug., from a letter of Lasiri's to Matthes, wherein he complains that the police arrest him when he is going about by night to make enquiries for Matthes about rare Bug. words: "If it is possible, (give me a letter of attestation)" = If possible + is it = bara maqkulle i. Mak., from the anonymous collection of Mak. Dialogues: "Do not shoot so hastily !" = Not you hastily shoot = teya ko karo-karo aqmaqdiliq.

II. The formative ba-, which can also combine with the formative r- into bar-, without change of meaning. This formative is widely distributed. But as a living formative if only exists in a few languages, for example in Day., which has both ba- and bar-, and in that Sumbawarese dialect which is known to us only by the Story of the Dog's Dung. This text only contains 27 lines, and yet there are to be found in it 5 distinct cases of verbs in ba-, such as ba-laṅan, "to go", ba-rari, "to run away"; but when we find in 27 lines 5 cases of verbs formed with ba-, we are entitled to regard that "formative as a living one. The Day. ba- forms intransitive verbs, and such too are the 5 Sumbawarese ones. — Apart from this there are isolated cases in many IN languages of verbs formed with ba.

Philippines, Bis.:	baigad, "to scrape".
Celebes, Mak.:	baloliq, "to roll up".
Java, Sund.:	bagěnah, "to be happy".
Islands at the back of Sumatra, Mentaway:	baliyu, "to fill".
Madagascar, Hova:	wawenti, "to be massive".

Note. — The WB of Bis. baigad, viz. igad, does not exist in Bis. itself, but is found in Iloko; similarly, the WB of Mentaway baliyu occurs in Mak.
53. Besides these active formatives there are very many others that occur occasionally in one language or another, e.g. Old Jav. a-, which alternates with ma-; Bug. keq-, which denotes possession; Day. me- or mi-, which has been discussed above, etc., etc.
54. Now of all these active formatives one language will possess a larger stock, another a smaller one. By way of example, let us enumerate all the living active formatives that are found in Toba:

Formative
ma-	maribak, "to be torn".
man-	mananto, "to pay attention".
mar-	marhosa, "to breathe".
masi-	masihoda, "to buy horses".
marsi-	marsibuni, "to hide oneself".
marha-	marhapili, "to be biassed".
marhu-	marhuraja, "to beseech".
maṅin-	maṅintubu, "to beget".
maṅun-	maṅunsande, "to lean against".
patu-	patuṅosṅos, "to clench the teeth (with pain)".
-um-	humordit, "to shiver".
-ar- or -al-	dumarede, "to trickle".

55. Now there are also cases in which the same WB and the same active formative run together through so many languages, that one is compelled to style the whole formation a Common IN one. Such a case is manali, "to bind" < maṅ + tali, "cord".

Philippines, Tag.:	manali
Borneo, Day.:	manali
Java, Old Jav.:	manali
Sumatra, Toba:	manali
Islands at the back of Sumatra, Nias:	manali
Madagascar, Hova:	manadi < manali.

56. The Causative. There is one Common IN causative formative, namely pa-. Proof that it is Common IN:

Formosa, Form, dialect:	paita, "to let see".
Philippines, Nabaloi:	pabunu, "to cause to kill".
Celebes, Bug.:	papole, "to cause to come".
Borneo, Tar.:	paakan, "to let eat".
Java, Sund.:	pasak, "to make well done {i.e.completely cooked)".
Islands near New Guinea, Kamberese:	palaku, "to let go".
Sumatra, Angkola:	pauli, "to make beautiful".
Islands behind Sumatra, Mentaway:	pakom, "to let eat".
Madagascar, Hova:	mam-paturi, "to let sleep".

57. Illustrations of the causative: Bont., from the Story of the Stars: "The mother made the brother fly" = Made + f. m. b. = inpatayaw ina kawwaan. Bug., from the Injilai: "He made them go aboard his vessel" = He made + mount them in ship his = na panoq i ri lopi na. Mentaway, from the Dialogues about the Priesthood: "They make them healthy" = Make -f healthy = paäru.
Note. — In inpatayaw the in- is the sign of the past tense. The WB of panoq is noq, "to mount up into". 58. In several languages the causative formative pa- takes one or other of the active formatives in front of it, without any modification of meaning; thus in Day. we find by the side of pa- a form mampa-, and it is a point not to be overlooked that Hova also has mampa-, whence the above mampaturi, "to let sleep".
59. Alongside of pa-, the chief formative of the causative, there is the less widely distributed secondary form paka-. We find it in the Philippines, Celebes, Java, and the islands at the back of Sumatra, e.g. in Nias under the form faqa–. — Illustration of this causative formative: Bareqe, from the Story of the Deer and the Water-Snail: "Pay particular attention !" = You let + be + alive ears yours = ni pakanaa talĭṅa mi. The WB is naa.
60. Besides the above-named formatives there are a considerable number of others forming the causative that occur more occasionally, in one language or another, e.g., pe-, pu-, and in Bug. -ěṅ, with which (as we already know) Nias -o is identical, etc.
61. As in the case of the active, so here too in that of the causative we will enumerate all the formatives that occur in a particular language. In this instance we select Bug., the examples are all from the Injilai:

Formative	pa-	panoq, "to let mount".
	maqpa-	maqpatěllon, "to erect".
	po-	powata, "to make a slave of".
	paka-	pakěda < paka + ěda, "to let speak ".
	-ěṅ	lěppěssěṅ, "to set free".

62. As in the case of the active, we will give an instance here of the same WB and the same causative formative running together through a number of languages. The WB is iram, which also appears under the forms idam, injam, etc., according to the phonetic peculiarities of the several languages. The WB means "loan", the causative, therefore, "to effect a loan", which expression is used sometimes for "borrowing" and sometimes for "lending".

Philippines. Tag.:	magpahiram
Celebes, Mak.:	painraṅ
Borneo, Bol.:	painjam
Near Java, Mad.:	apaenjham
Madagascar, Hova:	mampindrana.

63. The passive. There are two passive formatives that we can call Common IN : ta- and in. The latter, like the active formative um, is a prefix before words that begin with a vowel and almost always an infix in those that begin with a consonant. Proof that the passive formative ta- is Common IN:

Philippines, Bis.:	takilid, "to be inclined".
Celebes, Bungku:	tapeha, "to be broken".
Borneo, Tar.:	tadagu, "to be spoken".
Java, Old Jav.:	tawurag, "to be scattered".
Islands near New Guinea, Sawunese:	tabolo, "to be submerged".
Sumatra, Toba:	talentes, "to be opened".
Islands at the back of Sumatra, Mentaway:	taico, "to be seen".
Madagascar, Hova:	tahuruaka, "to be pierced".

64. The second passive formative is in. Proof that it is Common IN:

Celebes, Bulu:	winunu, "to be killed".
Borneo, Bol.:	jinawal, "to be lost".
Java, Old Jav.:	inambah, "to be trodden on".
Islands near New Guinea, Kupangese:	inka, "to be eaten".
Sumatra, Lampong:	tinabor, "to be strewn".
Islands at the back of Sumatra, Mentaway:	tinibo, "to be dried".
Madagascar, Hova:	tinapaka, "to be broken".

65. Other important passive formatives that are fairly widely distributed are, above all, ka-, next tar-, for which Bug. and Mak. have taq-, further -an, etc.

66. Illustrations of the passive formatives. Mak., from the Jayalangkara : “(Jayalangkara saw the snake), then he was frightened” = Was + frightened J. = taq-baṅka (mi) Jayalaṅkara. Old Malagasy, from the sermon Tonih Zańahary : “Noah was saved alive” = The N. was + saved + alive = ra Nuhu winelun. Basa Sangiang, from the 14th Dirge : “Supported by the sheath of a sword” = S. sword-sheath = kañokah kumpaṅ. Modern Jav., from the drama Prabu Dewa Sukma, 8th Act: " It is not known whether the corpse has been hacked to pieces or burnt " = Not + know, hacked + to + pieces, burnt = tambuh cinacah kabĕsmi.

67. Shades of meaning conveyed by the passive formatives. Of the two chief passive formatives, in and ta-, or tar-, the first generally forms a pure passive, the other often involves a suggestion of unintentional action, chance happening, or possibility. — Illustrations. Mak., from the third Elegy: “(I would keep her) in a casket that could not be opened” = In c. not to + be + opened that = ri patti ta taqsuńke ya. Mkb., from the Manjau Ari: "Innumerable is the number of the prawns " = Not to + be + counted number prawns = tidaq tabado bañaq udan.

68. As in the cases of the active and causative, we will here enumerate all the passive formatives that are to be found in a particular language; in this instance we select Day.:

Formative ba-⁠bakunci, " to be locked ".

i-⁠iagah, " to be led ".

ta-⁠talenten, " to be lopped ".

tar-⁠tar ajar, " to be teachable ".

tapa-⁠tapaisä, " to be counted ".

Note. — The Day. formative tapa- also appears, as tafa-, in Hova, a point not to be overlooked.

69. As in the cases of the active and causative, so here too in connexion with the passive we will give an instance of the same WB and the same formative running together through a number of languages. The WB is bunu, “to kill”, also “to fight”, or in other forms wunu, bono, etc.

Philiphines, Tag. : binono
Celebes, Bulu: winunu
Java, Old Jav. : winunuh
Sumatra, Toba: binunu
Madagascar, Hova: wununu.

Note. — In the Hova form the i of the infix has become assimilated to the u of the WB. In Section III we met with an analogous assimilation in Day., a point that is not to be overlooked.

70. Correlation of the passive with the active.

I. It very often happens that certain passive formations are closely connected with certain active ones. Old Jav. -um- forms actives with an aorist sense, and the like shade of meaning is indicated by the passive ka-; hence the active in -um- and the passive in ka-- are correlated together.

II. But just as often no such close connexion exists between active and passive forms respectively. Old Jav. has an active formative ma-, which mostly forms intransitive verbs, but also transitive ones : thus in the Śakuntalā there is the sentence: “No one did evil” = Not was man did evil = tātan hana wwań Jiiagawe hala. This active in ma- in Old Jav. has no specific passive correlative.

