CC BY
49
ЯЗЫКОЗНАНИЕ
УДК 81-112.2
Alexander Ogloblin
ON THE VOICE CONFORMITY IN WEST AUSTRONESIAN DOMAIN
To Gennadii Evgenjevich Rachkov, the founder of Philippine studies in Saint Petersburg
The existence of several passive voices in Philippine languages implies that the subject of a passive sentence corresponds to different object arguments (actants) in an active sentence. In Tagalog, such objects may denote undergoer/patient, addressee, beneficiary, instrument, stimulus or location.1 Examples (1-3) quoted from Gennadii Rachkov’s book on Tagalog morphology [5, p. 140] show the use of verb forms from the base tanong ‘ask (question), enquire’ which in fact is the verbal noun ‘question’2. It has passive forms with the prefix i- implying that it takes the question’s content as a subject, and with the suffix -in implying that the question’s addressee is the subject. The glossing in (1-3) follows Rachkov’s analysis. The forms in (1-3) belong, in his terms, to the indefinite aspect. It is opposed to the durative aspect, which is marked by partial reduplication of the root, as u-uwi ‘come back’ in (1). Indefinite and durative aspects are labelled perfective and imperfective respectively by other authors.
In sample sentence (1), the passive is marked by the prefix i-, and the infix -in- denotes the realis mood. The addressee argument of the asking action is marked by the preposition sa, and the content of the question is a clause in the subject function (preposition sa is also used with common nouns, meanwhile proper nouns, as in (3), have an oblique article instead).
(1) Gabi-gabiu ay i-t<in>anong nila
every: evening2 COP <REAL>-PV-ask 3:pl sa akin kung3 kailan ka u-uwi to 1sg when3,4 2 DUR-come back
1 The problem of subject in Tagalog has been widely discussed. Some criteria of the arguments with the article ang and its equivalents in Tagalog were proposed, see [1, p. 151 sqq]. In Malay and Javanese, the subject is readily identified through relativization. Cf. [2, p. 108; 3, p. 237; 4, p. 599].
2 Translations from Russian and accent symbols are author’s own. To be precise, verbs in different voices are derivatives, not inflectional forms. “[T]he voice affixes are essentially derivational when viewed from a cross-linguistic perspective" [6, p. 364]. The author uses the traditional label form.
© Alexander Ogloblin, 2011
‘Every: evening2 they ask me [lit. is asked by them to me], when will you be back’.
In (2), the infix -in- is again the marker of the realis mood, and the passive has the zero al-lomorph of the suffix -in.3 The addressee argument of the asking action is the subject marked by the article ang.
(2) T<in>anong-0 niya ang babae <REAL>-ask-PV 3sg ART woman
‘He asked [was asked by him] a woman’.
Both passive forms correspond to only one active form. The actor voice prefix nag- in (3) denotes the realis mood.4 The addressee Danding is the first object with the oblique (locative/dative) article kay. The second object is again a clause.
(3)Bawa’t1 isa2 ay nag-tanong kay Danding kay3 ano4 ang lagay everybody^ COP AV:REAL-ask Art PN how3,4 ART state ng kanya-ng ama-ng maysakit
GEN 3sg-LIG father-LIG ill
‘Everybody asked Danding about (how is) the state of his ill father’.
Tagalog passive forms are oriented to semantic functions of the subject. Thus, bivalent (semantically monotransitive) verbs of different meanings have different passive forms. The majority of them, e.g. those meaning ‘to use’, ‘to create’, ‘to call’, ‘to look for’, ‘to clean’, ‘to see’ have passive forms with the suffix -in or zero, whereas passive forms of the verbs denoting ‘to nurse’, ‘to study’ or ‘to burn’ have the suffix -an, and passive forms of the verb ‘to draw, to depict’ have the prefix i-. The subject of all these passive verbs corresponds to the object without preposition in the active. There are also bivalent verbs requiring prepositional object in the active (again with different passive markers -an, -in / zero or i-), verbs with variable government in the active, and verbs without active counterparts (passiva tantum) [5, p. 136-138].
