Negative Markers in Lamaholot , Eastern Flores – Indonesia : The Case of Kiwangona Dialect

Marcelinus Yeri Fernandez Akoli ✉ Johnson Welem Haan and Melania Tuto Keredok Lecturer, English Education Department, Faculty of Teacher Training and Educational Sciences, Universitas Nusa Cendana, Kupang, Indonesia Student, English Education Department, Faculty of Teacher Training and Educational Sciences, Universitas Nusa Cendana, Kupang, Indonesia Corresponding Author: Marcelinus Yeri Fernandez Akoli, E-mail: marcelinus.akoli@staf.undana.ac.id


Background of the Study
In human communication, saying 'no' is a common thing. We can easily observe how people say 'no' to others in order to show their disagreement, contradiction or denial of something. By grasping this idea, it is unsurprising that almost every language in the world possesses a 'no' expression. In linguistics, any expression of saying 'no' is called 'negation'. Negation, moreover, can be defined as a construction including either grammatical or semantic analysis showing typical expression marking either partial or total contradiction of a sentence's meaning (Crystal,2008). Morante and Sporleder (2012) are more syntactically-oriented when they consider negation as a grammatically-categorized process in which the truth value of a proposition is changed. In English negative construction, a number of negative markers like no, not, never, nothing and none are occupied with expressing contradiction. The usage of these negative markers can be exemplified as follows.
1. Sam showed no mercy to his opponent. 2. She did not study Biology last night. 3. Hector never plays Chess all his life. 4. They earned nothing from the job except the food. 5. None gave him permission to leave the class.
This article discusses negative markers in Lamaholot (ISO 639-3:slp), one of the Austronesian languages spoken in the eastern part of Indonesia, particularly on eastern parts of Flores island, Adonara island, Solor Island, Lembata island (except Kedang area), and some areas on Pantar island as well as surrounding islands. Typologically, Lamaholot belongs to the Flores-Lembata subgroup, under the branch of the Malayo-Polynesian group (Adelaar,2005). Besides Lamaholot, some surrounding languages such as Sikka and Kedang also belong to this subgroup. Administratively, it covers East Flores and Lembata regencies of East Nusa Tenggara province. This language is spoken by 150,000 to 200,000 speakers. As been claimed by Keraf (1978), there are 33 different dialects of Lamaholot. These 33 dialects are classified into three groups, that is, Western Lamaholot (23 dialects), Central Lamaholot (8 dialects) and Eastern Lamaholot (2 dialects). However, some other linguists recently believe that Lamaholot is best understood as a LINGUA FRANCA functioning to link a number of closely related and poorly documented languages, each with internal variation (Grimes et al., 1997). Under Keraf's dialectal classification, the data in this article were taken from the Kiwangona dialect, which belongs to the Western Lamaholot group. In the eastern borders of Lamaholot speaking areas further than Kedang areas covering Pantar and Alor islands, there live non-Austronesian languages of Timor-Alor-Pantar (TAP) subgroups. These speaking areas should be mentioned here as they play significant effects on the appearance of Lamaholot today, particularly in the issue of its negative constructions. Picture 1. Map of Lamaholot dialects in Solor archipelago (taken from Keraf, 1978)

