The phonology, morphology, and syntax of Sundanese

: Sundanese is an indigenous language that is spoken in West Java, Indonesia. This study considers the phonology, morphology, and syntax of Sundanese based on a corpus of more than two hundred words. It specifically aims to contribute to the literature of the language by providing some linguistic characteristics of Sundanese and comparing results to what has been previously introduced in the literature. Some previous studies of Sundanese have extensively covered the syntactic and some morphological structures of the language, but only a few studies have covered the phonological aspects. This study aims to provide a comprehensive analysis of the morphological, syntactic, and phonological aspects of Sundanese based on the production of a native speaker of the language. The participant translated all words into Sundanese and provided the singular and the plural forms for each word in the list. The participant also produced sentences using the same words in the corpus. Data were then observed, and patterns were grouped and categorized for analysis. Results showed some similarities and differences to the findings of previous work in the field. Conclusions were drawn and comparisons were made where appropriate.


Introduction
Sundanese, an indigenous language in Indonesia, is spoken in West Java and belongs to the family of Austronesian languages.Austronesian is one of the two largest language families in the world.The other language family is the Niger-Congo.Sundanese people form one of the largest ethnic groups in Indonesia, the second one after the Javanese.There are around 32 million speakers of Sundanese worldwide (Eberhard et al., 2019).One of the features of Sundanese is that it has speech levels.That is, it has different levels of politeness.To elders, parents, and younger generations, the polite form is preferred.When speaking to friends, a less polite speech is used.Impolite form of speech is only used when talking to animals, but not humans (Van Syoc, 1959).Recently, there has been a noticeable impact on the use of Sundanese speech levels by the younger generations which poses a threat to the stability of the language.Many Sundanese families prefer to teach their children the dominant languages, Indonesian and English, and gradually abandon their own native language (Indrayani, 2011).Wurm (2002) showed that a language becomes unsafe when it is used by the younger generations in limited domains only, not in all domains.In Wurm's terms, Sundanese would be considered unsafe and potentially an endangered language as the youth preferred using the dominant languages, not their own.Alwasilah (2008) also supported that most Sundanese people were showing negative attitudes towards their mother tongues.Families showed a strong tendency towards teaching their children the 'prestigious' English language at provided different sentences including these words.Data were then transcribed by the author and revised by the participant for accuracy purposes.Structures were observed and analyzed and patterns were highlighted to provide a thorough phonological, morphological, and syntactic analysis of Sundanese.The phonological analysis covered the language inventory and highlighted some phonological processes that took place in Sundanese.The morphological analysis covered the affixation system that nouns, verbs, and adjectives underwent.Finally, the syntactic analysis investigated the syntactic representations and word order of nouns and prepositional phrases, relative clauses, and passive constructions in Sundanese.

Results
Results are presented below in three subsections: the phonology of Sundanese, the morphology of Sundanese, and finally the syntax of Sundanese.

Consonants
Based on the transcription of the data provided, Table 1 presents the consonants that are found in the language inventory of Sundanese.Data included seven stops: the voiceless bilabial /p/, the voiced bilabial /b/, the voiceless alveolar /t/, the voiced alveolar /d/, the voiceless velar /k/, the voiced velar /g/, and the voiceless glottal stop / ʔ/.Some consonants were found to be inserted in certain environments, such as the glottal stop /ʔ/, and the two glides /w/ and /j/.These environments will be discussed later when phonological processes are considered.

Vowels
There are ten distinct vowels of Sundanese as presented in Figure 1.

ta.ʊn 'year'
There are some restrictions, however, on what segments can go in onset vs. coda positions in Sundanese roots.In CV.CVC syllable structure, almost any consonant is allowed to occur as the onset of either the first or second syllables, see Table 2.Only the glide /j/ does not appear as an initial onset in CVC syllable structure.3.3.Phonological processes

Dissimilation
The first phonological process observed in the data is dissimilation, where two sounds that are similar become different.Dissimilation in Sundanese takes place when a word's root contains the sound /r/.One of the possible ways of forming plurals in Sundanese is by inserting the infex /ar/ (few) or /arar/ (many).However, when a word's root has the sound /r/ in any position, /r/ dissimilates to /l/.

