Vol.The 16 Indonesian No. 1 (2015): 133–166 PhilippeWacana Grangé, verbal sufix “–nya“ 133 The Indonesian verbal sufix –nya Nominalization or subordination? Philippe Grangé Abstract The sufix ‑nya is one of the most frequent and polysemic sufixes in Indonesian. It can provide deinite determination and topicalization. The “Verb‑nya“, which often appears in a topicalized subject Noun Phrase (NP), is generally labelled as a deverbal noun. Nevertheless, many syntactic constraints set it apart from Indonesian deverbal nouns. “Verb-nya“ must be complemented by a NP, which can easily be reconstructed as a former subject: a sentence is topicalized and thus becomes a noun clause, generally the subject of the main clause Verb Phrase (VP). I argue that “Verb-nya“ is a subordinate noun clause, almost always conveying causality. This causal noun clause, an innovation in formal written Indonesian (especially in the media), seems to ill a “gap“: the impossibility of beginning a sentence with a subordinating morpheme (‘that’, ‘because’). Keywords Indonesian, Bahasa Indonesia, sufix ‑nya, nominalization, causality, aspect, noun clause, topic. Introduction In this paper, I examine a pattern which is becoming increasingly frequent in written Indonesian: a verb is sufixed with ‑nya and seemingly promoted as the head of a noun phrase (NP). At irst sight, the compound ”Verb‑nya” is a deverbal noun, which often appears in a topicalized subject NP. This pattern also occurs, less frequently, within object noun phrases and prepositional phrases. However, the speciic syntactic constraints that restrict the use of ”Verb-nya” compared to nouns, lead us to question its status as a deverbal noun. Moreover, the tendency to use ”Verb-nya” to express a cause, and its complex aspectual features, set it apart from other Indonesian nominalization Philippe Grangé holds a PhD in Linguistics, Université de Poitiers, under the supervision of Professor Jean Chuquet and Professor Hein Steinhauer: “Temps et aspect en Indonésien” [Tense and aspect in Indonesian] (Summa Cum Laude), 2006. After having taught French as a Foreign Language and French Literature in Indonesia and in France, he is since 2006 associate professor at the University of La Rochelle, France. He teaches general linguistics and Indonesian. His research activities focus on syntax and semantics of Indonesian, and the description of Lamaholot and Bajo languages. Philippe Grangé may be reached at: [email protected]. © 2015 Faculty of Humanities, University of Indonesia DOI: 10.17510/wjhi.v16i1.370 134 Wacana Vol. 16 No. 1 (2015) devices. In the irst section, I will deine what exactly ”Verb‑nya” represents, among the wide range of possibilities for -nya, one of the most frequent and polysemic sufixes in Indonesian. The second and third sections respectively examine the -nya sufixation applied to intransitive and transitive verbs. Section 4 appraises the syntactic constraints of the ”Verb-nya” pattern. Finally, Section 5 clariies its aspectual features, and challenges the claim that ”Verb‑ nya” pattern is a nominalization. 1. An overview of the suffix -nya In Indonesian, the sufix ‑nya appears in a wide range of syntactic structures, making it one of the most familiar and puzzling morphemes in Indonesian. Surprisingly, Englebretson (2003) was, to my knowledge, the irst linguist to propose a broad analysis of -nya, extremely frequent in his colloquial Indonesian corpus. He dealt mainly with the “epistemic -nya constructions”, while the pattern “Verb-nya”, analysed as a nominalization device, is evoked only in a few lines (Englebretson 2003: 168). Yap (2011) focuses on the wide use of -nya expressing the speaker’s subjectivity, especially in Malay (which differs slightly from Indonesian). Arka (2011) shows that -nya has implications regarding modality, initeness, and TAM (Tense‑Aspect‑Modality); I will further discuss the aspectual dimension of this sufix. In this section, I will describe various uses of -nya. Most of them are related to the main point of this paper, the pattern ”Verb-nya”. 1.1 From 3rd person pronoun to definite determiner The sufix ‑nya mostly occurs as 3rd person possessive determiner or 3rd person pronoun, successively illustrated in the following example:1 (1) Mobil car -nya masih baru tetapi dia men-jual -nya -3sg.poss still but 3sg av- new sell- -3sg ‘His/Her car is still new but he/she sold it.’ The 3sg personal pronoun -nya may have an object function and play the role of a Patient, like the second -nya in example (1), be the head of a prepositional phrase (PP), example (2), or play the role of an Agent in Undergoer Voice,2 1 Glossing: appl: applicative; atl: atelic; av: actor voice; uv: undergoer voice; auv: accidental undergoer voice; caus: causative; dem: demonstrative; det: determiner; excl: exclusive; incl: inclusive; intm: intermediated; nom: nominalizer; perf: perfective; impf: imperfective; stat: stative; rel: relative; poss: possessive; prn: proper noun; sg: singular: pl: plural. For all abbreviations used in the text see also the list at the end of the paper. 2 We avoid the ”Active versus Passive” terminology, which is more appropriate to European languages. Indonesian has “symmetric voices“, as argued by Himmelmann (2005), because none of the voices may be considered as the transformation of the other one ; moreover Undergoer Voice is almost as frequent as Actor Voice. Actor Voice is marked by a verbal nasal preix, men- (av) hence ”Agent men‑V Patient”, while Undergoer Voice is marked by the preix di- (uv), hence ”Patient di-V (Agent)” or ter- (auv, accidental/non-voluntary). Another pattern, often said to be an alternative ”passive form“, bound to 1st and 2nd persons, actually results Philippe Grangé, The Indonesian verbal sufix “–nya“ 135 example (4). (2) Saya ber- terima kasih kepada -nya 1sg receive give to atl- -sg ‘I am grateful to him/her.’ Arka (2011: 81) also evokes ”verbs of saying and feeling”,3 to which the subject pronoun -nya adds a ”past reference”, and gives the following example: (3) ‘Siapa itu?’ tanya =nya. Who that ask =nya ‘‘‘Who is that?“ he asked./ #he will ask.’ (4) Rumah itu di- jual -nya (kepada Anwar) house det uv- sell -3sg (to prn) ‘He/She sold this house (to Anwar).’ Literally: ‘This house was sold by him/ her (to Anwar).’ The Undergoer Voice structure di-V-nya (uv-V-3sg), example (4), is typical of classical Malay narrative texts, where it conveys an aoristic aspect. It is far less in use in contemporary Indonesian novels, and can be heard only in formal speech.4 There is a clear-cut syntactic distinction: in the Undergoer Voice, the 3sg pronoun -nya refers to an Agent, while in the Actor Voice it refers to a Patient. This sufix may also appear as head of a PP. In other words, ‑nya is always oblique, while the personal pronoun ia 3sg is always subject, and dia 3sg may be used in any case.5 However, Yap (2011: 644‑646) signals a handful of puzzling examples which seem to contradict this rule, and indeed some paradoxical expressions where -nya seems to refer to 2sg rather than 3sg. In Indonesian, a noun complementation expressing a possession (or genitive) relation needs no linking morpheme: the ”possessed”, whether a noun or a pronoun, is simply placed to the left of the ”possessor”. However in colloquial Indonesian, sentences such as example (6) are quite common: from the fronting of an oblique argument (generally the object), which entails the absence of any verbal preix and the inseparability of subject and verb: ”Patient Agent Ø‑V”. 3 In direct speech, these verbs entail subject inversion (that is “quoted speech“ V S), as in French. 4 Englebretson (2003: 169) observes that “there are no examples of this structure in [his] corpus [of colloquial Indonesian conversation]”. 5 Another distinction between these 3sg personal pronouns is of a semantic nature: ia and dia represent humans, while -nya may refer to humans, animates or inanimates. Note also that -nya may refer to a plural person, hence 3pl, although the pronoun mereka 3pl is increasingly preferred in standard and colloquial Indonesian. 136 Wacana Vol. 16 No. 1 (2015) (5) Saya suka rumah Lina Saya suka rumah -nya 1sg house prn 1sg like house -3sg.gen like ‘I like Lina’s house.’ (6) ‘I like her house.’ Saya suka rumah -nya Lina 1sg house -nya prn like ‘I like Lina’s house.’ One could analyse ”-nya” in example (6) as the genitive 3rd person pronoun -nya, redundant (cross-referenced) with the noun referring to the possessor, that is, in rumah-nya Lina ‘Lina’s house’, -nya and Lina would be co-referential. We believe rather that this -nya is a deinite determiner;6 in Indonesian, it is perfectly possible to ind a possessive determiner and a deinite determiner (spec) within the same NP. The constraint is that -nya must not be separated from the speciied head noun or core NP (head noun plus adjective). (7) Rumah itu. Rumah -nya Rumah biru house det house -det/3poss house blue -det/3poss ‘This house’ (8) ‘This/his house’ -nya ‘This/his blue house’ Rumah -nya itu. *Rumah itu -nya *Rumah Lina -nya house -nya det house det -nya house -nya ‘This/his house’ ‘This house’ prn ‘house of Lina’ In sum, beside the canonical use of -nya as a broad representative of 3rd person (possessive determiner or oblique pronoun), this sufix may also behave as a deinite determiner. Englebretson (2003) labels this particular use of -nya “identiiability marker”7 and points out that it tends to mark referents which are identiiable not by the immediate speech context or by anaphora, but through a wider context, involving the shared cultural and pragmatic knowledge of the speakers. Although possessive and determinative meanings of -nya are often mingled, the subject NP in example (9) should be understood as ‘the restaurants in Indonesia’ rather than ‘Indonesia’s restaurants’. (9) Di Indonesia, restoran -nya selalu ramai. in Indonesia, restaurant -det always crowded ‘In Indonesia, the restaurants are always crowded.’ The deinite determiner ‑nya may be considered as a further grammaticalization 6 Also labelled ”ligature -nya”, see Arka (2011: 79). Englebretson (2003: 161) quotes Chafe (1994: 93): “To put it simply, an identiiable referent is one the speaker assumes the listener will be able to identify”. 7 Philippe Grangé, The Indonesian verbal sufix “–nya“ 137 of the possessive determiner, as suggested by Yap (2011: 647). It is particularly frequent in colloquial Indonesian; it can determine common nouns of course, but also pronouns and even proper nouns, as illustrated in examples (10) and (11). -nya baru. Saya -nya yang malas. restauran -det new 1sg rel lazy (10) Restoran ‘It is I who is lazy.’ ‘The restaurant is new.’ (11) Hendra npr -det di mana? Hendra -nya di mana? at which npr -det at which ‘Where is Hendra?’ ‘Where is this Hendra again?’ The sufix ‑nya also forms numerous discourse adverbs, labelled “epistemic -nya constructions” by Englebretson (2003: 170-185) or “evidentiality” by Arka (2011: 83). (12) Nyata -nya, be.real -nya, -nya sendiri. sains me- milik -i dinamika science av- posses -lock dynamics -3sg.poss alone ‘In point of fact, science possesses its own dynamic.’ Arka (2011: 82) also describes “modal nominalization”: (13) Harus must =nya kamu datang =nya 2sg come ‘You should have come’ Other usages of -nya include adverbialisation: (14) Sedikit -nya 8.000 perawat ber- demo di depan DPR. be.few -nya 8.000 nurse demonstrate at front prn atl- ‘At least 8.000 nurses are demonstrating in front of the House of Representatives.’ Englebretson (2003) proposes a ine grained analysis of these constructions, and Yap (2011) shows that “epistemic -nya constructions” and adverbialization, which occur in contexts such as expressing an opinion, making a judgment or giving advice, relect the subjectivity of the speaker, see examples (12) to (14). I will not elaborate on this particular semantic facet of -nya in this paper. 1.2 Identifying the ‘Verb-nya’ pattern The possessive determiner -nya is sufixed to a noun, thus cannot be confused with the ”Verb-nya” pattern. The problem arises when -nya is afixed to a verb: is it a 3rd person subject pronoun, or another kind of morpheme (that I will provisionally label ”deinite determiner”)? The 3sg pronoun -nya always plays 138 Wacana Vol. 16 No. 1 (2015) a semantic role, either Patient, example (1), or Agent,8 examples (4) and (15). (15) Alasan kami di- tolak -nya reason 1pl.poss uv- refuse -3sg ‘Our justiication was rejected by him.’ In the Undergoer Voice, the pronoun -nya is of course optional, because the Agent can be omitted or referred to by a noun phrase,9 as illustrated by example (16). (16) Alasan kami di- tolak. reason 1pl.excl uv- refuse ‘Our justiication was rejected. / / di- tolak (oleh) M.A. uv- refuse (by) prn rejected by the Supreme Court.’ On the other hand, the ”Verb-nya” pattern occurs in a non-verbal phrase which behaves like a noun phrase (NP), where -nya is a mere linking morpheme and plays no semantic role, example (17). (17) Alasan di- tolak -nya banding Hartono memang wajar. reason uv- refuse -det appeal actually lawful prn ‘The reason of the rejection of Hartono’s appeal is actually lawful.’ In (17), the morpheme -nya does not refer anaphorically to any noun or NP, while without -nya, the sentence would be ungrammatical. The striking syntactic divergence between the uses of di-tolak-nya in examples (15) and (17) is represented in the two following constituent-based trees (Figures 1 and 2): S NP N VP Det V UV PP V prep NP pro Alasan reason kami di1pl.poss uv- tolak refuse Ø ‑nya -3sg Figure 1. Representation of the constituent-based tree in Example 15. Furthermore, in a PP this personal pronoun can play a Beneiciary role, example (2), among other roles such as location. 9 Obviously, -nya is not a cross reference or agreement morpheme, which would be a hapax rule for Indonesian. 8 Philippe Grangé, The Indonesian verbal sufix “–nya“ 139 S NP Det VP NP Adv V N Ø alasan reason (di-tolak-nya memang banding Hartono) (uv-refuse-det actually appeal npr) wajar lawful Figure 2. Representation of the constituent-based tree in Example 17. Another salient difference between -nya 3sg pronoun and -nya ”deinite determiner“ is that the compound ”Verb-nya” must be complemented by a NP. Without this complementation, phrases with ”Verb-nya” would become ungrammatical, or lead back to the interpretation of -nya as a 3sg pronoun, see example (18). (18) Alasan di- tolak -nya memang wajar. reason uv- refuse -3sg actually lawful ‘The reason it was rejected by him is actually lawful.’ A complementation as in example (17) (that is, ‘the fact that Hartono’s appeal has been rejected’) indicates that ”Verb-nya” and its complement can be inserted as a whole into a NP, complementing its head noun. (That is alasan ‘reason’). At irst sight, this syntactic function of ”Verb‑nya” and its complement is a NP, if we consider ”Verb-nya” as a deverbal noun. An alternative analysis will be suggested in Section 5. As noticed by Kaswanti Purwo (2008), the ”Verb-nya” pattern, usually regarded as deverbal nominalization, is one of the salient innovations of present‑day standard Indonesian. I could not ind a single example of this pattern in classical texts, mostly literary, from the database Malay Concordance Project, Proudfoot (2008). The ”Verb-nya” pattern is increasingly frequent in standard written Indonesian, especially in the media, and more generally in any formal written text, including ‘serious’ blogs and administrative or legal texts. Malaysian (contemporary Malay spoken in Malaysia) seems less concerned by this innovation, but this needs to be veriied. While this trend seems to be affecting all kinds of formal and written texts in Indonesian, it remains unusual in conversations, which is not surprising: oral discourse avoids long and complex sentences. Most of my examples were gathered 140 Wacana Vol. 16 No. 1 (2015) from online Indonesian media between 2008 and 2014. I shall now examine examples of this pattern, applied to various verb categories, before analysing its syntactic features. 2. Intransitive verbs suffixed by -nya 2.1 Stative verbs This ”Verb-nya” pattern is restricted to gradable stative verbs,10 as pointed out by Steinhauer (2008). I found no examples of this pattern such as *betulnya, *berkeluarganya or *baharinya, formed respectively on the non-gradable stative verbs betul ’be exact’, berkeluarga ’be married’11 and bahari ’be maritime’. Stative verbs sufixed by ‑nya occur in Noun Phrases, either subject or object, and Prepositional Phrases. These syntactic contexts will be examined successively. 2.1.1 Subject Noun Phrase (19) Mahal -nya pendidikan di Indonesia be.expensive -nya education in Indonesia -kan dampak dari pasar shape -appl impact from market me- rupa av- bebas. free ‘The expensiveness of education in Indonesia is a consequence of the free market.’ (20) Tinggi -nya inlasi meny- (s)ebab -kan nilai be.high -nya inlation av- cause -appl value Rupiah men- (t)urun. Rupiah av- decrease ‘The height [high rate] of inlation causes (that) the value of Rupiah to decrease.’ Roughly 90% of the occurrences of mahalnya ‘the expensiveness’ and 80% of tingginya ‘the height’ gathered from online media contents appear in this type of structure, as head of a subject NP. This results from the topicalization of a stative verb, that is, mahal ‘be expensive’ in pendidikan mahal ‘education is expensive’ fronted as mahal-nya pendidikan ‘the expensiveness of education’, or 10 In the traditional Indonesian grammar terminology, stative verbs are labelled adjektiva. However, the string ber-V-nya may well occur, especially when the sufix ‑nya conveys a possessive meaning (3sg.poss or 3pl.poss) within a NP encompassing a nominal complementation. For instance tempat ‘place’, kumpul ‘gather’ > tempat berkumpulnya mahasiswa ‘the students’ meeting place’. The sufix ‑nya applies to the NP head and its complement as a whole: {tempat berkumpul}-nya mahasiswa, a structure similar to organisasinya mahasiswa ‘the students’ organisation’. Some ambiguous examples remain puzzling, such as teman ‘friend’ > bertemannya mereka ‘their friendship’, or, as pointed out by Kaswanti Purwo (2008: 3), berkumpulnya para pekerja ‘the gathering of workers’. 11 Philippe Grangé, The Indonesian verbal sufix “–nya“ 141 tinggi ‘be high’ in inlasi tinggi ‘inlation is high’ fronted as tinggi-nya inlasi ‘the height of inlation’. Subsequently, the noun which was subject in the original sentence, for example, pendidikan ‘education’ or inlasi ‘inlation’ is demoted and becomes the complement of the ”Verb-nya” compound. A few stems preixed by the atelic preix ber- behave like gradable stative verbs and may also be sufixed with ‑nya. (21) Ber- kurang -nya vitamin C akan meng- akibat atl- less -nya vitamin C will kulit men-(t)ua dengan cepat skin av-be.old with be.fast av- -kan consequence -appl ‘A reduction of vitamin C entails that the skin will get older fast.’ ‘The fact that vitamin C lessens entails that the skin will get older fast.’ 2.1.2 Object Noun Phrase Some stative verbs may occur as the head of an object NP, although this is not common. This means that fronting is an optional context for ”Verb-nya”, which must still be complemented by a NP. (22) Bapepam akan selidik -i prn will investigate -appl harga akuisisi BUMI Angga Aliya. price purchase prn -nya mahal be.expensive -nya ‘The Bapepam will investigate the expensiveness of BUMI Angga Aliya purchase price.’ (23) Gubernur Jakarta meny- (s)esal -kan kurang -nya Governor Jakarta av- less -nya pasok -an gas. supply -nom gas regret -appl ‘The Governor of Jakarta regrets the shortage of natural gas supply.’ 2.1.3 Prepositional Phrase Some stative verbs sufixed by ‑nya may also become the head of a prepositional phrase. They generally occur in argumentative speech, usually preceded by prepositions such as karena ‘because’ and/or verbs expressing causality. (24) Harga minyak me- lonjak karena sedikit -nya cadangan AS. price oil av- jump because be.few -nya stock prn ‘The price of oil jumps because of the smallness of US stocks.’ However, I found a few examples where this pattern is compatible with PPs localizing the event in time, for example, di tengah bersemi-nya N ‘during the 142 Wacana Vol. 16 No. 1 (2015) germination of N’ from semi ‘germinate‘. When the main clause verb is in Undergoer Voice, there are also examples of ”Verb-nya” as head of an agentive PP, which may be introduced by oleh ‘by’, example (25), or simply juxtaposed paratactically, example (26). (25) Semua itu di- per- parah oleh buruk -nya aturan. all det uv- caus- be.grave by be.bad -nya regulation ‘All this is aggravated by the fact that the regulation is bad.’ (26) Hal itu di- picu banyak -nya problem det uv- trigger be.a.lot -nya Jakarta yang ber- libur di prn rel atl- holiday at penduduk inhabitants Bandung. prn ‘This problem is triggered (by) the large numbers of people from Jakarta who spend holidays in Bandung.’ The “Verb-nya“ pattern is mainly linked to the expression of causal relations, and this specialization calls into question the status of “Verb‑nya“ as a mere nominalization device, as will be discussed later. 2.2 Intransitive dynamic verbs 2.2.1 Nominalization of bare intransitive verbs As Steinhauer (2001: 597) notes, if we consider the Indonesian dynamic verbs, transitive verbs may be afixed with the ”deinite determiner” ‑nya only if they are in the Undergoer Voice, while intransitive dynamic verbs (which cannot be in the uv) may be sufixed with ‑nya, either on their bare form (stem) or after having been preixed with meN-, as will be shown in 2.2.2. Dynamic intransitive verbs in their bare form, when sufixed by ‑nya, always express completed events. This is especially the case for verbs of motion: turun ‘to descend’ > turun-nya ‘the fall /decrease of’ bangun ‘to rise’ > bangun-nya ‘the rise of’ masuk ‘to enter’ > masuk-nya ‘the entry (the fact of coming in)’ jatuh ‘to fall’ > jatuh-nya ‘the fall, the drop of’ datang ‘to come’ > datang-nya ‘the arrival of’ (27) Masuk -nya organisasi asing di Aceh di- enter -nya organization foreign in uv- prn uv- harap -kan mem- beri -kan kejelasan […] hope -apll av- give -apll transparency ‘The entry/access of foreign organizations into Aceh will hopefully bring transparency […]’ Philippe Grangé, The Indonesian verbal sufix “–nya“ 143 Other intransitive dynamic verbs sufixed by the determinative ‑nya include: mati ‘to die, to be dead’ > mati-nya ‘the death (the fact that … is dead)’ lahir ‘to be born, to appear’ > lahir-nya ‘the birth (the fact that … was born)’ timbul ‘to appear, to arise’ > timbul-nya ‘the appearance, the emergence’ padam ‘to switch off, to extinguish’ > padam-nya ‘the extinction/putting out (of ire, lights)’ The ”Verb-nya” pattern frequently occurs as head of the Subject NP, but may also appear within an Object NP, for instance matinya ‘the fact that something dies’ in example (28). (28) Perubahan iklim meny- (s)ebab -kan change climate av- cause -appl die terumbu karang. coral reef mati -nya -nya ‘Climate change causes the death of coral reefs.’ As already underlined, this ”Intransitive Verb-nya” string must be complemented by a NP. 2.2.2 Nominalization of affixed intransitive verbs It may seem odd that intransitive dynamic verbs should generally be preixed with meN‑, labelled ”Actor Voice” preix, because intransitive verbs do not undergo voice alternation. However, in this context, meN- signals dynamism and autonomous change,12 rather than voice. Intransitive dynamic verbs preixed by meN‑ and sufixed by ‑nya are always atelic verbs, illustrating a progressive and uncompleted change or motion, for example, menurun ‘to decrease’ from turun ‘to descend’. Interestingly, for some intransitive dynamic verbs, two competing patterns of sufixation with the ”deinite determiner” -nya are found: turun ‘to descend’ > turun-nya ‘the descent, the decrease (completed)’ men-(t)urun ‘to decrease’ > men-(t)urun-nya ‘the decrease (still ongoing at the moment of reference)’ This important aspectual feature will be discussed in Section 5. Most intransitive dynamic verbs can be sufixed according to the pattern meN-Verb12 The change is no longer autonomous when the verb is sufixed by the causative/ factitive/benefactive -kan. For instance, men-(t)urun ‘to decrease’, men-(t)urun-kan ‘to put down, to make something lower’. 144 Wacana Vol. 16 No. 1 (2015) nya, and a few of them may also occur without the preix, that is, Ø‑Verb‑nya. lonjak ‘to bounce’ > *rosot13 me-lonjak ‘to jump’ > me-lonjak-nya ‘the sharp rise’ > *lonjak-nya > me-rosot ‘to decline’ > me-rosot-nya ‘the deterioration’ > *rosot-nya *13 These meN-Verb-nya compounds express change and progress, their process type is ”dense” (see note 19). The same Aktionsart (inherent aspect) is found in verbs whose stem is a stative verb, transformed into dynamic verbs through the preix meN-: ‘be big, tall’ besar tinggi ‘be tall, high’ panas ‘be hot, warm’ ‘be bad’ buruk ‘be wide’ luas ‘be strong’ kuat ‘be weak’ lemah ‘be old’ tua > membesar ‘to grow’ > membesarnya ‘the growth, the expansion’ > meninggi ‘to increase’ > meningginya ‘the (ongoing) increase’ > memanas ‘to warm up’ > memanasnya ‘the warming’ > memburuk ‘to worsen’ > memburuknya ‘the worsening’ > meluas ‘to widen’ > meluasnya ‘the extension, the widening’ > menguat ‘to strengthen’ > menguatnya ‘the strengthening’ > melemah ‘to weaken’ > melemahnya ‘the weakening’ > menua ‘to get/grow old’ > menuanya ‘the ageing’ A more accurate translation of the ”Verb-nya” patterns above would be periphrastic: ‘the fact that … is …-ing’, for instance meluasnya banjir ‘the fact that the lood is extending’. (29) Me- luas -nya banjir lumpur ber- dampak av- -nya looding mud atl- impact be.wide terhadap ekonomi Surabaya. toward economy prn ‘The extension of the mud looding is having an effect on Surabaya’s economy.’ 13 The stem rosot is never used without afixation. Philippe Grangé, The Indonesian verbal sufix “–nya“ 145 Intransitive dynamic verbs preixed by meN- which do not involve progressivity can also be afixed by ‑nya in its ”deinite determiner/topicalizer” meaning, as in example (30), but cannot be the head of a NP, so example (31) is not acceptable. (30) Huh, kelamaan me- lamun -nya, aku harus lekas mandi. huh too.long av- daydream -det, 1sg must quick bath ‘Huh, this daydreaming was too long, I must have a bath quickly.’ (31) *Me- lamun av- -nya daydream -nya sopir meng- akibat -kan kecelakaan itu. driver av- -appl accident det cause ‘The driver’s daydreaming caused this accident.’ Why is ”Verb-nya” barred14 in example (31), while the same surface structure is perfectly acceptable in example (29)? The ”Verb-nya” pattern, head of a NP, is linked to the expression of causality, as pointed out in Sections 2.1.3 and 2.2.1. Verbs that express motion, change, and progressivity in gaining or losing a property (dense type of process) are liable to convey cause or consequence: they trigger an effect, or are triggered under some factors. Conversely, compact processes (Activities in Vendler’s terminology),15 for instance melamun ‘to daydream’ are homogeneous (similarity between all points of the process); nothing is supposed to change. This may be the reason why they are not compatible with the ”Verb-nya” pattern, whose core feature is to convey causal information. 3. Transitive verbs suffixed by -nya In order for the ”Verb-nya” pattern to apply to transitive verbs, these must be in the undergoer voice, that is, passivized with di- (uv) or ter- (auv). In other words, the Actor Voice is not compatible with the ”Verb-nya” pattern. (32) *Pasar Taiwan men- (t)olak -nya cabai Indonesia market prn av- refuse -nya chilli prn mem- (p)ukul usaha ekspor cabai. av- hit business export chilli ‘(the fact that) The Taiwan market rejects Indonesian chilli hits the chilli export business hard.’ The reason for this impossibility probably lies in the fact that when a sentence begins with S men-V-nya, the listener will immediately interpret -nya as a 3rd 14 Example (31) must be rearranged, for instance Ada kecelakaan karena sopir melamun ‘There was an accident because the driver was daydreaming.’ A verb like melamun ‘to daydream’ may also play the syntactic role of a NP subject, as in Melamun dapat mengakibatkan kecelakaan ‘Daydreaming may cause accidents’. 15 See Vendler 1967. 146 Wacana Vol. 16 No. 1 (2015) person object pronoun ‘him/her/it‘, corresponding to the standard Actor Voice structure SVO, as in Saya menolaknya (1sg reject 3sg) ‘I reject it’. Moreover, the verb sufixed with ‑nya in the undergoer voice must be fronted. As a consequence, the NP referring to the Patient cannot appear in its usual preverbal position, example (33), and must be placed between the ”Verb-nya” and the optional Agent, as in example (34). (33) *Cabai Indonesia di- tolak -nya oleh pasar Taiwan by market prn chilli prn uv- refuse -nya mem- (p)ukul usaha ekspor cabai. av- hit business export chilli ‘(the fact that) Indonesian chilli is rejected by the Taiwan market hits the chilli export business hard.’ (34) Di- tolak -nya cabai Indonesia oleh pasar Taiwan uv- refuse -nya chilli prn by market prn mem- (p)ukul usaha ekspor cabai. av- hit business export chilli ‘(the fact that) Indonesian chilli is rejected by the Taiwan market hits the chilli export business hard.’ Literally: ‘Rejected in Indonesian chilli by Taiwan market hits hard the business of exporting chilli.’ As already noted for intransitive verbs, the ”Verb-nya” pattern applied to transitive verbs must be complemented, for instance by the NP cabai Indonesia ‘Indonesian chilli’ in example (34). A structure like S di-V-nya as in example (33) would lead to an interpretation of -nya as a 3rd person pronoun ‘(by) him/her/it‘, referring to the Agent, as illustrated by example (35): (35) Cabai Indonesia di- tolak -nya. chilli prn uv- refuse -3sg ‘Indonesian chilli is rejected by him/her.’ Alternatively, a verb may be in the ”accidental undergoer voice” (so-called ”accidental passive” or ”agentless passive”), signalled by the preix ter-, before it occurs in a ”Verb-nya” pattern. (36) Ter- dampar -nya paus di pantai Trisik auv- strand -nya whale at beach prn meng- heran -kan warga se- tempat. av- amaze -appl people same- place ‘The fact that a whale is aground at Trisik beach amazes local people.’ Literally: ‘Stranded is a whale at Trisik beach amazes local people.’ Philippe Grangé, The Indonesian verbal sufix “–nya“ 147 To sum up, the patterns di-Verb-nya (uv- Transitive Verb -nya), ter-Verbnya (auv- Transitive Verb -nya) and more generally all ”Verb-nya” patterns imply that the NP begins with this string, followed by a complement NP, and optionally by a PP. Most frequently, the ”Verb-nya” pattern occurs as head of the subject NP, as schematized below: [ Intransitive Verb1 -nya NP (PP) ] Verb2 … [ uv-Transitive Verb1 -nya NP (PP) ] Verb2 … However, an uv-Verb-nya pattern may also occur as the head of an object NP, example (37): S Verb1 [uv-Transitive Verb2 -nya NP (PP)] (37) Persyaratan mutu yang sangat ketat meny- (s)ebab -kan Norm quality rel very strict av- -apll di- tolak -nya cabai Indonesia. uv- refuse -nya chili Indonesia. cause ‘Quality norms that are very strict caused the rejection of the Indonesian chilli.’ Again, this pattern is linked to the expression of a causal relation. Moreover, within the main clause, causal verbs often occur, for example, menyebabkan ’to cause’ as in example (37). This causal feature can also be observed when ”uv-Verb-nya” occurs as head of a prepositional phrase (PP). This PP is often fronted, in accordance with the standard cause-consequence order: (38) Dengan di- tolak -nya banding oleh M.A., Martias With uv- refuse -nya appeal by prn prn tetap di- hukum 1,5 tahun penjara. still uv- condemn 1,5 year jail ‘Following the rejection of the appeal by the Supreme Court, Martias is still sentenced to 1.5 years in jail.’ Literally: ‘With is rejected appeal by the Supreme Court…’ Other prepositional phrases whose head is a ”uv-Verb-nya” involve the morphemes karena ’because’, dengan ’with’, maka ’therefore’, sehingga ’as a result’, and setelah or sesudah ’following, afterwards’, all within a sentence expressing a causal relation. 4. ”Verb-nya” pattern: a nominalization? The ”Verb-nya” pattern is usually considered as a deverbal nominalization. The problem is that this so-called deverbal noun remains highly predicative, and does not share some essential features with Indonesian common nouns. 148 Wacana Vol. 16 No. 1 (2015) Before looking at its syntactic properties, a thorough examination is needed of the sufix ‑nya in its role as determiner and topicalizer. 4.1 -nya as determiner/topicalizer In Section 1.1 the sufix ‑nya has been decribed as a very polysemic morpheme. Besides its uses as oblique pronoun or genitive linker, it may be described not only as an ”identiiability marker”, see Englebretson (2003: 161‑162), but more widely as a deinite determiner and/or a topicalizing morpheme. Genitive interpretation can be ruled out, when for instance an apposition occurs between the irst person saya and orang-nya ’the person’, example (39), where obviously -nya cannot be interpreted as a 3rd person pronoun or a possessive determiner cross-referencing a 1sg subject. (39) Saya orang -nya manis, ramah, baik 1sg person -nya sweet good heart and tidak sombong. neg proud polite hati dan ‘As a person I am sweet, polite, good-hearted, and not proud.’ This kind of topicalization may also concern proper nouns and personal pronouns, as in examples (10), (11), and (40). (40) Pacar -ku ingin me- nikah, saya friend -1sg.poss desire av- marry 1sg tidak mau. neg want -nya yang -nya rel ‘My boyfriend wishes to marry, it’s me who doesn’t want to.’ Moreover, the deinite determiner ‑nya plays a part in the ”information packaging” of the sentence: it indicates where the topic is located. Stative verbs sufixed with ‑nya are commonly fronted (that is, a classic means of topicalization), carrying emphatic or exclamative meaning in a nominal sentence. (41) Berat -nya kopor itu!16 heavy -nya suitcase det ‘How heavy this suitcase is!’ Literally: ‘The heaviness of this suitcase!’ (42) Mahal -nya expensive -nya solar sekarang! gas.oil now ‘How expensive is diesel oil now!’ Literally: ‘The expensiveness of gas oil now!’ Philippe Grangé, The Indonesian verbal sufix “–nya“ 149 Kaswanti Purwo (2008: 2) notes that “the deverbal nominalization with -nya (for example, naiknya) [‘the rise’] may be in competition with the earlier (equivalent?) noun formation (for example, kenaikan) [‘the increase’] but the two are not interchangeable”. In effect, the role of -nya as topicalizer in the ”Verb-nya” pattern can be opposed to various nominalization devices in Indonesian, for instance the conix ke- -an, which does not place any emphasis on the action depicted by the deverbal noun. For instance, the intransitive dynamic verb datang ‘to arrive’ can be nominalized as kedatangan ‘arrival, coming’. 