Malay (Ml) and Old Javanese (OJv) are distant relatives of the Philippine languages; at present, Malay has two standard varieties, Indonesian and Malaysian. It is generally accepted that the proto-language of the Austronesian family was in existence several thousand years ago in the Taiwan area. From there, some Austronesians expanded in the southern direction which led to the formation of Malayo-Polynesian (MP), a major branch including all Austro-nesian languages outside Taiwan. The Philippine archipelago was the first vast region where the MP immigrants settled. They expanded into Indonesian archipelago through later migrations, presumably between 1500 and 500 BC. The island of Java was in all probability reached by MP speakers from the North.
3 The absence of the undergoer passive suffix -in in realis mood might have another interpretation if it is acknowledged that the infix -in- which is used only in the in realis mood passive, does have, besides the modal meaning, the function of a passive marker, with or without the suffix -in. According to Lina Shkarban, the modal meaning in this infix is not separated from that of voice [7, p. 99]. Some authors, as noted by Shkarban, refuse identifying grammatical meanings of mood, aspect and voice with separate affixes; “the verb form as a whole appears to represent a cluster of grammatical meanings” [7, p. 100] (translated from Russian by the author).
4 The realis prefix nag- alternates with the irrealis prefix mag-; both are not relevant to aspect. The same is true
for other pairs of alternating prefixes in various verbal derivations: maki------naki-, maka-------naka-, ma------na-, etc.
Compare p. 117 below.
In Ml and OJv, the subject in passive sentences generally corresponds only to the first object in non-passive sentences, providing that this first object denotes undergoer, addressee, beneficiary, instrument or place. Reduction of several Philippine passive forms is compensated by transitive derivational affixes, or applicative affixes according to current terminology. Thus, in Ml the passive prefix di- is indifferent to the semantic function of the passive subject and is applied to all transitive verbs with no exception. Semantic distinctions are related to transitive affixes that are part of the verb base. The most important are verbs with the suffixes -i and -kan, sometimes labelled the first and the second transitive respectively, depending on the suffix.
For instance, the Indonesian passive form di-tanam-i ‘to be planted upon’ requires the subject which denotes a location, such as garden or field, as in (4a). On the other hand, the passive form di-tanam (or di-tanam-kan which presupposes a deliberate or intensive action) ‘to be planted’ requires an undergoer of the planting action (e.g. rice or maize) to appear in the subject position, as in (4b).
(4a) Kebun itu di-tanam-i jagung oleh Amir garden that PV-plant-TR1 maize by PN ‘That garden was planted with maize by Amir’.
(4b) Jagung di-tanam(-kan) oleh Amir di kebun itu maize PV-plant(-TR2) by PN in garden that ‘Maize was planted by Amir in that garden’.
Both passive forms have active counterparts with the prefix meN-: menanam-i and mena-nam(-kan); the rule that “the subject in the passive is the first object in the active”5 applies to both.
(5a) Amir menanami [meN-tanam-i] kebun dengan jagung PN AV:plant-TR1 garden with maize
‘Amir plants the garden with maize’.
(5b) Amir menanam(kan) [meN-tanam-kan] jagung di kebun itu PN AV:plant-TR2 maize in garden that
‘Amir plants maize in the garden’.
From the base -baca ‘to read’ a benefactive -baca-kan ‘to read to/for someone’ is formed, and a derivation -baca-i which means ‘to read a lot’ and is used in the same construction as -baca.
(6a) Buku-buku itu di-baca / di-baca-i oleh Amir book-pl that PV-read / PV-read-TR1 by PN ‘These books were read by Amir’.
The active voice marker is the prefix mem-.
(6b) Amir mem-baca buku PN AV-read book ‘Amir is reading a book’.
5 N is the variable nasal consonant m, n, ny [n] or ng [q]. It adapts to initial phonemes of the root of which [p], [t], [k] and [s] are usually deleted. The prefix di- is restricted to the third person actor in standard language; the alternative prefix (for any person) is zero. Some Indonesian examples are constructed by the author according to well-known models. The tense inflection is absent; in translations it is conditioned by context, either real or presupposed.