III. In Bug. there is a passive derived from the WB gauq, “blue” , viz. rigauq, “to be coloured blue”, but there is no corresponding active. — Alongside of the Hova manduka, “to throw a spear”, from the WB luka, there are two passives: aluka, “to be thrown”, said of the spear, and lukana, “to be hit”, said of the person.

IV. In certain IN languages it has become a regular custom for the transitive active to be accompanied by two passives. The Mlg. grammar calls one of them simply “the passive” , the other “the relative”.

71. Use of the passive. The passive is used much more frequently in IN than it is in the better-known Indo-European languages. And this phenomenon is so widespread that we must style it Common IN. Proof:

I. By the evidence of translations. The expression in the Sanskrit original of the Prasthãnikaparwa : phalam prāpnōti is in the Old Jav. recension rendered by : phala pińangih.* The prāpnōti of the original is active, pińangih on the other hand is a passive, formed with the infix -in-. — The passage in Tell, where Stüssi says: "He is now gone to fetch the bride at Imisee ", is rendered by Rizal by the passive construction: “The bride at Imisee is now being fetched (= susunduin) by him”.

II. By the evidence of enumerations. In the Day. story of the Chopper and the Buffaloes there are upwards of 24 passives in something under 100 lines. — The short episode beginning “sěrta ditikamña” in the battle of the five friends with the pirates in the Hang Tuah contains 9 actives and 6 passives. — In the detailed account of the battle in the Old Jav. Mausalaparwa the number of actives and passives is approximately equal; e.g. the following passage occurs: “Reeds were pulled up, they were used as weapons, for they turned into clubs wherewith blows were dealt on the adversaries”.

72. The frequency of the passive is to be explained as follows : In all the IN languages it is a matter of great moment to emphasize by linguistic means that element of the sentence which is considered the most important one. These means include: intonation, unusual syntactical order, particles of emphasis, and also the choice between the active and the passive construction. If the subject is to be put into the foreground, the active is chosen ; if it be desired to lay stress on the object, recourse is had to the passive construction, i.e. the object is made the subject. In the Toba Sangmaima, therefore, the construction is not as might have been expected: “The mother went to cook and killed a fowl for the dinner”, but: “and a fowl was killed by her” ; for the point of importance is not that she killed, but rather what it was that she killed.

* I.e., “to get one's due”.

73. The verbal systems.

I. We have learnt that the IN languages often have several formatives that perform jDrecisely the same function. If, for example, we analyse the Hova descriptive piece " Fiana-kaviana " in Julien, we see that the two formatives mi and man < ma + ń occur in it particularly frequently. It contains 6 verbs in man- and 7 in mi-. In all the 13 cases the formative simply creates an active transitive verb, without any. special shade of meaning; manasa lamba means " to wash clothes ", but mitutu wari " to pound rice". — In Old Jav. the two active formatives a- and ma- can be used for one another at pleasure, and the same applies to the Bug. aq- and maq-. If we analyse the Prasthānikaparwa from that point of view, we see that e.g. "he made" is represented on some occasions by sira agawe and on others by sira magawe. The meaning is absolutely the same, and so is the situation : in both cases the word that precedes the verb ends in -a. II. On the basis of the condition sketched in I. above, several of the IN languages have elaborated a series of verbal systems running parallel and side by side with one another, much like the Latin conjugations in a and e. As a specimen I here exhibit the two systems of Mentaway :

The a- System.

The a- System.

The fullest development of this principle is the elaboration of the Bug. verbal systems in a and e, which I have dealt with in a former monograph.*

• [" Sprachvergleichendes Charakterbild eines indonesischen Idiomes", §§ 84-99.]

____________________________________________

SECTION V: THE MOODS.

74. Among the moods the imperative is the one that is most elaborated in IN; it displays the greatest number of formatives. The conjunctive is much more scantily equipped. The modal shades of meaning represented by "can, may, must, shall, and will " are mostly expressed with the aid of auxiliary words of form, though the conjunctive can also perform those functions. The same applies to the irrealis. And it often happens that the sentence contains no linguistic element at all, apart from intonation, whereby we can recognize the mood.

75. The imperative. Nearly all the IN languages possess imperative WB's, i.e. WB's that exist only as imperatives. Examples: Nias aine, "come!", Karo ota, "let us go!", Day. hua, "attention !" — Illustration, Karo, from the Story of the Glutton: "Let us go home !" = ota ku rumah.

It is deserving of particular notice that practically all the IN languages have an imperative word for the idea "lo !", "behold !", though each language has a different one: Bont. nay, Nias hiza, Hova indru, etc. — Illustrations. Nias, from the Consecration Song on the gold ornament: "Finished is the jewel, behold ! perfected is the glittering of the gold" = noaway ganaqa, hiza ! nomaulu zaquso. Hova, from the Testament of Umbiasa: "Behold, (my) son, the counsels" = indru anaka ni anatra.

76. Apart from these imperative WB's, the active imperative is formed, in the first place, by omitting the formatives, or to express it more accurately, by uttering the WB in a tone of command, request, entreaty, and the like, so as to express this mood. Thus in the Day. Story about saying "H'm" we find the sentence: "Fetch the sirih-vessel and bring it here !" = F. V. s., b. h. = duan saraṅan sipa, imbit katoh. The indicatives of these imperatives are manduan<man + duan and mimbit < ma + imbit. This kind of imperative formation is found in all the IN languages, and is accordingly Common IN; even languages that generally employ some other method always exhibit a few cases of the one just described. In Hova I know only the one instance: fuha, “wake up !”, the indicative of which is mifuha. — Illustrations. Old Jav., from the Kuñjarakarna: “Go, then, into the underworld !” = laku la mareṅ Yamaloka. Modern Jav., from the drama Prabu Dewa Sukma, Act I.: “And, elder brother, hasten ! ” = Ian, kakań, gupuh. Bug., from the Epigrams against Cowardice: “Retire, you cowards !” = E,. y. c. the = esaq ko kelow e. Hova, from the Oracular Formulas of Amurunkay : “Awake, oracle !” = fuha sikidi. Achinese, from the Story of the Wise Judge: “Speak the truth !” = S. with truth = kěhěn bak těpat.

77. Secondly, a very widespread mode of forming the imperative is to use the indicative as imperative without any change save in intonation. Of all the Mentaway texts the second Story of the Great Bear contains the greatest number of imperatives: there in we find formations like: “You, fish !” = äkär manuba. This manuba<man + tuba is also indicative. Alongside of it there are formations like pana, “shoot !” This pana is also indicative, so here there is nothing omitted in the imperative, as there was in the Day. duan above. — In Matthijsen's Tettum dialogues there is a passage: “Where shall I lay them ? Lay them here !” = Shall lay where ? Lay here ! = atu tau basa ? tau banee. Here the same word tau is both indicative and imperative, it is a simple, underived verb. — In the Tontb. Sacrificial Prayer there is a sentence “Drink there, you gods !” = mělěp ańe, e kasuruan ; and in the Dirge for the Dead Mother we find: " It is not yet time to drink " = Not + yet time drink = raqipeq toro mělěp. Here the same form mělěp is both indicative and imperative, and it is a derived verb, formed from the WB ělěp.

78. A third mode of forming the imperative consists in the use of specific imperative formatives differing from those of the indicative and conjunctive.

I. A fairly widely distributed formative is pa-, which in this case has, of course, no causative meaning. It is found in the Phihppines, in Magindanao; in Java, in Old Javanese; in Sumatra, in Toba; in the islands at the back of Sumatra, in Nias. —Illustrations. Old Jav., from the Kuñjarakarna: " Clasp his feet ! " = C. f. h. = pamĕkuli jön ira. Magin- danao, from Juanmarti's Collection of Dialogues: “Wake up ! I am awake already” = Awake ! Am + awake I already = pagedam ! nakagedam aku den.

II. Nias has the formatives a- and o-, i.e. the m of the indicative form beginning with ma- or mo- is omitted. Thus from a WB gule, “vegetables”, are formed an indicative mogule, “to cook vegetables”, and an imperative ogule. Illustration, from the Story of the Strange Cook: " Well, cook vegetables !" = lau, ogule.

We have noticed on several occasions that Hova has special relations with Day. But it also shares all sorts of peculiarities with the languages of Sumatra and the islands at the back of Sumatra. Thus the Nias mode of forming the imperative is also found in Hova: from leha, “step”, are formed the indicative mandeha, “to go”, and the imperative andeha. Illustration, from Rahidy's Fable of the Crocodile: “Let us (= isika) go (and) swear blood-brotherhood !” = andeha isika himafatidra.

79. A fourth kind of imperative is constituted by using the conjunctive as an imperative.

I. In Old Jav, -a forms the conjunctive, in Modern Jav.j the formations in -a are used as conjunctives and as imperatives, in the dialects of Madagascar only as imperatives. Still, even in Old Jav. we already find passages where -a has an imperative function. — Illustrations. Old Javanese, from the Āśramawasanaparwa : “Conclude an agreement!” = gumawayakĕna ṅ sandhi. Old Mlg., from the sermon Harireunau: “Assent!” = meteza hanaw (= 2nd person singular pronoun).

II. We have learnt that the formative -um- produces aorists, futures, and conjunctives. Hence in some languages it is also used to form imperatives as well, e.g. in Tontb.

Illustration, from the Dirge for the Dead Mother: "Step down here !" = tumuli mai. — In Tag. -umn- is the regular formative of the imperative, the indicative having a different one. — Illustration, from Tell: "Quick, old man, set to work!" =

dali, matanda, gumawa.

80. We know that the IN languages have no word corresponding to the Indo-European copula "to be". Therefore sentences in which, in our languages, the copula forms, or introduces, the predicate, have no verb in IN. In such cases the imperative is expressed merely by intonation. Illustration, Mentaway, from the Fishermen's Stories : "Be my bride!" = Bride my you = madi ku äkäw.