Assimilation
It seems that the insertion of the plural markers few and many, (-ar) and (-arar) respectively, not only dissimilates /r/ to /l/ when /r/ occurs anywhere in the root, but also assimilates to /al/ and /alal/ when /l/ starts a word.Data show that unlike /r/, the position of /l/ in a word's root matters.That is, when /l/ occurs word medially or finally, no changes take place (21).However, when /l/ occurs word initial, a process of assimilation takes place (22)

Nasal assimilation
When floating nasal consonants (such as the prefix/-ŋa/ or /-ŋ/) are attached to verb roots in the active form, the nasal segment assimilates with the first consonant of the root to share the same place of articulation.In Sundanese, the passive form is formed by the attachment of the prefix /-di/ to the underlying representation of the root.Thus, it is always helpful to consider the passive structure to explain any phonological process that take place in the language.As indicated in the language inventory, there are four nasal consonants in Sundanese /m/, /n/, /ŋ/, and /ɲ/.Examples for each case of nasal assimilation is presented below.
When a root starts with a voiceless bilabial stop /p/, the association of a nasal consonant will undergo an assimilation process and change to become a voiced nasal bilabial /m/ that shares the same place of articulation as /p/.Consider the following example: (23) The voiced nasal bilabial /m/ + /pikir/ mikir 'think' di-pikir passive form Similarly, when a root starts with a voiceless alveolar stop /t/, the nasal prefix /ŋa/ or /ŋ/ assimilates and changes to the voiced alveolar nasal /n/, which shares the same place of articulation as /t/, as presented in (24).( 24 Only one word in the data disagrees with this dissimilation process presented here, and that is the word: /mmam/ 'eat', where the passive form is /di-mam/.For this word, the nasal /m/ is allowed to occur as the first segment in the root in this exceptional case, and since it is the only word found in the data, it is not possible to make a generalization.

Insertion
Insertion or epenthesis is observed in four environments in the Sundanese data.The first environment is when the root of the verb starts with a vowel, a glottal stop is inserted when making the passive form by adding the prefix di-plus a glottal stop.The last environment of insertion is observed in the data when plural forms are considered.One way to form plurals is Sundanese is by reduplicating the first CV of the root and adding the suffix (-an) (more discussion on plurals is provided in the morphology subsection).Three different consonants might be inserted in the derived plural forms based on vowel quality: The glottal stop /ʔ/, and it is usually inserted as the coda for words that end with vowels, or used to separate two identical vowels, see ( 30 An interesting observation of this study is on the attachment of the suffix (-na), a possessive marker, to nouns to indicate possessiveness for the third person only, unless if it is followed by the first-person singular pronoun sim kuriŋ to indicate the first singular pronoun 'my'.Consider the following example where the singular possessive marker (-na) is attached to the word taŋkal 'tree': (37) taŋkal sim kuriŋ tree-possessive my 'My tree' taŋkal-na sim kuriŋ tree-possessive my (owned by me) 'My tree' Or just: (38) taŋkal-na tree-possessive 'His/Her tree' In example (38), the possessive marker (-na) is attached to the noun with no other pronouns following it.Thus, the meaning is understood to be for the third person pronouns only.Now consider the same sentences for the plural word ta-taŋkal-an 'trees' where the possessive marker /-na/ becomes nana for noun-possessive marker agreement: (note that the first syllable is reduplicated for plurality).( 39) ta-taŋkal-an sim kuriŋ tree-(plural) my 'My trees' ta-taŋkal-an-nana sim kuriŋ (more definite) tree(plural)-possessive my (owned by me) 'My trees' Or just: (40) ?ta-taŋkal-an-nana tree(plural)-possessive 'His/Her trees' But not: (41) * ta-taŋkal-an-na sim kuriŋ Furthermore, the previous examples show that there must be an agreement of nouns and possessive markers; that is, the singular possessive marker /-na/ must be attached to singular nouns, and the plural possessive marker /nana/ must be attached to plural nouns.Violations of this rule lead to ungrammatical sentences (41).
Not only agreement is important here, but also the order of the plural circumfix (reduplicated syllable+root+an) and the possessive marker /nana/.The plural possessive marker /nana/ must always follow the circumfix, and never precede it or even precede a portion of it.Consider the following: ( 42 Example (46) shows that it is ungrammatical to have a singular subject followed by a plural verb.Similarly, nouns must also agree in number with adjectives.See ( 47) and ( 48 Example (50) shows that adjectives can be pluralized by the infixation of (-ar/al) or (-arar/alal) after the first consonant of the word.
Moreover, to form verbs that are derived from adjectives (e.g., 'to become'), the circumfix (ŋa + Adj + an) is used.Note, however, that no objects must be present with this circumfix, as shown in the following sentences: (51) imah-na Abul ŋa-bodas-an house-Possessive name to become (causative)-white 'Abul's house becomes white' (52) bal ŋabɨrɨm-an ball to become (causative)-red 'The ball becomes red' The attachment of the possessive marker /-na/ to the word imah in example (51) indicates that it refers to the third person, which happens to be 'Abul' in this sentence or just 'his/her house' when no overt noun is present.
However, for causative forms, a different circumfix is used (ŋa + Adj + kɨn).This circumfix, which indicates the meaning of 'to make something be something', requires an object following it.Interestingly, the suffix /-na/ that is attached to the word badʒiŋ-na 'squirrel' in example (55) differs from the possessive marker /-na/./-na/ as it is used in example (55) serves a different function, and that is to refer to a specific entity.In other words, 'a specific squirrel' compared to badʒiŋ 'a squirrel'.
Furthermore, when comparing one item to another, the circumfix (laŋkuŋ + Adj + ti batan) is used.

