Bahasa Siladang (Siladang Language), henceforth SL, is a language spoken by Siladang people living in Sipapaga and Aek Banir villages in Panyabungan Sub-district, Mandailing Natal Regency, North Sumatra Province. Discussion on SL syntax under the headings of typological study is oriented to the basic structure of clauses is aimed at seeking whether the SL accentuates the subject or highlights the topic in its syntactical typology. This section at a glance discusses the pragmatic functions to support and bridge the discussion that departs from a simple sentence. The study of language typology proposed by Comrie (1989) is a form of reaction to the theory of generative transformation which is based on the English behaviour. The theory of language typology is claimed as a neutral theory for various languages. Testing the theory of typology is important to know whether the theory can be applied to determine the typology of SL.
This study is based on the Greenberg's rule (1963: 76-77) which sees the effect of word order on the formation of ad-position types (prepositions or postpositions) and nominal phrases involving both adjective and genitive forms. In this case, he developed a universal word order theory that divides the world's languages into three types, namely S-V-O, S-O-V, and V-S-O. He also proposed a typology which is called Basic Order which concludes that there are six sentence patterns: SVO, SOV, VSO, VOS, OSV, and OVS. Initial research by Pawiro and Siwi (2015) shows that typologically, the word order of SL is S-V-O as shown in the examples (1)-(2).
(1) Iənə bisə mam-bantu dong ku
[3SG] be able [PREF-help to [1SG]
S V Prep O
`He/She can help me'.
(2) Bopok-ku ma-mavi dong-ku sabuoh buku
S V Prep-OTL OL
`My father gives me a book'.
The nature of subject correlates to the link between the agent and the topic. Agents are grammatical roles based on semantic roles, while topics are on pragmatic functions. The discussion on pragmatic functions is intended to find out whether the SL is classified as the subject prominent language or grouped as the topic prominent language. The study is based on the theoretical framework of language typology concerning the difference between the language of the subject with the language that accentuates the topic as suggested by Li and Thompson (1976: 457-489). The topics of sentence are limited terms proposed by Comrie (1989: 64). He argues that the topics of sentence are concerned with what the sentence is about, or what the sentence talks about (Artawa, 1998: 65). Blake (1994) sets the topic boundary as `what is spoken to' which becomes the opposition to the comment, for example, what is said about the topics. Blake further said that the topics are usually the given information and are specifically expressed as subjects. Topics can also be positioned outside the actual clauses which are marked by changes of intonation.
2. Literary Review
The clause structure is one of the discussions in the field of linguistics, especially in syntax. Lyons (1987: 170) argues that syntax is only one of grammatical issues and one of the areas in linguistic studies focusing on the study of how words combine to build larger units (phrases, clauses, and sentences). Syntax is rather easily understood as the science or the study of clauses and/or sentences. The clause structure is a syntactic unit formed from the basic constituents in the form of syntactic and complementary units. In addition, the clause structure is also a bundle of interconnected meanings contained in the clause to form a complete meaning in a clause. Like many languages in the world, the predicate of a structure can be filled by verbal or nonverbal elements. Based on the predicate functions, the basic clause consists of two types, namely the basic clause with verbal and nonverbal predicates.
Gundel (1988) argues that the topic-syntactic construction includes the type of simple sentences in the standard variety of language which highlight the topics. However, such constructions exist only as a marked choice of form in the languages that significantly accentuates the subject. Artawa (1998: 68) states that the opinions and studies carried out by Gundel with respect to the topic-comment constructions support what Li and Thompson (1976) have proposed. In line with the study of language typology by these two linguists, Bahasa Indonesia is classified as a language that accentuates the subject; however, they argue that Tagalog, as one of the Austronesian languages, is a language that does not accentuate the subjects or topics.