81. The imperative of the causative is analogous to that of the active. In Hova maturi is "to sleep", mampaturi, "to cause to sleep"; the corresponding imperatives are: maturia and mampaturia. In Bug. the causative patĕtton is both indicative and imperative. — Illustration, from the first executioner's story in the Injilai: "Erect it (= the house) at midday!" = At midday of day the you erect it = ri tĕnĕsso na ĕssow e mu patĕtton i.

82. The imperative of the passive has the formative -en, in Tag. -in, in Toba -on, in conformity with the phonetic laws already stated. This formative has a considerable distribution, being found in the languages of the Philippines, Northern Celebes, Java, and Sumatra. — Illustrations. Tag., from Tell : " Forget it now (and live only for joy) !" = Must + be + forgotten by + you = limutin mo. Tontb., from the first Vampire Story: " Let us only look for crabs !" = Must + be + sought by -f us only c. = umuṅĕn ta reqe komaṅ. Toba, from the Sangmaima: "What then must be done by me?" = beha ma bahenon ku.

83. The imperative is often accompanied by particles, which make it stronger, or milder, or more polite, and so on. This usage may be called Common IN, In Old Jav. and Bont. ta is used in that way, in Day. has, etc., etc. Illustrations.

Old Jav., from Mpu Tanakung's Prosody: “ Do hurry!” = D. h. you = ta iṅgal kita. Bont., from the Story of the Brothers and the Rat: “ Do let us go into my house!” = Do go we i. h. m. = ta umüy tako is afoṅ ko. Day., from the Story of the Inner Bark of the Tree: “ Well,” said Hatalla, “ be it so!” = has, koan Hatalla, jadi.

84. Later on we shall meet with a widespread particle ma, mo, mě, ma-ma, ma-lah, etc., which serves to emphasize the predicate. It is also used extremely often with the imperative. Illustrations. Toba, from the Sangmaima: “ Prepare provisions!” = bahen ma bohal. Mkb., from the Manjau Ari: “ Do smoke tobacco!“ = isoq malah santo. Mak., from the Epic Maqdi: “ Only say it (and we will act according to your words) ” = Say only = maqkana mama.

85. The IN languages have two kinds of negatives, one for the indicative and another for the imperative. This phenomenon is so widespread that we must call it Common IN. Thus Masaretese has the two negatives mohe and bara.—Illustrations. Masaretese, from the Garuda Story: “ His children did not grow big ” = His children the not big = rinenake ariat ro mohe haat; “ Do not be malicious!” = Not you m. = bara kimi walekuk.

But in many IN languages the prohibitive negative takes the indicative, not the imperative. In Hova “ to rule ” is, in the indicative, mandzáka, the imperative is mandzaká < mandzaka + a. Now in the Testament of Umbiasa we find: “ Do not rule with the flesh, rule with the spirit! ” = aza mandzáka ami ni mifu, mandzaká ami ni fanahi.

86. The Conjunctive. Only a few IN languages have a formative for this mood; it is, therefore, not a Common IN phenomenon. Frequently it is not expressed at all; or else only by means of auxiliary words of form, such as the Mal. baraṅ, “ possibly ”. Special conjunctive formatives exist in particular in Old Jav. and Bont., the former using -a, the latter -ed, or after a vowel -d. As already mentioned, the formative -um- my also be used to form the conjunctive; that occurs in Tontb. I. The conjunctive in Bont. Illustration, from the Battle of Kaloqokan: “ We ought to go to Bontok ” = G. w. should to B. = umüy kami-d ad Funtok.

II. The conjunctive in Tontemboan. Illustration, from the Story of the Burning of the Vampire: “ Do go and tell them !” = Go do tell to them = maṅe oka kumua an isera. Here the conjunctive kumua, from the WB kua, is dependent on the imperative mane.

In Bont. and Tontb. the conjunctive is not often used; in the Battle of Kaloqokan — 192 lines of print — there are only 2 cases. But it occurs quite regularly in Old Jav.

87. The conjunctive in Old Javanese. Its use, whether as a dependent verb or independently, coincides ahiiost completely with the Latin usage.* Thus an analysis of the whole of the Kamahāyānikan — 63 printed pages — has yielded the following results:

I. The conjunctive of reserved utterance: “ (You have now been instructed, and so your defects) have probably disappeared ” hilaṅa, corresponding to the indicative hilaṅ.

II. The conjunctive of request: “ Let (rice, drink, etc.) be offered ” = wehakěna.

III. The conjunctive of condition: “ If (freedom from desire) be attained (then Buddhahood is also won) ” = an kapaṅguha.

IV. The concessive conjunctive: “ Even though (no beauty) is seen (in your teacher, nevertheless be amiable towards him) ” = yadyapi katona.

V. The conjunctive in sentences denoting intention: “ (Strive after Advaya), in order that (your defects) may disappear ” = yatānyan hilaṅa.

VI. The conjunctive after verbs of command: “ (The order shall be given) that (these men) be slain ” = pějahana. VII. The conjunctive after verbs of permission: “(It is not permitted) to indulge in (love in the temple)” = gumawa- yakĕna.

VIII. The conjunctive after verbs of doubt and hesitation : “(Do not hesitate) to practise (the holy Samaya)” = gumawa-yakĕna.

88. The Optative is either an imperative, or else it makes use of special auxiliary verbs, or, most frequently, the above-mentioned particles of emphasis, ma, mama, lah, malah, etc. — Illustration, Mkb., from the Manjau Ari: “May he quickly grow big !” — Quickly “lah” big = dareh lah gadaṅ.

89. The Potential has the formative maka-, which has a considerable distribution, being found in the Philippines, Celebes, and Madagascar. — Illustration, Bont., from the Story of the Stars: “But he cannot fly” = But not can fly = ya adi makatayaw. — Or else auxiliary verbs meaning “can” are used, e.g. in Karo banci. — Illustration, from the Story of the Glutton: “What then can (one) do ?” = kuga kin banci bahan.

90. The modal shade of meaning represented by “I will” is often expressed by the future, which in Nias for example has the formative da-; or by the conjunctive ; or by means of auxiliary verbs meaning “will”, e.g. in Gayo male. Illustrations. Nias, from the Story of the Old Cat: “Where is the old thing, I want to kick it to death” = Where old + one, “da” + I + kick + dead = hezo nina, da-u-hundrago. Gayo, from the small vignettes: " I will turn back " = aku male ulak.

91. The modal shades of meaning represented by “must, shall, may” are rendered by paraphrases like “it is necessary, it is good, it is seemly”, etc.; and this type of phrase is so widely distributed that we must style it Common IN. Illustration, Toba, from the Sangmaima: “The spear Siringis must not get lost” = Not good, lost + go s. S. = naso tupa maga hujur Siriṅis. 92. The Irrealis. On account of the interesting character of this mood from the point of view of general comparative philology, we must consider it in some detail. It has several different modes of formation, but none of them are Common IN.

I. Formation of the irrealis by reduplication of the first syllable of the verbal WB. This is found in Mentaway; but an analysis of all the Mentaway texts — 80 pages of print — has only yielded three instances of it. Illustration, from the Love Dialogues: “Then there would be naught good in me” = Not would + be good in + me = ta babara uktuk ku.

II. The irrealis is expressed with the same means as the conjunctive, future, or passive imperative. — Illustrations. Mkb., from the Manjau Ari: “Who should have taught me ?” = W. s. + h. + t. me = siya ka-maajari den. Karo, from the Story of the Glutton: “One would have thought he was dead” = mate ninĕn. — ka- is the sign of the future, -ĕn the sign of the passive imperative.

III. Mas has a special auxiUary word of form for the irrealis, viz., enao, which is put after the verb. Illustration, from the Story of Kawofo: “Fain would Kawofo have eaten” = F. K. eat “enao” = omasi Gawofo ia enao.

Note. — The initial of Gawofo follows from the laws of the “status constructus”, which will be dealt with in Section IX. 93. The IN languages have three means of forming tenses: formatives, auxiliary words of form, and reduplication.

94. The Present Active. We have given, in Section IV, the formation of the three genera. The verbal forms of the active, which we there ascertained, are in some languages presents, in others they have no implication of any particular time, and so can be used for the present. — To that rule there are, however, exceptions. We have already learnt that the formative -um- is used in certain languages as a future, in others as an imperative.

95. The Past Active. The past tense is formed either by means of formatives or with the help of auxiliary words of form. The first type of past has formatives which are characterized by the possession of the sound n. This mode of formation is found in the Philippines ; the intermediate islands south of the Philippines; in Northern Celebes; in Nias, at the back of Sumatra; and in Madagascar. So its distribution is quite a wide one; but, on the other hand, it is to be noted that this type is wanting in Old Jav.; and, moreover, the formative is not the same in all the above languages, though it everywhere contains an n.

96. Now the first way of forming the past tense is to add to the active form, as delineated in Section IV, the formative ni- or no- or in- or -in-.

Formosa, Form, dialect:	linummis, Pres. lummis, “ to glow ”.
Philippines, Bont.:	inumjanak, Pres. umjanak, “ to arrive ”.
Intermediate islands, Talautese:	inumire, Fut. umire, “ to nod ”.

Northern Celebes, Tontb.:	nimaali, Pres. maali, “ to bring ”.
Islands at the back of Sumatra, Nias:	nomofano, Pres. mofano, “ to start ”.
Madagascar, Hova:	nutunena, Pres. tunena, “ to be calmed ”.

Note I.—The Form, vocabulary only gives the forms, without telling us what tenses they represent; thus we simply find: lummis, linummis. But as the better known Magindanao conjugates precisely like this particular Form, dialect, we may conclude from it that linummis is a past tense:

Word-base	Present	Past
	Mangindanao
lutad, “ to lower ”	lumutad	linumutad
	Formosan dialect
lis, “ to glow ”	lummis	linummis

Note II.—Hova nu- and Nias no- are identical; in accordance with phonetic law Hova represents the o of other languages by u.

Note III.—In Talautese the form with an m is not a present but a future.

Note IV. —The past formative nu- is also found in Toba (in Sumatra) in the extended form nuṅ or nuṅa. This consists of nu and the emphatic particle ṅa or ṅĕ, which recurs in several IN languages; in Karo, which is closely related to Toba, ṅĕ after vowels also appears as ṅ, e.g. in the Story of the Glutton, l. 28.