Morphological and Syntactic Settings of Lamaholot Language
Morphologically, most Austronesian languages can be labelled as agglutinative-synthetic languages (Blust,2013). They do not possess many prefixes attached to their verbal and nominal phrases. In the case of Flores languages, Blust adds that there is a morphologically-extreme reduction undergone by Manggarai, Ngadha and Lio, showing these languages are more synthetic than other neighbouring languages in the Central-Eastern Malayo Polynesian sub-group. Similarly, Lamaholot also exhibits a few morphological phenomena (Kroon,2016). Regarding negative markers, it should be acknowledged that many languages in Flores and Sumba, such as Manggarai (Akoli,2021) and Kambera (Klamer,1998), mainly possess negators in pre-predicate position and double negation.
Syntactically, however, the morphosyntactic patterns found in Lamaholot negative expressions show unique configurations as they fill all three stages of Jaspersen's cycle, that is, negator in pre-predicate position, negator in clause-final position and double negation (Fricke,2017). Fricke claims that the existence of clause-final negator in Lamaholot is due to its contact with adjacent non-Austronesian languages. All of these considerations give us reasons to describe negative markers in Lamaholot not in the same way as Fricke (2017) did before but on a dialectal scale. This article seeks to discuss the negative markers in the Kiwangona dialect of the Lamaholot in terms of their forms, distributions and meanings.  Sulistiyono (2015) and a Grammar of Solor-Lamaholot: a Language of Flores -Eastern Indonesia by Kroon (2016). Beyond the most current pieces of works above, there are a number of either classic or influential Lamaholot works that cannot be denied, such as Morfologi Dialek Lamalera by Keraf (1978), Koda Kiwang: Dreisprachiges Worterbuch der Lamaholot Sprache, Dialekt von Lewolema by Pampus (2008) and A Grammar of Lamaholot, Eastern Indonesia: The Morphology and Syntax of the Lewoingu Dialect by Nishiyama and Kelen (2007). However, it should also be acknowledged that not all dialects and linguistic aspects of the language have been explored and reported due to its complexity. Particularly, this is also true for the issue of negative construction in Lamaholot.
It is possible that the most contextual piece of writing on the negative markers in Lamaholot can be found in Fricke's article entitled 'The Rise of Clause-Final Negation in Flores-Lembata, Eastern Indonesia' (2017). Fricke started her article with a proposal saying that the existence of clause-final negative construction in a number of observed Flores-Lembata languages possibly results from a historical contact between those languages with an unidentified non-Austronesian language located around Alor-Pantar islands. Using Jespersen Cycle, Fricke compared nine Flores -Lembata languages: Sikka, Hewa, Lewotobi, Lewoingu, Solor, Lamalera, Central Lembata, Kedang and Alorese, to map their negative structures inside the languages' clauses. As a result, Fricke found that the nine observed languages fill all three stages of a Jespersen cycle. In particular, Kedang and Sikka exhibit Pre-predicate negative contruction; Hewa, Lamalera and Central Lembata demonstrate double negative structure; Lewoingu, Lewotobi, Solor and Alorese show clause-final negative formation. Hewa In her investigation for describing the etymology of Flores-Lembata negators, Fricke found that there is a historical link between some of the languages in terms of Pre-Predicate negators, while for the clause-final negators, she theorized that they might have resulted from a grammaticalization of inherited words as well as spontaneous innovation. Moreover, the phenomenon of clausefinal negators in the Flores-Lembata subgroup can eventually be understood as 'structural borrowing' that was estimated to occur in a situation of language shift between the Austronesian language and an unknown Papuan language.

Data and Data Collection Technique
This research employs a qualitative method. The data were collected through a two-week field trip to Bayunta'a village of Ile Boleng subdistrict on Adonara island. This area once again speaks the Kiwangona dialect of Lamaholot. There were 5 language informants involved in this research. The data were obtained through elicitation processes and note-taking activities. During the process of elicitation, the data were recorded by using SONY IC Recorder ICD-UX200F. The recorded data were then transcribed and, together with the notes, gave contributions to the data analysis.

Data Analysis Technique
The data were analyzed by using Negation Patterns theorized by Mosel and Spriggs (1999). In the linguists' perspective, negative markers can typologically be grouped into five types, i.e., (1). yes/no questions ; (2). Negative existential such as none or nothing ; (3). Negation of verbal constituents ; (4). Negation of non-verbal constituents ; (5). Negative imperative. Finally, this article also gives comments on the existence of Lamaholot negators based on the Jespersen cycle (1917) in order to see whether this result agrees with Fricke's findings or not. The Lamaholot negative markers are presented and described in the next section based on the two ways of analysis above.

Results
Based on our investigation on Mosel and Sprigg's types of negative markers found in the Kiwangona dialect, there are eight negative markers identified and can be visualized as follows : The common negator in the Kiwangona dialect is 'take' that can stand alone. Moreover, together with 'waŋa' and 'amu', 'take' can also behave as a negative existential form meaning 'nothing'. To mark negations in both predicate and non-predicate formations, the speakers use 'hala' and 'rehi', with the latter is better understood as a negative auxiliary. Finally, 'ake' is used in Kiwangona's imperative clauses.
Moreover, our finding also proves that negative markers in the Kiwangona dialect of the Lamaholot fill all three stages of a Jespersen Cycle, that is, Pre-predicate position (take); double position (ake -hala) and clause-final position (take/hala/rehi/waŋa/amu). The Jespersen Cycle of Lamaholot negators can be presented in Table 3 below. As can be seen from the table above, the negative markers are mostly found in the clause-final position. In fact, there are 5 observed negators existing in the given positions. Moreover, pre-predicate and double negators only share one negative marker for each of them, that is, 'take' and 'ake-hala', respectively. Further analysis on Table 03 can be seen at 4.2.2 on three stages of the Jespersen Cycle in the Kiwangona dialect of the Lamaholot.