There is the tendency called dual subject construction which possesses special positions in the languages that highlight the topics. The construction of the whole sentence in this language group is usually indicated by one sentence having two adjacent NPs to the left of the predicates; one of the NPs carries the function of `topic' and the other brings the function of `subject'; thus, this condition creates the construction of double subjects. Consider the following examples in (3) and (4) from Li and Thompson (1976: 468).
(3) Sekana wa tai ga oisili (Japanese)
Fish TOP red biter SUB delicious
`The red fish (topic) biter is delicious'
(4) Néiki shu yézi da (Chinese)
The tree leaves big
`The liaves of the tree (topic) are big'
In the example in (3)-(4), both topics and subjects exist. Li and Thompson explains that in (3) the topic is marked by the particle wa and the subject by ga. In (4), the Néiki shu `the tree' becomes the topic and the NP yézi is the subject and in this sentence the topic can be removed with a pause. The topics in the example (3)-(4) above are positioned as the topics at beginning of the sentence. In this connection, the topic-comment constructions in the languages highlighting the topic are classified as the unmarked construction; however, the languages giving prime to the topic-comment constructions are determined as the marked construction.
The clausal topicalization is the derived clause theta which gets the effects of pragmatic functions in the grammatical level. The term `topicalization' is used here to refer to the syntactic construction (derivatives) in which the noun phrase (NP) in the basic construction (canonical) which takes its position in the post-verbal position (predicate) appears in the initial position before the subject or directly before the verb in the languages with verbal sequences. In this case, the subject appears in the object position. The sentence topic is related to the preposition expressed by the sentence. The definition of topic in terms of `contradiction' relation between the form and its preposition, in fact, is derived from traditional definition of `subject'. Topics must not necessarily be the grammatical subjects, and grammatical subjects do not necessarily have to be topics. The topics are also defined as foreground expression, or as an element that governs the framework of predictive distance, time, or features which are governed by main prediction. The topic refers to the sentence-relation which is pragmatically constructed by using pragmatic relation that must be understood as meanings that are contended by a particular context of discourse.
According to Artawa (1998: 68), traditionally the subject of a sentence is understood as an element that specifies what that phrase is. If this view is approved, it can be said that passive sentences should be understood as a matter of `patient' (not `agent') because passivity is a syntactic process that moves the patient into the subject and the agent become adjunct. In English, the subject is usually the initial argument. There is another construction that the initial argument is not the subject of a sentence. Such constructions are referred to as left-dislocation and topicalization.
So topicalization is a syntactic construction that place a constituent that normally follows a verb, moved forward preceding the subject NP. Topics are not necessarily all left-most sentence elements or the left-most NP on the outer structure, although this is probably the most common position. Not the entire initial argument is the subject. There is a syntactic construction that the original is not the subject of the sentence. See the following examples in (5) and (6) which are quoted from Artawa 1998: 68.
(5) Mary, she came yesterday.
(6) Mary, I know.
The construction in (5) is an example of a so-called left dislocation and the one in (6) is an example of topicalization. The difference between them is that in the left-dislocation construction there is a pronoun in the actual clause that refers to the initial noun phrase of the clause, whereas the topicalization construction does not have such phenomenon. In (5) the pronoun she is an anaphoric which refers to the NP Mary.
3. Research Method
This research is qualitative with descriptive approach and designed to make systematic, factual and accurate description about the SL characteristics. It uses typological method with comparative and inductive sub-methods; the nature of this study is empirical. Van Valin and La Polla (1999, 2002: 3) states that linguistic research is aimed at explaining linguistic phenomena. The natural data of this study were collected on the basis of existing facts or the phenomenon of language that was empirically used by SL speakers without considering right or wrong prescriptive grammar (Djajasudarma, 1993: 8). The data should be semantically and pragmatically grammatical and acceptable (Sudaryanto 1986: 62). Mithun (2001: 34-43) believes that the quality and quantity of data collection is highly dependent on the researchers and the time and skills of speakers.
