Free and Phonologically Conditioned Allomorphs of the Bidayuh-Somu Language

This study is on linguistics, that is, morphology. It aims to describe the morphemes and allomorphs of the Bidayuh-Somu Language. The study is descriptive, that is, documentary analysis. The data of this study is secondary, a-ready-to-use data, taken from a doctoral dissertation. It is an example sorted based on morphemes and allomorphs of the language. Since the language is indigenous and not in written form, the data is phonemic and morphemic transcription. As a limitation, the data is only morphemes with allomorphs. The sole morpheme with zero allomorphs is not included in this study. As an exemplification, this study describes the morphemes that are nominal and verbal. They are prefix {puN-} and circumfix {niN- ŋ eh}. The allomorph of the morpheme for this current study is the element of nasalization consisting of the phoneme /m-/, /n-/, / ŋ -/, and / ɲ -/, and alternants. The allomorphs of this language are free and phonologically conditioned.


I. INTRODUCTION
As mapped by Alloy et al. (2008), the Bidayuh-Somu Language, abbreviated BSL, is one of the languages and dialects that are spoken by natives in West Kalimantan Indonesia. By 2015, the language was spoken by 40,166 people primarily residing in two subdistricts, namely Kembayan and Noyan, which are situated along Indonesian Kalimantan and Malaysian Borneo boundary in the district Sanggau (Bunau, 2019). In terms of language family, Asmah (1983) classified the languages of Borneo or Kalimantan are of the Austronesian language family. Thus, this Bidayuh-Somu Language is a family member of the Austronesian and is subgrouped into the language of the Bidayuh or Bidayuhik sub-group.
In terms of language, Levi-Strauss (1969, pp. 56-57) stated: "... language is a social phenomenon. When we speak, we are unaware of our language's morphological laws and the phonemes we employ to convey different meanings… When they speak, they are unaware of their language's morphological laws or the phonemes that they employ to convey different meanings." This social phenomenon and the unconsciousness are perceived as similar to the people of Bidayuh-Somu and their language. As a concern towards that unconsciousness, this study attempts to identify and describe the morphemes and affixation process. It is a study of the complex word-formation of the language. In terms of that complex word formation, morphology studies words resulting from the process of affixation, compounding, and reduplication (Plag, 2014). Nevertheless, it is the affixation process that is sorted to describe in this article.
This study is linguistics and focused on affixation as one of the morphological processes of word formation. In line with this morphology, Denham and Lobeck (2010) stated that morphology is a study of the system of rules underlying our knowledge of the structure of the words. In brief, morphology is the study of word structure. Moreover, Bloomfiled (1933) had initially stated that the morphology of a language means the construction of words in which bound forms appear among the constituents. By definition, the resultant forms are either bound forms or expressions. That statement shows that the element of morphology is free and bound morpheme (Bunau & Yusof, 2018). The combination of the construction of free and bound morpheme then forms the complex word. Hence, this study structurally studies the constituents or parts of the word, that is, phonological and grammatical features found in the process of word-formation.
To have a broader context, morphology is a study of the internal structure of the word. It is a study on a form, or the relation between word and morpheme as an element of word, and word as an element or unit of phrase. Since words may function as free morphemes and minimal phrases (Asmah, 2013), therefore the word resulting from the process of morphology discussed in this study is classified into nominal and verbal classes. Since that morphology is a study of form and was put forward early by Bloomfield (1933), thus its process will result in grammatical and semantic functions. The grammatical function is the changes in word class or part of speech after the process of morphology, while the semantic function is the meaning shown after the process (Martín Arista & Vea Escarza, 2016). That morphological process applies by affix element to free morpheme that is root or base.
The affix element may consist of a bound morpheme and its variants. So that the bound morpheme may function as class-maintaining and class-changing. In this context, Booij and Booij (2019) provides an example by using the word buyer which is a noun. The word has two morphemes, that is /buy/ and -er. The morpheme buy is a verb and free, while -er is bound and cannot stand alone. Root word buy has lexical meaning, that is meaning that does not require grammar (Omar, 2014). The complex word buyer, on the contrary, has grammatical meaning, the meaning that is indicated after the process of morphology.
For it is a study of morphology, this affixation is structural, that is, word structure containing the combination of free and bound morpheme and word formation process. It is a study of the language that involves the feature of form and meaning (Bloomfiled, 1933). In short, the form of the language is to define its particular meaning. It means that in the process of morphology, the combination of free and bound morpheme is not aimed only to recognize its structure but also to define its meaning, that is, to maintain or to change word class. Phonetics or grammatical features determine word meaning based on its class or part of speech. It was explained by Bloomfield (1933, p. 159) a long time ago with his famous example of the phoneme /-y/ that is affixed to the words John, Bill, and Dan is a noun to form complex words Johnny, Billy, and Danny that is adjective. The phoneme /-y/ indicates meaning little.