97. Illustrations of the past formations of the preceding paragraph. Tontb., from the Story of the Founding of the Village of Kapoya: "But Asaq set forth from Sondĕr" = sapaka si Asaq (ya) nicumĕsot (ai) an Sondĕr. Nias, from the Story of Buruti: "My mother has gone away" = G. + a. mother my = nomofano nina gu. Toba, from the Sangmaima : "The spear is lost, (dragged away by wild pigs)" = "nuṅa" lose spear = nuṅa mago hujur. 98. Secondly, the past tense is formed by replacing by n the m of the present formatives ma-, mar-, mi-, etc., which we became acquainted with in Section IV.

Philippines, Bont.:	nalufug, Pres. malufug, “ to perish ”.
Intermediate islands, Talautese:	namali, Fut. mamali. “ to buy ”.
Madagascar, Hova:	natahutra, Pres. matahutra, “ to fear ”.

99. Illustrations of the past formation of the preceding paragraph. Bont., from the Lumawig: “ Then all the people had perished ” = keceṅ nalufug amin nan taku. Hova, from the Fable of the Crocodile: “ Then replied the hedgehog ” = dia namali ni sukina.

100. Thirdly, the past tense is indicated by auxiliary words of form. Nearly all these words mean “ finished, completed ”. This linguistic phenomenon may therefore be characterized as Common IN, But the same identical word of form seldom runs through several languages: thus Bug. uses pura, while the closely related Mak. has leqbaq, Old Jav. huwus, Kupangese hidi, etc. — Illustrations of this type of past tense. Old Jav., from the Kuñjarakarna: “ Your words have entered into my bones ” = huwus the w. your penetrate come into the b. = huwus ika pawarah ta anusup tĕka i ṅ tahulan. Kupangese, from the Story of the Fool: “ We have now made the hole ” = We make “ hidi ” hole now = kit sukun hidi bolo son.

101. Words of form such as have been given in the preceding paragraph are not, as a general rule, used to indicate merely that the action took place in the past; there is nearly always an idea of completion or of the pluperfect bound up with them. When it is merely a matter of past time, these words of form are hardly ever used, the present, or rather the verbal form implying no particular time, then does duty as a past tense. — Illustrations, from Bug. letters: “ I have drunk the whole of the medicine (and now my motions are really not so painful as they were) ” = Completely “ pura ” I drink m. = aṅkana piira uw inuṅ pabura. — “I have bought a house (and now I have not got enough money to pay for it)” = House I buy = bola uw ĕlli. — In the first case the idea of completion is emphasized, hence the use of pura; in the second case that notion is not present, so the word is not used. — In the sentence from the second executioner's story in the Injilai : “Scarcely had it taken (the poison, when it died)” = “pura” it takes = pura na ĕmmĕ, the word pura indicates the plu- perfect.

102. In the languages which form the past by means of genuine formatives, the formative may either simply indicate past time, or completion, or the pluperfect, as well. Should it, however, be desired to throw special emphasis on the fact of completion or the pluperfect, then these languages, besides using the formative, also add words of form meaning “finished, complete”, like those mentioned above. Thus Hova uses efa, which is identical with the Nias efa in aefa, “finished”. — Illustration from the Testament of Umbiasa: “When knowledge has entered into our mind (it cannot be taken away from us again)” — The k., when “efa” entered into m. = ni fahendrena, raha efa tafiditra an tsaina.

103. Those languages which possess genuine formatives for expressing the past employ them in a very consistent manner. So in the Hova sentence from the Fable of the Cuckoo : “They would not receive him” , both verbs, “would” and “receive”, are in the past tense: Not would have + received him = tsi neti nandray azi. — In this respect Nias is an exception; it does not use its past formative no- very frequently. In the Story of Buruti — 46 hnes of print — no- only occurs about half a dozen times. That, however, is without counting the phrase no-mege, “above mentioned”, as its meaning has faded and it almost does duty as an article.

'104. The languages which express the past by means of words of form employ them, as already stated, almost exclusively in cases where it is desired to express completion or the idea of the pluperfect; but even in such cases the words of form are often omitted. In the second executioner's story in the Injilai, we find in 147 lines of print less than half a dozen puras.

105. The Future Active. This tense has four modes of formation in IN: first, it is expressed by genuine formatives; secondly, by enclitic, mostly monosyllabic, words of form, which are on the way to become formatives ; thirdly, by disyllabic, somewhat more independent, words of form; and, fourthly, by syllabic reduplication.

106. First: the future made by means of formatives. None of these formatives has a wide distribution, none therefore can be styled Common IN.

I. Hova replaces the m of formatives beginning with m by h. Example: "to kiss" = miuruka, "to have kissed" = niuruka, "to be about to kiss" = hiuruka. This h-formation has originated by analogy with the hu-formation, which will be discussed hereafter.

II. Sund. forms the future by means of the prefix pi- and suffix -ön, e.g., pidataṅön, "to be about to come".

III. In some languages the aorist formative um is used to form the future, e.g., in Tontb. and Bont., thus Bont. umoto, "to be going to cook".

IV. Some languages employ for the future the formatives beginning with m-, which are elsewhere present or have no implication of time; so Talautese mamali, "to be going to buy".

V. Old Jav. uses its conjunctive formative -a as a future, e.g., matya < mati + a, "to be going to die".

107. Illustrations to the foregoing paragraph. Old Jav., Mausalaparwa, from the prophecy of the angry Brahman: "Baladewa will not die at the same time" = Baladewa tan ilu matya. Hova, from the Testament of Umbiasa: "I shall be gathered to the forefathers" (i.e., shall die) = S. + b. + g. to the f. I = hihauna ami ni razana ahu. 108. Secondly: the future made by means of enclitic words of form. These are all prepositions indicating the direction "whither". How they come to denote the future has been discussed in Section III.

I. The Western languages, those of Sumatra, the Malay Peninsula, and Madagascar, employ the preposition ka, which in extended form becomes kan and akan, and in variant form ku. Thus ka is used as a preposition, e.g., in Mal.; akan in Day.; ku in Karo. As a sign of the future ka is used, e.g., in Mkb.; akan in Day.; hu < ku in Hova.

II. The Eastern languages, including those of Southern and Central Celebes, employ the preposition la, which in certain languages has the form da in conformity with the RLD-law. As a preposition la occurs, e.g., in Kupangese; the sign of the future is la in Mak., da in Bareqe, and, far away from the Eastern group, in Nias.

III. Bont. uses ad as a preposition and as a sign of the future. It is conceivable that this ad is identical with the da = la in II. above. For, in the first place, metathesis is a very common phenomenon in the IN languages; and, secondly, another preposition in Bont., synonymous with ad, appears both as is and as si.

109. Like Hova, Day. also employs the preposition ku to indicate the future, but it has made it into a verb: maku. Precisely analogous cases are the Bug. matu < ma + tu and the Tettum atu < a + tu; the preposition tu exists independently, e.g., in Toba.

110. Illustrations to the two preceding paragraphs. Mkb., from the Story of Manjau Ari: "I shall go" = den ka pai. Mak., from the Jayalangkara : "I shall now go and sleep" = Shall go I sleep = la maṅey aq tinro. Bareqe, from the Story of the Migrating Mouse: "I shall emigrate" = S. e. I = da melinja (mo) yaku. Tettum, from Matthijsen's Dialogues: "To-morrow morning (the) horses will come" = awan saivan kuda atu mai. 111. The third way of indicating the future consists in the use of auxiliary words meaning "will" and the like, e.g. hěndaq in Mal., issa in Bont., etc. — In the use of the second and third methods we meet with the same state of things as we noticed in connexion with the past tense : the particles are often omitted. In the Mal. Epic Bidasari we ought to find the future expressly indicated more especially in Canto 5, which deals largely with the making of plans for future action; but we find hardly any cases of it.

112. Some languages employ several of the modes of forming the future, for instance Mak., which has both la and sallaṅ, and Bont. All the methods possible in Bont. are represented in the Story of the Battle of Kaloqokan:

I. The future is not indicated by any linguistic means: "When will they come ?" = When c. t. = kad (nan) alian ca ?

II. Future with formative -um-: "We shall run away" = S. + r. + a. we = lumayao kami.

III. Future with ad: "It will be much" = ad aṅsan.

IV. Future with issa: "You will come, the three of you" = W. y. c. the three = issa kayu ('d) sumaa ay tolo.

113. Among the languages which form their tenses with genuine formatives, there are some that have only elaborated two tenses : thus Magindanao only has the present and the past, and has to use a periphrasis for the future. Other languages form all the three tenses, and also bring the imperative into the ambit of this system of tenses. Such elaborated systems are found in Formosa, the Phihppines, the intermediate islands, e.g. Sangir, and Madagascar. Example:

	Hova	Tagalog
WB	tadi, "to bind"	tawag, "to call"
Present	manadi	tuṅmatawag
Past	nanadi	tuṅmawag
Future	hanadi	tatawag
Imperative	manadia	tumawag

114. Illustrations of the tense-system in Madagascar, from the Old Malagasy Sermons:

I. Present: "Abu Bekr, who fears the Lord" = The A. B., f. t. L. = r' Abubakiri matahutru an Dzaṅahari.

II. Past: "My heart has stored it up" = Heart my h. + st. + up = fu ku nitarimi.

III. Future: "They will not see me" = N. w. s. m. = tsi hahita ahi.

115. Tenses of the Causative. These follow the active closely. Example from Hova :

	"to speak"	"to cause to speak"
Present	miteni	mampiteni
Past	niteni	nampiteni
Future	hiteni	hampiteni

Illustration from Bont., which, as we already know, uses pa- as the sign of the causative and in- as the sign of the past tense; from the Story of the Stars: "Mother has made our brother fly" = Has + caused + to + fly m. b. o. = in- patayaw ina kawwaan mi.