The verb -baca-kan ‘to read for/to someone’ allows two passive constructions, both with the same passive prefix di-:
A. The beneficiary is the prepositional object and the undergoer is the subject. The difference from (6a) is that the beneficiary, i.e. the one listening to what is being read, is compulsory.
(7a) Surat Yasin di-baca-kan untuk Pak Harto chapter PN PV-read-TR2 for Mr. PN
‘The Yasin chapter [of the Qur’an] has been read for Mr Harto’ [8].
The beneficiary may be omitted but is still presupposed.
(7b) Vonis di-baca-kan oleh hakim sentence PV-read-TR2 by judge ‘The sentence was read (aloud) by the judge’.
B. The beneficiary is the subject, as with anak ‘child’ in (8a), and the undergoer (here cerita ‘story’) is the object.
(7c) Anak kecil di-baca-kan cerita tentang sejarah dan budaya Tionghoa child little PV-read-TR2 story about history and culture Chinese ‘Stories about Chinese history and culture are being read to little children’
[9].
Both constructions have active counterparts.
(8a) ia telah mem-baca-kan surat itu untuk ibu-nya 3sg PRF AV-read-TR2 letter that to mother-3 ‘He has read this letter to his mother’ [10, h. 62].
(8b) Saya tiap malam mem-baca-kan anak saya cerita 1sg every night AV-read-TR2 child 1sg story
‘I read stories to my children every night before they go to sleep’ [11, p. 227].
Sentences like (8b) are more common as text-book examples but they are also found in genuine texts, cf. men-cari-kan ‘to seek, (to try) to find’ in (9a).
(9a) Aku sendiri akan men-cari-kan-mu rumah 1sg oneself FUT AV-find-TR2-2sg house ‘I shall find a house for you myself’ [12].
Let us compare a passive sentence with the same verb.
(9b) Uang [...] malah di-cari-kan sponsor oleh Manajemen timnas money even PV-find-TR2 sponsor by management national.team ‘The national team’s management even tries to find a sponsor for money’; here the beneficiary ‘money’ is non-human [13].
In Old Javanese, standards are looser and less elaborate than in Indonesian, but some data demonstrates the same active-passive conformity. Sentences (10-11) contain passive and active forms of the benefactive verb -pajaraken ‘to tell someone, to refer to someone’s words; to convey’ which are derived with the circumfix pa- -aken from the root ajar ‘information, communication’. The undergoer and subject in (10a) is wuwus ‘words, speech’. The benefi-
ciary (whose words are conveyed) -ku ‘my’ is the attribute of the subject. The form pajarakena is irrealis passive (without a special passive marker).
(10a) iki wuwus-ku ya teki pajarakenanta [pa-ajar-aken-a-nta] these words-1 PTL PTL.this BEN-information-BEN-IRR-2 ‘Convey these words of mine’ (lit.: let be conveyed by you) [14, p. 32].
In (10b), the undergoer and object is sawuwusira ‘all words of. The beneficiary (whose words are conveyed) Jaratkaru is the attribute of the object.
(10b) Kari ta sang naginl [...] majaraken [m-pa-ajar-aken] sa-wuwus-ira sang Jaratkaru remain PTL ART female.snake AV-BEN-tell-BEN all-word-GEN ART PN ‘The female snake remained [alone]; [she] told everything what Jaratkaru had said’ [15, p. 28].
The root kalung ‘necklace’ is a base for two transitive constructions: one with -an/-i (the first transitive that is common with locative and addressive meanings) and -aken (the second transitive, common with instrumental and causative meanings). OJv passive forms are built with -in-, which has two alternative positions: as an infix with an initial consonant, and as a prefix to bases with an initial vowel (the same positions as in Tagalog). So the passive of the base -kalung-i ‘to be attired with a necklace’ is k-in-alung-an; it has as its subject “the one whose neck is adorned with a kalung”. The pronoun ya ‘she’ is the subject in (11).
(11) s<in>inghel-an k<in>alung-an ya
<PV>singhel-TR1 <PV>necklace-TR1 3
‘She was attired with a singhel and a necklace’ (singhel ‘a piece of garment hanging down’) [14, p. 1778).