This research focuses on the disclosure of basic clauses with tangible verbal language (logical meaningful words) that are naturally available and acceptable. Two of three kinds of data sources are used (Mallinson and Blake, 1981: 12-18). Four criteria of sentences, such as, well-formed and acceptable, and ill-formed and unacceptable sentences are used as the selection by the researcher (Haegemen and Gueron, 1999: 14-18). When a sentence is formed according to the rules of the internal grammar of the speaker's language, the sentence must be grammatical; when a sentence is not formed on the basis of grammatical rules, it must be ungrammatical. However, very often, native speakers use less well-structured sentences and having been cross-checked, the sentences are acceptable. The research instrument was the researcher himself who set the focus of the research and selected informants as the data source. Questionnaires were based on what Comrie (1983) proposed. Methods of data collection might include direct elicitation, recording, and checking elicitation (Mithun, 2001: 34-43). Techniques of data collection were the hearing and the speech. The agih method was used to analyze data and placed the parts of the language studied as a determinant of analysis (Djajasudarma, 1993: 60); (Sudaryanto, 1993: 31-100) and this method had seven advanced techniques (Sudaryanto, 1993: 36).
To know more about SL verbal clauses there are two parameters, namely semantic and syntactic parameters, which can be used as a guide in relation to verbs as a function of predicate clause filler. Both parameters are applied in an integrated manner so as to produce a classification of verbs or predicates that are not in a separate form between semantic and syntactic parameters. Referring to the semantic parameter, the verb is a category that refers to the meaning of the activity (action/action and process) and circumstances. Furthermore, verbs are examined from the semantic aspect of the number of participants or arguments involved in an event or circumstance. By contrast, syntactically, a verb is a core category that is present in the predicate clause position. In addition, verbs are viewed syntactically regarding the number of core arguments that must be present in a clause. Based on the syntactic distribution, the verb is the core of the verb phrase (VP) which is also at the same time the core constituent of a clause. Verbs also have a function to determine whether or not the presence of participants or arguments in a clause exists.
The verb has a syntactic feature as a predicate involving the presence of a number of arguments in its function to form a clause. Based on the number of arguments in a clause/phrase, verbs can be distinguished from intransitive, transitive, and ditransitive. The three types of verbs produce intransitive, transitive, and bipolar constructions. A more detailed discussion of the three types of verbal clauses can be seen in the followings.
Based on the main categories of function cluster predicate fillers, verbs can be classified into two, namely intransitive and transitive verbs. There are two parameters used to determine the classification of verbs (Budiarta, 2013: 97). Intransitive verbs or intransitive clauses in this study refer to the same terms because they are an integral as a whole. The use of the term intransitive verb or intransitive predicate has the same meaning. Both the verb and the intransitive predicate require a single core argument in a clause construct. Seen from its semantic nature, the intransitive verbs that populate the predicate of the intransitive clause can be distinguished from the experiential verb and the action verb. The verb is a verb that has a meaning of experience and a verb of action is a verb that has a meaning of action. The semantic role which becomes the only core argument in the intransitive clause and accepts the experiential verb is referred to the experiencer, while the only core argument in the predetermined clause of the action verb is the agent. If it refers to a grammatical function, the only argument in the intransitive clause occupies a function as subject (Artawa, 1998: 13). The intransitive clauses in SL can be seen in (7)-(11).
(7) Oku bapacu
(8) Iənə manari
(9) Ivɔng mɔlɔngkɔh ka sikola
3PL go to school
`they go to school'
(10) Si budɔk i jatoh
Child DEF fall
`A child falls on the ground'
(11) ɔkɔ tedə
The clauses in (7)-(11) are intransitive which are predicated by intransitive verbs. Some verbs populate the predicates of the intransitive clauses, for instance, the verbs bapacu in (7), manari (8), mɔlɔngkɔh (9), jatoh in (10), and tedə in (11). The core arguments are the oku, iənə, ivɔng, si budɔk and ɔkɔ.