116. The tenses of the passive sometimes follow those of the active pretty closely, as in Hova; in other cases, as in Tag., they diverge further from them. Examples:

I. Hova —

WB	ume, "to give"
Present	umena,"to be given"
Past	numena
Future	humena
Imperative	umeu

II. Tagalog —

WB	tawag, "to call"
Present	tinatawag, "to be called"
Past	tinawag
Future	tatawagin
Imperative	tawagin

'117. A peculiar way of forming the past passive is found in Bug. and the closely related Mak. Bug. pura and Mak. leqbaq, which mean “finished, completed”, are put before the simple underived WB; it is to be remembered that the WB, as already mentioned above, often has in itself a passive signification. Bug. example:

WB siyoq, “to bind”

Past Active pura maqsiyoq

Past Passive pura siyoq.

It is true that this formation is not found very frequently; in the Injilai the first instance occurs in the first executioner's story. — Illustration of this mode of forming the past tense, from the third executioner's story in the Injilai: "A man who had also been bitten" = worowane pura oqkoq to.

---

SECTION VII : THE PERSONS.

118. The IN languages often possess two parallel series of personal pronouns: full forms and short forms. Thus in very many languages the full form of "I" is aku, while its short form in Bont. is ak, in Mal. ku, in Old Jav. k. Full forms are found in all the languages, so that this linguistic fact must be called Common IN; the distribution of the short forms is less extensive, though we find them in nearly all the great areas of distribution: in Celebes, Bug. possesses them, but Tontb. does not. In certain languages the series of the short forms is incomplete; others, on the contrary, have two series of them.

Note. — With the etymological relation between the full and short forms of the pronouns we need not deal here, as this is a monograph on the verb. Nor need we speak of the relation between the short forms of the personal pronouns and the possessive pronouns, which also appear as a species of short forms; e.g. Bont.: "I": full form = saken; short form = ak; possessive, "my" = ko. — Besides, I have said something about this subject in a former monograph.*

119. The full forms of the personal pronouns accompany the verb, either as subject or as object, in precisely the same way as substantives do. Thus in the Day. Story of the Inner Bark of the Tree we find :

The Inner Bark went = I. + B. the w. = keaṅ-ñamo tä hagoet.

He went = iä hagoet.

Hence in what follows we shall have but little more to do with the full forms ; we simply refer the reader to Section IX. On the other hand the short forms are eminently deserving of the attention of linguistic students, more particularly of those

_{* ["Sprachvergleichendes Charakterbild eines indonesischen Idiomes", §§ 65, 157 seqq.]} who are concerned with the Indo-European languages and those who devote themselves to the study of languages in general, for the combination of these forms with the verb represents the commencement of a conjugation.

120. We have already learnt that of the languages which possess short forms of the personal pronouns some exhibit incomplete, others complete, series.

I. In Mal. only the pronouns " I " and " thou " have short forms :

	Full form	Short form
I	aku	ku
Thou	ěṅkaw	kaw

II. Mentaway has the two series complete, save that for the second person plural the full form and the short are identical :

	Full form	Short form
I	aku	ku
Thou	äkäw	nu
He	iña	i
We	sita	ta
You	kam	kam
They	sia	ra

Note. — The Mentaway grammar does not mention the short form i, but there are passages in the texts which admit of no other interpretation than the existence of such a short form. A passage of that sort occurs at the beginning of the first Story of the Great Bear: " Father, he twines yarn at home " = ukui i puputärä bakä ka uma.

III. Bug. in addition to a series of full forms possesses two complete series of short forms:

	Full form	Short form
I	iyaq	u or ku	aq
Thou	iko	mu	o or ko
He	iya	na	i
We	idiq	kiq	kiq
You	iko	mu	o or ko
They	iya	na	i

121. Use of the short forms. In this paragraph we always mean the use of the short form as subject; its function as object will be discussed later.

I. The full forms are employed when the emphasis is on the subject; when that is not the case, the short forms are used. In the Bug. letter from the Princess Weyanu (i.e., Princess X) to Matthes, which is about Bug. manuscripts, occurs the passage: “I myself will give orders to convey them there” = I self order, convey them = iyaq pa maqsuro panoq i. Here the word pa, “self”, shows that the emphasis is on the subject. In the same letter there is mention of some fragrant oils, but there we find : “I have not handed them over to him” = Not I have + handed + over him = deq u pati wiriw i. Here the important point is the predicate, and hence the short form u is used for the subject.

II. When the emphasis is on the subject it very often happens that both forms, the full and the short, are used together. So in the Mak. Jayalangkara we find the sentence : “(Jayalangkara was without fear, but his brothers cried:) 'We are exceedingly afraid!” = Fear exceedingly we we = mallaq duduw aq inakke. Here the emphasis is on the subject “we”, because of its antithesis to Jayalangkara, and it is expressed twice, by the short form aq and the full form inakke

III. In the case of the third person, when that is expressed by a substantive, the short form of the pronoun is often added as well. But this does not involve any emphasis or any other special effect. If we find in the above-mentioned letter about manuscripts: na ala i karaeṅ riy anu = “He has taken them, the Prince of X”, this means no more than: “The Prince of X has taken them”.

IV. There are also certain limitations in the use of the short forms, which vary from one language to another. Usually they only accompany the active and causative forms of the verb. Mal. employs the short forms only with transitive verbs. In Mak. this limitations does not hold good; illustration, from the dialogue of the cats in the Jayalangkara: “Come on, let us go” = unibamo kiq lampa. 122. We now come to the question : In what manner do the short forms of the personal pronouns combine with the verb?

I. Where the language in question has only one series of short forms: in some languages they precede, while in others they follow, the verb. This does not depend on the usual order of the subject, be it a substantive or a full form of pronoun. In Mal. the subject as a rule follows the predicate, but the short forms of the personal pronouns have to precede it. — They precede the verb in Toba, Mal., Bareqe, Tettum, etc., but follow it in Bont.

II. Where the language in question has two series of short forms: in that case the one series always precedes, the other follows, the verb; that holds good, e.g., of the two Bug. series given above, the u-series precedes, the aq-series follows.

123. Illustrations to the foregoing paragraph.

I. Position of the forms before the verb. Bareqe, from the Song about the Beloved Relatives: "I value (them) like gold" = ku timba ewa wuyawa. Mak., from the Sixth Elegy: "God, I pray" = Batara, ku kanro. Old Jav., from the fifth canto of the Rāmāyana: "He bent this bow" = r ayat ikanaṅ laras. Mentaway, from the Love Dialogues: "I will not" = Not I w. = ta k' oba < ta ku oba.

It chances that the short forms of all the personal pronouns occur in the Nias Story of Buruti and Futi:

I — "I know my mother well" = I k. w. mother my =u ila sa nina gu.

Thou — "Why dost thou steal my child ?" = W. t. s. c. my =hanawa o tago nono gu.

He — "The ghost Buruti spake" = He s. B. g. = i mane Buruti-beχu.

We — "We will speak" = We s. w. = ta waqo dania.

You — "Give me the child !" = You g. me c. = mi beqe χogu nono.

They — "They have stolen my child" = la tago nono gu.

II. Position of the short forms after the verb. Bont., from the Battle of Kaloqokan: "They run into the wood" = R. they i. w. = umüy ca id pagpag. 124. In languages that possess two series of short personal pronouns the speaker is free to choose between those that precede and those that follow the verb. In the Jayalangkara we find : "When you arrive at Masereq, you will ascend the mountain" = W. a. y. there at M., y. w. mount up m. = punna baltu ko maṅe ri Masereq, nuw eroq naiq ri moncoṅ, but it could equally have been: nu battu and eroq ko. That appears most plainly from an analysis of the dialogue of the two cats in the Jayalangkara, when they want to go to Masereq, for nowhere else in the whole of the Jayalangkara are the pronouns as frequent as in that passage: we see there that the two series are used indiscriminately.

125. When the short forms of the personal pronouns precede the verb, some languages omit the active formatives, others do not. In Mal. the word for "to see" is mĕlihat, but "I see" is ku lihat, the mĕ- being dropped. In Rottinese "to seek" is akaneni < aka + WB neni, "he seeks" is nakaneni, the aka- being retained. — Illustrations. Mal., from the Hang Tuah: "I have taken it away from you again" = + t.-a again from you = ku ambil pula daripada mu. Rottinese, from the Animal Play: "He seeks the man" = H. s. m. the = nakaneni touk a.

126. The most interesting question is as to the degree of intimacy that exists in the combination of short personal pronoun and verb. In some languages the connexion is a close one, in others it is looser.

127. The looser combination. This is found, e.g., in Bug. and Mak.:

I. In these languages it is not absolutely essential that the pronoun should come immediately next to the verb. In the Jayalangkara we find : "Go you!" == maṅe ma ko, where the verb and pronoun are separated by the emphatic particle ma.

II. Genuine suffixes effect a shifting of the accent, but pronouns put after the verb do not. III. Pronouns can be used with other parts of speech besides verbs.-Illustration, Bug., from the Injilai: "You are a man, I am a bird" = tau ko, ku manuqmanuq.

IV. The short form need not necessarily be the subject, it may be the object.

128. The closer combination. This is found, e.g., in Mai. Here the short form of the pronoun only goes with verbs; between the pronoun and the verb nothing can be interposed ; ku lihat can only mean "I see", not " see me". — The accent does not come into question here, for words that precede do not influence the accent of what follows.

129. But the most intimate combination of the short pronoun and the verb is to be found in Rottinese. The verbal formatives in Rottinese begin with a vowel; and the short forms of the pronouns, which precede the verb, have lost their vowels, e.g. "he" = n < na. Hence it has become possible for the short pronoun and the verb to coalesce into a real unit. For instance, the WB for "to flee" is lai, the verb alai, " he flees" is nalai. Specimen paradigm:

WB	hani
Verb	ahani
I wait	ahani
Thou waitest	mahani
He waits	nahani
We wait	tahani
You wait	mahani
They wait	lahani

Note I.—Rottinese usually also puts the full form of the pronoun before this conjugated verb, e.g. "I wait" = au ahani.