The active form is k-um-alung-i, and “the one to whom something is put around the neck” is the first object introduced by the preposition ri ‘at, in, on, to, with’ in (12a). The verb in (12a) is reduplicated, apparently with an expressive pejorative connotation.
(12a) Ika ng ratu Pariksit [...] k<um>alung-kalung-i ri bapa-ngku ri ng ula
that ART king PN <AV>necklace-PJR-TR1 on father-1 with ART snake ‘That king Pariksit has (impudently) put on a snake around my father’s neck’
[16, p. 93].
The second transitive verb from the same root is k-in-alung-aken in the passive and k-um-alung-aken in the active. The necklace (or something used as such) is subject with the former and first object with the latter. The active form in (12c) denotes a state and not an action.
(12b) hana ta wangke ni ng ula deles, ya ta [...]
there.be PTL corpse GEN ART snake deles it PTL k<in>alung-aken i gulu bhagawan Samlti <PV>necklace-TR2 on neck saint PN
‘There was the corpse of a deles snake, it was put [by him] around the neck of Saint Samiti’ [16, p. 92].
(12c) Hana ta sira mangke k<um>alung-aken ula ri gulu-nira be PTL 3 now <AV>necklace-TR2 snake on neck-3 ‘Now he is [still] having a snake around his neck’ [16, p. 93].
Such derivations are quite common; compare t-in-ambak-an ‘to be provided with a dam’ versus t-um-ambak-aken ‘to use as a dam’ and t-in-ambak-aken ‘to be used as a dam’ from tambak ‘dam, wall’ [14, p. 1916].
Other OJv passive derivatives are those with the prefix ka- and the circumfix ka- -an denoting uncontrolled results. These formations are usually labelled accidental passive. ka-derivatives are passive counterparts of transitives with the suffix -aken: compare s-um-or-aken ‘to defeat’ (from sor ‘lower, inferior’) and ka-sor ‘defeated’, ang-lumrah-aken ‘to spread’ — ka-lumrah ‘widely known, renowned’. Compare also sample sentences (13-14) with the verb bases -ton ‘to see’ and -pin-ton-aken ‘to show’.
(13a) larwa-larwan t<um>on dilah ni ng dipa flying.ant <AV>-see flame GEN ART lamp ‘Flying ants see the flame of the lamp’ [14, p. 2029].
(13b) t<in>-on-ira ta wangkay sang Samba <PV>-see-3 PTL corpse ART PN
‘He saw Samba’s corpse’ (lit.: was seen by him) [16, p. 116].
(13c) Ka-ton tawak [ta awak]-nira ngkane ng sawah de sang guru ACP-see PTL body-3 there.on ART field by ART teacher ‘He was noticed on the field by the teacher’ (lit.: his body) [16, p.75].
(14a) dateng aminton[aN-pinton]-aken rupa-nya come AV-show-TR2 appearance-3
‘[He] came showing his appearance’ [14, p. 2029].
(14b) p<in>inton-aken pwa gandewa-nira <PV>-show-TR2 PTL bow-3 ‘His bow was shown [by him]’ [Ibid].
(14c) ka-pinton-a kasatyan i nghulun ACP-show-IRR fidelity GEN 1 ‘Let my fidelity be shown’ [Ibid].
The accidental passive is also used with intransitives and nouns: lungha ‘go’ — ka-lungha-lungha ‘to go far, wander around’, sarira ‘body’ — ka-sarira ‘to embody, assume bodily shape’. The circumfix ka- -an corresponds to active verbs with the suffix -i: compare amepet-i [aN-pepet-i, from pepet ‘blocked, cut off’] ‘to block off’ and ka-pepet-an ‘blocked / cut off, covered completely’. However, some ka- -an forms have no corresponding -i-transitives, such as ka-paty-an ‘to lose/suffer a loss through someone’s death’ which is derived from the noun pati ‘(someone’s) death’, and ka-ton-an = ka-ton ‘to be seen/noticed, become visible’. Both ka- and ka- -an also derive from nouns, as in ka-pati ‘death’ and kasatyan [ka-satya-an] ‘fidelity’ (14c). The conformity of accidental passive to other constructions needs more research.
The active-passive conformity is uniform for all transitive verbs, either with root or affixal base. The choice of participant which is non-actor for the subject function has an impact on transitive markers but not on passive markers.