Structurally, the verbs in (7)-(11) may occupy certain positions, for instance, following or preceding the core arguments. Thus, the core arguments are pre-verbal or precede the verbs (predicates). Examples in (7)-(11) show that typologically the constituent order of the intransitive clause is an SV (Subject+Verbal). The clause structure (cl-str) of the intransitive verb in (9) is represented in the tree diagram 1.
Unlike an intransitive clause whose predicate requires one core argument, the transitive clause is a predicate clause (verb) requiring two core arguments. Based on the data, the SL has a predetermined mono-transitive clause or its verb requires the presence of two core arguments. In macro perspective, these two core arguments can be called actors and undergoers (Van Valin, 1999). See the examples of mono-transitive clauses in (12)-(16):
(12) Kəmi manyuan lasinə
1PL plant chillies
`We are plant chillies'
(13) Iənə mangkael ikan
3SG to fish fish
`He/She is fishing fish'
(14) Pɔlɔdɔng mamacul sɔvɔh
Farmers cultivate paddy-field
`The farmers are cultivating their paddy-fields'
(15) Amai mamali boju ka poken
Mother buy shirt to market
`(My) mother bought (my) shirt in the market'
(16) Kholilah manggotil podi
Kholilah cut paddy
`Kholilah is cutting paddy-plants'
The clause examples in (12)-(16) above are predicated by transitive verbs which occupy the predicates of the transitive clauses, for instance, manyuan, mangkael, mamacul, mamali, and manggotil. The verbs require the presence of two core arguments, for examples, the clauses in (12) have two core arguments, namely, kəmi and lasinə, in (13) iənə and ikan, in (14) pɔlɔdɔng and sɔvɔh, in (15) amai and boju, and in (16) Kholilah and podi. The clause structure with the transcontinental mono-transitive is shown in the tree diagram 2.
Structurally, the verbs in (12)-(16) occupy their positions between two core arguments which functionally become the subjects and objects. They are pre-verbal and functionally considered the subjects of transitive clauses and the core arguments in the post-verbal positions functionally serve as the objects of transitive clauses. The core arguments, for instance, kəmi, iənə, pɔlɔdɔng, amai, and Kholilah become the subjects. Meanwhile, the core arguments, for example, lasinə, ikan, sɔvɔh, boju, and podi function as objects.
In addition to an Di-transitive clause whose verbs require the presence of two core arguments in a construction, the BS also has a clause construct whose variant presents three core arguments or is called a transitive clause. Dixon (2010: 116-117) does not use the term ditransitive clause for clause construction that has more than two core arguments. Dixon uses an extended transitive term for a bipolar clause. The construction of a bipolar transition BS that presents three core arguments is seen in the following clause example.
(17) Oku mamalikan iənə boju
1SG buy 3SG shirt
`I bought him a shirt'
(18) Iənə mambəvəkan oku manggeh
3SG bring 1SG mangosteen
`He/She brings me mangosteens'
(19) Jalmɔ i mambaikan iənə pitis
Person DEF give him money
`That person gave him/her money'
(20) Amai manungkihkan oku aə nyiə
Mother make me water coconut
`(My) mother made green coconut-water for me'
The examples in (17)-(20) show three core arguments, namely, oku, iənə, ans boju in (17), iənə, oku, and manggeh in (18), jalmɔ i, iənə, and pitis in (19) and amai, oku, and aə nyiə in (20). Specifically, clause structure with di-transitive verb is represented in the tree diagram 3.
Furthermore, the examples in (17)-(20) show that the SL has clause constructions with both more than two core arguments and morphological markers, such as, ma - + - kan which are attached to the verbs that serve to increase the presence of arguments in those clauses. The examples in (17)-(20) have alternation structures as shown in the examples (21)-(24) below.