Verbs
Just like nouns and adjectives, plural verbs are formed by the insertion of (ar) 'few' and (arar) 'many' or their alternations (al) and (alal), respectively.In roots that start with consonants, (ar) and (arar) are inserted after the first consonants (61).On the other hand, when roots start with vowels, (ar) and (arar) are prefixed to the roots ( 62 In (72), there are two arguments following the imperative verb pasak 'cook': the direct object lauk 'fish', and the prepositional phrase kaŋgo bapa 'for father'.Thus, the suffix /-kɨn/ is attached.The same sentence will be ungrammatical if /-kɨn/ is not attached to the root of the verb, as shown in (73).
The benefactive verbs in Sundanese are formed in two ways depending on the position of the direct and the indirect objects, (DO) and (IO), respectively.The first way is formed following this structure: S V (active) DO [kaŋgo IO].This structure shows that when the direct object immediately follows the verb, the active form of the verb is used along with the word kango 'for' that should precede the indirect object, as exemplified in examples ( 74) and ( 75). ( 74 The second way to form benefactive verbs is by the insertion of the circumfix [maŋ + V(active) + k ɨn] when the benefactor appears between the verb and the direct object, following this structure: S maŋ-

Prepositional phrases (PPs)
In the data of this study, there are seven Sundanese prepositions as listed in ( 89 The adverb di luhur-ɨn indicates the meaning of 'something is on top of something else'.The underlying root luhur is attached to the suffix /-ɨn/, which also appears in other adverbs like hɑndap-ɨn and sɑbudɨr-ɨn.Note that when the root luhur appears as a verb or a verb-like, different suffixes are attached to it.See (96) below: (96) Katie ŋa-luhur-an Abul kaŋgo syntax name active-root-active name for syntax 'Katie gets higher grade than Abul in Syntax' In several sentences in the data, the word aja 'exist/here' shows up before prepositions, see ( 97) and (98).The presence of aja is important; otherwise, the meaning will not be conveyed.(97) andʒɨna aja di pasar She exist at market 'She is at the market' (98) sim kuriŋ aja di sakala I exist at school 'I am here at school' The word aja, which denotes the meaning of existence, seems to appear in sentences with no main verbs around; however, this is not consistent as there are instances where neither main verbs nor aja appear in a sentence (99).Thus, the presence of this word remains unknown.(99) lampu di luhur-ɨn andʒɨna light under you 'The light is above you' Interestingly, the word aja must agree with the subject in terms of number.In other words, singular subjects require singular form of aja, and plural subjects require plural form of aja.Since the word aja starts with a vowel, the infix (ar/arar) is prefixed to the word, instead of inserted.The only difference found between the reflexives ɲalira and diri is that ɲalira follows the pronoun it describes while diri always precedes it.It even precedes the word ɲalira when both occur in one sentence, See (112).
Furthermore, when reflexives refer to objects 'things' like the word masin 'machine', for example, only the reflexive word ɲalira is used with no repeated pronouns.Consider the following example: (114) masin parɨm ɲalira pas parantos atosan Machine switch off self-reflexive past finish 'The machine switches itself off when it's finished'.
Another interesting observation on reflexives is the use of intensive pronouns.When reflexives are used as intensive pronouns, reflexives must appear at the beginning of the sentence followed by a relative clause to indicate the meaning of 'myself is the one who made this', as exemplified below:

Passive construction
As indicated earlier in the morphology part, the passive form of the verb is formed by attaching the prefix (di-) to the underlying root, see ( 118 There are different affixes that are attached to the verb in its passive form.The first of which is the suffix /-an/.When it is added to the passive verb, the presence of agents becomes obligatory (120).Absence of agents in such cases will render ungrammatical sentences ( 121).( 120 Another important suffix that can be added to the verb in the passive form is the suffix /-kɨn/.The addition of this suffix requires the sentence to have a locative complement.Unlike the suffix /-an/, the presence of the agent here is optional.( 122 Similarly, the passive benefactive structure is formed by adding the prefix /di-/ to the root of the verb, and the presence of the agent is optional here, as exemplified in ( 124).( 124) lauk parantos di-pasak [ku sim kuriŋ] kaŋgo Katie fish past passive -cook by me for name 'The fish was cooked (by me) for Katie' However, when the circumfix (di + root + an) is used for passive form, the presence of the agent becomes obligatory, see ( 125).( 125 For the future tense, there are two adverbs used: bade and bakal.The only difference between the two is that the former is used to refer to immediate or near future actions ( 138) and ( 139), and the latter is used for later or far future actions ( 140) and ( 141).( 138

Relative clauses
The word anu 'that/who/which' is used to indicate relative clauses in Sundanese.It immediately follows the subject/object it relativizes.There are different structures of relatives depending on whether subjects or objects are being relativized.Only three possible structures of relative clauses are found in the data collected: In (142a), the object of the main clause is relativized to become the object of the relative clause.A topicalization takes place where the object of the relative clause is fronted to become the subject of the clause.In such case, two ways are possible to from the relative clause: (143) Object of the main clause is relativized, object in relative clause a. anu + di(passive)-V(root) + ku + agent (pronoun) b. anu + agent(pronoun) + V(root) *but not agent(noun) + V(root) The second relative clause structure, as illustrated in (142b), shows that the object of the main clause is relativized to become the subject of the clause.In this case, the word anu immediately follows the object of the main clause, and it must be followed by the verb and a complement if the verb has a transitive nature.The third and last structure of Sundanese relative clauses, as illustrated in (142c), occurs when the subject of the main clause is relativized to become the subject of the relative clause.In this case, the word anu comes right after the main subject followed by an obligatory verb and an object, if required:

Discussion
Regarding the three research questions of the study, the results' section provided detailed answers to the first two questions about the phonotactic constraints of the Sundanese sound system, and the formation of words, phrases, and sentences.The current discussion section will address the third question by providing an interpretation of the current study's findings and their relevance to the findings of previous work on Sundanese.
Phonologically, all the consonants found in the data were distinct phonemes, except for the glottal stop that was found to be inserted in certain environments.In Sundanese, glottal stops were inserted in three environments: at the beginning of words that start with vowels, at the end of words that end with vowels, and in the middle of some words to separate two identical vowels.As for vowels, data showed that, unlike findings that viewed the two vowels as distinct, the two vowels [o] and [ʊ] were acoustically similar with the former being more rounded.Regarding consonant and vowel clusters, data showed that both were allowed in Sundanese.Anderson (1972) explained that consonant clusters were not common in Sundanese, but he did not deny the possibility of their existence.Data from this study, although limited, showed that many words in Sundanese had clusters.This finding contradicts the claim of Anderson regarding the uncommonness of consonant clusters in the language.Some phonological processes such as assimilation and dissimilation were observed in the data when nasals or plural markers (-ar and -arar) were involved.Particularly, when words contained approximants or nasal consonants.Data also showed that nasal consonants affected the following and preceding vowels due to nasality spread.Similarly, Anderson (1972) discussed nasality spread in his study and showed that nasality spread to affect all vowels following the nasalized consonant when a word contained a nasal consonant.Sometimes, the spread extended to affect vowels of adjacent syllables.He supported this claim by citing examples from (Robins, 1957, p. 91 Most Sundanese words consist of two syllables.However, words with one or more than two syllables were allowed in the language.The minimum syllable structure that could exist in Sundanese was V. Hanafi (1997) and Müller (2001), among others, agreed with this finding.Almost any consonant is allowed to occur as an onset except for the glide /j/.Moreover, the coda position of the first and second syllables in Sundanese are restricted to nasals, and a few other segments in the language.Due to the limited set of data, other studies were observed to check consistency.Van Syoc (1959) confirmed that restrictions on consonants applied in Sundanese where certain sound combinations were not allowed in the language.
Morphologically, Wessing (1976) explained that all Sundanese nouns, verbs, and adjectives could be pluralized.He showed that nouns were pluralized by reduplication while adjectives and verbs were pluralized by the infixation of (-ar/al).He added that only a few nouns like mass nouns could be pluralized using the same infixes of verbs and adjectives.However, data collected in this paper contradicted this finding.Most Sundanese nouns, if not all, were pluralized by the infixations of (-ar/al) just like verbs and adjectives.Reduplication was also a way of making nouns plural in Sundanese.Furthermore, results showed that the agreement of nouns with verbs and adjectives in number as well as the order of affixes within words or sentences matter in the language.Some interesting observations that were noticed in the data included the attachments of certain affixes in Sundanese.Some affixes, like /an/, required verbs to have objects or complements, otherwise they would be ungrammatical.Other affixes were attached to verbs only when more than one argument was involved, for example, the suffix /-kɨn/.
Moreover, some forms of verbs require certain words to precede or follow them in Sundanese sentences.For example, in the benefactive verbs, the word kaŋgo 'for' must follow the verb and precede the indirect object.In the causative form of verbs, the word miwaraŋ 'make' preceded both the verb and the object.
Syntactically, results showed that nouns preceded all other determiners, adjectives, genitives, and possessive pronouns.Numbers and agentive modifiers, on the other hand, did not follow this rule; that is, they preceded other nouns.Furthermore, Sundanese is an SVO language; that is, subjects always precede verbs and objects while prepositional phrases or complements follow verbs.These results were similar to what have been found in previous literature (e.g., Hardjadibrata, 1985;Müller, 2001;and Doran and Bangga, 2022).Interestingly, the word aja 'exist/here', which denotes the meaning of existence, was present before prepositions in some sentences of the data, and it seemed to appear in sentences with no overt main verbs.This finding was not previously discussed in the literature of Sundanese, at least to the author's knowledge.Since only a few sentences contained this word, rules that govern the presence of this word remained unknown and required further investigations.
Regarding reflexives, both nouns and adjectives could be relativized in Sundanese.Noun phrases were relativized to indicate possessiveness while adjectives were relativized to distinguish between intentional and unintentional actions.Reflexives were always followed by the word ɲalira, which means 'self '.Interestingly, two other words were used in the data to indicate the reflexive 'self ': soraŋan and diri.The two words ɲalira and diri were used for people while soraŋan was used for non-humans.The only difference found between ɲalira and diri was that ɲalira followed the pronoun it described while diri always preceded it.It even preceded the word ɲalira when both occurred in one sentence.
Another interesting finding about Sundanese syntactic structures was observed in the use of Sundanese temporal adverbs.To indicate the past tense, the adverb parantos was used.However, when adjectives were involved, a different temporal adverb was used, kapuŋkur.For the future tense, two adverbs were used in the language: bade and bakal.The adverb bade was used for immediate or near future actions, while bakal was used for far future actions.