The voice alternation in Ml is inflectional; in OJv it is somewhere between inflection and derivation. In some cases, the application of a voice affix in itself is enough to derive OJv transitives from intransitive verbs or from nouns; the resulting forms have additional causative, instrumental or other meanings.
(15a) Sakweh sang watek dewata ma-hom ri pucak i ng Mahameruparwata
all ART group god INTR-held.meeting on top GEN ART PN.mountain ‘The whole company of gods held a meeting on the top of the M. mountain’ [16, p. 78].
(15b) Ika h<in>om-nira this <PV>-discuss-3
‘This was discussed by them (their aim was to get the nectar)’ [Ibid.].
Other examples: pes ‘weakness’ — in-apes ‘to be weakened’, ang-alih ‘to move to another place, seek a different place’ — in-alih ‘to be moved, transferred’, luput ‘to escape’ — l-um-uput ‘to let escape, free’, puket ‘drag-net’ — amuket [aN-puket], passive p-in-uket ‘to fish with a net’, talinga ‘ear’ — t-um-alinga, passive t-in-alinga ‘to listen attentively’, gajah ‘elephant’ — g-in-ajah ‘formed in the shape of an elephant’.
All Tagalog verbs have four inflectional forms of aspect, modality and tense expressed by alternation of partial reduplication, prefixal allomorphs with initial m- and n-, and the infixes/prefixes -um- and -in-; the latter two combine these aspectual-modal-temporal meanings with that of active resp. passive voice; see (1-3) above. The verb class is furthermore subdivided into several derivational subclasses connected to certain more or less specific semantic categories with different markers of the actor voice for each of them. So there are actor voice verbs with the prefixes mag-, maN-, um- (having a zero allomorph), magka-, maka-, maka-, maki-. For instance, mag-verbs express various meanings including causative, reflexive, intensive, denominal predicational; the prefiix maka- produces verbs of uncontrolled action, maki-denotes joint action [5, p.103; 7, p. 142].6 All of them pertain to irrealis mood and, except for -um-, are opposed to markers of the realis mood with initial n-, i.e. nag-, naN-, nagka- etc.; some of them may be partially reduplicated resulting in magkaka-, makiki-, etc. Rachkov has suggested a new interpretation of these actor voice prefixes.
“The prefixes mag / nag, mang / nang, um, magka / nagka, ma / na, maka / naka, maki / naki are traditionally treated as formal markers of the active voice, as if they combine the meanings of voice, aspect and partly those of modality and tense. But collating the active and passive forms and bearing in mind that voice and aspect have separate formal means in the passive, it may be argued that the formal expression of the active voice is nothing but the absence of passive affixes <...>”[5, p.128] (translated from Russian by the author) so that the “explicit” markers of the actor voice in fact convey only aspect, modality and tense. Such interpretation means, to my mind, that the active meaning is rendered by a zero prefix.
We can compare this complicated system with that of the OJv that is much simpler. Indeed, OJv also has more than one marker of the active, i.e. maN-, aN- and -um-. The prefixes maN- and aN-, as well as intransitive prefixes ma- and a-, display no semantic opposition and are merely varieties of the same prefix. However, historically they may be related to modality, since only aN- is allowed in certain imperative constructions.7 Thus, the OJv reflects the Philippine system in a formal way, although the original meaningful differences between maN-and aN- are no longer relevant.
6 Special subclass of potentive verbs meaning unintentional action and that of stative verbs are proposed by Himmelmann [6, p. 369-372]. As clearly demonstrated by Rachkov, verbs of different derivational subclasses, both active and passive, may have the same inflectional paradigms [5, p. 144 sqq].
7 This is an inference from Zoetmulder’s data: [17, p. 165].
In some OJv verbs, -um- is equivalent to (m)aN-. Both affixes derive synonymous forms from some bases: compare r-um-engo, mang-rengo from the root base rengo ‘hear’. However, some OJv verbs are opposed to one another, the ones with (m)aN- yielding transitive verbs, and those with -um- intransitive ones: ang-guyu ‘to laugh at’ — g-um-uyu ‘to laugh’.8
The position of -um-affix is analogous in Tagalog and OJv; in both languages, it is used for non-passive verbs, that is, for verbs that are active transitive and intransitive. This feature is historically table. From the structural point of view, the passive in both languages, as well as in Ml, is “more transitive” than the active.