Affixation, citing Denham and Lobeck (2010), is a process of attaching affixes or morphemes to other morphemes or words. Affixes are bound morphemes, including prefixes, suffixes, infixes, and circumfixes. In complex word formation, the word is a structure or construction consisting of the free and bound morpheme. Hence, affixation that employs bound morpheme to affix to free morpheme is a system of rules in morphology. Furthermore, it indicates that the important element that building a word is a morpheme, the smallest grammatical unit that has meaning (Omar & Yahaya, 2018). It means that affixation is a wordformation process that combines free and bound morphemes to form complex words. In other words, it is a study of word structure in which a complex word is a construction of constituents consisting of the free and bound morpheme.
In the affixation process, a bound morpheme consists of two, namely derivational and inflectional. A derivational morpheme is a bound morpheme that is used to derive new word that is complex words and may function to change word class or part of speech and to maintain word class or part of speech as well. A derivational morpheme is an affix that comprises prefixes, suffixes, infixes, and circumfixes. Therefore, derivational morphemes can also be class-changing or class-maintaining (Asmah, 2013). Unlike derivational morpheme, the inflectional morpheme, as in English, is an affix that is only a suffix. It does not function to change word class or part of speech. It only maintains word class or part of speech remaining to be verbs, nouns, or adjectives. Its suffixation process indicates tense for verbs, plurality for nouns, and comparison and superlative for adjectives.
The instance of derivational morpheme, as discussed in Booij (2019), is /-er/, as found in word /player/. Observing the process of its suffixation, it is obvious that derivational morpheme /-er/ functions to change /play/, that is a verb into /player/, which is a noun. Thus, the morpheme /-er/ is class-changing. Moreover, the instances of the inflectional morpheme, as discussed in Crosson et al. (2020), are /-s/ as in verb /tells/ and noun /balls/, /-es/ as in verb /watches/ and noun /boxes/, /-ing/ as in verb /reading/, /-er/, as in adjective /taller/, and /-est/ as in adjective /smallest/. As bound morphemes, these derivational and inflectional morphemes cannot stand alone as a word, nor can they stand alone in isolation. They should be attached to other morphemes to form a word.
Affixation is one morphological process in which the construction is of free and bound forms that have meaning. The form that has meaning is called morpheme and its variants (Stockbridge et al., 2021). A morpheme is an abstract unit; in language, its use is represented through phonological sound. That phonological sound is called allomorph, or variant of morpheme. It means that allomorph is a member of a morph or a form. As a variant of morpheme, allomorph is the real use of morpheme (Mel'cuk, 2016). Furthermore, Carstairs-McCarthy (2002) states that an allomorph is a phonological sound that is based on the context. The sound is distinctive according to the presence of sound of the initial or final consonant or vowel of the word or syllable of the word.
Moreover, Asmah (2013) said that morpheme is recognized from its allomorph, and each morpheme consists of at least one allomorph. The allomorph is divided into two, the phonological-conditioned and free. The phonological-conditioned allomorph is the representation of a morpheme that can be recognized based on its phonological, grammatical, and lexical features. In addition, in terms of selection, phonologically-conditioned allomorph selection refers to any case in which allomorphs are chosen based on the phonology of the stem, affix, or phonological word to which they attach (Carstairs-McCarthy, 2005).
The distribution characteristic of phonological-conditioned allomorphs is set aside or aloof to each other. It means that the initial or final phoneme of a free morpheme, root or base, is covert or overt when the allomorph is affixed. The free allomorph is an allomorph that is free to exchange with each other but still belongs to the same morpheme as a member or variant of the morpheme. In line with the affixation process, the bound morpheme is an affix consisting of prefixes, suffixes, infixes, and circumfixes (Lane et al., 2019). In terms of position in a word, the prefix is initial while the suffix is final. The infix is usually inserted between the first and second phonemes of the word. The position of circumfix is initial and final and affixed as a set to the word.
Derivational meaning or the derivational changes of meaning can be determined based on the elements or parts of a complex word form. Martini (2016) put forward that derivational meaning changes are subject to variations in transparency. For instance, the meaning of a complex word form /shortness/ is derived from its constituent parts, /short/ and /ness/. The constituent part /-ness/ that is derivational morpheme changes /short/ that is adjective to becoming /shortness/ that is noun. Moreover, Greshchuk (2019) stated that the meaning of a bound morpheme should be persistent. For example, the meaning of agentive morpheme /er/ in the words /singer/, /painter/, lover, and /worker/ is 'one who' or 'doer'. Yet, the meanings are also affected by the situational context and social usage of language (Smith, 2020).
There are two previous descriptive studies on the languages of Borneo or Kalimantan implemented through the use of a qualitative approach and field linguistic method taken as precedence for this study. The studies compiled the structural theory to identify the words and morphemes, and bound morphemes and allomorphs of the languages. The first is the study entitles Grammatical Description of the Iban Language of Sarawak by Asmah (2013). In terms of morphology, the study found that the word of the language is a complex word in which formation is by employing free and bound morpheme including its allomorph as a construction. Based on the class of words, it was found that the word and morpheme are verbal and nominal. It was also found that the bound morpheme is both class-maintaining and classchanging. Another descriptive study on the language of Borneo or Kalimantan is Six Bidayuhic Variants of the Sekadau River by Shin & Collins ( 2001). The study is qualitative and applied field linguistics as the method of study. The study studied six variants or dialects of the Bidayuh group along the Sekadau river in West Kalimantan. It was found that the words and morphemes of the six variants are distinctive although they are all grouped as Bidayuh.