(21) Oku mamali boju da iənə
1SG buy shirt Prep 3SG
`I bought shirt for him/her'
(22) Iənə mambəvə manggeh da oku
3SG bring mangosteen Prep 1SG
`He brought mangosteens for me'
(23) Jalmɔ i mambai pitis da iənə
Person DEF give money Prep him/her
`The person gave money to him/her'
(24) Amai manungkih aə nyiə da oku
Mother make water coconut Prep 1SG
`(My) mother made green coconut-water for me'
The examples in (21)-(24) are categorized as transitive clauses with predicates having two core arguments, for instance, oku and boju in (21), iənə and manggeh in (22), jalmɔ i and pitis in (23), and amai and aə nyiə in (24) and are in contrast to clauses in (17)-(20) in which their core arguments undergo changes in their grammatical functions from previously as objects, such as, iənə in (17), oku in (18), iənə in (19), and oku in (20) into obliques. These obliques are marked by the presence of the preposition da in clauses (21)-(24). The clauses in (17)-(20) are named the extended transitive because of having three core arguments while the examples in (21)-(24) are called the transitive clauses consisting of only two core arguments. The alternatives of clause structure from (17)-(20) to (21)-(22) are caused by the disappearance of the suffix marker –kan. Thus, it can be concluded that the marker is used to increase the number of core arguments from the verbs with two core arguments into three-core arguments. Pay attention to the tree diagram 4 which shows the clause structure alternatives from ditransitive into mono-transitive predicates.
In the review to find out and conclude whether SL can be classified as a language promoting the subject or a language accentuating the topic, there are three things to consider which were discussed by Artawa (1998), for instance, (i) left dislocation, (ii) fronting, and (iii) topicalization,
The examples of left-dislocation in SL can be seen in (7)-(9):
(25) Kepala kampung, iənə di surau
Head village, he in mushollah
`About the village head, he is in the musholah'
(26) Batang Gadis i, aəne janeh
Batang Gadis this-ART, water-POS3TG limpid
`About Batang Gadis river, its water is limpid'
(27) Uvong i, angginə ɔndɔh mamali oto
Person-ART brother-POS3TG already buy car
`In regard with that man, his younger brother has purchased a car'.
In clause (25) the noun phrase (NP) is dislocated to the left and is definite (in SL the name of person/greeting of honour/general noun need not be marked/zero marking). The NP is followed by a complete sentence whose subject refers to the NP which is transmigrated to the left as exemplified in pronoun iənə `he/she'. Furthermore, in clause (26), the construction of left-dislocation showing the possessor-possessee relationship is also prevalent in the SL. This means that such construction shows that there is possessor-possessee relationship between NPs which undergo left-dislocation in which the subject follows the NPs. In (26), the NP dislocated to the left also indicates the possessor-possessee relationship. The actual clauses in examples (25) and (26) have nonverbal predicates and the construction showing possessor-possessee relationship like this can also happen between NPs which are moved out to the left of their verbal predicate clauses as also shown in (27) which becomes transitive construction plus prefix ma- construction.
The core argument in SL is an unmarked element. Oblique, on the other hand, is marked by preposition and the relation of oblique can be moved out to the front position through the process of fronting. Here are two examples of fronting in (28)-(29).
(28) Amai tuhə mamitokkan galas da naik mejo
Mother old put glass Prep on table
`Grandmother put the glass on the table'.
(29) Da naik mejo amai tuhə mamitokkan galas
On table mother old put glass
`On the table grandmother put the glass'
In sentence (28), the noun phrase (NP) da naik mejo `on the table' has a preposition da and is termed as the `oblique'. The relation of oblique may be transformed to the fronting as shown in (29).
The following examples in (30)-(33) show the existence of core arguments being moved to the left. This transformation shows that the objects are topicalized; therefore, readers can learn about the topicalization in SL from the examples.