Relation of active/non-passive and passive markers to transitivity
T ransitive Intransitive
Active/Non-passive: Tag.-um-, mag-, maka-, maka-, maki-OJv -um-, maN-/ aN-, Ml meN- + +
Passive: Tag -in, -an, i-, (-in-), OJv -in-, Ml di- + -
Note that in Ml, as distinct from Tagalog, transitivity is defined on morphological grounds of regular active-passive conformity, since the former generally does not allow passive formation from intransitive verbs. In Tagalog, such definition doesn’t work and transitive and intransitive verbs in the active differ only by the presence/absence of the object (a convention may include objects with or without a preposition). Once again, the OJv is in-between: it disposes of transitive markers as Ml and allows passives from non-transitives (verbs and nouns) as Tagalog.
Ml went further away from Tagalog, since its “inequivocally” transitive passive has reduced its scope, being specialised for the third person of the actor (see note 5). The passive for the first and second persons has zero prefix instead of the prefix di-. Thus, (4a) for the meaning “planted (by me/us/you)” (and the like) would have the form 0-tanami and (6a) for the meaning “read (by me/us/you)” (and the like) would have the form 0-baca. Such zero forms are preceded by designations of the actor: aku, or its proclitical form ku-T, or kita ‘we’ (inclusive) or any other denoting the first and second person:
(4c) Kebun itu aku/ku-/kita tanam-i jagung
‘That garden was planted with maize by me/us’.
(6c) Buku-buku itu aku/ku-/kita baca ‘These books were read by me/us’.
The zero-form baca formally does not differ from intransitive verbs whose meaning may approach those of the non-passive or the passive: bangun ‘wake up’, tidur ‘sleep’, tutup ‘be closed’, robek ‘be torn’ and many others.
As shown by the comparison of Tagalog and OJv verbs, their affixes are formally similar (or even identical) but they are functionally different. The prefix maN- in Tagalog is found mostly in intransitive verbs of habitual or professional activity [5, p. 116-118]. The infix -in- is also semantically different. As mentioned above, in Tagalog it is used in passive voice forms of
8 In New Javanese, aN- survived in archaic literary style, the most common variety being the consonantal prefix N-, while -um- became obsolete in some intransitive verbs.
the realis series. In Old Javanese, it became a pure passive marker without aspectual or modal meanings. In contrast to Tagalog maka-, OJv maka- is no potentive prefix; it is used in possessive-reflexive verbs, ‘to have/make for oneself something/someone as denoted by the base’: stri ‘wife’ > makastri ‘marry, take as one’s wife’.
Aspectual meanings in OJv are rendered by optional auxiliaries, as in isolating languages: huwus ‘already’, turung ‘not yet’ and the like; tense is expressed by adverbs or just by the context and modality, along with the future, is expressed by the suffixes -a and -en. These suffixes are usually labelled the irrealis and render a large spectrum of non-indicative meanings, including the subjunctive and the imperative. Besides, there are several means expressing the imperative proper [18; 4, p. 620].9
Therefore, there are significant differences between Tagalog and OJv verbal affixation which might suggest a considerable time depth in the split-up of their ancestral languages.
The historic transition from the system of several passive voices to the one with only one passive voice and transitive derivational markers (i.e. from “multi-passive” to “multitransitive”) was labelled “the break in the verb morphology” by Ulo Sirk [20, p. 335]. The evidence for intermediary stages in this transition may be found on the periphery of the Philippine area and beyond. In North Borneo, there exist languages representing the transitional stage from three to two passives or even one passive voice. Such languages, according to Beatrice Clayre, “occupy an area of transition between the languages of Sabah to the north which have a developed focus (or voice) system, and the languages of Sarawak and Kalimantan to the south which have a much reduced, or even lost, focus (or voice) system” [21, p. 17]. So, the verbal system of the Lun Dayeh language suggests that besides undergoer and instrument voices, “there was at least one other voice which may have been locative or benefactive” [Ibid., p. 18].10 Regrettably, at present the author has no data on how instrumental or benefactive meaning is expressed in a North Bornean language with only one passive, but it seems that it is not with the use of transitive suffixes, so that word order and prepositions are preferred. Compare an example from another study by Clayre on Berawan, a language of North-East Kalimantan: akkohi merrah2 ubi3 la’am ‘I1 scatter food3 for4 the chickens^. Indeed, such suffixes, as noted by Sirk, are nearly absent in Kalimantan [24]. However, they are present in South and Southeast Sulawesi languages, such as Buginese or Tukang Besi.