16 (43) saya datang ke Holland 1sg arrive to → kenom- prn ‘I came to Holland’ (44) datang -an saya arrive -nom 1sg ke Holland to prn ‘my arrival in Holland’ saya datang ke Holland 1sg arrive to prn → datang -nya saya ke Holland arrive -nya 1sg to prn ‘I came to Holland’ ‘the fact that I came to Holland’ In example (44), the ”Verb-nya” pattern signals that the ‘arrival in Holland’ is the topic, which will be further commented on. It is foregrounded, usually standing for the cause of forthcoming events, or important circumstances for the following events. (45) Datang -nya saya ke Holland, justru mau arrive -nya 1sg to prn precisely want meng- hadir -i upacara pemberian av- present -loc ceremony offering penghargaan. award ‘The (reason why) I came to Holland was precisely to attend an award ceremony.’ On the other hand, the deverbal noun kedatangan ‘arrival, coming’ usually doesn’t prompt comments, and is likely to be followed by time and location information. Moreover, the ”Verb-nya” pattern conveys aspect, an issue I will return to in Section 5. 4.2 Syntactic constraints of ”Verb-nya” pattern The ”Verb-nya” pattern does not share some essential syntactic features of common nouns in Indonesian. I will examine a series of syntactic constraints, dealing with determination, negation, interrogation, relative clauses, adverbs of degree and aspect markers. 16 Example quoted from Steinhauer (2001: 597). 150 Wacana Vol. 16 No. 1 (2015) 4.2.1 Determination The Indonesian deinite determiners ini ‘this’ and itu ‘that’ are placed at the end (the right‑most position) of the NP. This deinite determination is optional for NPs headed by the ”Verb-nya” pattern. (46) Tinggi -nya angka golput high -nya number abstention (det) must buah pikiran. fruit thought (itu) harus men- jadi av- become ‘The (this) height [high level] of abstention must become food for thought.’ It seems that the ”Verb-nya” pattern itself never undergoes deinite determination, as may be the case for a deverbal noun, but this is hard to demonstrate, precisely because this pattern requires to be complemented (it is always the head of a NP). Numeral and indeinite determiners, which must be placed at the beginning of the NP, highlight the difference between ”Verb-nya” and a deverbal noun such as penolakan ‘refusal, rejection’, from the verb stem tolak ‘refuse, reject’. (47) Tiga/ Beberapa penolakan permintaan... Three/ some refusal demand ‘Three/Some refusals of the application’ (48) *Tiga/ *Beberapa di- tolak -nya permintaan... Three/ some uv- refuse -nya demand ‘*Three/*Some (fact that) the application is rejected’ ”Verb-nya” does not behave like a noun with respect to determination. This is possibly because the sufix ‑nya already acts as a deinite determiner, which blocks any further determination. 4.2.2 Negation Whenever a ”Verb-nya” pattern retains negation, it will be expressed by the predicative negation morpheme tidak, and not by the nominal negation morpheme bukan. For instance, the deverbal noun penolakan ‘refusal, rejection’, from the verb stem tolak ‘refuse, reject’, will be preceded by bukan, as opposed to ditolaknya ‘the fact that … is rejected’, which is negated by tidak. (49) Itu *tidak/ bukan penolakan permintaan kami. det neg/ neg refusal demand 1pl.excl.poss ‘This is not a rejection of our request.’ Philippe Grangé, The Indonesian verbal sufix “–nya“ (50) 151 Tidak/ *Bukan di- tolak -nya impor cabai di Taiwan neg/ neg uv- refuse -nya import chilli in prn men- jadi kabar gembira bagi petani Indonesia. av- become news happy for farmers prn ‘The fact that chilli importation to Taiwan is not rejected is great news for Indonesian farmers.’ This apparent paradox, and others to come, shows that this ”Verb-nya” pattern conserves basic predicate features. The predicative negation morpheme tidak also applies to this pattern when the verb is dynamic intransitive, example (51), and/or when it the head of an object NP, example (52). (51) Tidak masuk -nya anggota para KPU itu karena neg enter -nya member det prn because worry. det khawatir. ‘(The fact that) the Election Board members did not attend is because they were worried.’ (52) Masalah ini bisa ber- dampak tidak di- problem det can atl- impact neg uv- angkat -nya mereka pada rekruitmen PNS raise -nom 3pl at recruitment civil.servant ‘This problem may entail (that they will) not be appointed through the civil servant recruitment procedure.’ 4.2.3 Coordination Two ”Verb-nya” patterns can be coordinated by dan ‘and’ or atau ‘or’ as follows: (53) Yang mem- (p)utus -kan di- terima atau di- rel av- decide -appl uv- accept or uv- tolak -nya gugatan istri hanyalah hakim agama. refuse -nya complaint wife only judge religion ‘The only one who settles on whether the wife’s complaint is received or rejected is the religious judge.’ (54) langka -nya kayu mem- buat expensive and rare -nya wood av- make orang me- lirik bahan lain. people av- consider stuff other Mahal dan ‘The high price and the scarcity of wood makes people consider other materials.’ Literally: ‘Expensive and rare of the wood makes people consider …’ 152 Wacana Vol. 16 No. 1 (2015) Interestingly, the sufix ‑nya encompasses the coordinated verbs, for instance example (53) {diterima atau ditolak}-nya ‘that is received or rejected’ and example (54) {mahal dan langka}-nya ‘that is expensive and rare’. The strings diterima-nya atau ditolak-nya and mahal-nya dan langka-nya would be judged clumsy and unidiomatic, though not ungrammatical. In other words, it seems that instead of sufixing each one of the verbs, the serial construction is sufixed as a whole: {Verb and Verb}-nya, then complemented as a whole by a NP. Conversely, coordinated deverbal nouns must be afixed independently, on each stem, for instance penerimaan atau penolakan ‘acceptation or refusal’: peN-(t)erima-an atau peN-(t)olak-an. Again, the ”Verb-nya” pattern differs from Indonesian deverbal nouns. 4.2.4 Complementation As we have seen in example (18) and others, the ”Verb-nya” pattern must be complemented by a NP. This complement originates from a demoted subject NP, as illustrated in example (55). (55) Terumbu karang mati coral reef die → Mati die ‘coral reefs die’ -nya terumbu karang -nya coral reef ‘the death of coral reefs’ Literally: ‘the (fact that) die coral reefs’ The complement NP head referent is often inanimate, because ”Verb-nya” is generally used in argumentative discourse, dealing with social or scientiic issues for instance. As already mentioned above, causal relations are the favoured context for this pattern. However, the complement NP may be either a proper noun or a personal pronoun, with a human referent, for instance kami ‘we’ (1pl.excl) in example (56): (56) Masuk -nya enter -nya 1pl.excl to di- aku -i pesaing. uv- admit -appl competitor kami ke pasar ritel sudah market retail perf ‘The fact that we engaged in the retail market is acknowledged by our competitors.’ The ”Verb-nya” pattern can also occur as a non core argument in a NP, complementing a head noun. For instance, in example (57), datangnya tsunami ‘the surge of the tsunami’ complements the head noun trauma ‘trauma’. Philippe Grangé, The Indonesian verbal sufix “–nya“ (57) Trauma datang -nya tsunami tidak bisa begitu trauma come -nya tsunami neg can so saja hilang dari penduduk Aceh. simply disappear from inhabitant prn 153 ‘The trauma of the surge of the tsunami cannot fade away so easily for Aceh people.’ Even when occurring as a non-core argument, the ”Verb-nya” pattern must be complemented. In Indonesian, this constraint does not apply to nouns: they are optionally complemented by a NP. This again leads us to doubt that ”Verb-nya” is a nominalization pattern. 4.2.5 Equational structure This test is inspired by Arka (2011: 84). Adalah is a copula, whose usage is narrow and speciic, compared to the verb ”to be” and its equivalents in European languages. Adalah is bounded to equational structures and conveys deinitions: status or profession of a person, meaning of a word, comparison. Any isolated noun can be deined through adalah, example (58), but this is possible for ”Verb-nya” only when complemented, example (60). (58) Pen- (t)olak -an adalah per- buat -an nom- refuse -nom be nom- do nom- men- (t)olak. av- refuse ‘Refusal means the action of refusing.’ (59) *Di- tolak -nya adalah… uv- refuse -nya be... *‘The fact that is refused means…’ (60) Di- tolak -nya cabai Indonesia adalah proteksionisme. uv- refuse -nya chilli prn be protectionism ‘The fact that Indonesian chilli is rejected means protectionism.’ The whole NP headed by ”Verb-nya” may be deined or characterized via adalah, but not the string ”Verb-nya” alone, which behaves differently from common nouns. 4.2.6 Relative clause Another essential syntactic feature of nouns that the ”Verb-nya” pattern does not share is that it cannot be expanded by a relative clause introduced by yang, as illustrated by example (62). 