Conclusion
This paper provides a morphological, syntactic, and phonological analysis of Sundanese, an indigenous language that is spoken in West Java, Indonesia.Sundanese people, although they form one of the largest ethnic groups in Indonesia, showed a tendency towards teaching their younger generations Indonesian and English instead, the two official languages in the country.This, if continued, would pose a threat to the stability of Sundanese and result in causing the language to become endangered.More than two hundred words of the Swadesh list were translated, transcribed, and used in sentences by a male informant who is a native speaker of the language.The current study aims to contribute to the literature of the language by providing some linguistic characteristics of Sundanese and comparing results to what has been introduced in the literature.Data were observed, and patters were grouped and categorized for analysis.Results showed some similarities and differences to the findings of other previous works in the field.Conclusions were drawn and comparisons were made where appropriate.
) ta-taŋkal-an-nana sim kuriŋ tree(plural)-Possessive my (owned by me) 'My trees' (43) * ta-taŋkal-nana-an sim kuriŋ tree(plural)-Possessive my (owned by me) 'My trees' A final point to make here about nouns is the Subject-Verb agreement.Nouns in Sundanese must always agree with verbs; that is, singular subjects require singular verbs, and plural subjects require plural verbs, as the following examples indicate.(44) gogog lumpat di imah dog-singular run-singular in house 'The dog runs inside the house' runs inside the house' ) below.(47) Plural nouns and adjectives: iʔ-imah-an anu b-al-ərəsih house-SUF (plural) that-REL clean-Infix(plural) 'Many houses are clean' (48) Singular nouns and adjectives: imah anu bərəsih house(singular) that-REL clean-Adj(singular) 'A clean house'4.2.AdjectivesSundanese adjectives can also be singular or plural.Consider the following examples:
According to the data collected, most Sundanese words consist of two syllables.However, words with one or more than two syllables are also acceptable.The minimum syllable structure that can exist in Sundanese is V. Sundanese syllables can either be V, CV, VC, or CVC.Examples of each form are

Table 2 .
Sundanese onsets in initial and medial syllable structures.CVC.CVC is/are restricted to nasals (17), and few other segments such as: /r/, /h/, /l/ and /s/, as in (18) (these four words are the only ones found in the data that allow other segments to exist as codas).Most segments, on the other hand, can occur as codas for final syllables in Sundanese roots, with the exception of palatals /ɲ/ and /j/, affricates /tʃ/ and /dʒ/, the voiced stop /b/, and the glide /w/.
Moreover, the coda position of the first syllable in CVC.CVC syllable structure or the first and second syllables in CV(C).
To indicate Agentive nouns, nouns that do the action described by verbs, the word tukaŋ must precede the noun to modify it, see examples below.
However, the addition of this suffix requires verbs to be plural or to have plural objects (64).If the meaning is intended for a singular object, then /-an/ must not be attached to the verb (65).
Moreover, when an intransitive verb is used where only IO appears in a prepositional phrase preceded by kaŋgo, the following simple structure is used: S V(active) [kaŋgo IO].See examples below: Finally, to express the causative form of verbs in Sundanese, the word miwaraŋ 'make' is used before both the verb in its active form and the object of the sentence.Here are some examples:In NPs, nouns always precede all other determiners, adjectives, genitives, and possessive pronouns.
A final point to mention here is the use of tə to indicate negation in Sundanese.Two words in the data are found to function as negation markers 'not': həntə and tə.The former is used with verbs and adjectives, (102) and (103), while the latter only appears with the preposition tə disarəŋan 'without' (104).In the data collected, same personal pronouns that function as the main subject of the sentence are repeated and followed by the word ɲalira to indicate reflexiveness in Sundanese.Here are some examples: Interestingly, two other words that indicate the reflexive meaning 'self ' are found in the data: soraŋan and diri.The participant indicated that both words ɲalira and diri are reflexives of people (112) while The presence of the agent ku sim kuriŋ 'by me' following the verb di-dɔrɔŋ is optional.Also, it is optional to have compliments.
Another important issue to be covered here is the difference between passive verbs in terms of intentional vs. accidental events.To indicate the accidental form of the passive forms, the prefix /ka-/ is attached instead of /di-/.Consider the following accidental passive verbs:Temporal adverbs are adverbs that immediately precede main verbs or sometimes adjectives to indicate the time reference for the event in the sentence.Each tense has its own temporal adverbs in Sundanese.To indicate the past tense in Sundanese, the adverb parantos is used, as exemplified below:The temporal adverb nudʒu, on the other hand, is used to refer to the progressive form of verbs in Sundanese.Consider the two following examples: The man who cut my tree lives in that house.'A final point to make here is that both nouns and adjectives can be relativized in Sundanese.Noun phrases are relativized to indicate possessiveness (151) while adjectives are relativized to distinguish between intentional vs. unintentional actions (152).