Discussion 4.2.1 Forms, Meanings and Distributions of Negative Markers in Lamaholot 1. Expressing 'no'
The concept of 'no' can be generally compared to Lamaholot 'take' with the same meaning. Furthermore, 'take' can stand alone, making it a general negative marker in Lamaholot. Each time you express 'no' in Lamaholot, you will say 'take', as can be seen from Sentence (1) below.
Diachronically, 'take' must have been taken from *taq, a Proto Malayo Polynesian (PMP) form meaning 'no' or 'not' (Blust and Trussel,2010). In our evaluation, the *q sound in *taq has been strengthened into a [k] velar-plosive where the monosyllabic negator eventually becomes a disyllabic word when it becomes 'take'.

Expressing negative existential forms
Generally, negative existential forms can be glossed as 'nothing' or 'none'. Like many other Lamaholot dialects, the negative existential form in Kiwangona is 'take'. However, Kiwangone has two other negative existential forms, that is, 'waŋa' and 'amu'. These negators possess typically semantic elemens. In particular, 'waŋa' is understood as either 'none' or 'nothing', and 'amu' can be translated as 'nothing'.

a 'take.'
As can be seen from the sentences (2) and (3) below, 'take' interestingly can sit in either pre-predicate and clause-final position without changing the meaning of the clause. This results in a more flexible as well as richer grammatical construction of the negative clauses. In terms of the scope of negation, this phenomenon can be called 'movable negation' (Reesink,2002) because the negative marker can possibly be moved from the pre-predicate position into clause-final position or verse versa. Similarly to Fricke (2017), Reesink believes the characteristics of movable negation existing in Austronesian languages is due to contact with Papuan languages. The structures of negative clauses, therefore, where 'take' exists are as follows : (1)

2.b 'waŋa'
Diachronically, 'waŋa' may have been derived from *wada, a Proto Malayo-Polynesian (PMP) word meaning 'be; exist; have; wealthy; not exist; not have' (Blust and Trussel,2010). Moreover, it is semantically flexible. When referring to people, it means 'none', and it means 'nothing' when addressing things or activities. This negator is located in the final-clause position. In terms of the scope of negation, 'waŋa' can operate as both sentential negator as well as constituent negator by looking at sentences (4) and (5) below. In particular, it plays the 'sentence negator' role in sentence (4) as it determines the whole sentence's meaning. On the other hand, it functions as a constituent negator at the sentence (5) since it only grammatically exists in a supporting clause instead of a main verbal clause. Thus, in a sentence where 'waŋa' exists, it can normally be constructed either (1). Subj. + Pred. + Neg. Exist or (2). Subj.+ Pred. + Subj. + Neg.Exist.
In a predicate construction involving a verbal constituent, 'hala' obviously functions as a sentential negator as its appearance determines the negative meaning of the whole sentence. In terms of the scope of negation, 'hala' sits in a clause-final position which is totally different from 'take' and 'amu', where both of them can be moved from a pre-predicate position into a clause-final one. The possible grammatical construction where 'hala' exists is : Subj. + V.Pred + Obj. + Neg. (9). 'Nae gere tapo na rehi'. Nae gere tapo na rehi 3sg.subj V. N. 3sg. Aux.V.Neg he/she climb coconut (tree) cannot 'He cannot climb (a) coconut tree'.

3.b 'rehi'
'rehi' in Lamaholot is a negative auxiliary meaning 'cannot'. Moreover, learning from sentence (9), 'rehi' can be considered as a 'constituent negator' as it exists as a constituent in a separated clause than its main predicate. Morphosyntactically, this negator is tenseless, making it able to describe situations in the past, present or future. The grammatical structure where 'rehi' exists to indicate negative auxiliary is as follows : Subj. + V. + Obj. + Subj. + Aux.Neg. (10). Not only for verbal predicates, 'hala' also functions as a sentential negator in a clause where it gives negative identity to a nonverbal predicate. Sitting in the position of clause-final, 'hala' behaves like 'waŋa' and 'rehi'. Compared to 'take' and 'amu', it can be said that both negative markers are more movable than the other mentioned negators as they can sit in either position. The grammatical constituent where 'hala' exists in this non-verbal predicate is: Subj. + non-V.Pred + Prep. + Neg. (11). 'Ake tojo pita nire'. Ake Tojo pita Nire Neg. Imper V.

Expressing imperative form
N.

Do not
Push Door This 'Do not push this door !' 'ake' is an imperative negator in many Lamaholot dialects. Like English 'do not' or Indonesian 'Jangan', it exists in the initial position of a sentence. In our observation, it may have diachronically been weakened from 'take' by assuming that the [t] alveolar nasal sound in the initial position has been totally reduced. The clause construction where 'ake' presents can be drawn as follows: Neg.Imper + V. + Obj.