II. METHOD
The approach of this study is qualitative. Omar (2014) said that in morphology, the study describes the features or characteristics of morphemes and allomorphs. Data of this study is bound morphemes and allomorphs of the Bidayuh-Somu Language. The data were sorted from a doctoral dissertation, in which the data were taken from oral utterances spoken by informants, and the technique used to collect the data was recording and observing. The source material of oral utterances was oral traditions or folklores and conversations. In accordance with oral tradition or folklore, Maltseva ( 2018) stated that folklore or oral tradition is a source or material for data of language study. Moreover, the informants were native speakers of the language. To collect sufficient and proper data on utterances, the number of informants was extended by using the snowball sampling technique (Naderifar et al., 2017). Observing, that is participant observation (Baese-Berk, 2019), was applied by listening and noting utterances the people said during the activity of the forum or event. The sufficient data were then transcribed, and the transcription is phonemic. Its purpose, as argued by Vythelingum et al. ( 2018), is to provide the symbol of the sound of the utterance.
The method of data analysis, quoting Ulfsbjorninn (2020), is selection for morpheme and allomorph determiner, and equivalence, for glossing. The equivalence method was also implemented using the distributional model (Lenci et al., 2022), or the referential technique (Indrati, 2018), in which English is the reference. The technique of analyzing data is direct element classification (Cristina & Afriana, 2021), which is to classify the unit of lingual or language element of the Bidayuh-Somu Language into the nominal and verbal morphemes. In other words, the technique of data analysis was applied using the Item and Arrangement Model (Tomaschek et al., 2019) to analyze the structure of words. To analyze the process of word formation, the technique of the Item and Process Model (Tomaschek et al., 2019) was also applied. The unit of lingual or the item of the language element for this study is the complex words of the Bidayuh-Somu Language that is formed by the combination of free morpheme, and bound morpheme and its allomorph.

III. RESULTS AND DISCUSSION
Based on its construction, it was found that the derivational complex word of the Bidayuh-Somu Language is the combination of the free and bound morpheme. The bound morpheme, to refer to Hankamer and Mikkelsen (2018), is prefix, suffix, infix, and circumfix. For prefixation, the structure is bound morpheme + free morpheme, whereas, for suffixation, its structure is free morpheme + bound morpheme. For circumfixation, the structure is bound morpheme + free morpheme + bound morpheme. construction of infixation is similar to prefixation, which is bound morpheme + free morpheme. In infixation, it is also common to say that the infix is inserted between the first and second phoneme of a free morpheme.