OJv -aken is cognate to Old Malay -akan, Malay -kan, and similar suffixes exist in other languages of western Indonesia; compare Sundanese -kon in Jawa, and Karo Batak -ken and Toba Batak -hon in Sumatra. In Buginese (South Sulawesi) and Balinese, such transitive suffixes have different forms. This indicates a parallel but separate development which may be explained by different histories of these languages.
It should be noted that similar suffixes exist in Oceanic languages. The mainstream migrations of Austronesians to Oceania proceeded along the northern coast of New Guinea, with a starting points to the south of the Philippines, somewhere between North Borneo, North Su-
9 Some hybrid forms are also attested. On some OJv auxiliaries see [19].
10 Lun Dayeh suffix -a? is used in the imperative. It is, according to Clayre, a vestige of the third passive, the locative or benefactive. The “projective” (non-realis, non-indicative) suffix -a is reconstructed for the Proto-Malayo-Polynesian language. It was lost in some Philippine languages but preserved in others, as suffix -o in Kimaragang of Sabah (North-East Borneo), as well as -a in Old and New Javanese and in Old Malay in the 7th century. The suffix -a with the imperative meaning also exists in Malagasy, the language of Madagaskar. Malagasy is a descendant of the early Maanyan language of South-East Kalimantan, more than a thousand years ago and not far from Java. See [22; 23].
lawesi, Halmahera and the Moluccan archipelago, around or after 2000 BC [20].11 This may indicate a region of the original lower-order proto-language (or rather a system of dialects) in the process of a shift to a “multi-transitive” system. Among its descendants is the prehistorical Javanese (pre-Old Javanese) along with other languages of western Indonesia.12
References
1. Himmelmann N. P. Typological characteristics — Adelaar A. and N. P. Himmelmann (eds). The Austronesian languages of Asia and Madagascar. London: Routledge 2005. P. 110-181.
2. Ogloblin A. K. Ocherk diahroniceskoj tipologii malajsko-javanskih jazykov [Aspects of diachronic typology of Malayo-Javanic languages]. Moscow: Novoe tysyacheletie, 1996. 192 p.
3. Idem. Grammatika indonezijskogo literaturnogo jazyka [A grammar of standard Indonesian]. St. Petersburg: St Petersburg Univ. Press, 2008. 440 p.
4. Idem. Javanese. — Adelaar A. and N. P. Himmelmann (eds). Op. cit. P. 590-624.
5. Rachkov G. E. Vvedenie v morfologiju sovremennogo tagal'skogo jazyka [Introduction to morphology of contemporary Tagalog]. Leningrad: Leningrad Univ. Press, 1981. 216 p.
6. Himmelmann H. P. Tagalog — Adelaar, Himmelmann (eds). Op. cit. P. 350-376.
7. Shkarban L. I. Grammaticeskij stroj tagal'skogo jazyka [Grammatical system of Tagalog]. Moscow: Vostochnaja literatura, 1995. 248 p.
8. Koran Indonesia. 12.01.08.
9. http://community.siutau.com; quoted 30.09.2009.
10. Kamus Besar Bahasa Indonesia. Jakarta: Balai Pustaka, 1989. xix, 1090 p.
11. Johns J., in collab. Robyn Stokes. Introduction to Indonesian language and culture. Book 1. Botany (Australia), 1996. 388 p.
12. Adams C. H. Bung Karno penjambung lidah rakjat Indonesia. Alih bah. Abd. Bar Salim. Djakarta: Gunung Agung, 1966. xxiii, 470 p.