154 (61) Wacana Vol. 16 No. 1 (2015) Pen- (t)olak -an yang di- umum -kan … nom- refuse -nom rel uv- announce -appl ‘The rejection which has been announced …’ (62) ? Di- tolak -nya yang di- umum -kan … uv- refuse -nya rel uv- announce -appl *‘That is rejected which has been announced …’ or ‘He refuses what has been announced …’ The second translation of example (62) is, however, correct if we interpret -nya as 3sg pronoun, hence uv-refuse-3sg ‘She/He refuses’, literally ‘refused by her/him’. As we have already pointed out, the ”Verb-nya” pattern must be complemented by a NP which originates from a subject NP. This NP complement function cannot be illed by a relative clause deprived of any antecedent. 4.2.7 Interrogation The ”Verb-nya” pattern is not compatible with an interrogative structure, as illustrated by example (64). (63) Kapan pen- (t)olak -an cabai mem- (p)ukul usaha? When nom- refuse -nom chilli av- hit business ‘When did the rejection of the chilli hit the business?’ (64) *Kapan di- tolak -nya cabai mem- (p)ukul usaha? Whe uv- refuse -nya chilli av- hit business ‘When did the fact that chilli is rejected hit the business?’ Only one fronting movement is possible in a sentence, hence we cannot have simultaneously a wh-movement (marked by an interrogative morpheme) and the fronting of an object NP (becoming the head of the subject NP). This is probably why the ”Verb-nya” pattern cannot follow interrogative morphemes such as kapan ‘when’, kenapa ‘why’, di mana ‘where’, untuk apa ‘what for’, etcetera. 4.2.8 Resilience of adverbs of degree and aspect-mood markers The ”Verb-nya” pattern can retain adverbs of degree. For instance, sentence (65), when fronted as in example (66), will conserve the adverb makin ‘increasingly’, to the left of ”Verb-nya”. (65) Harga kayu makin mahal. price wood increasingly expensive ‘The price of wood is more and more expensive.’ Philippe Grangé, The Indonesian verbal sufix “–nya“ (66) 155 -nya harga kayu mem- buat orang increasingly expensive -nya price wood av- Makin mahal me- lirik av- bahan lain consider stuff untuk struktur other for do people atap rumah. structure roof house ‘The increasing price of wood makes people consider other materials for the house roof frames.’ The adverbs that can be found in these structures are mainly adverbs of high-degree (for example, terlalu ’too much’). Another feature showing that ”Verb-nya” remains highly predicative lies in the fact that the predicate may retain some aspect or mood markers.17 The continuative aspect marker masih in example (67) remains to the left of ”Verb-nya” in example (68). (67) Harga suku cadang masih tinggi. price reserve still high piece ‘The price of spare parts is still high.’ (68) Masih tinggi -nya harga suku cadang men- jadi still high -nya price piece reserve av- become alasan tarif tak bisa turun. reason price neg can descend ‘[The fact that] the price of spare parts is still high becomes the reason why the cost cannot decrease.’ Obviously, a noun could not be complemented by a degree adverb or an aspect marker, unless this noun was used predicatively;18 another indication that ”Verb-nya” is closer to a verb than to a nominal. There is no room here to deal with other syntactic features of the ”Verb-nya” pattern. For instance, when formed on a transitive verb, therefore with the di- / ter- undergoer voice, ”Verb-nya” can retain the agent complementation introduced by oleh ’by’, see examples (34) and (38). Furthermore, the ”Verb-nya” pattern can be pronominalized, generally by itu ‘that, this’. To sum up, ”Verb-nya” is deinitely not a deverbal noun, because it remains highly predicative, and does not share essential features of Indonesian deverbal nouns. 5. Aspectual and syntactic features of ”Verb-nya” pattern It has gone almost unnoticed, except by Arka (2011), that the ”Verb-nya” pattern retain aspectual features of its verb, because, as said above, it remains highly predicative. These aspectual features are originate not only from the 17 We are grateful to Bambang Kaswanti Purwo for this remark. For instance Iwan masih guru (prn still teacher) ‘Iwan is still teacher’ or Kamu pelucu banget (2sg joker very) ‘You are really a joker’. 18 156 Wacana Vol. 16 No. 1 (2015) verb’s Aktionsart, but further develop through sophisticated rules, involving valency and voice. These complex aspectual features set ”Verb-nya” pattern apart from deverbal nouns afixed by peN- -an and ke- -an. Having ruled out the hypothesis of the ”Verb-nya” pattern as a nominalization device, I will inally offer an alternative syntactic analysis. 5.1 Verbal suffix -nya and aspect Arka (2011: 81) has highlighted that ”verbs of saying” sufixed by ‑nya 3sg convey, in the absence of a time adverb or aspect marker, a “past temporal reference”, as in example (3): tanya ‘ask’ > tanya-nya ‘he/she asked’, and not ‘He/she asks’. Moreover, even without any aspectual marker, the ”Verb-nya” pattern includes an aspectual meaning under certain valency conditions. As shown in examples (43) to (45), the intransitive dynamic verb datang ‘to arrive’ can be nominalized as kedatangan ‘arrival, coming’, while datangnya ‘the fact that (someone) came’ not only signals that the ‘arrival’ is the topic, but also implies that this process is completed. Let us compare ”Verb-nya” pattern with the ”competing” nominalization conix peN- -an: turun ‘to descend, to decrease’ > turun-nya ’the descent’ (something has decreased, and is seen as stable at the moment of reference) > men-(t)urun-nya ’the decrease’ (something is decreasing at the moment of reference) > pen-(t)urun-an ’the decrease’ (something has decreased, is decreasing or may decrease in the future) These deverbal nouns,19 or whatever they are, could be translated as ’the fall, the drop, the descent, the decrease’; their difference in meaning relects the aspect of the process they refer to, as I have tried to illustrate by a paraphrase. The distinction rests on the respective Aktionsarts (inherent aspect) of turun and menurun. Turun ‘descend, get down’ can be seen as an accomplishment (the descent or decrease is a discrete,20 completed event), while menurun ‘descend, get down’ is atelic, dense and not completed at the time of reference. In effect, 19 There are some other nominalization patterns on the stem turun ‘to descend, to decrease’, which are not relevant here, for instance turunan ‘way down’, keturunan ‘descendants, lineage, heredity’. 20 Following the Théorie des Opérations Enonciatives, as developed by Culioli (1999), a discrete type of process means that this process is undividable and telic, for instance to die, to win (it recalls Vendler’s Achievement Aktionsart). A dense type of process means it can be divided, and no inal boundary is intended, for instance to walk. Finally, compact processes correspond to States in Vendler’s terminology, but also to any atelic Activity, for instance he repairs chairs if it refers an usual occupation. Culioli’s categories also apply to nouns (countable/uncountable referents for instance). Philippe Grangé, The Indonesian verbal sufix “–nya“ 157 the ”Verb-nya” pattern does not automatically mark a perfective aspect; it interferes with the inherent aspect (Aktionsart) of the verb stem. Let us consider intransitive verbs. On a stative verb stem, the Ø‑ ‑nya pattern simply indicates a state as in tinggi ‘high’ > tinggi-nya ‘the height’.21 For intransitive dynamic verbs, the Ø‑ ‑nya pattern indicates a perfective aspect as in turun-nya ‘the decrease (completed)’, while the meN- -nya pattern generally speciies a progressive (imperfect, uncompleted) process, as in example (29) and (69). (69) -nya Men- (t)urun av- descend -nya neraca perdagangan. ekspor akan mem- per- buruk export will bad av- intm- balance commerce ‘The fact that exports are decreasing will worsen the balance of commerce.’ For transitive verbs, which must be in the Undergoer Voice before being sufixed by ‑nya, a perfective aspect is conveyed: (70) Di- tolak -nya cabai Indonesia mem- (p)ukul … uv- refuse -nya chilli Indonesia av- hit ‘The fact that Indonesian chilli has been rejected hits […]’ (71) Pen- (t)olak -an cabai Indonesia nom- refuse -nom chilli Indonesia mem- (p)ukul … av- hit ‘The rejection of Indonesian chilli hits […]’ The di- -nya afixation signals that the process is completed. On the other hand, with peN- -an, example (71), we have no information about the completion of the process; in other words it remains ”aspect neutral”. The same aspectual opposition appears between ter- -nya (ter-: accidental undergoer voice) and the nominalization conix peN- -an : (72) Ter- dampar -nya paus ... Pen- dampar -an paus auv- strand -nya whale nom- strand -nom whale ‘The fact that a whale has been stranded.’ ‘The stranding of a whale.’ Interestingly, the ”Verb-nya” pattern must be consistent with time reference and aspect markers. The perfective aspect indicated by ”uv-Verbnya” is incompatible with events situated in the future, or with imperfective aspect markers. In the following example, relating the assassination of a famous workers’ union leader during a light to Amsterdam in 2004, this 21 This ”state of things” is, however, topicalized, and will be commented on, or regarded as the cause of forthcoming events. 158 Wacana Vol. 16 No. 1 (2015) tragic event is anticipated before its completion, therefore the ”Verb-nya” pattern is inappropriate. (73) Pen- (t)awar -an untuk duduk di kelas bisnis nom- propose -nom for sit in class business men- jadi langkah pertama *di- bunuh -nya/ pem- av- become step irst uv- kill -nya/ nom- bunuh -an Munir yang akan di- laku -kan kill -nom prn rel will uv- commit -appl oleh Pollycarpus. by prn ‘The suggestion that he should sit in business class was the irst step towards the assassination of Munir that was to be perpetrated by Pollycarpus.’ There is a clear-cut aspectual opposition between ”Verb-nya” and peN- -an, for example, bunuh ‘kill‘ > dibunuhnya Munir ’the fact that Munir has been killed’ versus pembunuhan Munir ’the murder of Munir’, the last being indifferent as to whether the event is completed or not. This is why both are acceptable in examples (74) and (75). (74) Kita belum 1pl not. yet know tahu motif di- bunuh -nya Munir. motive uv- kill -nya prn ‘We still don’t know the reason why Munir was murdered.’ (75) Kita belum tahu 1pl not. yet know motif pem- bunuh -an Munir. motive nom- kill -nom prn ‘We still don’t know the motive for Munir’s murder.’ Testing the compatibility between aspect markers and the ”Verb-nya” pattern conirms the aspectual meaning of ‑nya. Aspect interpretation varies with two factors: the valency of the verb (transitive/intransitive) and the voice marking (Actor Voice meN-/Undergoer Voice di- or ter‑/no voice preix Ø‑). In what follows (see Table 1) I will test only three aspect markers: sudah (perfective), sedang (imperfective, ‘be V-ing’), and akan (uncertain/future). Philippe Grangé, The Indonesian verbal sufix “–nya“ 159 Verb valency and voice Compatibility of ”Verb-nya”22 with aspect markers Aspect conveyed by ”Verb-nya” Stative verbs23 Ø‑V‑nya tinggi ‘be high’ > tinggi-nya ‘the height, the fact that … is high’ sudah/*sedang/*akan tingginya STAT meN–V–nya besar ‘be big’ > mem-besar-nya ‘the growth, the increase (not completed)’ *sudah/*sedang/*akan membesarnya IMPF Intransitive dynamic verbs Ø‑V–nya turun ‘to go down’ > turun-nya ‘the decrease (completed)’ sudah/*sedang/*akan turunnya PERF meN–V–nya turun ‘to go down’ > men-(t)urun- IMPF nya ‘the decrease (current, not completed)’ sudah/sedang akan menurunnya Transitive verbs bangun ‘to put up, build’ > diPERF bangun-nya ‘the fact that … is built’ sudah /? sedang/*akan dibangunnya di–V–nya ter–V–nya 2223 pilih ‘to choose’ > ter-pilih-nya ‘the choice (completed)’ sudah/*sedang / *akan terpilihnya PERF Table 1. Three aspect markers: sudah, sedang, and akan. For intransitive verbs which do not require voice marking, the picture is contrasted. Stative verbs will remain stative within the ”Verb-nya” pattern, but the perfective marker sudah shifts to perfect aspect, more accurately the perfect of result.24 For instance, tingginya ‘the height, the fact that … is high’ > sudah tingginya ‘the fact that … became high’. Conversely, intransitive dynamic verbs will become perfective through the ”Verb-nya” pattern, for instance turun ‘to go down’ > turun-nya ‘the decrease (completed)’, and the addition of the perfective marker sudah does not modify the aspect.25 On a stative verb stem, meN- indicates an imperfective aspect, for instance besar ‘be big’ > mem-besar ‘to grow, increase, swell’. The ”Verb-nya” pattern, that is, mem-besar-nya ‘the growth, the fact that (something) increases’ does not modify this aspect, and no aspect marker can be added; even sedang (impf) 22 All the “Verb-nya“ displayed in this column must be complemented by a NP, a noun or a pronoun. 23 Stative verbs are of course intransitive. The stative verb tinggi should be translated as ‘be high’ rather than ‘high’. 24 Following Comrie (1976: 56-58) who gives the following deinition of the ”perfect of result”: “a present state is referred to as being the result of some past situation”. 25 The marker sudah is an aspect and mood marker. In this example, while the aspect is already perfective, sudah conveys mood: the event was expected or feared by the speaker. For a detailed account of sudah and other aspect-mood markers in Indonesian, see Grangé (2013). 160 Wacana Vol. 16 No. 1 (2015) is barred, perhaps because it would be redundant. On an intransitive dynamic verb, while meN- leads to an imperfective aspect interpretation, for instance turun ‘go down’ > men-(t)urun ‘decrease’, the ”Verb-nya” pattern will not alter this aspectual meaning, and remains surprisingly compatible with any aspect marker. For transitive verbs, the system is simpler. They must be in the Undergoer Voice (di- or accidental undergoer voice ter‑) and their sufix ‑nya is compatible with the aspect marker sudah. This is consistent with the fact that ”uv-Verbnya” refers to completed events, and conveys a perfective aspect. On the other hand, the Actor Voice preix meN‑ seems to neutralize the default aspectual meaning of -nya. Moreover, when a perfective aspect is signalled by -nya, this will also lead to the localization of the event in the past, if no other indication of time is available in the context. For this reason, example (76) is to be interpreted as ‘the coral died because of this condition’, and (77) as ‘the coral usually dies in this condition’. (76) Per- ubah -an iklim meny- (s)ebab -kan mati nom- change -nom climate av- cause -appl die -nya terumbu -terumbu karang. -nya coral -pl reef ‘Climate change entailed the death of the coral reefs / that the coral reefs died.’ (77) Per- ubah -an iklim meny- (s)ebab -kan ke- nom- change -nom climate av- cause -appl nom- mati -an terumbu -terumbu karang. die -nom coral -pl reef ‘Climate change entails the death of coral reefs.’ To sum up, the nominalization and ”Verb‑nya” patterns can be classiied according the aspect they convey by default (see Table 2). Verb stem Aspect Stative aspect Imperfective aspect stative verbs Ø‑V‑nya meN-V-nya intransitive dyn. verbs meN-V-nya transitive verbs (Undergoer Voice only) Perfective aspect Ø‑V‑nya di-V-nya ter-V-nya Table 2. The nominalization and Verb‑nya patterns. The conixes peN- -an or ke- -an are mere nominalization devices: the deverbal nouns they form behave syntactically as bare common nouns, and do Philippe Grangé, The Indonesian verbal sufix “–nya“ 161 not convey any particular aspectual information. The verbal sufix ‑nya is relevant to the expression of aspect, which differentiates it even more from nominalization. It is noticeable that under the ”Verb-nya” pattern, transitive verbs always indicate perfective aspect. This is probably because ”uv-Verb-nya” and its complement generally refer to the cause (see Sections 2.1.3 and 2.2.1), while the consequence is expressed by the following VP. Logically, the cause precedes the consequence, which implies the completion of a causal event before the consequence occurs. The aspectual features of the ”Verb-nya” pattern may be linked to two problems outlined above. Firstly, among the stative verbs, this pattern only applies to gradable stative verbs, as noticed by Steinhauer (2008), see Section 2.1. Secondly, intransitive dynamic verbs that do not express progressivity never occur within this pattern; for instance melamun ‘to daydream’ can be sufixed by ‑nya as a deinite determiner, but cannot be topicalized, as head of a NP or clause, see examples (30) and (31). In both cases, it appears that the ”Verb-nya” pattern cannot apply to homogenous processes (compact processes), but only to discrete and dense processes. A forthcoming event is triggered by ”Verb-nya”, because the state of affairs changes, for instance something becomes more and more expensive, or warmer and warmer (stative gradable verbs, intransitive verbs preixed by meN-), for example tingginya ‘the height, the fact that … is high’ and men-(t)urun-nya ’the decrease’. A progressive change in the state of things is depicted, as a prelude to a consequence. On the contrary, non-gradable stative verbs, for instance bahari ‘be maritime’, or verbs which do not involve any change between the beginning and the end of the process, for instance melamun ‘to daydream’, seem less apt to trigger a consequence. 