Note II.—The transition from the short form of the pronoun with a vowel to a form without a vowel is neatly illustrated in Mentaway. Whereas in Mak. "I" is only ku and "thou" only nu, in Mentaway one can say either ku or k, nu or n, when a vowel immediately follows. Thus in the same Love Dialogue we find: " Why don't you care to? " = Why not you care = apa ta nu oba, and " Don't you care for me? " = ta n oba aku .

130. Illustrations of the Rottinese conjugation. It so happens that the whole conjugation is represented in the Animal Play:

I — " I say " = au ae.

Thou — " Don't you know, then? " = Then thou not know = le o ta malelak.

He — " He seeks the man " = nakaneni touk a .

We — " We flee " = ala talai.

You — " Eat! " = mua leonma.

They — " That they may not see me " = That they n. s. me = fo ala boso lita au.

131. In the languages in which the short pronouns are closely connected with the verb, the pronoun only does duty as the subject: Mal. ku lihat can never mean " see me ". In languages where the connexion is a less intimate one the pronoun can also serve as the object. — Illustration from the Mak. Jayalangkara : " (It were better that we roam around than that) the snake should eat us " = Us it eat, snake = kiq na kanre naga.

132. In the languages which possess two series of short pronouns, such as Bug., Mak., and Nias, when at the same time one pronoun is used before the verb and another after it, then the first one is the subject and the second the object. — Illustrations. Bug., from the Injilai: " I kill you " = u sampelle o. Nias, from the Story of Buruti : "I love you " = u omasiqo o.

____________________

133. We have learnt that the IN verb can express genus, mood, tense and person. That, however, does not conclude the cycle of its vital manifestations.

I. We have already heard that certain languages are able to express the beginning of an action, by means of the aorist formative um. — Now some languages can also indicate duration: thus Old Jav. has given the active formative man- and the passive formative -in- a durative tinge which did not originally belong to them. Illustration, from the Āśramawasanaparwa: "As long as Draupadi was being ill-treated" = D. as + long + as she was + being + ill-treated = Dropadī kāla nira winudan. — Other languages, again, are able to express the fact that something just intervenes during the continuance of an action. But here we do not meet with genuine formatives, but merely auxiliary words of form such as těnah, sědaṅ, or sadaṅ, and the like. Illustration, Mkb., from the Manjau Ari: "A woman was engaged in weaving" = A person woman "sadaṅ" wove = sa oraṅ padusi sadaṅ batanun. — Bont. has a formative naka- to indicate the conclusion of the action. Illustration, from the Story of the Rat and the Two Brothers: "Now they have finished eating" = Now h. + f. + e. they = keceṅ nakakanan ca.

II. Here and there we also find participial formatives. Nias has a present participle in s-, the formation of which has been discussed in Section III. — Illustration, from the Story of the Fish and the Rat: "A woman drawing water" = Woman d. w. =alawe sanaqu idano.

III. In several IN languages number can be expressed. Masaretese has a verbal formative which is da- in the singular, and du- in the plural. Illustration, from the History of the Tagalasi Tribe: "He saw the inhabitants of Tagalasi-Miten sitting there" = Saw i. the T.-M. sit = daaṅak geba ro Tagalasi-Miten duptea. — Nias uses an infixed g as a sign of the plural. Illustration, from the Story of Buruti:

Thou weepest	mee o
They weep always	mege-ege ira.

In Gayo the formative i indicates the plural either of the subject or of the object, the latter in the following sentence from the Story of the Blue Princess: "She had all her clothes on" = All had + on clothes = mbeh sěloki pěkayan.

Note.—All the phenomena mentioned in this appendix to Section VII occur only sporadically; we cannot draw from them any conclusions as regards Common IN.

SECTION VIII : VERBAL PHRASES AND OTHER

MODES OF EXPRESSION.

134. IN often uses the verb in cases where the Indo-European languages with which we are more generally familiar employ a substantive, adverb, etc. ; but the opposite also holds good.

135. IN forms abstract substantives just as Indo-European does. Thus from the WB ro,"to come", which is also used without any formative as a verb, Toba derives the substantive haroro, "arrival" < ha + ro reduplicated. — Illustration from the Sangmaima : "In order that they may know the time of my arrival" = In + order + that be + known time of arrival my = asa diboto bakta ni haroro nku.

136. Now the IN languages often use a substantival costruction in cases where as a general rule the better known Indo-European languages adopt the verbal construction; and that applies, in particular, to the verbs "to do ", " to intend ", "to think", "to say", "to be named ". This phenomenon can be styled Common IN. The IN substantives in question are either substantival WB's like Old Jav. don, Tag. ibig, "intention", or else derivative substantives like Old Jav. pagaway, "the making", which exists alongside of the WB gaway and the verb magaway.

I. Substantival construction with the ideas of "doing", "making". Old Jav., from the Prosody of Mpu Tanakung: "Well, what had you to do?" = What then making your = mapa kari pagaway ta.

II. With the idea of intention. Tag., from Tell: "Why do you crowd upon me (in the open road) ?"— What the intention your with me = ano an ibig niniyo sa akin. Old Jav. from the Āsramawasanaparwa: "That is what they desire to attain" = (That that) may + be + attained, (is) aim that = kapaṅgiha don ika.

III. With the idea of saying. Tontb., from the Story of the Demon that haunts women in confinement: "Then they exclaimed" = Then speech their = siituoka kiia era (o). Nias, from the Story of Buruti: "Then she spake thus" = Then thus speech her again = ba simane li nia zui. Day., from the Story of the Inner Bark: "Then said he to himself" = Then (was) word of the heart = tä koa n huaṅ.

IV. With the idea of thinking. Karo, from the Story of the Glutton: "But I think the Glutton is dead" = Heart mine yet, that dead the Glutton = ate ku min, ĕṅgo mate si Lagaman.

V. With the idea of being called. Old Jav., from the Śakuntala: "There was once a king, who was called Duswanta = Was a k., D. name his = hana sira mahārādja Duswanta naran ira. Tontb., from the Description of the Sacrificial Feast: " This sacrificial feast is called the offering for the plants " = Name of s. + f. this (is) plant- offering = naran i papeliqin itu manusew.

137. On the other hand, a verb is often used in IN in cases where the Indo-European languages with which we are more generally acquainted would employ some other part of speech. All the instances here enumerated are Common IN.

I. The verb replaces an indefinite pronoun. The verb that serves this purpose is the verb " to be, to be in existence ", Old Jav. hana, Mal. ada, Nias so. — Illustrations. Old Jav., from the Kunjarakarna: " Some had their heads chopped off " = Were, (whose) chopped -j- off -f- were heads their = hana winadun kapala na. Mai., from Abdullah's Journey: " There were several islands, some big and some small " = Were several piece islands, were small, were big = ada beberapa buwah pulaw, ada kecil, ada besar.

II. The verb replaces a preposition, namely, the prepositions "at" or "about" etc., with verbs denoting an emotion. The verbs doing duty in these cases are "to see" and "to hear".—Illustrations. Day., from the second Sangumang Story: "His mother wondered at what Sangumang said" = M. his wondered hearing words S. = indu e heṅan mahiniṅ auh Saṅumaṅ. Old Jav., from the Āsramawasanaparwa: "The king wept, being touched by his condition" = Wept k. t. seeing c. his = manaṅis mahārāja kasrĕpan tumon gati nira.

III. The verb replaces adverbs such as "up", "down", "out", "back", and the like.—Illustrations. Kupangese, from the Story of the Fool: "Then he stepped in into the midst" = Then he s. entering into the m. = ti un laok tama se tlala. Day., from the second Sangumang Story: "I walked along by the side of the gigantic chopper" = I w. going + along (by) g. + c. = aku mananjoṅ mahoroy pahera.

IV. The verb replaces the affirmative 'particle.—Illustration, Nias, from the Story of Futi: "Have you heard the words of the chief? She said: 'Yes!'" = You h. w. c.? She s.: 'I heard' = o rono li razo? i mane: u roṅo.

138. With respect to several of the passive forms of the verb there is a controversy as to whether they should not rather be regarded as substantival forms. My view is the following:

I. The name "passive forms" is given by the grammars to certain linguistic phenomena which were undoubtedly originally substantives. In the Mkb. work Manjau Ari there is a sentence: "It grows in the field, surrounded by trees". "Surrounded by trees" is: diliṅkuaṅ kayu. The form diliṅkuaṅ is explained by the traditional grammar as a passive, and it is further added that the agent kayu is annexed without the preposition "by". But the WB liṅkuaṅ is also a substantive in Mkb., meaning "something that surrounds"; and di is also a preposition; so I could also take di liṅkuaṅ kayu as meaning "in a ring-fence of trees", for the genitive relation is expressed in many IN languages, and in Mkb. in particular, by the mere order of the words without the intervention of any preposition. Thus in my view diliṅkuaṅ kayu was originally a substantival construction.

II. In the Old Jav. Kuñjarakarna we find the sentence: "He came upon a door" = A d. was met by him = babahan kapangih de nira. Here, no doubt, one might also regard kapangih as a substantive, for ka- serves in many IN languages to form substantives as well as verbs. But here the agent is attached by means of the preposition de, which never introduces a genitive relation. I cannot, therefore, without putting a strain upon it, construe the sentence as: "the door was a find of his". On the contrary, de corresponds rather with the kind of prepositions that are used in our languages to introduce the agent in passive sentences; that is shown by active constructions like the following from the Āśramawasanaparwa: "To undergo pain through you" = maněmu lara de ñu.

III. In not a few IN languages the agent can be introduced in both ways, either genitively or by a preposition meaning "by, through", as in Mal., where the preposition is oleh. An analysis of the whole of the Mal. work Hang Tuah has resulted in showing that though the construction with the preposition preponderates, the genitive construction is freely represented there too, and no difference in meaning is perceptible as between the two modes of expression. For example: "It was heard by the chief" = didĕnar batin; but: "It was heard by the mother" = didĕnar oleh ibu. This interchangeability of the two forms indicates, to my mind, the occurrence of a transformation in the mental attitude with which they are regarded: what was originally substantival has gradually come to be felt as verbal.