13. http://forum-milanisti.com; quoted 30.09.2009.
14. Zoetmulder P. J. with the collaboration of S. O. Robson. Old Javanese-English dictionary. Vol. 1-2. ‘s-Gravenhage: Nijhoff, 1982. xxxi, 2368 p.
15. Juynboll H. H.(ed.) Adiparwa. Oudjavaansch prosageschrift. ’s-Gravenhage: Nijhoff, 1906. 214 p.
16. Zoetmulder P. J. Sekar Sumawur. Bunga rampai bahasa Djawa Kuno. I. Djakarta, 1958. 150 p.
17. Zoetmulder P. De taal van het Adiparwa. Een grammaticale studie van het Oudjavaans. Dordrecht, 1983. viii, 261 p.
18. Ogloblin A. K. Povelitel'nye predlozhenija v drevnejavanskom jazyke [Imperative sentences in Old Javanese] — Kasevich V. B. and Kaplun M. I. (eds). The languages of the Far East, South-East Asia and West Africa. Proceedings of the 7th International Conference. Moscow, Sept. 16-19, 2003. Pt. 2. Moscow, 2003. P. 21-30.
19. Idem. Taksis v drevnejavanskom jazyke [Taxis in Old Javanese] — Khrakovsky V. S. (ed.) Tipologija taksisnyh konstrukcij. Moscow: “Znak”, 2009. P. 624-650.
20. Sirk Yu. H. Avstronezijskie jazyki: Vvedenie v sravnitel'no-istoriceskoe izucenie [Austronesian languages: Introduction to comparative study]. Moscow: Vostochnaja literatura, 2008. 480 p.
11 Based on the way these suffixes are formed, they presumably originated from post-verbal prepositions [20, p. 336 sqq]; in more detail [25]. For a detailed analysis of transitivity in Indonesian, see [26].
12 My cordial thanks go to Sander Adelaar for his comments and editorial corrections to the first draft of this paper. All mishits and errors are mine, of course.
21. Clayre B. Kelabitic languages and the fate of ‘focus’: evidence from the Kerayan — I Wayan Arka and M. Ross (eds). The many faces of Austronesian voice systems: some new empirical studies. Canberra: Pacific Linguistics, 200З. P. 17-З7. On the following example see Clayre B. Verb affixes in Berawan // Ode C., Stokhof W., with assist. C. Baak (eds.). Proceedings of the Seventh International Conference on Austronesian Linguistics. Amsterdam etc.: Rodopi, 1997. P. 237.
22. Adelaar S. Les langues austronesiennes et la place du Malagasy dans leur ensemble — Archipel. 1989. T. 38. P. 2З-З2.
23. Idem. Asian roots of the Malagasy: a linguistic perspective — Bijdragen tot de taal-, land- en vol-kenkunde. 199З. D. 1З1, afl. 3. P. 32З-3З6.
24. Sirk Yu. K istoriceskoj grammatike avstronezijskih jazykov [On historical grammar of Austronesian languages] — Oriental studies I. Tartu University scholarly papers. Issue 201. Tartu, 1968. P. 38-З0.
2З. Sirk U. On the history of transitive verb suffixes in the languages of western Indonesia. — Stein-hauer H. (ed.) Papers in Austronesian linguistics No. 3 / Pacific linguistics. Canberra, 1996. P. 191-20З.
26. Alieva N. F. Indonezijskij glagol: Kategorija perehodnosti [Indonesian verb: The category of transitivity]. Moscow: Nauka, 197З. 142 p.
Abbreviations
ACP — accidental passive
ART — article
AV — active/actor voice
BEN — benefactive
COP — copula
DUR — durative
FUT — future
GEN — genitive
INTR — intransitive
IRR — irrealis
LIG — link, ligature
pl — plural
PJR — pejorative
PN — proper name
PRF — perfect
PTL — particle
PV — passive voice
REAL — realis
sg — singular
TR1 — 1st transitive
TR2 — 2nd transitive
1/2/3 — 1st / 2nd / 3rd person
<> used for infixes and their glosses
Статья поступила в редакцию 23 сентября 2010 г.