5.2 “Verb-nya“ pattern: a subordinate clause As we have seen above, the ”Verb-nya” pattern has little to do with nominalization, and retains many predicative features. This pattern allows to topicalize a ”circumstance” that will be further commented on, or a cause followed by its consequence. It often relects the promotion of a causal PP to either a subject NP or an object NP. See for instance the transformation of sentence (78) into sentence (79). (78) Usaha ekspor cabai di- pukul karena pasar Taiwan business export chilli uv- hit market prn men- (t)olak cabai Indonesia. av- refuse chilli prn because ‘The chilli export business is hit hard because Taiwan’s market rejects Indonesian chilli.’ Extracted from the PP karena pasar Taiwan menolak cabai Indonesia ’because 162 Wacana Vol. 16 No. 1 (2015) Taiwan’s market rejects Indonesian chilli’, the string pasar Taiwan menolak cabai Indonesia is fronted and promoted as the new subject NP: (79) Di- tolak -nya cabai Indonesia oleh pasar Taiwan uv- refuse nya chilli prn by market prn mem- (p)ukul usaha ekspor cabai. av- hit business export ‘The rejection of Indonesian chilli by Taiwan’s market hits the chilli export business hard.’ Rather than a nominalization device, it would be more accurate to regard the ”Verb-nya” pattern as a subordinate clause (dependent clause). Syntactically, it behaves like a NP, but if we consider that the verb remains predicative (not a deverbal noun), it should be considered as a noun clause expressing cause. This ”noun causal clause” does not begin with a subordinating morpheme (like bahwa ‘that’). However, it would be ungrammatical to begin a sentence with a noun clause introduced by bahwa ‘that’: (80) *Bahwa Rupiah men- (t)urun (itu) me- rugi that prn av- det av- harm -appl ekonomi Indonesia economy prn go.down -kan *‘That Rupiah falls harms the Indonesian economy.’ It is also impossible to begin the sentence with a causal adverbial clause, introduced by karena ‘because’. But through the ”Verb-nya” pattern, it becomes easy to topicalize the ‘cause’, in the form of a causal noun clause. (81) Men- (t)urun -nya Rupiah me- rugi av- go.down -nya prn av- harm -appl ekonomi Indonesia. economy prn -kan ‘The fact that Rupiah falls harms the Indonesian economy.’ The compulsory complement NP of ”Verb-nya” originates from a former subject. Intransitive and transitive sentences will be transformed into subordinate noun clauses as follows (S stands for Subject, V for verb, O for object): → [V-nya S] SV SVO → [uv-V-nya S PP], where PP is the demotion of O We have seen that the sufix ‑nya has a wide scope, and as a deinite Philippe Grangé, The Indonesian verbal sufix “–nya“ 163 determiner, can topicalize any noun, NP or even verb. It does the same for a clause. Within the ”Verb-nya” pattern, the verbal sufix ‑nya may be considered as a subordinating morpheme, labelled ”Sub” in the constituent-based tree (here S stands for Sentence), illustrating example (81), see Figure 3. S SubP Sub VP NP VP Voice V N AV meNRupiah turun merugikan ekonomi Indonesia. -nya Figure 3. The consituent-based tree of example 81. The verbal sufix ‑nya, if we regard it as a subordinating morpheme, cannot be placed at the beginning of the noun clause. It ”moves” to the right of the verb, in a position where we could not ind its homonymous 3sg pronoun -nya ‘he/she’. Another noticeable movement is the subject inversion within the noun clause, that is Rupiah in example (81), which must move to the right of Verb-nya. Subject inversion is, however, frequent in Indonesian with verbs of saying, for instance jawab-nya ‘he answered’. Therefore, Rupiah in example (81) should still be regarded as the subject NP within the noun clause. Considering the sentence as a whole, a subordinate noun clause can occupy the same functions as a NP. This is the case with the ”Verb-nya” noun clause: although very often subject (and topic) of the sentence, it also occurs as an object or within a PP, and may be pronominalized. The ”Verb-nya” pattern is an innovation in Indonesian, which occurs mostly in written, formal Indonesian. Journalists, administrative oficers, and report writers are extremely fond of this pattern. Apparently, it ”ills a gap” in Indonesian syntax, which does not allow sentences beginning by subordinating morphemes like ‘that’ or ‘because’. 164 Wacana Vol. 16 No. 1 (2015) 6. Conclusion The ”Verb-nya” pattern is a new trend in Indonesian, increasingly frequent in formal written Indonesian today, especially in the media. It is traditionally labelled as a nominalization device, a claim that is challenged in this paper. Syntactic tests (determination, negation, coordination, complementation, relatives, etcetera) show that ”Verb-nya” behaves very differently from nouns; the Indonesian nominalization conixes ke- -an and peN- -an form deverbal nouns that are syntactically very close to plain, unafixed nouns. If ”Verb-nya” is not a nominalization, what is it? Using the general ability of -nya to topicalize a NP, a SV or SVO sentence can be topicalized, and become a subordinate noun clause. This -nya noun clause is very often fronted and stands for the subject, but other functions are possible: object or PP. The ”Verb-nya” pattern, or -nya noun clause, has two important characteristics: causal and aspectual features. The information provided in this noun clause will either be commented on, or trigger a consequence, especially when this clause is fronted. But even within a PP, the ”Verb-nya” pattern still expresses cause; it can be labelled a causal noun clause. Aspect is also conveyed, under speciic syntactic constraints, and ‑nya interacts with the Aktionsart (inherent aspect) of the verb. Depending on the valency of the verb, on its voice preix (meN- actor voice, di- or terundergoer voice, or none of them), and excluding homogenous (non-gradable or non-progressive) processes, ”Verb-nya” can signal stative, perfective or imperfective aspect. This causal noun clause seems to ill a ”gap” in Indonesian syntax: the impossibility of beginning a sentence with a subordinating morpheme (bahwa ‘that’, karena ‘because’). Abbreviations used appl : applicative atl : atelic auv : accidental undergoer voice av : actor voice caus : causative dem : demonstrative det : determiner excl : exclusive gen : genitive impf : imperfective incl : inclusive intm : intermediated nom : nominalizer np : noun phrase Philippe Grangé, The Indonesian verbal sufix “–nya“ o : object perf : perfective pl : plural poss : possessive pp : prepositional phrase prn : proper noun rel : relative s : subject sg : singular stat : stative tam : Tense-Aspect-Modality uv : undergoer voice v : verb vp : verb phrase 165 References Arka, I Wayan. 2011. “On modality and initeness in Indonesian; Complexities of =nya nominalization”. [Paper, Workshop on TAM markers and evidentiality in Indonesian Languages, Tokyo University of Foreign Studies, 17‑18 February; Http://lingdy.aacore.jp/en/contact/index.html.] Chafe, Wallace L. 1994. Discourse, consciousness, and time; The flow and displacement of conscious experience in speaking and writing. Chicago: University of Chicago Press. Comri, Bernard. 1976. Aspect; An introduction to the study of verbal aspect and related problems. Cambridge: Cambridge University Press. Culioli, Antoine. 1999. Pour une linguistique de l’énonciation; Formalisation et opérations de repérage (Tome 2). Gap: Ophrys. Englebretson, Robert. 2003. Searching for structure - the problem of complementation in colloquial Indonesian conversation. Amsterdam: John Benjamins. Grangé, Philippe. 2013. “Aspect in Indonesian; Free versus bound markers”, NUSA, Linguistic Studies of Indonesian and Other Languages in Indonesia 55: 57-79. Himmelmann, Nikolaus P. 2005. “The Austronesian languages of Asia and Madagascar; Typological characteristics”, in: K. Alexander Adelaar and Nikolaus P. Himmelmann (eds), The Austronesian languages of Asia and Madagascar, pp. 110-181. New York: Routledge. Kaswanti Purwo, Bambang. 2008. “Contemporary Indonesian syntax; Some evidence of innovations and language change”. [Paper, Twelfth International Symposium on Malay/Indonesian Linguistics (ISMIL 12), Leiden, 25-27 June.] Proudfoot, Ian. 2008. “Malay Concordance Project”. [Http://mcp.anu.edu. au/, consulté le /; accessed on 21‑02‑2013.] Steinhauer, Hein. 2001. Leerboek Indonesisch. Leiden: KITLV Press. 166 Wacana Vol. 16 No. 1 (2015) Steinhauer, Hein. 2008. “Does Indonesian have adjectives?”. [Paper, Twelfth International Symposium on Malay/Indonesian Linguistics (ISMIL 12), Leiden, 25-27 June.] Vendler, Zeno. 1967. Linguistics in philosophy. Ithaca: Cornell University Press. Yap, Foong Ha. 2011. “Referential and non‑referential uses of nominalization constructions in Malay”, in: F.H. Yap, Karen Grunow-Hårsta, and Janick Wrona (eds), Nominalization in Asian Languages; Diachronic and typological perspectives, pp. xvii, 627-658. Amsterdam: John Benjamins.