IV. In Bug. the agent is never introduced genitively, but always by the preposition ri, which never indicates the genitive. I have analysed the whole of the Paupau Rikadong — 29 pages of print — from that point of view, and have not found a single exception. That, to my mind, shows that what the Bug. grammar calls " passive " is really felt to be passive, even though one or other of the forms of the passive may have been originally substantival.

V. In individual cases it will often be difficult to put oneself into the mental attitude of the IN native so as to be able to determine whether, for him, a given linguistic phenomenon which the ordinary grammar calls a passive is really substantival or verbal. — Speaking with some reserve, I incline to the view that the construction under I.: diliṅkuaṅ kayu is really felt as a passive by the Mkb. people of to-day, and is, therefore, a verbal phenomenon, as the ordinary grammar says it is.

---

SECTION IX : THE VERB IN THE SENTENCE.

139. Among the linguistic means employed by IN for linking the several parts of a sentence together the following are of articular importance: prepositions, the copula, the status constructus, and the syntactical order of the words.

140.Prepositions.

I. In Common IN the genitive relation has the preposition n or ni. This reposition links substantive with substantive. There are no active or causative verbs that " govern " the genitive. But, as we have heard, the agent is often linked with the passive predicate like a genitive. — We have also learnt that the genitive can be expressed by the mere order of the words, without any preposition at all.

II. In some languages the dative relation has a special preposition, e.g. in Karo: man; in other languages the dative is expressed by the same prepositions as the adverbial.

III. The accusative relation is very rarely indicated by a preposition; as a general rule, the syntactical order suffices.

IV. In Common IN the adverbial is introduced by prepositions. Of these the two prepositions i and ri have a particularly wide distribution.

141.The copula.

I. The Indo-European copula, the verb " to be ", has nothing exactly corresponding to it in IN. Hence the sentence, " What is the reason that it is so ?" is expressed in the Kuñjarakarṇa by: " What reason of it, so?" = apa dumeh ña maṅkana. In Achinese, in the Story of the Pehcan, we find the sentence, " Then exclaimed all the fish: ' It is good so ' " rendered by: " Then exclaimed all fish: ' Thus good ' " = těmar sěut bandum ěṅkut : ño mĕnan. In Bug., in a letter of the Princess Aru Panchana we find the expression: "What is the price of this gold thread?" = What price of it gold thread this = siaga ělli na wěnampulawěn ede.

II. IN possesses a feature which the grammars rightly call a copula. Only it is not a verb, but a particle, viz., i or ay or ya, etc., which links different parts of the sentence, particularly subject and predicate, together. The copula is found in the Philippines, in Northern Celebes and in Madagascar; it cannot, therefore, be called Common IN.

142. The status constructus, in Nias.

It is formed principally in two ways: words that begin with a vowel take n or g before that vowel; words that begin with a surd turn it into a sonant. Many words do not form the status constructus at all. Examples:

	Mother	Rat
Status absolutus	ina	tequ
Status constructus	nina	dequ

The status constructus serves the same purpose as the copula, it links the several parts of the sentence together, especially the subject with the predicate. "Rat" is tequ and "to go" is moi, and in the Story of the Rat and the Fish the phrase "the rat goes" is: moi dequ.

Note.—The first method of forming the status constructus may be thus explained: the sounds n and g are the prepositions n and ka, respectively. The preposition n has been mentioned at the beginning of this section; ka, which, in accordance with the Nias phonetic laws stated in Section III had to become ga, has been repeatedly referred to. But we have also learnt that n is a Common IN preposition for the genitive relation and ka for the place "whither". We must, therefore, assume that the two prepositions have considerably enlarged their sphere of action, or have made it more general. This assumption is rendered credible by the fact that there are parallel processes in Mentaway, which is a near neighbour to Nias geographically and shares a number of special features with it. Now in Mentaway the preposition ka has also considerably extended its functions, so that it is now able to introduce almost any syntactical relation; and moreover before vowels it often appears in the abbreviated form k.*

143. The emphasizing of the predicate.

I. Nearly all IN languages possess particles which serve to emphasize some particular part of the sentence, so that this phenomenon must be styled Common IN. The most widely distributed is the particle ma, which also appears as mo, mĕ, mama, and man; it occurs in the Philippines, in Celebes, in the islands lying near New Guinea, and in Sumatra. Mal. has an emphatic particle lah, Mkb. ma + lah, etc.

Note.—Though some languages have ma, others mo, others again mĕ, this change of vowel does not as a rule correspond with the phonetic laws of the languages in question; we must, therefore, provisionally call it variation.

II. Now though it is true that these particles can be used to emphasize any part of the sentence, yet they are most frequently put after the predicate. In the Sumbawarese text about Dog's Dung — 27 lines of print — mo occurs eleven times, of which nine are cases where it follows the verb.

III. A minority of the languages makes but sparing use of the particles of emphasis; thus Kupangese, for example, where in the Story of the Fool — 10 pages of print — ma occurs only once, viz., in the phrase baku ma, “it is enough”. The majority use them very plentifully, e.g. Toba. In the Toba Story of Sangmaima the particle ma occurs in the ordinary course of the narrative after nearly every predicate. Illustration: “Sangmaima ate, and then took his provisions and went into the depths of the forest” = Then ate the S.,

* [The second mode of formation of the status constructus is explained by the author in his monograph “Indonesisch und Indogermanisch im Satzbau”, § 180, as resulting from the carrying on of the “voice” of the final vowel (with which all Nias words end) onto the initial consonant of the following word, when it is closely connected with what immediately precedes, thus changing the unvoiced consonant into a voiced one. See also Essay IV, § 302, II.] then were + taken p. his, then went into forest deep = asa maṅan ma si Saṅmaima, asa diboan ma bohal na, asa laho ma tu tombak loṅo-loṅo.

144. Sentences that have a predicate but no subject.

I. The indefinite pronouns “it” or “one”, used as subjects, are not as a rule represented at all in IN. — Illustrations. Mentaway, from the Contest between Sun and Moon: “I am well, it is raining” = aku maäru, urat. Bareqe, from the Song to the Moon: “It gets dark, before one wends one's way homewards” = Gets dark, before go + home = maweṅi (mo) nepa jela.

It is true that the subject is sometimes indicated in cases of that kind. In the Bont. Story of the Stars the expression “it is growing dark” (i.e., night is coming on) is sometimes rendered by malafi and at other times by malafi nan talon, “the daytime is becoming night”.

II. When the verb is in the imperative the pronominal subject may be added or omitted. Languages that possess short forms of pronouns are fond of adding the subject in such cases. Thus in the Bug. Paupau Rikadong the king says to his servants: “Go, then !” = Go then you = lao sa o. In Mak., in the expressions “don't” = teya ko and “don't let us” = teya kiq, by means of which the prohibitive is formed, the pronoun always appears; an analysis of the whole of the extensive work Jayalangkara has hardly revealed a single exception to this rule.

145. Having in the preceding paragraph dealt with sentences that have no subject, we have now to speak of sentences that have no verbal predicate.

I. It has already been mentioned that IN possesses no verb corresponding to the Indo-European copula.

II. When in an IN sentence there is an adverb or a preposition indicating a direction in space, the verb of going, coming, or remaining, which would be the predicate, is often omitted. — Illustrations. Old Malagasy, from the Sermon Tonih Zaṅahary; “Where art thou, Moses ?” = Where thou, the M. = aiza hanaw, ra Musa. Day., from the Story of the Inner Bark of the Tree: “For what purpose do you come?” = For what you = akan kwe kaw.

III. In IN sentences we very often find the predicate accompanied by an interjection, mostly an onomatopœic one. — Illustrations. Old Jav., from the Kuñjarakarna: “Hey presto ! (she) was at the door” = rĕp datĕn ri ṅ lawaṅ. Rottinese, from the Animal Play: “There is a flash; bang ! the musket rings out” = nandela; daṅ ! sisilo nali.

Now in such sentences as these the verb may also be omitted, so that the interjection by itself plays the part of a predicate. Illustration, Toba, from the Sangmaima: “Then cried the kite ‘ hulishulis ’ ” = Then “hulishulis” k. = asa hulishulis (ma) lali.

The phenomena mentioned in this paragraph are to be regarded as Common IN.

146. Linking the subject with the predicate. First method: the syntactical order.

I. The predicate precedes the subject; this rule is Common IN. — Illustrations. Old Form., from Vlis’s Collection of Dialogues: “You have evidently been sleeping” = Have + slept y. e. = nimesip kaw lawa. Old Jav., from the Tantri Fables: “Thus spake the goose, then answered the tortoise” = Thus word of the g., a. the t. = maṅkana liṅ n ikaṅ haṅśa, sumahur ikaṅ pās. Toba, Sangmaima, from the Burning of the Book of Magic: “Then his book of magic was burnt, but a leaf of it fell behind his house” = Then burned magic + book his, fell at back of house his one leaf magic + book = asa gor (ma) pustaha na, timpal (ma) tu pudi ni ruma na sa lompit pustaha.

II. But this order is not absolutely obligatory. If special emphasis is to be laid on the subject, it may precede. — Enumeration: the Mentaway Story of the Spirit of the Palm Toddy contains 25 lines and three instances of the order subject + predicate; and in each case this occurs because the subject is to be emphasized, on account of an antithesis. III. Certain isolated languages, which have no close connexion with one another, follow the opposite order as a general rule. To this category belong Day. and Masaretese.— Illustrations. Masaretese, from the Story of the Forest Spirit: “Meanwhile men ate pouched rats”* = Meanwhile m. the a. p. + r. = gamdi geba ro ka tonal. Day., from the first Story of Sangumang: “The buffaloes were penned by me” = B. the w. + p. by + me = hadaṅan tä kuroṅ ku. — Enumeration. In the Masaretese Legend of the People of Tagalasi the order subject + predicate is strictly maintained.
IV. The order of the short forms of the pronouns has been dealt with above.
147. Linking the subject with the predicate.
Second method : The copula. This copula, the particle i or ya or dia, etc., is interposed between subject and predicate, thus linking them together. Illustrations. Tag., from Tell: “That cries to Heaven” = iya i sumisigaw sa laṅit. Tontb., from the Story of the Water Snail and the Antelope: “I have been requested by the antelope” = I "“ya”" h. + b. + r. by a. = aku ya tinaqaran i tuqa. Hova, from the Testament of Umbiasa: “The body requires nourishment” = ni nufu dia mila hanina.
148. Linking the subject with the predicate. Third method: the status constructus in Nias. The subject, which follows, is put into the status constructus. — Illustration. from the Kawofo: “Then appeared Kawofo” = ba so Gawofo.
149. The agent relation in the passive sentence. This has been discussed in Section VIII.
150. Predicate and predicative. The predicative is simply added, without more, to the verb of the predicate. This phenomenon is to be regarded as Common IN. — Illustrations. Modern Jav., from the History of the State of Kěḍiri : “He was made commander-in-chief” = kadadosakěn senapati. Sund., from Van der Ent's Descriptions of Animals

* Cuscus moluccana.

and Plants: “Their leaf ribs are manufactured into brooms” = L.t.a. + m.b. = nere na dijiyön sapu.