(30) Udin mamacah botu gɔdɔng
Udin Active-pecah stone big
`Udin broke big stone'
(31) Botu gɔdɔng Udin pacah
Stone big Udin break
`The big stone Udin broke'
(32) Sisi budɔk mamokol beduk
Active-pommel beduk (traditional; drum)
`The kids are pommeling the beduk'
(33) Beduk sisi budɔk pokol
Beduk kids pommel
`Beduk the kids are pommeling'
In (30) the NP botu gɔdɔng `big stone' is the core argument (object/patient) and so is the NP beduk in (32). These two core arguments become the initial elements in each of the clauses in (31) and (33). Thus, the object/patient in SL can be topicalization leads to the loss of the prefix marker ma- from the verbs.
The sentence `topic' is a type of sentence in which its pattern structure consists of (topic + comment). The `topic' element in this case includes the NP and the `comment' has a sentence consisting of subject and predicate. A `topic' sentence can be construction from a basic sentence by moving one of the NPs to the starting position, making it the topic of the new phrase. The initial position is filled with pronouns representing the NP; consider the following examples in (34) and (35) below.
(34) Dageng sibudɔk i kuvɔng togɔ
Body kid DEF less healthy
`The kid is not well'
(35) Sibudɔk i dagengnə kuvɔng togɔ
Kid DEF body-POS3TG less healthy
`The kid's health is not good'
Gundel (1988) believed that the rule of subject forming becomes the most common features in languages that highlight subjects as it is encountered in English. In such a type of language, the topic-comment' structure does not typically specify the syntactic structure. The tendency for phrases that refer to the `topic' in the initial position of a sentence occurs primarily by the close relationship between the subject and the topic. However, the rules of subject forming are more limited, if they also occur, in languages that powerfully feature the topics, such as Lisu and Mandarin languages.
It is concluded that the structure of clauses in SL consists of nonverbal and verbal predicates. Nonverbal structures may appear in adjectival, nominal, numerical predicate clauses, and in predicate clauses with prepositional phrases. The SL does not have copulative verbs and its verbal predicates of clause structure consist of intransitive, mono-transitive, and di-transitive (or sometimes called extended transitive). Semantically, there is only argument whose function is as the agent in the intransitive clause. If the verb that occupies the predicate position of the intransitive clause is a verb of situation (or non verbal predicate), the grammatical subject in the transitive clause generally becomes the agent; the grammatical object in the transitive clause is generally the patient. Subject as the core grammatical relation appears in the pre-verbal position of the canonical structure. An adverb can be inserted between pre-verbal and verbal arguments. In the canonical position, the pre-verbal argument becomes the subject so between the subject and its verb may have adverb. The object is the argument that directly follows the verbal or post-verbal argument in the canonical structure.
Based on the above description, the applicative construction of SL is categorized to tend to obey the rules of object forming. These rules appear more prevalently that the rules of subject creation that occur only in a small portion of the application; it also occurs in a number of verbs that have a certain semantic aspect. Referring to Artawa (1998) when a productive language has its rules of subject forming and of left-dislocation in its marked sentence structure, this language is classified as the one which accentuates the subject.
There are, in fact, several reasons why SL can be regarded as the language that accentuates the subject. First, the grammatical requirements of its basic structure are determined as the subject-predicate construction. The second reason is there is active-passive construction in SL with high productivity. Third, even though SL can topicalize the patient/objects, it is not basic construction but a derivative construction. The fourth reason is that the rules of object forming are not considered topicalization because the sentence element that makes the object is oblique-related element. Fifth, the topic-oriented structure of `double subject' in SL is not the basic sentence structure. Finally, there is the absence of morpho-syntactic markers for the elements to be considered as the topic. Topicalization becomes the marked sentence structure as well as the left-dislocation. However, the subject-predicate construction also becomes the unmarked sentence structure. With respect to pragmatic functions, the SL is typologically determined as the subject prominent language and as a result, its basic structure is oriented to the subject-predicate construction.