The result of data and analysis show that the Bidayuh-Somu Language has 22 (twenty-two) bound morphemes, not to include nasalization. The nominal morpheme with element of nasalization consists of prefix {puN-}, and circumfixes {kuN-ŋeh} and {puN-ŋeh}. The verbal morpheme containing nasalization is {buN-}, {kuN-}, {niN-}, and {tiN-}. The verbal morpheme has two circumfixes, namely {niN-ŋeh} and {tiN-ŋeh}.
It was identified that there are 10 morphemes present with the element of morphophoneme. As morphophonemic, it is the relation between morpheme and phoneme. In the context of this study, the archiphoneme N is to represent nasal phonemes consisting of /m/, /n/, /ŋ/, and /ɲ/. In word formation, the process is nasalization. As a result, the allomorphs of the morphemes with archiphoneme N element are phonological-conditioned (Asmah, 2013). The allomorph is set aside or aloof when it is affixed to root or base started with particular phonemes.
As a limitation, the description presented in this article only covers the nominal morpheme {puN-} and its allomorphs, and the verbal morpheme {niN-ŋeh} and its allomorphs.

A. Morpheme {puN-}
This morpheme {puN-} is a prefix and presents with a root or base that is a verb. Consequently, this morpheme is class-changing. This morpheme can be affixed to a free morpheme started with all consonant and vowel phonemes. The morphophonemic or nasalization process is additive and replacive. It means that the first phoneme of the root or base is nasalized. If it is additive, the process of the morphophonemic or nasalization will result in double homorganic nasal-oral consonants to root or base, and the first phoneme is overt.
1. Allomorph /pu-/ This allomorph enters the root that is the verb, and therefore, it is class-changing. It derives noun from a verb. The example of its affixation process goes in the following Table I:  From the example above, it is obvious that this allomorph is affixed to roots started with liquid, nasal /m/, and plosive /b/ consonants.

Allomorph /pum-/
As shown by the data, the allomorph /pum-/ is only affixed to the root that is a verb. So, this allomorph functions as class-changing. Look at the examples of its affixation in Table II below: This allomorph changes verb into a noun. Since the nasal is present, thus this allomorph is phonologically-conditioned. The examples indicate that allomorph /pum-/ only enters the root in which the first phoneme is a plosive consonant. The process of affixation of this allomorph results in double homorganic nasal-consonant /mp/ to base.
3. Allomorph /pun-/ As a variant or member of morpheme {puN-}, this allomorph /pun-/ is found to present with the root that is a verb. It is a class-changing and phonologically-conditioned allomorph. In the following Table III is an example of its affixation process: Like allomorph /pum-/ that creates double homorganic nasal-oral consonant to base, this allomorph creates the same, that is /nc/ and /nt/. Moreover, based on the example, it is identified that this allomorph is prefixed to root initiated with plosive /t/ and affricate /c/ consonants.

Allomorph /puŋ-/
Unlike other previous 4 allomorphs of morpheme {puN-}, this allomorph /puŋ-/ is present with verb and noun, and so it is class-changing and class-maintaining. This allomorph is phonologically-conditioned, and the process of its attachment results in double homorganic nasal-oral consonant /ŋk/ to the base. Please see the process of its prefixation in following Table IV:  The nasal element is additive as seen in examples number 1, 3, 4, and 5, and is replacive as seen in example number 2. This allomorph is affixed to a root beginning with a vowel and plosive /k/. Based on the analysis, it is also found that the allomorph /puŋ-/ may vary with alternant /pur-/ (see example number 3-5 above) without changing its meaning when it is affixed to a root initiated with a vowel.

Allomorph /puɲ-/
Based on the data, it is found that allomorph /puɲ-/ is a prefix attached to a root that is a verb and noun. As a result, the allomorph is class-maintaining and class-changing. In the following Table V is the process of its affixation: Its analysis shows that the allomorph is phonologically-conditioned since the first phoneme of the root is covert when the allomorph is affixed. In addition, this allomorph is attached to the root starting with affricate /j/ and fricative /s/ consonants.