151. Predicate and infinitive. In this case also there is, as a rule, merely juxtaposition. — Illustration, Bug., from the letter of Princess X, wherein she asks Matthes for a copy of the Jayalangkara : “That is what I wish to say to you” = That I wish say to you = iya uw aqkatta powadada r' idiq.
Bont., in this case, employs the copula. Illustration, from the Battle of Kaloqokan: “We are going to take” = Going we "ay" take = umüy kami ay umala.
152. Predicate and direct object.
I. The Common IN rule is that the direct object is added, without more, to the predicate. — Illustrations. Simalurese, from Westenenk's small Collection of Dialogues: “May I take these coconuts ?” = M. I t. c. these = dai u abe bonol ereh. Sangirese, from Adriani's Songs: “He longs for the absent ones” = Desires men absent the = měkati tau tadi e. Banggayan, from Riedel's text: “We looked for bulbous roots” = ikami moṅombolii baku.
This order is not absolutely obligatory. If the object is to be emphasized it may precede. In the Masaretese Story of the Forest Spirit there is one case of the order object + predicate, occasioned by the object being emphasized on account of antithesis: “(Hereafter I shall only come with my voice), but you shall not be able to behold my body again” = But body the you get see no more = bu fatan di kimi beta aṅak mela beka.
II. In a very small number of languages the direct object is linked to the predicate by means of a preposition. This is done in Bont. by means of is and in Hova, in certain cases, by means of ani. — Illustrations. Bont., from the Battle of Kaloqokan: “Then we buy the cake” = T. b. we the c. = keceṅ lumago kami is nan kankanen. Hova, from the Testament of Umbiasa: "Men have begotten thee" = ulumbeluna (nu) niteraka ani ialahi.
III. In Nias the direct object is in the status constructus. Illustration, from the Story of the Woman who wanted to eat the Lightning: “Wrap up the dog !” = You w. + u. d. = mi fanombo nasu. — The status absolutus is asu.
153. It very often happens that verbs which are transitive in the Indo-European languages most familiar to us also take a direct object in IN. I have analysed from this point of view the Old Jav, tale which is embodied in Mpu Tanakung's Prosody, and the result is as follows. Many verbs take the accusative that also take it in German; but contrary to the German idiom are: “Helfen gegen Liebespein” (To be of help against the pangs of love) = atuluṅ rimaṅ; and “Den Vergnügen nachjagen” (To race after pleasures) = aṅrarah rüm.
154. All IN languages have verbs of motion which are construed with the accusative, particularly the verb "to go into". — Illustrations. Karo, from the Story of the Glutton : "They went into their house" = T. w. h. their = si dahi rumah na. Mal., from the Hang Tuah: “He went into the house” = masuq rumah.
Amongst all the IN languages of which I have analysed texts of some length, Day. is the one that displays this phenomenon most frequently, and I have found it occurring oftenest in the two Sangumang Stories:

palus huma	to enter into the house
tamä huma	to go into the house
lumpat huma	to mount into the house
blua huma	to come out of the house
huli lewu	to go back to the village
sampay kaleka	to arrive at the place
mahoroy pahera	to walk along the gigantic chopper.

155. In many IN languages we find the phenomenon of a preposition, particularly the preposition i, coalescing with the verb. Such verbs in that case require no further linguistic means to link up the object, though per se it be an indirect one, or even the adverbial ; in other words, they are construed transitively. Example, Bug.:

To go on a path:	joppa ri lalěṅ.
To tread a path :	joppai lalěṅ.

I have dealt with this matter in considerable detail in a former monograph.*

156. The object used with reflexive verbs demands special consideration. The IN methods differ to this extent from those of the Indo-European languages best known to us in that they do not say, for example, "I betake myself" but "I betake my body" or "my person". This phenomenon is Common IN. "Body" or "person" is awak in many languages, ale in Bug., droi in Achinese, and so on. — Illustrations. Basa Sangiang, from the First Dirge for the Dead: "Remove yourselves upstream !" = Remove person your upstream = tasat arep m ṅaju-ṅaju. Bug., from the first executioner's story in the Injilai: "It is to be feared that the king will repent him of it" = To + fear that repent person his king the = ajaqke na sěssěi ale na aruṅ e. Achinese, from the Story of the Wise Judge : "He made himself in shape hke a man" = He shaped person like mankind = ji pěrupa droi sěpěrti manusiya. — Enumeration. In Section I mention was made of an enumeration which showed that reflexive verbs are rarer in Mal. than in German, for instance. An analysis of the whole of the Jayalangkara has had a similar result; yet in the highly coloured description of the fight waged by the hero with the inhabitants of Masereq, for example, there are three consecutive cases: "to guard oneself from", "to hurl oneself against", "to throw oneself upon".
157. It is a neat coincidence that not a few IN reflexive verbs find their pendants in French. Bug., as the above example shows, has a parallel to "se repentir", Mal. to "se taire", Mak. to "s'évanouir", and Mkb. to "s'agenouiller".
158. The indirect object. It is the Common IN rule that a preposition is used to link the indirect object with the predicate. — Illustrations. Mak., Jayalangkara, from the king's speech if " This I lay in charge upon you" = iya (mi) ku

• ["Sprachvergleichendes Charakterbild eines indonesischen Idiomes", §§ 116-123.]
  

† Matthes ed., pp. 143 seqq. palaq ri kaw. Old Jav., from the Āśramawasanaparwa : "To reflect on death" = atutura ri pati. Bont., from the Kolling: "Tell it to our mother !" = T. you to m. our = kana m ken ina ta. Hova, from the Testament of Umbiasa: "He who follows after the low fellows becomes a low fellow (himself)" = izay miaraka ami ni amhualambu, dia amhualambu.
159. I have analysed the three books of the Old Jav. Mahābhārata entitled Āsramawasanaparwa, Mausalaparwa, and Prasthānikaparwa from the point of view of studying the dative object, which in Old Jav. is construed with the two synonymous prepositions i and ri, and the result is as follows. The dative object occurs especially:
I. With verbs of saying, asking, answering, commanding, and greeting; e.g., maněmbah ri, "to ask someone respectfully".
II. With verbs of thinking, knowing, remembering, and forgetting; e.g., atutur ri, "to think of".
III. With verbs of desiring, rejoicing, being content, and being sorrowful; e.g., alara ri, "to mourn over".
160. Since, as has been shown above, many IN languages possess formatives which make verbs, that would otherwise take an indirect object or an adverbial, into transitive verbs, it follows that in such languages the indirect object seldom appears. In the Karo Story of the Glutton the first dative object, introduced by the preposition man, does not occur till 1. 100: "This stone is suitable for a seat" = Suits stone this for seat = měhuli batu ndai man pěrkundulkundulěn. On the other hand, by way of contrast, the quite short Tontb. Story of the Python contains half a dozen cases of the dative object.
161. Nearer and remoter object together in a sentence. From what has been said above it follows that the former comes next to the predicate without the intervention of any preposition — and the latter then follows, accompanied by a preposition. I have analysed the whole of Jonker's Book of Laws on this point and find a great many cases of agreement with the corresponding German idiom, e.g., to give refuge to
an offender; to give wages to the man; to lend a weapon to an offender; to deliver, to pledge, etc., a thing to someone. But contrary to the German idiom, "to accuse something to a person", meaning "to accuse a person of something".
162. The adverbial is worked into the sentence by means of prepositions. Among the widely distributed prepositions are di, ri, "in, on, at", etc., and ka, "to, towards"; Tag. has sa, "in, at"; Timorese bi, "in"; etc., etc. — Illustrations. Lampong, from Ophuijsen's Collection of Mousedeer Stories: "If you come down to the water I shall catch you" = If y. descend to w., you I c. = asal niku turun di way, niku ku tĕkĕp. Běsěmah, from Half rich's Collection of Proverbs: "Where is there any ivory that has no flaw?" = In what i., not flawed = di manĕ gadiṅ diq běrětaq. Kangeanese, from the Story of Kandhulok: "On his way Kandhulok arrived at a ricefield" = O.-h.-w. the K. a. at r. = sa-jhalan-jhalan-na se Kandhulok těppaq ka saba. Timorese, from Jonker's text: "What is smelling in the room in there?" = saan nafo hi keen nanan. Togianese, from Adriani's small Collection of Texts: "I will not live in my village here any longer" = Not + more I will live h. in v. my = tamo ku poru maroro iriqi ri lipu ṅku. Nabaloi, from Scheerer's Collection of Dialogues: "We eat on the march" = Eat we during m. = maṅan tayo chi chalan.* Tag., from Tell: "You are my guest at Schwyz, I am yours at Lucerne" = You g. mine at S., I the yours at L. = kayo i panaohin ko sa Schwyz, ako aṅ iniyo sa Luserna.

• Scheerer gives no description of his ch, but as he has based his spelling on the Spanish usage it seems likely that his ch is identical with the sound rendered in these Essays by c.

1. _{* [As to the abbreviations, see Essay II, § 15.]}
2. [The analogy of the English "I am going to dine" is so close that the author would doubtless have mentioned it if he had been writing in English or for English readers.]