Three Stages of Jespersen Cycle in Kiwangona dialect of the Lamaholot (a) First Stage : pre-predicate position (take & amu)
As been highlighted by Dahl (1979:89), the prototypical location of the negative adverb in a clause cross-linguistically is prepredicate. Reesink (2002:244) strengthened Dahl's concept by looking at the data from a number of Austronesian languages and acknowledged that the original place of a negative adverb in Austronesian languages is, again, pre-predicate.
In the Lamaholot context where the Kiwangona dialect exists, two negative markers sitting in the pre-predicate position are 'take' and 'amu'. Following the history of Jespersen's cycle, these pre-predicate negators are then able to move to the clause-final position, allowing them to possibly exist in either position.

(b) Second Stage : double position (ake-hala)
In Jespersen's perspective (1917 :4), the general history on the development of double negatives among languages around the world starts with the following fixed pattern : 'the original negative adverb is first weakened, then found insufficient and therefore strengthened, generally through some additional word, and this in its turn may be felt as the negative proper and may then in the course of time be subject to the same development as the original word'.
As been discussed before, the double position in the Kiwangona context looks to agree with Jespersen's idea above. Firstly, 'take' is reduced into 'ake', which then functions as an imperative negator locating in the initial position of a clause. Linguists generally accept this process as 'grammaticalization'. After that, this reduced negator is fortified by another additional negator located in the clause-final position, that is, 'hala'. 'ake-hala' moreover means a 'must'. It can cross-linguistically be compared with 'neka-toe', a Manggaraian double negation with the same meaning (Akoli,2021). Sentences (12) and (13)  In a broader map of Austronesian languages, Blust (2013), after conducting research on 33 languages in 2009, argued that the existence of double negation in this language phylum is a cross-linguistically unusual phenomenon, although its exception can be identified from Kambera in Sumba island and Teop, an Oceanic language spoken in the West Salomons. In our evaluation, Blust's argument above, however, is, exactly resulted from a lack of data since a latter survey done by Vossen and van der Auwera (2014) toward 409 Austronesian languages revealed that 20 percent of the total observed languages possesses double negation. Therefore, instead of accepting it as an unusual phenomenon, we can consider double negation in Austronesian languages as a fairly common negation pattern since it does not only exist in Kambera (Klamer, 1998) and Teop but also in a number of languages we observe such as Balinese (Arka et al.,2018), Manggarai (Akoli,2021) and Lamaholot.

(c) Third Stage : Clause-Final position (hala/take/amu/waŋa/rehi)
In Austronesian languages with SVO word-order, the commonplace of negative adverb is pre-predicate. In contrast, many Papuan languages are dominated by SOV word-order. Consequently, the canonical place of negative adverb for these Papuan languages is in the clause-final position. Therefore, it is quite surprising to see that the number of Lamaholot negators located in clause-final position is more numerous than those in other positions. In our analysis, as both Reesink (2002) and Fricke (2017) agree that this phenomenon is due to a historical contact between Lamaholot and an unidentified Papuan language since earlier age, there is a grammatical impact from the Papuan language to Lamaholot enabling the Austronesian language to have a higher degree of 'movability' of its negative markers in its clauses. Therefore, 'take', and 'amu' can move from its canonical place, that is, prepredicate position, into the clause-final one. Moreover, 'hala' has nowadays emerged as a single post-predicate negator as a result of reaching the final stage of the Jespersen Cycle compared to 'ake -hala' as double negation. In our assumption, 'waŋa' and 'rehi' also become post-predicate negators when this pattern has regularly been used by its speakers as a common construction.

Conclusion
After discussing negation patterns and stages of the Jespersen cycle found in the Kiwangona dialect of the Lamaholot, there are a number of important points that should be stated here. The first is that negative markers in Kiwangona fill the five types of negation patterns theorized by Mosel and Spriggs. However, the most important one is that this finding agrees with Fricke (2017) that the existence of negative markers in the Kiwangona dialect follows three stages of the Jespersen Cycle of negation patterns in Flores-Lembata languages. Moreover, if Fricke compared nine different languages to obtain the complete stages of the Jespersen Cycle, the data we got from Kiwangona, in contrast, suggest that the three stages of the Jespersen Cycle can only be found in one single dialect like Kiwangona. Fourth, there are two grammatical aspects that observably resulted from the contact. The first one is grammaticalization that can be found at 'ake', which might have been weakened from 'take'; and the second one is 'higher degree of negator's movability' enabling negators like 'take' and 'amu' to move from the pre-predicate position into clause-final one. Looking at these findings, the Kiwangona dialect may be one of the Lamaholot dialects with interestingly complex aspects of negative construction. A more linguistically comprehensive investigation is needed to conduct in order to get a better picture of negation in Lamaholot dialects.