B. Morpheme {niN-ŋeh}
This study finds that the circumfix morpheme {niN-ŋeh} is present before and after free morpheme that is verb and noun. Consequently, this verbal circumfix is class-maintaining and class-changing. Based on the analysis, it is also identified that this circumfix derives from a passive verb. It is listed that the morpheme {niN-ŋeh} has the member, that is allomorph /ni-ŋeh/, /nim-ŋeh/, /nin-ŋeh/, /niŋ-ŋeh/, and /niɲ-ŋeh/. The allomorph /ni-ŋeh/ is free, and the other four are phonologically-conditioned allomorphs.
Based on the process of affixation, it is found that the initial phoneme of root morpheme may firstly be overt and covert as well before accepting this circumfix. The analysis of examples shows that the allomorph is phonologically-conditioned. When it is overt, the initial phoneme of the root morpheme creates a double homorganic nasal-oral consonant, then the morpheme {niN-ŋeh} is affixed to the base morpheme. When it is covert, the initial phoneme of the root morpheme is replaced with nasal phonemes. Moreover, as the analysis shows, this circumfix morpheme indicates meaning 'to do something for someone', 'to have something done', and to signify participle adjective.
1. Allomorph /ni-ŋeh/ The formation of a complex word shows that allomorph /ni-ŋeh/ of the Bidayuh-Somu Language is affixed as a set before and after root morpheme that is verb and noun. An example of its circumfixation is in Table VI below: In relation to its function, this allomorph is class-maintaining (please see example number 1-3) and classchanging (see example number 4-5). It is found that this circumfix is attached to a root that is begun with consonant liquid and nasal /m/.

Allomorph /nim-ŋeh/
The process of affixation shows that this circumfix allomorph /nim-ŋeh/ is affixed to the root morpheme initiated with the plosive /p/ consonant. Table VII below is a very limited example identified for its circumfixation: The implemented study on the Bidayuh-Somu Language shows that this allomorph /nim-ŋeh/ is phonologically-conditioned and present with root morpheme that is verb and noun. Therefore, this allomorph functions as class-maintaining (please see example number 1) and class-changing (please see example number 2).
3. Allomorph /nin-ŋeh/ Based on the structure, the allomorph /nin-ŋeh/ of the language is circumfixed as a set before and after root morpheme that is verb and noun. The example of its circumfixation is listed in Table VIII.
The description of the example shows that this allomorph is class-maintaining (please look at example number 1-2) and class-changing (please look at example number 3). From the analysis, it is recognized that the allomorph /nin-ŋeh/ is phonologically-conditioned and present with root morpheme started from consonant of plosive /t/, and affricate /c/ and /j/.

Allomorph /niŋ-ŋeh/
This circumfix allomorph is affixed as a set before and after root morpheme that is begun with consonant plosive /k/ and vowel /i/. Please see the example in the following Table IX:  The study on the Bidayuh-Somu Language shows that the allomorph /niŋ-ŋeh/ is phonologicallyconditioned, and functions class-maintaining as seen in examples number 1-2, and class-changing as found in example number 3.
5. Allomorph /niɲ-ŋeh/ As a member of morpheme {niN-ŋeh}, this allomorph /niɲ-ŋeh/ is found to present as a set before and after root morpheme that is only a verb. Hence, this circumfix allomorph is only class-maintaining. Here in Table X is a very limited example available of its circumfixation process: Based on its structure, this allomorph is phonologically-conditioned and attached to root morpheme initiated with affricate /c/ and fricative /s/ consonants.

IV. CONCLUSION
Based on the description, it is explicitly concluded that the morphemes of the Bidayuh-Somu Language with nasals elements are free and phonologically conditioned. The morphemes and their allomorphs are nominal and verbal. The phonologically-conditioned allomorph of this language is additive and replacive to the first or initial phonemes of the root or base word. When it is additive, the first phoneme is overt and creates double homorganic nasal-oral consonants. When it is replacive, the first phoneme is covert and replaced with nasals consonants.
The prefix morpheme {puN-} and its allomorph are nominal and are class-changing and classmaintaining. They enter root or base morpheme started with all consonants and vowels phonemes. Moreover, the circumfix morpheme {niN-ŋeh} and its allomorph are verbal and are class-changing and class-maintaining. They are affixed before and after root or base words beginning with all consonants and vowels phonemes.